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

Add SDPA support for M2M100 (#33309)

Browse Source 
* Add SDPA support for M2M100

* [run_slow] m2m_100, nllb
This commit is contained in:
Isaac Schifferer
2024-09-25 10:04:42 -07:00
committed by GitHub
parent 68049b17a6
commit 61e98cb957
4 changed files with 178 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
docs/source/en/model_doc/m2m_100.md
View File

@@ -163,3 +163,21 @@ Below is an expected speedup diagram that compares pure inference time between t
<div style="text-align: center">
<img src="https://huggingface.co/datasets/visheratin/documentation-images/resolve/main/nllb-speedup.webp">
</div>
## Using Scaled Dot Product Attention (SDPA)
PyTorch includes a native scaled dot-product attention (SDPA) operator as part of `torch.nn.functional`. This function
encompasses several implementations that can be applied depending on the inputs and the hardware in use. See the
[official documentation](https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html)
or the [GPU Inference](https://huggingface.co/docs/transformers/main/en/perf_infer_gpu_one#pytorch-scaled-dot-product-attention)
page for more information.
SDPA is used by default for `torch>=2.1.1` when an implementation is available, but you may also set
`attn_implementation="sdpa"` in `from_pretrained()` to explicitly request SDPA to be used.
```python
from transformers import M2M100ForConditionalGeneration
model = M2M100ForConditionalGeneration.from_pretrained("facebook/m2m100_418M", torch_dtype=torch.float16, attn_implementation="sdpa")
...
```
For the best speedups, we recommend loading the model in half-precision (e.g. `torch.float16` or `torch.bfloat16`).

18
docs/source/en/model_doc/nllb.md
View File

@@ -188,3 +188,21 @@ Below is an expected speedup diagram that compares pure inference time between t
<div style="text-align: center">
<img src="https://huggingface.co/datasets/visheratin/documentation-images/resolve/main/nllb-speedup.webp">
</div>
## Using Scaled Dot Product Attention (SDPA)
PyTorch includes a native scaled dot-product attention (SDPA) operator as part of `torch.nn.functional`. This function
encompasses several implementations that can be applied depending on the inputs and the hardware in use. See the
[official documentation](https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html)
or the [GPU Inference](https://huggingface.co/docs/transformers/main/en/perf_infer_gpu_one#pytorch-scaled-dot-product-attention)
page for more information.
SDPA is used by default for `torch>=2.1.1` when an implementation is available, but you may also set
`attn_implementation="sdpa"` in `from_pretrained()` to explicitly request SDPA to be used.
```python
from transformers import AutoModelForSeq2SeqLM
model = AutoModelForSeq2SeqLM.from_pretrained("facebook/nllb-200-distilled-600M", torch_dtype=torch.float16, attn_implementation="sdpa")
...
```
For the best speedups, we recommend loading the model in half-precision (e.g. `torch.float16` or `torch.bfloat16`).