SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity
Files
b4115a426eb284ccf4494d90aba3d264ccbd6f25
HuggingFace_transformer/tests/models
History
Anton Vlasjuk b4115a426e
Some checks failed
Release - Conda / build_and_package (push) Has been cancelled
Details
Secret Leaks / trufflehog (push) Has been cancelled
Details
[Ernie 4.5] Add ernie text models (#39228) 
* init

* copied from remote

* add proper structure and llama like structure

* fixup

* revert to state that works

* get closer to llama

* slow and steady

* some removal

* masks work

* it is indeed the rope implementation, how dafuq does it mesh with the cache now hmm

* nice

* getting closer

* closer to transformers style

* let's simplify this, batching works now

* simplified

* working version with modular

* it is indeed the rotation per weights, make it complete llama style

* cleanup conversion, next to look at -> tokenizer

* remove llama artefacts

* fix modeling tests (common ones)

* style

* integration test + first look into tokenization (will need more work, focussing on modeling other models first)

* style

* working moe version, based on remote

* lets keep it simple and go step by step - transformers annotations for modular and transformers style rope (complex view)

* more cleanup

* refactor namings and remove addition forXXX classes

* our moe won't cut it it seems, correction bias seems to be missing in remote code version

* tokenization change (remote)

* our moe version works when adding normalization :D

* cleanup moe

* nits

* cleanup modeling -> let's get to modular next

* style

* modular v1

* minor things + attempt at conversion (which doesn't work)

* no conversion follow glm, fixup modular and other nits

* modular cleanup

* fixes

* tests, tests, tests + some moe dtype forcing

* simplify modular, fix fatal fa2 bug, remaining tests

* fix import issue?

* some initial docs, fix bnb faulty behavior --> needs to fix some tests because of gate needing to be float

* fix sdpa test, load on init dtype only

* fixup post merge

* style

* fix doc links

* tokenization cleanup beginnings

* simplify tokenizer by a lot as its basically llama

* tokenizer is full llama with different defaults + extra special tokens

* sync og special tokens of ernie

* fix decoding with numbers (also in remote done what a timing), begin of tok tests

* align with remote and preserve special tokens, adjust tests to ernie legacy behavior, warning for questionable behavior (also in llama)

* nits

* docs

* my daily post merge it is

* check

* tokenization update with explanations and conversion script

* review on modular (til), revert some tokenizer things i did prior, remove mtp comment (low prio)

* post merge fixes

* fixup tokenization, llama fast is the way to go

* more fixups

* check

* import fixes

* correction bias following the paddle code

* fix

* fix TP plan, fix correction bias sharding during forward

* style

* whoops

* fix tied weights

* docs and last nit

* license

* flasky tests

* move repo id, update when merged on the hub
2025-07-21 19:51:49 +02:00
..
aimv2
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
albert
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
align
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
altclip
🚨 Don't use cache in non-generative models (#38751)
2025-07-01 09:08:21 +00:00
arcee
Add Arcee model support (#38621)
2025-06-24 15:05:29 +02:00
aria
Update expected values (after switching to A10) - part 6 (#39207)
2025-07-03 22:45:30 +02:00
audio_spectrogram_transformer
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
auto
Enable some ruff checks for performance and readability (#39383)
2025-07-17 13:21:59 +00:00
autoformer
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
aya_vision
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
bamba
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
bark
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
bart
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
barthez
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
bartpho
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
beit
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
bert
[cache] make all classes cache compatible finally (#38635)
2025-07-16 14:00:17 +02:00
bert_generation
Fix typos in strings and comments (#37799)
2025-04-28 11:39:11 +01:00
bert_japanese
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
bertweet
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
big_bird
[cache] make all classes cache compatible finally (#38635)
2025-07-16 14:00:17 +02:00
bigbird_pegasus
[generation] bring back tests on vision models (#38603)
2025-06-06 08:23:15 +00:00
biogpt
Bart: new cache format (#35314)
2025-05-16 13:26:54 +02:00
bit
Add ImageProcessorFast to BiT processor (#37180)
2025-04-14 17:07:48 +02:00
bitnet
Add Bitnet model (#37742)
2025-04-28 15:08:46 +02:00
blenderbot
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
blenderbot_small
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
blip
Fix more flaky test_initialization (#38932)
2025-06-20 17:28:32 +02:00
blip_2
🚨🚨 Fix and simplify attention implementation dispatch and subconfigs handling (#39423)
2025-07-18 13:41:54 +02:00
bloom
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
bridgetower
Add Idefics2/3 and SmolVLM Fast image processors + improvements for fast image processors (#38157)
2025-06-23 14:17:25 +00:00
bros
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
byt5
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
camembert
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
canine
Skip torchscript tests for 2 models (#38643)
2025-06-06 20:17:37 +02:00
chameleon
fix xpu failures on PT 2.7 and 2.8 w/o IPEX and enable hqq cases on XPU (#39187)
2025-07-08 10:18:26 +02:00
chinese_clip
Add Fast Chinese-CLIP Processor (#37012)
2025-04-15 18:31:20 +02:00
clap
Fix missing initializations for models created in 2023 (#39239)
2025-07-21 14:43:52 +02:00
clip
Raise TypeError instead of ValueError for invalid types (#38660)
2025-07-21 12:42:00 +00:00
clipseg
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
clvp
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
code_llama
remove unhandled parameter (#38145)
2025-06-02 15:57:32 +02:00
codegen
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
cohere
Remove all traces of low_cpu_mem_usage (#38792)
2025-06-12 16:39:33 +02:00
cohere2
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
colpali
[vlm] fix loading of retrieval VLMs (#39242)
2025-07-15 17:23:54 +02:00
colqwen2
Update some tests for torch 2.7.1 (#38701)
2025-06-10 11:46:52 +02:00
conditional_detr
Update expected values (after switching to A10) (#39157)
2025-07-01 20:54:31 +02:00
convbert
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
convnext
Update expected values (after switching to A10) (#39157)
2025-07-01 20:54:31 +02:00
convnextv2
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
cpm
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
cpmant
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
csm
[chat template] return assistant mask in processors (#38545)
2025-07-18 12:23:20 +00:00
ctrl
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
cvt
Update expected values (after switching to A10) (#39157)
2025-07-01 20:54:31 +02:00
d_fine
Update expected values (after switching to A10) (#39157)
2025-07-01 20:54:31 +02:00
dab_detr
Update expected values (after switching to A10) (#39157)
2025-07-01 20:54:31 +02:00
dac
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
data2vec
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
dbrx
Refactor DBRX tests to use CausalLMModelTest base classes (#38475)
2025-06-13 16:22:12 +01:00
deberta
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
deberta_v2
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
decision_transformer
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
deepseek_v2
Add DeepSeek V2 Model into Transformers (#36400)
2025-07-09 17:04:28 +02:00
deepseek_v3
Skip sdpa dispatch on flash test due to unsupported head dims (#39010)
2025-06-24 20:16:56 +02:00
deformable_detr
Update expected values (after switching to A10) (#39157)
2025-07-01 20:54:31 +02:00
deit
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
depth_anything
Skip some export tests on torch 2.7 (#38677)
2025-06-12 12:47:15 +02:00
depth_pro
Update expected values (after switching to A10) - part 7 (#39218)
2025-07-04 12:48:10 +02:00
detr
Update expected values (after switching to A10) (#39157)
2025-07-01 20:54:31 +02:00
dia
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
diffllama
Remove all traces of low_cpu_mem_usage (#38792)
2025-06-12 16:39:33 +02:00
dinat
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
dinov2
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
dinov2_with_registers
Fix more flaky test_initialization (#38932)
2025-06-20 17:28:32 +02:00
distilbert
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
dit
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
doge
Add Doge model (#35891)
2025-07-08 11:44:29 +02:00
donut
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
dots1
fix dots1 tests (#39088)
2025-06-27 16:54:11 +02:00
dpr
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
dpt
Update expected values (after switching to A10) - part 3 (#39179)
2025-07-02 22:48:30 +02:00
efficientnet
Add EfficientNet Image PreProcessor (#37055)
2025-04-16 21:59:24 +02:00
electra
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
emu3
Update expected values (after switching to A10) - part 5 (#39205)
2025-07-03 19:56:02 +02:00
encodec
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
encoder_decoder
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
eomt
🚨🚨🚨 [eomt] make EoMT compatible with pipeline (#39122)
2025-07-02 12:25:26 +01:00
ernie
Remove old code for PyTorch, Accelerator and tokenizers (#37234)
2025-04-10 20:54:21 +02:00
ernie4_5
[Ernie 4.5] Add ernie text models (#39228)
2025-07-21 19:51:49 +02:00
ernie4_5_moe
[Ernie 4.5] Add ernie text models (#39228)
2025-07-21 19:51:49 +02:00
esm
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
falcon
🚨 🚨 Inherited CausalLM Tests (#37590)
2025-05-23 18:29:31 +01:00
falcon_h1
Update expected values (after switching to A10) - part 5 (#39205)
2025-07-03 19:56:02 +02:00
falcon_mamba
Refactor MambaCache to modeling_mamba.py (#38086)
2025-07-21 14:59:36 +02:00
fastspeech2_conformer
fix fastspeech2_conformer tests (#39229)
2025-07-07 15:04:26 +02:00
flaubert
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
flava
Add Idefics2/3 and SmolVLM Fast image processors + improvements for fast image processors (#38157)
2025-06-23 14:17:25 +00:00
fnet
🚨 rm already deprecated pad_to_max_length arg (#37617)
2025-05-01 15:21:55 +02:00
focalnet
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
fsmt
Fix fsmt tests (#38904)
2025-06-19 10:56:34 +02:00
funnel
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
fuyu
🔴 [VLM] Add base model without head (#37033)
2025-05-07 17:47:51 +02:00
gemma
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
gemma2
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
gemma3
[gemma3] support sequence classification task (#39465)
2025-07-21 11:03:20 +02:00
gemma3n
Several fixes for Gemma3n (#39135)
2025-07-01 10:34:53 +02:00
git
[CI] Fix partially red CI (#39448)
2025-07-16 15:53:43 +02:00
glm
Update expected values (after switching to A10) - part 5 (#39205)
2025-07-03 19:56:02 +02:00
glm4
enable misc test cases on XPU (#38852)
2025-06-18 09:20:49 +02:00
glm4_moe
GLM-4 Update (#39393)
2025-07-21 13:24:34 +02:00
glm4v
fix Glm4v batch videos forward (#39172)
2025-07-10 10:44:28 +02:00
glpn
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
got_ocr2
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
gpt2
fix a bunch of XPU UT failures on stock PyTorch 2.7 and 2.8 (#39069)
2025-06-27 14:01:53 +02:00
gpt_bigcode
🚨🚨 Fix and simplify attention implementation dispatch and subconfigs handling (#39423)
2025-07-18 13:41:54 +02:00
gpt_neo
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
gpt_neox
🚨 🚨 Inherited CausalLM Tests (#37590)
2025-05-23 18:29:31 +01:00
gpt_neox_japanese
Remove old code for PyTorch, Accelerator and tokenizers (#37234)
2025-04-10 20:54:21 +02:00
gpt_sw3
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
gptj
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
granite
switch to device agnostic device calling for test cases (#38247)
2025-05-26 10:18:53 +02:00
granite_speech
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
granitemoe
switch to device agnostic device calling for test cases (#38247)
2025-05-26 10:18:53 +02:00
granitemoehybrid
switch to device agnostic device calling for test cases (#38247)
2025-05-26 10:18:53 +02:00
granitemoeshared
switch to device agnostic device calling for test cases (#38247)
2025-05-26 10:18:53 +02:00
grounding_dino
Raise TypeError instead of ValueError for invalid types (#38660)
2025-07-21 12:42:00 +00:00
groupvit
Fix pylint warnings (#39477)
2025-07-21 12:38:05 +00:00
helium
Update expected values (after switching to A10) - part 5 (#39205)
2025-07-03 19:56:02 +02:00
herbert
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
hgnet_v2
Add D-FINE Model into Transformers (#36261)
2025-04-29 12:17:55 +01:00
hiera
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
hubert
Fix bad tensor shape in failing Hubert test. (#39502)
2025-07-21 12:25:52 +01:00
ibert
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
idefics
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
idefics2
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
idefics3
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
ijepa
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
imagegpt
[test] update test_past_key_values_format (#37614)
2025-04-22 11:07:34 +01:00
informer
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
instructblip
Skip test_eager_matches sdpa generate and update an integration test for blip-like models (#39248)
2025-07-08 10:38:25 +02:00
instructblipvideo
[refactor] set attention implementation (#38974)
2025-07-15 09:34:06 +02:00
internvl
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
jamba
Expectation fixes and added AMD expectations (#38729)
2025-06-13 16:14:58 +02:00
janus
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
jetmoe
🚨 🚨 Inherited CausalLM Tests (#37590)
2025-05-23 18:29:31 +01:00
kosmos2
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
kyutai_speech_to_text
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
layoutlm
🚨 Don't use cache in non-generative models (#38751)
2025-07-01 09:08:21 +00:00
layoutlmv2
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
layoutlmv3
fix layoutlmv3 tests (#39050)
2025-06-26 20:07:17 +02:00
layoutxlm
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
led
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
levit
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
lfm2
Fix Lfm2 and common tests (#39398)
2025-07-14 12:02:59 +02:00
lightglue
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
lilt
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
llama
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
llama4
fix UT failures on XPU w/ stock PyTorch 2.7 & 2.8 (#39116)
2025-06-30 11:49:03 +02:00
llava
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
llava_next
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
llava_next_video
Test fixes for Aria (and some Expectation for llava_next_video) (#39131)
2025-07-02 23:41:14 +02:00
llava_onevision
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
longformer
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
longt5
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
luke
Enable some ruff checks for performance and readability (#39383)
2025-07-17 13:21:59 +00:00
lxmert
Remove all traces of low_cpu_mem_usage (#38792)
2025-06-12 16:39:33 +02:00
m2m_100
🔴🔴🔴 [Attention] Refactor Attention Interface for Bart-based Models (#38108)
2025-05-22 17:12:58 +02:00
mamba
Refactor MambaCache to modeling_mamba.py (#38086)
2025-07-21 14:59:36 +02:00
mamba2
Fix missing initializations for models created in 2024 (#38987)
2025-07-02 15:03:57 +02:00
marian
Remove all traces of low_cpu_mem_usage (#38792)
2025-06-12 16:39:33 +02:00
markuplm
No more Tuple, List, Dict (#38797)
2025-06-17 19:37:18 +01:00
mask2former
Update expected values (after switching to A10) - part 2 (#39165)
2025-07-02 22:47:55 +02:00
maskformer
Update expected values (after switching to A10) - part 2 (#39165)
2025-07-02 22:47:55 +02:00
mbart
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
mbart50
Use lru_cache for tokenization tests (#36818)
2025-03-28 15:09:35 +01:00
megatron_bert
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
megatron_gpt2
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
mgp_str
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
mimi
Add kyutai stt (#38909)
2025-06-24 18:01:15 +02:00
minimax
Refactor the way we handle outputs for new llamas and new models (#39120)
2025-07-05 11:34:28 +02:00
mistral
fix mistral and mistral3 tests (#38978)
2025-06-23 17:07:18 +02:00
mistral3
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
mixtral
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
mlcd
Add MLCD model (#36182)
2025-04-15 11:33:09 +01:00
mllama
[cache] make all classes cache compatible finally (#38635)
2025-07-16 14:00:17 +02:00
mluke
Fix typos in strings and comments (#37799)
2025-04-28 11:39:11 +01:00
mobilebert
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
mobilenet_v1
Update expected values (after switching to A10) - part 2 (#39165)
2025-07-02 22:47:55 +02:00
mobilenet_v2
Add segmentation_maps support to MobileNetV2ImageProcessor (#37312)
2025-07-07 13:34:59 -04:00
mobilevit
Add segmentation_maps support to MobileNetV2ImageProcessor (#37312)
2025-07-07 13:34:59 -04:00
mobilevitv2
Update expected values (after switching to A10) - part 2 (#39165)
2025-07-02 22:47:55 +02:00
modernbert
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
modernbert_decoder
Add ModernBERT Decoder Models - ModernBERT, but trained with CLM! (#38967)
2025-07-15 10:40:41 +02:00
moonshine
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
moshi
Remove all traces of low_cpu_mem_usage (#38792)
2025-06-12 16:39:33 +02:00
mpnet
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
mpt
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
mra
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
mt5
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
musicgen
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
musicgen_melody
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
mvp
[cache] make all classes cache compatible finally (#38635)
2025-07-16 14:00:17 +02:00
myt5
🚨 🚨 Setup -> setupclass conversion (#37282)
2025-04-08 17:15:37 +01:00
nemotron
switch to device agnostic device calling for test cases (#38247)
2025-05-26 10:18:53 +02:00
nllb
Use lru_cache for tokenization tests (#36818)
2025-03-28 15:09:35 +01:00
nllb_moe
Fix typos in strings and comments (#37799)
2025-04-28 11:39:11 +01:00
nougat
add fast image processor nougat (#37661)
2025-06-27 14:39:43 +00:00
nystromformer
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
olmo
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
olmo2
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
olmoe
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
omdet_turbo
Fix missing initializations for models created in 2024 (#38987)
2025-07-02 15:03:57 +02:00
oneformer
Update expected values (after switching to A10) - part 3 (#39179)
2025-07-02 22:48:30 +02:00
openai
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
opt
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
owlv2
Delete deprecated stuff (#38838)
2025-07-10 05:18:44 +00:00
owlvit
Delete deprecated stuff (#38838)
2025-07-10 05:18:44 +00:00
paligemma
Update expected values (after switching to A10) - part 5 (#39205)
2025-07-03 19:56:02 +02:00
paligemma2
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
patchtsmixer
🔴🔴🔴 [Attention] Refactor Attention Interface for Bart-based Models (#38108)
2025-05-22 17:12:58 +02:00
patchtst
Force torch>=2.6 with torch.load to avoid vulnerability issue (#37785)
2025-04-25 16:57:09 +02:00
pegasus
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
pegasus_x
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
perceiver
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
perception_lm
[cache] make all classes cache compatible finally (#38635)
2025-07-16 14:00:17 +02:00
persimmon
🚨 🚨 Inherited CausalLM Tests (#37590)
2025-05-23 18:29:31 +01:00
phi
🚨 🚨 Inherited CausalLM Tests (#37590)
2025-05-23 18:29:31 +01:00
phi3
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
phi4_multimodal
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
phimoe
Fix MoE gradient test (#38438)
2025-05-28 16:44:20 +01:00
phobert
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
pix2struct
Fix more flaky test_initialization (#38932)
2025-06-20 17:28:32 +02:00
pixtral
Add Idefics2/3 and SmolVLM Fast image processors + improvements for fast image processors (#38157)
2025-06-23 14:17:25 +00:00
plbart
🔴🔴🔴 [Attention] Refactor Attention Interface for Bart-based Models (#38108)
2025-05-22 17:12:58 +02:00
poolformer
Update expected values (after switching to A10) - part 3 (#39179)
2025-07-02 22:48:30 +02:00
pop2piano
[tests] remove TF tests (uses of require_tf) (#38944)
2025-06-25 17:29:10 +00:00
prompt_depth_anything
Skip some export tests on torch 2.7 (#38677)
2025-06-12 12:47:15 +02:00
prophetnet
[cache] make all classes cache compatible finally (#38635)
2025-07-16 14:00:17 +02:00
pvt
Update expected values (after switching to A10) - part 3 (#39179)
2025-07-02 22:48:30 +02:00
pvt_v2
Update expected values (after switching to A10) - part 3 (#39179)
2025-07-02 22:48:30 +02:00
qwen2
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
qwen2_5_omni
Fix Qwen Omni integration test (#39553)
2025-07-21 14:11:46 +02:00
qwen2_5_vl
[qwen2 vl] fix packing with all attentions (#39447)
2025-07-21 12:19:15 +02:00
qwen2_audio
Remove dead protected imports (#38980)
2025-06-23 13:44:50 +02:00
qwen2_moe
Fix MoE gradient test (#38438)
2025-05-28 16:44:20 +01:00
qwen2_vl
[qwen2 vl] fix packing with all attentions (#39447)
2025-07-21 12:19:15 +02:00
qwen3
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
qwen3_moe
Fix qwen3_moe tests (#38865)
2025-06-18 14:36:03 +02:00
rag
Fix rag (#38585)
2025-06-23 17:42:46 +02:00
recurrent_gemma
adjust input and output texts for test_modeling_recurrent_gemma.py (#39190)
2025-07-07 15:13:25 +02:00
reformer
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
regnet
Update expected values (after switching to A10) - part 8 - Final (#39220)
2025-07-04 13:35:53 +02:00
rembert
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
resnet
Update expected values (after switching to A10) - part 3 (#39179)
2025-07-02 22:48:30 +02:00
roberta
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
roberta_prelayernorm
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
roc_bert
Remove old code for PyTorch, Accelerator and tokenizers (#37234)
2025-04-10 20:54:21 +02:00
roformer
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
rt_detr
Update expected values (after switching to A10) - part 2 (#39165)
2025-07-02 22:47:55 +02:00
rt_detr_v2
Update expected values (after switching to A10) - part 2 (#39165)
2025-07-02 22:47:55 +02:00
rwkv
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
sam
Raise TypeError instead of ValueError for invalid types (#38660)
2025-07-21 12:42:00 +00:00
sam_hq
Raise TypeError instead of ValueError for invalid types (#38660)
2025-07-21 12:42:00 +00:00
seamless_m4t
[seamless_m4t] Skip some tests when speech is not available (#38430)
2025-06-02 09:17:28 +00:00
seamless_m4t_v2
[seamless_m4t] Skip some tests when speech is not available (#38430)
2025-06-02 09:17:28 +00:00
segformer
Update expected values (after switching to A10) - part 2 (#39165)
2025-07-02 22:47:55 +02:00
seggpt
Update expected values (after switching to A10) - part 3 (#39179)
2025-07-02 22:48:30 +02:00
sew
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
sew_d
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
shieldgemma2
[chat template] return assistant mask in processors (#38545)
2025-07-18 12:23:20 +00:00
siglip
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
siglip2
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
smollm3
Add SmolLM3 (#38755)
2025-06-25 15:12:15 +00:00
smolvlm
[smolvlm] fix video inference (#39147)
2025-07-02 12:05:10 +02:00
speech_encoder_decoder
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
speech_to_text
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
speecht5
Refactor embedding input/output getter/setter (#39339)
2025-07-21 18:18:14 +02:00
splinter
🚨 Don't use cache in non-generative models (#38751)
2025-07-01 09:08:21 +00:00
squeezebert
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
stablelm
🚨 🚨 Inherited CausalLM Tests (#37590)
2025-05-23 18:29:31 +01:00
starcoder2
Update expected values (after switching to A10) - part 5 (#39205)
2025-07-03 19:56:02 +02:00
superglue
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
superpoint
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
swiftformer
Fix missing initializations for models created in 2023 (#39239)
2025-07-21 14:43:52 +02:00
swin
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
swin2sr
Update expected values (after switching to A10) - part 3 (#39179)
2025-07-02 22:48:30 +02:00
swinv2
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
switch_transformers
Update expected values (after switching to A10) - part 3 (#39179)
2025-07-02 22:48:30 +02:00
t5
Fix convert_and_export_with_cache failures for GPU models (#38976)
2025-07-17 13:12:32 +00:00
t5gemma
fix failing test_sdpa_can_dispatch_on_flash (#39259)
2025-07-11 16:30:56 +02:00
table_transformer
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
tapas
[tests] remove tests from libraries with deprecated support (flax, tensorflow_text, ...) (#39051)
2025-06-26 16:25:00 +01:00
textnet
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
time_series_transformer
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
timesfm
Remove dead protected imports (#38980)
2025-06-23 13:44:50 +02:00
timesformer
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
timm_backbone
Fix missing initializations for models created in 2023 (#39239)
2025-07-21 14:43:52 +02:00
timm_wrapper
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
trocr
🚨 🚨 Setup -> setupclass conversion (#37282)
2025-04-08 17:15:37 +01:00
tvp
Add Optional to remaining types (#37808)
2025-04-28 14:20:45 +01:00
udop
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
umt5
[generation] bring back tests on vision models (#38603)
2025-06-06 08:23:15 +00:00
unispeech
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
unispeech_sat
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
univnet
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
upernet
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
video_llava
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
videomae
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
vilt
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
vipllava
[VLMs] support passing embeds along with pixels (#38467)
2025-07-01 11:33:20 +00:00
vision_encoder_decoder
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
vision_text_dual_encoder
[tests] remove tests from libraries with deprecated support (flax, tensorflow_text, ...) (#39051)
2025-06-26 16:25:00 +01:00
visual_bert
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
vit
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
vit_mae
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
vit_msn
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
vitdet
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
vitmatte
Add Idefics2/3 and SmolVLM Fast image processors + improvements for fast image processors (#38157)
2025-06-23 14:17:25 +00:00
vitpose
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
vitpose_backbone
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
vits
Expectation fixes and added AMD expectations (#38729)
2025-06-13 16:14:58 +02:00
vivit
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
vjepa2
Refactor the way we handle outputs for new llamas and new models (#39120)
2025-07-05 11:34:28 +02:00
voxtral
Add voxtral (#39429)
2025-07-18 00:02:04 +00:00
wav2vec2
Add torchcodec in docstrings/tests for datasets 4.0 (#39156)
2025-07-08 17:06:12 +02:00
wav2vec2_bert
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
wav2vec2_conformer
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
wav2vec2_phoneme
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
wav2vec2_with_lm
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
wavlm
Remove script datasets in tests (#38940)
2025-06-25 14:31:20 +00:00
whisper
Rename _supports_flash_attn_2 in examples and tests (#39471)
2025-07-21 14:02:57 +02:00
x_clip
Update expected values (after switching to A10) - part 4 (#39189)
2025-07-03 15:13:06 +02:00
xglm
Expectation fixes and added AMD expectations (#38729)
2025-06-13 16:14:58 +02:00
xlm
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
xlm_roberta
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
xlm_roberta_xl
Remove old code for PyTorch, Accelerator and tokenizers (#37234)
2025-04-10 20:54:21 +02:00
xlnet
Deprecate TF + JAX (#38758)
2025-06-11 17:28:06 +01:00
xmod
Remove old code for PyTorch, Accelerator and tokenizers (#37234)
2025-04-10 20:54:21 +02:00
yolos
🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)
2025-05-23 17:17:38 +02:00
yoso
Use Python 3.9 syntax in tests (#37343)
2025-04-08 14:12:08 +02:00
zamba
Remove all traces of low_cpu_mem_usage (#38792)
2025-06-12 16:39:33 +02:00
zamba2
fix UT failures on XPU w/ stock PyTorch 2.7 & 2.8 (#39116)
2025-06-30 11:49:03 +02:00
zoedepth
Skip some export tests on torch 2.7 (#38677)
2025-06-12 12:47:15 +02:00
__init__.py
…