huggingface/transformers

Pull Requests Commits

Update get device properties and types (#38723)

ivarflakstad committed 1 year ago

Verified d3639d20

Small fixes amd (#38700)

remi-or committed 1 year ago

Verified 42207080

Added Expectations for AMD (mostly bnb) (#38697)

remi-or committed 1 year ago

Verified 122d85a4

Fixed device_propreties unpacking in common tests (#38698)

remi-or committed 1 year ago

Verified 228dc492

Image processor compile fix (#38540)

remi-or committed 1 year ago

Verified 18f810f1

Disable custom MRA kernels for ROCm (#38578)

ahadnagy committed 1 year ago

Verified 0006491f

Add Expectations for three AMD tests (#38581)

remi-or committed 1 year ago

Verified cdce7663

Pin Scipy version to >=1.12.0 (#38469)

ahadnagy committed 1 year ago

Verified 75dbc099

Janus seamless transfer (#38580)

remi-or committed 1 year ago

Verified 3febf1e9

Expectation changes and more AMD expectations (#38529)

remi-or committed 1 year ago

Verified 66dfa8c6

Relaxed the output_attention condition for ValueError (#38560)

remi-or committed 1 year ago

Verified 2bd967a3

Added guards against device mismatch errors in some LogitsProcessor classes (#38558)

remi-or committed 1 year ago

Verified f5eb52a2

Fixed a multiple-devices issue in SmolVLMModel (#38557)

remi-or committed 1 year ago

Verified 1e06c9cb

Uninstall `kernels` for AMD docker images (#38354)

ydshieh committed 1 year ago

Verified 6e306342

Hot fix for AMD CI workflow (#38349)

ydshieh committed 1 year ago

Verified 4a03044d

new failure CI reports for all jobs (#38298)

ydshieh committed 1 year ago

Verified d0c9c66d

[docs]: update roformer.md model card (#37946)

KsuParkhamchuk committed 1 year ago

Verified 31f8a0fe

docs(swinv2): Update SwinV2 model card to new standard format (#37942)

BryanBradfo committed 1 year ago

Verified 36f97ae1

Update BioGPT model card (#38214)

Aguedoom committed 1 year ago

Verified 33d23c39

Remove duplicate docstring: resample (#38305)

qqii committed 1 year ago

Verified dffb1180

Never fallback to eager implicitly (#38327)

Cyrilvallez committed 1 year ago

Verified e0aad278

Use Gradient Checkpointing Layer in Jamba & Blip Related Models (#38310)

alex-jw-brooks committed 1 year ago

Verified e64ed030

:rotating_light: :rotating_light: Inherited CausalLM Tests (#37590)

Rocketknight1 committed 1 year ago

Verified 53fb245e

Enhance Model Loading By Providing Parallelism, Uses Optional Env Flag (#36835)

inf3rnus committed 1 year ago

Verified d5f992f5

[`FlexAttention`] Reenable flex for encoder-decoder and make the test more robust (#38321)

vasqu committed 1 year ago

Verified 1ed19360

refactor can_save_slow_tokenizer (#37722)

itazap committed 1 year ago

Verified bb567d85

[performance_optim] reduce frequency of declaring attention_mask in Ascend NPU flash attention (#38278)

FightingZhen committed 1 year ago

Verified 3c289e21

🚨Early-error🚨 config will error out if `output_attentions=True` and the attn implementation is wrong (#38288)

ArthurZucker committed 1 year ago

Verified f5d45d89

Fix some tests (especially compile with fullgraph=True on Python<3.11) (#38319)

Cyrilvallez committed 1 year ago

Verified 896833c1

add `vasqu` to `self-comment-ci.yml` (#38324)

ydshieh committed 1 year ago

Verified a63bc174

Older