Text Generation
Transformers
PyTorch
RefinedWeb
custom_code
text-generation-inference
psinger commited on
Commit
d9f7eff
•
1 Parent(s): 20f87e8

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -14,7 +14,7 @@ license: apache-2.0
14
 
15
  The following changes have been made:
16
  - Fixing generation configuration setting
17
- - Model now properly uses specified ```attention_mask``` when calling ```scaled_dot_product_attention``` - this also allows to specify custom attention masks and work with left-padded input.
18
 
19
 
20
  # 🚀 Falcon-40B
 
14
 
15
  The following changes have been made:
16
  - Fixing generation configuration setting
17
+ - Model now properly uses specified ```attention_mask``` when calling ```scaled_dot_product_attention``` - this also allows to specify custom attention masks and work with left-padded input. However, this will disable additional memory and flash optimization.
18
 
19
 
20
  # 🚀 Falcon-40B