update readme

This commit is contained in:
OlivierDehaene 2022-12-15 17:03:37 +01:00
parent 291722cb48
commit 82706a651c

View File

@ -17,6 +17,7 @@ to power Bloom, BloomZ and MT0-XXL api-inference widgets.
- 45ms per token generation for BLOOM with 8xA100 80GB - 45ms per token generation for BLOOM with 8xA100 80GB
- Logits warpers (temperature scaling, topk ...) - Logits warpers (temperature scaling, topk ...)
- Stop sequences - Stop sequences
- Log probabilities
## Officially supported models ## Officially supported models