ErrLookup › huggingface/pytorch-image-models

huggingface/pytorch-image-models

The largest collection of PyTorch image encoders / backbones. Including train, eval, inference, export scripts, and pretrained weights -- ResNet, ResNeXT, EfficientNet, NFNet, Vision Transformer (ViT), MobileNetV4, MobileNet-V3 & V2, RegNet, DPN, CSPNet, Swin Transformer, MaxViT, CoAtNet, ConvNeXt, and more · Python · 301 source files

Analyzed at 9a5261e31b on 2026-08-27. 144 documented errors.

Code / MessageTypeSeverityTags
Input image must have positive dimensions, got H={height}, W
validation error timm, transforms, image-size, validation
Invalid class map file, expected a dict ({class_map_path}).
validation error timm, class-map, pickle, dataset
Dataset length is unknown, please pass `num_samples` explici
validation error timm, huggingface, dataset, num-samples
Found 0 images in subfolders of {root}. Supported image exte
error_code critical timm, dataset, image-folder, no-samples
Invalid or corrupt tar info cache file {cache_path}.
validation error timm, tar-dataset, cache-corruption
split {split} not found in info ({info.get('splits', {}).key
error_code error timm, webdataset, split, info-json
Please install webdataset 0.2.x package `pip install git+htt
error_code error timm, webdataset, missing-dependency, pip
Invalid split definition, num_samples not specified in train
error_code error timm, webdataset, num-samples, training
ScheduledBatchSampler requires a sampler with a length.
validation error timm, sampler, pytorch, type-validation
ScheduledBatchSampler requires a non-empty sampler.
validation critical timm, sampler, empty-dataset
batch_sizes must contain at least one value.
validation error timm, sampler, batch-size, config
All scheduled batch sizes must be positive integers.
validation error timm, sampler, batch-size, validation
num_batches must be a positive integer when specified.
validation error timm, sampler, num-batches, validation
choice_schedule must be 'constant' or 'progressive'.
validation error timm, sampler, schedule, invalid-argument
A progressive schedule requires at least two choices.
validation error timm, sampler, progressive-schedule, validation
schedule_epochs must be a positive integer for a progressive
validation error timm, sampler, schedule-epochs, validation
schedule_spread must be non-negative.
validation error timm, sampler, schedule-spread, validation
schedule_random_mix must be between 0 and 1.
validation error timm, sampler, random-mix, validation
No full scheduled batch fits the sampler; reduce the batch s
validation error timm, sampler, batch-size, dataset-size
choice_weights and batch_sizes must have the same length.
validation error timm, sampler, choice-weights, validation
Model architecture ({arch_name}) has no pretrained cfg regis
exception error timm, model-registry, pretrained, config
features_only not implemented for Vision Transformer models.
exception error timm, coat, features-only, unsupported-operation
features_only not implemented for Vision Transformer models.
exception error timm, convit, features-only, unsupported-operation
features_only not implemented for ConvMixer models.
exception error timm, convmixer, features-only, unsupported-operation
Invalid local_mbconv_norm={local_mbconv_norm!r}; expected on
exception error timm, cpubone, validation, constructor
features_only not implemented for Vision Transformer models.
exception error timm, crossvit, features-only, unsupported-operation
Memory Efficient not supported in JIT
exception error timm, densenet, torchscript, gradient-checkpointing
Token mixer type: {} not supported
exception error timm, fastvit, architecture, validation
Expected input ndim in (3, 4, 5); got {x.ndim}.
exception error timm, gemma4-vit, tensor-shape, input-validation
patch_coord is required for pre-patchified input.
exception error timm, gemma4-vit, patch-coordinates, missing-argument
Cannot pool {N} tokens with k={k}: N must be divisible by k^
exception error timm, gemma4-vit, pooling, image-size
Image size ({H}, {W}) must be divisible by (patch_size * poo
exception error timm, gemma4-vit, image-size, pooling, validation
Gemma4VitEncoder does not support classification use cases.
exception error timm, gemma4-vit, encoder-only, unsupported-operation
output_fmt='NCHW' requires a raw image (B, C, H, W) input.
exception error timm, gemma4-vit, feature-extraction, tensor-shape
num_branches({}) <> num_blocks({})
exception error timm, hrnet, architecture, config-mismatch
MobileNetV5Encoder does not support classification use cases
exception error timm, mobilenetv5, encoder-only, unsupported-operation
Unsupported model {mode}
exception error timm, mvitv2, constructor, unsupported-operation
Cannot initialize position embeddings without grid_size.Plea
validation error timm, naflexvit, pos-embed, config
Patch interpolation is not supported by this embedding confi
exception error timm, naflexvit, patch-embed, runtime-check
Unknown rope_type: {cfg.rope_type}
validation error timm, naflexvit, rope, config
output_fmt="NCHW" is not supported for NaFlex (dict) inputs,
validation error timm, naflexvit, features, output-format
NaFlex forward_intermediates with active patch dropout requi
validation error timm, naflexvit, patch-dropout, features
Invalid net configuration
validation error timm, selecsls, variant, config
Unrecognized union:
validation error timm, sequencer, config, invalid-argument
The output channel {2 * self.output_size} is different from
validation error timm, sequencer, shape-mismatch, config
The output channel {self.output_size} is different from the
validation error timm, sequencer, shape-mismatch, config
features_only not implemented for Vision Transformer models.
exception error timm, visformer, features-only, unsupported
Unsupported patch embedding rank in {checkpoint_path}: {embe
validation error timm, vit, checkpoint, jax
Patch embedding shape mismatch in {checkpoint_path}: checkpo
validation error timm, vit, checkpoint, shape-mismatch
Cannot infer position grid from {pos_embed_w.shape[1]} token
validation error timm, vit, checkpoint, pos-embed
Unsupported position embedding shape in {checkpoint_path}: {
validation error timm, vit, checkpoint, pos-embed
{name} must be a scalar or scalar tensor.
validation error timm, optimizer, hyperparameter, type-error
Invalid {name}: {value}
validation error timm, optimizer, hyperparameter, range-check
Invalid beta parameter at index 0: {}
validation error optimizer, adamw, hyperparameters, validation
Invalid beta parameter at index 1: {}
validation error optimizer, adamw, hyperparameters, validation
AdamW does not support sparse gradients
validation error optimizer, adamw, sparse-gradients, embedding
API has changed, `state_steps` argument must contain a list
validation error optimizer, adamw, functional-api, pytorch-version
Invalid learning rate: {}
validation error optimizer, adan, hyperparameters, validation
Invalid epsilon value: {}
validation error optimizer, adan, hyperparameters, validation
Invalid beta parameter at index 0: {}
validation error optimizer, adan, hyperparameters, validation
Invalid beta parameter at index 1: {}
validation error optimizer, adan, hyperparameters, validation
Invalid beta parameter at index 2: {}
validation error optimizer, adan, hyperparameters, validation
lr as a Tensor is not supported for capturable=False and for
validation error optimizer, adopt, tensor-lr, foreach, capturable
Tensor lr must be 1-element
validation error optimizer, adopt, tensor-lr, shape
Invalid learning rate: {lr}
validation error optimizer, adopt, hyperparameters, validation
Invalid epsilon value: {eps}
validation error optimizer, adopt, hyperparameters, validation
Invalid beta parameter at index 0: {betas[0]}
validation error optimizer, adopt, hyperparameters, validation
Invalid beta parameter at index 1: {betas[1]}
validation error optimizer, adopt, hyperparameters, validation
Invalid weight_decay value: {weight_decay}
validation error optimizer, adopt, weight-decay, validation
ADOPT does not support sparse gradients
validation error optimizer, adopt, sparse-gradients, embedding
`requires_grad` is not supported for `step` in differentiabl
validation error optimizer, adopt, differentiable, autograd, meta-learning
Momentum {momentum} must be in the range [0,1]
validation error optimizer, madgrad, momentum, hyperparameter-validation
Learning rate {lr} must be positive
validation error optimizer, madgrad, learning-rate, hyperparameter-validation
Weight decay {weight_decay} must be non-negative
validation error optimizer, madgrad, weight-decay, hyperparameter-validation
Eps must be non-negative
validation error optimizer, madgrad, epsilon, hyperparameter-validation
momentum != 0 is not compatible with sparse gradients
exception error optimizer, madgrad, sparse-gradients, momentum
weight_decay option is not compatible with sparse gradients
exception error optimizer, madgrad, sparse-gradients, weight-decay
Invalid learning rate: {}
validation error optimizer, mars, learning-rate, hyperparameter-validation
Invalid epsilon value: {}
validation error optimizer, mars, epsilon, hyperparameter-validation
Invalid beta parameter at index 0: {}
validation error optimizer, mars, betas, hyperparameter-validation
Invalid beta parameter at index 1: {}
validation error optimizer, mars, betas, hyperparameter-validation
Adam does not support sparse gradients, please consider Spar
exception error optimizer, mars, sparse-gradients
Tensor must have at least 2 dimensions, got {tensor.ndim}
validation error optimizer, muon, shape-validation, parameter-routing
Unknown mode: {mode}
validation error optimizer, muon, enum-validation
Invalid conv_mode: {conv_mode}
validation error optimizer, muon, conv-mode, enum-validation
Invalid algo: {algo}. Must be 'muon' or 'adamuon'
validation error optimizer, muon, algo, enum-validation
adamw_lr is not supported with tensor lr; use fallback_lr_sc
validation error optimizer, muon, deprecation, tensor-lr
Cannot compute fallback_lr_scale from adamw_lr when lr=0
validation error optimizer, muon, deprecation, division-by-zero
Muon does not support sparse gradients
exception error optimizer, muon, sparse-gradients
Coefficient must be length-3 of real numbers, got: {x!r}
validation error optimizer, muon, newton-schulz, coefficients, validation
Unknown coefficients preset '{value}'. Valid options: {valid
validation error optimizer, muon, newton-schulz, preset, enum-validation
Preset '{value}' is empty or invalid
validation error timm, muon-optimizer, preset-validation, valueerror
Coefficients must be a preset name (str), a 3-sequence (a,b,
validation error timm, muon-optimizer, typeerror, argument-validation
Item {i} is not a sequence: {item!r}
validation error timm, muon-optimizer, typeerror, nesting
Coefficient list cannot be empty
validation error timm, muon-optimizer, empty-list, valueerror
Invalid learning rate: {}
validation error timm, nadam, learning-rate, valueerror
Invalid learning rate: {lr}
validation error timm, nadamw, learning-rate, valueerror
Invalid epsilon value: {eps}
validation error timm, nadamw, epsilon, valueerror
Invalid beta parameter at index 0: {betas[0]}
validation error timm, nadamw, betas, valueerror
Invalid beta parameter at index 1: {betas[1]}
validation error timm, nadamw, betas, valueerror
Invalid weight_decay value: {weight_decay}
validation error timm, nadamw, weight-decay, valueerror
NAdamW does not support sparse gradients
exception error timm, nadamw, sparse-gradients, embedding
API has changed, `state_steps` argument must contain a list
exception error timm, nadamw, functional-api, state-steps, pytorch-version-change
Invalid learning rate: {}
validation error timm, nvnovograd, learning-rate, valueerror
Invalid epsilon value: {}
validation error timm, nvnovograd, epsilon, valueerror
Invalid beta parameter at index 0: {}
validation error timm, nvnovograd, betas, valueerror
Invalid beta parameter at index 1: {}
validation error timm, nvnovograd, betas, valueerror
Sparse gradients are not supported.
exception error timm, nvnovograd, sparse-gradients, embedding
RAdam does not support sparse gradients
exception error timm, radam, sparse-gradients, embedding
Invalid alpha value: {}
validation error timm, rmsprop-tf, alpha, valueerror
RMSprop does not support sparse gradients
exception error timm, rmsprop-tf, sparse-gradients, embedding
Unsupported distill_type '{distill_type}'. Must be 'soft' or
validation error distillation, config-validation, valueerror, timm
No node names found matching {names}.
exception error attention-extraction, fx-graph, name-matching, timm
No module names found matching {names}.
exception error attention-extraction, module-names, hooks, timm
Please provide `hook_fns` for each `hook_fn_locs`, their len
validation error hooks, argument-validation, timm
You have provided a batch norm layer as the `root module`. P
exception error batchnorm, freezing, fine-tuning, timm
Error processing sample index {idx}. Error: {e}. Skipping sa
console warning dataset, robust-loading, naflex, timm
Calculated batch size <= 0 (seq_len={seq_len}, remaining={re
console warning naflex, batch-schedule, defensive-check, timm
Rank {self.rank}: Canonical schedule accounts for {total_sch
console warning naflex, batch-schedule, distributed, timm
Rank {self.rank}: Number of indices for this rank ({len(indi
console warning naflex, distributed, epoch-schedule, timm
Rank {self.rank}: Ran out of samples ({idx_pos}/{effective_s
console warning naflex, distributed, schedule-mismatch, timm
Rank {self.rank}: Assigned {scheduled_samples_count} samples
console warning naflex, distributed, schedule-mismatch, timm
Transform returned None for index {idx}. Skipping sample.
console warning naflex, transforms, robust-loading, timm
IndexError encountered for index {idx} (possibly due to padd
console warning naflex, index-error, padding, timm
{name.capitalize()} range reversed. Swapping.
console warning naflex, transforms, range-validation, timm
final_scale_range values should ideally be between 0.0 and 1
console warning naflex, transforms, scale-range, timm
Final scale randomization ({scale_factor:.2f}) resulted in s
console warning naflex, transforms, max-seq-len, timm
range should be of kind (min, max)
console warning transforms, range-validation, augmentation, timm
This version of pytorch does not have F.scaled_dot_product_a
console warning attention, pytorch-version, feature-detection, timm
DropBlock2d() got unexpected keyword argument '{k}'
console warning dropblock, deprecated-args, kwargs, timm
mean is more than 2 std from [a, b] in nn.init.trunc_normal_
console warning weight-init, trunc-normal, statistics, timm
Overwriting {model_name} in registry with {fn.__module__}.{m
console warning timm, registry, duplicate, model-registration
Mapping deprecated model name {deprecated_name} to current {
console info timm, deprecation, model-name, migration
CSATv2 is designed for 3-channel RGB input. in_chans={in_cha
console warning timm, csatv2, in-chans, input-shape
Importing from {__name__} is deprecated, please import via t
console warning timm, deprecation, import, futurewarning
Importing from {__name__} is deprecated, please import via t
console warning timm, deprecation, import, features
Importing from {__name__} is deprecated, please import via t
console warning timm, deprecation, import, torch-fx
Importing from {__name__} is deprecated, please import via t
console warning timm, deprecation, import, builder
Importing from {__name__} is deprecated, please import via t
console warning timm, deprecation, import, hub
Importing from {__name__} is deprecated, please import via t
console warning timm, deprecation, import, layers
Importing from {__name__} is deprecated, please import via t
console warning timm, deprecation, import, registry
It is highly recommended to have 'opt_einsum' installed for
console warning timm, optimizer, kron, performance, einsum
adamw_lr is deprecated, use fallback_lr_scale=adamw_lr/lr in
console warning timm, optimizer, muon, deprecation, api-change
Importing from {__name__} is deprecated, please import via t
console warning timm, deprecation, import, optimizer-factory