huggingface/transformers

Pull Requests Commits

Fix FA performance regression (#47134)

andreasgoulas committed 61 days ago

Verified 1567e34b

tests only for the top 10 download models (#47244)

3outeille committed 61 days ago

Verified 7ca3c8d4

Fix InputTokensDetails missing cache_write_tokens for openai>=2.34.0 (#47248)

ydshieh committed 61 days ago

Verified 2593eed7

Revert "Trigger a scheduled run" (#47249)

ydshieh committed 61 days ago

Verified ae228e18

Fix CI read-only cache failures by patching cached_files in conftest (#47043)

ydshieh committed 61 days ago

Verified 750b08f5

Remove Executorch from CI until latest version is supported and fully tested on CI env (#47242)

IlyasMoutawwakil committed 61 days ago

Verified 8f9c9d15

Be more defensive with `remap_legacy_layer_types` for custom models (#47245)

hmellor committed 61 days ago

Verified 59a725dc

Fix DistributedConfig docstring for unimplemented sp_plan (#47237)

3outeille committed 61 days ago

Verified 0066dee2

Switch mlinter to 0.1.2 (#47172)

tarekziade committed 62 days ago

Verified 0bc35541

trainer: clear MPS graph cache via torch_empty_cache_steps (#45818)

anagnorisis2peripeteia committed 62 days ago

Verified e3e82230

Trigger a scheduled run (#47209)

ydshieh committed 62 days ago

Verified 464f4730

Make executorch exporter tests always use xnnpack backend (#47201)

tarekziade committed 62 days ago

Verified bcccc3e7

No agent PR descriptions (#45790)

Rocketknight1 committed 62 days ago

Verified 8aa70f78

Clarify that max_steps is required for datasets without __len__ (#47155)

albertvillanova committed 62 days ago

Verified fab8993d

Cleanup pipelines, stop materializing generators (#47142)

Rocketknight1 committed 62 days ago

Verified 1293adfd

[Fix] Remove deprecated argument from `kernels` call (#47100)

remi-or committed 62 days ago

Verified 3fa92a08

Fix device_map computation when the no_split_modules have different sizes (#47203)

Cyrilvallez committed 62 days ago

Verified 6d960ca0

Remove executorch from all-latest-gpu image + add torch smoke test (#47196)

tarekziade committed 62 days ago

Verified be5e160c

Fix crash in greedy assisted generation with different tokenizers (#46936)

Sunt-ing committed 63 days ago

Verified 8ca0fbec

[Fix] Make DeepGEMM triton fallback more robust (#47126)

remi-or committed 63 days ago

Verified 17a09eeb

[Generation] Add static ensemble verification for lossy speculative decoding (#45979)

kasakh committed 63 days ago

Verified 753de4f6

Add native FSDP2 module + migration (#46707)

3outeille committed 63 days ago

Verified be7a11a8

Fix experts implementation in two spots (#47097)

remi-or committed 63 days ago

Verified ba6f4983

[Fix] Remove old automatic cross attn pattern from output recorders (#47117)

remi-or committed 63 days ago

Verified 4d828999

🌐 [i18n-KO] Translate accelerator_selection.md to Korean (#47157)

kkwjk2718 committed 63 days ago

Verified d45139cf

[i18n-KO] Translate optimum.md to Korean and fix Furiosa typo (#47156)

kkwjk2718 committed 63 days ago

Verified eb36335d

[docs] fix curly quotes rendering to straight quotes (#47135)

clijo committed 63 days ago

Verified 1f2fd058

Fix custom code which doesn't know about the new linear layer type names (#47174)

hmellor committed 63 days ago

Verified 4db1cacb

Reject path traversal in the `transformers_weights` config field (#46890)

LinZiyuu committed 63 days ago

Verified bce8fd08

[docs] Custom code conversion mapping (#47114)

stevhliu committed 63 days ago

Verified 3bbac3e0

Newer Older