lhallee commited on
Commit
4602900
·
verified ·
1 Parent(s): 430aa02

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -2
README.md CHANGED
@@ -70,8 +70,9 @@ Average MSE for FP16: 0.00000140
70
 
71
  Average MSE for BF16: 0.00004125
72
 
73
- ### FlashAttention2
74
- Requires PyTorch 2.5+ for the most savings, see [SDPA](https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html).
 
75
 
76
  ### Citation
77
  ```
 
70
 
71
  Average MSE for BF16: 0.00004125
72
 
73
+ ### Inference speed
74
+ We look at various ESM models and their throughput on an H100. FastESM is over twice as fast as ESM2-650 with longer sequences. Requires PyTorch 2.5+ for the most savings, see [SDPA](https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html).
75
+ ![image/png](https://cdn-uploads.huggingface.co/production/uploads/62f2bd3bdb7cbd214b658c48/PvaBGfuJXEW2v_WLkt63y.png)
76
 
77
  ### Citation
78
  ```