SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Add sdpa and fa2 the Wav2vec2 family. (#30121)

Browse Source 
* add sdpa to wav2vec.
Co-authored-by: kamilakesbi <kamil@huggingface.co>
Co-authored-by: jp1924 <jp42maru@gmail.com>

* add fa2 to wav2vec2

* add tests

* fix attention_mask compatibility with fa2

* minor dtype fix

* replace fa2 slow test

* fix fa2 slow test

* apply code review + add fa2 batch test

* add sdpa and fa2 to hubert

* sdpa and fa2 to data2vec_audio

* sdpa and fa2 to Sew

* sdpa to unispeech + unispeech sat

* small fix

* attention mask in tests

Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

* add_speedup_benchmark_to_doc

---------

Co-authored-by: kamil@huggingface.co <kamil.akesbi@gmail.com>
Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>
This commit is contained in:
Kamil Akesbi
2024-04-22 19:30:38 +02:00
committed by GitHub
parent 367a0dbd53
commit 569743f510
11 changed files with 2406 additions and 100 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
docs/source/en/perf_infer_gpu_one.md
View File

@@ -70,6 +70,12 @@ FlashAttention-2 is currently supported for the following architectures:
* [Qwen2](https://huggingface.co/docs/transformers/model_doc/qwen2#transformers.Qwen2Model)
* [Qwen2MoE](https://huggingface.co/docs/transformers/model_doc/qwen2_moe#transformers.Qwen2MoeModel)
* [Whisper](https://huggingface.co/docs/transformers/model_doc/whisper#transformers.WhisperModel)
* [Wav2Vec2](https://huggingface.co/docs/transformers/model_doc/wav2vec2#transformers.Wav2Vec2Model)
* [Hubert](https://huggingface.co/docs/transformers/model_doc/hubert#transformers.HubertModel)
* [data2vec_audio](https://huggingface.co/docs/transformers/main/en/model_doc/data2vec#transformers.Data2VecAudioModel)
* [Sew](https://huggingface.co/docs/transformers/main/en/model_doc/sew#transformers.SEWModel)
* [UniSpeech](https://huggingface.co/docs/transformers/v4.39.3/en/model_doc/unispeech#transformers.UniSpeechModel)
* [unispeech_sat](https://huggingface.co/docs/transformers/v4.39.3/en/model_doc/unispeech-sat#transformers.UniSpeechSatModel)
You can request to add FlashAttention-2 support for another model by opening a GitHub Issue or Pull Request.
@@ -203,6 +209,13 @@ For now, Transformers supports SDPA inference and training for the following arc
* [Qwen2MoE](https://huggingface.co/docs/transformers/model_doc/qwen2_moe#transformers.Qwen2MoeModel)
* [Musicgen](https://huggingface.co/docs/transformers/model_doc/musicgen#transformers.MusicgenModel)
* [MusicGen Melody](https://huggingface.co/docs/transformers/model_doc/musicgen_melody#transformers.MusicgenMelodyModel)
* [wav2vec2](https://huggingface.co/docs/transformers/model_doc/wav2vec2#transformers.Wav2Vec2Model)
* [Hubert](https://huggingface.co/docs/transformers/model_doc/hubert#transformers.HubertModel)
* [data2vec_audio](https://huggingface.co/docs/transformers/main/en/model_doc/data2vec#transformers.Data2VecAudioModel)
* [Sew](https://huggingface.co/docs/transformers/main/en/model_doc/sew#transformers.SEWModel)
* [UniSpeech](https://huggingface.co/docs/transformers/v4.39.3/en/model_doc/unispeech#transformers.UniSpeechModel)
* [unispeech_sat](https://huggingface.co/docs/transformers/v4.39.3/en/model_doc/unispeech-sat#transformers.UniSpeechSatModel)
<Tip>