improve readme

This commit is contained in:
OlivierDehaene 2023-02-14 12:13:38 +01:00
parent 97f9ae6a6d
commit 7a0bbf0994

View File

@ -49,17 +49,17 @@ to power LLMs api-inference widgets.
- Log probabilities - Log probabilities
- Distributed tracing with Open Telemetry - Distributed tracing with Open Telemetry
## Officially supported models ## Officially supported architectures
- [BLOOM](https://huggingface.co/bigscience/bloom) - [BLOOM](https://huggingface.co/bigscience/bloom)
- [BLOOMZ](https://huggingface.co/bigscience/bloomz) - [BLOOMZ](https://huggingface.co/bigscience/bloomz)
- [MT0-XXL](https://huggingface.co/bigscience/mt0-xxl) - [MT0-XXL](https://huggingface.co/bigscience/mt0-xxl)
- ~~[Galactica](https://huggingface.co/facebook/galactica-120b)~~ (deactivated) - [Galactica](https://huggingface.co/facebook/galactica-120b)
- [SantaCoder](https://huggingface.co/bigcode/santacoder) - [SantaCoder](https://huggingface.co/bigcode/santacoder)
- [GPT-Neox 20B](https://huggingface.co/EleutherAI/gpt-neox-20b) - [GPT-Neox 20B](https://huggingface.co/EleutherAI/gpt-neox-20b)
- [FLAN-T5-XXL](https://huggingface.co/google/flan-t5-xxl) - [FLAN-T5-XXL](https://huggingface.co/google/flan-t5-xxl)
Other models are supported on a best effort basis using: Other architectures are supported on a best effort basis using:
`AutoModelForCausalLM.from_pretrained(<model>, device_map="auto")` `AutoModelForCausalLM.from_pretrained(<model>, device_map="auto")`
@ -191,7 +191,7 @@ Be aware that the official Docker image has them enabled by default.
### Download ### Download
First you need to download the weights: It is advised to download the weights ahead of time with the following command:
```shell ```shell
make download-bloom make download-bloom