Hugging Face's logo

ctranslate2-4you
/
Llama-2-7b-chat-hf-ct2-int8

Transformers
ctranslate2
Inference Endpoints
Model card Files Community
ctranslate2-4you commited on
Commit
b6e1344
·
1 Parent(s): 54a6f3d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -0
README.md CHANGED
@@ -62,3 +62,4 @@ Information:
62
  | `int8` | 25% | 1.0 | Lower precision, suitable for whole numbers within a specific range. Often used where memory is crucial. |
63
 
64
  See what level of "compute" your Nvidia GPU supports: https://en.wikipedia.org/wiki/CUDA#GPUs_supported
 
 
62
  | `int8` | 25% | 1.0 | Lower precision, suitable for whole numbers within a specific range. Often used where memory is crucial. |
63
 
64
  See what level of "compute" your Nvidia GPU supports: https://en.wikipedia.org/wiki/CUDA#GPUs_supported
65
+ Also see: https://en.wikipedia.org/wiki/Bfloat16_floating-point_format#bfloat16_floating-point_format