Skip to content

epfl-dlab/dlab-utils

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 

Repository files navigation

DLAB utils

Python Library for DLAB utils

Installation

You can install the dlabutils package using pip:

pip install git+https://github.com/epfl-dlab/dlab-utils.git

Usage

Model Paths

The model_path function returns the path to the model weights given the huggingface identifier, if the model is available in the DLAB LLM weights directory, and defaults to the hf model name if it is not. The available_models function returns a list of all available models. Here's a quick tutorial:

from dlabutils import model_path, available_models
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(model_path("meta-llama/Meta-Llama-3-8B-Instruct"))
tokenizer = AutoTokenizer.from_pretrained(model_path("meta-llama/Meta-Llama-3-8B-Instruct"))

print(available_models())

About

Python Library for DLAB utilities

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages