Pavel Iakubovskii 3249c5dc15 Refactor attention for SigLIP based models (#36981) ...

* Update Siglip attention implementation

* Update tests for Siglip

* Remove one level of indentation

* Update test to be more specific

* Fixup

* Idefics2

* Idefics3

* Emu3

* SmolVLM

* Phi4 (just init small update)

* Idefics2 (test fix)

* Update siglip2 tests

* Update eager

* trigger

* Clean up

* Transfer inputs to device in test

* Fixing test

* Fixing test

* Revert contiguous

* Remove unused is_flash_attn_2_available

* Move flaky to specific models

2025-04-01 15:37:25 +02:00

..

albert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

align

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

altclip

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

aria

Support return_tensors in audio chat templates (#34601)

2025-03-25 11:08:47 +01:00

audio_spectrogram_transformer

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

auto

chore: fix typos in tests directory (#36785)

2025-03-18 10:31:13 +01:00

autoformer

Set weights_only in torch.load (#36991)

2025-03-27 14:55:50 +00:00

aya_vision

[Generation, Gemma 3] When passing a custom generation_config, overwrite default values with the model's base generation_config (#36684)

2025-03-15 12:40:09 +00:00

bamba

Expectations test utils (#36569)

2025-03-18 23:39:50 +01:00

bark

fix can_generate (#36570)

2025-03-17 14:56:18 +01:00

bart

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

barthez

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

bartpho

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

beit

[tests] Parameterized test_eager_matches_sdpa_inference (#36650)

2025-03-14 14:41:27 +00:00

bert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

bert_generation

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

bert_japanese

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

bertweet

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

big_bird

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

bigbird_pegasus

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

biogpt

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

bit

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

blenderbot

[tests] remove tf/flax tests in /generation (#36235)

2025-02-17 14:59:22 +00:00

blenderbot_small

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

blip

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

blip_2

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

bloom

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

bridgetower

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

bros

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

byt5

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

camembert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

canine

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

chameleon

fix typos in the tests directory (#36932)

2025-03-25 10:49:24 +01:00

chinese_clip

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

clap

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

clip

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

clipseg

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

clvp

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

code_llama

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

codegen

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

cohere

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

cohere2

[generate] ✨ vectorized beam search ✨ (#35802)

2025-03-18 18:39:36 +00:00

colpali

[tests] Parameterized test_eager_matches_sdpa_inference (#36650)

2025-03-14 14:41:27 +00:00

conditional_detr

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

convbert

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

convnext

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

convnextv2

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

cpm

Make pipeline able to load processor (#32514)

2024-10-09 16:46:11 +01:00

cpmant

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

ctrl

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

cvt

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

dab_detr

fix typos in the tests directory (#36932)

2025-03-25 10:49:24 +01:00

dac

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

data2vec

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

dbrx

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

deberta

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

deberta_v2

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

decision_transformer

[CI] Check test if the GenerationTesterMixin inheritance is correct 🐛 🔫 (#36180)

2025-02-21 10:18:20 +00:00

deepseek_v3

[WIP] add deepseek-v3 (#35926)

2025-03-28 15:56:59 +01:00

deformable_detr

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

deit

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

depth_anything

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

depth_pro

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

detr

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

diffllama

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

dinat

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

dinov2

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

dinov2_with_registers

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

distilbert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

dit

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

donut

[tests] further fix Tester object has no attribute '_testMethodName' (#35781)

2025-01-29 16:05:33 +01:00

dpr

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

dpt

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

efficientnet

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

electra

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

emu3

🔴 VLM: compile compatibility (#35724)

2025-02-14 15:23:49 +01:00

encodec

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

encoder_decoder

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

ernie

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

esm

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

falcon

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

falcon_mamba

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

fastspeech2_conformer

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

flaubert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

flava

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

fnet

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

focalnet

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

fsmt

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

funnel

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

fuyu

Update ruff to 0.11.2 (#36962)

2025-03-25 16:00:11 +01:00

gemma

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

gemma2

[generate] ✨ vectorized beam search ✨ (#35802)

2025-03-18 18:39:36 +00:00

gemma3

Deprecate #36741 and map Causal to Conditional (#36917)

2025-03-25 09:13:56 +01:00

git

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

glm

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

glpn

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

got_ocr2

Add Got-OCR 2 Fast image processor and refactor slow one (#36185)

2025-03-01 00:56:00 -05:00

gpt2

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

gpt_bigcode

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

gpt_neo

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

gpt_neox

Update ruff to 0.11.2 (#36962)

2025-03-25 16:00:11 +01:00

gpt_neox_japanese

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

gpt_sw3

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

gptj

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

granite

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

granitemoe

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

granitemoeshared

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

grounding_dino

[GroundingDino] Fix grounding dino loss 🚨 (#31828)

2025-02-27 19:15:58 +00:00

groupvit

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

helium

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

herbert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

hiera

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

hubert

[tests] remove pt_tf equivalence tests (#36253)

2025-02-19 11:55:11 +00:00

ibert

[warnings] fix E721 warnings (#32223)

2024-07-25 15:12:23 +02:00

idefics

Set weights_only in torch.load (#36991)

2025-03-27 14:55:50 +00:00

idefics2

Refactor attention for SigLIP based models (#36981)

2025-04-01 15:37:25 +02:00

idefics3

SmolVLM2 (#36126)

2025-02-20 15:00:26 +01:00

ijepa

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

imagegpt

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

informer

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

instructblip

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

instructblipvideo

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

jamba

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

jetmoe

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

kosmos2

Remove hardcoded slow image processor class in processors supporting fast ones (#36266)

2025-03-12 18:39:25 -04:00

layoutlm

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

layoutlmv2

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

layoutlmv3

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

layoutxlm

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

led

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

levit

[tests] further fix Tester object has no attribute '_testMethodName' (#35781)

2025-01-29 16:05:33 +01:00

lilt

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

llama

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

llava

Fix llava xpu tests. (#37130)

2025-04-01 11:10:13 +02:00

llava_next

Set weights_only in torch.load (#36991)

2025-03-27 14:55:50 +00:00

llava_next_video

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

llava_onevision

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

longformer

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

longt5

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

luke

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

lxmert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

m2m_100

[generate] beam search -- fix output cropping (#37080)

2025-03-28 18:57:51 +01:00

mamba

Remove deprecated batch_size parameter (#37007)

2025-03-27 15:01:56 +00:00

mamba2

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

marian

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

markuplm

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

mask2former

Update ruff to 0.11.2 (#36962)

2025-03-25 16:00:11 +01:00

maskformer

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

mbart

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

mbart50

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

megatron_bert

[CI] Check test if the GenerationTesterMixin inheritance is correct 🐛 🔫 (#36180)

2025-02-21 10:18:20 +00:00

megatron_gpt2

Skip tests properly (#31308)

2024-06-26 21:59:08 +01:00

mgp_str

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

mimi

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

mistral

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

mistral3

Deprecate #36741 and map Causal to Conditional (#36917)

2025-03-25 09:13:56 +01:00

mixtral

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

mllama

fix: loss computation after embeddings resize - mllama (#36840)

2025-03-21 14:47:59 +01:00

mluke

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

mobilebert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

mobilenet_v1

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

mobilenet_v2

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

mobilevit

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

mobilevitv2

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

modernbert

[ModernBERT] Never save 'reference_compile' config; should be set based on end user (#36305)

2025-04-01 14:14:39 +02:00

moonshine

[CI] Check test if the GenerationTesterMixin inheritance is correct 🐛 🔫 (#36180)

2025-02-21 10:18:20 +00:00

moshi

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

mpnet

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

mpt

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

mra

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

mt5

[tests] remove test_export_to_onnx (#36241)

2025-02-17 16:52:44 +00:00

musicgen

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

musicgen_melody

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

mvp

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

myt5

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

nemotron

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

nllb

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

nllb_moe

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

nougat

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

nystromformer

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

olmo

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

olmo2

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

olmoe

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

omdet_turbo

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

oneformer

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

openai

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

opt

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

owlv2

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

owlvit

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

paligemma

Fix typos (#36910)

2025-03-24 14:08:29 +00:00

paligemma2

Fix typos (#36910)

2025-03-24 14:08:29 +00:00

patchtsmixer

Set weights_only in torch.load (#36991)

2025-03-27 14:55:50 +00:00

patchtst

Set weights_only in torch.load (#36991)

2025-03-27 14:55:50 +00:00

pegasus

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

pegasus_x

[CI] Check test if the GenerationTesterMixin inheritance is correct 🐛 🔫 (#36180)

2025-02-21 10:18:20 +00:00

perceiver

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

persimmon

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

phi

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

phi3

Export for Phi4-mini (#36780)

2025-03-25 10:46:38 +01:00

phi4_multimodal

Add Phi4 multimodal (#36939)

2025-03-25 09:55:21 +01:00

phimoe

[generate] remove cache v4.47 deprecations (#36212)

2025-02-17 13:55:03 +00:00

phobert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

pix2struct

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

pixtral

Add Mistral3 (#36790)

2025-03-18 12:04:42 +01:00

plbart

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

poolformer

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

pop2piano

[CI] Check test if the GenerationTesterMixin inheritance is correct 🐛 🔫 (#36180)

2025-02-21 10:18:20 +00:00

prompt_depth_anything

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

prophetnet

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

pvt

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

pvt_v2

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

qwen2

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

qwen2_5_vl

Support return_tensors in audio chat templates (#34601)

2025-03-25 11:08:47 +01:00

qwen2_audio

Support return_tensors in audio chat templates (#34601)

2025-03-25 11:08:47 +01:00

qwen2_moe

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

qwen2_vl

Support return_tensors in audio chat templates (#34601)

2025-03-25 11:08:47 +01:00

qwen3

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

qwen3_moe

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

rag

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

recurrent_gemma

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

reformer

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

regnet

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

rembert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

resnet

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

roberta

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

roberta_prelayernorm

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

roc_bert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

roformer

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

rt_detr

fix typos in the tests directory (#36932)

2025-03-25 10:49:24 +01:00

rt_detr_v2

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

rwkv

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

sam

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

seamless_m4t

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

seamless_m4t_v2

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

segformer

[tests] remove pt_tf equivalence tests (#36253)

2025-02-19 11:55:11 +00:00

seggpt

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

sew

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

sew_d

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

shieldgemma2

Support return_tensors in audio chat templates (#34601)

2025-03-25 11:08:47 +01:00

siglip

Refactor attention for SigLIP based models (#36981)

2025-04-01 15:37:25 +02:00

siglip2

Refactor attention for SigLIP based models (#36981)

2025-04-01 15:37:25 +02:00

smolvlm

Mark 2 tests as flaky for now (#37038)

2025-03-27 10:59:47 +01:00

speech_encoder_decoder

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

speech_to_text

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

speecht5

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

splinter

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

squeezebert

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

stablelm

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

starcoder2

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

superglue

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

superpoint

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

swiftformer

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

swin

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

swin2sr

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

swinv2

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

switch_transformers

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

t5

Export T5 (encoder-decoder) to ExecuTorch (#36486)

2025-03-31 12:10:26 +02:00

table_transformer

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

tapas

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

textnet

Fix tests for vision models (#35654)

2025-02-13 10:28:37 +00:00

time_series_transformer

Set weights_only in torch.load (#36991)

2025-03-27 14:55:50 +00:00

timesformer

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

timm_backbone

Less flaky for TimmBackboneModelTest::test_batching_equivalence (#35971)

2025-01-30 16:56:26 +01:00

timm_wrapper

Enable auto task for timm models in pipeline (#35531)

2025-01-08 15:14:17 +00:00

trocr

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

tvp

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

udop

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

umt5

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

unispeech

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

unispeech_sat

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

univnet

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

upernet

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

video_llava

fix typos in the tests directory (#36932)

2025-03-25 10:49:24 +01:00

videomae

Set weights_only in torch.load (#36991)

2025-03-27 14:55:50 +00:00

vilt

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

vipllava

Fix typos (#36910)

2025-03-24 14:08:29 +00:00

vision_encoder_decoder

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

vision_text_dual_encoder

[tests] remove flax-pt equivalence and cross tests (#36283)

2025-02-19 15:13:27 +00:00

visual_bert

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

vit

Fix tests for vision models (#35654)

2025-02-13 10:28:37 +00:00

vit_mae

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

vit_msn

[tests] Parameterized test_eager_matches_sdpa_inference (#36650)

2025-03-14 14:41:27 +00:00

vitdet

Fixed VitDet for non-squre Images (#35969)

2025-02-25 19:31:24 +00:00

vitmatte

chore: fix typos in the tests directory (#36813)

2025-03-21 10:20:05 +01:00

vitpose

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

vitpose_backbone

skip test_initialization for VitPoseBackboneModelTest for now (#36154)

2025-02-12 18:24:24 +01:00

vits

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

vivit

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

wav2vec2

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

wav2vec2_bert

Support return_tensors in audio chat templates (#34601)

2025-03-25 11:08:47 +01:00

wav2vec2_conformer

Fix typos in tests (#36547)

2025-03-05 15:04:06 -08:00

wav2vec2_phoneme

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

wav2vec2_with_lm

use torch.testing.assertclose instead to get more details about error in cis (#35659)

2025-01-24 16:55:28 +01:00

wavlm

Add missing atol to torch.testing.assert_close where rtol is specified (#36234)

2025-02-17 14:57:50 +01:00

whisper

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

x_clip

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

xglm

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

xlm

Fix more inefficient PT operations (#37060)

2025-03-31 16:31:24 +01:00

xlm_roberta

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

xlm_roberta_xl

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

xlnet

Use lru_cache for tokenization tests (#36818)

2025-03-28 15:09:35 +01:00

xmod

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

yolos

Add common test for torch.export and fix some vision models (#35124)

2025-02-11 11:37:31 +00:00

yoso

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

zamba

Mark 2 tests as flaky for now (#37038)

2025-03-27 10:59:47 +01:00

zamba2

CI: avoid human error, automatically infer generative models (#33212)

2025-02-13 16:27:11 +01:00

zoedepth

Remove low_cpu_mem_usage and _fast_init (#36963)

2025-03-31 17:18:43 +02:00

__init__.py

…