ErrLookup › sgl-project/sglang

sgl-project/sglang

SGLang is a high-performance serving framework for large language models and multimodal models. · Python · 3,558 source files

Analyzed at 0132848349 on 2026-08-28. 3398 documented errors.

Code / Message	Type	Severity	Tags
No frames were recorded	error_code	error	ngram, ffi, buffer-size, cuda-kernel, tensor-shape
This browser cannot encode H.264 MP4	error_code	error	ngram, config, parsing, speculative-decoding
H.264 encoder did not return MP4 decoder config	error_code	error	ngram, config, parsing, range-validation
This browser does not support gzip stream decoding	error_code	error	ngram, config, range, bounds-check
delta payload size mismatch: expected ${expectedSize}, got $	error_code	error	ngram, config, overlap, parsing
Missing previous frame for delta payload	error_code	error	suffix-automaton, ngram, state-machine, api-misuse
Unsupported content type ${header.content_type}	error_code	error	diffusion, mesh-inpainting, invalid-argument, cpp-extension
/v1/models ${response.status}	http	error	radix-tree, hicache, not-implemented, mem-cache
Unsupported msgpack byte ${b}	error_code	error	radix-tree, hicache, not-implemented, kv-cache-load
This browser does not support gzip stream decoding	error_code	error	radix-tree, hicache, not-implemented, io-commit
delta payload size mismatch: expected ${expectedSize}, got $	error_code	critical	rocm, allreduce, tensor-parallel
Missing previous frame for delta payload	error_code	error	rocm, allreduce, deterministic, buffer-registration
Previous frame size does not match current delta payload	error_code	error	rocm, allreduce, deterministic, alignment, float32
This browser does not support worker image decoding	error_code	critical	rocm, allreduce, deterministic, tensor-parallel, float32
Unsupported content type ${header.content_type}	error_code	error	rocm, allreduce, deterministic, alignment, float16
Generate subcommand is not yet supported for model: {model_p	exception	critical	rocm, allreduce, deterministic, tensor-parallel, float16
Error: --model-type requires a value.	validation	error	rocm, allreduce, deterministic, alignment, bfloat16
Error: --model-type requires a non-empty value.	validation	critical	rocm, allreduce, deterministic, tensor-parallel, bfloat16
Reserved serve backend names cannot be used: {names}	validation	error	rocm, allreduce, deterministic, dtype
Out-of-tree serve backends cannot replace reserved or built-	exception	critical	rocm, allreduce, quick-allreduce, tensor-parallel
Unknown serve backend {name!r}. Available values: {available	validation	error	sglang, cli, entry-points, plugin-registry, invalid-argument
Multiple distributions register serve backend {name!r}: {pro	exception	error	sglang, cli, entry-points, duplicate-registration, plugin-conflict
Failed to load serve backend {name!r} from {self._entry_poin	exception	critical	sglang, cli, entry-points, plugin-load-failure, import-error
Serve backend {name!r} factory returned {type(backend).__nam	exception	critical	sglang, cli, plugin-api-mismatch, type-check
Serve backend {name!r} uses API version {backend.api_version	exception	critical	sglang, cli, version-mismatch, plugin-abi
Multiple serve backends matched this request: {names}. Selec	exception	error	sglang, cli, auto-detection, ambiguity, model-type
Usage: sglang serve --model-path <model-name-or-path> [addit	exception	info	sglang, cli, usage, help, missing-argument
Error: --model-path is required. Please provide the path to	exception	error	sglang, cli, missing-argument, model-path
Sync request failed: {error}	exception	critical	sglang, deep-gemm, compilation, http-500, kernel-build
Waiting for main node timeout!	exception	error	sglang, deep-gemm, timeout, multi-node, compilation
DeepGEMM Kernels compilation timeout.\n\nFeel free and pleas	exception	critical	sglang, deep-gemm, timeout, server-startup, compilation
The pointers must be multiple of 16 bytes.	validation	error	sglang, cuda-kernel, alignment, silu, shape-validation
The last dimension ({input.shape[-1]}) x itemsize ({input.dt	validation	error	sglang, cuda-kernel, alignment, quick-gelu, shape-validation
Can not import FA3 in sgl_kernel. Please check your installa	exception	critical	sglang, flash-attention, import-error, native-extension, environment
v_cache must be provided	validation	error	sglang, flash-attention, kv-cache, missing-argument, validation
k_cache can only be None when only_qv=True	validation	error	sglang, flash-attention, kv-cache, invalid-argument
q can only be None when only_qv=True	validation	error	sglang, flash-attention, invalid-argument, validation
q must be provided unless qv is provided with only_qv=True	validation	error	sglang, flash-attention, missing-argument, validation
{_METALLIB_NAME} not found next to the native Metal extensio	exception	error	metal, macos, installation, import-error, sgl-kernel
rope_pool_fused expects q/k/v to be 3-D	validation	error	shape-validation, rope, metal, sgl-kernel, tensor-dims
rope_pool_fused expects positions/slots to be 1-D	validation	error	shape-validation, rope, metal, positions, sgl-kernel
rope_pool_fused expects pool tensors to be 3-D	validation	error	shape-validation, kv-cache, metal, rope, sgl-kernel
q shape must be [num_tokens, num_qo_heads, head_dim], got {q	validation	error	shape-validation, rope, gqa, metal, sgl-kernel
k shape must be [num_tokens, num_kv_heads, head_dim], got {k	validation	error	shape-validation, gqa, rope, metal, sgl-kernel
v shape must match k shape, got {v.shape} vs {k.shape}	validation	error	shape-validation, rope, kv-projection, metal, sgl-kernel
positions/slots must have one entry per token	validation	error	shape-validation, positions, kv-cache-slots, rope, sgl-kernel
k_pool has incompatible shape {k_pool.shape}	validation	error	shape-validation, kv-cache, gqa, metal, sgl-kernel
v_pool shape must match k_pool shape, got {v_pool.shape} vs	validation	error	shape-validation, kv-cache, metal, rope, sgl-kernel
q/k/v dtypes must match	validation	error	metal, rope, dtype-mismatch, apple-silicon
pool dtypes must match q/k/v dtype	validation	error	metal, rope, pool, dtype-mismatch
Input probs contains NaN.	validation	error	musa, sampling, top-p, nan
Invalid filter_apply_order: {filter_apply_order}	validation	error	musa, sampling, invalid-argument, enum-value
scalar_type_id {scalar_type_id} doesn't exists.	validation	error	scalar-type, registry, version-mismatch, quantization
setup_metal.py only supports macOS (Apple Silicon).	console	error	metal, build, platform-check, apple-silicon
Apple toolchain not found. Install the Xcode Command Line To	console	error	metal, build, xcode, missing-toolchain, macos
Apple Metal shader compiler not found. Install a full Xcode	console	error	metal, build, xcode, shader-compiler, macos
metal_shader_sources is empty; nothing to compile	exception	error	metal, build, empty-input, shader
metal shader source not found: {metal_src}	exception	error	metal, build, file-not-found, shader
{self._op_label()}: no triton backend	exception	error	fused-op, backend-dispatch, not-implemented, triton
Cannot find NVIDIA Math-DX (cuBLASDx) headers. Install the `	exception	error	cuda, jit-kernel, missing-dependency, nvidia, build
Cannot find CUTLASS headers required for JIT compilation. Pl	exception	error	cutlass, jit-kernel, missing-dependency, flashinfer, deep-gemm, build
Unsupported type {type(data)}	validation	error	cutedsl, dtype, type-mismatch, kernel
num_heads must be divisible by num_epi_subtiles	validation	error	cutedsl, kernel-config, shape-validation, attention
num_heads // num_epi_subtiles must be divisible by 4 (FMA un	validation	error	cutedsl, kernel-config, shape-validation, attention
Unexpected initial_state_source shape: {initial_state_source	validation	error	gdn, linear-attention, shape-validation, cutedsl
Unexpected A_log shape: {A_log.shape}; expected numel={HV}	validation	error	kda, linear-attention, shape-validation, cutedsl
Unexpected dt_bias shape: {dt_bias.shape}; expected numel={H	validation	error	kda, linear-attention, shape-validation, cutedsl
Unexpected a shape for varlen: {a.shape}	validation	error	kda, varlen, shape-validation, cutedsl
Unexpected a shape for dense: {a.shape}	validation	error	kda, dense-decode, shape-validation, cutedsl
sparse_attn_v4_paged_decode expects fp16/bf16 q, got {q.dtyp	validation	error	attention, dtype, triton, deepseek, gpu
kv_scales supplied but unified_kv is {unified_kv.dtype}, exp	validation	error	attention, fp8, quantization, kv-cache, triton
kv_scales must be fp32, got {kv_scales.dtype}	validation	error	fp8, kv-cache, dtype, scales
D={D_check} must be divisible by GROUP_SIZE={_FP8_GROUP_SIZE	validation	error	fp8, head-dim, kv-cache, shape-validation
kv_scales shape {tuple(kv_scales.shape)} does not match expe	validation	error	fp8, kv-cache, shape-validation, scales
unified_kv dtype mismatch: kv={unified_kv.dtype}, q={q.dtype	validation	error	attention, dtype, kv-cache, triton
Triton sparse_attn_v4_paged_prefill requires CUDA/HIP tensor	error_code	error	attention, triton, device, cpu-vs-gpu
sparse_attn_v4_paged_prefill expects fp16/bf16 q, got {q.dty	validation	error	attention, dtype, triton, prefill
unified_kv dtype mismatch: kv={unified_kv.dtype}, q={q.dtype	validation	error	attention, dtype, kv-cache, prefill
kv dtype mismatch: kv={kv.dtype}, q={q.dtype}	validation	error	attention, dtype, extend, kv
head_dim mismatch: unified_kv={unified_kv.size(-1)}, kv={kv.	validation	error	attention, head-dim, shape-validation, kv-cache
bad compress_ratio {compress_ratio}	validation	error	deepseek, sparse-attention, indexing, config-validation
kernel dispatch requires at least one tensor argument	validation	error	kernel-dispatch, api-misuse, assertion, triton
(head_dim, head_dim_v)=({head_dim}, {head_dim_v}) exceeds SM	validation	error	flash-attention, sm120, shared-memory, head-dim, blackwell
SM120 relative bias currently supports head_dim and head_dim	validation	error	flash-attention, sm120, relative-bias, head-dim
SM120 relative bias requires tile_mn=(64, 128)	validation	error	flash-attention, sm120, tile-config, relative-bias
The requested FlashAttention forward configuration exceeds S	validation	error	flash-attention, sm120, shared-memory, config-validation
The requested SM120 sheared-bias specialization exceeds shar	validation	error	flash-attention, sm120, shared-memory, relative-bias
head_first is deprecated and will be removed in a future ver	exception	warning	fla, gated-delta-rule, deprecation, tensor-layout
The batch size is expected to be 1 rather than {q.shape[0]}	validation	error	fla, gated-delta-rule, varlen, tensor-shape
The number of initial states is expected to be equal to the	validation	error	fla, gated-delta-rule, varlen, batch-mismatch, shape-validation
Unsupported input shape {g.shape}, which should be (B, T, H,	validation	error	fla, cumsum, shape-validation, head-first
This layer norm doesn't support feature dim >= 64KB.	validation	error	fla, rms-norm, triton, feature-dim-limit
Unsupported activation: {self.activation}	validation	error	fla, rms-norm, activation-validation, config
`mixed_qkv` must be a 2D tensor (got ndim={mixed_qkv.ndim}).	validation	error	fla, fused-recurrent, decode, shape-validation
`mixed_qkv` must be contiguous in the last dim.	validation	error	fla, fused-recurrent, contiguity, stride-check
`a` and `b` must be 2D tensors (got a.ndim={a.ndim}, b.ndim=	validation	error	fla, fused-recurrent, shape-validation
`a`/`b` must be contiguous in the last dim.	validation	error	fla, fused-recurrent, contiguity, stride-check
`A_log`/`dt_bias` must be 1D tensors.	exception	error	fla, fused-recurrent, shape-validation, mamba-params
`A_log`/`dt_bias` must be contiguous.	exception	error	fla, fused-recurrent, contiguity, mamba-params
`ssm_state_indices` must be 1D for packed decode (got ndim={	exception	error	fla, fused-recurrent, state-cache, shape-validation
`out` must be contiguous.	exception	error	fla, fused-recurrent, contiguity, output-buffer
All inputs must be on the same device.	exception	error	fla, fused-recurrent, device-mismatch, multi-gpu
Mismatched batch sizes: mixed_qkv.shape[0]={B}, a.shape[0]={	exception	error	fla, fused-recurrent, batch-mismatch
`ssm_state_indices` must have shape [B] (got {tuple(ssm_stat	exception	error	fla, fused-recurrent, state-cache, shape-validation
`initial_state` must be a 4D tensor (got ndim={initial_state	exception	error	fla, fused-recurrent, state-cache, shape-validation
`initial_state` must be contiguous in the last dim.	exception	error	fla, fused-recurrent, state-cache, contiguity
`a`/`b` must have shape [B, HV] with HV={HV} (got a.shape={t	exception	error	fla, fused-recurrent, head-mismatch, tensor-parallel
`A_log` and `dt_bias` must have {HV} elements (got A_log.num	exception	error	fla, fused-recurrent, mamba-params, tensor-parallel
`out` must have shape {(B, 1, HV, V)} (got out.shape={tuple(	exception	error	fla, fused-recurrent, output-buffer, shape-validation
Invalid packed `mixed_qkv` last dim={qkv_dim} for HV={HV}, V	exception	error	shape-validation, triton-kernel, gated-delta-rule, packed-decode
Invalid packed Q size {q_dim}: must be divisible by K={K}.	exception	error	shape-validation, triton-kernel, gated-delta-rule, packed-decode
Invalid head config inferred from mixed_qkv: H={H}, HV={HV}.	exception	error	shape-validation, gqa, triton-kernel, packed-decode
Packed decode kernel only supports NK=1 (got K={K}, BK={BK})	exception	error	kernel-limit, triton-kernel, packed-decode
`a` must have shape [B, HV*K] with HV={HV}, K={K} (got a.sha	exception	error	shape-validation, kda, gqa
`b` must have shape [B, HV] with HV={HV} (got b.shape={tuple	exception	error	shape-validation, kda, gqa
`A_log` must have {HV} elements (got {A_log.numel()}).	exception	error	shape-validation, model-weights, kda
`dt_bias` must have {HV * K} elements (got {dt_bias.numel()}	exception	error	pytorch, tensor-shape, kda, linear-attention, validation
Invalid packed Q size {q_dim}: must be divisible by K={K}. K	exception	error	pytorch, tensor-shape, gqa, kda, packed-qkv
Backward pass is not implemented yet and we do not have plan	exception	error	pytorch, autograd, backward, linear-attention, not-implemented
The batch size is expected to be 1 rather than {q.shape[0]}	exception	error	pytorch, variable-length, fla, linear-attention, validation
The number of initial states is expected to be equal to the	validation	error	pytorch, variable-length, state-management, gdn, validation
The number of intermediate state indices is expected to be e	validation	error	pytorch, state-management, gdn, validation
`mixed_qkv` must be 2D (got ndim={mixed_qkv.ndim}).	validation	error	pytorch, tensor-shape, replayssm, decode, validation
`mixed_qkv` must be contiguous in the last dim.	validation	error	pytorch, contiguity, triton, replayssm
`b` must be 2D (got b.ndim={b.ndim}).	validation	error	pytorch, tensor-shape, replayssm, gate
`A_log` must be a 1D tensor.	validation	error	pytorch, tensor-shape, replayssm, decay
`initial_state` must be 4D (got ndim={initial_state.ndim}).	validation	error	pytorch, tensor-shape, replayssm, state-management
`out` must be contiguous.	validation	error	pytorch, contiguity, replayssm, output-buffer
`write_pos` must be a 1D int32 tensor.	validation	error	pytorch, dtype, replayssm, ring-buffer
The batch size is expected to be 1 rather than {q.shape[0]}	validation	error	kda, fla, varlen, batch-shape, attention
This layer doesn't support feature dim >= 64KB.	validation	error	triton, l2norm, feature-dim, kda, attention
This layer norm doesn't support feature dim >= 64KB.	validation	error	rms-norm, triton, feature-dim, gated, kda
Unknown flash attention version {ver}	validation	error	flash-attention, version-dispatch, attention, invalid-argument
flash_attn at sgl-kernel is only supported on sm90 and above	validation	error	flash-attention, fa3, gpu-compatibility, sm90, hardware-unsupported
FlashAttention-4 CUTE is not available. Install flash-attn-4	validation	error	flash-attention, fa4, import-error, missing-dependency
FA4 does not support updating KV cache in-place.	validation	error	flash-attention, fa4, kv-cache, unsupported-operation
FA4 path does not support rotary embedding.	exception	error	flash-attention, fa4, rotary-embedding, unsupported-operation
FA4 path does not support non-consecutive batch indices or l	exception	error	flash-attention, fa4, batch-indices, left-padding, unsupported-operation
out must not require gradients	exception	error	flash-attention, fa4, autograd, out-tensor, validation
out must have stride 1 in the last dimension	exception	error	flash-attention, fa4, stride, memory-layout, validation
FlashAttention-4 CUTE is not available. Install flash-attn-4	exception	error	flash-attention, fa4, sm120, import-error, missing-dependency
FA4 does not support updating KV cache in-place.	exception	error	flash-attention, fa4, sm120, kv-cache, unsupported-operation
FA4 path does not support rotary embedding.	exception	error	flash-attention, fa4, sm120, rotary-embedding, unsupported-operation
FA4 path does not support non-consecutive batch indices or l	exception	error	flash-attention, fa4, sm120, batch-indices, unsupported-operation
Unsupported tcgen05 MMA op kind: {type(op).__name__}	exception	error	cutlass, cute, tcgen05, blackwell, mma, unsupported-dtype
{tensor_name}{context_clause} with shape {tensor.shape} cann	validation	error	block-sparse, attention, shape-mismatch, broadcast
{name}_block_cnt and {name}_block_idx must both be provided	validation	error	block-sparse, attention, paired-arguments, validation
{name}_block tensors must have dtype torch.int32	validation	error	block-sparse, attention, dtype, int32-required
{name}_block_cnt and {name}_block_idx must be on the same de	validation	error	block-sparse, device-mismatch, attention, cuda
{name}_block tensors must live on CUDA	validation	error	block-sparse, cuda, cpu-tensor, attention
{name} must have dtype torch.int32	validation	error	block-sparse, dtype, int32, metadata
{name} must be on the same device as block sparse tensors	validation	error	block-sparse, device-mismatch, metadata
{name} must live on CUDA	validation	error	block-sparse, cuda, metadata, cpu-tensor
Block sparse tensors{context} require BLOCK_SIZE_KV={base_n_	validation	error	block-sparse, block-size, configuration, attention
mask_block_cnt and mask_block_idx must be provided for block	validation	error	block-sparse, missing-argument, validation
Block sparse tensors{context} require explicit sparse_block_	validation	error	block-sparse, ambiguous-config, shape-inference
Block sparse tensors{context} have block size {sparse_block_	validation	error	block-sparse, block-size, alignment
Block sparse tensors{context} must have shapes (B, H, M) and	validation	error	block-sparse, shape-mismatch, rank
Block sparse tensors{context} {dim_name} dim must be {tgt} o	validation	error	block-sparse, shape-mismatch, broadcasting
Block sparse tensors{context} must share the same m-block di	validation	error	block-sparse, shape-mismatch, consistency
Block sparse tensors{context} n-block dimension must be <= {	validation	error	block-sparse, shape-mismatch, kv-length
Block sparse tensors{context} m-block dimension {num_m_block	validation	error	block-sparse, block-size, shape-mismatch
All block sparse tensors must be on the same device	validation	error	block-sparse, device-mismatch
spt must be a bool when provided	validation	error	block-sparse, type-error, spt
spt requires dq_write_order to be provided	validation	error	block-sparse, spt, missing-argument, backward
SplitKV partial output (mO) must be Float32	validation	error	cuda, dtype, flash-attention, cutlass, split-kv
All tensors must have the same data type	validation	error	cuda, dtype, flash-attention, cutlass, type-mismatch
Only Float16 or BFloat16 is supported	validation	error	cuda, dtype, flash-attention, cutlass, unsupported-dtype
LSE tensor must be Float32	validation	error	cuda, dtype, flash-attention, cutlass, lse
cu_seqlens_q tensor must be Int32	validation	error	cuda, dtype, flash-attention, varlen, int32
cu_seqlens_k tensor must be Int32	validation	error	cuda, dtype, flash-attention, varlen, int32
seqused_q tensor must be Int32	validation	error	cuda, dtype, flash-attention, masking, int32
seqused_k tensor must be Int32	validation	error	cuda, dtype, flash-attention, masking, int32
module {__name__!r} has no attribute {name!r}	exception	error	python, import, lazy-import, attributeerror, refactoring
O partial tensor must match dtype_partial	validation	error	cuda, dtype, flash-attention, split-kv, combine
O tensor must match dtype	validation	error	flash-attention, dtype-mismatch, cuda-kernel, validation
LSE partial tensor must be Float32	validation	error	flash-attention, dtype-mismatch, numerical, validation
LSE tensor must be Float32	validation	error	flash-attention, dtype-mismatch, validation
O partial tensor must have 4 or 5 dimensions: (num_splits, b	validation	error	flash-attention, shape-mismatch, rank-error, validation
LSE partial tensor must have 3 or 4 dimensions: (num_splits,	validation	error	flash-attention, shape-mismatch, validation
O tensor must have 3 or 4 dimensions: (batch, seqlen, nheads	validation	error	flash-attention, shape-mismatch, validation
LSE tensor must have 2 or 3 dimensions: (batch, seqlen, nhea	validation	error	flash-attention, shape-mismatch, validation
Type mismatch: {self.q_dtype} != {self.k_dtype}	validation	error	flash-attention, dtype-mismatch, sm100, fp8
Type mismatch: {self.q_dtype} != {self.v_dtype}	validation	error	flash-attention, dtype-mismatch, fp8, sm100
Type mismatch: {self.sfq_dtype} != {self.sfk_dtype}	validation	error	flash-attention, fp8, mxfp8, scale-factor, dtype-mismatch
The layout of mBias is wrong	error_code	error	flash-attention, tensor-layout, bias, sm100
Block sparsity + paged KV not supported on SM100	exception	error	flash-attention, block-sparse, paged-kv, sm100, unsupported-feature
Block sparsity + sheared bias is not supported on SM90	exception	error	flash-attention, block-sparse, bias, sm90, unsupported-feature
Invalid arch format: {arch_str}	validation	error	gpu-arch, validation, configuration, cuda
FA4 CuTe FP8 backward is not supported yet (forward-only).	exception	error	flash-attention, fp8, autograd, not-implemented
out must not require gradients	validation	error	flash-attention, autograd, out-tensor, validation
out must have stride 1 in the last dimension	validation	error	flash-attention, stride, memory-layout, out-tensor
Custom user-provided score_mod is not supported on SM8x arch	exception	error	flash-attention, score-mod, flex-attention, sm80, ampere, unsupported-feature
Unsupported compute capability: {arch}. Supported: 8.x, 9.x,	validation	error	flash-attention, gpu-arch, unsupported-hardware, cuda
out is only supported for forward-only inference	validation	error	autograd, attention, inference, flash-attention
FlashAttention combine kernel cannot be implemented with giv	error_code	critical	cuda, attention, split-kv, kernel-config
Unsupported CUTLASS scalar type for A/B: {cutlass_type!r}	exception	error	cutlass, sm100, dtype, tensor-core
Unsupported CUTLASS scalar type for accumulator: {cutlass_ty	exception	error	cutlass, sm100, accumulator, dtype
M must be 64, 128 or 256	validation	error	cutlass, sm100, mma, tile-shape
N must be a multiple of 8 in the range 8…256	validation	error	cutlass, sm100, mma, alignment
Unexpected swizzle shift – want S==3 for M==4	validation	error	cutlass, sm100, swizzle, shared-memory
Only Swizzle<2,5,2> supported for 128B_BASE32B	validation	error	cutlass, sm100, swizzle, shared-memory
Unsupported swizzle triple for UMMA smem descriptor	validation	error	cutlass, sm100, swizzle, shared-memory
Not a canonical UMMA_MN Layout: Expected profile failure.	validation	error	cutlass, sm100, layout, shared-memory
Not a canonical UMMA_MN Layout: Expected stride failure.	validation	error	cutlass, sm100, layout, stride
SWIZZLE_128B_BASE32B is invalid for Major-K	validation	error	cutlass, sm100, swizzle, layout
Not a canonical UMMA_K Layout: Expected MN-size multiple of	validation	error	cutlass, sm100, layout, alignment
Not a canonical UMMA_K Layout: Expected profile failure.	validation	error	cutlass, sm100, layout, shared-memory
Not a canonical UMMA_K Layout: Expected stride failure.	validation	error	cutlass, sm100, layout, stride
hd256 forward varlen expects q rank 3 or 5, got rank {q_rank	exception	error	cuda, attention, tensor-rank, varlen, head-dim-256
hd256 forward non-varlen expects q rank 4 or 5, got rank {q_	exception	error	cuda, attention, tensor-rank, batched, head-dim-256
hd256 forward varlen expects k rank 3 or 5, got rank {k_rank	exception	error	cuda, flash-attention, tensor-shape, sm100, cutlass
hd256 forward non-varlen expects k rank 4 or 5, got rank {k_	exception	error	cuda, flash-attention, tensor-shape, sm100
The layout of q is not supported	exception	error	cuda, flash-attention, memory-layout, sm100
The layout of k is not supported	exception	error	cuda, flash-attention, memory-layout, sm100, kv-cache
The layout of v is not supported	exception	error	cuda, flash-attention, memory-layout, sm100, kv-cache
Type mismatch: {self.q_dtype} != {self.k_dtype}	exception	error	cuda, flash-attention, dtype-mismatch, sm100
Type mismatch: {self.q_dtype} != {self.v_dtype}	exception	error	cuda, flash-attention, dtype-mismatch, kv-cache
flashinfer_sparse_mla supports only GLM DSA with FP8 KV cach	validation	error	attention-backend, config-validation, sm120, fp8, glm
GLM DSA with FP8 KV cache on NVIDIA SM120/SM121 supports onl	validation	error	attention-backend, config-validation, sm120, fp8, glm
`mixed_qkv` must be a 2D tensor (got ndim={mixed_qkv.ndim}).	validation	error	kda, mamba, tensor-shape, helion, triton
`mixed_qkv` must be contiguous in the last dim.	validation	error	kda, mamba, contiguity, helion
`a` and `b` must be 2D tensors (got a.ndim={a.ndim}, b.ndim=	validation	error	kda, mamba, tensor-shape, helion
`a`/`b` must be contiguous in the last dim.	validation	error	kda, mamba, contiguity, helion
`A_log`/`dt_bias` must be 1D tensors.	validation	error	kda, mamba, tensor-shape, parameters, helion
`A_log`/`dt_bias` must be contiguous.	validation	error	kda, mamba, contiguity, parameters, helion
`ssm_state_indices` must be 1D for packed decode (got ndim={	validation	error	kda, mamba, tensor-shape, state-indices, helion
`out` must be contiguous.	validation	error	kda, mamba, contiguity, output-buffer, helion
All inputs must be on the same device.	validation	error	kda, mamba, device-mismatch, helion, multi-gpu
Mismatched batch sizes: mixed_qkv.shape[0]={B}, a.shape[0]={	validation	error	kda, mamba, batch-mismatch, helion
`ssm_state_indices` must have shape [B] (got {tuple(ssm_stat	validation	error	kda, mamba, shape-mismatch, state-indices, helion
`initial_state` must be a 4D tensor (got ndim={initial_state	validation	error	kda, helion, shape-validation, tensor-ndim
`initial_state` must be contiguous in the last dim.	validation	error	kda, helion, contiguity, tensor-stride
Helion KDA decode requires power-of-two key and value head d	validation	error	kda, helion, power-of-two, head-dim, model-config
`a` must have shape [B, HV*K] with HV={HV}, K={K} (got a.sha	exception	error	kda, helion, shape-validation, packed-layout
`b` must have shape [B, HV] with HV={HV} (got b.shape={tuple	exception	error	kda, helion, shape-validation, packed-layout
`A_log` must have {HV} elements (got {A_log.numel()}).	exception	error	kda, helion, parameter-shape, model-config
`dt_bias` must have {HV * K} elements (got {dt_bias.numel()}	exception	error	kda, helion, parameter-shape, model-config
`out` must have shape {(B, 1, HV, V)} (got out.shape={tuple(	exception	error	kda, helion, shape-validation, output-buffer
Invalid packed `mixed_qkv` last dim={qkv_dim} for HV={HV}, V	exception	error	kda, helion, packed-layout, qkv, shape-validation
Invalid packed Q size {q_dim}: must be divisible by K={K}. K	exception	error	kda, helion, gqa, head-config
Invalid head config inferred from mixed_qkv: H={H}, HV={HV}.	exception	error	kda, helion, head-config, shape-validation
varlen KDA requires batch size 1	exception	error	kda, helion, varlen, batch-shape
KDA prefill requires an indexed initial-state pool	exception	error	kda, helion, prefill, state-pool, required-argument
g and beta must cover every q token	exception	error	kda, helion, prefill, length-mismatch
KDA `a` must be a contiguous 2D or 3D tensor.	exception	error	kda, replayssm, contiguity, tensor-ndim
KDA `dt_bias` must be a contiguous 1D or 2D tensor.	exception	error	kda, replayssm, contiguity, parameter-shape
`write_pos` must be a 1D int32 tensor.	exception	error	kda, replayssm, dtype, int32, tensor-ndim
`write_pos` must have shape {(batch,)}.	exception	error	kda, replayssm, shape-validation, batch-mismatch
`force_flush` must be a length-B int32 tensor or None.	exception	error	kda, replayssm, dtype, int32, optional-argument
ReplaySSM cache length must be at least 1.	exception	error	kda, replayssm, cache-allocation, zero-size
`d_cache` must have shape [slots, HV, L, V].	exception	error	helion, kda, replayssm, tensor-shape, cache-layout
`k_cache` must have shape [slots, H, L, K].	exception	error	helion, kda, replayssm, tensor-shape, cache-layout
`g_cache` must have shape [slots, HV, L, K].	exception	error	helion, kda, replayssm, tensor-shape, gate-cache
`g_cache` must have dtype torch.float32.	exception	error	helion, kda, replayssm, dtype, gate-cache
ReplaySSM inputs must be on the same device.	exception	error	helion, kda, replayssm, multi-gpu, device-placement
`force_flush` must be on the same device as the inputs.	exception	error	helion, kda, replayssm, device-placement, force-flush
MXFP8 fused prologue requires head_dim-aligned Q/K/V.	exception	error	mxfp8, quantization, head-dim, attention-prologue, inkling
MXFP8 fused prologue requires K/V scale buffers.	exception	error	mxfp8, quantization, scale-buffers, attention-prologue, inkling
MXFP8 fused prologue requires interleaved K/V scale buffers	exception	error	mxfp8, scale-buffers, tensor-shape, kv-cache, inkling
MXFP8 fused prologue requires contiguous interleaved SFK/SFV	exception	error	mxfp8, scale-buffers, contiguity, kv-cache, inkling
MXFP8 fused decode prologue requires head_dim-aligned Q/K/V.	exception	error	mxfp8, quantization, head-dim, decode, inkling
MXFP8 fused decode prologue requires K/V scale buffers.	exception	error	mxfp8, scale-buffers, decode, kv-cache, inkling
MXFP8 fused decode prologue requires interleaved K/V scale b	exception	error	mxfp8, scale-buffers, tensor-shape, decode, inkling
MXFP8 fused decode prologue requires contiguous interleaved	exception	error	mxfp8, scale-buffers, contiguity, decode, inkling
KDA cutedsl: safe_gate (lower_bound) not yet supported	exception	error	kda, cutedsl, safe-gate, not-implemented, linear-attention
Unsupported integer dtype: {dtype}	exception	error	kda, cutlass, dtype, index-tensor, prefill
eqlen with B>1 and T % {BT} != 0 not supported (got B={B}, T	exception	error	kda, linear-attention, shape-mismatch, not-implemented
kda_prefill is the inference forward path: cp_context, and d	exception	error	kda, training-vs-inference, not-implemented, invalid-argument
allow_neg_eigval=True requires 2*sigmoid(beta), which is not	exception	error	kda, beta, flag-conflict, not-implemented
Flash attention currently only supported for compute capabil	exception	error	cuda, gpu-capability, lightning-attn, hardware-unsupported
Unsupported split_k: {split_k}	exception	error	nsa, triton, split-k, invalid-argument
Unsupported d_qk: {d_qk}. Expected {DSV4_D_QK} (DeepSeek V4)	exception	error	nsa, triton, deepseek, head-dim, model-mismatch
qprep_bf16_fp8_sm90 requires an SM90 (Hopper) GPU	exception	error	cuda, sm90, hopper, jit-kernel, hardware-unsupported
{name} must have shape {tuple(shape)}, got {tuple(t.shape)}	validation	error	shape-validation, output-buffer, sparse-mla, fp8
{name} must have dtype {dtype}, got {t.dtype}	validation	error	dtype-validation, output-buffer, sparse-mla, fp8
{name} must be on device {device}, got {t.device}	validation	error	device-validation, multi-gpu, output-buffer, sparse-mla
{name} must be contiguous	validation	error	contiguity, output-buffer, sparse-mla, strides
q must have shape (s_q, h_q, d_qk), got {tuple(q.shape)}	validation	error	rank-validation, shape-validation, sparse-mla, q8kv8
kv must have shape (s_kv, h_kv, d_qk), got {tuple(kv.shape)}	validation	error	rank-validation, shape-validation, sparse-mla, kv-cache
indices must have shape (s_q, h_kv, topk), got {tuple(indice	validation	error	rank-validation, indices, sparse-mla, topk
q must be a CUDA tensor	validation	error	device-validation, cuda, cpu-tensor, sparse-mla
kv must be a CUDA tensor	validation	error	device-validation, cuda, mixed-device, sparse-mla
indices must be a CUDA tensor	validation	error	device-validation, cuda, indices, sparse-mla
kv must be on q's device {device}, got {kv.device}	validation	error	device-validation, multi-gpu, tensor-parallel, sparse-mla
indices must be on q's device {device}, got {indices.device}	validation	error	cuda, device-mismatch, sparse-attention, sglang
q must be torch.float8_e4m3fn, got {q.dtype}	validation	error	dtype, fp8, sparse-attention, sglang
kv must be torch.float8_e4m3fn, got {kv.dtype}	validation	error	dtype, fp8, kv-cache, sparse-attention
q must be contiguous	validation	error	contiguity, cuda, sparse-attention
kv must be contiguous	validation	error	contiguity, kv-cache, cuda
indices must be contiguous	validation	error	contiguity, indices, sparse-attention
kv d_qk must match q d_qk={d_qk}, got {kv_d_qk}	validation	error	shape-mismatch, kv-cache, sparse-attention
sparse_mla_q8kv8_prefill_fwd requires h_q padded to a positi	validation	error	shape-validation, tensor-parallel, sparse-attention
sparse_mla_q8kv8_prefill_fwd requires h_kv=1, got {h_kv}	validation	error	shape-validation, mla, sparse-attention
sparse_mla_q8kv8_prefill_fwd supports d_qk=512/576, got {d_q	validation	error	shape-validation, mla, unsupported-dim
indices must have shape ({s_q}, {h_kv}, topk), got {tuple(in	validation	error	shape-mismatch, indices, sparse-attention
indices must be int32, got {indices.dtype}	validation	error	dtype, indices, sparse-attention
Q8KV8 sparse-prefill topk width must be a positive multiple	validation	error	shape-validation, topk, sparse-attention
topk_length must be int32 with shape ({s_q},), got {tuple(to	validation	error	dtype, shape-mismatch, topk
topk_length must be a CUDA tensor	validation	error	cuda, device-mismatch, topk
topk_length must be on q's device {device}, got {topk_length	validation	error	cuda, device-mismatch, topk
topk_length must be contiguous	validation	error	contiguity, topk
topk_length values must satisfy 0 <= topk_length <= topk ({t	validation	error	value-validation, topk, sparse-attention
sparse_mla_q8kv8_prefill_fwd only supports d_v=512, got {d_v	validation	error	shape-validation, mla, unsupported-dim
attn_sink requires topk_length to be provided as well	validation	error	argument-validation, attention-sink, sparse-attention
attn_sink must be float32 with shape ({h_q},), got {tuple(at	validation	error	mla, sparse-attention, tensor-validation, dtype
attn_sink must be a CUDA tensor	validation	error	mla, sparse-attention, tensor-validation, cuda-device
attn_sink must be on q's device {device}, got {attn_sink.dev	validation	error	mla, sparse-attention, multi-gpu, device-mismatch
attn_sink must be contiguous	validation	error	mla, sparse-attention, tensor-validation, contiguity
{name} must be a torch.Tensor	validation	error	mla, quantization, scale-validation, tensor-validation
{name} must be a CUDA tensor	validation	error	mla, quantization, cuda-device, scale-validation
{name} must be on q's device {device}, got {scale.device}	validation	error	mla, quantization, multi-gpu, device-mismatch
{name} must be float32, got {scale.dtype}	validation	error	mla, quantization, dtype, scale-validation
{name} must be a scalar tensor, got shape {tuple(scale.shape	validation	error	mla, quantization, scale-validation, shape
out, max_logits and lse must not alias each other	validation	error	mla, sparse-attention, buffer-aliasing, output-buffers
Unsupported fused vision RoPE inputs: q={q.shape}/{q.dtype}/	validation	error	vision-rope, triton, gpu-capability, tensor-validation
In-place vision RoPE requires complex64 frequencies, got {fr	validation	error	vision-rope, dtype, complex-tensor
missing value for {a} (expected e.g. `{a} 2,4`)	validation	error	cli-parsing, multi-gpu, argument-validation
Invalid number of GPUs requested: {N} (available: {num_devic	validation	error	multi-gpu, cli-validation, cuda-devices
module {__name__!r} has no attribute {name!r}	exception	error	import, lazy-loading, attribute-error, diffusion
unsupported input for Sana fused bias-SiLU	exception	error	sana, diffusion, triton, memory-format, channels-last
unsupported input for Sana fused bias-GLU	exception	error	sana, diffusion, glu, channels-last, triton
unsupported input for packed fused SiLU-mul	exception	error	swiglu, silu, triton, strides, bit-exact
combined_history=True requires direction=0 (bidi)	validation	error	gdn, linear-attention, triton, argument-validation
unsupported dtype for causal Conv3D cat/pad: {x.dtype}	exception	error	cuda, dtype, conv3d, diffusion
unsupported input for causal Conv3D cat/pad CUDA	exception	error	cuda, contiguity, conv3d, input-validation
q, k, and v must have the same 3D shape	validation	error	attention, ulysses, shape-mismatch, sequence-parallelism
q, k, and v must be CUDA tensors	validation	error	cuda, device, ulysses, attention
q, k, and v must have the same device and dtype	validation	error	dtype, device, mismatch, ulysses
q, k, and v must be contiguous in head_size	validation	error	stride, contiguity, ulysses, triton
world_size must be positive and divide global_heads	validation	error	distributed, world-size, divisibility, ulysses
out must be a contiguous tensor with the expected shape, dev	validation	error	out-buffer, shape-mismatch, allocation, ulysses
Unsupported usp_merge_heads dtype: {dtype}	exception	error	dtype, jit, ulysses, cuda
unsupported input for usp_merge_heads CUDA	exception	error	cuda, fallback, ulysses, input-validation
timestep must have shape [B, S, 9 * D]	validation	error	shape, ltx2, adaln, diffusion
timestep must be a CUDA bfloat16 tensor	validation	error	dtype, cuda, bfloat16, ltx2
timestep must be contiguous	validation	error	contiguity, triton, ltx2
scale_shift_table must have shape [9, D]	validation	error	shape, ltx2, checkpoint, adaln
scale_shift_table must be CUDA, bf16/fp32, last-dim contiguo	validation	error	device, dtype, contiguity, ltx2
timestep last dim must equal 9 * hidden	validation	error	shape, ltx2, adaln, dimension-mismatch
hidden size is outside the supported LTX2 fast-path range	validation	error	hidden-size, kernel-limits, ltx2, triton
Unsupported modulate_scale_shift dtype: {dtype}	exception	error	dtype, jit, modulate, cuda
unsupported input for modulate_scale_shift CUDA	exception	error	cuda, modulate, fallback, input-validation
Unsupported residual_gate_add dtype: {dtype}	exception	error	dtype, jit, residual, cuda
Validate failed: unsupported dtype: {t.dtype}	validation	error	dtype, validation, cuda-kernel, diffusion
Validate failed: unsupported tensor shape: {t.shape}.	validation	error	shape, validation, cuda-kernel, diffusion
Validate failed: not contiguous on dim D.	validation	error	contiguity, stride, validation, cuda-kernel
Validate failed: S({S}) must be divisible by F({F}).	validation	error	shape, video-diffusion, modulation, validation
D={D} not supported, must be multiple of 256 and <= 8192	validation	error	shape-constraint, cuda-kernel, diffusion, norm
norm_type must be one of "layer" and "rms"	validation	error	enum, norm, validation
unsupported input for wan_rmsnorm_silu	validation	error	memory-format, triton, vae, wan, validation
QKV tensors must have shape [B, S, H, D]	validation	error	shape, rope, attention, hunyuan, diffusion
QKV tensors must be CUDA bfloat16 tensors	validation	error	dtype, device, rope, bfloat16, hunyuan
QKV tensors must be on the same CUDA device	validation	error	device, multi-gpu, rope, hunyuan
image QKV shapes must match	validation	error	shape, attention, gqa, hunyuan, rope
text QKV shapes must match	validation	error	shape, attention, hunyuan, rope
QKV last dimensions must be contiguous	validation	error	triton, rope, tensor-contiguity, hunyuan
head_dim must be positive, even, and <= 128	validation	error	rope, head-dim, shape-validation, hunyuan
cos and sin must have matching [S, D/2] shapes	validation	error	rope, cos-sin, shape-validation
cos/sin shape does not cover image tokens and head_dim	validation	error	rope, cos-sin, bounds-check, multimodal
cos and sin must be CUDA and last-dim contiguous	validation	error	rope, cuda, device-placement, contiguity
QKV and cos/sin tensors must be on the same CUDA device	validation	error	cuda, multi-gpu, device-mismatch, rope
Unsupported interleaved_rope_fp64 dtype: {dtype}	validation	error	jit, dtype, bfloat16, rope
Unsupported ltx25_decoder_rope dtype: {dtype}	validation	error	jit, dtype, bfloat16, ltx, rope
unsupported input for LTX2 QKNorm split-RoPE CUDA	validation	error	ltx, qknorm, rope, cuda, input-validation
LTX2 split RoPE shape mismatch: x={tuple(x.shape)}, cos={tup	validation	error	rope, ltx, shape-validation
cta_n={cta_n} invalid for use_2cta={use_2cta}: bf16 K-major	validation	error	gemm, blackwell, cutedsl, tile-config, mma
TGV cute_ext backend supports {list(_TORCH_TO_CUTLASS_DTYPE)	validation	error	gemm, tgv, dtype, cutlass
TGV cute_ext output supports {list(_TORCH_TO_CUTLASS_OUT_DTY	validation	error	gemm, tgv, output-dtype, cutlass
TGV cute_ext tactic {tactic} out of range [0, {len(_TGV_CUTE	validation	error	gemm, tgv, tactic, out-of-range
cutedsl_bf16_gemm requires an SM10x GPU	error_code	error	gemm, sm100, blackwell, gpu-architecture
dsv3_fused_a_gemm requires SM90 (Hopper) or later	error_code	error	gemm, dsv3, sm90, hopper, gpu-architecture
fp8_blockwise_scaled_mm JIT kernel requires SM120 (Blackwell	error_code	error	gemm, fp8, blockwise, sm120, blackwell, gpu-architecture
LoRA batch_info must provide max_len or seg_lens.	validation	error	lora, batch-info, missing-config
tiny_gemm: no valid split_n for N={n}, K={k}, max_m={max_m};	error_code	error	gemm, tiny-gemm, split-n, block-size
tiny_k_gemm: no valid split_n for N={n}, K={k}	error_code	error	cuda, gemm, shape-validation, tiny-kernel
LoRA batch_info must provide max_len or seg_lens.	validation	error	lora, trtllm, missing-argument, validation
attn_res_fused_tma requires SM100+ excluding SM12x; SM{major	error_code	error	cuda, tma, sm100, architecture-check, jit
attn_res: nvb must be in [1, {_MAX_BANK_ROWS}], got {nvb}	validation	error	kimi-k3, tuning, argument-validation, range-check
gemm_ar: M={m} outside [1, {MAX_TOKENS}]	validation	error	gemm, all-reduce, token-limit, shape-validation
expected q/k shape {expected_shape}	validation	error	kda, mtp, shape-validation, dspark
expected v/g shape {expected_shape}	validation	error	kda, mtp, shape-validation, dspark
expected beta shape {(1, T, H)}	validation	error	kda, mtp, shape-validation, beta-decay
DSpARK KDA MTP requires a fixed 1 + num_spec dense tokens pe	validation	error	kda, mtp, cu-seqlens, uniform-batch
expected recurrent state layout [pool, H, V=128, K=128]	validation	error	kda, mtp, recurrent-state, dtype-stride
kv-canary: launch_canary_plan_kernels requires full_to_swa_i	validation	error	kv-canary, sliding-window, argument-validation
kv-canary: launch_canary_plan_kernels verify_capacity={verif	validation	error	kv-canary, capacity-mismatch, argument-validation
kv-canary: launch_plan_entries_kernel requires req_to_verify	validation	error	kv-canary, plan-entries, argument-validation
kv-canary: offsets kernel bs must be in [0, {_PLAN_BS_BLOCK_	validation	error	kv-canary, batch-size, bounds-check
kv-canary: write_offsets_len must be positive, got {write_of	validation	error	kv-canary, offsets, argument-validation
kv-canary: write_req_capacity must be non-negative, got {wri	validation	error	kv-canary, capacity, argument-validation
kv-canary: verify_capacity must be non-negative, got {verify	validation	error	kv-canary, capacity, argument-validation
kv-canary: req_to_token_stride0 must be positive, got {req_t	validation	error	kv-canary, stride, argument-validation
kv-canary: lut_len must be non-negative, got {lut_len}	validation	error	kv-canary, lut, argument-validation
kv-canary: has_swa_lut must be bool, got {type(has_swa_lut).	validation	error	kv-canary, type-check, argument-validation
kv-canary: lut_len must be positive when has_swa_lut is True	validation	error	kv-canary, lut, swa
kv-canary: lut_len must be 0 when has_swa_lut is False	validation	error	kv-canary, lut, swa
kv-canary: write_offsets_len must equal write_req_capacity +	validation	error	kv-canary, offsets, off-by-one
kv-canary: bs={bs} exceeds write_req_capacity={write_req_cap	validation	error	kv-canary, capacity, bounds-check
kv-canary: req_to_token_stride0={req_to_token_stride0} does	validation	error	kv-canary, stride, layout
kv-canary: {name} must have dtype {dtype}, got {tensor.dtype	validation	error	kv-canary, dtype, validation
kv-canary: {name} must be 1-D, got shape {tuple(tensor.shape	validation	error	kv-canary, shape, validation
kv-canary: {name} must be 2-D, got shape {tuple(tensor.shape	validation	error	kv-canary, shape, validation
kv-canary: {name} length must be {expected}, got {actual}	validation	error	kv-canary, length-mismatch, validation
kv-canary: {name} length must be >= {minimum}, got {actual}	validation	error	kv-canary, length-mismatch, lut
kv-canary: {name} must be on {reference_name}'s device {refe	validation	error	kv-cache, device-mismatch, torch, validation
kv-canary: launch_canary_plan_kernels_torch_reference verify	validation	error	kv-cache, shape-mismatch, validation, speculative-decoding
kv-canary: launch_canary_plan_kernels_torch_reference requir	validation	error	kv-cache, missing-argument, ragged-tensor, validation
kv-canary: SWA slot {slot} is outside full_to_swa_index_mapp	validation	error	kv-cache, index-out-of-range, sliding-window, lut
kv-canary: scatter_req_token_ids flat_in must be 1-D, got sh	validation	error	kv-cache, shape-validation, tensor-rank
kv-canary: scatter_req_token_ids offsets must be 1-D, got sh	validation	error	kv-cache, shape-validation, tensor-rank
kv-canary: scatter_req_token_ids req_pool_indices must be 1-	validation	error	kv-cache, shape-validation, tensor-rank
kv-canary: scatter_req_token_ids pool_out must be 2-D, got s	validation	error	kv-cache, shape-validation, tensor-rank
kv-canary: scatter_req_token_ids flat_in must be int64, got	validation	error	kv-cache, dtype-validation, torch
kv-canary: scatter_req_token_ids offsets must be int64, got	validation	error	kv-cache, dtype-validation, torch
kv-canary: scatter_req_token_ids req_pool_indices must be in	validation	error	kv-cache, dtype-validation, torch
kv-canary: scatter_req_token_ids pool_out must be int32, got	validation	error	kv-cache, dtype-validation, torch
kv-canary: scatter_req_token_ids offsets length {offsets.sha	validation	error	kv-cache, shape-validation, csr-offsets, off-by-one
kv-canary: scatter_req_token_ids bs+1={bs + 1} exceeds BATCH	validation	error	kv-cache, batch-size-limit, triton, capacity
kv-canary: {name} must be contiguous	validation	error	kv-cache, contiguity, triton, strides
kv-canary: RealKvSource.page_size must be >= 1, got {self.pa	validation	error	kv-cache, config-validation, page-size
kv-canary: RealKvSource.num_bytes_per_token must be a positi	validation	error	kv-cache, alignment, byte-width, validation
kv-canary: RealKvSource.read_bytes must be a positive multip	validation	error	kv-cache, alignment, validation, sampling
kv-canary: RealKvSource.tensor must be at least 2-D, got sha	validation	error	kv-cache, shape-validation, tensor-rank
kv-canary: RealKvSource.tensor dim-1 byte width must be a mu	validation	error	kv-cache, alignment, strides, validation
kv-canary: VerifyPlan verify_capacity must be positive, got	validation	error	kv-canary, argument-validation, capacity
kv-canary: at most {consts.MAX_REAL_KV_SOURCES} RealKvSource	validation	error	kv-canary, cuda-abi, limit-exceeded
kv-canary: real_kv_sources[{i}].tensor (viewed as uint8) mus	validation	error	kv-canary, shape-validation, tensor-layout
kv-canary: canary_buf slot stride must hold at least 4 int64	validation	error	kv-canary, buffer-layout, reference-implementation
kv-canary: WritePlan write_req_capacity must be positive, go	validation	error	kv-canary, argument-validation, capacity
kv-canary: at most {consts.MAX_REAL_KV_SOURCES} RealKvSource	validation	error	kv-canary, cuda-abi, limit-exceeded
kv-canary: expected input tensors are required when enable_w	validation	error	kv-canary, argument-validation, assert-mode
kv-canary: expected input tensors must be None when enable_w	validation	error	kv-canary, argument-validation, mutually-exclusive
kv-canary: canary_buf slot stride must hold at least 4 int64	validation	error	kv-canary, buffer-layout, reference-implementation
kv-canary: expected input tensors are required when enable_w	validation	error	kv-canary, reference-implementation, assert-mode
kv-canary: expected input tensors must be None when enable_w	validation	error	kv-canary, reference-implementation, mutually-exclusive
unknown q-prep variant {variant!r} (SGLANG_OPT_Q8KV8_QPREP_V	validation	error	q8kv8, env-var, variant-selection, mla
unknown absorbed-bmm K variant: {variant!r}	validation	error	q8kv8, triton, variant-selection, internal-invariant
Unsupported dtype {k.dtype}. Supported: bfloat16, float16	validation	error	fp8, kv-cache, dtype-validation
HiSparse speculative swap requires 2-4 steps, got {num_steps	validation	error	hisparse, speculative-decoding, shape-validation
miss_src, miss_dst, and miss_count must be provided together	validation	error	hisparse, argument-validation, all-or-none
miss_src must be int64 and miss_dst must be int32.	validation	error	hisparse, dtype-validation, miss-plan
miss_count must be int32.	validation	error	hisparse, dtype-validation, miss-plan
speculative miss_src/miss_dst must have shape [batch, >= ste	validation	error	hisparse, shape-validation, miss-plan
speculative miss_count must have shape [batch].	validation	error	hisparse, shape-validation, miss-plan
rmsnorm_hf: unsupported hidden_size={hidden_size} (must be a	validation	error	rmsnorm, shape-validation, cuda-kernel, layernorm
unknown gpu '{gpu}', expected one of {sorted(GPU_BUDGETS_BYT	validation	error	gpu, lookup-table, shared-memory, lplb
fused IPM kernel needs {used/1024:.1f} KiB of shared memory	validation	error	shared-memory, capacity, lplb, cuda-kernel
LPLB fused solver unavailable: {_unavailable_reason()}	exception	critical	lplb, backend-unavailable, jit, cuda
LPLB fused solver requires CUDA tensors; got A on {A.device}	validation	error	device-mismatch, cuda, lplb
LPLB fused solver requires float32; got A.dtype={A.dtype}.	validation	error	dtype-validation, float32, lplb
Unsupported dtype {dtype}. Supported: float16, bfloat16, flo	validation	error	dtype-validation, mamba, causal-conv1d, jit-kernel
{fn_name}: dst entry dims (dims {entry_start_dim}..{dst.ndim	validation	error	mamba, contiguity, strides, triton
dst and src must be on the same device. {dst.device=} {src.d	validation	error	device-mismatch, mamba, scatter, cuda
num_token_non_padded must be a torch.Tensor	exception	error	moe, triton, type-validation, tensor-shape
num_token_non_padded must be a single-element tensor, got sh	exception	error	moe, triton, tensor-shape, validation
num_token_non_padded must be an integer tensor, got {num_tok	exception	error	moe, dtype, triton, validation
num_token_non_padded and x must be on the same device	exception	error	moe, device-mismatch, cuda, validation
Unsupported activation: {ACTIVATION_TYPE}	exception	error	moe, triton, activation, unsupported-operation
topk kernels only support k <= 32: {k=}	exception	error	moe, topk, triton, capacity-limit
topk kernels only support streaming implementation: {_impl=}	exception	error	moe, topk, not-implemented, feature-flag
topk_ids must be int32, got {topk_ids.dtype}	exception	error	moe, dtype, int32, triton
topk_ids must be a CUDA tensor	exception	error	moe, cuda, device-mismatch, validation
native MXFP8 MoE only supports gated swiglu-oai, got {activa	exception	error	moe, mxfp8, rocm, not-implemented, activation
Type must match: {self.a_dtype} != {self.b_dtype}	exception	error	nvfp4, quantization, dtype-mismatch, cutlass, gpu-kernel
expected a tensor with at least one dimension	validation	error	nvfp4, weight-loading, shape-validation
dimension {dim} size {dim_size} must be divisible by 2 * gro	validation	error	nvfp4, shape-alignment, weight-loading, swiglu
nvfp4_gemm_swiglu_nvfp4_quant currently supports NVFP4 input	validation	error	nvfp4, dtype-validation, quantization
nvfp4_gemm_swiglu_nvfp4_quant requires CUDA tensors	validation	error	nvfp4, cuda, device-placement, gpu-kernel
nvfp4_gemm_swiglu_nvfp4_quant requires SM100, got SM{major}{	validation	error	nvfp4, sm100, blackwell, gpu-architecture, cuda
Shape mismatch: A K={k}, B K={b.shape[1] * 2}	validation	error	nvfp4, shape-mismatch, gemm, quantization
Interleaved FC1 N must be even, got {n}	validation	error	nvfp4, shape-validation, swiglu, weight-layout
Output N={n_out} must be divisible by sf_vec_size={sf_vec_si	validation	error	nvfp4, shape-alignment, scale-factor, quantization
Unsupported nvfp4_gemm_swiglu_nvfp4_quant configuration: sha	validation	error	nvfp4, cutlass, tiling, shape-alignment, sm100
int32-packed scale buffers require scale_ue8m0=True	validation	error	quantization, fp8, ue8m0, scale-factor, dtype-mismatch
scale_ue8m0=True requires an int32-packed output_s	validation	error	quantization, fp8, ue8m0, scale-factor, dtype-mismatch
Unsupported output_s dtype {output_s.dtype}	validation	error	quantization, fp8, scale-factor, dtype-validation
Unsupported dtype {dtype}. Supported: float16, bfloat16, flo	validation	error	fp8, quantization, jit, unsupported-dtype
probs must be 2D, got shape={tuple(probs.shape)}	validation	error	sampling, top-p, top-k, renorm, shape-validation
renorm kernels require a CUDA/HIP tensor	validation	error	sampling, top-p, cuda, device-placement
top_p must be scalar or have one value per row, got {top_ps.	validation	error	sampling, top-p, batch-size-mismatch, validation
top_p values must be in (0, 1]	validation	error	sampling, top-p, value-range, validation
top_k must be scalar or have one value per row, got {top_ks.	validation	error	sampling, top-k, batch-size-mismatch, validation
probs must be 2D, got shape={tuple(probs.shape)}	validation	error	sampling, top-p, renorm, shape-validation
Invalid stacked fused KV projection shape: got {tuple(kv.sha	validation	error	shape-validation, speculative-decoding, fused-kernel
Invalid fused KV projection shape: got {tuple(kv.shape)}, ex	validation	error	shape-validation, speculative-decoding, fused-kernel
Invalid fused KV rotary/head dim pair: rotary_dim={rotary_di	validation	error	config-validation, rope, speculative-decoding
Invalid stacked k_norm_weight shape for fused KV materializa	validation	error	shape-validation, rmsnorm, speculative-decoding
Invalid stacked eps shape for fused KV materialization: got	validation	error	shape-validation, rmsnorm, speculative-decoding
Invalid k_out shape for fused KV materialization: got {tuple	validation	error	shape-validation, kv-cache, speculative-decoding
Invalid k_out device/dtype for fused KV materialization: got	validation	error	device-dtype-validation, kv-cache
Invalid v_out shape for fused KV materialization: got {tuple	validation	error	shape-validation, kv-cache
Invalid v_out device/dtype for fused KV materialization: got	validation	error	device-dtype-validation, kv-cache
Only neox-style RoPE is supported.	validation	error	unsupported-feature, rope, speculative-decoding
Invalid fused KV rotary/head dim pair: rotary_dim={self.rota	validation	error	config-validation, rope
num_kv_heads mismatch across layers for fused KV path: expec	validation	error	config-validation, gqa, speculative-decoding
head_dim mismatch across layers for fused KV path: expected	validation	error	config-validation, attention
RoPE config mismatch across layers for fused KV path: expect	validation	error	config-validation, rope
RoPE cos/sin cache is too short for fused KV materialization	validation	error	rope, context-length, kv-cache
positions must match ctx_hidden token count for fused KV mat	validation	error	shape-validation, speculative-decoding
Unknown match_type: '{match_type}'. Must be 'BFS' or 'PROB'.	validation	error	argument-validation, ngram, speculative-decoding
External ngram corpus exceeds the remaining token budget ({m	validation	error	resource-limit, ngram, corpus-loading
Conflicting kernel registration for op {spec.op!r}, backend	validation	error	kernel-registry, duplicate-registration
No '{backend.value}' backend registered for op {op!r}	exception	error	kernel-registry, missing-registration
No kernels registered for op {op!r}	exception	error	kernels, registry, invalid-argument, sglang
No '{backend.value}' backend registered for op {op!r}	exception	error	kernels, backend-selection, registry, sglang
op {op!r} has no backend usable on device {platform.device.v	validation	error	kernels, device-eligibility, environment, missing-dependency, sglang
op {op!r} has multiple backends usable on device {platform.d	validation	error	kernels, backend-selection, ambiguity, sglang
KernelSpec.target must be 'module:attr', got {self.target!r}	validation	error	kernels, spec-validation, configuration, sglang
Crusoe API key required. Pass api_key= or set CRUSOE_API_KEY	validation	error	frontend, api-key, missing-env-var, crusoe, sglang
This use case is not supported if api speculative execution	exception	error	frontend, openai, chat-model, program-structure, sglang
Unknown dtype: {sampling_params.dtype}	validation	error	frontend, openai, dtype, invalid-argument, sglang
This use case is not supported. For OpenAI chat models, sgl.	exception	error	frontend, openai, chat-model, streaming, program-structure, sglang
select/choices is not supported for chat models. Please try	exception	error	frontend, openai, choices, chat-model, not-supported, sglang
Invalid dtype: {sampling_params.dtype}	exception	error	frontend, dtype, regex, invalid-argument, sglang
Initialization failed. Please see the error messages above.	exception	critical	frontend, server-startup, spawn, oom, sglang
Server failed to start within the timeout period.	exception	error	frontend, timeout, server-startup, spawn, sglang
Failed to get server info. {error_data['error']['message']}	error_code	error	http, server-info, startup, network
GenerativeModel	error_code	error	import-error, optional-dependency, vertexai, environment
Unconditional token logprobs are required for this method.	validation	error	validation, logprobs, argument-validation, choices
Timeout while waiting for event '{name}'	error_code	error	timeout, concurrency, async, meta-info
Unknown type: {type(other)}	validation	error	type-error, interpreter, dsl, validation
Wrong type of stop in sampling parameters.	validation	error	sampling-params, validation, stop-sequences
Tried to append None to state.	validation	error	none-check, operator-overload, dsl, validation
Invalid join mode: {mode}	validation	error	validation, fork-join, dsl, enum-value
Invalid value: {other}	validation	error	operator-overload, fork-join, type-error, dsl
Given arguments mismatch the SGL function signature	validation	error	sglang, batch, arguments, signature-validation
Cannot put argument inside a f-string. This is not compatibl	validation	error	sglang, f-string, tracer, typeerror
name must be provided	validation	error	sglang, reasoning, naming, validation
Ray is required for --use-ray mode. Install it with: pip ins	error_code	critical	sglang, ray, importerror, dependency, server-launch
Unsupported model type: {model_type}	validation	error	comfyui, diffusion, model-loading, unsupported-architecture
Failed to get model info: {str(e)}	error_code	error	network, http-client, model-info, sglang-server
Prompt cannot be empty	validation	error	validation, prompt, image-generation
Image file not found: {image_path}	validation	error	file-io, image-edit, path-validation
Mask file not found: {mask_path}	validation	error	file-io, mask, inpainting, path-validation
Failed to edit image: {str(e)}	error_code	error	network, http, image-edit, timeout, sgldiffusion
Failed to generate image: {str(e)}	error_code	error	network, http, image-generation, timeout, retry, sgldiffusion
Video generation failed: {error_msg}	error_code	error	video-generation, server-side-failure, polling, sgldiffusion
Lost connection to server after {consecutive_errors} consecu	error_code	error	network, connection-lost, video-generation, polling, sgldiffusion
Network error after {consecutive_errors} consecutive failure	error_code	error	network, timeout, video-generation, polling, retry, sgldiffusion
Video generation timed out after {max_wait_time} seconds	error_code	error	timeout, video-generation, polling, sgldiffusion
Failed to generate video: {str(e)}	error_code	error	network, http, video-generation, submit, sgldiffusion
No image data in response	validation	error	validation, empty-response, image-decoding, sgldiffusion
Image index {index} out of range	validation	error	validation, index-out-of-range, image-decoding, sgldiffusion
No base64 image data found	validation	error	validation, base64, image-decoding, response-schema, sgldiffusion
lora_nickname cannot be empty	validation	error	validation, lora, empty-parameter, sgldiffusion
Failed to set LoRA adapter: {str(e)}	error_code	error	network, http, lora, sgldiffusion
Failed to unset LoRA adapter: {str(e)}	error_code	warning	network, http, lora, sgldiffusion
Prompt cannot be empty	validation	error	validation, comfyui, empty-prompt, image-generation
Failed to generate image: {str(e)}	error_code	error	comfyui, image-generation, error-wrapping, network
No image data in response	validation	error	comfyui, empty-response, base64, image-generation
Failed to generate video: {str(e)}	exception	error	comfyui, video-generation, error-wrapping, network
fl2va requires first_frame, last_frame, or both	validation	error	comfyui, sgldiffusion, minimax-h3, input-validation
ref2va requires at least one of reference_image, reference_v	validation	error	comfyui, sgldiffusion, minimax-h3, input-validation
t2va takes no conditioning inputs; pick another task	validation	error	comfyui, sgldiffusion, minimax-h3, input-validation
Failed to generate MiniMax-H3 video: {str(e)}	exception	error	comfyui, sgldiffusion, rpc, video-generation
Krea-2 sequence parallelism does not support ragged/padded m	validation	error	sglang, krea-2, sequence-parallelism, multi-prompt, batching, diffusion
actions must be a list[list[str]]	validation	error	sglang, lingbot-world, type-validation, actions, embodied-ai
Unknown token_type {token_type}, only support "text" or "ima	validation	error	sglang, longcat-image, token-type, position-ids, invalid-argument
Cannot duplicate reference image of batch size {latent_condi	validation	error	sglang, longcat-image, batch-size, reference-image, image-to-image
padding_side must be 'left' or 'right', got {padding_side}	validation	error	sglang, ltx-2, padding-side, text-embedding, invalid-argument
Unsupported text encoder output: expected `hidden_states`.	validation	error	sglang, ltx-2, text-encoder, hidden-states, attribute-error, mocking
num_inference_steps must be positive, got {steps}	validation	error	sglang, ltx-2, num-inference-steps, sigma-schedule, diffusion, invalid-argument
LTX-2 SP time-sharding for packed token latents currently re	validation	error	sglang, ltx-2, sequence-parallelism, patch-size, video, latent-sharding
Expected {seq_len=} > 0 for packed token latents.	validation	error	sglang, ltx-2, sequence-length, latent-packing, sequence-parallelism, video
Invalid {self.vae_scale_factor=}. Must be > 0.	validation	error	ltx-2, video-generation, config-validation, sequence-parallelism
Invalid {self.patch_size=}. Must be > 0.	validation	error	ltx-2, video-generation, config-validation, sequence-parallelism
Invalid latent H/W computed from batch.height/width: {batch.	validation	error	ltx-2, video-generation, resolution-validation, sequence-parallelism
Invalid spatial patching for packed token latents. Expected	validation	error	ltx-2, video-generation, resolution-validation, divisibility
Invalid tokens_per_frame={tokens_per_frame} from {latent_hei	validation	error	ltx-2, video-generation, defensive-check, sequence-parallelism
LTX-2 token latents seq_len={seq_len} is not divisible by to	validation	error	ltx-2, video-generation, sequence-parallelism, tensor-shape-mismatch
MiniMax-H3 quality="high" is validated only for the strict 4	validation	error	minimax-h3, hardware-validation, quality-mode, video-generation
MiniMax-H3 on MPS requires synchronous layerwise offload for	validation	error	minimax-h3, mps, apple-silicon, memory-offload, server-args
MiniMax-H3 MPS execution does not support torch.compile; pas	validation	error	minimax-h3, mps, torch-compile, server-args
MiniMax-H3 ring parallelism requires the FlashAttention back	validation	error	minimax-h3, ring-parallelism, attention-backend, server-args
Unsupported image type: {type(image)}	validation	error	multimodal, input-validation, type-error, image-processing
QwenImageEditPlus expects either one shared condition image	validation	error	qwen-image, image-editing, batch-validation, multimodal
QwenImage RoPE text cache overflow before denoising: require	validation	error	qwen-image, rope, sequence-length, cache-overflow, multimodal
QwenImage text conditioning mask has shape {tuple(mask.shape	validation	error	qwen-image, shape-mismatch, mask-validation, text-embedding
Cannot duplicate `image` of batch size {latent_condition.sha	validation	error	qwen-image, batch-mismatch, image-latents, img2img
Qwen-Image-Layered requires generated latent shapes.	validation	error	qwen-image, layered-generation, latent-shapes, missing-metadata
Qwen-Image-Layered generated latent shapes must match, got {	validation	error	qwen-image, layered-generation, shape-mismatch, latent-shapes
num_frames must be positive	validation	error	sana-video, video-generation, argument-validation, num-frames
SANA-WM height/width must be divisible by the LTX-2 spatial	validation	error	sana-wm, vae-stride, resolution-validation, video-generation
SD3 CLIP postprocessing requires hidden_states from encoder	validation	error	stable-diffusion-3, clip, text-encoding, hidden-states
Z-Image text embeddings must have shape [seq, dim] or [batch	validation	error	z-image, multimodal, tensor-shape, text-embeddings, validation
Unable to infer Z-Image caption length for rotary embeddings	validation	error	z-image, rotary-embeddings, batch-state, multimodal
Z-Image transformer has no `rotary_emb`. It likely loaded vi	validation	critical	z-image, model-loading, fallback, rotary-embeddings, diffusers
rollout_noise_level must be a number, got {noise!r}	validation	error	rl-rollout, sampling-params, type-validation, config
rollout_noise_level must be finite, got {noise!r}	validation	error	rl-rollout, sampling-params, nan-infinity, validation
rollout_noise_level must be non-negative, got {noise!r}	validation	error	rl-rollout, sampling-params, range-validation
rollout_sde_type must be one of {_VALID_ROLLOUT_SDE_TYPES},	validation	error	rl-rollout, sde, enum-validation, config
Nunchaku SVDQuant is only supported on NVIDIA CUDA GPUs (Amp	validation	error	nunchaku, svdquant, cuda, platform-support, quantization
Nunchaku SVDQuant is currently only supported on Ampere (SM8	validation	error	nunchaku, svdquant, gpu-compatibility, ampere, hopper, quantization
--enable-svdquant requires --transformer-weights-path to be	validation	error	nunchaku, svdquant, missing-path, quantization, startup-config
flash-attn is not installed. Please install it, e.g., `pip i	exception	critical	flash-attn, import-error, dependency-missing, attention, cuda
Invalid threshold_type for topk: {threshold_type}. Choose 'q	validation	error	validation, value-error, moba, attention, config
Invalid threshold_type: {threshold_type}. Choose 'query_head	validation	error	validation, value-error, moba, attention, config
Invalid select_mode: {select_mode}. Choose 'topk' or 'thresh	validation	error	validation, value-error, moba, attention, select-mode
chunk_size should be a int, or a tuple of length 2 or 3, now	validation	error	validation, value-error, chunk-size, moba, shape
module {__name__!r} has no attribute {name!r}	exception	error	attribute-error, env-vars, lazy-loading, config
pipeline_cls must inherit from ComposedPipelineBase	validation	error	type-error, registry, pipeline, validation, subclass
pipeline_config_cls must inherit from PipelineConfig	validation	error	type-error, registry, pipeline-config, validation, subclass
Pipeline '{pipeline_name}' is already registered; pass overw	validation	error	registry, duplicate, value-error, pipeline, idempotency
Model path '{model_path}' is already registered	validation	error	registry, duplicate, value-error, model-path, hf-hub
Model path '{model_path}' is already registered for pipeline	validation	error	registry, pipeline-registration, diffusers, duplicate-entry
Z-Image caption tensor must have rank 2 or 3	validation	error	tensor-shape, cuda-graph, z-image, multimodal, padding
{name}	exception	error	attribute-access, partial-initialization, cuda-graph, wrapper
cache_dit_params must be a dict, got {type(raw).__name__}.	validation	error	cache-dit, request-validation, type-error, config
Unknown cache_dit_params keys: {sorted(unknown)}. Valid keys	validation	error	cache-dit, unknown-key, request-validation, config
cache_dit_params['secondary'] must be a dict, got {type(seco	validation	error	cache-dit, secondary-cache, type-error, request-validation
Unknown cache_dit_params['secondary'] keys: {sorted(unknown)	validation	error	cache-dit, secondary-cache, unknown-key, request-validation
Transformer {transformer.__class__.__name__} has no attribut	validation	error	cache-dit, block-adapter, model-internals, integration
num_inference_steps is required for transformer-only mode. P	validation	error	cache-dit, missing-config, inference-steps, transformer-only
{transformer_cls_name} is not officially supported by cache-	validation	error	cache-dit, dit, block-adapter, unsupported-model, valueerror
Dual-transformer cache-dit is only supported for {sorted(DUA	validation	error	cache-dit, dual-transformer, model-name-registry, valueerror
num_inference_steps is required for dual-transformer mode. P	validation	error	cache-dit, dual-transformer, missing-config, num-inference-steps
Dual transformers for {model_name} must expose cache-dit blo	validation	error	cache-dit, dual-transformer, model-introspection, attribute-mismatch
num_instances must be >= 1, got {num_instances}	validation	error	disaggregation, dispatch-policy, config-validation, empty-instances
Unknown dispatch policy '{name}'. Available: {list(policies.	validation	error	disaggregation, dispatch-policy, unknown-policy-name, factory
Duplicate request_id: {request_id}	validation	error	disaggregation, request-state, duplicate-id, idempotency
Unknown request_id: {request_id}	validation	error	disaggregation, request-state, unknown-id, stale-request
Cannot transition {request_id} from terminal state {old_stat	validation	warning	disaggregation, request-state, state-machine, terminal-state, race-condition
Invalid transition for {request_id}: {old_state.value} -> {n	validation	error	disaggregation, request-state, state-machine, invalid-transition
Group {group_name} is destroyed.	validation	critical	distributed, collective, process-group, weakref
world_size ({world_size}) is less than tensor_parallel_degre	exception	critical	distributed, parallelism, configuration, startup
runtime.response_format must be 'envelope' or 'raw'	validation	error	api, request-validation, response-format, http-400
{field_name} is not valid JSON	validation	error	api, json, request-validation, multipart
{field_name} must be a JSON object	validation	error	api, json, type-validation, http-400
{detail}	http	error	api, http-400, request-validation, error-wrapper
Cosmos3 action input accepts one image field; use a list or	validation	error	cosmos3, observation, image-input, validation
Cosmos3 observation image arrays must use uint8 dtype	validation	error	cosmos3, numpy, dtype, image-input
Cosmos3 observation image arrays must have shape [H, W] or [	validation	error	cosmos3, numpy, shape, image-input
Cosmos3 action prompt must be a string or non-empty list	validation	error	cosmos3, prompt, validation
Cosmos3 action prompt list must contain only strings	validation	error	cosmos3, action-endpoint, prompt-validation, type-validation
Cosmos3 batched action input requires one prompt per image,	validation	error	cosmos3, batching, prompt-validation, cardinality-mismatch
Cosmos3 forward_dynamics produces video; use /v1/videos inst	validation	error	cosmos3, action-mode, endpoint-routing, video-generation
Cosmos3 action endpoint supports action_mode='policy' or 'in	validation	error	cosmos3, action-mode, enum-validation
action_horizon must be a positive integer	validation	error	cosmos3, action-horizon, numeric-validation
Cosmos3 requires num_frames == action_horizon + 1, got num_f	validation	error	cosmos3, num-frames, action-horizon, constraint-violation
Cosmos3 action num_frames must be greater than 1	validation	error	cosmos3, num-frames, numeric-validation
Cosmos3 action_horizon must be divisible by 4 so num_frames	validation	error	cosmos3, temporal-vae, divisibility, action-horizon
Cosmos3 policy input requires an observation image	validation	error	cosmos3, policy-mode, missing-image, input-validation
Cosmos3 inverse_dynamics input requires an observation video	validation	error	cosmos3, inverse-dynamics, missing-video, input-validation
Cosmos3 action requests accept either an image or a video	validation	error	cosmos3, mutually-exclusive-inputs, input-validation
Cosmos3 action batch size {batch_size} exceeds --batching-ma	validation	error	cosmos3, batching, capacity-limit, server-args
Cosmos3 action requests require domain_name or domain_id	validation	error	cosmos3, domain-config, missing-required-field
raw_action_dim is required when only domain_id is provided	validation	error	cosmos3, domain-config, action-dim, missing-required-field
Cosmos3 inverse_dynamics prompt must be a string	validation	error	cosmos3, inverse-dynamics, prompt-validation, type-validation
Unsupported dtype: {obj.dtype}	validation	error	numpy, msgpack, serialization, dtype-validation
image payload requires b64_json	validation	error	image-payload, base64, input-validation, action-endpoint
Action endpoint requires SamplingParams or ActionSamplingPar	validation	error	action-endpoint, sampling-params, model-registry, subclass-validation
output_format must be 'list' or 'numpy'	validation	error	action-endpoint, output-format, enum-validation
Action endpoint is not implemented for {sampling_params_cls.	validation	error	action-endpoint, not-implemented, model-support, dispatch
{response.error}	exception	error	action-inference, scheduler, runtime-error, propagated-error
action policy returned no output	exception	error	action-inference, empty-output, scheduler
action output dimensions must be non-zero, got {tuple(action	validation	error	action-inference, shape-validation, numpy
action output must have shape [H, D] or [B, H, D], got {tupl	validation	error	action-inference, shape-validation, numpy
--diffusers-kwargs must be valid JSON. Got: {args.diffusers_	console	error	cli, json, diffusers, argument-validation
Number of gpus must be positive	console	error	cli, gpu, argument-validation
Config file not found: {args.config}	console	error	cli, config-file, argument-validation
Config file not found: {args.config}	console	error	cli, config-file, serve, argument-validation
The {option_string} option is not yet implemented	console	warning	cli, not-implemented, argparse
Could not connect to remote scheduler at {self.server_args.s	exception	critical	connection, scheduler, remote-mode, startup
When using multiple prompts with multiple input images, prov	validation	error	multimodal, input-validation, image-input, diffusion
Cannot use multiple prompts with a fixed output_file_name. E	validation	error	input-validation, file-output, batch-generation
{output_batch.error}	exception	critical	scheduler, inference-failure, diffusion, runtime-error
generate_action requires an ACTION pipeline, got {sampling_p	validation	error	input-validation, action-pipeline, data-type, diffusion
action policy returned no output	exception	error	action-policy, empty-output, scheduler, diffusion
Prompt text file not found: {path}	exception	error	file-not-found, prompt-file, input-validation, filesystem
No prompts found in file: {path}	validation	error	input-validation, prompt-file, empty-input
Expected {request_count} outputs, got {output_count} from sc	exception	critical	scheduler, consistency-check, batch-generation, diffusion
{failure_msg}: {error_msg}	exception	error	lora, scheduler, adapter-loading, runtime-error
Failed to process image source: {str(e)}	http	error	http, upload, mesh-generation, client-error
Invalid request body: {e}	http	error	http, request-validation, pydantic, mesh-generation
An input image is required for mesh generation	http	error	http, missing-parameter, mesh-generation, unprocessable-entity
Mesh not found	http	error	http, not-found, mesh-generation, job-store
Mesh has been uploaded to cloud storage. Please use the clou	http	warning	http, cloud-storage, redirect, mesh-generation
Generation is still in-progress	http	warning	http, async-job, polling, mesh-generation
prompt event payload must be a string	validation	error	realtime, event-validation, prompt, websocket
camera_actions event payload must be list[list[str]]	validation	error	realtime, event-validation, camera-actions, type-mismatch
unsupported event kind: {event.kind}	validation	error	realtime, event-routing, unsupported-event, websocket
prompt event payload must be a non-empty string	validation	error	realtime, event-validation, prompt, empty-value
composite_input event payload must be a map	validation	error	realtime, event-validation, composite-input
composite_input event payload requires non-empty input_types	validation	error	realtime, event-validation, composite-input, missing-field
composite_input input_types must contain non-empty strings	validation	error	realtime, event-validation, composite-input, type-mismatch
composite_input event payload requires {input_type}	validation	error	realtime, event-validation, composite-input, missing-field
unsupported composite_input type: {input_type}	validation	error	realtime, event-validation, composite-input, unsupported-type
pass only one of camera_actions or action	validation	error	realtime, mutually-exclusive, condition-inputs, sana
Invalid request body: {e}	http	error	http-400, request-validation, video-generation, openai-api
{e}	http	error	http-400, sampling-params, video-generation, out-of-range
Video not found	http	warning	http-404, video-generation, job-not-found, polling
--disagg-server-addr is required for --disagg-role {role_typ	validation	error	disaggregated-serving, cli-args, missing-argument, validation
Role {role_type.value} rank {rank_idx} failed to initialize.	exception	critical	disaggregated-serving, multi-gpu, worker-init, nccl, tensor-parallel
Unknown disagg_role: {role}	validation	error	cli-args, dispatch, enum-validation, disaggregated-serving
Unknown approximate mode: {approximate}	validation	error	activation, gelu, argument-validation, torch
Activation function {act_fn_name!r} is not supported.	validation	error	activation, registry, lookup-failed, model-config
Mode must be one of {valid_modes}, got {mode}	validation	error	sta, attention, mode-validation, argument-validation
mask_candidates is required for STA_searching mode	validation	error	sta, missing-argument, kwargs-validation
mask_search_files_path is required for STA_tuning mode	validation	error	sta, missing-argument, file-path, pipeline-ordering
mask_candidates is required for STA_tuning mode	validation	error	sta, missing-argument, kwargs-validation
mask_search_files_path_pos, mask_search_files_path_neg, and	validation	error	sta, attention, sparse-tuning, kwargs-validation, multimodal
mask_candidates is required for STA_tuning_cfg mode	validation	error	sta, attention, sparse-tuning, kwargs-validation
load_path is required for STA_inference mode	validation	error	sta, attention, sparse-tuning, kwargs-validation
AITer backend does not have a metadata builder.	exception	error	aiter, attention-backend, rocm, not-implemented
AITer backend requires num_heads ({num_heads}) to be a multi	validation	error	aiter, gqa, attention-backend, rocm, shape-validation
AITER Sage backend does not have a metadata builder.	exception	error	aiter, sage-attention, attention-backend, rocm, not-implemented
AITER Sage attention is not available, please update AITER v	exception	critical	aiter, sage-attention, rocm, dependency-version, import-error
{name} is required for NPU packed attention	validation	error	npu, ascend, varlen, attention, packed-sequences
{name} must be a 1D int32 or int64 tensor	validation	error	npu, ascend, varlen, dtype-validation, tensor-shape
{name} and its host copy must have the same length	validation	error	npu, ascend, varlen, host-device-sync, validation
{name} must start with 0 and contain at least one sequence	validation	error	npu, ascend, varlen, packed-sequences, validation
{name} must end at the packed token count {total_tokens}, go	validation	error	npu, ascend, varlen, packed-sequences, shape-validation
{name} must be non-decreasing	validation	error	npu, ascend, varlen, packed-sequences, validation
NPU packed attention requires q, k, and v in [T, N, D] layou	validation	error	npu, ascend, varlen, tensor-layout, shape-validation
NPU packed attention requires q, k, and v on the same NPU; i	validation	error	npu, ascend, device-placement, validation
NPU packed attention requires q, k, and v with the same dtyp	validation	error	npu, ascend, dtype, mixed-precision, validation
NPU packed attention requires matching K/V token and head co	validation	error	npu, ascend, varlen, shape-validation, kv-cache
NPU packed attention requires matching Q/K head dimensions	validation	error	npu, ascend, head-dim, shape-validation
cu_seqlens_q and cu_seqlens_k must describe the same batch	validation	error	npu, ascend, varlen, batch-mismatch, validation
NPU packed attention does not support a sequence that is emp	exception	error	npu, ascend, varlen, ring-attention, not-implemented
Unexpected Ascend TND softmax LSE shape: expected {(q.shape[	exception	critical	ascend, npu, flash-attention, lse, shape-mismatch
{type(self).__name__} does not implement packed varlen atten	exception	error	attention-backend, varlen, not-implemented, abstract-method
{type(self).__name__} does not implement ring KV-chunk atten	exception	error	attention-backend, ring-attention, not-implemented, distributed
Invalid attention metadata values.Sparsity should be in [0,	validation	error	block-sparse, attention, metadata, validation, value-out-of-range
flash_attn_varlen_func_op is out-only op; return_softmax_lse	validation	error	flash-attention, varlen, api-misuse, lse
flash_attn_varlen_func_op_lse is out+lse op; return_softmax_	validation	error	flash-attention, varlen, api-misuse, lse
f"flash attention version {fa_ver} is not supported."	validation	error	flash-attention, version-mismatch, dispatch, config
FlashAttention did not return the softmax LSE required by ri	exception	critical	flash-attention, ring-attention, lse, contract-violation
The required 'attentions' package is not installed. Install	exception	error	import-error, npu, ascend, missing-dependency, laser-attention
Invalid attention metadata values.Sparsity should be in [0,	validation	error	rain-fusion, sparse-attention, metadata, validation, value-out-of-range
GQA/MQA requires query heads to be a multiple of KV heads, g	validation	error	sage-attention, gqa, head-mismatch, validation
f"seq_len {item} not supported for STA"	validation	error	sliding-tile-attention, seq-len, unsupported-value, key-error
st attn not supported	validation	error	sliding-tile-attention, missing-dependency, native-extension, init
SGLANG_DIFFUSION_ATTENTION_CONFIG is not set	validation	error	sliding-tile-attention, missing-env-var, config, init
mask_strategy cannot be None for SlidingTileAttention	validation	error	sliding-tile-attention, mask-strategy, none-check, forward
mask_strategy[0] cannot be None for SlidingTileAttention	validation	error	sliding-tile-attention, mask-strategy, validation, forward
forward_batch cannot be None	validation	error	sliding-tile-attention, forward-context, null-check, runtime-state
Invalid STA_param	validation	error	sliding-tile-attention, metadata, index-out-of-range, prefix-parsing
Unsupported sol_attn dense_backend={dense_backend!r}; expect	validation	error	sol-attn, config, invalid-value, enum
Sol-Attn requires head_size={_SOL_ATTN_HEAD_DIM}, got {head_	validation	error	sol-attn, head-size, unsupported-dimension, init
f"Sol-Attn requires bfloat16 activations, got {q.dtype}"	validation	error	dtype, bfloat16, attention-backend, gpu
f"Unknown feature map: {feature_map}"	validation	error	config, feature-map, validation, constructor
Missing required argument for SparseVideoGen2Attention: {nam	validation	error	kwargs, metadata, attention-backend, validation
raw_latent_shape must be (T, H, W) or (B, C, T, H, W) for SA	validation	error	shape-validation, latent-shape, video-generation
raw_latent_shape must be divisible by patch_size for SAP att	validation	error	shape-validation, patch-size, divisibility, video-generation
Sparse Video Gen 2 attention does not support causal attenti	validation	error	causal-mask, attention-backend, config
Sparse Video Gen 2 attention backend requires svg package to	exception	critical	missing-dependency, installation, attention-backend, optional-package
f"Invalid prefix for SparseVideoGen2AttentionImpl: {prefix}"	validation	error	prefix, layer-index, weight-loading, validation
n_q/n_k must be one of {VALID_N}, got n_q={n_q}, n_k={n_k}	validation	error	hyperparameter, validation, sparse-attention, constructor
All ranges must be within [0, {max_seqlen}], got {range_valu	validation	error	attention, varlen, mask-metadata, range-validation, multimodal
Attention backend override '{target}' resolved to '{resolved	validation	error	attention, backend-override, configuration, enum-mismatch
UlyssesAttention's all-to-all spans the combined sequence pa	exception	critical	attention, ring-parallelism, sequence-parallel, distributed, not-implemented
K/V-gather SP does not support varlen UlyssesAttention.	exception	error	attention, sequence-parallel, kv-gather, varlen, not-implemented
Replicated Q, K, and V must be provided together.	validation	error	attention, replicated-tokens, argument-validation, multimodal
K/V-gather SP does not support video sparse attention.	exception	error	video-sparse-attention, sequence-parallel, kv-gather, not-implemented
Ring Attention requires a backend whose kernel exposes the s	error_code	critical	attention, ring-parallelism, backend-support, lse, initialization
Varlen USPAttention does not support ring parallelism yet.	exception	error	attention, varlen, ring-parallelism, sequence-parallel, not-implemented
USPAttention's masked path does not support replicated prefi	exception	error	attention, sequence-parallel, attention-mask, replicated-tokens, not-implemented
USPAttention masked path supports ring parallelism only for	exception	error	attention, ring-parallelism, attention-mask, batch-size, fa-backend, not-implemented
{selection_error}{component_suffix}	validation	critical	attention-backend, config, multimodal, sglang
No compatible attention backend is available{component_suffi	validation	critical	attention-backend, no-backend-available, multimodal, sglang
Attention backend '{selected_backend}' is not supported by t	validation	error	attention-backend, fail-closed, config-validation, multimodal
Invalid attention backend for {current_platform.device_name}	validation	error	attention-backend, platform-mismatch, device-support
{debug_name} requires cache_head_start when cache heads ({nu	validation	error	kv-cache, gqa, head-slicing, shape-mismatch
Invalid {debug_name} write range: local=[{local_start_index}	error_code	critical	kv-cache, index-out-of-range, assertion, chunking
recent_window_tokens must be non-negative or None	validation	error	kv-cache, sliding-window, argument-validation
Quant-VideoGen KV-cache quantization requires its optional r	exception	error	missing-dependency, optional-extra, kv-cache, quantization, pip-install
QVGPackedCausalKVCache does not support pinned-sink (longliv	exception	error	kv-cache, not-implemented, feature-incompatibility, quantization
{debug_name}: cache_head_start required for head slice	validation	error	kv-cache, gqa, head-slicing, argument-validation
{debug_name}: current-chunk rewrite size changed	exception	error	kv-cache, diffusion, rewrite, shape-mismatch
{debug_name}: non-sequential write current_start={current_ch	exception	error	kv-cache, sequential-write, not-implemented, chunking
recent_window_tokens must be >= 0 or None	validation	error	kv-cache, sliding-window, argument-validation
Expected hidden_size to be {self.hidden_size}, but found: {h	validation	error	layernorm, shape-mismatch, hidden-size, model-config
Expected hidden_size to be at least {self.variance_size_over	validation	error	layernorm, shape-mismatch, variance-override, model-config
Norm type {self.norm_type} not implemented	exception	error	layernorm, norm-type, config-validation, not-implemented
Only gate value of 1 is supported for int type, but got {gat	exception	error	layernorm, gate, argument-validation, cuda-kernel
Gate type {type(gate)} not supported	exception	error	layernorm, gate, type-error, argument-validation
The quantization method `{quantization}` is already exists.	exception	error	quantization, registry, duplicate, config
The quantization config must be a subclass of `QuantizationC	exception	error	quantization, type-validation, subclass, config
Invalid quantization method: {quantization}	exception	error	quantization, lookup, invalid-argument, config
SGLang diffusion currently supports AutoRound auto_gptq chec	exception	error	quantization, auto-round, checkpoint, packing-format
AutoRound fused module {target!r} has inconsistent shard con	exception	error	quantization, auto-round, fused-modules, checkpoint
The input size is not aligned with the quantized weight shap	exception	error	quantization, bitsandbytes, shape-mismatch, alignment
Parameter {param_name} not found in the model.	exception	error	quantization, bitsandbytes, parameter-mapping, checkpoint
bitsandbytes 4-bit TP only supports column-parallel output s	exception	error	quantization, bitsandbytes, tensor-parallel, not-implemented
bitsandbytes 4-bit TP does not support nested quant states.	exception	error	quantization, bitsandbytes, nested-quant, tensor-parallel, not-implemented
bitsandbytes 4-bit TP shard is not aligned to quantization b	exception	error	quantization, bitsandbytes, tensor-parallel, alignment
Comfy full_precision_matrix_mult does not support fused line	validation	error	quantization, fp8, comfy, fused-layers
Unsupported Comfy FP8 layer formats: {unsupported}	validation	error	quantization, fp8, comfy, mixed-precision
ComfyFp8Config must be constructed from safetensors layer ma	validation	error	quantization, fp8, comfy, api-misuse
Comfy INT8 embedding weights support lookup only	validation	error	quantization, int8, embedding, not-implemented
Comfy full_precision_matrix_mult does not support fused line	validation	error	quantization, nvfp4, comfy, fused-layers
Unsupported Comfy NVFP4 companion for {prefix!r}: {marker_fo	validation	error	quantization, nvfp4, comfy, mixed-precision
Comfy NVFP4 layer {prefix!r} must request full_precision_mat	validation	error	quantization, nvfp4, comfy, metadata-validation
comfy_nvfp4 is inferred from per-layer checkpoint metadata;	validation	error	quantization, nvfp4, comfy, api-misuse
Unsupported quantized embedding marker for {prefix!r}: {mark	validation	critical	quantization, checkpoint, embedding, model-load
Unsupported quantized linear marker for {prefix!r}	validation	critical	quantization, checkpoint, linear-layer, model-load
kitchen_int8 group_size must be one of {_SUPPORTED_GROUP_SIZ	validation	error	quantization, config-validation, group-size
Unsupported Comfy INT8 format for {prefix!r}: {marker.get('f	validation	critical	quantization, checkpoint, config-validation
Serialized kitchen_int8 layer {prefix!r} must set convrot=tr	validation	critical	quantization, checkpoint, convrot
Serialized kitchen_int8 layer {prefix!r} must declare convro	validation	critical	quantization, convrot, group-size, checkpoint
Serialized kitchen_int8 layer {prefix!r} has input size {lay	validation	critical	quantization, shape-mismatch, convrot
Serialized W4A4 checkpoints are not supported on MPS	validation	critical	quantization, mps, platform-support, apple-silicon
Serialized W4A4 checkpoints require CUDA compute capability	validation	critical	quantization, cuda, compute-capability, gpu-hardware
Unsupported Comfy W4A4 format for {prefix!r}: {marker_format	validation	critical	quantization, checkpoint, config-validation
kitchen_w4a4 is inferred from per-layer checkpoint metadata;	validation	error	quantization, api-misuse, config
Serialized W4A4 layer {prefix!r} has input size {layer.input	validation	critical	quantization, shape-mismatch, w4a4
Serialized W4A4 layer {prefix!r} has unsupported convrot_gro	validation	critical	quantization, convrot, group-size, validation
Serialized W4A4 layer {prefix!r} has unsupported linear_dtyp	validation	critical	quantization, dtype, validation, w4a4
Serialized W4A8 checkpoints are not supported on MPS	validation	critical	quantization, mps, platform-support, w4a8
Serialized W4A8 checkpoints require CUDA compute capability	validation	critical	quantization, cuda, compute-capability, w4a8
Unsupported Comfy W4A8 format for {prefix!r}: {marker_format	validation	critical	quantization, checkpoint, validation, w4a8
Serialized W4A8 layer {prefix!r} must set convrot=true	validation	critical	quantization, convrot, checkpoint, w4a8
kitchen_w4a8 is inferred from per-layer checkpoint metadata;	validation	error	quantization, api-misuse, w4a8
Unsupported quantized embedding marker for {prefix!r}: {mark	validation	critical	quantization, embedding, checkpoint, w4a8
Unsupported quantized linear marker for {prefix!r}	validation	error	quantization, checkpoint, config-validation
Serialized W4A8 layer {prefix!r} has input size {layer.input	validation	error	quantization, dimension-mismatch, checkpoint
Invalid precision: {self.precision}. Must be 'int4' or 'nvfp	validation	error	quantization, config-validation, nunchaku
Rank must be positive, got {self.rank}	validation	error	distributed, config-validation, nunchaku
QuantoInt8Config must be constructed from safetensors metada	validation	error	quantization, quanto, config-loading
Quanto checkpoint is missing quantization_map_base64	validation	error	quantization, quanto, safetensors, checkpoint-metadata
Invalid Quanto quantization_map_base64	validation	error	quantization, quanto, base64, json, checkpoint-metadata
Quanto quantization map must be a non-empty object	validation	error	quantization, quanto, json-validation, checkpoint-metadata
Quanto quantization map entries must be named objects	validation	error	quantization, quanto, json-validation, checkpoint-metadata
Quanto tensor/map prefixes do not match: missing metadata={s	validation	error	quantization, quanto, checkpoint, int8, multimodal
Unsupported Quanto weight type for {prefix!r}: {quantization	validation	error	quantization, quanto, int8, unsupported-dtype
Quanto activation quantization is not supported for {prefix!	validation	error	quantization, quanto, activations, weight-only
Quanto layer {prefix!r} is missing tensors: {sorted(missing)	validation	error	quantization, quanto, missing-tensor, checkpoint
Quanto layer {prefix!r} contains both packed and dense weigh	validation	error	quantization, quanto, duplicate-weights, checkpoint
Quanto layer {prefix!r} needs a 2D I8 weight, got {data_slic	validation	error	quantization, quanto, dtype, shape-validation
Quanto layer {prefix!r} has incompatible scale {scale_slice.	validation	error	quantization, quanto, scale, shape-validation
Quanto auxiliary scale {scale_name!r} must be a float scalar	validation	error	quantization, quanto, scale, scalar-validation
Quanto layers collide after parameter mapping at {mapped_pre	validation	error	quantization, quanto, name-mapping, collision
Weight input_size_per_partition = {input_size_per_partition}	validation	error	fp8, quantization, tensor-parallel, block-size
Weight output_partition_size = {output_partition_size} is no	validation	error	fp8, quantization, tensor-parallel, shape-mismatch
GGUFConfig must be constructed from a GGUF checkpoint	validation	error	gguf, quantization, config, unsupported-operation
f"Unsupported patch_size type: {type(patch_size)}"	exception	error	config, multimodal, validation, constructor
f"Expected camera embedding shape [B, C, F, H, W], got {tupl	exception	error	multimodal, tensor-shape, validation
f"Input shape {tuple(x.shape)} must be divisible by patch_si	exception	error	multimodal, tensor-shape, patch-embedding
f"The class {type(quant_method).__name__} must implement the	exception	error	quantization, embedding, not-implemented, constructor
Model config does not contain a _class_name attribute. Only	exception	error	model-loading, diffusers, adapter, config
f"Adapter weights at '{component_weights_path}' do not match	exception	error	model-loading, state-dict, adapter, weight-mismatch
Model config does not contain a _class_name attribute. Only	exception	error	model-loading, diffusers, bridge, config
f"No safetensors files found in {component_model_path}"	exception	error	model-loading, safetensors, missing-weights
f"Cannot parse checkpoint quantization for {component_name!r	exception	error	quantization, bitsandbytes, config, model-loading
f"Transformers-managed {component_name!r} quantization requi	exception	error	quantization, bitsandbytes, config, model-loading
{component_name!r} does not support an explicit quantization	validation	error	quantization, server-args, config, model-loading
Failed to load customized {component_name}; native fallback	error_code	critical	model-loading, fallback, quantization, wrapper-exception
Unsupported library: {transformers_or_diffusers}	validation	error	model-loading, config, dispatch
Cannot parse checkpoint quantization metadata for {component	validation	error	quantization, config, model-loading, fail-closed
{component_name!r} checkpoint declares quantization metadata	validation	error	quantization, model-loading, fail-closed, state-dict
Model config does not contain a _class_name attribute. Only	validation	error	model-loading, diffusers, decoder, config
Cannot load PE model: 'model_max_length' not found in {os.pa	error_code	error	model-loading, tokenizer, missing-file, config
The SRT encoder checkpoint adapter supports only serialized	validation	error	quantization, text-encoder, fp8, model-loading
Serialized quantized component weights cannot use a stacked	validation	error	quantization, state-dict, parameter-mapping, text-encoder
A GGUF encoder checkpoint cannot be combined with a second q	validation	error	gguf, quantization, text-encoder, multimodal, checkpoint
{component_name!r} manages its own checkpoint quantization a	validation	error	quantization, online-quantization, text-encoder, unsupported-operation
Cannot configure checkpoint quantization for {component_name	validation	error	quantization, checkpoint-parsing, config-error, text-encoder
{component_name!r} already declares checkpoint quantization;	validation	error	quantization, online-quantization, conflicting-config, text-encoder
Online quantization {explicit_quantization!r} is not support	validation	error	quantization, online-quantization, unsupported-format, text-encoder
A quantized {component_name!r} checkpoint requires an in-tre	validation	error	quantization, architecture-unsupported, text-encoder, tensor-parallel
Cannot parse checkpoint quantization for {component_name!r}:	validation	error	quantization, checkpoint-parsing, gguf, text-encoder
Online quantization for {component_name!r} requires an in-tr	validation	error	quantization, online-quantization, architecture-unsupported, text-encoder
A quantized {component_name!r} checkpoint requires an in-tre	validation	error	quantization, architecture-unsupported, checkpoint, text-encoder
The {component_name!r} checkpoint declares quantization, but	validation	error	quantization, model-mismatch, linear-layers, text-encoder, silent-failure-guard
Rank-local FSDP shard produced for non-DTensor parameter {ta	exception	error	fsdp, dtensor, distributed, weight-loading, sharding
Rank-local TP shard produced for DTensor parameter {target_p	exception	error	tensor-parallel, fsdp, dtensor, distributed, weight-loading
GGUF tensor {tensor.name} declares original shape {logical_s	validation	error	gguf, checkpoint, tensor-shape, validation, diffusion
GGUF tensor {tensor.name} is quantized, but diffusion GGUF c	validation	error	gguf, quantization, tensor-layout, diffusion
GGUF tensor {tensor.name} has inner dimension {inner_dim}, w	validation	error	gguf, quantization, block-alignment, tensor-shape
GGUF tensor {tensor.name} is not aligned to {_GGML_SUPER_BLO	validation	error	gguf, quantization, super-block, alignment
GGUF tensors collide after parameter mapping at {alias!r}	validation	error	gguf, parameter-mapping, name-collision, checkpoint
MiniMax-H3 adaln_t_table must have shape [N, D] with N >= 2,	validation	error	minimax-h3, safetensors, checkpoint, tensor-shape, validation
MiniMax-H3 checkpoint shards disagree on adaln_t_table shape	validation	error	minimax-h3, safetensors, multi-shard, shape-mismatch
Unsupported Comfy NVFP4 companion format(s): + ", ".join(sor	validation	error	minimax-h3, nvfp4, quantization, unsupported-format
MiniMax-H3 NVFP4 metadata requires checkpoint files	validation	error	minimax-h3, nvfp4, missing-argument, api-misuse
--quantization {server_args.quantization} cannot be combined	validation	error	gguf, quantization, config-conflict, transformer-loader
--enable-svdquant cannot be combined with a GGUF transformer	validation	error	gguf, svdquant, nunchaku, config-conflict
GGUF diffusion checkpoints require CUDA; the GGML kernels ha	validation	error	gguf, cuda, platform-unsupported
GGUF diffusion checkpoints are incompatible with FSDP infere	validation	error	gguf, fsdp, distributed, config-conflict
LoRA is not supported on a GGUF transformer: an adapter cann	validation	error	gguf, lora, unsupported-feature
--minimax-h3-adaln-online rebuilds AdaLN outputs from the sa	validation	error	gguf, minimax-h3, adaln, config-conflict
--minimax-h3-adaln-cache-path requires the unquantized trans	validation	error	gguf, minimax-h3, adaln-cache, config-conflict
Resolved GGUF path is not a GGUF file: {resolved}	validation	error	gguf, file-validation, checkpoint
no safetensors files found in {quantized_path or component_m	validation	error	safetensors, checkpoint, file-not-found
GGUF and safetensors quantization metadata conflict	validation	error	gguf, quantization, metadata-conflict, load-spec
Checkpoint quantization is encoded in per-layer metadata; do	validation	error	quantization, config-conflict, checkpoint, server-args
Per-layer checkpoint quantization and Nunchaku are mutually	validation	error	nunchaku, quantization, mutually-exclusive, config-conflict
GGUF is selected by passing the checkpoint itself, not `--qu	validation	error	gguf, quantization, cli-usage, weights-path
Checkpoint at '{model_path}' is incomplete — the following s	exception	critical	safetensors, checkpoint, download, corrupt-checkpoint
No safetensors files found in {model_path}	validation	error	safetensors, missing-weights, model-path
Found {len(safetensors_files)} safetensors files in {model_p	validation	error	safetensors, sharding, ambiguous-checkpoint
unknown weight reader {requested!r}; available: {available_r	validation	error	weight-reader, registry, invalid-name, validation
Duplicate tensor names detected across safetensors files. Re	validation	critical	safetensors, duplicate-keys, checkpoint, weights
Found {len(corrupted_files)} corrupted safetensors file(s).	exception	critical	safetensors, corruption, download, retry
batching config rule from {source} must be an object, got {t	validation	error	batching, config, schema-validation, json
batching config rule requires max_batch_size	validation	error	config, batching, validation, multimodal
batching config rule cannot set both model and model_contain	validation	error	config, batching, validation, mutually-exclusive
batching config rule requires model or model_contains	validation	error	config, batching, validation, selector
batching config rule max_batch_size must be >= 1	validation	error	config, batching, validation, range-check
batching config rule max_cost must be > 0	validation	error	config, batching, validation, range-check
batching config rule device_memory_gb_min must be <= device_	validation	error	config, batching, validation, range-check, gpu-memory
batching config {source} does not contain any rules	validation	error	config, batching, empty-config, startup
batching config schema_version must be 1	validation	error	config, schema-version, batching, version-mismatch
batching config must be a {'schema_version': 1, 'rules': [..	validation	error	config, json, batching, format
batching config rule from {source} contains unknown key(s):	validation	error	config, batching, typo, unknown-key
cannot parse boolean batching config value: {value!r}	validation	error	config, batching, boolean-parsing, validation
Grouped pipeline returned fewer outputs than requests.	error_code	critical	runtime, pipeline, inference, internal-error, multimodal
Expected {len(reqs)} grouped outputs, got {len(output_batch.	error_code	critical	runtime, pipeline, inference, assertion, multimodal
Component {component_name!r} resolved to layerwise-offload,	validation	error	memory, offload, config, component-residency, startup
Component {component_name!r} resolved to component-offload,	validation	error	memory, offload, fsdp, component-residency, config
Invalid component residency assignment: {value!r}	validation	error	config, component-residency, type-error, validation
Component residency must use COMPONENT=MODE, got {assignment	validation	error	config, component-residency, parsing
Invalid component residency assignment: {raw_selector!r}={ra	validation	error	config, component-residency, type-error, validation
Component residency selector cannot be empty	validation	error	config, component-residency, parsing, empty-value
Invalid component residency mode {raw_mode!r} for {selector!	validation	error	config, component-residency, parsing, enum
Invalid layerwise offload component name: {raw_component}.	validation	error	config-validation, layerwise-offload, type-error
failed to move modules to {device}; rollback finished: error	error_code	critical	cuda-oom, device-movement, rollback, memory-offload
sleep/wake does not support FSDP inference	validation	error	fsdp, sleep-wake, feature-unsupported
Cannot update weights while the server is sleeping. Call res	error_code	error	sleep-wake, weight-update, rl-workflow
Server is sleeping. Call resume_memory_occupation first.	error_code	error	sleep-wake, generation, invalid-state
Scheduler terminated after {self._max_consecutive_errors} co	error_code	critical	event-loop, crash-loop, scheduler, circuit-breaker
Expected x.shape[-1] to be even for split rotary, got {last}	validation	error	rope, shape-validation, ltx-2
Only 'rms_norm_across_heads' is supported as a valid value f	validation	error	qk-norm, unsupported-feature, ltx-2
{rope_type=} not supported. Choose between 'interleaved' and	validation	error	rope, enum-validation, ltx-2
The `hidden_states` sequence length {hidden_states.shape[1]}	validation	error	learnable-registers, shape-validation, ltx-2
LTX2DurationHead requires at least one of video_tokens / aud	validation	error	multimodal, duration-head, argument-validation, ltx-2
predict_num_frames supports a single prediction only, got sh	validation	error	batching, duration-head, shape-validation, ltx-2
Unknown interaction strategy: {strategy}	validation	error	config-validation, enum-value, dual-tower, mova
num_heads ({self.num_heads}) must be divisible by tp_size ({	validation	critical	tensor-parallel, model-config, startup, tp-sharding
Invalid direction: {direction}	validation	error	enum-value, routing, dual-tower, forward-pass
head_dim must be a multiple of 8, got {head_dim}.	validation	error	rope, model-config, shape-validation, ltx-2
dim {dim} must be divisible by head_dim {head_dim}.	validation	error	attention, model-config, divisibility, ltx-2
Neighborhood attention requires each dim to be at least its	validation	error	attention, input-shape, video-generation, ltx-2
decoder_model_output_type must be 'x0' or 'v', got {arch.dec	validation	error	config-validation, diffusion, model-config, ltx-2
decoder_stage_channels[{stage_idx + 1}] must be {expected},	validation	error	model-config, channel-dimensions, config-validation, ltx-2
Subclasses of BaseDiT must define '{attr}' class variable	validation	error	sglang, dit, subclass-contract, class-attribute, import-time
Subclass {self.__class__.__name__} must define _supported_at	validation	error	sglang, dit, attention-backend, init-validation
Subclasses of BaseDiT must define '{attr}' instance variable	validation	error	sglang, dit, post-init, attribute-validation
Cosmos3CausalAttention requires num_attention_heads divisibl	validation	error	sglang, cosmos3, tensor-parallel, attention-heads, config
Cosmos3CausalAttention requires num_key_value_heads divisibl	validation	error	sglang, cosmos3, tensor-parallel, gqa, kv-heads
Cosmos3CrossAttention requires num_attention_heads divisible	validation	error	sglang, cosmos3, tensor-parallel, cross-attention
Cosmos3CrossAttention requires num_key_value_heads divisible	validation	error	sglang, cosmos3, tensor-parallel, cross-attention, kv-heads
Cosmos3 requires text_ids and text_mask to be passed	validation	error	sglang, cosmos3, text-conditioning, forward-validation, required-argument
Cosmos3 action generation does not support sequence parallel	validation	error	sglang, cosmos3, action-generation, sequence-parallel, unsupported-feature
Invalid mode: {mode}, must be one of 'write', 'read', 'skip'	validation	error	sglang, glm-image, kv-cache, mode-validation, enum-value
unknown qk_norm: {qk_norm}. Should be one of None, 'layer_no	validation	error	config, validation, diffusion, qk-norm
unknown norm_type {norm_type}	validation	error	config, validation, diffusion, norm
Unknown history_scale_mode: {history_scale_mode}	validation	error	config, validation, attention, diffusion
encoder_hidden_states is required when encoder_key_value is	validation	error	runtime, attention, missing-argument, diffusion
Hidden size {hidden_size} must be divisible by num_heads {nu	validation	error	config, validation, transformer, diffusion
Got {axes_dim} but expected positional dim {pe_dim}	validation	error	config, validation, rope, diffusion
Didn't get guidance strength for guidance distilled model.	validation	error	runtime, diffusion, guidance-distillation, missing-argument
Hunyuan3D reference attention requires a shared cache.	validation	error	runtime, diffusion, reference-attention, missing-cache
Reference attention was not initialized.	panic	error	runtime, initialization, reference-attention, diffusion
Multiview attention was not initialized.	panic	error	runtime, initialization, multiview, diffusion
Expected BasicTransformerBlock, got {type(transformer).__nam	validation	error	initialization, diffusers, type-mismatch
Unexpected SD2 mid block: {type(mid_block).__name__}.	validation	error	initialization, diffusers, type-mismatch
Position map {height}x{width} is not divisible by {grid_reso	validation	error	validation, shape-mismatch, mask, diffusion
Hunyuan3D Paint does not use extra UNet conditioning.	validation	error	runtime, api-misuse, diffusion, unet
Hunyuan3D Paint does not use added conditioning.	validation	error	runtime, api-misuse, diffusion, unet
Hunyuan3D Paint expects square latents and a matching view c	validation	error	runtime, shape-mismatch, multiview, diffusion
Got {config.rope_axes_dim} but expected positional dim {pe_d	validation	error	config, validation, rope, video-diffusion
teacache is not supported yet for HunyuanVideo	validation	error	unsupported-feature, teacache, video-diffusion, runtime
vis_freqs_cis is required for fused QK-Norm + RoPE kernel	validation	error	runtime, rope, missing-argument, diffusion
vis_freqs_cis must be a 2D cos_sin_cache tensor	validation	error	runtime, rope, shape-mismatch, diffusion
Fused QK-Norm + RoPE kernel only supports float16/bfloat16,	validation	error	dtype, rope, fused-kernel, joyimage, multimodal
txt_freqs_cis must be a 2D cos_sin_cache tensor	validation	error	rope, shape-validation, joyimage, multimodal
Hidden size {self.hidden_size} must be divisible by num_atte	exception	critical	config-validation, attention-heads, init-time, joyimage
JoyImage conditioning batch mismatch: hidden_states batch={b	exception	error	batch-mismatch, cfg-conditioning, joyimage, multimodal
LingBotVideoBlock expects token-level temb6 with shape (B*S,	exception	error	shape-validation, timestep-embedding, lingbot, video-diffusion
c2ws_plucker_emb shape must match hidden_states shape, got {	exception	error	shape-validation, camera-conditioning, plucker, lingbot
num_heads ({self.num_heads}) must be divisible by ulysses_de	exception	critical	parallelism, ulysses, attention-heads, lingbot, config-validation
LingBot causal sequence sharding currently requires kv_cache	exception	error	not-implemented, sequence-parallelism, kv-cache, lingbot
LingBot causal sequence sharding requires forward_batch.sequ	exception	error	sequence-parallelism, missing-attribute, forward-batch, lingbot
Unsupported qk_norm: {qk_norm}	exception	critical	config-validation, qk-norm, init-time, lingbot
LingBotWorld requires cross_attn_norm=True	exception	critical	config-validation, cross-attention, init-time, lingbot
LingBot causal sequence sharding currently supports ulysses_	exception	error	not-implemented, sequence-parallelism, ring-attention, lingbot
LingBot causal sequence sharding currently requires ulysses_	exception	error	sequence-parallelism, config-mismatch, lingbot
Expected x.shape[-1] to be even for split rotary, got {last}	exception	error	rope, shape-validation, ltx2, even-dimension
{rope_type=} not supported. Choose between 'interleaved' and	exception	critical	config-validation, rope, init-time, ltx2
Modality {modality} is not supported. Supported modalities a	exception	critical	config-validation, modality, init-time, ltx2
Unknown activation function: {act_fn}	exception	critical	config-validation, activation, init-time, ltx2
Invalid {tp_size=}. Expected tp_size >= 1.	exception	error	parallelism, tensor-parallel, init-order, distributed, ltx2
LTX2Attention requires heads divisible by tp_size, got {self	exception	critical	parallelism, tensor-parallel, attention-heads, config-validation, ltx2
LTX2Attention requires inner_dim divisible by tp_size, got {	exception	critical	parallelism, tensor-parallel, inner-dim, config-validation, ltx2
out_channels must be divisible by tp_size for TP-sharded out	exception	error	tp-sharding, divisibility, ltx-2, config-validation, tensor-parallel
audio_out_channels must be divisible by tp_size for TP-shard	exception	error	tp-sharding, audio, divisibility, ltx-2, config-validation
num_frames/height/width must be provided for RoPE coordinate	exception	error	rope, missing-argument, ltx-2, forward, video
audio_num_frames must be provided for RoPE coordinate genera	exception	error	rope, audio, missing-argument, ltx-2, forward
perturbation_configs length must match batch size, got {len(	exception	error	batch-mismatch, perturbation, flow-matching, ltx-2, forward
Incomplete Diffusers H3 fused parameters: {incomplete}	exception	error	checkpoint-loading, weight-mapping, diffusers, minimax-h3, state-dict
MiniMaxH3DiTModel.forward requires kwarg {key!r}	exception	error	missing-argument, forward-contract, minimax-h3, kwargs-validation
qkv weight has incompatible output dim for grouped checkpoin	exception	error	qkv, gqa, checkpoint-loading, weight-reorder, shape-mismatch
img_position_ids must be [1, S, 3], got {list(img_position_i	exception	error	rope, position-ids, shape-mismatch, minimax-h3, input-validation
MiniMax H3 ring parallelism requires the FlashAttention back	exception	error	ring-parallelism, attention-backend, flashattention, minimax-h3, not-implemented
MiniMax H3 requires subblock_sparse_query_block_mask when Su	exception	error	minimax-h3, sparse-attention, dit, mask-required
MiniMax H3 attention heads must be divisible by TP size: {ar	exception	critical	minimax-h3, tensor-parallel, config-validation, startup
adaln out_features mismatch: {out_features} != {expand_ratio	validation	error	minimax-h3, adaln, shape-mismatch, config-validation
MiniMax H3 AdaLN cache takes exactly one of path (prebuilt s	validation	error	minimax-h3, adaln-cache, argument-validation
MiniMax H3 AdaLN cache max_plans must be positive	validation	error	minimax-h3, adaln-cache, argument-validation
MiniMax H3 AdaLN cache max_plan_width must be positive; set	validation	error	minimax-h3, adaln-cache, server-args, argument-validation
MiniMax H3 AdaLN cache does not exist: {self.path}	validation	error	minimax-h3, adaln-cache, file-not-found
MiniMax H3 AdaLN cache has an unsupported or missing format_	validation	error	minimax-h3, adaln-cache, version-mismatch, safetensors
MiniMax H3 AdaLN cache model_variant does not match the load	validation	error	minimax-h3, adaln-cache, variant-mismatch
MiniMax H3 AdaLN cache has invalid timestep plans	validation	error	minimax-h3, adaln-cache, corrupt-cache, validation
TP size must be positive.	validation	critical	config, tensor-parallel, minimax-h3, validation, init
num_attention_heads must be positive.	validation	critical	config, model-architecture, validation, minimax-h3
hidden_size must be positive.	validation	critical	config, model-architecture, validation, minimax-h3
attention_head_dim must be positive.	validation	critical	config, model-architecture, validation, minimax-h3
ffn_hidden_size must be positive.	validation	critical	config, model-architecture, validation, minimax-h3
MiniMax H3 {name}={value} must be divisible by TP size {tp_s	validation	critical	tensor-parallel, divisibility, config, minimax-h3, launch
MiniMax H3 Ulysses size must be positive.	validation	critical	sequence-parallel, ulysses, config, validation
MiniMax H3 ring size must be positive.	validation	critical	ring-attention, sequence-parallel, config, validation
MiniMax H3 TP-local heads {local_heads} must be divisible by	validation	critical	ulysses, sequence-parallel, divisibility, tensor-parallel
MiniMax H3 packed sequence alignment {MINIMAX_H3_PACKED_SEQU	validation	critical	sequence-parallel, ulysses, ring-attention, alignment, divisibility
MiniMax H3 AdaLN cache is only compatible with unquantized w	validation	error	adaln, quantization, checkpoint, config-conflict
MiniMax H3 pruned curve checkpoints cannot use a separate Ad	validation	error	adaln, checkpoint, config-conflict, minimax-h3
{name} must stay fp32 after load, got {param.dtype}.	validation	error	dtype, fp32, weight-loading, adaln
{name} must stay fp32 with curve AdaLN, got {param.dtype}.	validation	error	dtype, fp32, adaln, weight-loading
rope.inv_freq must stay fp32 after load, got {rope_inv_freq.	validation	error	dtype, fp32, rope, weight-loading
{key}.position_ids is required	validation	error	input-validation, position-ids, multimodal, forward
{key}.{field} is required	validation	error	input-validation, multimodal, missing-field
refiner cu_seqlens live text length must be in [1, {int(prom	validation	error	input-validation, cu-seqlens, off-by-one, batching
packed seq_len {seq_len} not divisible by the combined seque	validation	error	sequence-parallel, ulysses, ring-attention, divisibility, padding
refiner cu_seqlens live text length must be in [1, {int(text	validation	error	minimax-h3, dit, shape-validation, text-embeddings
refined prompt embeddings must have hidden width {self.hidde	validation	error	minimax-h3, hidden-size, embedding-width
MiniMaxH3DiTModel.forward received unexpected kwargs: {unexp	validation	error	minimax-h3, kwargs-contract, api-signature
x must be [1, S, C], got {list(x.shape)}	validation	error	minimax-h3, input-shape, packed-sequence
token_tags must cover the full packed sequence ({seq_len}),	validation	error	minimax-h3, token-tags, packed-sequence
inverse_indices must be [{seq_len}], got {list(inverse_indic	validation	error	minimax-h3, inverse-indices, packed-sequence
subblock_sparse_query_block_mask must be a tensor	validation	error	minimax-h3, sparse-attention, type-validation
TP-local heads {local_heads} not divisible by Ulysses world	validation	error	minimax-h3, ulysses, attention-heads, tensor-parallel
block_token_tags must cover the rank-local packed sequence (	validation	error	minimax-h3, token-tags, sequence-parallelism
update_mask length mismatch: {update_mask.shape[0]} != {vide	validation	error	minimax-h3, update-mask, tensor-parallel, logits
Invalid VAE type: {self.vae_type}	validation	error	mova-audio-dit, vae-type, unsupported-config
num_heads ({self.num_heads}) must be divisible by tp_size ({	validation	error	mova-video-dit, tensor-parallel, attention-heads, divisibility
When additional_t_cond is True, addition_t_cond must be prov	validation	error	qwen-image, timestep-conditioning, missing-argument
image_rotary_emb must be cos_sin_cache tensors	validation	error	qwen-image, rope, cos-sin-cache, format-validation
SANA forward pass requires encoder_hidden_states	validation	error	sana, encoder-hidden-states, missing-argument
SANA-Video checkpoints with embedded guidance are not suppor	validation	error	sana-video, guidance-embeds, unsupported-checkpoint
SANA-Video requires encoder_hidden_states	validation	error	sana-video, encoder-hidden-states, missing-argument
SANA-WM plucker_embedder is not initialized.	validation	error	sana-wm, plucker-embedder, camera-conditioning, uninitialized-module
plucker_emb token count {plucker_emb.shape[1]} != latent tok	validation	error	sana-wm, shape-mismatch, camera-embedding, validation
SANA-WM forward requires encoder_hidden_states.	validation	error	sana-wm, missing-argument, required-parameter
SANA-WM forward requires timestep.	validation	error	sana-wm, missing-argument, timestep
SANA-WM camera_conditions must be sampled at latent frames:	validation	error	sana-wm, camera-conditions, shape-mismatch, latent-frames
SANA-WM forward_long requires encoder_hidden_states.	validation	error	sana-wm, forward-long, missing-argument
SANA-WM forward_long requires timestep.	validation	error	sana-wm, forward-long, timestep, missing-argument
chunk_size must be > 0, got {chunk_size}.	validation	error	sana-wm, chunking, invalid-argument, validation
T must be > 0, got {T}.	validation	error	sana-wm, chunking, empty-tensor, temporal-dim
Unknown chunk_split_strategy '{strategy}'. Supported: unifor	validation	error	sana-wm, chunking, invalid-enum, config
Either chunk_index or chunk_size must be provided.	validation	error	sana-wm, chunking, missing-argument
chunk_index must be strictly increasing, got {normalized}.	validation	error	sana-wm, chunking, invalid-argument, ordering
Unsupported SANA-WM update_rule: {self.update_rule}	validation	error	sana-wm, gdn, invalid-enum, config
Unsupported SANA-WM cam_update_rule: {self.cam_update_rule}	validation	error	sana-wm, gdn, camera-branch, invalid-enum
Unsupported SANA-WM gdn_backend: {self.gdn_backend}. Expecte	validation	error	sana-wm, gdn, backend-selection, invalid-enum
SANA-WM Triton GDN backend unavailable: {reason}	error_code	error	sana-wm, gdn, triton, backend-fallback, cuda
SANA-WM Triton camera GDN backend unavailable: {precheck_rea	error_code	error	sana-wm, gdn, triton, camera-branch, no-grad
SANA-WM Triton camera GDN backend unavailable: {reason}	error_code	error	sana-wm, gdn, triton, camera-branch, shape-constraints
Unexpected RoPE rank: {cos.ndim}	validation	error	sana-wm, rope, tensor-rank, refiner
num_frames/height/width are required when hidden_states is p	validation	error	sana-wm, refiner, missing-argument, pre-packed-latents
The native SD2 UNet currently supports only the Hunyuan3D fo	validation	error	stable-diffusion, unet, config-validation, hunyuan3d
Hunyuan3D SD2.1 UNet requires four channel stages.	validation	error	stable-diffusion, unet, config-validation, hunyuan3d
Hunyuan3D SD2.1 UNet requires two ResNet layers and one tran	validation	error	stable-diffusion, unet, config-validation, hunyuan3d
Hunyuan3D SD2.1 checkpoints require linear projection.	validation	error	stable-diffusion, unet, config-validation, hunyuan3d
Expected a 2D, 3D, or 4D attention mask, got {attention_mask	validation	error	attention-mask, shape-validation, stable-diffusion
Unsupported native SD cross-attention arguments: {sorted(uns	validation	error	cross-attention, unsupported-argument, stable-diffusion
The Hunyuan3D SD2.1 UNet has no added conditioning.	validation	error	unet, unsupported-argument, hunyuan3d, stable-diffusion
T2I adapter residuals are not supported by Hunyuan3D.	validation	error	t2i-adapter, unsupported-argument, hunyuan3d
ControlNet down and mid residuals must be provided together.	validation	error	controlnet, argument-pairing, unet
class_labels are required by this UNet.	validation	error	unet, class-conditional, missing-argument
encoder_hidden_states must be provided.	validation	error	stable-diffusion-3, missing-argument, text-embedding
pooled_projections must be provided.	validation	error	stable-diffusion-3, missing-argument, pooled-embedding
QK Norm type not supported	exception	critical	wanvideo, qk-norm, config-validation, init
Z-Image expects one caption embedding per image, got {len(al	validation	error	z-image, batching, shape-validation
Z-Image batch must contain at least one image latent	validation	error	z-image, empty-batch, validation
caption_valid_mask must have one row per Z-Image caption	validation	error	z-image, mask-validation, batching
Cannot pad RoPE freqs of length {cos.shape[0]} to shorter ta	validation	error	z-image, rope, shape-validation
Unsupported encoder folding mode: {mode!r}	validation	error	encoder-folding, config-validation, tp-group
Subclass {self.__class__.__name__} must define _supported_at	validation	error	encoder, subclass-contract, attention-backend
You have to specify input_ids	validation	error	clip, text-encoder, input-validation, multimodal
The original encoder only has {num_hidden_layers} layers, bu	validation	error	clip, config, model-init, layer-count
You must specify exactly one of input_ids or inputs_embeds	validation	error	gemma2, encoder, input-validation, mutually-exclusive
H3 conditioning projection {bias_name} has shape {tuple(bias	validation	critical	minimax-h3, conditioning-projection, shape-mismatch, checkpoint
H3 conditioning projection contains unsupported tensors: {so	validation	critical	minimax-h3, conditioning-projection, unsupported-tensors, checkpoint
H3 conditioning projection has neither W nor an MLP	exception	critical	minimax-h3, conditioning-projection, empty-checkpoint
H3 conditioning projection MLP outputs width {layer_input_di	exception	critical	minimax-h3, conditioning-projection, shape-mismatch, width-mismatch
H3 conditioning projection W has shape {tuple(self.weight.sh	exception	critical	minimax-h3, conditioning-projection, shape-mismatch, transposed-weight
H3 conditioning projection expects width {self.input_dim}, g	exception	critical	minimax-h3, conditioning-projection, forward, width-mismatch
H3 conditioning projection produced no output	exception	error	minimax-h3, conditioning-projection, no-output, defensive
MiniMax H3 Qwen3-VL encoders smaller than 32B require --comp	exception	critical	minimax-h3, conditioning-projection, missing-component, server-args
H3 conditioning projection expects encoder width {input_dim}	exception	critical	minimax-h3, conditioning-projection, width-mismatch, config-validation
H3 conditioning projection must output width {MINIMAX_H3_QWE	exception	critical	minimax-h3, conditioning-projection, width-mismatch, config-validation
H3 conditioning projection tap {tap} is outside the selected	exception	critical	minimax-h3, conditioning-projection, layer-index, out-of-range
MiniMax H3 Qwen3-VL language-layer configuration is inconsis	exception	critical	minimax-h3, config, layer-count, inconsistent-config
input_ids must be 1-D, got {list(input_ids.shape)}	exception	error	minimax-h3, encode-ids, input-shape, rank-mismatch
pixel_values and image_grid_thw must be given together	exception	error	minimax-h3, encode-ids, paired-args, image-input
pixel_values_videos and video_grid_thw must be given togethe	exception	error	minimax-h3, encode-ids, paired-args, video-input
unexpected hidden shape {list(hidden.shape)}, expected {expe	exception	error	minimax-h3, encode-ids, output-shape, sanity-check
Unexpected MiniMax H3 Qwen3-VL checkpoint weight: {name} (ma	exception	critical	minimax-h3, load-weights, unknown-key, checkpoint
Failed to load MiniMax H3 Qwen3-VL weight {name!r}: checkpoi	exception	critical	minimax-h3, load-weights, weight-loading, tensor-parallel
You must specify exactly one of input_ids or inputs_embeds	exception	error	mistral-3, forward, mutually-exclusive-args, input-validation
Qwen-VL position_ids do not match the attention input	exception	error	qwen-vl, rope, position-ids, shape-mismatch, multimodal
shard_offset and shard_size must be provided	exception	error	weight-loading, column-parallel, shard, checkpoint
Error raised in subprocess: {returned.stderr.decode()}	exception	error	registry, subprocess, model-loading, import-error
Expected a string in the format `<module>:<class>`	validation	error	registry, model-registration, lazy-import, validation
Model architectures {architectures} failed to be inspected.	validation	error	registry, model-resolution, architecture-unsupported
Model architectures {architectures} are not supported for no	validation	error	registry, unsupported-architecture, model-resolution
Unsupported model architecture: {arch}. Registered architect	exception	error	registry, alias, architecture, unsupported-architecture
Subclasses of BaseScheduler must define '{attr}' property	exception	error	scheduler, abstract-base, subclass-contract, diffusion
pair_postprocess must be callable or None	validation	error	scheduler, type-error, postprocess, flow-match
Scheduler not initialized; call set_timesteps() first	exception	error	scheduler, initialization-order, flow-match, state-error
pairs must be a torch.Tensor of shape [N, 2]	validation	error	pytorch, tensor-shape, flow-matching, scheduler, validation
pairs must be a torch.Tensor	validation	error	pytorch, type-check, scheduler, sigma-shift, validation
pairs length must be greater than 0	validation	error	pytorch, empty-tensor, scheduler, sigma-shift, validation
source must be 'timesteps' or 'sigmas'	validation	error	scheduler, enum-validation, api-misuse, sigma-shift
shift must be positive	validation	error	scheduler, config-validation, sigma-shift, flow-matching
denoising_strength must be positive	validation	error	scheduler, flow-matching, validation, diffusion, config
exponential_shift enabled but exponential_shift_mu is missin	exception	error	scheduler, flow-matching, missing-parameter, exponential-shift
Unknown pair_postprocess name: {name}	validation	error	scheduler, invalid-name, enumeration
vec must be 1D	validation	error	scheduler, tensor-shape, validation
pair_postprocess must return a torch.Tensor	validation	error	scheduler, type-mismatch, callback
pair_postprocess must return the same shape as input	validation	error	scheduler, tensor-shape, callback
Must pass a value for `mu` when `use_dynamic_shifting` is Tr	validation	error	scheduler, diffusion, missing-parameter, dynamic-shifting
Passing integer indices as timesteps is not supported. Pass	validation	error	scheduler, timestep-type, type-mismatch
Passing integer indices as timesteps is not supported.	validation	error	scheduler, timestep-type, consistency-model
Only one of `config.use_beta_sigmas`, `config.use_exponentia	validation	error	scheduler, config, mutually-exclusive, diffusion
`time_shift_type` must either be 'exponential' or 'linear'.	validation	error	scheduler, invalid-enum-value, config
Unknown time_shift_type: {self.config.time_shift_type}	validation	error	scheduler, invalid-enum-value, config-mutation
`mu` must be passed when `use_dynamic_shifting` is set to be	validation	error	scheduler, diffusion, missing-parameter, dynamic-shifting
`sigmas` and `timesteps` should have the same length	validation	error	scheduler, length-mismatch, validation
`sigmas` and `timesteps` should have the same length as num_	validation	error	scheduler, length-mismatch, validation
Either num_inference_steps, sigmas, or timesteps must be pro	validation	error	scheduler, missing-argument, validation
Passing integer indices (e.g. from `enumerate(timesteps)`) a	validation	error	scheduler, timestep-type, flow-matching
Number of inference steps is 'None', run 'set_timesteps' fir	validation	error	diffusion, scheduler, state-not-initialized, helios
Scheduler type '{self.config.scheduler_type}' not implemente	exception	error	diffusion, scheduler, not-implemented, config-mismatch
{name} must be finite	validation	error	diffusion, scheduler, nan, numerical-stability, validation
{name} must be a torch.Tensor	validation	error	diffusion, scheduler, type-validation, timestep
{name} must be a floating point tensor	validation	error	diffusion, scheduler, dtype, timestep
{beta_schedule} is not implemented for {self.__class__}	validation	error	scheduler, diffusion, config-validation, unipc
{solver_type} is not implemented for {self.__class__}	validation	error	scheduler, diffusion, solver, unipc
{self.config.timestep_spacing} is not supported. Please make	validation	error	scheduler, diffusion, timesteps, unipc
`final_sigmas_type` must be one of 'zero', or 'sigma_min', b	validation	error	scheduler, diffusion, sigmas, karras, unipc
missing `sample` as a required keyword argument	validation	error	scheduler, diffusion, api-misuse, unipc
prediction_type given as {self.config.prediction_type} must	validation	error	scheduler, diffusion, prediction-type, unipc
prediction_type given as {self.config.prediction_type} must	validation	error	scheduler, diffusion, prediction-type, unipc
missing `order` as a required keyword argument	validation	error	scheduler, diffusion, api-misuse, unipc
missing `last_sample` as a required keyword argument	validation	error	scheduler, diffusion, api-misuse, unipc
missing `this_sample` as a required keyword argument	validation	error	scheduler, diffusion, api-misuse, unipc
Number of inference steps is 'None', you need to call 'set_t	validation	error	scheduler, diffusion, initialization, unipc
Unsupported dims: {self.dims}	validation	error	upsampler, multimodal, video, tensor-shape
Unsupported scale {scale}. Choose from {list(mapping.keys())	validation	error	upsampler, scale, config, value-error
Either spatial_upsample or temporal_upsample must be True	validation	error	upsampler, config, init, value-error
A dict of processors was passed, but the number of processor	validation	error	attention, processor, autoencoder, value-error
Cannot call `set_default_attn_processor` when attention proc	validation	error	attention, processor, fusion, state-error
`fuse_qkv_projections()` is not supported for models having	validation	error	fusion, attention, optimization, value-error
A dict of processors was passed, but the number of processor	validation	error	attention, processor, flux2, value-error
Cannot call `set_default_attn_processor` when attention proc	validation	error	attention, processor, flux2, state-error
Cosmos3 AVAE dec_strides product must equal hop_size: produc	validation	error	avae, config, cosmos3, init, value-error
Unsupported AVAE normalization_type={self.normalization_type	validation	error	avae, normalization, decode, config
Unsupported topk_mode {topk_mode}	validation	error	flashvdm, topk, hunyuan3d, config, value-error
Please install diso via `pip install diso`, or set mc_algo t	validation	error	python, import-error, missing-dependency, diso, marching-cubes, hunyuan3d
Unsupported mc_algo {mc_algo}, available: {list(SurfaceExtra	validation	error	python, value-error, invalid-argument, mc-algo, hunyuan3d, flashvdm
Unsupported down_block_type: {down_block_type}	validation	error	python, value-error, model-config, hunyuan-vae, unsupported-block
Unsupported time_compression_ratio: {temporal_compression_ra	validation	error	python, value-error, model-config, hunyuan-vae, compression-ratio
Unsupported up_block_type: {up_block_type}	validation	error	python, value-error, model-config, hunyuan-vae, unsupported-block
Unsupported time_compression_ratio: {time_compression_ratio}	validation	error	python, value-error, model-config, hunyuan-vae, compression-ratio
Unsupported LTX-2.3 encoder block: {block_name}	validation	error	python, value-error, model-config, ltx-2-3, encoder-block
Unsupported latent_log_var: {latent_log_var}	validation	error	python, value-error, model-config, ltx-2-3, latent-log-var
Invalid causality_axis: {causality_axis}	validation	error	python, value-error, invalid-argument, ltx-2-audio, causal-conv
Unknown Pi05 Gemma variant: {variant}	validation	error	pi05, gemma, config, invalid-variant
dimension ({dimension}) must be divisible by 2	validation	error	pi05, sinusoidal-embedding, dimension-validation
time must have shape [batch]	validation	error	pi05, tensor-rank, input-validation
pad_masks and att_masks must be [batch, seq]	validation	error	pi05, attention-mask, tensor-rank, input-validation
Invalid Pi05 precision: {precision}	validation	error	pi05, dtype, precision, config
Unsupported Pi05 dtype: {dtype_name}	validation	error	pi05, dtype, checkpoint, config
Pi05 weight load failed: {len(missing)} missing weights, {mi	error_code	critical	pi05, weight-loading, checkpoint, state-dict
Pi05 action state is missing on single-rank run	error_code	error	pi05, distributed, sequence-parallel, runtime-state
Pi05 action state broadcast returned None	error_code	error	pi05, distributed, broadcast, sequence-parallel, nccl
Pi05 action fallback must run on the action root	error_code	error	pi05, distributed, action-root, sequence-parallel
Expected Hunyuan3D2PipelineConfig, got {type(config)}	validation	error	hunyuan3d, pipeline-config, type-mismatch, sglang
server_args is required to resolve Ideogram4 NVFP4 paths	validation	error	ideogram, nvfp4, missing-argument, lazy-init, sglang
JoyEchoPipeline requires JoyEchoPipelineConfig, got {type(co	validation	error	joyecho, pipeline-stages, type-mismatch, sglang
--load-diffusion-decoder was requested, but this checkpoint	validation	error	ltx2, diffusion-decoder, checkpoint-manifest, sglang
--model-variant {server_args.model_variant} requires '{cls._	validation	error	ltx2, model-variant, partial-download, checkpoint, sglang
Invalid ltx2_two_stage_device_mode={mode!r}. Expected one of	validation	error	ltx2, device-mode, env-var, invalid-value, sglang
{self.pipeline_name} requires --spatial-upsampler-path (comp	validation	error	ltx2, spatial-upsampler, missing-path, component, sglang
{self.pipeline_name} requires --distilled-lora-path (compone	validation	error	ltx2, distilled-lora, missing-path, component, sglang
MiniMax H3 requires ffmpeg and ffprobe for media processing	exception	critical	minimax-h3, ffmpeg, missing-system-dependency, docker, sglang
MiniMax H3 model variant must be a non-empty string	validation	error	minimax-h3, model-variant, input-validation, sglang
unsupported MiniMax H3 model variant {variant!r}; supported:	validation	error	minimax-h3, model-variant, unsupported-value, sglang
MiniMax H3 --model-variant and --model-subfolder select diff	validation	error	minimax-h3, model-variant, model-subfolder, conflicting-config, sglang
MiniMax H3 loaded checkpoint partition does not match --mode	validation	error	minimax-h3, checkpoint-mismatch, model-variant, integrity-check, sglang
MiniMaxH3Pipeline only supports monolithic deployment; disag	validation	error	minimax-h3, disaggregation, monolithic-only, sglang
Pi05Pipeline v1 supports same-process execution only. Use pr	validation	error	pi05, vla, disaggregation, monolithic-only, sglang
VLA action expert should not share the prefix TP layout. Use	validation	error	pi05, vla, parallelism-strategy, tp, config, sglang
Encoded prompt has {tensor.shape[1]} tokens, expected at lea	validation	error	sana-video, prompt-window, sequence-length, validation, sglang
SANA-WM does not support tensor parallelism yet. Use --num-g	validation	error	sana-wm, tensor-parallelism, unsupported-feature, sglang
SANA-WM does not support temporal sequence parallelism yet.	validation	error	sana-wm, sequence-parallelism, unsupported-feature, sglang
Unknown image_vae_encoding_position: {image_vae_encoding_pos	validation	error	config-validation, pipeline, multimodal, typo
[pred_noise_to_pred_video] Invalid timestep shape: {timestep	validation	error	tensor-shape, diffusion, scheduler, validation
Error on rank 0	exception	critical	distributed, multi-gpu, rank-failure, error-propagation
Error on rank 0: {broadcasted_batch}	exception	critical	distributed, multi-gpu, rank-failure, error-propagation
safetensors metadata {key!r} must be a positive integer	validation	error	lora, safetensors, metadata, peft
conflicting safetensors LoRA alpha metadata: {declared}	validation	error	lora, safetensors, metadata, conflict
PEFT adapter_config.json must contain a JSON object	validation	error	lora, peft, json, config
adapter_config.json lora_alpha conflicts with safetensors me	validation	error	lora, peft, metadata, conflict
PEFT lora_alpha must be a positive integer	validation	error	lora, peft, validation, config
Invalid LoRA merge mode: {merge_mode}. Valid modes: {LORA_ME	validation	error	lora, config-validation, diffusion-pipeline
Length mismatch: lora_nicknames has {len(lora_nicknames)} it	validation	error	lora, argument-validation, length-mismatch
Length mismatch: lora_nicknames has {len(lora_nicknames)} it	validation	error	lora, argument-validation, length-mismatch
Dit target weight name {target_name} already exists in lora_	validation	error	lora, checkpoint, duplicate-key
Invalid target(s): {invalid_targets}. Valid targets: {self.V	validation	error	lora, config-validation, invalid-target
Adapter {nickname} not found in the pipeline. Please provide	validation	error	lora, missing-adapter
Dynamic LoRA currently supports only one adapter per target.	validation	error	lora, merge-mode, multi-adapter
'Req' object has no attribute 'sampling_params'	validation	error	attribute-access, req, serialization
'{}' object has no attribute '{}'	validation	error	attribute-access, req
{} did not declare component use: {}	validation	error	stage, component-declaration, pipeline
{} verification failed for {}: Failed fields: {}\nDetails: {	validation	error	stage, verification, schema-validation
shot_prompts must be non-empty	validation	error	prompt-validation, causal-denoising
num_blocks must be positive	validation	error	prompt-validation, causal-denoising, block-config
shot_durations must match shot_prompts length	validation	error	prompt-validation, length-mismatch, causal-denoising
causal block prompt count must match causal block count, got	validation	error	causal-denoising, prompt-validation, block-config
realtime_causal_sink_size must be non-negative	validation	error	causal-denoising, kv-cache, config-validation
realtime_causal_kv_cache_num_frames must be positive	validation	error	causal-denoising, kv-cache, config-validation
{} does not support QVG KV-cache quantization	validation	error	kv-cache, quantization, unsupported-feature
num_frames must be divisible by num_frames_per_block for cau	validation	error	causal-denoising, frame-count, divisibility
(num_frames - 1) must be divisible by num_frame_per_block wh	validation	error	causal-denoising, frame-count, divisibility
Expected packed image latents [B, S0, D].	validation	error	ltx-2, image-encoding, latent-shape, packing, torch
LTX-2 conditioning token count mismatch: {packed.shape[1]=}	validation	error	ltx-2, token-count, conditioning, resolution, shape-mismatch
Generator must be provided	validation	error	generator, vae-sampling, determinism, batch-construction
Could not access latents of provided encoder_output	validation	error	vae, encoder-output, attribute-access, diffusers-compat
dynamic_batch_seeds must be a list with one seed per prompt	validation	error	seed, batch-validation, input-validation
seed list length must match num_outputs_per_prompt ({num_vid	validation	error	seed, input-validation, num-outputs
Either `prompt` or `prompt_embeds` must be provided	validation	error	prompt-validation, task-type, input-validation
For classifier-free guidance, either `negative_prompt` or `n	validation	error	cfg, negative-prompt, input-validation
Number of inference steps must be positive, but got {batch.n	validation	error	inference-steps, input-validation, range-check
Guidance scale must be positive, but got {batch.guidance_sca	validation	error	cfg, guidance-scale, input-validation, range-check
Server was launched with --enable-cfg-parallel but this requ	validation	error	cfg, classifier-free-guidance, cfg-parallel, server-args, request-validation
Height and width must be provided	validation	error	latent-preparation, height-width, missing-dimensions, validation
You have passed a list of generators of length {len(generato	validation	error	generator, batch-size, seed, latent-preparation
Cosmos3 accepts either --image-path (I2V) or --video-path (V	validation	error	cosmos3, i2v, v2v, mutually-exclusive, conditioning
Cosmos3 I2V image list is empty	validation	error	cosmos3, i2v, empty-list, image-input
No frames decoded from video: {video_path!r}	validation	error	cosmos3, v2v, video-decode, ffmpeg, corrupt-file
condition_video_keep must be 'first' or 'last', got {keep!r}	validation	error	cosmos3, v2v, condition-video-keep, enum-validation
Cosmos3TokenizationStage requires a tokenizer; expected the	validation	error	cosmos3, tokenizer, qwen2, checkpoint, init-validation
Cosmos3 prompt batch must not be empty	validation	error	cosmos3, tokenization, empty-prompt, empty-batch
Unexpected return type from apply_chat_template: {type(resul	error_code	error	cosmos3, tokenizer, apply-chat-template, transformers, type-mismatch
Cosmos3 batched prompts must tokenize to the same length bec	validation	error	cosmos3, multimodal-gen, batching, tokenization, validation
condition_frame_indexes={cond_indexes} exceeds the latent fr	validation	error	cosmos3, video, frame-index, out-of-range, validation
sound generation was requested (sound_duration > 0) but the	validation	error	cosmos3, audio, checkpoint-capability, validation
action_mode is set but the loaded Cosmos3 checkpoint has no	validation	error	cosmos3, action-generation, checkpoint-capability, validation
domain_id must be non-negative, got {domain_id}	validation	error	cosmos3, action-generation, domain-id, validation
Unknown action domain name {domain_name!r}. Valid names: {so	validation	error	cosmos3, action-generation, domain-name, lookup, validation
Cosmos3 action generation requires --domain-id or --domain-n	validation	error	cosmos3, action-generation, missing-parameter, domain-id
Unsupported action_mode={sp.action_mode!r}; expected one of	validation	error	cosmos3, action-generation, enum, invalid-argument-value
action_mode='forward_dynamics' requires an 'action' array (l	validation	error	cosmos3, action-generation, missing-parameter, forward-dynamics
action must have shape [T, D], got {tuple(action.shape)}	validation	error	cosmos3, action-generation, tensor-shape, validation
action_mode={mode!r} requires --raw-action-dim.	validation	error	cosmos3, action-conditioning, server-args, configuration
raw_action_dim must be in [1, {action_dim}], got {raw_action	validation	error	cosmos3, validation, range-check, action-dim
Cosmos3 rollout supports T2V/T2I only; I2V/V2V conditioned-f	validation	error	cosmos3, rollout, i2v, sde, rl-sampling
Cosmos3 rollout does not support action/sound modalities.	validation	error	cosmos3, rollout, action-latents, sound-latents, not-implemented
Cosmos3 action generation does not support CFG parallel yet	error_code	error	cosmos3, cfg-parallel, action-generation, not-implemented
Cosmos3 action request produced no action tensor	exception	error	cosmos3, action-generation, runtime, internal-error
No raw action dim for Cosmos3 embodiment {embodiment!r}. Exp	validation	error	cosmos3, embodiment, lookup, validation
width and height must be non-zero, got width={width}, height	validation	error	cosmos3, resolution, aspect-ratio, validation
Action normalization stats not found at {stats_path}.	exception	error	cosmos3, file-not-found, normalization-stats, checkpoint
Unknown action normalization method {method!r}.	validation	error	cosmos3, normalization, enum-validation, action-stats
GLM-Image AR returned too few output_ids: got {actual_output	exception	critical	glm-image, autoregressive, token-length, image-generation, runtimeerror
I2I mode is not supported yet via external SGLang encoder UR	validation	error	glm-image, external-encoder, image-to-image, notimplemented
GLM-Image AR batch returned an unexpected response: expected	exception	error	glm-image, external-server, batch-mismatch, response-validation, runtimeerror
Cannot split GLM-Image AR output for sequential inference: e	exception	error	glm-image, sequential-inference, batch-split, shape-mismatch, runtimeerror
`negative_prompt` should be the same type to `prompt`, but g	validation	error	glm-image, negative-prompt, type-mismatch, input-validation, typeerror
`negative_prompt`: {negative_prompt} has batch size {len(neg	validation	error	glm-image, negative-prompt, batch-size, input-validation, valueerror
You have passed a list of generators of length {len(generato	validation	error	glm-image, generator, batch-size, latents, input-validation, valueerror
`callback_on_step_end_tensor_inputs` has to be in {self._cal	validation	error	glm-image, callback, tensor-inputs, input-validation, valueerror
Cannot forward both `prompt`: {prompt} and `prompt_embeds`:	validation	error	glm-image, prompt-embeds, mutual-exclusion, input-validation, valueerror
Provide either `prompt` or `prompt_embeds`. Cannot leave bot	validation	error	glm-image, prompt-embeds, required-argument, input-validation, valueerror
Hunyuan3D requires 'image_path' input.	validation	error	hunyuan3d, input-validation, multimodal, missing-argument
Hunyuan3D only supports a single image input.	validation	error	hunyuan3d, input-validation, single-image-constraint
Hunyuan3D expects image_path as str, got {type(batch.image_p	validation	error	hunyuan3d, type-error, input-validation
Image path not found: {batch.image_path}	validation	error	hunyuan3d, file-not-found, filesystem, docker
Hunyuan3D only supports num_outputs_per_prompt=1.	validation	error	hunyuan3d, input-validation, unsupported-parameter
Timesteps must be provided	validation	error	hunyuan3d, pipeline-order, missing-state, scheduler
Latents must be provided	validation	error	hunyuan3d, pipeline-order, missing-state, latents
Conditioning (prompt_embeds) must be provided	validation	error	hunyuan3d, pipeline-order, missing-state, conditioning
Mesh generation failed: surface extraction returned None. Th	exception	error	hunyuan3d, mesh-extraction, marching-cubes, degenerate-output
Ideogram4DenoisingStage applies its custom scheduler step	exception	error	ideogram, scheduler, unsupported-operation, diffusers
prompt has {num_text_tokens} tokens, exceeds max_text_tokens	validation	error	ideogram, prompt-length, token-limit, validation
height/width must be between 256 and 2048	validation	error	ideogram, image-resolution, validation, out-of-range
height/width must be divisible by patch_size*ae_scale_factor	validation	error	ideogram, image-resolution, divisibility, validation
Unknown Ideogram 4 preset {preset!r}; expected one of {sorte	validation	error	ideogram, preset, invalid-enum, validation
SP DMD renoise requires packed video `batch.raw_latent_shape	validation	error	joyecho, sequence-parallel, dmd, latent-shape, validation
SP DMD renoise requires `batch.sp_audio_orig_num_frames`.	validation	error	joyecho, sequence-parallel, audio, dmd, validation
JoyEcho requires audio latents for denoising.	validation	error	joyecho, audio, missing-latents, denoising
JoyEcho audio scheduler was not prepared.	validation	error	joyecho, audio, scheduler, initialization
Expected scheduler.sigmas to be a tensor for JoyEcho.	validation	error	joyecho, scheduler, sigmas, type-mismatch
latent_window_size must be positive, got {latent_window_size	validation	error	joyecho, memory, window-size, validation
memory_position_mode must be one of {'reference', 'legacy',	validation	error	joy-echo, memory, rope, config-validation, multimodal
Invalid latent grid for memory RoPE: {latent_height=} {laten	validation	error	joy-echo, memory, rope, latent-grid, video
memory_video_len must be a multiple of latent_height * laten	validation	error	joy-echo, memory, rope, alignment, video-tokens
Expected [F, H, W, C] uint8 video, got shape={tuple(video_ui	validation	error	joy-echo, video, tensor-shape, nhwc, pil
Expected RGB video with trailing channel dim 3, got shape={t	validation	error	joy-echo, video, rgb, channels, tensor-shape
Expected batch size 1 for decoded audio, got shape={tuple(wa	validation	error	joy-echo, audio, waveform, batching, memory-slot
Expected decoded audio with 1, 2, or 3 dims, got shape={tupl	validation	error	joy-echo, audio, waveform, tensor-shape
Expected audio_latent shape [B, T, C], got shape={tuple(audi	validation	error	joy-echo, audio, latent, tensor-shape, memory-slot
paired audio memory slot requires audio_latent	validation	error	joy-echo, audio, memory-slot, missing-argument
All memory audio latents must share batch and channel dimens	validation	error	joy-echo, audio, latent, memory-bank, concat
SP-sharded LTX-2 TI2V expected raw seq_len divisible by toke	exception	critical	ltx-2, sequence-parallel, video-generation, shape-validation
Cannot repeat tensor with batch={tensor.shape[0]} to target_	exception	error	ltx-2, batch-dim, cfg-guidance, shape-validation
Unexpected audio latents rank: {audio_latent_model_input.ndi	exception	error	ltx-2, audio-latents, rank-check, video-generation
LTX-2 audio scheduler was not prepared.	exception	critical	ltx-2, audio-scheduler, initialization, pipeline-ordering
LTX-2 requires audio latents for denoising.	exception	critical	ltx-2, audio-latents, required-input
Expected scheduler.sigmas to be a tensor for LTX-2.	exception	error	ltx-2, scheduler, sigmas, type-validation
LTX2 stage-1 CFG parallel degree exceeds guidance pass count	exception	error	ltx-2, cfg-parallel, config-mismatch, distributed
auto_duration was requested but this checkpoint has no durat	exception	error	ltx-2, auto-duration, checkpoint-capability, version-mismatch
keyframe resolved_frame_index values disagree with semantic	exception	error	minimax-h3, keyframes, validation, pipeline
cached keyframe preparation disagrees with the resolved plan	exception	error	minimax-h3, cache-coherence, pipeline, keyframes
fl2va keyframe preparation requires cached pre-queue probe a	exception	error	minimax-h3, fl2va, pipeline-ordering, missing-metadata
noise_aug must be in [0, 1], got {noise_aug}	validation	error	minimax-h3, noise-aug, parameter-validation, range-check
data URI must contain a comma separator	validation	error	minimax-h3, data-uri, material-io, parsing
data URI header is too large	validation	error	minimax-h3, data-uri, header-limit, material-io
data URI must use ;base64 encoding	validation	error	minimax-h3, data-uri, base64, material-io
base64 URI header is too large	validation	error	minimax-h3, base64-uri, header-limit, material-io
material URI has an invalid percent escape	validation	error	minimax-h3, percent-encoding, material-io, parsing
material URI base64 payload must be ASCII	validation	error	minimax-h3, base64, ascii, material-io
material URI has an invalid base64 character {character!r}	validation	error	minimax-h3, base64, alphabet, material-io
unsupported tar material URI	validation	error	minimax-h3, tar-uri, scheme, material-io
tar material URI must contain '<tar_path>:<encoded_header>'	validation	error	minimax-h3, tar-uri, parsing, material-io
tar material URI encoded header is too large	validation	error	minimax-h3, tar-uri, header-limit, material-io
tar material URI has an invalid encoded header	validation	error	minimax-h3, tar-uri, base64, json, material-io
tar material URI header must be a JSON object	validation	error	minimax-h3, tar-uri, json, schema, material-io
unsupported tar material header schema: {header.get('schema'	validation	error	minimax-h3, tar-uri, schema-version, material-io
tar material header requires integer offset_data and size	validation	error	minimax-h3, tar-uri, json-fields, material-io
tar material offset_data and size must be non-negative	validation	error	minimax-h3, tar, material-uri, validation
{label} does not exist or is not a file: {path}	validation	error	minimax-h3, file-not-found, material-localization
{label} is empty: {path}	validation	error	minimax-h3, empty-file, material-validation
MiniMax H3 image material is invalid	validation	error	minimax-h3, image-decode, pil, corrupt-file
MiniMax H3 image material uses an unsupported format	validation	error	minimax-h3, image-format, unsupported-format
MiniMax H3 image material has no positive display geometry	validation	error	minimax-h3, image-geometry, degenerate-image
unsupported MiniMax H3 condition type {condition_type!r}	validation	error	minimax-h3, condition-type, enum-validation
MiniMax H3 media material is invalid	validation	error	minimax-h3, ffprobe, ffmpeg-missing, media-validation
MiniMax H3 media container format is not allowed	validation	error	minimax-h3, container-format, ffmpeg, unsupported-format
MiniMax H3 video material has no video stream	validation	error	minimax-h3, video-stream, ffprobe
MiniMax H3 audio material has no audio stream	validation	error	minimax-h3, audio-stream, ffprobe
MiniMax H3 video material has invalid dimensions	validation	error	minimax-h3, video-dimensions, ffprobe, metadata
MiniMax H3 video material has no positive dimensions	validation	error	minimax-h3, video-dimensions, ffprobe
MiniMax H3 video material has no usable frame rate	validation	error	minimax-h3, frame-rate, ffprobe, video
MiniMax H3 audio material has invalid metadata	validation	error	minimax-h3, audio-metadata, ffprobe
MiniMax H3 audio material has no usable sample rate	validation	error	minimax-h3, sample-rate, ffprobe, audio
MiniMax H3 audio material has no usable channel count	validation	error	minimax-h3, channels, ffprobe, audio
MiniMax H3 media material has no positive duration	validation	error	minimax-h3, duration, ffprobe, media
material URI has an invalid base64 payload	validation	error	minimax-h3, base64, material-uri, urlsafe
material URI has invalid base64 padding	validation	error	minimax-h3, base64, padding, material-uri
material URI has data after base64 padding	validation	error	base64, minimax-h3, material-uri, validation
material URI base64 payload is empty	validation	error	base64, empty-payload, minimax-h3
material URI has an invalid base64 payload length	validation	error	base64, length-validation, minimax-h3
material URI decoded payload is empty	validation	error	base64, empty-decode, minimax-h3
MiniMax H3 base64 decoded size {total} != expected {decoded_	validation	error	base64, internal-consistency, minimax-h3
tar material payload is empty	validation	error	tar, minimax-h3, material-uri
tar material source does not exist or is not a file: {source	validation	error	tar, file-not-found, minimax-h3
tar material payload is truncated: expected {size} bytes, on	validation	error	tar, truncation, minimax-h3
tar material payload is truncated with {remaining} bytes lef	validation	error	tar, truncation, race-condition
tar material reader returned too many bytes	validation	error	tar, io-contract, defensive
HTTP material response.read() must return bytes, got {type(c	validation	error	http, type-error, mocking
HTTP material body is empty: {uri}	validation	error	http, empty-response, minimax-h3
condition URI must be a non-empty string	validation	error	validation, minimax-h3, material-uri
file URI host must be local, got {parsed.netloc!r}	validation	error	file-uri, minimax-h3, material-uri
MiniMax H3 s3:// material URIs require a configured artifact	validation	error	s3, minimax-h3, material-uri
MiniMax H3 material localization does not support URI scheme	validation	error	uri-scheme, minimax-h3, material-uri
MiniMax H3 material localization completed without cached pr	validation	error	internal-state, minimax-h3, probe
keyframe_frame_indices must be omitted when keyframe cond is	validation	error	config-validation, minimax-h3, packed-sequence
strict fl2va packed layout requires keyframe_frame_indices	validation	error	missing-parameter, minimax-h3, packed-sequence
strict fl2va packed layout requires integer keyframe_frame_i	validation	error	type-validation, numpy, minimax-h3
{path}.kind must be a non-empty string	validation	error	validation, schema, multimodal, minimax-h3
{path}.kind unsupported for ref2va: {kind!r}	validation	error	validation, enum, multimodal, minimax-h3
seq_len {seq_len} < used rows {used}	validation	error	validation, sequence-length, alignment, minimax-h3
hybrid ref2va layout only supports first/last keyframe ancho	validation	error	validation, keyframe, video, minimax-h3
{name} must have length {length}, got {list(value)!r}	validation	error	validation, shape, patchify, minimax-h3
{name} values must be positive, got {list(value)!r}	validation	error	validation, shape, positive-check, minimax-h3
{name} must be rank {rank}, got shape={list(tensor.shape)}	validation	error	validation, tensor-shape, rank, minimax-h3
video latent spatial/time dims must be divisible by patch_si	validation	error	validation, tensor-shape, patchify, divisibility, minimax-h3
video token dim {int(rows.shape[-1])} != patch volume * chan	validation	error	validation, tensor-shape, unpatchify, minimax-h3
video rows {int(rows.shape[0])} must be divisible by thw {	validation	error	validation, tensor-shape, unpatchify, row-count, minimax-h3
{context}video block token counts and timestamps must align	validation	error	minimax-h3, ref2va, video-presentation, validation
{context}video block token count must be positive	validation	error	minimax-h3, video-presentation, validation
prompt must be non-empty	validation	error	minimax-h3, text-encoding, validation
image_token_counts must be non-empty	validation	error	minimax-h3, multi-image, validation
image_token_count must be positive	validation	error	minimax-h3, multi-image, validation
{name} must be an int or a sequence of ints	validation	error	minimax-h3, type-validation, ref2va
{name} must be a sequence	validation	error	minimax-h3, type-validation, video-presentation
{name} must not mix nested and flat entries	validation	error	minimax-h3, shape-validation, video-presentation
video block token counts and timestamps must align	validation	error	minimax-h3, ref2va, alignment
image_token_count required for an image reference	validation	error	minimax-h3, ref2va, alignment
video reference requires block token counts and timestamps	validation	error	minimax-h3, ref2va, video-presentation, alignment
unsupported ref2va condition type {cond_type!r}	validation	error	minimax-h3, ref2va, enum-validation
unused image_token_count entries	validation	error	minimax-h3, ref2va, alignment
unused video block token count entries	validation	error	minimax-h3, ref2va, alignment
reference image width and height must be positive finite num	validation	error	minimax-h3, image-shape, validation
reference image ratio must be within the inclusive range 1:4	validation	error	minimax-h3, aspect-ratio, image-validation
reference image target dimensions must be positive	validation	error	minimax-h3, resize, argument-validation
reference image target dimensions must be aligned to {MINIMA	validation	error	minimax-h3, alignment, resize
reference audio duration bound must be positive	validation	error	minimax-h3, audio, duration-validation
reference audio start time must be non-negative	validation	error	minimax-h3, audio, timestamp-validation
reference audio sample rate must be positive	validation	error	minimax-h3, audio, sample-rate
unsupported MiniMax H3 audio material chain {material_chain!	validation	error	minimax-h3, material-chain, unsupported-operation
reference audio is empty: {audio_path}	validation	error	minimax-h3, audio, empty-media
reference video has no frames: {video_path}	validation	error	minimax-h3, video, empty-media, ffmpeg
ref2va requires at least one image reference	validation	error	minimax-h3, ref2va, missing-input
ref2va video preparation requires a video or video_audio ref	validation	error	minimax-h3, ref2va, missing-input, not-implemented
task {task!r} is not served by MiniMax H3 partition {self.pa	validation	critical	minimax-h3, model-index, config-validation
task {task!r} resolves outside partition {self.partition!r}	validation	error	minimax-h3, task-routing, partition
{path} must be a non-empty list	validation	critical	minimax-h3, model-index, config-validation
{path} must contain non-empty strings	validation	critical	minimax-h3, model-index, config-validation
{path} must not contain duplicates	validation	critical	minimax-h3, model-index, config-validation, duplicates
model_index.json._minimax_h3 must be an object	validation	critical	minimax-h3, model-index, config-validation
model_index.json._minimax_h3.schema_version must be 1	validation	critical	minimax-h3, model-index, schema-version
model_index.json._minimax_h3.partition must be one of fl2va,	validation	critical	minimax-h3, model-index, partition, config-validation
model_index.json._minimax_h3.task_aliases must map strings t	validation	error	minimax-h3, model-index, task-aliases, config-validation
model_index.json._minimax_h3.sigma_shift_scales must be an o	validation	error	minimax-h3, sigma-shift-scales, model-index, config-validation
model_index.json._minimax_h3.sigma_shift_scales requires num	validation	error	minimax-h3, sigma-shift-scales, numeric-coercion, model-index
tasks must contain canonical task names, got {task!r}	validation	error	minimax-h3, task-names, canonicalization, model-index
task {task!r} does not belong to partition {partition!r}	validation	error	minimax-h3, partition, task-routing, model-index
task alias {alias!r} targets undeclared task {target!r}	validation	error	minimax-h3, task-aliases, dangling-reference, model-index
unsupported task alias mapping {alias!r} -> {target!r}	validation	error	minimax-h3, task-aliases, canonicalization, model-index
MiniMax H3 request task must be a non-empty string	validation	error	minimax-h3, task-required, request-validation, sampling-params
MiniMax H3 requires num_inference_steps >= 2 because its vid	validation	error	minimax-h3, num-inference-steps, request-validation, sigma-schedule
quality must be one of {list(QUALITY_LEVELS)}, got {quality!	validation	error	minimax-h3, quality-level, request-validation, sampling-params
MiniMax-H3 quality="high" requires a resolved request plan	validation	error	minimax-h3, quality-high, request-plan, pipeline-order
MiniMax-H3 quality="high" is validated only for {_MINIMAX_H3	validation	error	minimax-h3, quality-high, workload-validation, golden-config
{path} must be a non-empty string	validation	error	minimax-h3, request-validation, string-field, canonical-request
{path} must be an integer	validation	error	minimax-h3, request-validation, integer-field, type-check
{path} must be a number	validation	error	minimax-h3, request-validation, float-field, type-check
{path} must be a positive finite number	validation	error	minimax-h3, request-validation, positive-float, nan-inf
{path} must be a non-negative finite number	validation	error	minimax-h3, request-validation, non-negative-float, nan-inf
{path} is required and must be an object	validation	error	minimax-h3, request-validation, target-object, canonical-request
{path}.short_edge must be positive, got {short_edge}	validation	error	minimax-h3, validation, request, short-edge
{path}.aspect_ratio must be "auto" for task {profile.task!r}	validation	error	minimax-h3, aspect-ratio, validation
{path}.aspect_ratio for task {profile.task!r} must be 'auto'	validation	error	minimax-h3, aspect-ratio, enum-validation
{path}.duration_seconds is required	validation	error	minimax-h3, duration, required-field
{path}.duration_seconds must be a number	validation	error	minimax-h3, type-mismatch, duration
{path}.duration_seconds must be positive	validation	error	minimax-h3, duration, positive-value
{path}.duration_seconds must be in [{MINIMAX_H3_MIN_DURATION	validation	error	minimax-h3, duration, range-check
{path} must be a list	validation	error	minimax-h3, conditions, type-mismatch
{path} must be empty for task {profile.task!r} (got {len(con	validation	error	minimax-h3, conditions, task-profile
{path} requires at least one entry for task {profile.task!r}	validation	error	minimax-h3, conditions, required-field
{path} requires at least {profile.min_condition_count} entri	validation	error	minimax-h3, conditions, count-constraint
{path} allows at most {profile.max_condition_count} entries	validation	error	minimax-h3, conditions, count-constraint
{cpath} must be an object	validation	error	minimax-h3, conditions, type-mismatch
{cpath} has unknown fields: {sorted(unknown)}	validation	error	minimax-h3, conditions, unknown-fields, schema
{cpath}.role must be keyframe or reference, got {role!r}	validation	error	minimax-h3, conditions, role, enum-validation
{cpath}: {exc}	validation	error	minimax-h3, conditions, role-type-mapping
{cpath}.frame_index requires a resolved target duration	validation	error	minimax-h3, frame-index, duration, dependency-order
{cpath}.frame_index must be -1 or in [0, {aligned_frame_coun	validation	error	minimax-h3, frame-index, bounds-check, frame-alignment
{cpath}.start_time_seconds is only allowed for video or vide	validation	error	minimax-h3, request-validation, video, conditions
conditions for task {task!r} must include one or two ordered	validation	error	minimax-h3, keyframes, fl2va, ref2va, request-validation
ref2va keyframes require at least one reference condition; u	validation	error	minimax-h3, ref2va, task-validation, conditions
conditions[{index}]: video references are not supported in v	validation	error	minimax-h3, video-input, unsupported-type, request-validation
target.duration_seconds is required, or exactly one audio re	validation	error	minimax-h3, duration, target, request-validation
target.duration_seconds is required when multiple audio-bear	validation	error	minimax-h3, duration, ambiguous-source, request-validation
seed must be non-negative, got {normalized_seed}	validation	error	minimax-h3, seed, validation
seed must not exceed the signed int64 maximum, got {normaliz	validation	error	minimax-h3, seed, int64-overflow, validation
target.aspect_ratio must be 'W:H' or 'auto', got {value!r}	validation	error	minimax-h3, aspect-ratio, parsing, resolved-plan
target.aspect_ratio must be integer 'W:H', got {value!r}	validation	error	minimax-h3, aspect-ratio, integer-parse, resolved-plan
target.aspect_ratio components must be positive, got {value!	validation	error	minimax-h3, aspect-ratio, positive-validation, resolved-plan
target.short_edge must be an integer, got {value!r}	validation	error	minimax-h3, short-edge, type-error, spatial
target.short_edge must be a positive integer, got {value!r}	validation	error	minimax-h3, short-edge, integer-validation, spatial
shape width and height must be positive finite numbers	validation	error	minimax-h3, spatial, type-error, dimensions
shape ratio must be a positive finite number	validation	error	minimax-h3, spatial, ratio, floating-point
adapt_shape_v1 ratio must be within the inclusive range 1:4	validation	error	minimax-h3, aspect-ratio, range-check, spatial
canonical request must be a mapping	validation	error	minimax-h3, type-error, mapping, resolved-plan
canonical request has unknown fields: {sorted(unknown)}	validation	error	minimax-h3, unknown-fields, schema-drift, resolved-plan
canonical request missing {key!r}	validation	error	minimax-h3, schema-validation, required-field
{profile.task} ResolvedPlan requires one or two ordered imag	validation	error	minimax-h3, keyframe, frame-index, plan-validation
conditions[{index}].frame_index is required	validation	error	minimax-h3, frame-index, required-field
conditions[{index}].frame_index must be -1 or in [0, {frame_	validation	error	minimax-h3, frame-index, out-of-range
conditions[{index}].frame_index resolves to {resolved_frame_	validation	error	minimax-h3, frame-index, duplicate
MiniMaxH3AudioEncodingStage direct audio tokenizer encode re	exception	error	minimax-h3, audio-encoding, legacy-api, not-implemented
MiniMax H3 audio encode failed on rank 0: {owner_error}	exception	critical	minimax-h3, audio-encoding, distributed, rank0
unsupported MiniMax H3 decoder task {task_value!r}	validation	error	minimax-h3, task, enum-value-invalid
MiniMax H3 tasks require the video_vae output decoder	error_code	critical	minimax-h3, vae, missing-module, decoding
video_vae became unavailable during decode	error_code	critical	minimax-h3, vae, race-condition, component-registry
MiniMax H3 audio decode failed on rank 0: {owner_error}	error_code	critical	minimax-h3, audio-decode, distributed, rank0
MiniMax H3 audio decode produced no output payload	error_code	critical	minimax-h3, audio-decode, payload-format
keyframe condition rows require plan.task='fl2va' or 'ref2va	validation	error	minimax-h3, keyframe, task-mismatch
fl2va denoising requires encoded keyframe condition rows	validation	error	minimax-h3, fl2va, keyframe, missing-input
encoded keyframe condition rows must be a mapping	validation	error	minimax-h3, keyframe, type-validation
keyframe denoising requires semantic_frame_indices in {MINIM	validation	error	minimax-h3, keyframe, signature-validation
keyframe payload requires an integer frame_count	validation	error	minimax-h3, keyframe, type-validation, frame-count
keyframe payload frame_count must be greater than one	validation	error	minimax-h3, keyframe, frame-count, boundary
keyframe denoising requires pixel_frame_indices resolved fro	validation	error	minimax-h3, keyframe, index-consistency
keyframe denoising requires one encoded keyframe per semanti	validation	error	minimax-h3, keyframe, cardinality, type-validation
MiniMax H3 denoise state must be a mapping	validation	error	minimax-h3, pipeline, batch-state, validation
MiniMax H3 initial_video_rows must be a rank-2 tensor	validation	error	minimax-h3, tensor-shape, batch-state
MiniMax H3 initial_audio_rows must be a rank-2 tensor	validation	error	minimax-h3, tensor-shape, batch-state
MiniMax H3 latent preparation requires pre-queue resolved_v2	validation	error	minimax-h3, geometry, plan-validation, pipeline
MiniMax H3 latent preparation requires pre-queue resolved te	validation	error	minimax-h3, temporal-dimensions, plan-validation
aligned video noise shape {list(video_noise.shape)} != [{vid	validation	critical	minimax-h3, noise-generation, internal-invariant
replica broadcast of batch.extra[{key!r}] got None	error_code	error	minimax-h3, replica-broadcast, distributed, distributed-communication
MiniMaxH3TextEncodingStage requires the pipeline processor c	validation	error	minimax-h3, pipeline-components, model-index, init-validation
MiniMaxH3TextEncodingStage direct Qwen3VL encoder forward re	validation	error	minimax-h3, legacy-api, not-implemented, request-format
MiniMax H3 text encode produced no native payload	validation	error	minimax-h3, text-encoding, payload-validation, data-parallel
MiniMax H3 text encode failed on rank {owner}: {owner_error}	error_code	error	minimax-h3, data-parallel, error-propagation, text-encoding
MiniMax H3 text payload broadcast failed	error_code	critical	minimax-h3, broadcast, tensor-dict, collective, distributed-communication
MiniMax H3 text payload must contain positive.hidden_states	validation	error	minimax-h3, text-encoding, tensor-shape, payload-validation
MiniMax H3 text payload positive.text_len must match the hid	validation	error	minimax-h3, text-encoding, length-mismatch
MiniMax H3 text encoding requires an ordered keyframe signat	validation	error	minimax-h3, keyframes, plan-validation, fl2va
task {plan.task!r} cannot carry image.target_canvas material	validation	error	minimax-h3, task-validation, materials, plan-validation
MiniMaxH3TextEncodingStage direct encode requires a text_enc	validation	error	minimax-h3, missing-component, text-encoding
MiniMax H3 text_encoder component must expose callable encod	validation	error	minimax-h3, duck-typing, text-encoding, type-validation
MiniMaxH3TextEncodingStage direct encode requires a tokenize	validation	error	minimax-h3, missing-component, tokenizer
fl2va Qwen preparation requires one or two ordered images wi	validation	error	minimax-h3, fl2va, keyframes, validation
MiniMaxH3VisualEncodingStage cannot encode material chains {	validation	error	minimax-h3, visual-encoding, unsupported-feature, validation
keyframe visual preparation requires one or two ordered imag	validation	error	minimax-h3, keyframes, fl2va, payload-validation
prepared reference videos payload must carry a non-empty 'vi	validation	error	minimax-h3, ref2va, reference-video, payload-validation
MiniMax H3 task must be a non-empty string	validation	error	minimax-h3, task-validation, invalid-argument
unsupported MiniMax H3 task {task!r}	validation	error	minimax-h3, task-validation, unknown-task
task {self.task!r} does not allow condition role={role!r} ty	validation	error	minimax-h3, condition-rules, task-profile, validation
unknown minimax_h3 task {task!r}; supported: {sorted(MINIMAX	validation	error	minimax-h3, task-profile, unknown-task, registry-lookup
MiniMax H3 shift_scale must be > 0	validation	error	minimax-h3, sampling, shift-scale, invalid-argument
MiniMax H3 num_steps must be > 0	validation	error	minimax-h3, sampling, num-steps, invalid-argument
task is required for MiniMax H3; supported tasks: fl2va, ref	validation	error	minimax-h3, task-validation, missing-parameter, video-adapter
MiniMax H3 SGLang backend only supports output_mode='decoded	validation	error	minimax-h3, video-generation, output-mode, validation
MiniMax H3 does not support enable_frame_interpolation: the	validation	error	minimax-h3, frame-interpolation, video-generation, validation
MiniMax H3 does not support enable_upscaling: the accepted d	validation	error	minimax-h3, upscaling, video-generation, validation
MiniMax H3 DiffGenerator requires save_output=True and a non	validation	error	minimax-h3, sampling-params, save-output, validation
queued MiniMax H3 jobs require pre-queue resolved_v2 geometr	validation	critical	minimax-h3, queue-invariant, geometry, internal
queued MiniMax H3 jobs require pre-queue resolved temporal d	validation	critical	minimax-h3, queue-invariant, frame-count, internal
MiniMax H3 video generation produced {len(output_paths)} out	exception	error	minimax-h3, output-count, validation, video-generation
generated MiniMax H3 outputs have inconsistent media metadat	exception	error	minimax-h3, metadata-consistency, multi-output, validation
MiniMax H3 final output ffprobe timed out after 30 seconds	exception	error	minimax-h3, ffprobe, timeout, subprocess
ffprobe is required to validate final MiniMax H3 output	exception	error	minimax-h3, ffprobe, missing-binary, environment
ffprobe failed for final MiniMax H3 output{suffix}	exception	error	minimax-h3, ffprobe, corrupt-output, validation
ffprobe returned invalid JSON for MiniMax H3 output	exception	error	minimax-h3, ffprobe, json-parse, validation
ffprobe returned invalid stream metadata	exception	error	minimax-h3, ffprobe, stream-metadata, validation
generated MiniMax H3 MP4 must contain exactly one video stre	exception	error	minimax-h3, mp4, stream-layout, validation
generated MiniMax H3 MP4 has invalid size {width}x{height}	exception	error	minimax-h3, mp4, dimensions, validation
generated MiniMax H3 MP4 size does not match the resolved re	exception	error	minimax-h3, resolution, geometry-mismatch, validation
generated MiniMax H3 MP4 frame rate must be {MINIMAX_H3_SUPP	exception	error	minimax-h3, video, fps, mp4, validation
You have passed a list of generators of length {len(generato	validation	error	mova, generator, batch-size, diffusers, validation
MOVA requires reference image latents for denoising	validation	error	mova, image-latent, conditioning, validation
Pi05 v1 expects one prompt per action request	validation	error	pi05, vla, prompt-validation, batch-size
Pi05 v1 expects one state vector per request	validation	error	pi05, vla, state-input, batch-size, torch
Pi05 state dim must be <= {self.config.state_dim}, got {stat	validation	error	pi05, vla, state-dim, config-mismatch, torch
Pi05 noise must have shape {expected}, got {tuple(noise_tens	validation	error	pi05, vla, noise, shape-validation, flow-matching, torch
Qwen-Image-Layered requires a non-empty image_path.	validation	error	qwen-image, layered-editing, image-path, validation
Only one of `timesteps` or `sigmas` can be passed. Please ch	validation	error	qwen-image, diffusers, scheduler, timesteps, sigmas, mutually-exclusive-args
The current scheduler class {scheduler.__class__}'s `set_tim	validation	error	qwen-image, diffusers, scheduler, timesteps, scheduler-unsupported
The current scheduler class {scheduler.__class__}'s `set_tim	validation	error	qwen-image, diffusers, scheduler, sigmas, scheduler-unsupported
Cannot duplicate `image` of batch size {image_latents.shape[	validation	error	qwen-image, diffusers, batch-size, latents, image-editing
You have passed a list of generators of length {len(generato	validation	error	qwen-image, diffusers, generator, reproducibility, batch-size
{error_label} must be list[list[str]]	validation	error	validation, camera-actions, sana-wm
action string is empty	validation	error	validation, action-string, sana-wm
invalid action segment {segment!r}; expected '<keys>-<frames	validation	error	validation, action-string, sana-wm
invalid duration in action segment {segment!r}	validation	error	validation, action-string, duration, sana-wm
unknown action keys {bad}; allowed keys are {sorted(_SANA_WM	validation	error	validation, action-string, whitelist, sana-wm
camera trajectory must have shape (F, 4, 4); got {c2w.shape}	validation	error	numpy, camera-trajectory, shape-validation, sana-wm
unsupported intrinsics shape {arr.shape}; expected (4,), (3,	validation	error	numpy, intrinsics, shape-validation, sana-wm
SANA-WM stage-1 expects exactly one Gemma-2 text encoder.	validation	error	model-config, text-encoder, sana-wm
SANA-WM denoising requires initialized latents.	validation	error	pipeline-order, latents, sana-wm
SANA-WM denoising expects 5D latents shaped (B, C, T, H, W),	validation	error	latents, shape-validation, sana-wm
SANA-WM denoising requires prepared timesteps.	validation	error	scheduler, timesteps, pipeline-order, sana-wm
SANA-WM denoising requires positive prompt embeds.	validation	error	prompt-embeds, pipeline-order, sana-wm
SANA-WM CFG requires negative prompt embeds.	validation	error	cfg, negative-prompt, sana-wm
Unsupported VAE encode output for SANA-WM first-frame condit	validation	error	vae, encode-output, compatibility, sana-wm
SANA-WM generator list must not be empty.	validation	error	generator, noise, validation, sana-wm
SANA-WM generator list length must match latent batch size;	validation	error	generator, batch-size, noise, sana-wm
SANA-WM seed list must not be empty.	validation	error	seed, generator, validation, sana-wm
SANA-WM seed list length must be 1 or match latent batch siz	validation	error	seed, batch-size, generator, sana-wm
condition_image tensor must be CHW or HWC with 1, 3, or 4 ch	validation	error	image-preprocessing, channels, shape-validation, sana-wm
camera_conditions must have shape (T,20) or (B,T,20), got {t	validation	error	sglang, sana-wm, camera-conditions, shape-validation, video-generation
camera_conditions batch dimension must be 1 or match request	validation	error	sglang, sana-wm, camera-conditions, batch-size-mismatch
camera_conditions must have last dimension 20, got {tuple(ca	validation	error	sglang, sana-wm, camera-conditions, feature-dimension
Prepacked latent-frame camera_conditions require chunk_pluck	validation	error	sglang, sana-wm, chunk-plucker, prepacked-conditions, world-model
SANA-WM action and camera_to_world/camera_path are mutually	validation	error	sglang, sana-wm, action-conditioning, mutually-exclusive-args, camera-control
chunk_plucker must have shape (48,T,H,W) or (B,48,T,H,W), go	validation	error	sglang, sana-wm, chunk-plucker, tensor-rank, shape-validation
chunk_plucker batch dimension must be 1 or match request bat	validation	error	sglang, sana-wm, chunk-plucker, batch-size-mismatch
chunk_plucker shape mismatch for SANA-WM: expected {expected	validation	error	sglang, sana-wm, chunk-plucker, latent-resolution, shape-mismatch
SANA-WM first-frame conditioning failed; refusing to continu	exception	critical	sglang, sana-wm, first-frame-conditioning, fail-fast, runtime-wrapper
SANA-WM is a TI2V world model and requires condition_image f	validation	error	sglang, sana-wm, condition-image, missing-input, ti2v
SANA-WM refiner requires a string prompt or one prompt per b	validation	error	sana-wm, refiner, prompt-validation, batch-mismatch, valueerror
SANA-WM refiner text encoder must return per-layer hidden_st	exception	error	sana-wm, text-encoder, hidden-states, runtimeerror, model-output
Stage-1 latent has {z.shape[2]} frames but sink_size={sink_s	validation	error	sana-wm, refiner, latent-shape, sink-frame, valueerror
SANA-WM refiner requires batch.latents from stage 1.	validation	error	sana-wm, refiner, missing-latents, pipeline-order, valueerror
SANA-WM refiner expects 5D latents shaped (B, C, T, H, W), g	validation	error	sana-wm, refiner, latent-shape, ndim, valueerror
SANA-WM refiner decoding expects decoded video shaped (B, C,	validation	error	sana-wm, refiner, vae, decode-shape, valueerror
SANA-WM refiner decoding expected a sink frame plus refined	validation	error	sana-wm, refiner, decode, temporal-length, valueerror
denoising_step_list must end with 0, got {schedule}	validation	error	sana-wm, self-forcing, sigma-schedule, denoising-steps, valueerror
SANA-WM realtime denoising expects this tick's pre-noised ch	validation	error	sana-wm, streaming, realtime, latents, valueerror
SANA-WM realtime denoising requires a realtime session	validation	error	sana-wm, streaming, session, realtime, valueerror
chunk plan {plan} does not cover the incoming {incoming.shap	validation	error	sana-wm, streaming, chunk-plan, frame-count, valueerror
SANA-WM streaming does not support CFG parallel; run replica	exception	error	sana-wm, streaming, cfg-parallel, notimplementederror, server-args
SANA-WM streaming requires positive prompt embeds.	validation	error	sana-wm, streaming, prompt-embeds, conditioning, valueerror
SANA-WM streaming CFG requires negative prompt embeds.	validation	error	sana-wm, streaming, cfg, negative-prompt, valueerror
SANA-WM streaming denoising expects 5D latents (B, C, T, H,	validation	error	sana-wm, streaming, offline, latent-shape, valueerror
streaming needs >= {num_frame_per_block} latent frames, got	validation	error	sana-wm, streaming, frame-count, block-size, valueerror
SANA-WM streaming decode requires AutoencoderKLCausalLTX2Vid	validation	error	sana-wm, streaming, vae, component-paths, ltx2
unsupported capture mode: {mode}	validation	error	sana-wm, streaming-refiner, kv-capture, mode-string, valueerror
missing captured KV on {attr}	exception	error	sana-wm, streaming-refiner, kv-capture, runtimeerror, state-lifecycle
Unsupported LTX-2 RoPE type: {attn.rope_type}	validation	error	sana-wm, ltx2, rope, attention-config, valueerror
Module(s) requested for update not found in pipeline: {unkno	validation	error	weights-update, module-name, validation, multimodal
Missing tensor payload for module(s): {missing}. Provided mo	validation	error	weights-update, payload, validation, multimodal
Ambiguous tensor payload for multi-module update. Provide a	validation	error	weights-update, payload, ambiguous, multimodal
flattened_bucket payload must be a dict with 'flattened_tens	validation	error	weights-update, flattened-bucket, validation, multimodal
flattened_bucket payload missing 'flattened_tensor' or 'meta	validation	error	weights-update, flattened-bucket, missing-key, multimodal
Unsupported module payload type for load_format={load_format	validation	error	weights-update, payload-type, validation, multimodal
flattened_bucket 'flattened_tensor' must be a torch.Tensor.	validation	error	weights-update, flattened-bucket, dtype, multimodal
flattened_bucket 'metadata' must be a list.	validation	error	weights-update, flattened-bucket, metadata, multimodal
Unsupported dtype in flattened_bucket metadata: {dtype!r}	validation	error	weights-update, dtype, flattened-bucket, multimodal
Unsupported activation type: {act_type}	validation	error	realesrgan, activation, config, postprocess
Unsupported RRDBNet conv_first input channels: {in_channels}	validation	error	realesrgan, checkpoint, architecture, postprocess
All frames in a batch must have the same resolution	validation	error	realesrgan, batch, resolution, postprocess
Failed to load Real-ESRGAN checkpoint from '{resolved_path}'	exception	error	realesrgan, checkpoint, torch-load, postprocess
Real-ESRGAN weight file '{resolved_path}' is not compatible	exception	error	realesrgan, checkpoint, architecture, postprocess
RealESRGAN batch upscale did not produce all frames	exception	error	realesrgan, batch, invariant, postprocess
huggingface_hub is required to download Real-ESRGAN weights.	exception	error	realesrgan, huggingface, dependency, postprocess
Failed to download Real-ESRGAN weights from HuggingFace repo	exception	error	realesrgan, huggingface, download, postprocess
RIFE weight file not found: {flownet_path} Expected layout:	exception	error	rife, weights, file-not-found, postprocess
{kind} state payload requires transitions	validation	error	realtime, control-events, schema, validation
{kind} transition must be a map	validation	error	realtime, control-events, schema, validation
{kind} transition actions must be a list	validation	error	control-events, schema-validation, realtime
control signal kind {item.kind!r} does not match queue kind	validation	error	control-signals, queue, type-mismatch
Invalid backend: {value}. Must be one of: {', '.join([m.valu	validation	error	config, enum-validation, backend-selection
lora_alpha must be a positive integer	validation	error	lora, config-validation, server-args
scheduler_rpc_timeout must be None or an integer between 1 a	validation	error	timeout, rpc, scheduler, config-validation
--bcg-text-buckets must contain at least one positive intege	validation	error	cuda-graph, buckets, config-validation
Invalid ltx2_two_stage_device_mode={mode!r}. Expected one of	validation	error	ltx2, device-placement, config-validation
ltx2_two_stage_device_mode=resident conflicts with explicit	validation	error	ltx2, config-conflict, device-placement
Ring Attention requires one of the ring-capable backends ({'	validation	error	ring-attention, attention-backend, distributed, config-validation
Attention backend name must be a string	validation	error	attention-backend, type-validation, config
Invalid attention backend '{backend}'. Available options are	validation	error	attention-backend, enum-validation, config
{option} must be a dict or a comma-separated component=value	validation	error	config-parsing, component-map, type-validation
{option} must use component=value entries	validation	error	config, cli, validation, layerwise-offload
unknown residency policy {policy!r} for component {component	validation	error	config, validation, layerwise-offload, residency
component_attention_backends must be a dict or a comma-separ	validation	error	config, type-error, attention-backend
component_attention_backends must use component=backend entr	validation	error	config, cli, validation, attention-backend
Component attention backend key must be a string	validation	error	config, type-error, attention-backend
Component attention backend key must not be empty	validation	error	config, validation, attention-backend
Invalid --warmup-mode {self.warmup_mode!r}; expected one of	validation	error	config, cli, warmup, validation
--warmup-num-frames must be a positive integer.	validation	error	config, cli, warmup, validation
{name} port {port} is unavailable and --strict-ports is enab	exception	critical	network, ports, startup, config
{name} port {port} duplicates {seen_ports[port]} port and --	exception	critical	network, ports, startup, config
kv_gather_degree does not compose with ulysses_degree or rin	validation	error	parallelism, config, validation
MPS currently supports only --num-gpus 1	validation	error	platform, mps, gpu, config
MPS supports only resident or layerwise-offload component re	validation	error	platform, mps, residency, config
{feature_name} requires {component_name!r} to be resident; g	validation	error	config, residency, feature-conflict
Could not parse attention backend config: {config_str}	validation	error	config, attention-backend, parsing
Failed to find available port after {max_attempts} attempts	exception	error	network, port-allocation, server-startup, configuration
error: unrecognized arguments: {' '.join(remaining)}	console	error	cli, argparse, argument-validation, startup
kv_cache_quant_config must be QVGKVQuantArgs or a dict	validation	error	configuration, type-validation, quantization, kv-cache
Removed server argument(s): {replacements}	validation	error	api-migration, configuration, breaking-change, argument-validation
--gpu-ids contains a non-integer GPU id: {token}	validation	error	gpu, cli-arguments, validation
--gpu-ids GPU ids must be non-negative: {gpu_id}	validation	error	gpu, cli-arguments, validation
--gpu-ids contains duplicate GPU ids: {parsed}	validation	error	gpu, cli-arguments, validation, duplicates
{field_name} is required	validation	error	zmq, endpoint, config, validation
{field_name} must be formatted as tcp://host:port or host:po	validation	error	zmq, endpoint, validation
{field_name} must include both host and port: {value!r}	validation	error	zmq, endpoint, validation
{field_name} port must be an integer: {port_str}	validation	error	zmq, endpoint, port, validation
{field_name} port must be between 0 and 65535: {port}	validation	error	zmq, endpoint, port, validation
invalid IPv6 address format: missing ']'	validation	error	ipv6, distributed, network, validation
invalid IPv6 address: {host}	validation	error	ipv6, distributed, network, validation
received IPv6 address format: expected ':' after ']'	validation	error	ipv6, distributed, network, validation
a port must be specified in IPv6 address (format: [ipv6]:por	validation	error	ipv6, distributed, network, port, validation
invalid port in IPv6 address: '{port_str}'	validation	error	ipv6, distributed, port, validation
Unsupported socket type: {socket_type}	validation	error	zmq, socket, validation
Multi-output conditioning requires prompt text so the prompt	validation	error	batching, sampling, multimodal, validation
Multi-output conditioning requires at least one prompt.	validation	error	batching, sampling, validation, empty-input
{name} has batch dim {current_batch_size} (shape {tuple(valu	validation	error	batching, tensor-shape, validation
{name} must be a tensor, list of tensors, or None.	validation	error	batching, type-error, validation
{name} entries must be tensors or None.	validation	error	batching, type-error, validation
{name}[{index}] has {len(sequence_lengths)} entries; expecte	validation	error	batch-validation, conditioning, multimodal
{field_name} must be a tensor, list of tensors, list of sequ	validation	error	type-validation, conditioning, multimodal
The size of ({name}) is ({self.name_to_size[name]}), but you	exception	critical	distributed, parallelism, config
GGUF models are not supported.	validation	error	model-format, gguf, unsupported
Failed to load diffusers config from {file_path}: {e}	exception	error	config, json, corrupt-cache
Specified lora_weight_name '{weight_name}' not found in {loc	validation	error	lora, file-not-found, local-model
Native diffusion LoRA loading requires a safetensors file, g	validation	error	lora, safetensors, unsupported-format
Resolved LoRA weight {selected_file!r} was not downloaded to	validation	error	lora, download, cache
Model directory {model_path} does not contain model_index.js	validation	error	model-format, diffusers, validation
model_index.json does not contain _diffusers_version	validation	error	diffusers, config-validation
Model directory {model_path} is missing required component d	validation	error	diffusers, incomplete-download, model-files
Model directory {model_path} does not contain a transformer/	validation	error	diffusers, model-files, directory-layout
model_index.json for {model_name_or_path} does not contain _	validation	error	diffusers, config-validation
Failed to find config.json for {model_name_or_path} after fa	validation	error	model-repo, diffusers, config-not-found
{exc}	exception	error	file-not-found, huggingface, modelscope
No cached files for {repo_id} match {allow_patterns or '*/	exception	error	offline-cache, snapshot-download, huggingface, modelscope
Failed to decode base64 image. Expected format: `data:[<medi	validation	error	base64, data-uri, image-input, validation
{b64_format_hint} (missing ;base64 marker)	validation	error	base64, data-uri, image-input, validation
{b64_format_hint} (empty data payload)	validation	error	base64, data-uri, empty-payload, validation
Failed to decode base64 image: {str(exc)}	exception	error	base64, decode-error, image-input
Image is fully transparent	validation	error	image-processing, alpha-channel, rgba, mesh3d, validation
input image is empty	validation	error	image-processing, mask, segmentation, mesh3d, validation
Comfy layer {prefix!r} is missing checkpoint tensors: {sorte	validation	error	quantization, checkpoint, safetensors, validation
Comfy W4A8 layer {prefix!r} has invalid group_size={group_si	validation	error	quantization, w4a8, group-size, validation
Comfy W4A8 layer {prefix!r} needs I8 weights and FP8 group s	validation	error	quantization, dtype-mismatch, w4a8, safetensors
Comfy W4A8 layer {prefix!r} needs F32 channel scales, got {c	validation	error	quantization, dtype-mismatch, w4a8, channel-scales
Comfy W4A8 layer {prefix!r} needs a 2D packed weight, got {w	validation	error	quantization, shape-mismatch, w4a8, safetensors
Comfy W4A8 layer {prefix!r} has incompatible weight/scale sh	validation	error	quantization, shape-mismatch, w4a8, group-size
Comfy W4A8 layer {prefix!r} needs an F32[16] codebook	validation	error	quantization, codebook, dtype-mismatch, w4a8
Comfy W4A8 layer {prefix!r} has an incompatible correction t	validation	error	quantization, shape-mismatch, w4a8, correction-tensor
Comfy W4A4 layer {prefix!r} needs I8 packed weights and F32	validation	error	quantization, dtype-mismatch, w4a4, safetensors
Comfy W4A4 layer {prefix!r} has incompatible weight/scale sh	validation	error	quantization, shape-mismatch, w4a4, scales
Comfy W4A4 layer {prefix!r} has unsupported convrot_groupsiz	validation	error	quantization, comfy, w4a4, checkpoint-validation, safetensors
Comfy W4A4 layer {prefix!r} has input size {logical_input_si	validation	error	quantization, comfy, w4a4, shape-mismatch, checkpoint-validation
Comfy NVFP4 layer {prefix!r} needs U8 packed weights and FP8	validation	error	quantization, nvfp4, fp8, dtype-mismatch, checkpoint-validation
Comfy NVFP4 layer {prefix!r} needs a scalar F32 weight_scale	validation	error	quantization, nvfp4, scalar-scale, dtype-mismatch, checkpoint-validation
Comfy NVFP4 layer {prefix!r} needs a 2D packed weight, got {	validation	error	quantization, nvfp4, rank-mismatch, conv-weights, checkpoint-validation
Comfy NVFP4 layer {prefix!r} has incompatible weight/scale s	validation	error	quantization, nvfp4, shape-mismatch, block-scale, checkpoint-validation
Comfy NVFP4 layer {prefix!r} has an incompatible pre_quant_s	validation	error	quantization, nvfp4, pre-quant-scale, shape-mismatch, checkpoint-validation
Comfy tensorwise INT8 layer {prefix!r} needs a 2D weight, go	validation	error	quantization, int8, tensorwise, rank-mismatch, checkpoint-validation
Comfy INT8 layer {prefix!r} needs I8 weights and F32 scales,	validation	error	quantization, int8, dtype-mismatch, checkpoint-validation
Comfy INT8 layer {prefix!r} has incompatible weight/scale sh	validation	error	quantization, int8, rowwise, shape-mismatch, checkpoint-validation
Cannot collate mixed VLA state presence	validation	error	vla, batching, robotics, validation
Cannot collate mixed VLA noise presence	validation	error	vla, diffusion-noise, batching, validation
Invalid VLA prefix cache layer: {layer_idx}	exception	error	vla, prefix-cache, kv-cache, index-error
Warmup image path is required for image-input model	exception	error	warmup, image-input, server-startup, configuration
Invalid Hugging Face {field_name}: {path!r}	validation	error	huggingface, weights, url-parsing, path-traversal
Weight URL pins revision {url_revision!r}, which conflicts w	validation	error	huggingface, weights, revision-conflict, configuration
Only huggingface.co weight URLs are supported; use a local p	validation	error	huggingface, weights, unsupported-host, url-parsing
Diffusion weights must come from a Hugging Face model repo	validation	error	huggingface, weights, wrong-repo-type
Hugging Face weight URL has no model repo: {source!r}	validation	error	huggingface, weights, malformed-url
Unsupported Hugging Face weight URL: {source!r}	validation	error	huggingface, weights, unsupported-url-shape
Hugging Face weight URL has no filename: {source!r}	validation	error	huggingface, weights, url-parsing
Weight source {source!r} is neither a local path nor an owne	validation	error	weights, source-parsing, huggingface
Weight file {source.filename!r} was not found in {source.rep	exception	error	huggingface, weights, file-not-found
Weight subfolder {source.subfolder!r} was not found in {sour	exception	error	huggingface, weights, subfolder
Weight path does not exist: {local_path}	exception	error	local-filesystem, weights, path-not-found
Requested weight {weight_name!r} was not found	exception	error	weights, file-selection
Weight name {weight_name!r} matches multiple files: {list(ba	validation	error	weights, ambiguous-selection
Source contains no recognized weight files	exception	error	weights, file-selection, no-candidates
Source contains multiple independent weight files; select on	validation	error	weights, ambiguous-selection
--num-inference-steps must be at least 2	validation	error	cli, validation, diffusion
--mode {args.mode} requires {mode_variant}	validation	error	cli, validation, config-mismatch
MiniMax H3 AdaLN cache must be built on CUDA	validation	error	cuda, environment, cli
AdaLN cache must cover at least one timestep plan	validation	error	cli, validation, diffusion, timesteps
Could not resolve a transformer directory from: {path}	exception	error	modelopt, fp8, path-resolution
Could not resolve backbone.pt from: {path}	exception	error	modelopt, fp8, path-resolution
Expected an index file or a single safetensors shard in {mod	validation	error	modelopt, fp8, safetensors, weight-map
Only per-tensor FP8 scales are supported for diffusion check	validation	error	modelopt, fp8, quantization
Expected a flat quantization_config dict in the ModelOpt exp	validation	error	modelopt, fp8, config
This tool only supports ModelOpt diffusers FP8 exports (quan	validation	error	modelopt, fp8, quantization, config
BF16 fallback patterns are enabled, but --base-transformer-d	validation	error	modelopt, fp8, cli, missing-argument
DeepSeekV4 only supports interleave CP strategy, got {cfg.cp	validation	error	deepseek, context-parallel, sglang, config-validation
DeepSeekV4 CP supports moe_a2a_backend in {supported_a2a_bac	validation	error	deepseek, moe, a2a-backend, context-parallel, sglang
Invalid expert_pack configuration:\n{details}	validation	error	gguf, expert-pack, sglang, path-validation
HiSparse supports DSA {label} backend(s) {sorted(allowed_bac	validation	error	hisparse, dsa, attention-backend, kv-cache-dtype, sglang
HiSparse requires one of {HISPARSE_KV_CACHE_DTYPES} KV cache	validation	error	hisparse, kv-cache-dtype, sglang, config-validation
--enable-hisparse is not supported with the unified-KV path	validation	error	hisparse, rocm, flashmla, env-var, sglang
Kimi-K3 DCP with decode_attention_backend='cutedsl_mla' requ	exception	error	kimi-k3, flashinfer, version-mismatch, dcp, sglang
Kimi-K3 DCP with decode_attention_backend='cutedsl_mla' requ	exception	error	kimi-k3, flashinfer, signature-check, dcp, sglang
Kimi-K3 DCP + DSPARK currently requires SGLANG_RAGGED_VERIFY	validation	error	kimi-k3, dspark, speculative-decoding, env-var, sglang
Decode attention backend for Kimi-K3 DCP must be 'cutedsl_ml	validation	error	kimi-k3, attention-backend, dcp, sglang
GptOssForCausalLM on Intel XPU only supports bfloat16 dtype,	validation	error	gpt-oss, intel-xpu, dtype, sglang
MiniCPM does not support DP attention	validation	error	minicpm, dp-attention, sglang, config-validation
MiniCPM SALA does not support hierarchical cache	validation	error	minicpm, hierarchical-cache, hicache, sglang
MiniCPM sparse attention does not support PD disaggregation	validation	error	minicpm, pd-disaggregation, sparse-attention, sglang
TensorRT-LLM MLA backend only supports kv-cache-dtype of fp8	validation	error	sglang, trtllm, mla, kv-cache-dtype, config-validation
tokenspeed_mla backend is only supported on Blackwell GPUs (	validation	error	sglang, tokenspeed, mla, hardware-gpu, blackwell
tokenspeed_mla backend requires kv-cache-dtype=fp8_e4m3, got	validation	error	sglang, tokenspeed, mla, kv-cache-dtype, config-validation
CuteDSL MLA backend is only supported on Blackwell GPUs (SM1	validation	error	sglang, cutedsl, mla, prefill, hardware-gpu
CuteDSL MLA backend only supports kv-cache-dtype of fp8_e4m3	validation	error	sglang, cutedsl, mla, kv-cache-dtype, config-validation
Dual chunk attention is enabled, but attention backend is se	validation	error	sglang, dual-chunk-attention, attention-backend, config-conflict
--quantization nvfp4_online is supported only on NVIDIA Blac	validation	error	sglang, nvfp4, quantization, hardware-gpu, blackwell
--quantization nvfp4_online supports only --moe-runner-backe	validation	error	sglang, nvfp4, moe-runner-backend, quantization, config-conflict
{}: {} not model-overridable; declarations are limited to th	validation	error	sglang, model-overrides, declarations, publish-gate, internal-api
PD decode DCP requires --disaggregation-transfer-backend moo	validation	error	sglang, pd-disaggregation, dcp, transfer-backend, config-validation
PD decode DCP currently requires chunk cache; --disaggregati	validation	error	sglang, pd-disaggregation, dcp, radix-cache, config-conflict
PD decode DCP currently requires chunk cache; --enable-hiera	validation	error	sglang, pd-disaggregation, dcp, hierarchical-cache, config-conflict
--disaggregation-decode-enable-radix-cache is incompatible w	validation	error	sglang, pd-disaggregation, radix-cache, hisparse, config-conflict
--disaggregation-decode-enable-radix-cache is incompatible w	validation	error	sglang, pd-disaggregation, radix-cache, fake-backend, config-conflict
--disaggregation-decode-enable-radix-cache is incompatible w	validation	error	sglang, pd-disaggregation, radix-cache, speculative-decoding, config-conflict
SGLANG_DISAGG_STAGING_BUFFER requires disaggregation_transfe	validation	error	sglang, pd-disaggregation, staging-buffer, environment-variable, config-validation
SGLANG_RUST_SERVER serves the PD KV bootstrap registry on th	validation	error	sglang, pd-disaggregation, rust-server, bootstrap-port, port-conflict
Gemma4AssistantForCausalLM draft requires --speculative-algo	validation	error	sglang, speculative-decoding, eagle3, gemma4, model-incompatibility
--speculative-draft-window-size must be positive, got {}.	validation	error	sglang, speculative-decoding, window-size, argument-validation
DFLASH speculative decoding only supports CUDA and NPU devic	validation	error	sglang, dflash, speculative-decoding, device-support, hardware-gpu
Currently DFLASH speculative decoding does not support dp at	validation	error	speculative-decoding, dflash, dp-attention, server-args
Currently DFLASH speculative decoding only supports pp_size	validation	error	speculative-decoding, dflash, pipeline-parallel, server-args
DFLASH speculative decoding requires setting --speculative-d	validation	error	speculative-decoding, dflash, draft-model, missing-argument
DFLASH requires --speculative-dflash-block-size to be positi	validation	error	speculative-decoding, dflash, argument-validation, block-size
Both --speculative-num-draft-tokens and --speculative-dflash	validation	error	speculative-decoding, dflash, conflicting-arguments, block-size
--speculative-draft-window-size must be >= --speculative-num	validation	error	speculative-decoding, dflash, window-size, argument-validation
DSpark speculative decoding only supports CUDA or NPU device	validation	error	speculative-decoding, dspark, device-support, cuda, npu
DSpark with dp attention requires --enable-dp-lm-head.	validation	error	speculative-decoding, dspark, dp-attention, lm-head
DSpark with dp attention supports moe_a2a_backend 'none' (bu	validation	error	speculative-decoding, dspark, moe, a2a-backend, dp-attention
DSpark with dp attention + moe_a2a_backend={} requires SGLAN	validation	error	speculative-decoding, dspark, env-var, ragged-verify, moe
DSpark with dp attention does not support context parallel (	validation	error	speculative-decoding, dspark, context-parallel, dp-attention
DSpark ignores --speculative-moe-a2a-backend; with dp attent	validation	error	speculative-decoding, dspark, moe, a2a-backend, conflicting-arguments
Currently DSpark speculative decoding only supports pp_size	validation	error	speculative-decoding, dspark, pipeline-parallel
DSpark dense speculative decoding requires setting --specula	validation	error	speculative-decoding, dspark, draft-model, missing-argument
DSpark requires --speculative-dspark-block-size to be positi	validation	error	speculative-decoding, dspark, argument-validation, block-size
DSpark speculative_num_draft_tokens must equal gamma + 1 (=	validation	error	speculative-decoding, dspark, conflicting-arguments, num-draft-tokens
DSpark could not resolve speculative_num_draft_tokens; set -	validation	error	speculative-decoding, dspark, missing-argument, num-draft-tokens
DSpark speculative_num_draft_tokens must be >= 2 (= gamma +	validation	error	speculative-decoding, dspark, num-draft-tokens, argument-validation
Currently standalone speculative decoding does not support d	validation	error	speculative-decoding, standalone, dp-attention
trtllm_mha backend only supports topk = 1 for speculative de	validation	error	speculative-decoding, attention-backend, trtllm-mha, eagle, topk
--speculative-use-rejection-sampling is only supported for E	exception	error	speculative-decoding, rejection-sampling, server-args, eagle
--speculative-use-rejection-sampling requires --speculative-	validation	error	speculative-decoding, rejection-sampling, eagle-topk, server-args
--speculative-use-rejection-sampling is incompatible with --	validation	error	speculative-decoding, rejection-sampling, accept-threshold, server-args
--speculative-use-rejection-sampling is incompatible with --	validation	error	speculative-decoding, rejection-sampling, determinism, server-args
--speculative-use-rejection-sampling with multi-layer EAGLE	validation	error	speculative-decoding, rejection-sampling, multi-layer-eagle, server-args
speculative_eagle_topk > 1 with page_size > 1 is only suppor	validation	error	speculative-decoding, eagle-topk, page-size, attention-backend, server-args
Ngram speculative decoding only supports CUDA or CPU devices	validation	error	speculative-decoding, ngram, device-support, rocm, server-args
--speculative-ngram-external-sam-budget must be positive whe	validation	error	speculative-decoding, ngram, external-corpus, server-args, validation
--speculative-ngram-external-corpus-max-tokens must be posit	validation	error	speculative-decoding, ngram, external-corpus, server-args, validation
speculative_ngram_external_sam_budget must be less than or e	validation	error	speculative-decoding, ngram, external-corpus, draft-tokens, server-args, validation
quantize_and_serve requires ModelOpt quantization (set with	validation	error	quantization, modelopt, config-validation, sglang
quantize_and_serve functionality is currently disabled due t	exception	error	quantization, modelopt, not-implemented, feature-disabled, sglang
Config list contains configs from 2 methods, must be only 1	exception	error	quantization, config-conflict, model-config, sglang
Quantization method specified in the model config ({quant_me	exception	error	quantization, config-mismatch, cli-args, sglang
Unknown quantization method: {self.quantization}. Must be on	exception	error	quantization, typo, unsupported-method, sglang
{self.quantization} quantization is currently not supported	exception	error	quantization, rocm, amd, hardware-support, sglang
Unknown dtype: {dtype}	exception	error	dtype, config-validation, sglang
Grammar mask max_rows must be positive, got {max_rows}	exception	error	grammar, constrained-decoding, validation
--enable-strict-thinking requires a grammar backend with tok	validation	critical	grammar, strict-thinking, xgrammar, startup
--enable-strict-thinking requires a grammar backend that sup	validation	critical	grammar, configuration, strict-thinking, startup
Invalid grammar backend: {name}	validation	error	grammar, configuration, startup
{matcher_error}	exception	error	grammar, llguidance, structured-output
think_end_token '{reasoning_parser.detector.think_end_token}	validation	error	reasoning, tokenizer, grammar, startup
Strict reasoning format requested but the grammar backend do	validation	error	reasoning, grammar, configuration
think_excluded_token '{token}' could not be encoded by the t	validation	error	reasoning, tokenizer, grammar
Unknown value for {flag}: {name}. Available: {list(mapping.k	validation	error	cli, preset, validation, whitelist
--diff-threshold with a single argument must be a float shor	validation	error	cli, parsing, float, threshold
--diff-threshold expects a single float shorthand or (regex	validation	error	cli, parsing, arity, threshold
tensor {name!r} matched no --diff-threshold pattern ({[rule.	validation	error	regex, threshold, pattern-matching, fullmatch
invalid predicate {expr!r}: {e}	validation	error	predicate, dsl, syntax-error, eval
invalid predicate {expr!r}: {e}; allowed names are {ALLOWED_	validation	error	predicate, dsl, name-error, whitelist
Length mismatch: {details}	validation	error	validation, length-mismatch, dataclass, invariants
{label}: directory {directory} has no .pt files at top level	validation	error	filesystem, ambiguity, directory-layout, dump
{label}: no .pt files found in {directory} or any of its sub	validation	error	filesystem, missing-files, dump, not-found
{cls_name}.{f.name}: expected {expected.__name__}, got {type	validation	error	type-check, dataclass, runtime-validation, config
Invalid config pair (missing '='): {pair!r}	validation	error	config, validation, debug-utils
Unknown config key {key!r}. Valid keys: {sorted(defaults)}	validation	error	config, validation, unknown-key
{key}: expected {field_type}, got {value!r}	validation	error	config, type-coercion, validation
cannot mix lambda extractor with static kwargs	exception	error	decorator, api-misuse, debug-utils
must provide either a lambda or static kwargs	exception	error	decorator, api-misuse, missing-argument
Unknown mode {mode!r}	exception	error	config, enum-value, debug-utils
[Grafter] tags={tags} matched BOTH grafter_b2t_filter and gr	exception	error	distributed, grafter, filter-config
requires #senders == #recvs but got #senders={len(received_l	exception	error	distributed, collective, grafter
requires matching shapes but received_list[{my_recv_rank}].s	exception	error	distributed, tensor-shape, grafter
Unknown dumper control method: {method!r}	exception	error	rpc, http-control, method-dispatch
RPC error on {self._debug_name}: {response['error']}	exception	error	rpc, debug-utils, remote-error, sglang
_load_function expects 'pkg.module.symbol', got {path!r} (mi	exception	error	import, config-validation, debug-utils, sglang
PR #{pr_num} revert is not registered; available: {sorted(_P	exception	error	env-var, configuration, debug-utils, sglang
Unknown router: {name}	exception	error	cli-arguments, simulation, schedule-simulator, sglang
Unknown scheduler: {name}	exception	error	cli-arguments, simulation, schedule-simulator, sglang
Unknown stop criteria: {self.stop_criteria}	exception	error	configuration, simulation, schedule-simulator, sglang
could not import any module prefix of '{qualified_name}'	exception	error	import, patching, source-patcher, sglang
resolved target '{qualified_name}' is not callable: {type(ta	exception	error	patching, type-check, source-patcher, sglang
empty match text	exception	error	patching, validation, source-patcher, sglang
match text not found in source:\n{preview}\n\nsource_len={le	exception	error	patching, text-matching, source-patcher, sglang
match text found multiple times ({len(found_indices)} occurr	exception	error	source-patching, text-match, ambiguity, sglang
only one of 'replacement', 'prepend', 'append' may be set, g	validation	error	validation, edit-spec, mutually-exclusive, sglang
Unknown data: {df.columns}. You may need to set `--data-type	exception	error	data-schema, polars, cli, text-comparison
Ascend PD transfer does not support HiSparse destination dev	exception	error	ascend, npu, disaggregation, hicache, not-implemented
Unsupported DisaggregationMode: {disaggregation_mode}	exception	critical	ascend, disaggregation, config, init
Ascend Transfer Engine initialization failed.	exception	critical	ascend, npu, transfer-engine, initialization, runtime
Unsupported DisaggregationMode: {self.disaggregation_mode}	exception	critical	disaggregation, config, init, pd
PD peers must connect matching DCP ranks, got prefill={self.	exception	critical	disaggregation, dcp, parallelism, bootstrap
Unsupported PD DCP topology: {self.dcp_size} -> {dst_dcp_siz	exception	critical	disaggregation, dcp, topology, parallelism
PD DCP source/destination KV geometry differs: src={src_toke	exception	critical	disaggregation, dcp, kv-cache, geometry-mismatch
Page size mismatch: prefill server has page_size={info.page_	exception	critical	disaggregation, pd-disagg, page-size, config-mismatch, kv-cache
KV cache dtype mismatch: prefill server has kv_cache_dtype={	exception	critical	disaggregation, pd-disagg, kv-cache-dtype, config-mismatch, quantization
PD decode DCP requires an MLA or hybrid-MLA KV pool.	exception	critical	disaggregation, dcp, context-parallel, mla, unsupported-feature
PD decode DCP currently requires prefill attention CP=1, got	exception	critical	disaggregation, dcp, attention-parallelism, prefill, config-mismatch
torch.distributed must be initialised before CommonKVManager	exception	critical	disaggregation, multi-node, torch-distributed, initialization-order
Unexpected compressed-MLA dst_kv_ptrs length {len(dst_kv_ptr	exception	error	disaggregation, mla, pipeline-parallel, kv-pointers, internal-invariant
Cannot resolve total_kv_heads: kv_args has neither total_kv_	exception	error	disaggregation, staging-buffer, kv-heads, missing-metadata
Staging is enabled but kv_manager._staging_ctx.allocator is	exception	error	disaggregation, staging-buffer, env-var, allocator, initialization
Staging is enabled but kv_manager.kv_buffer_tensors is None.	exception	error	disaggregation, staging-buffer, kv-buffer-tensors, initialization
[Staging] KV transfer via staging buffer failed: {e}. sessio	exception	error	disaggregation, staging-buffer, transfer-failure, wrapper-exception
group_concurrent_contiguous requires equal-length src/dst in	exception	error	disaggregation, kv-transfer, index-mismatch, validation
PD DCP transfer requires decode_prefix_len to align to the v	validation	error	disaggregation, dcp, alignment, prefix-cache
num_kv_tokens must fit in the provided source pages, got tok	validation	error	disaggregation, dcp, capacity-check, kv-transfer
Insufficient destination DCP pages: required={required_pages	validation	error	disaggregation, dcp, destination-pages, allocation
SGLANG_DISAGG_STAGING_BUFFER is designed for non-MLA models	validation	error	disaggregation, staging-buffer, mla, env-var, unsupported-feature
PP consensus is required when pp_size > 1	validation	error	disaggregation, pipeline-parallel, consensus, api-misuse
rids_to_check cannot be used in PP mode	validation	error	disaggregation, pipeline-parallel, api-misuse, argument-conflict
DSV4 HiSparse direct PD transfer currently requires the Moon	validation	error	disaggregation, hisparse, dsv4, mooncake, unsupported-backend
Unsupported KV cache type for decode offload	validation	error	disaggregation, kv-offload, hicache, unsupported-pool-type
Invalid hicache storage backend extra config JSON: {e}	validation	error	hicache, json-config, decode-offload, config-parse-error
Grid dim ({_mm_grid_attrs[modality]}) not found in {mm_input	validation	error	multimodal, preprocessor, grid-metadata, kimi, validation
Invalid Kimi image grid metadata: {values}; expected [h, w]	validation	error	multimodal, kimi, grid-metadata, shape-validation
grpcs:// is not supported; use grpc://	validation	error	grpc, url-scheme, tls, network, configuration
Invalid modality: {modality}	validation	error	multimodal, modality, enum-validation, receiver
mm_content_hashes has {len(image_hashes)} entries for {image	validation	error	multimodal, content-hash, request-validation, receiver
mooncake encoder_transfer_backend requires HTTP encoders; us	validation	error	mooncake, rdma, grpc, disaggregation, configuration, epd
gRPC encode only supports IMAGE modality, got: {non_image}	validation	error	grpc, multimodal, modality, not-implemented, encoder
EPD MMReceiver: http mode requires http:// encoder URLs. Set	validation	error	configuration, url-scheme, env-var, grpc, http, epd
Unsupported transport_mode: {transport_mode}	validation	error	configuration, transport-mode, env-var, receiver
{error_msg}	http	error	encoder, pipeline, error-wrapper, epd, multimodal
encode metadata not ready	http	error	timeout, metadata, encoder, disaggregation
no staged embedding for /send req_id={req_id} (already relea	http	error	encoder, send, request-lifecycle, race-condition
launch_local_runtime requires --dp-size 1; got dp_size={get_	validation	critical	startup, config, dp-size, parallelism
Encoder DP mode requires --dp-size > 1 and --tp-size 1; got	validation	critical	startup, config, dp-size, tp-size, parallelism
Feature attrs ({_mm_feature_attrs[modality]}) not found in {	validation	error	multimodal, schema, preprocessor, validation
No embedding available for request: {state.req_id}	http	error	internal, encoder, state-machine, race-condition
Encoder request was released: {state.req_id}	http	error	lifecycle, race-condition, encoder, send
Inconsistent receive_count for req_id={req_id}: registered {	http	error	validation, refcount, dp, tensor-parallel
Encoder produced {mm_embedding.shape[0]} tokens, but preproc	http	error	multimodal, token-count, encoder, preprocessor, mismatch
Encoder produced {mm_embedding.shape[0]} tokens, but preproc	http	error	disaggregation, multimodal, encoder, shape-mismatch
No embedding available for Mooncake GPU-direct transfer: {re	http	critical	mooncake, disaggregation, encoder, gpu-direct
Mooncake transfer_sync failed for {req_id} (session={session	http	critical	mooncake, rdma, transfer, network
Rank 0 produced no embedding for {ctx.req_id}	http	critical	encoder, disaggregation, mooncake, staging
Invalid endpoint: must contain 'inproc' or 'tcp'	validation	error	kv-events, zmq, endpoint, config
Unknown event publisher '{kind}'	validation	error	kv-events, config, registry
[Staging] Bulk RDMA transfer failed with ret={ret}. src_ptr=	exception	critical	mooncake, rdma, kv-cache, staging
--enable-unified-memory does not support different prefill /	exception	error	pd-disagg, unified-memory, tensor-parallel, mamba
PD KV layout mismatch on the whole-envelope path: prefill ha	exception	error	pd-disagg, unified-memory, kv-layout, page-size
Mamba state layouts differ between prefill and decode (src i	exception	error	pd-disagg, mamba, tp-degree-mismatch, unified-memory
PD Disaggregation does NOT support PD different TP sizes for	exception	error	pd-disagg, tp-degree-mismatch, hybrid-model, non-mla
{st.upper()} state index length mismatch: prefill={len(src_i	exception	critical	pd-disagg, state-index, kv-corruption-guard
PD disagg: PP>1 not supported for MiniMax sparse index yet.	exception	error	pd-disagg, minimax, pipeline-parallel, sparse-attention
PD disagg: heterogeneous TP not supported for MiniMax sparse	exception	error	pd-disagg, minimax, tp-degree-mismatch
HiSparse destination device indices are not supported by PD	exception	error	pd-disagg, dcp-relayout, hicache, kv-cache
Transfer thread failed because of {e}. Prefill instance with	exception	critical	pd-disagg, transfer-thread, mooncake, wrapper
KVTransferError	exception	critical	pd-disagg, kv-transfer, error-propagation
Source and destination groups must have the same length	validation	error	mori, index-plan, validation
KV memory descriptors are empty on prefill side	exception	critical	mori, kv-cache, descriptor, initialization-order
Destination KV descriptors do not match prefill pp configura	validation	error	mori, pipeline-parallel, layer-mismatch, kv-cache
Destination MLA KV descriptors do not match prefill pp confi	validation	critical	disaggregation, prefill-decode, pipeline-parallel, mla, kv-cache
Head slice size evaluates to zero	validation	critical	disaggregation, tensor-parallel, kv-cache, integer-division
Slice size exceeds destination token capacity for TP slice t	validation	critical	disaggregation, tensor-parallel, heterogeneous-tp, kv-cache
PD state transfer failed: kv_args.state_types is empty but s	exception	error	disaggregation, hybrid-model, mamba, state-transfer, configuration
PD state transfer failed: state component count mismatch (lo	exception	error	disaggregation, hybrid-model, state-transfer, version-mismatch
PD state transfer failed: unknown state_type={st}	exception	error	disaggregation, state-transfer, dispatch, unsupported-feature
PD state transfer failed: mamba requires single state index,	exception	error	disaggregation, mamba, state-transfer, batching
PD state transfer does not support TP-mismatched non-MLA SWA	exception	error	disaggregation, swa, tensor-parallel, heterogeneous-tp, unsupported-feature
PD disagg: PP>1 not supported for MiniMax sparse index yet.	exception	error	disaggregation, minimax, pipeline-parallel, sparse-attention, unsupported-feature
PD disagg: heterogeneous TP not supported for MiniMax sparse	exception	error	disaggregation, minimax, tensor-parallel, heterogeneous-tp, sparse-attention
NIXL PD transfer does not support HiSparse combined with dec	exception	error	nixl, disaggregation, speculative-decoding, hicache, pd-disaggregation
NIXL KV transfer has no KV memory segments	validation	error	nixl, disaggregation, hicache, memory-registration, pd-disaggregation
NIXL heterogeneous-TP direct-to-host KV transfer is not impl	exception	error	nixl, disaggregation, heterogeneous-tp, hicache, pd-disaggregation
Missing NIXL destination KV memory kind	exception	error	nixl, disaggregation, internal-invariant, transfer-worker
Missing aux index for last chunk	exception	error	nixl, disaggregation, chunked-transfer, internal-invariant
NIXL transfer encountered ERR room={room}	exception	error	nixl, rdma, network, disaggregation, transfer-failure
NIXL memory registration failed for {mem_kind} kv tensors	exception	critical	nixl, memory-registration, startup, rdma, disaggregation
NIXL memory registration failed for aux tensors	exception	critical	nixl, memory-registration, startup, disaggregation
NIXL memory registration failed for state tensors	exception	critical	nixl, memory-registration, hybrid-model, startup, disaggregation
KVTransferError(self.bootstrap_room, failure_reason)	exception	critical	sglang, nixl, kv-transfer, disaggregation, distributed-inference
NIXL KVSender Exception	exception	critical	sglang, nixl, kv-transfer, disaggregation, tp-rank-crash
NIXL KVReceiver Exception	exception	critical	sglang, nixl, kv-receive, disaggregation, distributed-inference
SGLANG_DISAGG_STAGING_BUFFER is designed for non-MLA models	exception	error	sglang, disaggregation, env-var, mla, config-validation
SGLANG_DISAGG_STAGING_BUFFER requires a positive chunked_pre	exception	error	sglang, disaggregation, chunked-prefill, page-size, config-validation
SGLANG_DISAGG_STAGING_BUFFER with pp_size > 1 is only suppor	validation	error	sglang, disaggregation, pipeline-parallelism, mooncake, nixl, config-validation
SGLANG_DISAGG_STAGING_BUFFER does not support prefill contex	validation	error	sglang, disaggregation, context-parallelism, staging-buffer, config-validation
top_logprobs_num {top_logprobs_len} exceeds disaggregation m	validation	error	sglang, disaggregation, logprobs, metadata-buffer, capacity
return_sampling_mask with disaggregation requires SGLANG_DIS	validation	error	sglang, disaggregation, sampling-mask, env-var, config-validation
Sampling mask length {mask_len} exceeds disaggregation metad	validation	error	disaggregation, sampling, buffer-capacity, pd-disaggregation
Unsupported transfer backend: {transfer_backend}	validation	error	disaggregation, transfer-backend, invalid-argument, configuration
PD disaggregation for MiniMax sparse layers with index value	validation	error	disaggregation, minimax, sparse-attention, not-implemented
DSV4 draft state transfer expects SWA-only NextN layers	validation	error	disaggregation, deepseek-v4, speculative-decoding, swa
DSV4 target and draft pools must use the same unified-KV mod	validation	error	disaggregation, deepseek-v4, unified-kv, configuration-mismatch
DSV4 target and draft pools must share SWA ring geometry: ta	validation	error	disaggregation, deepseek-v4, swa, geometry-mismatch
DSV4 target and draft pools must share the SWA index mapping	validation	error	disaggregation, deepseek-v4, swa, index-mapping
DSV4 target and draft pools must share paged SWA geometry: t	validation	error	disaggregation, deepseek-v4, page-size, sliding-window
--enable-tp-lm-head-all-to-all requires an available PyNCCL	validation	critical	distributed, nccl, p2p, tp, lm-head
The memory capacity is unbalanced. Some GPUs may be occupied	error_code	error	distributed, tp, gpu-memory, resource-conflict
CUDART error: {error_str}	error_code	critical	cuda, gpu, out-of-memory, invalid-device, runtime
Error happened when batch testing peer-to-peer access from {	exception	error	cuda, p2p, custom-all-reduce, nccl, subprocess
File {normalized_input} does not exist.	validation	error	config, mooncake, ib-devices, file-not-found
Failed to parse JSON content from file {normalized_input}	validation	error	json, config, mooncake, ib-devices
Failed to read JSON file {normalized_input}: {exc}	validation	error	config, file-permissions, mooncake, io
Invalid JSON mapping: {normalized_input}	validation	error	json, config, mooncake, cli
Invalid format: expected a mapping from GPU id to IB device	validation	error	json, config, schema, mooncake
Invalid format: keys must be integers (or string representat	validation	error	json, config, schema, mooncake, type-mismatch
No valid GPU mappings found in JSON	validation	error	json, config, empty-config, mooncake
No IB devices configured for GPU {gpu_id}. Available GPUs: {	validation	error	config, mooncake, ib-devices, gpu-mapping
Please install mooncake by following the instructions at htt	exception	critical	mooncake, dependency, import-error, kv-cache-transfer, sglang
Mooncake's batch register requires a newer version of moonca	exception	error	mooncake, version-mismatch, batch-register, upgrade-required, sglang
Mooncake Transfer Engine initialization failed.	exception	critical	mooncake, rdma, initialization, native-return-code, sglang
Mooncake's batch transfer requires mooncake-transfer-engine	exception	error	mooncake, version-mismatch, batch-transfer, upgrade-required, sglang
Failed to import 'set_transfer_engine' from 'mooncake.pg'. P	exception	error	mooncake, version-mismatch, elastic-ep, import-error, sglang
NCCL only supports CUDA, ROCm and MUSA backends.	validation	error	nccl, pytorch, cuda, backend-not-supported, distributed, sglang
world_size ({world_size}) is not equal to tensor_model_paral	exception	error	parallelism, tensor-parallel, pipeline-parallel, world-size, config-validation, sglang
decode_context_parallel_size ({decode_context_parallel_size}	exception	error	parallelism, decode-context-parallel, config-validation, sglang
Decode context parallel (decode_context_parallel_size > 1) i	exception	error	parallelism, decode-context-parallel, platform-support, cuda, rocm, sglang
tensor_model_parallel_size ({tensor_model_parallel_size}) mu	exception	error	parallelism, decode-context-parallel, tensor-parallel, divisibility, config-validation, sglang
thinking.budget_tokens must be >= 1024 (got {})	validation	error	anthropic, thinking, validation, request-validation, pydantic
thinking.budget_tokens is not allowed when thinking.type is	validation	error	anthropic, thinking, validation, request-validation, pydantic
thinking.display is not allowed when thinking.type is 'disab	validation	error	anthropic, thinking, display, validation, request-validation
thinking.budget_tokens is not allowed when thinking.type is	validation	error	anthropic, thinking, adaptive, validation, request-validation
Model is required	validation	error	anthropic, model, validation, request-validation, missing-field
max_tokens must be positive	validation	error	anthropic, max-tokens, validation, request-validation
Anthropic redacted_thinking history is not supported	http	error	anthropic, redacted-thinking, conversation-history, request-conversion
tool_choice references tool {tool_name!r} but it is not in t	http	error	anthropic, tool-choice, tools, validation, request-conversion
tool_choice={tc_type!r} requires at least one custom tool; a	http	error	anthropic, tool-choice, built-in-tools, request-conversion
No tool call found	exception	error	agent, tool-call, dispatch, no-op
SGLANG_RUST_SERVER is not supported with the offline Engine	exception	error	sglang, environment-variable, offline-engine, startup
routed_dp_rank={routed_dp_rank} out of range [0, {dp_size})	exception	error	sglang, data-parallel, argument-validation
Multi-node weight cache daemons (nnodes > 1) require --dist-	exception	error	sglang, multi-node, distributed, weight-cache, startup
Weight cache daemon for pp_rank={pp_rank} tp_rank={tp_rank}	exception	error	sglang, weight-cache, timeout, startup, daemon
Weight cache daemon (pid={p.pid}) exited prematurely with co	exception	critical	sglang, weight-cache, daemon-crash, startup
engine_info_bootstrap_port {bootstrap_port} is already in us	exception	error	sglang, port-conflict, multi-instance, startup, bootstrap
Initialization failed. Please see the error messages above.	exception	critical	sglang, scheduler, startup, subprocess
{e}	http	error	sglang, http-400, bootstrap-server, transfer-engine
Invalid rank parameter	http	error	sglang, http-400, parameter-validation, bootstrap-server
No transfer engine info for rank {rank}	http	warning	sglang, http-404, race-condition, bootstrap-server
Unknown req_type: {req_type!r} (expected 'generate' or 'embe	exception	error	grpc, request-validation, dispatch, sglang
gRPC mode requires the smg-grpc-servicer package. If not ins	exception	critical	grpc, dependency, import-error, installation, sglang
--enable-metrics requires smg-grpc-servicer ≥ 0.5.3 (the ver	exception	error	grpc, metrics, version-mismatch, dependency, sglang
No call message found for {call_id}	exception	error	harmony, tool-calls, conversation-history, validation, sglang
Unknown input type: {response_msg['type']}	exception	error	harmony, input-validation, discriminator, sglang
Unknown output type: {type(output)}	exception	error	harmony, output-parsing, type-dispatch, sglang
Invalid number of contents in browser message	exception	error	harmony, browser-tool, message-structure, validation, sglang
Unknown browser action: {recipient}	exception	error	harmony, browser-tool, unknown-action, sglang
Unknown recipient: {message.recipient}	exception	error	harmony, recipient-routing, message-parsing, sglang
Unknown channel: {message.channel}	exception	error	harmony, channel-routing, message-parsing, sglang
Invalid thinking_mode `{thinking_mode}`	exception	error	deepseek, chat-template, thinking-mode, validation
Invalid message for role `{role}`: {msg}	exception	error	deepseek, developer-message, message-validation
Invalid messages at {index}: {assistant_msg}	exception	error	deepseek, tool-calls, conversation-history
No tool calls but found tool output	exception	error	deepseek, tool-calls, validation
ThinkingMode: {thinking_mode}, invalid message without reaso	exception	error	deepseek, thinking-mode, multi-turn
Unknown role: {role}	exception	error	deepseek, unknown-role, chat-template
Tool call format error	exception	error	deepseek, tool-parsing, model-output
Assistant tool call function.arguments must be a JSON object	validation	error	deepseek, tool-calls, json-validation
`task` requires at least one message with role='user' or 'de	validation	error	deepseek-v4, task, message-validation
Invalid reasoning effort profile: {reasoning_effort_profile!	validation	error	deepseek-v4, reasoning-effort, validation
Invalid reasoning effort {reasoning_effort!r} for profile {r	validation	error	deepseek-v4, reasoning-effort, validation
deepseek_v4 merges tool messages into user; please preproces	validation	error	deepseek-v4, tool-calls, preprocessing
Unknown role: {role}	validation	error	deepseek-v4, unknown-role, message-validation
max_tokens must be positive	validation	error	openai-api, max-tokens, pydantic-validation
thinking parts require exactly one of 'thinking' or 'text'	validation	error	openai-api, thinking-part, pydantic-validation
'role' must be one of {allowed} (case-insensitive).	validation	error	openai-api, role, validation
'role' must be a string	validation	error	openai-api, role, type-error
thinking content parts are only valid in assistant messages	validation	error	openai-api, thinking-part, role-validation
reasoning_effort must not be a boolean	validation	error	openai-api, reasoning-effort, type-error
invalid reasoning effort: {effort!r}	validation	error	openai-api, reasoning, validation, sglang
tool_choice 'required' or a named tool cannot be combined wi	validation	error	openai-api, tool-calling, structured-output, constrained-decoding, sglang
Value error, parameter top_n should be larger than 0.	validation	error	embeddings, validation, openai-api, sglang
Exactly one of 'prompt' or 'messages' must be provided.	validation	error	tokenization, validation, openai-api, sglang
Function tools must include a name.	validation	error	responses-api, tool-calling, validation, sglang
Cannot combine tool calls with constrained decoding (text.fo	validation	error	responses-api, tool-calling, constrained-decoding, structured-output, sglang
Invalid X-Data-Parallel-Rank header: must be an integer, got	http	error	http-header, data-parallel, routing, sglang
Assistant tool call function.arguments must be valid JSON.	validation	error	tool-calling, json, chat-template, openai-api, sglang
Assistant tool call function.arguments must be a JSON object	validation	error	tool-calling, json, validation, openai-api, sglang
Inkling reasoning_effort must not be a boolean	validation	error	inkling, reasoning, type-validation, sglang
Inkling reasoning_effort must be in [0.0, 0.99]	validation	error	inkling, reasoning, parameter-out-of-range, sglang
invalid Inkling reasoning_effort: {value!r}	validation	error	inkling, reasoning, validation, sglang
SGLANG_INKLING_DEFAULT_REASONING_EFFORT must be numeric	exception	error	environment-variable, inkling, reasoning, server-config, sglang
SGLANG_INKLING_DEFAULT_REASONING_EFFORT must be in [0.0, 0.9	exception	error	environment-variable, inkling, reasoning, server-config, sglang
Harmony does not support reasoning effort {reasoning_effort}	validation	error	gpt-oss, harmony, reasoning, openai-api, sglang
return_prompt_token_ids is not supported with streaming. Ple	validation	error	streaming, token-ids, openai-api, sglang
return_token_ids is not supported with streaming on /v1/chat	validation	error	streaming, token-ids, openai-api, sglang
return_meta_info is not supported with streaming. Please set	validation	error	streaming, meta-info, openai-api, sglang
{template_error}	validation	error	jinja, chat-template, openai-api, bad-request
expected a JSON array of tool calls, got {type(tool_call_dat	exception	error	tool-calls, json-parsing, openai-api
every tool call must be a JSON object with a 'name'	exception	error	tool-calls, json-parsing, validation
Cannot rewrap thinking history: no reasoning detector is con	exception	error	reasoning, chat-history, server-config
Anthropic thinking is not supported for models without a rea	validation	error	anthropic-api, reasoning, server-config
Reasoning parser '{self.reasoning_parser}' is always-on and	validation	error	anthropic-api, reasoning, toggle
Anthropic thinking is not supported for reasoning parser '{s	validation	error	anthropic-api, reasoning, unsupported-feature
id2label mapping is missing	exception	critical	classify, model-config, startup
schema_ is required for json_schema response format request.	validation	error	json-schema, response-format, validation
embed_override_token_id is required when embed_overrides is	validation	error	embeddings, request-validation
embed_override_token_id requires embed_overrides to be provi	validation	error	embeddings, request-validation
{template_error}{suffix}	validation	error	jinja, embeddings, chat-template
Failed to render chat template for embedding input: {templat	validation	error	jinja, embeddings, chat-template, type-error
--sidecar requires importable module {module_name!r} with a	exception	critical	sidecar, module-import, startup, sglang, cli
--sidecar requires module {module_name!r} to expose a callab	exception	critical	sidecar, callable, startup, sglang
No browser tool call found	validation	error	tool-calling, browser, validation, sglang
browser.search requires a query	validation	error	browser, search, argument-validation, tool-calling
browser.find requires a pattern	validation	error	browser, find, argument-validation, tool-calling
Unknown browser action: {recipient}	validation	error	browser, unknown-action, tool-calling, dispatch
browser.open requires a cursor or url	validation	error	browser, open, argument-validation, tool-calling
Unknown browser cursor: {cursor}	validation	error	browser, cursor, state, tool-calling
No URL recorded for browser cursor: {cursor}	validation	error	browser, state-corruption, url, tool-calling
{e}	http	warning	http, loads, bad-request, metrics, sglang
structure_info not used for JSON schema constraints	exception	error	function-call, not-implemented, json-schema, parser
Kimi K3 uses its model-native structural tag implementation	exception	error	kimi-k3, function-call, not-implemented, structural-tag
Kimi K3 additional parameter schema accepts no values	validation	error	kimi-k3, json-schema, tool-calling, validation
Kimi K3 tool parameters 'properties' must be an object	validation	error	kimi-k3, json-schema, tool-calling, validation, properties
Kimi K3 tool parameters 'required' must be a string list	validation	error	kimi-k3, json-schema, required-fields, validation
Kimi K3 required parameters are missing schemas: {sorted(mis	validation	error	kimi-k3, json-schema, required-fields, validation
Kimi K3 tool property schemas must be JSON schemas	validation	error	kimi-k3, json-schema, tool-calling, validation
Kimi K3 required parameter {key!r} accepts no values	validation	error	kimi-k3, json-schema, required-fields, validation
Kimi K3 tool parameters 'additionalProperties' must be a sch	validation	error	kimi-k3, json-schema, additional-properties, validation
Kimi K3 strict tool {tool.function.name!r} must define param	validation	error	kimi-k3, json-schema, tool-calling, validation, parameters
sgl_kernel.metal is importable, but the native Metal extensi	error_code	critical	mlx, metal, aot-kernel, apple-silicon, install
AttentionOffsetCache should not store data	error_code	error	mlx, kv-cache, api-misuse
WindowedAttentionKVCache holds only the trailing window and	error_code	error	mlx, kv-cache, sliding-window, attention-mask
BatchedDecodeContext requires full_kv_pool_index_by_layer wh	validation	error	mlx, speculative-decode, aot-kernel, kv-cache, config-validation
Cannot determine attention scale for {type(inner).__name__}	error_code	error	mlx, attention, model-compatibility, patching
Cannot determine attention head counts for {type(inner).__na	error_code	error	mlx, attention, model-compatibility, patching
Unexpected q_proj output shape {q_proj_output.shape} for {ty	error_code	error	mlx, decode, attention, shape-mismatch
MLX auxiliary-state radix cache does not support enable_mamb	error_code	error	mlx, mamba, unsupported-feature, server-args
Layer count and attention attribute count differ: {len(layer	validation	error	mlx, kv-cache, layout, validation
MLX model has no supported attention layers	error_code	error	mlx, kv-cache, layout, model-compatibility
no_rope_layers contains non-binary entries {bad_flags}; each	validation	error	config-validation, mlx, rope, muse-glimmer
layer_types has {len(self.layer_types)} entries but num_hidd	validation	error	config-validation, layer-types, muse-glimmer
layer_types contains unknown entries {bad}; expected only 'f	validation	error	config-validation, layer-types, enum-values
layer_types disagrees with no_rope_layers (NoPE layers must	validation	error	config-validation, consistency-check, layer-types
muse_glimmer_mlx_format {self.muse_glimmer_mlx_format} is no	validation	error	version-mismatch, packaging, mlx
config.json claims a packaged Muse Glimmer MLX artifact (mus	validation	error	checkpoint-format, packaging, weight-keys
not a complete raw Muse Glimmer HF checkpoint: {len(missing)	validation	error	checkpoint-integrity, missing-keys, weight-loading
embed_tokens.weight has shape {embed_shape} but config says	validation	error	shape-mismatch, config-validation, embedding
raw q_proj.weight has shape {raw_q_shape}, expected ({H * D}	validation	error	shape-mismatch, fused-gate, packaging
{gate_name} has shape {tuple(g.shape)}, expected ({H * D}, {	validation	error	shape-mismatch, attention-gate, weight-loading
SGLANG_USE_MLX requires stable Torch 2.13.x and MLX >= 0.32.	exception	error	mlx, apple-silicon, dependency-missing, mps, sglang
SGLANG_USE_MLX requires stable Torch 2.13.x and MLX >= 0.32.	exception	error	mlx, version-mismatch, torch, mps, sglang
SGLANG_USE_MLX requires an available PyTorch MPS device	exception	error	mlx, mps, torch, apple-silicon, device-unavailable
SGLANG_USE_MLX requires an available MLX Metal device	exception	error	mlx, metal, apple-silicon, device-unavailable, sglang
MLX async runner does not support forward mode: {forward_mod	validation	error	mlx, forward-mode, async-scheduler, sglang, unsupported-feature
Unknown MLX async mode: {launch.mode}	validation	error	mlx, async-scheduler, internal-contract, sglang
Only ver=3 is supported for MUSA FA3.	validation	error	musa, flash-attention, moore-threads, sglang
CP attention for non-FIA path on Ascend is not yet implement	exception	error	ascend, npu, context-parallel, fia, not-implemented, huawei
The 'enable-mixed-chunk' feature is currently unsupported in	exception	error	ascend, npu, mixed-chunk, mla, deepseek, not-implemented, huawei
n must be a positive power of 2, got {n}	exception	error	ascend, dsv4, hadamard, validation, npu, sglang
GPTQ act_order on XPU requires each group_size block of inpu	exception	error	xpu, gptq, quantization, tensor-parallel, act-order, not-implemented
kv-canary: {name} must be positive, got {value}	exception	error	kv-canary, validation, capacities, value-error
kv-canary: req_to_token_pool_size must be positive, got {req	exception	error	kv-canary, validation, req-pool, value-error
kv-canary: max_seq_len_per_req must be positive, got {max_se	exception	error	kv-canary, validation, seq-len, value-error
kv-canary: pool_slot_count must be positive, got {pool_slot_	exception	error	kv-canary, validation, kv-cache-pool, value-error
kv-canary: cuda_graph_max_bs must be non-negative, got {cuda	exception	error	kv-canary, validation, cuda-graph, value-error
kv-canary: speculative_num_draft_tokens must be non-negative	exception	error	kv-canary, validation, speculative-decoding, value-error
kv-canary: max_prefill_tokens must be positive, got {max_pre	exception	error	kv-canary, validation, prefill, scheduler, value-error
kv-canary: kv_canary must be one of none/log/raise, got {mod	exception	error	kv-canary, config, cli-args, enum-validation
kv-canary: launch_per_forward not supported on sweep endpoin	exception	error	kv-canary, endpoint, sweep-kernel, not-implemented
kv-canary: read_bytes must be non-negative, got {requested}	exception	error	kv-canary, validation, valueerror, read-bytes
kv-canary: read_bytes must be <= num_bytes_per_token ({num_b	exception	error	kv-canary, validation, valueerror, read-bytes
kv-canary: read_bytes must be a multiple of {_REAL_KV_READ_A	exception	error	kv-canary, alignment, validation, valueerror
kv-canary: {type(obj).__name__} missing required method {met	exception	error	kv-canary, monkey-patching, attributeerror
kv-canary: {type(obj).__name__}.{method_name} already wrappe	exception	error	kv-canary, monkey-patching, idempotency, runtimeerror
walk_radix_cache_for_canary does not support {cache_type.__n	exception	error	kv-canary, radix-cache, notimplementederror, type-dispatch
walk_radix_cache_for_canary does not support {type(radix_cac	exception	error	kv-canary, radix-cache, notimplementederror, locking
Activation function {act_fn_name!r} is not supported.	validation	error	activation, config, valueerror, model-loading
intermediate_size must be specified for scaled activation fu	validation	error	activation, quantization, fp8, config, valueerror
trtllm_mla backend can only be used with MLA models.	validation	error	attention-backend, trtllm, mla, server-args, valueerror
trtllm_mla cannot serve decode context parallelism with spec	validation	critical	attention-backend, trtllm-mla, context-parallelism, speculative-decoding, mla, sglang
tokenspeed_mla backend can only be used with MLA models.	validation	error	attention-backend, tokenspeed-mla, mla, model-arch-mismatch, sglang
cutedsl_mla backend can only be used with MLA models.	validation	error	attention-backend, cutedsl-mla, mla, model-arch-mismatch, sglang
trtllm_mha backend can only be used with non-MLA models.	validation	error	attention-backend, trtllm-mha, mla, model-arch-mismatch, sglang
hpc_ops backend can only be used with non-MLA models.	validation	error	attention-backend, hpc-ops, mla, model-arch-mismatch, npu, sglang
Cross attention is not supported in the hpc_ops attention ba	validation	error	attention-backend, hpc-ops, encoder-decoder, cross-attention, npu, sglang
hpc_ops backend does not support speculative decoding for no	validation	error	attention-backend, hpc-ops, speculative-decoding, npu, sglang
Short-conv hybrid models (ZAYA1 CCA, LFM2 / LFM2-MoE) are no	validation	critical	npu, ascend, short-conv, lfm2, zaya1, attention-backend, not-implemented, sglang
Expected hybrid GDN or NemotronH models, but got unknown mod	validation	error	hybrid-model, linear-attention, gdn, nemotron-h, model-registry, sglang
kv-canary: forward_batch.batch_size={bs} exceeds pre-allocat	validation	error	kv-canary, capacity, cuda-graph-max-bs, batch-size, sglang
kv-canary: forward_batch token count={num_tokens} exceeds pr	validation	error	kv-canary, capacity, chunked-prefill, max-prefill-tokens, sglang
DeepSeek-V4 flashmla_sparse_q8 prefill requires SM90 CUDA GP	exception	error	deepseek-v4, flashmla, q8-kv, sm90, gpu-compatibility, sglang
DeepSeek-V4 flashmla_sparse_q8 prefill requires d_v=512, got	exception	error	deepseek-v4, flashmla, head-dim, model-config-mismatch, sglang
DSV4 ragged verify does not support context parallel (CP); s	exception	critical	deepseek-v4, ragged-verify, context-parallelism, env-var, speculative-decoding, sglang
DSV4 ragged verify does not support online c128 MTP; set SGL	exception	critical	deepseek-v4, ragged-verify, mtp, online-compress, env-var, sglang
Dots SWA latent decode requires page_size=64, got {backend.p	exception	error	dots-hybrid, swa-mla, page-size, attention-backend, sglang
SGLANG_DSA_TOPK_BROADCAST requires PyNCCL during CUDA graph	exception	critical	dsa, pynccl, cuda-graph, broadcast, tp, env-var, sglang
DSA indexer weights_proj LoRA is incompatible with piecewise	exception	error	dsa, lora, piecewise-cuda-graph, prefill, deepseek, sglang
DSA indexer only supports CUDA, HIP, and NPU	exception	critical	dsa, device-support, xpu, not-implemented, deepseek, sglang
Invalid version: {self.fa_impl_ver=}	validation	error	flash-attention, version, init, sglang
MXFP8 KV cache requires the FA4 backend.	exception	error	mxfp8, kv-cache, flash-attention-4, quantization, sglang
MXFP8 KV cache requires per-token Q scales (q_descale) from	exception	error	mxfp8, q-descale, quantization, flash-attention-4, sglang
score_mod is only supported by the FA4 backend.	exception	error	score-mod, flash-attention-4, extend, sglang
rel_bias (sheared bias) is only supported by the FA4 backend	exception	error	rel-bias, flash-attention-4, extend, sglang
The hpc_ops attention backend does not support logit cap.	validation	error	attention-backend, hpc-ops, logit-cap, unsupported-feature, sglang
The hpc_ops attention backend only supports the default soft	validation	error	attention-backend, hpc-ops, softmax-scaling, head-dim, sglang
The hpc_ops attention backend with an fp8_e4m3 KV cache requ	exception	error	hpc-ops, fp8, kv-cache-dtype, hunyuan, attention-backend, sglang
Invalid forward mode: {forward_batch.forward_mode=}	validation	error	forward-mode, hybrid-linear-attention, mamba, metadata, sglang
Invalid forward mode: {forward_mode=}	validation	error	cuda-graph, forward-mode, hybrid-linear-attention, capture, sglang
Mamba2AttnBackend's forward is called directly instead of th	exception	error	mamba, hybrid-linear-attention, interface-contract, not-implemented, sglang
spec_info is unset in TARGET_VERIFY mode; the extend_* metad	exception	error	speculative-decoding, target-verify, spec-info, intel-amx, metadata, sglang
FlashInfer GDN prefill is not supported with --enable-determ	validation	error	gdn, linear-attention, deterministic-inference, flashinfer, triton, config-validation, sglang
CuteDSLKDAKernel does not support target_verify	exception	error	sglang, kda, speculative-decoding, not-implemented, linear-attention
FlashInfer KDA kernel (recurrent_kda) is not available. Requ	exception	error	sglang, flashinfer, kda, gpu-compatibility, sm100
f"recurrent_kda needs a [N, HV, V, K] state pool; got shape	validation	error	sglang, kda, tensor-shape, validation, state-pool
f"recurrent_kda state inner strides must be compact (V*K, K,	validation	error	sglang, kda, tensor-stride, contiguity, validation
f"recurrent_kda state pool breaks the compiled stride contra	validation	error	sglang, kda, alignment, memory-layout, flashinfer
FlashInfer KDA verify kernel only supports topk=1 (retrieve_	exception	error	sglang, kda, speculative-decoding, topk, flashinfer
f"KDA verify needs {draft_token_num} scratch steps, but inte	exception	error	sglang, kda, speculative-decoding, buffer-sizing
FlashInfer KDA verify requires an identity intermediate row-	exception	error	sglang, kda, speculative-decoding, index-mapping, invariant
FlashInferKDAKernel has no prefill kernel; keep prefill on T	exception	error	sglang, flashinfer, kda, prefill, not-implemented
The 'flashkda' KDA prefill backend requires the flash_kda mo	exception	error	sglang, flashkda, kda, missing-dependency, pip-install
FlashKDAKernel only supports prefill (extend)	exception	error	sglang, flashkda, kda, decode, not-implemented
The Helion package is required when a KDA backend is set to	exception	error	sglang, helion, kda, missing-dependency, version-pin
f"SGLang KDA state must be [B,H,V,K] with (V,K)={expected},	validation	error	sglang, nvidia, kda, tensor-shape, layout
NVIDIA KDA state must be [B,H,K,V] with (K,V)={expected}, go	validation	error	sglang, nvidia, kda, tensor-shape, layout
NvidiaKDAKernel is prefill-only	exception	error	sglang, nvidia, kda, decode, not-implemented
NvidiaKDAKernel does not support target_verify	exception	error	sglang, nvidia, kda, speculative-decoding, not-implemented
PtxKDAKernel is prefill-only	exception	error	sglang, ptx, kda, gb300, decode, not-implemented
PtxKDAKernel does not support target_verify	exception	error	sglang, ptx, kda, speculative-decoding, not-implemented
{self.__class__.__name__} does not support target_verify	exception	error	sglang, linear-attention, speculative-decoding, not-implemented, base-class
Lightning (seg_la) linear-attention backend does not support	exception	error	sglang, lightning, seg-la, speculative-decoding, topk, linear-attention
MiniCPM fused top-k only supports bfloat16 and float16, got	validation	error
trtllm_mla does not forward the cyclic DCP metadata to its d	exception	error	sglang, attention-backend, context-parallelism, not-implemented, mla, speculative-decoding
output_ws should be prepared for cuda-graph mode	exception	error	sglang, vision-transformer, cuda-graph, kwargs-validation, multimodal
cuda-graph mode cu_seqlens should be a list	exception	error	sglang, vision-transformer, cuda-graph, type-validation, cu-seqlens
VisionFlash3Attention is only available for cuda or musa	exception	error	sglang, vision-transformer, flash-attention, platform-support, hardware-compat
VisionFlash4Attention is only available for cuda	exception	error	sglang, vision-transformer, flash-attention-4, platform-support, hardware-compat
VisionFlashInferAttention is only available for cuda	exception	error	sglang, vision-transformer, flashinfer, platform-support, hardware-compat
sequence_lengths should be prepared for vision flashinfer_cu	exception	error	sglang, vision-transformer, flashinfer-cudnn, kwargs-validation, multimodal
max_seqlen should be prepared for vision flashinfer_cudnn at	exception	error	sglang, vision-transformer, flashinfer-cudnn, kwargs-validation, max-seqlen
flashinfer_cudnn expects packed indptrs as a torch.Tensor	exception	error	sglang, vision-transformer, flashinfer-cudnn, type-validation, cu-seqlens
FlashInfer allreduce fusion mnnvl backend requires a Blackwe	validation	error	flashinfer, allreduce-fusion, gpu-architecture, mnnvl, backend-selection
cuMemGetAllocationGranularity failed for FlashInfer workspac	error_code	error	cuda-driver, flashinfer, workspace-preflight, granularity, driver-version
cuMulticastGetGranularity failed for FlashInfer workspace pr	error_code	error	cuda-driver, multicast, flashinfer, workspace-preflight, nvswitch
Pack: Only supports tensors with dimensions not greater than	validation	error	int4, quantization, tensor-shape, packing, weight-loading
Expected hidden_size to be {self.hidden_size}, but found: {h	validation	error	layernorm, shape-mismatch, hidden-size, validation
Expected hidden_size to be at least {self.variance_size_over	validation	error	layernorm, variance-override, shape-validation, gdn
Unknown Shard Id {shard_id}	validation	error	weight-loading, shard-id, qkv-fusion, quant-scale
{loaded_weight} are not all equal	validation	error	npu, quant-scale, weight-loading, per-tensor-quant, allclose
Shard id with multiple indices is not supported in weight_lo	validation	error	weight-loading, shard-id, merged-column, api-version
Expected scalar scale for fused-in-checkpoint merged-column	validation	error	weight-loading, per-tensor-scale, quantization, fused-checkpoint, shard-id
expert-pack header coverage is inconsistent	exception	critical	moe, expert-pack, binary-format, header-validation, sglang
expert-pack is not identity triplet layout	exception	critical	moe, expert-pack, binary-format, flags, layout-mismatch
expert-pack alignment is invalid	exception	critical	moe, expert-pack, binary-format, alignment, corruption
expert-pack data offset is invalid	exception	critical	moe, expert-pack, binary-format, offset, alignment
expert-pack index is truncated	exception	critical	moe, expert-pack, binary-format, truncated-file, corruption
expert-pack index role or rank is invalid	exception	critical	moe, expert-pack, binary-format, index-entry, version-skew
expert cache and staging budgets, and read splits, must be p	exception	error	moe, expert-pack, configuration, vram-budget, argument-validation
expert-pack stats flush interval cannot be negative	exception	error	moe, expert-pack, configuration, interval, argument-validation
expert-pack direct I/O is unavailable on this platform	exception	error	moe, expert-pack, direct-io, platform-support, configuration
expert-pack or manifest is missing: {self.path}, {self.manif	exception	critical	moe, expert-pack, file-not-found, manifest, path-resolution
Kimi active routed MoE layers must be exactly 1..92	validation	critical	kimi, moe, expert-pack, manifest-validation
Kimi manifest expert-pack path does not match pack_path	validation	critical	kimi, moe, expert-pack, path-mismatch
Kimi expert-pack size does not match its manifest	validation	critical	kimi, moe, expert-pack, file-size-mismatch
Kimi expert-pack physical role order is unsupported	validation	critical	kimi, moe, expert-pack, role-order
Kimi expert-pack {role} quant type is unsupported	validation	critical	kimi, moe, quantization, unsupported-dtype
Kimi expert-pack role sizes do not match object bytes	validation	critical	kimi, moe, expert-pack, size-consistency
Kimi expert-pack header is truncated	validation	critical	kimi, moe, expert-pack, truncated-file
Kimi expert-pack header does not match its manifest	validation	critical	kimi, moe, expert-pack, header-validation
Kimi expert-pack index is truncated	validation	critical	kimi, moe, expert-pack, truncated-index
Kimi expert-pack identity mismatch at index {index}	validation	critical	kimi, moe, expert-pack, index-integrity
Kimi expert-pack range mismatch at index {index}	validation	critical	kimi, moe, expert-pack, range-validation
Kimi expert object is not contiguous at index {index}	validation	critical	kimi, moe, expert-pack, layout
Kimi expert-pack file has trailing or missing bytes	validation	critical	kimi, moe, expert-pack, file-size-mismatch
Kimi expert-pack index SHA-256 does not match manifest	validation	critical	kimi, moe, expert-pack, sha256, integrity
full pack verification requested, but manifest has no full S	validation	error	kimi, moe, expert-pack, sha256, missing-field
Kimi expert-pack SHA-256 does not match manifest	validation	critical	kimi, moe, expert-pack, sha256, integrity
Unsupported cute dtype {input.dtype}	validation	error	flashinfer, cutedsl, moe, unsupported-dtype
CuteDSL masked MoE supports activation 'silu' (gated) or 're	validation	error	flashinfer, cutedsl, moe, activation, unsupported-config
Can't import trtllm_fp8_block_scale_moe from flashinfer. Ple	exception	error	flashinfer, trtllm, moe, fp8, import-error, version-mismatch
Can't import trtllm_fp8_block_scale_routed_moe from flashinf	exception	error	flashinfer, trtllm, moe, fp8, import-error, version-mismatch
The hpc_ops MoE runner backend does not support fused shared	validation	error	sglang, moe, hpc-ops, shared-experts, config-validation
The hpc_ops MoE runner backend does not support apply_router	validation	error	sglang, moe, hpc-ops, router-weight, config-validation
The hpc_ops MoE runner backend does not support no_combine (	validation	error	sglang, moe, hpc-ops, no-combine, config-validation
The hpc_ops MoE runner backend runs a plain SiLU-and-mul; it	validation	error	sglang, moe, hpc-ops, swiglu, activation, config-validation
The hpc_ops MoE runner backend only supports FP8-quantized M	validation	error	sglang, moe, hpc-ops, fp8, quantization, config-validation
Unknown gemm type: {gemm_type}	validation	error	sglang, moe, humming, gemm, dispatch
Unsupported activation: {self.activation}	validation	error	sglang, moe, humming, activation, config-validation
cannot found moe_block_size for shape {valid_shape_m}	validation	error	sglang, moe, humming, tuning-config, batch-size, index-out-of-range
DeepEP returned FP8 input while Humming is configured for BF	validation	error	sglang, moe, humming, deepep, fp8, dtype-mismatch, distributed
Humming expected DeepEP FP8 hidden states and group-128 scal	validation	error	sglang, moe, humming, deepep, fp8, dtype-mismatch, distributed
aiter is required when SGLANG_USE_AITER is set to True	validation	critical	moe, rocm, aiter, import-error, env-var, amd
fuse_swiglu_interleaved set on an incompatible fused_moe cal	validation	error	moe, triton, swiglu, quantization, dtype, feature-guard
Unsupported activation: {activation=}, with {is_gated=}	validation	error	moe, activation, fallback-kernel, validation
Unsupported ascend_dispatcher_output_dtype: {self.ascend_dis	validation	error	ascend, npu, moe, dispatcher, dtype, quantization
combine() called before dispatch()	error_code	error	ascend, npu, dispatcher, lifecycle, state-machine
unsupported mode	validation	error	deepep, moe, dispatcher, enum, cuda-graph
DeepEP is not installed. Please install DeepEP package from	validation	critical	deepep, moe, import-error, expert-parallel, distributed
Ascend A2/A3 NPU does not support nvfp4 deepep_dispatcher_ou	validation	error	ascend, npu, deepep, nvfp4, quantization, hardware-support
triton runner was supported but it's temporarily disabled	error_code	error	deepep, deepgemm, triton, moe, not-implemented, feature-flag
Invalid deepep_mode: {self.deepep_mode}	validation	error	deepep, moe, dispatcher, enum, validation, version-skew
Invalid quantization method: {quantization}. Available metho	validation	error	quantization, configuration, startup, sglang
Invalid quantization method on CPU: {quantization}. Availabl	validation	error	quantization, cpu, amx, platform-support
Unsupported weight_bits: {weight_bits}, currently only suppo	validation	error	quantization, auto-round, weight-bits
Unsupported data_type: {data_type}, currently only support	validation	error	quantization, auto-round, data-type
Unsupported packing_format: {packing_format}, currently only	validation	error	quantization, auto-round, packing
Unsupported backend: {backend}, currently only support {se	validation	error	quantization, auto-round, backend
Fused MoE layer '{layer_name}' requires consistent quant con	validation	error	quantization, auto-round, moe, fused-layer
Fused module '{layer_name}' requires consistent quant config	validation	error	quantization, auto-round, fused-module, qkv
SGLang's AutoRound CPU inference path currently supports onl	validation	error	quantization, auto-round, cpu, amx, weight-bits
SGLang's AutoRound GPTQ loader supports desc_act=False only.	validation	error	quantization, auto-round, gptq, desc-act
The Triton WNA16 MoE backend only supports symmetric INT4 gr	validation	error	quantization, moe, triton, int4, compressed-tensors
The W8A8Int8 Fused MoE scheme is implemented only for NPU fo	exception	error	quantization, moe, int8, npu, hardware-support
The W4A8Int8 Fused MoE scheme is implemented only for NPU fo	exception	error	quantization, moe, w4a8, npu, hardware-support
Unsupported FusedMoe scheme: {weight_quant}, {input_quant}	exception	critical	quantization, moe, unsupported-scheme, compressed-tensors
{scheme.__class__.__name__} is not supported on XPU (no XPU	exception	error	xpu, intel-gpu, quantization, hardware-support, fp8
Block-quantized lm_head is not supported; use channel or ten	exception	error	quantization, lm-head, block-quantization, weight-loading, tp-sharding
A scheme must be defined for each layer	validation	error	quantization, scheme, linear-layer, internal-invariant
For Fused MoE layers, only {CompressionFormat.pack_quantized	validation	error	quantization, moe, mxint4, compressed-tensors, model-config
Current platform does not support NVFP4 quantization. Please	validation	error	nvfp4, gpu-hardware, blackwell, quantization, moe
Unsupported weight strategy={self.strategy}, supported strat	validation	error	quantization, fp8, w8a16, compressed-tensors, strategy
Unknown quantization strategy {self.strategy}	validation	error	quantization, fp8, w8a8, strategy, post-load
For FP8 Fused MoE layer, we require either per tensor or cha	validation	error	quantization, fp8, moe, static-scales, input-quantization
The output_size of gate's and up's weight = {intermediate_si	validation	error	quantization, fp8, moe, block-quantization, tensor-parallel, shape-mismatch
The input_size of down's weight = {intermediate_size_per_par	validation	error	quantization, fp8, moe, block-quantization, tensor-parallel, shape-mismatch
Unsupported weight quantization strategy: {self.weight_quant	validation	error	quantization, fp8, moe, strategy, compressed-tensors
QuantConfig has static quantization, but found activation sc	validation	error	quantization, fp8, moe, missing-scales, checkpoint-corruption
Unknown quantization strategy {self.strategy}	validation	error	quantization, int8, strategy, w8a8
Static compressed-tensors scheme is not yet supported on NPU	exception	error	npu, int8, quantization, static-scales, not-implemented
For INT8 Fused MoE layers, we require channelwise, dynamic p	validation	error	quantization, int8, moe, strategy
For INT8 Fused MoE layers, we require channelwise, dynamic p	validation	error	quantization, int8, moe, static-scales
Marlin kernels require group quantization or channelwise qua	validation	error	quantization, marlin, int4, group-quantization, strategy
Unsupported num_bits = {num_bits}. Supported num_bits = {WNA	validation	error	quantization, marlin, bit-width, unsupported-format
For Fused MoE layers, only {CompressionFormat.pack_quantized	validation	error	quantization, moe, marlin, format, compressed-tensors
Found different quantization schemes for {shard_proj_names}	validation	error	quantization, compressed-tensors, fused-layers, config-mismatch
Unable to find matching target for {layer_name} in the compr	validation	error	quantization, compressed-tensors, layer-matching, config-mismatch
expert-pack requires --disable-shared-experts-fusion so the	validation	error	expert-pack, moe, shared-experts, launch-flag
expert-pack v1 supports only single-GPU TP=EP=1	validation	error	expert-pack, moe, tensor-parallel, single-gpu
Unknown KV cache quantization method: '{name}'. Available: {	validation	error	quantization, kv-cache, config-validation, registry
DeepSeek-V4 FP4 experts require torch.float4_e2m1fn_x2 suppo	exception	critical	quantization, fp4, pytorch-version, deepseek
Unsupported activation scheme {activation_scheme}	validation	error	quantization, fp8, config-validation, checkpoint-config
The block-wise quantization only supports fp8-serialized che	validation	error	quantization, fp8, block-quant, checkpoint-config
The quantization block size of weight must have 2 dimensions	validation	error	quantization, fp8, config-validation, shape-validation
The block-wise quantization only supports dynamic activation	validation	error	quantization, fp8, block-quant, activation-scheme
MXFP8 requires weight_block_size=[1, 32].	validation	error	quantization, mxfp8, config-validation
Weight input_size_per_partition = {input_size_per_partition}	validation	error	quantization, fp8, tensor-parallel, shape-validation
Weight output_partition_size = {output_partition_size} is no	validation	error	quantization, fp8, tensor-parallel, shape-validation
--fp8-gemm-backend=deep_gemm cannot serve MXFP8 weight shape	exception	error	quantization, mxfp8, deep-gemm, hardware-compatibility, gemm-backend
The output_size of gate's and up's weight = {intermediate_si	validation	error	quantization, fp8, moe, tensor-parallel, shape-validation
The input_size of down's weight = {intermediate_size_per_par	validation	error	quantization, fp8, moe, tensor-parallel, shape-validation
Found static activation scheme for checkpoint that was not s	validation	error	quantization, fp8, moe, activation-scheme, checkpoint-config
MXFP8 MoE quantization requires SM100 or ROCm gfx95 (gfx942	exception	error	quantization, mxfp8, moe, hardware-compatibility, rocm
QuantConfig has static quantization, but found activation sc	validation	error	quantization, fp8, moe, missing-weights, activation-scheme
The hpc_ops MoE runner backend requires static activation sc	validation	error	quantization, fp8, moe, moe-runner-backend, activation-scheme
The hpc_ops MoE runner backend does not support MoE GEMM bia	validation	error	quantization, moe, moe-runner-backend, unsupported-feature
Unsupported runner backend: %s	exception	error	quantization, moe, dispatch, version-skew
MXFP8 dense GEMM requested via --fp8-gemm-backend=flashinfer	error_code	error	quantization, mxfp8, gemm-backend, hardware-compatibility, flashinfer
MXFP8 dense GEMM requested via --fp8-gemm-backend=flashinfer	error_code	error	quantization, mxfp8, gemm-backend, flashinfer, hardware-compatibility
Currently, only 4bits is supported on CPU with AMX.	validation	error	gptq, quantization, cpu, amx, unsupported-operation
Currently, gptq_v2 is not supported on CPU with AMX.	validation	error	gptq, checkpoint-format, gptq-v2, cpu, amx
The input size is not aligned with the quantized weight shap	validation	error	gptq, tensor-parallel, shape-mismatch, quantization, cpu, amx
The output size is not aligned with the quantized weight sha	validation	error	gptq, tensor-parallel, shape-mismatch, pack-factor, cpu, amx
The input size is not aligned with the quantized weight shap	validation	error	gptq, tensor-parallel, shape-mismatch, group-size
The output size is not aligned with the quantized weight sha	validation	error	gptq, tensor-parallel, shape-mismatch, pack-factor
Currently, desc_act (True) is not supported by GPTQ quantiza	validation	error	gptq, npu, ascend, desc-act, act-order
Humming quantization requires `humming-kernels`. Please inst	exception	error	humming, quantization, missing-dependency, import-error, installation
W4AFP8 group_size must be a positive integer, got {group_siz	validation	error	humming, group-size, config-validation, quantization
W4AFP8 shape_k = {shape_k} must be divisible by group_size =	validation	error	humming, tensor-parallel, shape-mismatch, group-size, quantization
W4AFP8 shape_k = {shape_k} must be divisible by 8 for int32	validation	error	quantization, w4afp8, humming, tensor-parallel, shape-validation
FP8 weight_block_size must contain two positive integers, go	validation	error	quantization, fp8, config-validation, humming
Humming quantization for MoE only supports moe_runner_backen	validation	error	moe, humming, runner-backend, config-validation
Humming does not support DeepEP {output_dtype} dispatch; use	validation	error	deepep, humming, moe, dtype, config-validation
Humming FP8 dispatch requires {sublayer_name} K={shape_k} to	validation	error	deepep, humming, fp8, tensor-parallel, shape-validation
{self.__class__.__name__}.apply should not be called.	exception	error	kv-cache, quantization, api-misuse, not-implemented
Only support per-tensor scaling factor for fp8 KV cache	validation	error	kv-cache, fp8, scale-format, checkpoint-validation
NVFP4 global scale tensor must already be on the KV tensor d	validation	error	nvfp4, kv-cache, device-mismatch, quantization
num_bits must be 4 or 8, got {}	validation	error	marlin, awq, gptq, bit-width, kernel-support
Currently, only group size 128 and -1 (channelwise) is suppo	validation	error	marlin, gptq, awq, group-size, config-validation
The params dtype must be float16, but got {params_dtype}	validation	error	marlin, dtype, float16, config-validation
Weight output_size_per_partition = {output_size_per_partitio	validation	error	marlin, tensor-parallel, shape-validation, gptq, awq
Weight output_size_per_partition = {output_size_per_partitio	validation	error	marlin, tensor-parallel, pack-factor, shape-validation
Weight input_size_per_partition = {input_size_per_partition}	validation	error	marlin, tensor-parallel, shape-validation, gptq, awq
Weight input_size_per_partition = {input_size_per_partition}	validation	error	marlin, tensor-parallel, group-size, shape-validation
Each permutation group must reside on the same gpu	validation	error	marlin, tensor-parallel, tile-alignment, shape-validation
Weight output_size_per_partition = {output_size_per_partitio	validation	error	marlin, gptq, tensor-parallel, shape-validation
Weight input_size_per_partition = {input_size_per_partition}	validation	error	marlin, gptq, tensor-parallel, shape-validation
NVFP4 embedding is gather-only. Reaching here means a tied l	exception	error
ModelOptMixedPrecisionConfig only supports MIXED_PRECISION c	validation	error
No ModelSlim MoE scheme found for layer {prefix}	validation	error	modelslim, quantization, moe, ascend-npu, config
Mismatched ModelSlim quantization for W13 in layer {prefix}:	validation	error	modelslim, quantization, moe, fused-weights, config-mismatch
Missing ModelSlim MoE quantization description for layer {pr	validation	error	modelslim, quantization, moe, missing-keys, config
Unsupported ModelSlim MoE schemes for layer {prefix}: W13='{	validation	error	modelslim, quantization, moe, unsupported-scheme, version-mismatch
Detected some but not all shards of {prefix} are quantized.	validation	error	modelslim, quantization, fused-layers, mixed-precision, config
A scheme must be defined for each layer	validation	error	modelslim, quantization, scheme-uninitialized, runtime
weight_prefix must be 'w13' or 'w2', got '{weight_prefix}'	validation	error	modelslim, mxfp8, moe, constructor-validation, ascend-npu
weight_prefix must be 'w13' or 'w2', got '{weight_prefix}'	validation	error	modelslim, int4, moe, constructor-validation, ascend-npu
weight_prefix must be 'w13' or 'w2', got '{weight_prefix}'	validation	error	modelslim, mxfp4, moe, constructor-validation, ascend-npu
weight_prefix must be 'w13' or 'w2', got '{weight_prefix}'	validation	error	modelslim, int8, moe, constructor-validation, ascend-npu
weight_prefix must be 'w13' or 'w2', got '{weight_prefix}'	validation	error	quantization, moe, npu, modelslim, validation
Unsupported params_dtype: {params_dtype}	validation	error	quantization, dtype, npu, modelslim, w8a8
weight_prefix must be 'w13' or 'w2', got '{weight_prefix}'	validation	error	quantization, moe, npu, modelslim, validation
num_bits must be 4 or 8, got {}	validation	error	quantization, moe, gptq, awq, marlin
The quantization method moe_wna16 + awq is not supported for	validation	error	quantization, awq, moe, gpu-capability, hardware
moe_wna16 only support gptq and awq.	validation	error	quantization, moe, gptq, awq, config-validation
The package `amd-quark` is required to use MX-FP4 models. Pl	error_code	error	mxfp4, amd, quark, missing-dependency, quantization
Petit is not installed. Please install it with `pip install	exception	error	quantization, nvfp4, petit, missing-dependency, python
{error_msg}	exception	error	quantization, nvfp4, petit, config-validation, python
Online MXFP4 requantization from compressed-tensors NVFP4 ch	exception	error	quantization, mxfp4, nvfp4, compressed-tensors, quark, not-implemented, python
MIXED_PRECISION layer group {tail!r} has inconsistent quant	exception	error	quantization, mixed-precision, quark, mxfp4, config-validation, python
MIXED_PRECISION layer group {tail!r} uses unsupported quant	exception	error	quantization, mixed-precision, quark, unsupported-algo, mxfp4, python
Unsupported online_scheme: {online_scheme}	exception	error	quantization, quark, config-validation, online-requantization, python
Either quant_config or online_scheme must be provided	exception	error	quantization, quark, constructor, required-argument, python
MIXED_PRECISION checkpoint has no NVFP4 layers to requantize	exception	error	quantization, quark, mxfp4, mixed-precision, nvfp4, python
Requantization into {config['requantization_method']} is not	exception	error	quantization, quark, requantization, not-implemented, checkpoint, python
Online MXFP4 quantization for MoE layers requires an AMD ROC	exception	critical	quantization, mxfp4, amd, rocm, moe, hardware-unsupported
use_mxfp8=True is not supported in Quark MXFP4 requantizatio	exception	error	quantization, mxfp8, mxfp4, quark, config-mismatch
Only block_quant=True is supported in Quark MXFP4 requantiza	exception	error	quantization, fp8, block-quantization, quark, moe
Requantization in QuarkW4A4MXFp4MoE from {self.dequantizatio	exception	error	quantization, quark, unsupported-format, moe
Online MXFP4 quantization for MoE is only supported on AMD G	exception	critical	quantization, mxfp4, aiter, rocm, dependency-missing
Cannot restore flashinfer TRT-LLM BF16 MoE weight shape for	exception	error
backend must be a non-empty string	validation	error	sampler, validation, configuration, sglang
Failed to load LoRA adapter {lora_ref.lora_name} because it	validation	error	lora, duplicate, adapter, sglang
LoRA adapter {lora_ref.lora_name} with rank {lora_config.r}	validation	error	lora, memory-pool, rank, config, sglang
Failed to load LoRA adapter {lora_ref.lora_name} as a pinned	validation	error	lora, pinned, capacity, config, sglang
Failed to load LoRA adapter {lora_ref.lora_name}: {result.er	exception	critical	lora, startup, load-failure, sglang
Mixed shared-outer LoRA formats detected across loaded adapt	exception	error	lora, moe, shape-mismatch, sglang
SGLang does not recognize target_modules='{config.target_mod	validation	error	lora, target-modules, peft-config, sglang
SGLang currently only supports inferring LoRA target modules	validation	error	lora, target-modules, type-error, peft-config, sglang
LoRA adapter '{lora_name}' contains target modules {sorted(u	validation	error	lora, target-modules, subset-validation, sglang
LoRA targets the DSA indexer ({sorted(indexer_targets)}), wh	validation	error	lora, dsa, indexer, fusion, env-var, sglang
LoRA on Intern-S2-Mobius model.meta_mlp routed banks is not	validation	error	lora, intern-s2-mobius, moe, unsupported, sglang
LoRA with name {lora_name} does not exist. Loaded LoRAs: {se	validation	error	lora, registry, not-found, unload, sglang
The following requested LoRA adapters are not loaded: {name}	validation	error	lora, registry, not-found, request, sglang
LoRA with name {lora_ref.lora_name} already exists. Loaded L	validation	error	lora, registry, duplicate, register, sglang
experimental_sgl_marlin LoRA requires --lora-use-virtual-exp	validation	error	lora, marlin, moe, virtual-experts, experimental, sglang
experimental_sgl_marlin EP requires --moe-a2a-backend none	validation	error	moe, expert-parallelism, a2a, marlin, experimental, sglang
experimental_sgl_marlin LoRA requires --lora-use-virtual-exp	validation	error	lora, marlin, virtual-experts, startup-validation, experimental, sglang
experimental_sgl_marlin LoRA requires --lora-backend triton	validation	error	lora, backend, triton, marlin, experimental, sglang
experimental_sgl_marlin EP requires trivial expert placement	validation	error	moe, expert-parallelism, eplb, marlin, experimental, sglang
experimental_sgl_marlin configuration is unsupported: + ";	validation	error	gpu, compute-capability, marlin, hardware, experimental, sglang
LoRA pinned weight cache key collision for {cache_key!r}: ca	validation	error	lora, cache, shape-mismatch, pin-memory, sglang
Every extra_key should be a string.	validation	error	sglang, extra-key, cache-key, input-validation, batching
extra_key should be a list or a string.	validation	error	sglang, extra-key, type-error, input-validation
The length of cache_salt should be equal to the batch size.	validation	error	sglang, cache-salt, batching, input-validation
Every cache_salt should be a string.	validation	error	sglang, cache-salt, type-error, input-validation
cache_salt should be a list or a string.	validation	error	sglang, cache-salt, type-error
At least one of text, input_ids, or image should be provided	validation	error	sglang, empty-input, request-validation
text and input_ids cannot be provided at the same time	validation	error	sglang, input-conflict, tokenization, request-validation
lora_path list length ({len(self.lora_path)}) must match bat	validation	error	sglang, lora, batching, input-validation
return_flat_raw_top_logprobs requires rectangular top logpro	validation	error	sglang, logprobs, data-shape, input-validation
MM inputs where only some items are precomputed.	exception	error	sglang, multimodal, precomputed-embeddings, not-implemented
Input 'data' must be a torch.Tensor, but got {type(data)}	validation	error	sglang, torch, type-error, multimodal
Shared memory {name} not found	exception	critical	sglang, shared-memory, startup, race-condition, file-not-found
No processor registered for architecture: {hf_config.archite	validation	error	sglang, multimodal, unsupported-model, version-mismatch
SGLANG_RUST_SERVER does not yet apply --preferred-sampling-p	validation	error	sglang, rust-server, config-conflict, startup
SGLANG_RUST_SERVER=1: no native Rust MM pipeline for model_t	error_code	error	sglang, rust-server, multimodal, unsupported-model, startup
return_hidden_states must be a boolean or the string literal	validation	error	sglang, hidden-states, enum-validation, input-validation
Model vocab_size ({vocab_size}) exceeds MM_PAD_SHIFT_VALUE (	validation	critical	sglang, multimodal, vocab-size, constant-mismatch, startup
Invalid modality string: {modality_str}. Valid modalities ar	validation	error	sglang, modality, enum-validation
Unknown CacheAware Policy: {policy=}	validation	error	scheduling, config, enum-validation
Unknown CacheAgnostic Policy: {policy=}	validation	error	scheduling, config, enum-validation
Unknown schedule_policy: {policy=}	validation	error	startup, config, validation
[Elastic EP] WORLD MLP sync dp_size is out of sync: rank={to	panic	critical	distributed, elastic-ep, dp-attention, topology
[Elastic EP] WORLD MLP sync dp_size exceeds WORLD size: rank	panic	critical	distributed, elastic-ep, dp-size, topology
additional customized generation output is not supported by	validation	error	rust-egress, output-streaming, feature-incompatibility
[weight_cache] {op} of model weights is not supported while	error_code	error	weight-cache, cuda-ipc, memory-management, feature-incompatibility
Not enough data points for quadratic fitting ({len(L)} < 8).	validation	error	pipeline-parallel, profiling, chunked-prefill
Failed to fit coefficients: insufficient rank	validation	error	pipeline-parallel, profiling, linear-algebra
Failed to fit f(l) = al^2 + bl + c: {e}	validation	error	pipeline-parallel, profiling, linear-algebra
Fitted quadratic coefficient a={fitted_a:.2e} is not positiv	validation	error	pipeline-parallel, profiling, data-quality
Calculated target_latency={self.target_latency:.2f}ms is not	validation	error	pipeline-parallel, profiling, data-quality
LoRA is not enabled. Please set `--enable-lora` to enable Lo	http	error	lora, config, http-api
Didn't find any LoRA adapters when trying to evict LRU LoRA	validation	error	lora, eviction, capacity
Error while unloading LRU LoRA adapter '{lru_lora_name}': {u	validation	error	lora, eviction, unload
Streaming sessions are disabled. Please relaunch with --enab	validation	error	sessions, streaming, config, http-api
token_ids_logprob must be a flat list of integers.	validation	error	sglang, validation, logprob, request-validation
token_ids_logprob contains out-of-vocabulary token id {token	validation	error	sglang, vocab, out-of-range, logprob
The input_ids {seq} contains values greater than the vocab s	validation	error	sglang, input-ids, vocab, batch
The input_ids {input_ids} contains values greater than the v	validation	error	sglang, input-ids, vocab, validation
input contains {len(positions)} occurrences of embed_overrid	validation	error	sglang, embed-overrides, multimodal, count-mismatch
For multimodal input processing do not set `enable_tokenizer	validation	error	sglang, config, multimodal, batch-encode
Batch tokenization is not needed for pre-tokenized input_ids	validation	error	sglang, config, pretokenized, batch-encode
Batch tokenization is not needed for input_embeds. Do not se	validation	error	sglang, config, input-embeds, batch-encode
{finish_reason["message"]}	exception	error	sglang, abort, bad-request, non-streaming
LoRA adapter '{first_adapter}' was requested, but LoRA is no	validation	error	lora, adapter, server-args, configuration
Received request with {len(unique_lora_paths)} unique loras	validation	error	lora, multi-lora, limits, configuration
Got LoRA adapter that has never been loaded: {lora_path}\nAl	validation	error	lora, eviction, state-desync, cache
Failed to implicitly load LoRA adapter {lora_path}: {load_re	validation	error	lora, adapter-load, oom, incompatible-weights
Duplicate request ID detected: {rid}	validation	error	request-id, duplicate, retry, batching
Invalid prompts type for score_prompts.	validation	error	scoring, type-validation, input-format
{label} contains {len(positions)} occurrences of embed_overr	validation	error	embeddings, overrides, count-mismatch, validation
label_token_ids is required for generation (CausalLM) models	validation	error	scoring, causal-lm, labels, validation
items must be provided	validation	error	scoring, missing-argument, validation
embed_override_token_id is required when query_embed_overrid	validation	error	scoring, embeddings, overrides, validation
item_first is not supported when embeddings are supplied	validation	error	sglang, scoring, embedding-overrides, argument-validation
item_embed_overrides length ({len(item_embed_overrides)}) mu	validation	error	sglang, scoring, length-mismatch, embedding-overrides
Token ID {token_id} is out of vocabulary (vocab size: {vocab	validation	error	sglang, scoring, tokenizer, out-of-vocabulary
Invalid combination of query/items types for score_request.	validation	error	sglang, scoring, type-validation
return_pooled_hidden_states is not supported for CausalLM mo	validation	error	sglang, scoring, hidden-states, model-architecture
return_pooled_hidden_states is not supported for {archs[0]}.	validation	error	sglang, scoring, cross-encoder, hidden-states
Out of memory. Try to lower your batch size.\nTry to allocat	exception	critical	sglang, memory, kv-cache, allocation
Prefill out of memory. Try to lower your batch size.\nTry to	exception	critical	sglang, memory, paged-kv, prefill
alloc_req_slots runs out of memory. Please set a smaller num	exception	critical	sglang, memory, request-pool, capacity
Decode out of memory. Try to lower your batch size.\nTry to	exception	critical	sglang, memory, decode, paged-kv
--hicache-host-memory-mode buffer_only does not support side	validation	error	hicache, buffer-only, deepseek-v4, sidecar-pool, configuration
--hicache-host-memory-mode buffer_only on SWA models require	validation	error	hicache, buffer-only, swa, sliding-window, unified-kv, configuration
--hicache-host-memory-mode buffer_only requires an SWA host	validation	error	hicache, buffer-only, swa, memory-sizing, validation
Unknown retraction backup backend: {backend}	validation	error	retraction, backend-dispatch, invalid-value, internal
HiCache native hash is only supported on little-endian Linux	exception	error	native-extension, platform-support, linux-only, endianess
Failed to load HiCache native hash extension	exception	error	native-extension, build-failure, openssl, toolchain, cpp-extension
online c128 does not support MTP	validation	error	deepseek-v4, c128, speculative-decoding, mtp, incompatible-features
Use get_key_buffer instead.	exception	error	deepseek-v4, memory-pool, api-misuse, not-implemented
Short read for {suffixed}	error_code	error	hicache, file-io, cache-corruption, truncated-file
HiRadixCache only supports MHA, MLA, DSA, and MSA models	validation	error	hicache, hiradixcache, unsupported-architecture, model-support
Unsupported config file {path} (config format: {ext})	validation	error	hicache, config, file-format, extension
prefetch_threshold must be int, got {type(prefetch_threshold	validation	error	hicache, config-validation, prefetch, type-error
prefetch_timeout_base must be number, got {type(prefetch_tim	validation	error	hicache, config-validation, prefetch, type-error
prefetch_timeout_per_ki_token must be number, got {type(pref	validation	error	hicache, config-validation, prefetch, type-error
prefetch_timeout_max must be number, got {type(prefetch_time	validation	error	hicache, config-validation, prefetch, type-error
hicache_storage_pass_prefix_keys must be bool, got {type(hic	validation	error	hicache, config-validation, boolean, type-error
HiSparse device KV transfer requires sgl_kernel.kvcacheio (C	exception	critical	sgl-kernel, cuda, rocm, platform-support, hisparse
Dynamic HiCache sidecars require HostPoolGroup.	validation	error	hicache, hybrid-cache, sidecar, type-error
--enable-unified-memory with PD disaggregation does not supp	validation	error	unified-memory, pd-disaggregation, hybrid-swa, kv-cache, boot-config
--enable-unified-memory only supports hybrid Mamba and hybri	validation	error	unified-memory, model-architecture, kv-cache, boot-config
Speculative decoding with --enable-unified-memory is only su	validation	error	speculative-decoding, unified-memory, hybrid-swa, kv-cache
--prefill-only-disable-kv-cache expected NoOpMHATokenToKVPoo	exception	error	prefill-only, kv-cache, pool-family, boot-config
--prefill-only-disable-kv-cache is not supported for {unsupp	exception	error	prefill-only, mamba, fp4-kv, kv-cache
--enable-linear-replayssm-spec with DSPARK/DFLASH requires a	validation	error	speculative-decoding, kda, kimi-linear, dspark, dflash, boot-config
dcp_kv_mask is not supported for FP4 KV cache.	exception	error
NoOpMHATokenToKVPool.set_kv_buffer was called. This pool is	exception	critical	kv-cache, attention-backend, embedding, prefill-only
page-major layout has no per-layer contiguous regions; KV tr	exception	error	kv-cache, page-major, disaggregation, not-implemented
CPU offloading is unsupported under the page-major layout (T	exception	error	kv-cache, cpu-offload, page-major, not-implemented
prefix-valid commit is unsupported under the page-major layo	exception	error	kv-cache, prefix-cache, page-major, not-implemented
MXFP8 KV cache requires head_dim divisible by {self.MXFP8_SC	validation	error	kv-cache, mxfp8, quantization, head-dim
MXFP8 KV cache requires v_head_dim divisible by {self.MXFP8_	validation	error	kv-cache, mxfp8, quantization, v-head-dim
MXFP8 KV cache requires torch.float8_e8m0fnu support.	exception	error	mxfp8, kv-cache, pytorch-version, dtype
MXFP8 KV cache does not support SGLANG_USE_HND_KVCACHE.	validation	error	kv-cache, mxfp8, hnd-layout, env-var
MXFP8 KV cache does not support DCP KV masks.	exception	error	kv-cache, mxfp8, dcp-mask, not-implemented
MXFP8 KV cache requires K and V scale tensors.	validation	critical	mxfp8, kv-cache, quantization, scale-tensors, sglang
prefix-valid commit is unsupported for MXFP8 KV cache (it do	exception	error	mxfp8, kv-cache, prefix-caching, not-implemented, sglang
{layer_id=} not in full attention layers: {self.full_attenti	validation	error	kv-cache, hybrid-attention, layer-id, mapping, sglang
MHATokenToKOnlyPool does not allocate V	exception	error	kv-cache, k-only-pool, sparse-attention, minimax, sglang
MHATokenToKOnlyPool: use set_index_k_buffer on the parent Mi	exception	error	kv-cache, k-only-pool, sparse-attention, minimax, api-misuse, sglang
layer_id={layer_id} does not have an index V cache (either d	validation	error	kv-cache, sparse-attention, layer-id, mapping, minimax, sglang
layer_id={layer_id} is not a sparse attention layer; sparse	validation	error	kv-cache, sparse-attention, layer-id, mapping, minimax, sglang
layer.layer_id={layer.layer_id} does not have an index V cac	validation	error	kv-cache, sparse-attention, layer-id, write-path, minimax, sglang
layer.layer_id={layer.layer_id} is not in the K-only sparse	validation	error	kv-cache, sparse-attention, layer-id, write-path, minimax, sglang
move_kv_cache is not yet supported for MiniMaxSparseKVPool:	exception	error	sglang, kv-cache, not-implemented, speculative-decoding, minimax
LogicalHostPool size must be page-aligned, got size={size},	validation	error	sglang, memory-pool, page-alignment, validation
LogicalHostPool allocation must be page-aligned, got need_si	validation	error	sglang, memory-pool, allocation, page-alignment
Not enough host memory for V4 paged pool {pool_name}. Reques	validation	error	sglang, host-memory, hicache, out-of-memory, capacity
Unsupported layout: {self.layout}	validation	error	sglang, memory-pool, layout, invalid-config
Unsupported V4 paged host layout/backend: {self.layout}/{io_	validation	error	sglang, hicache, io-backend, layout, unsupported
Not enough host memory for DSA indexer hierarchical cache. R	validation	critical	memory, dsa-hicache, host-memory, startup
Unsupported layout: {self.layout}	validation	error	config, dsa-hicache, layout
Index buffer transfer expects page-aligned indices for DSA.	validation	error	dsa-hicache, page-alignment, transfer
Unsupported IO backend: {io_backend}	validation	error	dsa-hicache, io-backend, config
Layer-sharded DSA indexer HiCache backup with page_first lay	validation	error	dsa-hicache, layout, layer-sharding
Layer-sharded direct DSA indexer backup only supports layer_	validation	error	dsa-hicache, layout, io-backend
Mamba storage zero-copy requires page_first layout, got {sel	validation	error	sglang, mamba, layout, zero-copy, hierarchical-cache
Unsupported layout: {self.layout}	validation	error	sglang, kv-cache, layout, host-pool, initialization
Unsupported IO backend: {io_backend}	validation	error	sglang, io-backend, kv-cache, hierarchical-cache, configuration
Unsupported layout for models with head_dim != v_head_dim: {	validation	error	sglang, mla, kv-cache, layout, host-pool
Unsupported layout for models with head_dim != v_head_dim an	validation	error	sglang, mla, hicache, io-backend, layout
Unsupported layout for models with head_dim != v_head_dim an	validation	error	sglang, mla, hicache, io-backend, layout
Unsupported IO backend for models with head_dim != v_head_di	validation	error	sglang, mla, hicache, io-backend
get_split_heads_page_buffer_meta requires layout='page_head'	validation	error	sglang, hicache, not-implemented, kv-cache
Unsupported layout for models with head_dim != v_head_dim: {	validation	error	sglang, hicache, layout, kv-cache
Unsupported layout: {self.layout}	validation	error	sglang, mla, hicache, layout, init
Unsupported IO backend: {io_backend}	validation	error	sglang, mla, hicache, io-backend
Layer-sharded MLA HiCache backup with page_first layout requ	validation	error	sglang, mla, hicache, jit-kernel, sgl-kernel, build
Layer-sharded HiCache backup does not support layout: {self.	validation	error	hicache, mla, layout, context-parallelism, sglang
Layer-sharded direct HiCache backup only supports layer_firs	validation	error	hicache, direct-io, layout, sglang
Layer-sharded HiCache backup does not support IO backend: {i	validation	error	hicache, io-backend, context-parallelism, sglang
RadixKey index out of range: {idx}	exception	error	radix-cache, index-error, off-by-one, prefix-cache
RadixKey slice step must be 1	validation	error	radix-cache, slice, value-error, prefix-cache
RadixKey operations require matching extra_key, but got {sel	validation	error	radix-cache, extra-key, lora, value-error, prefix-cache
RadixKey operations require matching cache_salt, but got {se	validation	error	radix-cache, cache-salt, multi-tenant, value-error, prefix-cache
Host reference counter is already zero.	exception	error	radix-cache, refcount, hicache, double-release, runtime-error
cache_salt is not supported by the experimental C++ radix tr	validation	error	radix-cache, cpp-backend, cache-salt, unsupported-feature, value-error
Host cache is not supported yet	exception	error	radix-cache, cpp-backend, hicache, not-implemented, host-offload
Host-pool retraction does not support Mamba models.	validation	error	disaggregation, retraction, mamba, hybrid-ssm, unified-cache, value-error
Host-pool retraction does not support pure-SWA models.	validation	error	disaggregation, retraction, sliding-window, swa, unified-cache, value-error
--radix-cache-backend={name!r} is not registered. Registered	validation	error	radix-cache, backend-registry, plugin, server-args, value-error
--hicache-host-memory-mode buffer_only is only implemented f	validation	error	hicache, buffer-only, unified-cache, server-args, value-error
--enable-session-radix-cache requires UnifiedRadixCache, but	validation	error	session-cache, unified-cache, server-args, value-error
forward_batch with seq_lens is required for TopK retrieval	validation	error	sparse-attention, retrieval, forward-batch, seq-lens, value-error
Required: indexer, forward_batch, x, q_lora, positions	validation	error	deepseek, dsa, sparse-attention, retrieval, value-error
Quest query hidden size {hidden} not divisible by head_dim {	validation	error	quest, sparse-attention, head-dim, shape-mismatch, value-error
Unsupported query shape for Quest: {queries.shape}	validation	error	quest, sparse-attention, tensor-rank, shape-mismatch, value-error
Query heads {q_heads} not divisible by KV heads {kv_heads}	validation	error	quest, sparse-attention, gqa, head-mismatch, value-error
Unknown sparse algorithm: {algorithm_name}	validation	error	sparse-attention, factory, algorithm-name, value-error
Unsupported KV cache type {type(kvcache).__name__}: expected	validation	error	flexkv, kv-cache, attributeerror, attention-backend
Tag mismatch: expected CMD_LAYERWISE, got {payload.get('cmd'	exception	critical	flexkv, pipeline-parallel, protocol-mismatch, distributed
store_kv: token_ids has {n} entries but kv_indices has {len(	validation	error	flexkv, kv-cache, validation, off-by-one
Tag mismatch: expected CMD_PUT_META, got {payload.get('cmd')	exception	critical	flexkv, pipeline-parallel, protocol-mismatch, distributed
Tag mismatch: expected CMD_STORE_COMPLETE, got {payload.get(	exception	critical	flexkv, pipeline-parallel, protocol-mismatch, distributed
[FlexKV] Failed to connect to eventfd socket {self._layerwis	exception	critical	flexkv, unix-socket, worker-startup, connection-refused
Timed out waiting for ACK from FlexKV layerwise worker	exception	critical	flexkv, timeout, eventfd, worker-hang
FlexKV layerwise worker NACK'd eventfd transfer (ack={ack!r}	exception	critical	flexkv, eventfd, nack, file-descriptors
[FlexKV] Failed to send eventfds to {self._layerwise_socket}	exception	critical	flexkv, eventfd, retry-exhausted, unix-socket
hf3fs_fuse.io is not available. Please install the hf3fs_fus	exception	critical	hf3fs, importerror, missing-dependency, installation
Hf3fsClient.check: {offsets=}, {sizes=}	validation	error	hf3fs, validation, alignment, batch-io
Rank {rank} namespace '{namespace}' not initialized. Please	http	error	hf3fs, metadata-server, http-404, initialization-order
Failed to connect to metadata server: {e}	exception	critical	hf3fs, metadata-server, connection-refused, retries-exhausted
Namespace '{namespace}' for rank {rank} not initialized	exception	error	hf3fs, metadata, initialization-order, local-client
MLA model is not supported without global metadata server, p	validation	error	hf3fs, mla, missing-config, env-var
Failed to load config from {config_path}: {str(e)}	exception	error	hf3fs, config-file, json-parse, invalid-path
Missing required keys in config: {missing_keys}	validation	error	hf3fs, config-validation, missing-keys
LMCache is not installed. Please install it by running `pip	exception	critical	lmcache, importerror, missing-dependency, installation
MP mode requires --lmcache-config-file (the YAML supplies mp	validation	error	lmcache, mp-mode, missing-config, server-args
Failed to create shm file: {e}	exception	critical	shm, tmpfs, disk-full, hicache, docker
The installed Mooncake version does not support tenant_id in	exception	error	mooncake, version-mismatch, tenant, dependency
Failed to setup Mooncake Embedding Store: {ret_code}	exception	critical	mooncake, setup, transfer-engine, config
Invalid global_segment_size: missing number before 'gb'	validation	error	config, parsing, mooncake, validation
Config file path not set. Please set {envs.SGLANG_HICACHE_MO	exception	error	env-var, mooncake, config, missing-configuration
Failed to load config from {file_path}: {str(e)}	exception	error	config, json, file-io, mooncake
Either master_server_address or client_server_address is req	validation	error	config, mooncake, validation, missing-field
Either the environment variable 'MOONCAKE_MASTER' or 'MOONCA	validation	error	env-var, mooncake, missing-configuration
Either master_server_address or client_server_address is req	validation	error	config, mooncake, validation, missing-field
Please install mooncake by following the instructions at htt	exception	error	mooncake, import, missing-dependency, installation
Mooncake store is not initialized.	exception	error	mooncake, initialization, state, runtime-check
Failed to register buffer to Mooncake Store, error code: {re	exception	critical	mooncake, rdma, memory-registration, transfer-engine
MooncakeStore with standalone_storage=True requires Mooncake	exception	error	mooncake, allocator, standalone-storage, config
The installed Mooncake version does not support tenant_id in	exception	error	mooncake, version-mismatch, tenant, dependency
Failed to setup Mooncake store, error code: {ret_code}	exception	critical	mooncake, setup, transfer-engine, network
L3 cleaner low_watermark must be lower than high_watermark (	validation	error
Failed to register buffer to SiMM	exception	critical	rdma, memory-registration, hierarchical-cache, sim
SiMM Register Buffer Error.	exception	error	rdma, type-error, sim, memory-registration
mori.umbp is not available. Build mori with BUILD_UMBP=ON or	exception	critical	import-error, umbp, mori, native-dependency
UMBPHostTensorAllocator only supports CPU host memory, got d	validation	error	umbp, device-mismatch, validation
UMBPHostMemAllocator.alloc({} bytes) failed (requested_backi	exception	critical	umbp, hugepages, numa, out-of-memory
extra_config[{!r}] must be a boolean-like value (true/false,	validation	error	umbp, config-validation, boolean
{} must not be None	validation	error	umbp, config-validation, missing-value
{} must not be empty	validation	error	umbp, config-validation, empty-list
{} has {} entries, but rank_index={}	validation	error	umbp, config-validation, rank-mismatch
extra_config['ssd_io_backend'] must be one of: posix, io_uri	validation	error	umbp, config-validation, io-backend
extra_config['ssd_durability_mode'] must be one of: strict,	validation	error	umbp, config-validation, durability
extra_config['ssd_backend'] must be one of: file, spdk, spdk	validation	error	umbp, config-validation, ssd
extra_config['spdk_passthrough'] must be a dict of spdk_* fi	validation	error	umbp, spdk, config-validation
spdk_passthrough: unknown SSD config field {!r} (must be an	validation	error	umbp, spdk, config-validation
Unregistered UMBP hybrid pool: {}	validation	error	umbp, hybrid-pool, registration
Layer {} is not backed by an MLA KV pool	exception	critical	swa, mla, type-mismatch, memory-pool
SWA Radix tree sanity check failed, ping @hanming-lu: {e}	exception	critical	swa, radix-cache, internal-bug, sanity-check
evictable_size() is not implemented; use full_evictable_size	exception	error	swa, radix-cache, not-implemented, api-misuse
protected_size() is not implemented; use full_protected_size	exception	error	swa, radix-cache, not-implemented, api-misuse
CUDA error {int(result[0])}({_cudaGetErrorString(result[0])}	panic	critical	cuda, cuda-graph, gpu, driver
{source} must be a mapping or expose to_dict(), got {type(va	validation	error	quantization, type-validation, checkpoint, metadata
Download failed for {model_name_or_path} after {max_retries}	panic	critical	network, download, huggingface, retry-exhausted, weights
Downloaded model files are still corrupted for {model_name_o	panic	critical	download, corruption, huggingface, validation, weights
gguf package does not provide the DeepSeek name map	panic	error	gguf, deepseek, version-mismatch, dependency
DeepSeek-V4 GGUF mapping collision: {other!r} and {tensor_na	panic	error	gguf, deepseek, name-collision, weight-mapping
No DeepSeek-V4 checkpoint mapping for {len(missing)} GGUF te	panic	error	gguf, deepseek, unmapped-tensors, weight-mapping
GGUF BF16 payload does not have a byte-pair layout	validation	error	gguf, bf16, data-layout, weights
invalid compressor checkpoint name: {checkpoint_name}	validation	error	gguf, deepseek, compressor, name-mapping
quantized tensor maps to a non-weight parameter: {tensor.nam	validation	error	gguf, deepseek, quantization, weight-mapping
invalid Kimi-K3 attention-residual target {target_index}	validation	error	kimi-k3, gguf, weight-conversion, validation
Kimi-K3 GGUF ssm_a must contain finite floating values	validation	error	kimi-k3, gguf, kda, mamba, nan
Kimi-K3 GGUF ssm_a must contain only -exp(A_log) values	validation	error	kimi-k3, gguf, ssm, a-log
Kimi-K3 manifest format is unsupported	validation	error	kimi-k3, gguf, manifest, version-mismatch
Kimi-K3 manifest is incomplete	validation	error	kimi-k3, gguf, manifest, incomplete-conversion
ModelOpt is not available. Please install modelopt.	exception	error	modelopt, quantization, import-error, dependency, sglang
Failed to set up ModelOpt quantization: {e}	exception	error	modelopt, quantization, wrapper-exception, chained-exception, sglang
ModelOpt export functionality is not available. Please ensur	exception	error	modelopt, export, version-mismatch, import-error, sglang
Invalid quantization choice: '{quant_choice_str}'. Available	exception	error	modelopt, quantization, invalid-argument, config-validation, sglang
ModelOpt quantization config '{quant_cfg_name}' not found. P	exception	error	modelopt, version-mismatch, attributeerror, quantization, sglang
Runai Model Streamer Loader does not support ModelOpt quanti	exception	error	runai-streamer, modelopt, not-implemented, unsupported-combination, sglang
Failed to import sglang.private.private_model_loader	exception	error	load-format, private-module, import-error, sglang, internal-build
Post-load processing produced a meta tensor	exception	critical	meta-tensor, post-load, uninitialized-weights, model-loader, sglang
AfmoeConfig must define `num_experts`.	exception	critical	moe, config-validation, afmoe, model-loading
Tensor parallel size {self.tp_size} is greater than the numb	exception	critical	moe, tensor-parallel, afmoe, startup
Unsupported activation: {hidden_act}. Only xIELU is supporte	exception	critical	activation, config-validation, apertus
Self attention has no KV cache scaling factor attribute!	exception	error	fp8, kv-cache, quantization, apertus
Unsupported activation: {hidden_act}. Arcee model in SGLang	exception	critical	activation, config-validation, arcee
Self attention has no KV cache scaling factor attribute!	exception	error	fp8, kv-cache, quantization, arcee
Unsupported activation: {hidden_act}. Only silu is supported	validation	critical	activation, config-validation, baichuan
Unsupported activation. Only silu is supported for now.	validation	critical	activation, config-validation, bailing-moe
Unsupported activation: {config.hidden_act}. Only silu is su	validation	critical	activation, config-validation, bailing-moe, moe
num_nextn_predict_layers is not in the config	validation	critical	mtp, speculative-decoding, bailing-moe, weight-loading
Unsupported attention type: {config.attention_type}	validation	critical	attention, config-validation, bailing, hybrid-model
num nextn_predict_layers is not in the config	validation	critical	mtp, speculative-decoding, bailing, weight-loading
num nextn_predict_layers is not in the config	validation	critical	mtp, speculative-decoding, bailing, weight-loading
Unsupported attention type: {config.attention_type}	validation	critical	attention, config-validation, bailing, hybrid-model
num_fused_shared_experts > 1 ({self.num_fused_shared_experts	validation	critical	moe, shared-experts, cuda, platform-limit, bailing
DSPARK requires explicit layer_ids for aux hidden capture.	validation	error	dspark, hidden-states, capture, bailing
Only 'absolute' position_embedding_type is supported	validation	critical	bert, embeddings, config-validation
Either input_ids or inputs_embeds must be provided.	validation	error	clip, input-validation, embeddings
Sequence length {seq_length} exceeds the maximum {max_positi	validation	error	clip, sequence-length, input-validation
The original encoder only has {num_hidden_layers} layers, bu	validation	error	clip, vision-encoder, config-validation
Incorrect type of pixel values. Got type: {type(pixel_values	validation	error	multimodal, vision, type-validation, deepseek-ocr
Incorrect type of image sizes. Got type: {type(images_spatia	validation	error	multimodal, vision, type-validation, deepseek-ocr
Incorrect type of image crop. Got type: {type(images_crop)}	validation	error	multimodal, vision, type-validation, deepseek-ocr
Some weights are not initialized from checkpoints: {unloaded	error_code	critical	checkpoint-loading, weights, model-init, deepseek-ocr
Image aspect ratio must be smaller than 200	validation	error	dots3, vision, image-preprocessing, aspect-ratio, multimodal
Expected a PIL image, got {type(image)}	validation	error	dots3, type-error, pil, multimodal, input-validation
Unsupported activation: {hidden_act}. Only silu is supported	validation	error	exaone, activation, model-config, unsupported-operation
forward_deepep branch not implemented yet	exception	error
Inkling relative attention requires the vendored FA4 CUTE in	exception	critical	sglang, import-error, flashattention, cute, cuda, inkling
n must be a positive integer	validation	error	sglang, value-error, config-validation, model-loading, inkling
patch_size must be greater than 1, otherwise this doesn't ma	validation	error	sglang, value-error, vision-encoder, config-validation, inkling
InklingBatchDenseMLPWithLoRA is ineligible: {joined problems	validation	error	sglang, lora, triton-backend, bf16, eligibility-check, inkling
Inkling shared-sink LoRA requires four 4D MoE buffers	validation	error	sglang, lora, shape-validation, moe, inkling
Inkling shared-sink LoRA outer factors must have expert dime	validation	error	sglang, lora, shape-validation, moe, shared-experts, inkling
Inkling shared-sink gate-up A and down B must use the same e	validation	error	sglang, lora, shape-validation, consistency-check, moe, inkling
Inkling shared-sink LoRA expert count does not match	validation	error	sglang, lora, shape-validation, moe, shared-experts, inkling
Inkling shared-sink LoRA rank dimensions do not match	validation	error	sglang, lora, rank-validation, shape-validation, inkling
Shared-sink LoRA pool shape changed after initialization: ga	exception	error	sglang, lora, pool-allocation, shape-validation, runtime-error, inkling
Shared-sink LoRA slot out of range: {sorted(slot_ids)}	exception	error	lora, index-error, slot-management, sglang
Shared-sink down LoRA-A width must be divisible by {self.n_s	validation	error	lora, shape-mismatch, weight-loading, moe
Shared-sink gate/up LoRA-B height must be divisible by {self	validation	error	lora, shape-mismatch, weight-loading, moe
Unsupported activation: {activation_type}	exception	error	activation, moe, config-mismatch, sglang
RMSNorm expected hidden size {self.hidden_size}, got {origin	exception	error	rmsnorm, shape-mismatch, model-architecture
Inkling only supports group size 16 for NVFP4	exception	error	quantization, nvfp4, config-mismatch, inkling
InklingNvfp4MoEMethod is the dense shared-expert method; rou	exception	error	quantization, nvfp4, moe, not-implemented, inkling
Cannot deinterleave odd gate/up dimension {dim}: {tuple(weig	exception	error	weight-loading, layout-conversion, shape-mismatch, interns2
Unsupported activation: {hidden_act}. Only silu is supported	exception	error	activation, config-mismatch, internlm2, weight-loading
{config.text_config.architectures[0]} is not implemented.	exception	error	multimodal, architecture-not-supported, config-mismatch, interns1
Mobius fused gate/up destination is missing: {parameter_name	exception	error	weight-loading, key-mapping, mobius, interns2
Expected {num_experts} experts in {name}, got {loaded_weight	exception	error	weight-loading, expert-count-mismatch, mobius, moe
Intern-S2-Mobius requires at least one physical routed-exper	exception	critical	model-config, moe, interns2-mobius
num_attention_heads must be divisible by attention TP	exception	error	tensor-parallel, attention, launch-config
num_key_value_heads must be divisible by attention TP	exception	error	gqa, kv-heads, tensor-parallel
attention TP must be divisible by num_key_value_heads	exception	error	gqa, kv-heads, tensor-parallel
Intern-S2-Mobius baseline does not support pipeline parallel	exception	error	pipeline-parallel, unsupported-feature, launch-config
Unsupported Mobius layer type: {checkpoint_type}	exception	error	layer-type, checkpoint-compat, model-config
Intern-S2-Mobius baseline does not support PP tensors	exception	error	pipeline-parallel, runtime-misuse
Load Intern-S2-Mobius through its conditional-generation wra	exception	error	weight-loading, api-misuse
You have to specify pixel_values or pixel_embeds	exception	error	multimodal, vision, missing-input
wrong pixel_values size: {pixel_values.shape}	exception	error	multimodal, vision, input-shape
language_model does not support get_embed_and_head().	exception	error	speculative-decoding, kimi, attribute-error, model-loading
language_model does not support set_embed_and_head().	exception	error	speculative-decoding, kimi, attribute-error, model-loading
Eagle3 MLA layer requires q_lora_rank in the draft config	exception	error	eagle3, speculative-decoding, mla, config-validation, kimi
EAGLE3 currently only supports 1 layer	exception	error	eagle3, speculative-decoding, config-validation, kimi
Eagle3 MLA draft post_load_weights only supports float dtype	exception	error	eagle3, weight-loading, dtype, quantization, speculative-decoding
Unsupported activation: {hidden_act}	exception	error	activation, config-validation, model-loading, kimi
DSPARK aux hidden capture requires PP=1.	exception	error	dspark, pipeline-parallel, speculative-decoding, kimi
DSPARK requires explicit layer_ids for aux hidden capture.	exception	error	dspark, speculative-decoding, argument-validation, kimi
Kimi-K3 MLA K projection must remain GGUF Q4_0	exception	error	gguf, quantization, weight-loading, kimi, mla
Kimi-K3 MLA V projection must remain GGUF Q2_K	exception	error	gguf, quantization, weight-loading, kimi, mla
get_input_embeddings() is not available in encoder-only mode	exception	error	kimi-k3, encoder-only, attribute-error, embeddings
lm_head is not available in encoder-only mode	exception	error	kimi-k3, encoder-only, lm-head, attribute-error
DSPARK layer capture is not available in encoder-only mode	exception	error	kimi-k3, encoder-only, dspark, attribute-error
Kimi-K3 encoder mode supports image input only	exception	error	kimi-k3, multimodal, modality-mismatch, validation
Kimi-K3 encoder preprocessing needs an image processor	exception	error	kimi-k3, image-processor, missing-argument
Kimi-K3 expects one vision grid per MultimodalDataItem; spli	exception	error	kimi-k3, multimodal, grid-thws, data-shape
Kimi-K3 cannot mix local preprocessed and deferred images	exception	error	kimi-k3, multimodal, preprocessing-mismatch
Kimi-K3 image feature must be a torch.Tensor, got {type(item	exception	error	kimi-k3, type-error, feature-tensor, multimodal
Kimi-K3 deferred GPU preprocessing produced wrong grids	exception	error	kimi-k3, gpu-preprocessing, grid-mismatch, sanity-check
Unsupported Kimi-K3 deferred preprocessing backend: {backend	exception	error	kimi-k3, preprocessing-backend, unsupported-value
Kimi-K3 deferred feature length does not match image grids	exception	error	kimi-k3, multimodal, shape-mismatch, vision
Not support pos_emb_type: {pos_emb_type}	exception	error	kimi-k3, vision, config-validation, not-implemented
Not support norm_type: {norm_type}	exception	error	kimi-k3, vision, config-validation, not-implemented
Unsupported Kimi-K3 vision attention backend: {attention_bac	exception	error	kimi-k3, attention-backend, env-var, startup-validation
Not support merge_type: {self.merge_type}	exception	error	kimi-k3, vision, config-validation, not-implemented
Not support activation_func: {activation_func}	exception	error	kimi-k3, vision, activation, config-validation
KDA num_heads ({num_heads}) must be divisible by shard tp_si	exception	error	kimi-linear, tensor-parallel, divisibility, startup-validation
Unsupported activation: {config.hidden_act}. Only silu is su	exception	error	kimi-linear, activation, config-validation
DSPARK aux hidden capture requires PP=1.	exception	error	kimi-linear, dspark, pipeline-parallel, not-implemented
DSPARK requires explicit layer_ids for aux hidden capture.	exception	error	kimi-linear, dspark, argument-validation
Unsupported activation: {hidden_act}. Only silu is supported	exception	error	laguna, activation, config-validation
TP size {self.tp_size} > num_experts {config.num_experts}.	exception	error	laguna, moe, tensor-parallel, startup-validation
Checkpoint provides gate weight {name!r} but the model built	exception	critical	laguna, weight-loading, config-mismatch, gating
{len(missing)} routed-expert tensors were not loaded (sample	exception	critical	laguna, moe, weight-loading, missing-weights
DFLASH requires explicit layer_ids for aux hidden capture.	exception	error	laguna, dflash, argument-validation
Tensor parallel size {self.tp_size} is greater than the numb	exception	error	lfm2, moe, tensor-parallel, startup-validation
Expected a 3D packed tensor for {name}, got {loaded_weight.d	exception	critical	lfm2, moe, weight-loading, tensor-shape
Invalid gate_up_proj shape for {name}: {tuple(loaded_weight.	exception	critical	lfm2, moe, weight-loading, shape-validation
Missing rope_parameters[{layer_type}] for Mellum layer {laye	exception	critical	mellum, config-validation, rope, model-loading
Missing config.sliding_window for Mellum sliding_attention l	exception	critical	mellum, sliding-window, config-validation
Expected len(mlp_layer_types) == num_hidden_layers, got {len	exception	critical	mellum, config-validation, moe, layer-config
Unsupported mlp_layer_types[{lid}]={mlp_type}; expected 'spa	exception	critical	mellum, config-validation, moe
Sparse MLP requested but num_experts <= 0 in Mellum config	exception	critical	mellum, moe, config-validation, experts
Unexpected arguments: `**rope_kwargs` and `config` are mutua	exception	error	mimo-audio, rope, api-misuse, argument-validation
Feature size mismatch: {features.size(0)} vs {lengths.sum().	exception	error	mimo-audio, tensor-shape, batching, audio
Invalid projection layers: {config.projection_layers}	exception	critical	mimo-audio, config-validation, audio-encoder, version-mismatch
No model weights found in {path} (expected model.safetensors	exception	critical	mimo-audio, model-loading, file-not-found, huggingface, weights
MiMoV2 fused qkv_proj checkpoint is TP={expected_fused_tp_si	exception	critical	mimo-v2, tensor-parallel, weight-loading, checkpoint-layout
qkv_proj scale_inv {name}: shape mismatch {tuple(loaded_weig	exception	error	quantization, tensor-parallel, weight-loading, mimo
qkv_proj weight {name}: unexpected shape {tuple(loaded_weigh	exception	error	weight-loading, shape-mismatch, checkpoint, tensor-parallel
qkv_proj weight {name}: unexpected shape {tuple(loaded_weigh	exception	error	weight-loading, shape-mismatch, gqa, tensor-parallel
Cannot resolve deferred scale_inv {scale_name}: weight {weig	exception	error	quantization, weight-loading, naming, mimo
Unsupported activation: {hidden_act}. Only silu is supported	exception	error	config-validation, activation, mimo
Tensor parallel size {self.tp_size} is greater than the numb	exception	error	moe, tensor-parallel, expert-parallel, launch-config
Unsupported activation: {config.hidden_act}. Only silu is su	exception	error	config-validation, activation, moe, mimo
Self attention has no KV cache scaling factor attribute!	exception	error	kv-cache, fp8, quantization, attention-backend
forward() is not supported in encoder_only mode. Use get_aud	exception	error	asr, encoder-only, api-misuse, multimodal
No model architectures are specified	exception	error	config-validation, mindspore, model-loading
Invalid qk_norm_type: {self.qk_norm_type}	validation	error
projection_cls = {projection_cls}, not implemented	validation	error	phi4, multimodal, projection, not-implemented, config-validation
audio_projection_mode = {audio_projection_mode} not implemen	validation	error	phi4, audio, runtime-dispatch, value-error
Unsupported activation type {self.glu_act}	validation	error	phi4, activation, glu, config-validation
T5 attention bias with bucketed positions is not yet tested	validation	error	phi4, t5, attention-bias, not-implemented
No striding allowed for non-symmetric convolutions!	validation	error	convolution, causal, stride, padding, phi4
Invalid padding param: {padding}!	validation	error	convolution, causal, padding, type-validation
Sampling factor should be a multiply of 2!	validation	error	subsampling, convolution, audio, config-validation, phi4
subsampling_conv_chunking_factor should be -1, 1, or a power	validation	error	subsampling, chunking, audio, config-validation, phi4
DFLASH requires explicit layer ids for aux hidden capture.	validation	error	dflash, speculative-decoding, hidden-states, qwen3
Tensor parallel size {self.tp_size} is greater than the numb	validation	error	moe, tensor-parallel, expert-parallel, qwen3
DFLASH requires explicit layer_ids for aux hidden capture.	validation	error	dflash, speculative-decoding, qwen3-moe
Qwen3-Next shared expert fusion currently supports exactly o	validation	error	qwen3-next, moe, shared-expert, fusion
DFLASH requires explicit layer_ids for aux hidden capture.	validation	error	dflash, qwen3-next, speculative-decoding
Qwen3-Next MTP shared expert fusion currently supports exact	validation	error	qwen3-next, mtp, shared-expert, speculative-decoding
SinusoidsPositionEmbedding needs even channels input	validation	error	qwen3-omni, embedding, shape-validation
Explicit vision TP cannot be combined with data parallel	validation	error	qwen3-vl, vision, data-parallel, tensor-parallel
Vision tp_size and tp_rank must be set together	validation	error	qwen3-vl, vision, tensor-parallel, argument-validation
DFLASH requires explicit layer_ids for aux hidden capture.	validation	error	dflash, qwen3-vl, speculative-decoding
Unknown forward method: {forward_method}	validation	error	sarvam-moe, attention-backend, config
Tensor parallel size {self.tp_size} > num_experts {config.nu	validation	error	sdar-moe, tensor-parallel, moe, launch-config
The original encoder only has {num_hidden_layers} layers, bu	validation	error	siglip, vision-encoder, layer-override
Packed pixel_values token count does not match spatial_shape	validation	error	siglip2, multimodal, preprocessing, shape-mismatch
embed_dim must be divisible by num_heads (got `embed_dim`: {	validation	error	siglip2, vision-config, divisibility
`max_possible_layers` must be provided alongside `select_lay	validation	error	siglip2, api-misuse, layer-selection
Expected encoder_outputs to be a list when select_layers is	validation	error	siglip2, api-misuse, layer-selection
The original encoder only has {num_hidden_layers} layers, bu	validation	error	siglip2, vision-encoder, layer-override
Unsupported activation: {hidden_act}. Only silu is supported	validation	error	solar, activation, config
Self attention has no KV cache scaling factor attribute!	error_code	error	solar, fp8, kv-cache-scales, version-mismatch
Unsupported Spark2_5 layer_type: {layer_type}	validation	error	spark2-5, layer-type, config
hidden_size must be divisible by num_heads (got `hidden_size	validation	error	stablelm, tensor-parallel, divisibility
Unsupported activation: {hidden_act}. Only silu is supported	validation	error	step3-vl, activation, config
Tensor parallel size {self.tp_size} is greater than the numb	validation	error	step3-vl, tensor-parallel, moe
DeepEP MoE is not supported yet in Step3 model.	validation	error	step3-vl, deepep, moe-backend
Step3-VL image item is missing num_patches.	validation	error	step3-vl, multimodal, metadata
Step3-VL image item has num_patches > 0 but no patch_pixel_v	validation	error	step3-vl, multimodal, patches
use_rope2d must be True	validation	error	step3-vl, vision-config, rope
Step3-VL image item is missing num_patches.	validation	error	step3-vl-10b, multimodal, metadata
Step3-VL image item has num_patches > 0 but no patch_pixel_v	validation	error	step3-vl-10b, multimodal, patches
Weight {name} not found in params_dict	validation	critical	checkpoint-loading, vision-tower, weight-mapping
Tensor parallel size {self.tp_size} is greater than the numb	validation	critical	tensor-parallel, moe, config-validation
Only 1 nextn layer is supported for Step3p5 checkpoints.	validation	error	speculative-decoding, mtp, checkpoint-loading
Some parameters like {param_name_example} are not in the che	validation	critical	checkpoint-loading, mtp, random-init-guard
Weight {name} not found in params_dict	validation	critical	checkpoint-loading, vision-tower
Unsupported activation: {hidden_act}. Only silu is supported	validation	error	activation-function, config-validation, llama
Unsupported parallel style type {type(style)}, expected str	validation	error	type-validation, tensor-parallel, transformers-backend
Unsupported TP style '{style}' for Transformers backend.	validation	error	tensor-parallel, transformers-backend, tp-plan
Model {model_cls} does not support custom attention backends	validation	error	transformers-backend, attention-backend, compatibility
{type(self.model)} does not support tensor parallel yet!	validation	critical	tensor-parallel, transformers-backend, tp-plan
{type(self.model)} does not support pipeline parallel yet!	validation	critical	pipeline-parallel, transformers-backend
Pipeline parallel with multiple ModuleList blocks is not sup	validation	error	pipeline-parallel, transformers-backend
Could not find ModuleList in {type(self.model)}.	validation	error	pipeline-parallel, transformers-backend
No encoder method found for modality '{modality_name}'	validation	error	multimodal, transformers-backend, encoder-discovery
Empty multimodal encoder output.	validation	error	multimodal, empty-batch, encoder-output
Only 2D tile_tag is supported currently, got: {self.tile_tag	validation	error	config-validation, ocr, vision
Incorrect type of pixel values. Got type: {type(pixel_values	validation	error	type-validation, multimodal, pixel-values
Incorrect type of image sizes. Got type: {type(images_spatia	validation	error	type-validation, multimodal, spatial-crop
Incorrect type of image crop. Got type: {type(images_crop)}	validation	error	type-validation, multimodal, image-crop
Kimi-K3 image processor is missing deferred-preprocessing co	validation	error	multimodal, kimi-k3, config-validation, preprocessing
Unsupported Kimi-K3 encoder media item: {image}	validation	error	multimodal, kimi-k3, input-validation
Unsupported Kimi-K3 image channel count: {channels}	validation	error	multimodal, image-processing, numpy, kimi-k3
capacity must be positive, got {capacity}	validation	error	cuda-graph, kimi-k3, config-validation, vision-tower
min_hits must be positive, got {min_hits}	validation	error	cuda-graph, kimi-k3, config-validation
max_seqlen must be positive, got {max_seqlen}	validation	error	cuda-graph, kimi-k3, config-validation
mm_content_hashes has {len(content_hashes)} entries for {med	validation	error	multimodal, artifact-cache, kimi-k3, input-validation
content hash mismatch for media_data[{index}]: expected {cal	validation	error	multimodal, artifact-cache, content-hash, cache-invalidation
Error while loading data {data_str}: {e}	exception	error	multimodal, loading, http, input-validation
{modality.name} must be a list or None, got {type(data_list)	validation	error	multimodal, type-validation, input-validation
For {modality}, when providing a 'processor_output' or 'prec	validation	error	multimodal, precomputed-embedding, input-validation
An exception occurred while loading {modality.name} data at	exception	error	multimodal, loading, input-validation
An exception occurred while loading multimodal data: {e}	exception	error	multimodal, prompt-template, placeholder-mismatch, legacy
prompt has {num_placeholders} image placeholder token(s) but	validation	error	multimodal, tokenization, placeholder-mismatch, input-validation
Unknown multimodal item type: {type(item)}	validation	error	multimodal, routing, input-validation
processor image placeholder count mismatch: processor={proce	validation	error	multimodal, processor-override, tokenization, placeholder-mismatch
seq must be positive, got {seq}	validation	error	multimodal, video, config-validation, valueerror
max_new_tokens must be non-negative, got {max_new_tokens}	validation	error	multimodal, sampling-params, valueerror
max_new_tokens must leave room for input: max_new_tokens={ma	validation	error	multimodal, sequence-length, valueerror
audio_cap must be non-negative, got {audio_cap}	validation	error	multimodal, audio, config-validation, valueerror
audio_sr must be positive, got {audio_sr}	validation	error	multimodal, audio, sample-rate, valueerror
k_mode must not be empty	validation	error	multimodal, video, config-validation, valueerror
Dots omni audio must be mono, got shape={tuple(waveform.shap	validation	error	multimodal, audio, mono, valueerror
Unsupported preprocessed video item: {item_type}	validation	error	multimodal, video, schema-mismatch, version-drift
Image placeholder count does not match image_data	validation	error	multimodal, placeholder-mismatch, images, valueerror
Audio placeholder count does not match audio_data	validation	error	multimodal, placeholder-mismatch, audio, valueerror
Dots note omni requires a text prompt for multimodal request	validation	error	multimodal, request-format, typeerror-contract, valueerror
Unsupported dots note omni video_config fields: {sorted fiel	validation	error	multimodal, video-config, unknown-fields, valueerror
Dots note omni video preprocessing requires one request's sa	validation	error	multimodal, sampling-params, type-mismatch, valueerror
Video placeholder count does not match video_data: {len(vide	validation	error	multimodal, video, placeholder-mismatch, valueerror
Dots note omni expanded prompt is too long: {len(input_ids)}	validation	error	multimodal, context-length, prompt-too-long, valueerror
bad metadata: dur={duration} h={original_height} w={original	validation	error	multimodal, video, corrupt-metadata, decoding
unsupported time format: {time_format!r}	validation	error	multimodal, config-validation, video-qa, dots-note-omni
unsupported audio interleave mode: {ai_k_mode!r}	validation	error	multimodal, config-validation, audio-interleave, dots-note-omni
encounter invalid h_bar: {h_bar}, w_bar: {w_bar}	validation	error	multimodal, image-preprocessing, ernie-4-5-vl, pixel-limits
nframes should in interval [{FRAME_FACTOR}, {total_frames}],	validation	error	multimodal, video-preprocessing, ernie-4-5-vl, frame-sampling
InklingMultimodalProcessor: required config field {where}.{n	validation	critical	multimodal, model-config, inkling, missing-config-field
InklingMultimodalProcessor: {n_img_ph} image placeholder tok	validation	error	multimodal, inkling, placeholder-mismatch, input-validation
InklingMultimodalProcessor: {n_aud_ph} audio placeholder tok	validation	error	multimodal, inkling, placeholder-mismatch, audio
InklingMultimodalProcessor v1 requires pre-rendered input_id	validation	error	multimodal, inkling, pre-tokenized-input, missing-tokenizer
[internvl] Cannot process raw images/videos with pre-tokeniz	validation	error	multimodal, internvl, pre-tokenized-input, dynamic-tiling
[internvl][qwen] image_data provided but no images parsed fr	validation	error	multimodal, internvl, qwen, placeholder-mismatch
[internvl][internlm2] image_data provided but no images pars	validation	error	multimodal, internvl, image-placeholders, prompt-validation
Unsupported grid type for kimi image tokens: {type(grid_thw)	validation	error	kimi, multimodal, type-validation, grid-metadata
Invalid grid metadata for kimi image tokens: {vals} (expecte	validation	error	kimi, multimodal, grid-metadata, validation
The number of image placeholders exceeds img_grid_thw entrie	validation	error	kimi, multimodal, placeholder-count, validation
The number of image placeholders does not match img_grid_thw	validation	error	kimi, multimodal, placeholder-count, validation
Expected {len(image_token_counts)} image placeholder token(s	validation	error	kimi, k25, multimodal, placeholder-count, tokenization
Kimi GPU preprocessing expects raw uint8 pixels, got {image.	validation	error	kimi, k25, multimodal, dtype-validation, preprocessing
Kimi image placeholders must map one-to-one to image data: e	validation	error	kimi, k25, multimodal, placeholder-count, validation
Kimi image placeholders must map one-to-one to image data: e	validation	error	kimi, k25, multimodal, loader, count-mismatch
Expected one original size for each K3 image.	validation	error	kimi, k3, multimodal, metadata-mismatch, validation
Expected {len(image_token_counts)} image placeholder token(s	validation	error	multimodal, kimi-k3, placeholder-mismatch, validation
Expected {len(image_token_counts)} image placeholder(s), fou	validation	error	multimodal, kimi-k3, cpu-fallback, string-split, validation
Kimi-K3 processor feature length does not match image grids:	validation	error	multimodal, kimi-k3, processor-mismatch, shape-mismatch
Expected one Kimi-K3 image span for each image	validation	error	multimodal, kimi-k3, deferred-preprocessing, span-mismatch
Kimi image placeholders must map one-to-one to image data: e	validation	error	multimodal, kimi-vl, image-input, validation
Invalid image data: {image_data}	validation	error	multimodal, llava, image-input, type-validation
Cannot find corresponding multimodal processor registered in	validation	error	multimodal, llava, model-loading, unsupported-model
Required `vision_config.model_type` is not found in hf_confi	validation	error	multimodal, config, model-loading, missing-field
audio must be a str, bytes, tuple, torch.Tensor, or np.ndarr	validation	error	multimodal, audio, type-validation, mimo
audio must be a tuple of (waveform-T, original_sr-int/float)	validation	error	multimodal, audio, tuple-validation, mimo
waveform must be a 1D tensor, but got {self.audio[0].ndim}D	validation	error	multimodal, audio, tensor-shape, mimo
original_sr must be a positive number, but got {self.audio[1	validation	error	multimodal, audio, sample-rate, mimo
audio must be a 2D tensor, but got {self.audio.ndim}D tensor	validation	error	multimodal, audio, tensor-shape, mimo
torchaudio is required for audio inputs; install torchaudio	error_code	error	multimodal, audio, missing-dependency, mimo
content must be a ImageInput, but got {type(self.content)}	validation	error
ffprobe not found; install ffmpeg	error_code	error	ffmpeg, subprocess, missing-dependency, multimodal, video
Unsupported video input type for EPD encoder: {type(video_da	validation	error	video, input-validation, type-error, multimodal
Unsupported modality for EPD preprocessing: {modality}	validation	error	modality, dispatch, input-validation, multimodal
Video sampling strategy not specified	validation	error	video, config, sampling, missing-value
No frames before start_time {start_time} in all_timestamps {	validation	error	video, timestamps, segment, out-of-range
video must be a tuple of (video_tensor, timestamps), but got	validation	error	video, contract, type-error, decoding
Error processing video at index {idx}: {e}	error_code	error	video, parallel, executor, error-chaining, batch
Unknown visual_type: {visual_type}	validation	error	visual, dispatch, input-validation, version-skew
absolute aspect ratio must be smaller than 200, got {max(hei	validation	error	image, aspect-ratio, resize, input-validation
Unsupported image type: {type(img)}. Expected torch.Tensor o	validation	error	image, type-error, transform, input-validation
Unrecognized image input, support local path, http url, base	validation	error	image, loading, base64, input-validation
{name} must be a dict-like config, got {type(config)}	validation	error	config, type-error, initialization
processor_config.{key} must be set for MiMo-V2	validation	error	config, missing-key, initialization, checkpoint
{name} placeholder/data mismatch: {placeholder_count} placeh	validation	error	validation, placeholders, prompt-template, multimodal
audio_sampling_rate must be set in processor_config or audio	validation	error	audio, config, sampling-rate, initialization
Video file is corrupted or cannot be decoded	http	error	video, decode, http-exception, corrupt-file
unsupported audio item: loaded={loaded_type}, raw={raw_type}	exception	error	audio, input-validation, type-error, multimodal
Multimodal data is corrupted or cannot be decoded: {e}	exception	error	multimodal, wrapper, decode, request
Request mixes standalone audio and video-with-audio; EPD mer	exception	error	audio, video, not-implemented, epd, multimodal
tokenizer missing required special token {name!r}; checkpoin	exception	critical	tokenizer, checkpoint-mismatch, vocab, asr, initialization
unsupported audio item: loaded={loaded_type}, raw={raw_type}	exception	error	multimodal, audio, type-validation, asr
Multimodal data is corrupted or cannot be decoded: {e}	exception	error	multimodal, audio, decode, asr, corrupt-file
Expected CHW image tensor, got shape {shape}	exception	error	multimodal, image-processing, tensor-shape, step3-vl
Expected CHW image tensor with 1 or 3 channels, got shape {s	exception	error	multimodal, image-processing, channels, step3-vl
Unsupported image type: {type}	exception	error	multimodal, image-processing, type-validation, step3-vl
The number of placeholders does not match the number of repl	exception	error	multimodal, prompt-template, placeholder-mismatch, step3-vl
Unknown image_mode '{mode}'. Supported: {supported}	exception	error	multimodal, ocr, invalid-argument, config-validation
image_mode='{mode}' is not supported with multiple images (g	exception	error	multimodal, ocr, multi-image, config-validation
Language '{language}' not recognized. Use full name (e.g., '	exception	error	whisper, audio, language-code, validation
Language '{language}' is not in this Whisper model's vocabul	exception	error	whisper, audio, tokenizer-vocab, model-version
Whisper expects exactly 1 audio input, got {len}	exception	error	whisper, audio, single-input-constraint
total_pool_size must be positive	validation	error	cuda-ipc, memory-pool, config-validation, multimodal-transport
tokenizer_worker_num must be positive	validation	error	cuda-ipc, memory-pool, config-validation, multimodal-transport
Input 'data' must be a torch.Tensor, but got {type}	exception	error	cuda-ipc, type-validation, torch-tensor, multimodal-transport
Cannot resolve the {transport_name} consumer rank before par	exception	critical	distributed, parallel-state, cuda-ipc, initialization-order
{transport_name} consumer rank {rank} is outside [0, {total_	error_code	critical	distributed, rank-validation, cuda-ipc, multimodal-transport
total_consumer_count must be positive	validation	error	cuda-ipc, config-validation, multimodal-transport
{self.transport_name} acknowledgements support one consumer	validation	error	cuda-ipc, acknowledgement, protocol-validation, multimodal-transport
memory_size must be positive	validation	error	memory-pool, config-validation, cuda-ipc, multimodal-transport
consumer_count must be positive	validation	error	memory-pool, config-validation, cuda-ipc, multimodal-transport
max_inflight_slices must be positive	validation	error	memory-pool, config-validation, cuda-ipc, multimodal-transport
recycle_interval must be positive	validation	error	multimodal, memory-pool, argument-validation
byte_tensor must be a sufficiently large contiguous uint8 te	validation	error	multimodal, cuda, tensor-validation, memory-pool
{transport_name} pool is too small after control metadata: p	validation	error	multimodal, memory-pool, capacity-planning
{self.transport_name} pool slot generation exhausted	error_code	error	multimodal, memory-pool, counter-overflow, long-running
Cannot release inactive {self.transport_name} pool lease (sl	error_code	error	multimodal, memory-pool, double-free, lease-lifecycle
{self.transport_name} requires a CUDA tensor	validation	error	multimodal, cuda, tensor-device
{self.transport_name} cannot transport an empty tensor	validation	error	multimodal, empty-tensor, validation
ViT CUDA graph does not support attention backend: {backend}	error_code	error	multimodal, vit, cuda-graph, attention-backend
deepstack_visual_indexes exists but deepstack_merger_list is	error_code	error	multimodal, vit, deepstack, cuda-graph, config-mismatch
Missing required field: sm_group_num	validation	error	pdmux, config, yaml, missing-field
sm_group_num must be >= 3	validation	error	pdmux, config, validation
manual_divisions must have {expected} entries, but got {len(	validation	error	pdmux, config, partitioning
Unsupported compute capability: {major}.{minor}	validation	error	pdmux, gpu-architecture, compatibility
No valid partitions found for total SMs {total_sms} with con	validation	error	pdmux, gpu, partitioning, capacity
Invalid stream index: {idx}	validation	error	pdmux, stream, index-validation
RayPrometheusMetric requires Ray to be installed. Install wi	exception	error	observability, ray, metrics, missing-dependency
Number of labels must match the number of tag keys. Expected	validation	error	observability, ray, metrics, labels
labels() cannot be called on an already-labeled metric.	validation	error	observability, ray, metrics, api-misuse
opentelemetry package is not installed!!! Please not enable	error_code	error	observability, tracing, opentelemetry, missing-dependency
initialize opentelemetry error:{e}. Please set correct otlp	error_code	error	observability, tracing, opentelemetry, otlp-endpoint
Unsupported OTLP protocol '{protocol}' configured. Supported	validation	error	opentelemetry, tracing, configuration, env-var
Invalid style: {self.sep_style}	validation	error	conversation, chat-template, parser
Found more '{modality_token}' placeholders in input prompt t	validation	error	multimodal, vision, prompt, validation
The messages should be a list of dict.	validation	error	chat-api, request-validation, messages
The system message should be a single text.	validation	error	chat-api, system-message, validation
The assistant's response should be a single text.	validation	error	chat-api, assistant-message, validation
Unknown role: {msg_role}	validation	error	chat-api, role-validation
Inkling thinking parts require role='assistant'	validation	error	inkling, reasoning, content-parts, parser
assistant reasoning_content must be a string for Inkling ren	validation	error	inkling, reasoning, type-error
assistant message cannot mix reasoning_content with ordered	validation	error	inkling, reasoning, conflicting-fields
unsupported Inkling render part kind: {kind!r}	validation	error	inkling, internal, parser
message content must be a string or a sequence of parts	validation	error	inkling, type-error, content
content part must be mapping, got {type(part).__name__}	validation	error	inkling, type-error, content-parts
Inkling thinking part payload must be a string	validation	error	inkling, thinking, type-error
unsupported content part type: {ptype!r}	validation	error	inkling, content-parts, unsupported-type
Inkling reasoning_effort must be a number	validation	error	inkling, reasoning-effort, type-error
Inkling reasoning_effort must be finite and in [0.0, 0.99]	validation	error	inkling, reasoning-effort, range-validation
unsupported Inkling message role {role!r}; expected one of {	validation	error	inkling, role-validation
expected mapping, got {type(value).__name__}	validation	error	inkling, tool-calls, type-error
tool call function name must be a string	validation	error	inkling, tool-calls, type-validation, message-rendering
tool call function arguments must decode to an object	validation	error	inkling, tool-calls, json-decode, arguments
unknown Inkling special token: {token!r}	validation	error	inkling, special-tokens, tokenizer, key-error
text must be str, got {type(text).__name__}	validation	error	inkling, tokenizer, type-error, encode
Invalid content format: {content_format}	validation	error	jinja, chat-template, content-format, validation
Model type must be specified	validation	error	reasoning-parser, model-type, constructor
Unsupported model type: {model_type}	validation	error	reasoning-parser, model-type, unsupported-model
Chat template {chat_template_arg} is not a built-in template	error_code	error	chat-template, template-manager, file-not-found, server-args
Completion template {completion_template_arg} is not a built	error_code	error	completion-template, template-manager, file-not-found, server-args
Unknown separator style: {template['sep_style']}	validation	error	chat-template, json-template, sep-style, enum-key
config namespace {path!r} has no leaf/subgroup {name!r}	validation	error
Global server args is not set yet!	exception	error
config not published; cannot read a config leaf	validation	error	config, runtime-context, initialization-order, sglang
{name!r} is not a config leaf (no NS namespace)	validation	error	config, typo, field-mapping, sglang
subgroup {seg!r} missing under {path!r}	validation	error	config, namespace, projection-mismatch, sglang
override_server_args: unknown ServerArgs field(s): {sorted(u	validation	error	config, override, typo, sglang
SGLANG_ROLE_NAMESPACES={value!r} is not one of off / record	validation	error	environment-variable, config-validation, typo, sglang
publish role {role!r} has no ROLE_NAMESPACE_SETS entry; decl	validation	error	config, role-based-access, enforcement, sglang
config not published for role {role!r}: {detail}. The proces	exception	error	runtime-context, initialization-order, roles, sglang
invalid Rust extension build mode {mode!r}; expected auto, n	validation	error	rust-extension, build-mode, env-var, validation, sglang
cannot force-build {python_module} after it has been importe	exception	error	rust-extension, import-cache, force-rebuild, python, sglang
{crate.python_module} is not bundled or cached, and Rust ext	exception	error	rust-extension, missing-module, build-cache, env-var, sglang
Rust sources under {crate.workspace} changed during the buil	exception	error	rust, build-cache, concurrency, cargo
Rust workspace for {python_module} was not found at {workspa	exception	error	rust, file-not-found, discovery, cargo
{lockfile} is required for reproducible `cargo build --locke	exception	error	rust, cargo, lockfile, reproducibility
{manifest} declares python-module {python_module!r} but must	validation	error	rust, cargo, manifest, validation
no Cargo package under {workspace} declares `[package.metada	exception	error	rust, discovery, module-not-found, metadata
multiple Cargo packages under {workspace} declare python mod	validation	error	rust, discovery, ambiguity, metadata
failed to query the Rust toolchain with `{command} {' '.join	exception	error	rust, toolchain, environment, subprocess
Python did not report an EXT_SUFFIX for native extensions	exception	error	python, sysconfig, environment, native-extension
failed to build {crate.python_module} with Cargo	exception	error	rust, cargo, build-failure, compile
Cargo completed but did not produce the expected artifact {a	exception	error	rust, cargo, artifact, build-output
could not create an import spec for {module_name} at {path}	exception	error	python, import, importlib, native-extension
auxiliary output does not support pipeline-parallel transpor	exception	error	pipeline-parallel, sampling, type-validation, distributed
auxiliary PP output must contain at least one tensor	exception	error	pipeline-parallel, sampling, empty-payload, validation
auxiliary PP tensor names must be non-empty strings	exception	error	pipeline-parallel, sampling, validation, keys
auxiliary PP output {name!r} is not a tensor	exception	error	pipeline-parallel, sampling, torch, validation
duplicate auxiliary PP tensor {name!r}	exception	error	pipeline-parallel, sampling, duplicate-keys, validation
received auxiliary PP output without a sampling observer	exception	error	pipeline-parallel, sampling, observer, configuration
sampling observer does not support pipeline-parallel transpo	exception	error	pipeline-parallel, sampling, observer, type-validation
received a non-tensor auxiliary PP output	exception	error	pipeline-parallel, sampling, torch, validation
sampling observer did not reconstruct its PP output	exception	error	pipeline-parallel, sampling, observer, reconstruction
beam_width must be at least 1, got {self.beam_width}.	validation	error	sampling-params, beam-search, validation, sglang
temperature must be a non-negative finite number, got {self.	validation	error	sampling-params, temperature, nan, validation, sglang
top_p must be in (0, 1], got {self.top_p}.	validation	error	sampling-params, top-p, nucleus-sampling, validation, sglang
min_p must be in [0, 1], got {self.min_p}.	validation	error	sampling-params, min-p, validation, sglang
top_k must be -1 (disable) or at least 1, got {self.top_k}.	validation	error	sampling-params, top-k, validation, sglang
frequency_penalty must be in [-2, 2], got {self.frequency_pe	validation	error	sampling-params, frequency-penalty, validation, sglang
presence_penalty must be in [-2, 2], got {self.presence_pena	validation	error	sampling-params, presence-penalty, validation, sglang
repetition_penalty must be in (0, 2] (1.0 = no penalty), got	validation	error	sampling-params, repetition-penalty, validation, sglang
min_new_tokens must be in [0, max_new_tokens], got {self.min	validation	error	sampling-params, min-new-tokens, validation, sglang
max_new_tokens must be at least 0, got {self.max_new_tokens}	validation	error	sampling-params, validation, max-new-tokens, sglang
min_new_tokens must be in [0, max_new_tokens({self.max_new_t	validation	error	sampling-params, validation, min-new-tokens, sglang
logit_bias must has keys in [0, {vocab_size - 1}], got {toke	validation	error	sampling-params, logit-bias, vocab-size, validation, sglang
stop={stop_strs!r} is unavailable when skip_tokenizer_init=T	validation	error	skip-tokenizer-init, stop-strings, sampling-params, sglang
stop_regex={stop_regex_strs!r} is unavailable when skip_toke	validation	error	skip-tokenizer-init, stop-regex, sampling-params, sglang
min_new_tokens={min_new_tokens} is unavailable when skip_tok	validation	error	skip-tokenizer-init, min-new-tokens, sampling-params, sglang
resolution already failed on this ServerArgs; the handlers t	exception	error	server-args, resolution-pipeline, state-corruption, sglang
return_hidden_states_mode must be one of: None, 'last', or '	validation	error	server-args, hidden-states, cli-validation, sglang
Decode context parallel size (--dcp-size / --decode-context-	validation	error	server-args, dcp, parallelism, validation, sglang
--dcp-comm-backend {cfg.dcp_comm_backend} only affects the d	validation	error	server-args, dcp, comm-backend, parallelism, sglang
--dcp-comm-backend fi_a2a delegates the exchange to FlashInf	validation	error	sglang, distributed, dcp, flashinfer, mnnvl, hardware-requirement
--dcp-replicate-q-proj requires --dcp-size > 1.	validation	error	sglang, distributed, dcp, argument-validation
--dcp-replicate-q-proj only applies to the a2a/fi_a2a DCP co	validation	error	sglang, distributed, dcp, argument-validation, incompatible-flags
Invalid disaggregation_mode={cfg.disaggregation_mode!r}	validation	error	sglang, disaggregation, pd-disaggregation, argument-validation
--ssl-keyfile requires --ssl-certfile to be specified as wel	validation	error	sglang, ssl, tls, server-config, argument-validation
--ssl-certfile requires --ssl-keyfile to be specified as wel	validation	error	sglang, ssl, tls, server-config, argument-validation
--ssl-ca-certs has no effect without --ssl-certfile and --ss	validation	error	sglang, ssl, tls, argument-validation
--ssl-keyfile-password has no effect without --ssl-certfile	validation	error	sglang, ssl, tls, secrets, argument-validation
SSL key file not found: '{cfg.ssl_keyfile}'. Please check th	validation	error	sglang, ssl, tls, file-not-found, deployment
SSL certificate file not found: '{cfg.ssl_certfile}'. Please	validation	error	sglang, ssl, tls, file-not-found, deployment
SSL CA certificates file not found: '{cfg.ssl_ca_certs}'. Pl	validation	error	ssl, tls, certificates, server-args, startup-validation
--enable-ssl-refresh requires --ssl-certfile and --ssl-keyfi	validation	error	ssl, tls, certificate-rotation, server-args
--http2-max-concurrent-streams must be between 1 and 4294967	validation	error	http2, server-args, configuration-validation
--enable-http2 requires the 'granian' package. Install it wi	validation	error	http2, missing-dependency, pip, server-args
--enable-ssl-refresh is not supported with --enable-http2. G	validation	error	http2, ssl, certificate-rotation, incompatible-flags
mm_preprocess_cache_size_mb must be non-negative	validation	error	multimodal, cache, server-args, validation
mm_process_config must be a dict, but got {type(cfg.mm_proce	validation	error	multimodal, json, server-args, type-error
mm_process_config['{key}'] must be a dict, but got {type(cfg	validation	error	multimodal, json, server-args, type-error
SGLANG_GRPC_WORKER_THREADS ({cfg.grpc_worker_threads}) must	validation	error	grpc, config-validation, environment-variable, server-args
--sidecar-args requires --sidecar.	validation	error	grpc, sidecar, config-validation, server-args
--sidecar-args must be a JSON array of strings.	validation	error	grpc, sidecar, json-validation, config-validation
--sidecar must not be empty.	validation	error	grpc, sidecar, config-validation
--sidecar requires SGLang's native gRPC server; it cannot be	validation	error	grpc, sidecar, flag-conflict, config-validation
--sidecar requires --grpc-port or SGLANG_GRPC_PORT.	validation	error	grpc, sidecar, config-validation
--grpc-port is not supported with --use-ray: the Ray serve l	validation	error	grpc, ray, flag-conflict, config-validation
--grpc-port is not supported with --encoder-only: encoder di	validation	error	grpc, encoder-only, flag-conflict, config-validation
Native gRPC does not yet support --tokenizer-worker-num > 1.	validation	error	grpc, tokenizer-workers, config-validation
--grpc-port is incompatible with --api-key/--admin-api-key:	validation	error	grpc, security, api-key, flag-conflict
--cuda-graph-config[{phase}].backend={backend!r} not allowed	validation	error	cuda-graph, config-validation, backend-selection
Intern-S2-Mobius does not support: " + "; ".join(unsupported	validation	error	model-support, pipeline-parallelism, expert-parallelism, config-validation
--enable-dsa-cache-layer-split is only supported for DSA (De	validation	error	dsa, deepseek, model-support, config-validation
--enable-cp-decode-attn-tp is only supported for models whos	validation	error	context-parallel, model-support, config-validation
--enable-two-batch-overlap is not supported with DSA index-t	validation	error	tbo, dsa, deepseek, flag-conflict
--enable-dsa-cache-layer-split is not supported on decode wo	validation	error	dsa, pd-disaggregation, config-validation
--enable-dsa-cache-layer-split is only supported on PD prefi	validation	error	dsa, pd-disaggregation, config-validation
--enable-dsa-cache-layer-split requires --enable-prefill-cp	validation	error	dsa, context-parallel, config-validation
--enable-dsa-cache-layer-split currently only supports the m	validation	error	dsa, pd-disaggregation, transfer-backend, config-validation
--enable-dsa-cache-layer-split is not supported with pipelin	validation	error	dsa, pipeline-parallelism, config-validation
MiMoV2ForCausalLM requires effective attention TP size {expe	validation	critical	sglang, tensor-parallel, model-config, dp-attention, mimo
TRTLLM MHA backend for prefill requires Hopper (SM90), Black	validation	error	sglang, gpu-architecture, attention-backend, trtllm, sm90
TRTLLM FMHAv2 prefill on SM120 does not support fp8_e4m3 KV	validation	error	sglang, sm120, fp8-kv-cache, attention-backend, trtllm, env-var
TRTLLM MHA backend for decode is only supported on Hopper (S	validation	error	sglang, gpu-architecture, attention-backend, trtllm, decode
Prefill context parallelism with the TRTLLM MHA prefill back	validation	error	sglang, context-parallel, attention-backend, trtllm, sm100, prefill
intel_xpu backend is only supported on decode for MLA models	validation	error	sglang, intel-xpu, mla, attention-backend, prefill-decode-split
--kv-cache-dtype mxfp8 requires an SM100+ (Blackwell) GPU fo	validation	error	sglang, kv-cache-dtype, mxfp8, blackwell, gpu-architecture
--kv-cache-dtype=nvfp4 requires Blackwell SM100 or SM120. Us	validation	error	sglang, nvfp4, kv-cache-dtype, blackwell, gpu-architecture
KV4 is not tested on non-CUDA platforms.	validation	error	sglang, kv4, platform-support, rocm, non-cuda, kv-cache-dtype
--mamba-cache-philox-rounds must be non-negative.	validation	error	sglang, mamba, argument-validation, philox, cli-args
--mamba-max-states-per-path must be -1 (unlimited) or a posi	validation	error	sglang, mamba, server-args, validation, startup
Stochastic rounding for the Mamba SSM cache requires --mamba	validation	error	sglang, mamba, stochastic-rounding, dtype, server-args
Stochastic rounding for the Mamba SSM cache is only supporte	validation	error	sglang, mamba, cuda-only, platform, rocsm, server-args
Stochastic rounding for the Mamba SSM cache with --mamba-bac	validation	error	sglang, mamba, triton, sm100, gpu-architecture, stochastic-rounding
{flashinfer_error}	validation	error	sglang, mamba, flashinfer, dependency, import-error
--enable-int8-mamba-checkpoint is not supported together wit	validation	error	sglang, mamba, int8, hierarchical-cache, incompatible-flags, server-args
--enable-int8-mamba-checkpoint only supports the built-in ma	validation	error	sglang, mamba, int8, radix-cache, incompatible-flags, server-args
--linear-attn-decode-backend flashkda is not supported: Flas	validation	error	sglang, linear-attention, flashkda, prefill-only, server-args
--linear-attn-decode-backend flashinfer on SM100+ requires -	validation	error	sglang, linear-attention, flashinfer, sm100, bfloat16, dtype
--linear-attn-verify-backend flashinfer on SM100+ requires -	validation	error	sglang, linear-attention, flashinfer, blackwell, dtype-validation
--linear-attn-prefill-backend flashinfer on SM100+ requires	validation	error	sglang, flashinfer, cuda-version, blackwell, linear-attention
--enable-linear-replayssm requires Triton, or Helion for KDA	validation	error	sglang, replayssm, linear-attention, backend-validation
--enable-linear-replayssm requires --mamba-radix-cache-strat	validation	error	sglang, replayssm, mamba-radix-cache, config-conflict
--enable-linear-replayssm is not supported under PD disaggre	validation	error	sglang, replayssm, pd-disaggregation, unsupported-feature
--linear-replayssm-cache-len must be >= 1, got {cfg.linear_r	validation	error	sglang, replayssm, argument-validation
--enable-linear-replayssm-spec requires a linear draft chain	validation	error	sglang, replayssm, speculative-decoding, eagle, config-conflict
--enable-linear-replayssm-spec requires the triton or flashi	validation	error	sglang, replayssm, speculative-decoding, backend-validation
--enable-linear-replayssm-spec with SGLANG_RAGGED_VERIFY_MOD	validation	error	sglang, replayssm, ragged-verify, environment-variable, kda
--enable-linear-replayssm-spec is not supported on a PD pref	validation	error	sglang, replayssm, pd-disaggregation, speculative-decoding
DeepEP v2 MoE is not validated for {architecture!r}; support	validation	error	moe, deepep, a2a-backend, server-args, model-architecture
DeepEP v2 MoE is not validated as a speculative draft backen	validation	error	speculative-decoding, deepep, moe, draft-model, server-args
flashinfer_cutedsl FP4 MoE only supports DeepEP low_latency	validation	error	moe, flashinfer, cutedsl, fp4, deepep-mode, server-args
DeepEP v2 does not forward deterministic=True to ElasticBuff	validation	error	deepep, determinism, moe, server-args, elastic-buffer
DeepEP v2 MoE currently supports only --moe-runner-backend d	validation	error	moe, deepep, runner-backend, deep-gemm, server-args
DeepEP v2 MoE has not implemented the TBO/SBO overlap hooks	validation	error	deepep, tbo, sbo, overlap, moe, server-args
DeepEP v2 MoE has not validated fused shared experts yet. Re	validation	error	deepep, shared-experts, fusion, moe, server-args
moe_a2a_backend='pplx' only supports low-latency mode; set -	validation	error	pplx, deepep-mode, moe, a2a-backend, server-args
--ep-dispatch-algorithm {cfg.ep_dispatch_algorithm} picks a	validation	error	eplb, dispatch-algorithm, moe, a2a-backend, server-args
SGLANG_ENABLE_EPLB_BALANCEDNESS_METRIC is no longer supporte	validation	error	env-var, eplb, metrics, migration, server-args, deprecated
--prefill-only-disable-kv-cache currently requires --is-embe	validation	error	sglang, kv-cache, prefill, embedding, server-args
--prefill-only-disable-kv-cache does not currently support -	validation	error	sglang, kv-cache, fp4, quantization, server-args
--prefill-only-disable-kv-cache does not currently support -	validation	error	sglang, kv-cache, mxfp8, quantization, server-args
--prefill-only-disable-kv-cache requires --chunked-prefill-s	validation	error	sglang, chunked-prefill, kv-cache, flashattention, server-args
--prefill-only-disable-kv-cache requires --disable-radix-cac	validation	error	sglang, radix-cache, prefix-cache, kv-cache, server-args
--prefill-only-disable-kv-cache is incompatible with --attn-	validation	error	sglang, context-parallelism, kv-cache, distributed, server-args
--prefill-only-disable-kv-cache is incompatible with --enabl	validation	error	sglang, context-parallelism, prefill, kv-cache, server-args
--prefill-only-disable-kv-cache is incompatible with --enabl	validation	error	sglang, hisparse, sparse-attention, kv-cache, server-args
--prefill-only-disable-kv-cache currently requires the FA pr	validation	error	sglang, attention-backend, flashattention, kv-cache, server-args
hicache_host_memory_mode must be 'cache' or 'buffer_only', g	validation	error	sglang, hicache, host-memory, config-validation, server-args
--mm-feature-transport=cuda_ipc requires NVIDIA CUDA.	validation	error	sglang, cuda, multimodal, ipc, hardware-requirement
--mm-feature-transport=cuda_ipc only supports a single node.	validation	error	sglang, cuda-ipc, multi-node, multimodal, distributed
--enable-deepseek-v4-fp4-indexer requires SM100 or SM120 GPU	validation	error	sglang, deepseek, fp4, gpu-architecture, deepgemm
--disaggregation-decode-retraction-backup=host_pool is only	validation	error	sglang, pd-disaggregation, retraction, host-pool, config-validation
--disaggregation-decode-retraction-backup=host_pool does not	validation	error	sglang, pd-disaggregation, dcp, retraction, host-pool
--disaggregation-decode-retraction-backup=host_pool requires	validation	error	sglang, pd-disaggregation, priority-scheduling, preemption, retraction
The arguments enable-hierarchical-cache and disable-radix-ca	validation	error	sglang, hicache, radix-cache, mutually-exclusive, config-validation
The argument disaggregation-decode-enable-offload-kvcache is	validation	error	sglang, pd-disaggregation, kv-offload, decode-only, config-validation
The argument disaggregation-decode-enable-offload-kvcache is	validation	error	sglang, pd-disaggregation, kv-offload, hicache, missing-backend
The arguments disaggregation-decode-enable-offload-kvcache a	validation	error	sglang, pd-disaggregation, retraction, kv-offload, mutually-exclusive
--swa-full-tokens-ratio should be in range (0, 1.0].	validation	error	sglang, server-args, swa, validation, config
Currently only {RADIX_SUPPORTED_DETERMINISTIC_ATTENTION_BACK	validation	error	sglang, deterministic-inference, mla, attention-backend, deepseek
Deterministic inference with absorbed-MLA models on the fa4	validation	error	sglang, fa4, cuda-arch, deterministic-inference, blackwell
--enable-unified-memory supports monolithic (decode) cuda-gr	validation	error	sglang, unified-memory, cuda-graph, prefill, piecewise
--asr-max-buffer-seconds must be positive (got {cfg.asr_max_	validation	error	sglang, asr, transcription, server-args, validation
--asr-max-concurrent-sessions must be positive (got {cfg.asr	validation	error	sglang, asr, concurrency, server-args, validation
--prefill-decode-interval must be non-negative.	validation	error	sglang, scheduler, prefill-decode, server-args, validation
--default-chat-template-kwargs must decode to a JSON object	validation	error	sglang, chat-template, json, server-args, validation
Invalid modality '{modality}' in --limit-mm-data-per-request	validation	error	sglang, multimodal, limit-mm-per-request, server-args, validation
When enabling two batch overlap without an EP a2a backend (m	validation	error	sglang, two-batch-overlap, dp-attention, moe, server-args
--load-publish-endpoint needs an active --kv-events-config p	validation	error	sglang, kv-events, config-validation, load-publish
{reason}	validation	error	sglang, kv-events, endpoint-mismatch, dp-size
Invalid type for item in --lora-paths list: {type(lora_path)	validation	error	sglang, lora, cli-parsing, type-validation
Invalid type for --lora-paths: {type(cfg.lora_paths)}. Expec	validation	error	sglang, lora, cli-parsing, type-validation
LoRA is only compatible with NGRAM, EAGLE, NEXTN, EAGLE3, DF	validation	error	sglang, lora, speculative-decoding, feature-incompatibility
LoRA with EAGLE/NEXTN/EAGLE3 speculative decoding {reason}.	validation	error	sglang, lora, eagle, speculative-decoding, feature-incompatibility
--decoupled-spec-bind-endpoint, --decoupled-spec-connect-end	validation	error	sglang, speculative-decoding, decoupled, ipc, missing-argument
Missing adaptive runtime state for steps={speculative_num_st	exception	error	sglang, speculative-decoding, adaptive-runtime, state-management
BS {key}: candidate_steps must be a list of non-negative int	validation	error	sglang, speculative-decoding, config-validation, adaptive
speculative_adaptive_config must contain at least one intege	validation	error	sglang, speculative-decoding, config-validation, adaptive, key-format
HiCache does not support Inkling MTP draft state yet.	exception	error	sglang, speculative-decoding, hicache, mtp, not-implemented
External ngram corpus path does not exist: {path}	validation	error	sglang, ngram, speculative-decoding, file-not-found, validation
A tokenizer is required to load an external ngram corpus.	validation	error	sglang, ngram, tokenizer, validation, null-argument
External ngram corpus max tokens must be positive.	validation	error	sglang, ngram, validation, invalid-argument
Invalid JSON in external ngram corpus at line {line_no}: {e.	validation	error	sglang, ngram, json, jsonl, corrupt-data
Invalid external ngram corpus record at line {line_no}: expe	validation	error	sglang, ngram, json, jsonl, schema-validation
External ngram corpus exceeds the configured token limit ({m	validation	error	sglang, ngram, token-limit, budget-exceeded
External corpus '{corpus_id}' already exists. Remove it befo	validation	error	sglang, ngram, duplicate-key, idempotency
Invalid decoupled draft scheduler rid: {rid}	validation	error	sglang, speculative-decoding, decoupled, rid-parsing, validation
VerifyCommit committed_tokens must be non-empty: request_id=	validation	error	sglang, speculative-decoding, decoupled, protocol-validation, empty-collection
draft_token_num must be positive, got {draft_token_num}.	validation	error	speculative-decoding, dflash, argument-validation
next_token_logits row count mismatch for DFlash verify adjus	validation	error	speculative-decoding, dflash, shape-mismatch
num_target_layers must be positive, got {num_target_layers}.	validation	error	speculative-decoding, dflash, config-validation
num_draft_layers must be positive, got {num_draft_layers}.	validation	error	speculative-decoding, dflash, config-validation
DFlash layer selection requires num_target_layers >= 4. Got	validation	error	speculative-decoding, dflash, layer-selection, config-validation
DFLASH config.layer_types must be a sequence of attention ty	validation	error	speculative-decoding, dflash, hf-config, type-validation
DFLASH sliding_attention layers require config.sliding_windo	validation	error	speculative-decoding, dflash, sliding-window, hf-config
Invalid {field_name}={value!r}.	validation	error	speculative-decoding, dflash, config-parsing, type-validation
{field_name} must be {comparator}, got {parsed}.	validation	error	speculative-decoding, dflash, config-validation, range-check
DFLASH requires draft num_hidden_layers in config. Got confi	validation	error	speculative-decoding, dflash, missing-config-field
next_token_logits row count mismatch. Expected {bs * draft_t	validation	critical	sglang, speculative-decoding, dflash, shape-mismatch, tensor-validation
candidates and next_token_logits must be on the same device,	validation	error	sglang, speculative-decoding, dflash, cuda, device-mismatch
uniform_samples shape mismatch. Expected {(bs, draft_token_n	validation	error	sglang, speculative-decoding, dflash, shape-mismatch, rng
uniform_samples_for_final_sampling shape mismatch. Expected	validation	error	sglang, speculative-decoding, dflash, shape-mismatch, rng
Nemotron 3.5 DFLASH draft requires its checkpoint embedding.	exception	critical	sglang, speculative-decoding, dflash, nemotron, checkpoint, embedding
{context} expected 2D positions, got shape={tuple(pos2d.shap	exception	error	sglang, speculative-decoding, dflash, internal-invariant, shape-mismatch
{context} mask/position shape mismatch: {tuple(mask.shape)}	exception	error	sglang, speculative-decoding, dflash, internal-invariant, mask-shape
{context} req_to_token table is empty but gather mask is non	exception	error	sglang, speculative-decoding, dflash, kv-cache, memory-pool, internal-invariant
DFLASH mask_token must be a non-empty string, got {mask_toke	validation	error	sglang, speculative-decoding, dflash, config-validation, tokenizer
DFLASH mask_token_id is outside the target vocab size. mask_	validation	critical	sglang, speculative-decoding, dflash, vocab-size, embedding, model-loading
draft sampler set but the draft forward has no hidden_states	exception	critical	sglang, speculative-decoding, cuda-graph, draft-sampler, hidden-states, runtime
DSpark speculative_num_draft_tokens must be >= 2 (= gamma +	validation	error	sglang, dspark, speculative-decoding, config-validation, gamma
RaggedVerifyLayout requires at least one request	validation	error	speculative-decoding, validation, batch-empty
every request must verify the anchor (verify_len >= 1), got	validation	error	speculative-decoding, validation, off-by-one
capped layout has a row exceeding cap={self.cap}: {verify_le	validation	error	speculative-decoding, validation, capacity-limit
total_verify_tokens {total_verify_tokens} != sum(verify_lens	validation	error	speculative-decoding, validation, invariant
total_verify_tokens {total_verify_tokens} exceeds graph_num_	validation	error	speculative-decoding, cuda-graph, validation
capture layout needs 1 <= num_slots <= num_tokens, got num_s	validation	error	speculative-decoding, cuda-graph, validation
capture layout cannot pack num_tokens={num_tokens} into {num	validation	error	speculative-decoding, cuda-graph, validation
Unknown speculative algorithm name: {name}	validation	error	speculative-decoding, configuration, enum-lookup
Speculative algorithm {self.name} does not support overlap s	validation	error	speculative-decoding, overlap-scheduling, configuration
{spec_class.__name__} is missing duck-typed methods from Spe	validation	error	speculative-decoding, plugin-api, duck-typing
'{upper}' is a reserved speculative algorithm name; cannot b	validation	error	speculative-decoding, plugin-api, name-collision
Speculative algorithm '{upper}' already registered.	validation	error	speculative-decoding, plugin-api, duplicate-registration
Unknown speculative phase: {phase}	validation	error	speculative-decoding, validation, enum-lookup
Invalid simulate_acc_method: {simulate_acc_method}	validation	error	speculative-decoding, validation, configuration
STANDALONE speculative decoding requires the draft model to	validation	critical	speculative-decoding, vocabulary-mismatch, model-config
STANDALONE speculative decoding requires the draft model to	validation	critical	speculative-decoding, tokenizer-mismatch, vocabulary-mismatch
{start_len=} must be non-negative	validation	error	state-capture, validation, off-by-one
torch_npu detected, but NPU device is not available or visib	exception	error	npu, environment, hardware-detection
Unknown device module: {device}	validation	error	device, torch, validation
Unsupported device type: {device!r}. If this is an OOT platf	validation	error	platform-plugin, device, configuration
NPU detected, but torchair package is not installed. Please	exception	error	npu, ascend, torchair, torch-compile, import-error
CUDA error: {err}	exception	critical	cuda, gpu, memory-allocation, driver
temp_set_env should not be used for sglang env vars	validation	warning	environment-variables, testing, conventions, sglang
allowed media domains must be strings	validation	error	validation, media, security, ssrf
allowed media domains cannot be empty	validation	error	validation, media, security, config
Invalid allowed media domain {domain!r}: provide a hostname	validation	error	validation, media, security, ssrf, url-parsing
Invalid allowed media domain {domain!r}: ports are not suppo	validation	error	validation, media, security, ssrf
Invalid allowed media domain {domain!r}	validation	error	validation, media, security, idna, dns
media_url_max_file_size_mb must be non-negative	validation	error	validation, media, security, config
Invalid media URL: {url!r}	validation	error	media, security, ssrf, url-validation
Media URL domain is not allowed. Allowed domains: {sorted(_a	validation	error	media, security, ssrf, allowlist, network
media URL timeout must be positive	validation	error	validation, media, timeout, network
Invalid media URL: {current_url!r}	validation	error	media, url-validation, network, security
Remote media exceeds the {max_bytes} byte download limit	validation	error	media, size-limit, download, security
Invalid audio format: {audio_file}	validation	error	audio, multimodal, validation, input-validation
Could not decode audio: {e}	validation	error	audio, multimodal, decode, libsndfile
Invalid image: {image_file}	validation	error	image, multimodal, validation, input-validation
Unsupported video input type: {type(video_file)}	validation	error	video, multimodal, input-validation, valueerror
Could not decode video: {e}	validation	error	video, codec, decode, torchcodec, decord
{pkg} is installed with version {installed_version}, which i	exception	error	versioning, dependencies, packaging, pip
{pkg} with minimum required version {min_version} is not ins	exception	error	dependencies, missing-package, packaging
kill_process_tree: {len(alive)} process(es) not reaped withi	exception	error	process-management, sigkill, subprocess, cuda, zombie-process
Setting SGLANG_LOGGING_CONFIG_PATH from env with {SGLANG_LOG	exception	error	logging, configuration, env-var, startup
serialized_named_tensors entries must be base64 strings or b	validation	error	serialization, tensors, base64, typeerror
Blocked unsafe class loading ({module}.{name}), to prevent e	exception	error	security, pickle, cve, deserialization
func_path should contain both module name and func name (suc	validation	error	dynamic-import, configuration, valueerror
CUDA VMM POSIX FD broker failed	exception	critical	cuda, vmm, ipc, file-descriptor, broker
CUDA VMM POSIX FD broker returned no file descriptor	exception	critical	cuda, vmm, ipc, socket, file-descriptor
memory_size must be positive	validation	error	cuda, vmm, validation, constructor
consumer_count must be positive	validation	error	cuda, vmm, validation, constructor
recycle_interval must be positive	validation	error	cuda, vmm, validation, constructor
CUDA VMM multimodal transport selected POSIX_FD, but this po	exception	critical	cuda, vmm, fabric, gpu-topology, runtimeerror
memory_size={memory_size} is smaller than CUDA VMM granulari	validation	error	cuda, vmm, memory-allocation, granularity
CUDA VMM multimodal pool failed	exception	critical	cuda, vmm, pool, lazy-failure, runtimeerror
CUDA VMM multimodal pool is closing	exception	error	cuda, vmm, shutdown, race-condition, lifecycle
CUDA VMM pool has no occupied slice at control offset {contr	exception	error	cuda, vmm, double-free, memory-pool
CUDA VMM recycler did not stop	exception	error	cuda, vmm, shutdown, threading, timeout
CUDA VMM proxy has no shareable handle	exception	error	cuda, vmm, ipc, handle, multiprocessing
attention group range [{group_start}, {group_end}) is outsid	validation	error	cuda, vmm, parallelism, world-size-mismatch
consumer_count must be 1, the attention TP size, or the full	validation	error	cuda, vmm, validation, parallelism
CUDA VMM tensor has already released its pool slice	exception	error	cuda, vmm, use-after-free, lifecycle
Packed CUDA VMM features must be reconstructed before releas	exception	error	cuda, vmm, packed-tensors, api-misuse
A CUDA VMM-enabled model must provide a multimodal processor	exception	error	cuda, vmm, multimodal, config
CUDA VMM feature transport requires each feature field to co	exception	error	cuda, vmm, multimodal, validation, type-error
Failed to cancel {len(errors)} VMM transport slice(s)	exception	error	cuda, vmm, aggregate-error, dispatch
cuda.bindings.driver is required for CUDA VMM operations	exception	critical	cuda, vmm, import-error, environment, dependencies
{label}: {err}	exception	error	cuda, driver-api, vmm, error-code
Invalid graph capture input size: {nbytes}	exception	error	cuda, vmm, cuda-graph, validation
cuMemGetAddressRange: {err}	exception	error	cuda, vmm, cuda-graph, invalid-pointer
graph capture input at {ptr} is outside VMM allocation [base	exception	error	cuda, vmm, cuda-graph, pointer-range
no supported CUDA VMM allocation handle type	exception	critical	cuda, vmm, handle, driver-capability, environment
handle_types must be 'auto', an integer, or None	validation	error	cuda, vmm, validation, handle, type-error
invalid CUDA handle-type value: {handle_type_value}	validation	error	cuda, vmm, validation, handle, enum-value
{len(extents)} extents exceed BUMPARENA_MAX_EXTENTS ({BumpAr	validation	error	cuda, vmm, memory, capacity-limit
mapping [{offset}, {offset + size}) is outside reservation [	validation	error	cuda, vmm, offset-out-of-range, validation
VmmReservation.map_existing after close	exception	error	cuda, vmm, use-after-close, lifecycle
sendmsg sent {sent} bytes, expected {len(header)}	exception	error	network, unix-socket, scm-rights, fd-passing
received truncated fd header: {len(data)} < {_FD_HEADER_BYTE	exception	error	network, unix-socket, protocol-mismatch, fd-passing
expected one fd, got header={fd_count}, ancillary={len(fds)}	exception	error	network, scm-rights, fd-passing, validation
VMM handle export failed: FABRIC export failed on at least o	exception	critical	cuda, distributed, fabric, p2p, driver-support
duplicate fd for {key}	exception	error	distributed, fd-passing, duplicate-key, protocol
timed out waiting for POSIX fd exchange	exception	error	distributed, timeout, unix-socket, fd-passing
POSIX fd exchange receive failed	exception	error	distributed, fd-passing, wrapper-error
must be list or null; got {type(v).__name__}	validation	error	validation, config, type-error
row {i}: {e}	validation	error	validation, config, nested-list
elements must be int or list; got {type(v[0]).__name__}	validation	error	validation, type-error, config
No file matching quant type {quant_type!r} in {repo_id}. Ava	exception	error	gguf, huggingface, model-loading
Quant type {quant_type!r} is ambiguous in {repo_id}: {sorted	exception	error	gguf, huggingface, ambiguous-match
{model} contains {len(candidates)} .gguf files; name the one	exception	error	gguf, huggingface, model-selection
model_config_parser={model_config_parser!r} is incompatible	exception	error	gguf, config, server-args
Can't get gguf config for {config.model_type}. Place a confi	exception	error	gguf, config, unsupported-architecture
No pre-tokenizer regex known for tokenizer.ggml.pre={pre_nam	exception	error	gguf, tokenizer, version-mismatch
Found unknown quantization='{quantization}' in config	exception	error	mistral, quantization, config
File not found {model}, {file_name}	exception	error	mistral, config, file-not-found
Failed to load mistral '{config_file_name}' config for model	exception	error	mistral, config, corrupt-file
Unsupported image processor backend: {backend}. Expected one	validation	error	multimodal, image-processor, server-args
use_fast={use_fast} conflicts with image_processor_backend={	validation	error	multimodal, image-processor, conflicting-args
Cannot determine processor class for {model_path}	exception	error	multimodal, processor, model-loading
Failed to load image_processor for {model_path}: {e}. This m	exception	error	multimodal, image-processor, dependencies
Failed to load the tokenizer. If you are using a LLaMA V1 mo	exception	error	tokenizer, huggingface, typeerror
Failed to load the tokenizer. If the tokenizer is a custom t	exception	error	tokenizer, trust-remote-code, huggingface
Retry with use_fast=False for {tokenizer_name} also failed (	exception	error	tokenizer, transformers, version-mismatch
The fastokens package is required when --tokenizer-backend=f	exception	error	tokenizer, missing-dependency, installation
Cannot use the fast tokenizer in slow tokenizer mode.	validation	error	tokenizer, configuration, sglang
fastokens failed to load tokenizer for {tokenizer_name!r}. T	exception	error	tokenizer, sglang, backend
Integrity check failed: {joined errors}	exception	critical	integrity, checksum, model-files
No model files found in {model_path}	exception	error	model-files, checksum, path
No files found in HF repo {repo_id}.	exception	error	huggingface, checksum, model-files
Cannot msgpack encode object of type {type(obj)} with enc_ho	exception	error	serialization, msgpack, sglang
Cannot msgpack decode object of type {type(obj)} as {tp} wit	exception	error	serialization, msgpack, version-mismatch
Unhandled known MessagePack extension code: {code}	exception	error	serialization, msgpack, ipc, version-mismatch
Expected base64-encoded bytes	validation	error	serialization, base64, msgspec
{port_name} has invalid port number {port}. Valid TCP port r	validation	error	network, port, validation
{port_name} at {port} is not available in {timeout_s} second	exception	error	network, port, timeout, ci
Could not bind port {port} on any configured address family	exception	error	network, socket, bind
Unsupported socket type: {socket_type}	validation	error	zmq, network, socket
Environment variable SGLANG_LOCAL_IP_NIC requires package ne	exception	error	network, environment, dependency
Can not get local ip	exception	error	network, ip, container, distributed
Invalid port number: {s!r}	validation	error	network, parsing, address
Port out of range (0-65535): {port}	validation	error	network, port, parsing
Cannot resolve host {host!r}: {e}	exception	error	network, dns, distributed
Empty address string	validation	error	network, parsing, validation
Missing closing bracket in IPv6 address: {addr!r}	validation	error	network, ipv6, parsing
Invalid IPv6 address inside brackets: {host!r}	validation	error	network, ipv6, address-parsing, sglang
Expected ':port' after closing bracket, got: {rest!r}	validation	error	network, ipv6, port, address-parsing, sglang
Missing port in address (expected host:port): {addr!r}	validation	error	network, port, address-parsing, sglang
Empty host in address: {addr!r}	validation	error	network, address-parsing, hostname, sglang
Bare IPv6 address without brackets is ambiguous: {addr!r}. U	validation	error	network, ipv6, address-parsing, sglang
NUMA node {node} has no CPU cores allowed by the current aff	exception	warning	numa, cpu-affinity, cgroup, sglang, resource-binding
nvImageCodec could not decode the JPEG image	exception	error	nvjpeg, image-decoding, gpu, multimodal, sglang
nvImageCodec returned an invalid JPEG tensor: shape={tuple(i	exception	error	nvjpeg, image-decoding, tensor-shape, multimodal, sglang
Module instance {} is not unique	validation	error	nvtx, pytorch-hooks, shared-weights, profiling, sglang
Unknown {old_param_type=} {old_param=}	validation	error	offloading, meta-device, pytorch, parameters, sglang
Invalid device_uuid=	exception	error	cuda, device-uuid, gpu, torch-patch, sglang
Unknown type: {device_maybe_uuid=}	exception	error	cuda, device, type-error, torch-patch, sglang
No trace files found for profile_id: {self.profile_id}	validation	error	profiling, trace-merge, chrome-trace, sglang
manually start is only supported yet	exception	error	profiling, not-implemented, sglang
manually stop is only supported yet	exception	error	profiling, not-implemented, sglang
unsupported profile stage: {forward_mode=}	exception	error	profiling, forward-mode, scheduler, internal-error, sglang
rank_consensus() got unexpected keyword argument(s): {list(k	validation	error	decorator, distributed, rank-consensus, api-misuse, sglang
{name} must be True / False / str / list[str], got {value!r}	validation	error	decorator, distributed, type-validation, rank-consensus, sglang
Invalid --log-requests-level: {self.log_requests_level=}	validation	error	logging, cli-args, request-logging, sglang
Trace file is empty.	exception	error	profiling, rocm, rpd, chrome-trace, sglang
SGLANG_LOG_SCHEDULER_STATUS_TARGET is set but --enable-metri	validation	error	sglang, scheduler, metrics, env-var, startup-config
Multiple config files specified! Only one allowed.	validation	error	sglang, cli, config, yaml
No config file specified after --config flag!	validation	error	sglang, cli, missing-argument
Config file must contain a dictionary at root level	validation	error	sglang, yaml, config, validation
Config file must be YAML format, got: {path.suffix}	validation	error	sglang, yaml, file-extension, config
Config file not found: {file_path}	validation	error	sglang, config, file-not-found, yaml
Unsupported config option '{key_norm}' with action '{action.	validation	error	sglang, yaml, config, argparse, unsupported-option
The MLX tensor bridge requires MLX >= 0.32.0	exception	error	sglang, mlx, dependency, import-error, macos
MLX 0.32 does not support complex128; convert the Torch tens	validation	error	sglang, mlx, dtype, complex-numbers, torch
The MLX tensor bridge supports CPU and MPS tensors, got {ten	validation	error	sglang, mlx, device, torch
MlxTensorView requires a Torch MPS tensor, got {owner.device	validation	error	sglang, mlx, mps, device, torch
borrow_torch_tensors requires MPS tensors, got {devices}	validation	error	sglang, mlx, mps, batch, device
The MLX tensor bridge supports CPU and MPS targets, got {tar	validation	error	sglang, mlx, device, target
MLX float64 arrays cannot be exported to a Torch MPS tensor;	validation	error	sglang, mlx, float64, mps, dtype
mlx_call_multi operation must return a non-empty tuple or li	validation	error	sglang, mlx, callback, return-type, contract
mlx_call_multi outputs must be MLX arrays	validation	error	sglang, mlx, type-check, callback, return-type
pattern must contain at least one token	validation	error	sglang, tokens, pattern-matching, validation, kmp
[IpcModelLoader] Error communicating with daemon at {self.so	error_code	critical	weight-cache, ipc, daemon, socket
[weight_cache:{where}] quantization method {quant_method!r}	validation	error	quantization, weight-cache, ipc, config
Connection closed while reading message header	error_code	error	protocol, socket, eof, weight-cache
Message size {length} exceeds {MAX_MSG_SIZE} byte cap	validation	error	protocol, framing, size-limit
Connection closed while reading message body	error_code	error	protocol, socket, eof, truncated
{env_field.name}={template!r} must contain '{{global_rank}}'	validation	error	config, env-var, weight-cache, tensor-parallel
Weight cache daemon for rank {global_rank} is already runnin	exception	error	daemon, weight-cache, lifecycle, stale-process
sendmsg sent {sent} bytes, expected {len(payload)}	exception	error	ipc, fd-passing, sendmsg, weight-cache
received truncated fd header: {len(data)} < {_FD_INDEX_STRUC	exception	error	ipc, fd-passing, protocol
expected one fd, got {len(fds)}	exception	error	ipc, fd-passing, protocol
weight cache transport backend {VMM_FD_BACKEND!r} is not imp	exception	error	transport, not-implemented, weight-cache, config
Unknown weight cache transport backend {name!r}	exception	error	config, transport, invalid-argument
Cannot create empty tensor bucket	validation	warning	validation, weight-sync, empty-input
Must provide either named_tensors or both flattened_tensor a	validation	error	validation, weight-sync, constructor
Cannot parse schema {json_schema}. The schema must be either	validation	error	json-schema, structured-output, validation
Could not open video file:{video_path}	exception	error	video, opencv, file-io, multimodal
No free port available.	exception	error	network, port, test-infra
Command contains only environment variable assignments, no e	validation	error	shell, subprocess, validation
Server process exited with code {return_code}	exception	critical	server, lifecycle, startup, subprocess
Server process exited	exception	critical	server, lifecycle, oom, subprocess
num_tokens ({num_tokens}) exceeds num_max_dispatch_tokens_pe	exception	critical	moe, all-to-all, token-dispatcher, pplx, batch-size, capacity
mega MoE: num_tokens={num_tokens} exceeds SGLANG_OPT_DEEPGEM	exception	critical	moe, deepgemm, kimi-k3, mega-moe, buffer-capacity, env-var
Error in stream_executor: {get_exception_traceback()}	console	error	interpreter, stream-executor, async, frontend, nested-exception
`sglang.bench_offline_throughput` is deprecated and will be	console	warning	deprecation, benchmark, offline-throughput, future-warning
`sglang.bench_one_batch` is deprecated and will be removed i	console	warning	deprecation, benchmark, one-batch, future-warning
`sglang.bench_one_batch_server` is deprecated and will be re	console	warning	deprecation, benchmark, one-batch-server, future-warning
`sglang.bench_serving` is deprecated and will be removed in	console	warning	deprecation, benchmark, serving, future-warning
Triton is not supported on current platform, roll back to CP	console	warning	triton, cuda, device-detection, cpu-fallback, fla
The parameter max_tokens will be overwritten by speculated n	console	warning	openai-backend, speculative-decoding, sampling-params, max-tokens, ignored-parameter
EBNF is not officially supported by OpenAI endpoints. Ignori	console	info	ebnf, openai, grammar-constraint, warning, backend
Both dtype and regex are set. Only dtype will be used. dtype	console	warning	dtype, regex, structured-output, conflict, sampling-params
Warning: system prompt is not supported in VertexAI.	console	warning	vertexai, system-prompt, chat-roles, warning
Regular expression is not supported in the OpenAI backend.	console	warning	regex, openai, structured-output, constraint-dropped, warning
Regular expression is not supported in the VertexAI backend.	console	warning	regex, vertexai, structured-output, constraint-dropped, warning
Regular expression is not supported in the Anthropic backend	console	warning	regex, anthropic, structured-output, constraint-dropped, warning
Regular expression is not supported in the LiteLLM backend.	console	warning	regex, litellm, structured-output, constraint-dropped, warning
'python -m sglang.launch_server' is still supported, but 'sg	console	info	deprecation, cli, launch-server, entrypoint, migration
FusedScaleResidualNormScaleShift cuda not available, using n	console	warning	cuda, kernel-fallback, layernorm, performance, shape-constraint
FusedNormScaleShift cuda not available, using native fallbac	console	warning	cuda, kernel-fallback, layernorm, performance, shape-constraint
CUDA coredump env var {key} is already set to '{os.environ[k	console	info	cuda-coredump, environment-variables, debugging, configuration
'data_parallel_rank' is deprecated, use 'routed_dp_rank' ins	console	warning	deprecation, data-parallel, engine-api, migration
'data_parallel_rank' is deprecated, use 'routed_dp_rank' ins	console	warning	deprecation, openai-api, data-parallel, protocol, migration
Invalid value for {self.name}: {e}, using default "{default}	console	warning	environment-variables, configuration, parsing, fallback
Environment variable '{self.deprecated_name}' is deprecated;	console	warning	deprecation, environment-variables, alias, migration
Environment variable {old_name} is deprecated. Please use {s	console	warning	deprecation, environment-variables, migration, startup
Environment variable {key} is deprecated, please use {new_ke	console	warning	deprecation, environment-variables, prefix-rewrite, migration
attention-backend='nsa' is deprecated; use 'dsa' instead. Th	console	warning	deprecation, attention-backend, nsa, dsa, migration
sglang.srt.layers.attention.nsa is deprecated; use sglang.sr	console	warning	deprecation, import, module-rename, nsa, dsa
sglang.srt.layers.attention.nsa.dequant_k_cache is deprecate	console	warning	deprecation, sglang, nsa, import
sglang.srt.layers.attention.nsa.index_buf_accessor is deprec	console	warning	deprecation, sglang, nsa, import
sglang.srt.layers.attention.nsa.nsa_backend_mtp_precompute i	console	warning	deprecation, sglang, mtp, import
sglang.srt.layers.attention.nsa.nsa_indexer is deprecated; u	console	warning	deprecation, sglang, indexer, import
sglang.srt.layers.attention.nsa.quant_k_cache is deprecated;	console	warning	deprecation, sglang, quantization, import
sglang.srt.layers.attention.nsa.tilelang_kernel is deprecate	console	warning	deprecation, sglang, tilelang, import
sglang.srt.layers.attention.nsa.transform_index is deprecate	console	warning	deprecation, sglang, nsa, import
sglang.srt.layers.attention.nsa.triton_kernel is deprecated;	console	warning	deprecation, sglang, triton, import
sglang.srt.layers.attention.nsa.utils is deprecated; use sgl	console	warning	deprecation, sglang, utils, import
sglang.srt.layers.attention.nsa_backend is deprecated; use s	console	warning	deprecation, sglang, attention-backend, import
VisionAttention(head_size=...) is deprecated; use head_dim=.	console	warning	deprecation, sglang, vision, api-rename
{name} is deprecated; use {replacement} instead.	console	warning	deprecation, sglang, dcp, distributed
Only CUDA, HIP and XPU support AWQ currently.	console	warning	sglang, awq, quantization, platform-support, hardware
HIP does not support fused_marlin_moe currently.	console	warning	sglang, awq, marlin, rocm, moe, platform-support
Only CUDA, MUSA and NPU support GGUF quantization currently.	console	warning	sglang, gguf, quantization, platform-support
Only CUDA and MUSA support GGUF quantization currently.	console	warning	sglang, gguf, rocm, quantization, platform-support
MultiPlatformOp is deprecated; subclass sglang.kernels.fused	console	warning	deprecation, sglang, operators, subclassing
'data_parallel_rank' is deprecated, use 'routed_dp_rank' ins	console	warning	deprecation, sglang, request-api, data-parallel
skip_attn_backend_init is deprecated and will be removed; pr	console	warning	deprecation, sglang, scheduler, attention-backend
mean is more than 2 std from [a, b] in nn.init.trunc_normal_	console	warning	pytorch, initialization, numerics, trunc-normal
In ps_version 'v1', the height and width have not been swapp	console	warning	internvl, vision-encoder, pixel-shuffle, multimodal, deprecation
get_num_tokens_per_bs_for_target_verify is deprecated; use g	console	info	speculative-decoding, deprecation, api-rename, python-warnings
get_num_tokens_per_bs_for_target_verify is deprecated; use g	console	info	speculative-decoding, deprecation, api-rename, registry
LONG GARBAGE COLLECTION DETECTED \| Generation {} \| Duration:	console	warning	gc, latency-jitter, performance, scheduler, python
Using a slow tokenizer. This might cause a significant slowd	console	warning	tokenizer, performance, huggingface, startup
The 'checksums' format is deprecated. Please regenerate with	console	info	model-verification, manifest, deprecation, checksums
Including the scheme in --host ('{host}') is deprecated. Pas	console	info	url, deprecation, host-config, networking, tests
batch_draft_token_num config value ${config} must be less th	exception	error	config-validation, ngram, speculative-decoding, constructor
startExternalCorpusLoad called while another load is in prog	exception	error	ngram, corpus-loading, concurrency, state-machine
appendExternalCorpusTokens called without startExternalCorpu	exception	error	ngram, corpus-loading, api-misuse, state-machine
finishExternalCorpusLoad called without startExternalCorpusL	exception	error	ngram, corpus-loading, api-misuse, state-machine
External corpus is empty — no tokens were loaded.	exception	error	ngram, corpus-loading, empty-input
External corpus '${corpus_id}' already exists. Remove it bef	exception	error	ngram, corpus-loading, duplicate-key
batchMatch expects state_ids, tokens, and total_lens to matc	exception	error	ngram, batch-validation, argument-mismatch
Unknown match_type: '${param_.match_type}'. Must be 'BFS' or	exception	error	ngram, config-validation, enum-value
batchMatch received an empty token tail	exception	error	ngram, empty-input, batch-validation
out_tokens buffer too small: ${out_tokens.size(0)} < ${resul	exception	error	ngram, buffer-overflow, ffi, tensor-shape