ErrLookup › deepseek-ai/DeepSeek-V3

deepseek-ai/DeepSeek-V3

· Python · 5 source files

Analyzed at 9b4e9788e4 on 2026-08-14. 16 documented errors.

Code / MessageTypeSeverityTags
Vocabulary size must be divisible by world size (world_size=
exception critical distributed, tensor-parallelism, embedding, model-construction, deepseek
Output features must be divisible by world size (world_size=
exception critical distributed, tensor-parallelism, linear-layer, model-construction, deepseek
Input features must be divisible by world size (world_size=$
exception critical distributed, tensor-parallelism, linear-layer, model-construction, deepseek
Number of experts must be divisible by world size (world_siz
exception critical distributed, moe, experts, model-construction, deepseek
Prompt length exceeds model maximum sequence length (max_seq
validation error generation, context-length, kv-cache, validation, deepseek
Number of prompts exceeds maximum batch size (${args.max_bat
validation error batching, generation, cli, validation, deepseek
Either input-file or interactive mode must be specified
validation error cli, argument-validation, usage-error, deepseek
Input tensor must be contiguous
validation error triton, fp8, quantization, tensor-layout, deepseek
Last dimension size must be divisible by block_size (block_s
validation error triton, fp8, quantization, shape-validation, deepseek
Input tensors must be contiguous
validation error triton, fp8, dequantization, tensor-layout, deepseek
Input tensors must have 2 dimensions
validation error triton, fp8, dequantization, shape-validation, deepseek
Scaling factor tensors must be contiguous
validation error triton, fp8, gemm, tensor-layout, deepseek
Key ${key} not found in mapping
validation error checkpoint-conversion, safetensors, huggingface, key-mapping, deepseek
Dimension ${dim} must be divisible by ${mp}
validation error checkpoint-conversion, model-parallelism, sharding, validation, deepseek
Number of experts must be divisible by model parallelism
validation error checkpoint-conversion, cli, moe, model-parallelism, deepseek
Warning: Missing scale_inv tensor for ${weight_name}, skippi
console warning fp8, dequantization, checkpoint-conversion, safetensors, silent-corruption, deepseek