llama-recipes

llama-recipes copied to clipboard

FLOPs counter seems doesn't work

8

### 🚀 The feature, motivation and pitch I am able to run the training with the FSDP. But then add the "--flop_counter" flag. It gives the following issue. Could someone...

mathmax12

RuntimeError: probability tensor contains either `inf`, `nan` or element < 0

5

### System Info Version I'm using: ``` Python 3.10.11 torch==2.4.1 torchaudio==2.4.1 torchvision==0.19.1 nvidia-cuda-cupti-cu12==12.1.105 nvidia-cuda-nvrtc-cu12==12.1.105 nvidia-cuda-runtime-cu12==12.1.105 2 GPU NVIDIA RTX 6000 Ada Gen of 50Gb each (total 100GB) Ubuntu 22.04 ```...

himanshushukla12

Adding data prep recipes from data-prep-kit

1

# What does this PR do? This PR adds a data prep recipe from data-prep-kit toolkit https://github.com/IBM/data-prep-kit Fixes # (issue) ## Feature/Issue validation/testing This is a notebook that has been...

Bytes-Explorer

Fine-tuning with ``--run_validation False`` still calls ``get_custom_dataset`` and does not save the model

3

### System Info llama-recipes v0.0.3 torch 2.4.1 ### Information - [X] The official example scripts - [X] My own modified scripts ### 🐛 Describe the bug I'm encountering two issues...

lucas-ventura

# What does this PR do? Adds a new recipe showcasing how to use Llama for automating data analytics and reporting tasks on an open-source repo. ## Before submitting -...

subramen

Updated typing extension library to be using latest so notebooks work with google collab # What does this PR do? Fixes # (issue) ## Feature/Issue validation/testing Please describe the tests...

varunfb

# What does this PR do? This PR loads a model with torch_dtype=auto instead of bfloat16 when we do not specify train_config.use_fp16. For llama models this will not make a...

mreso

Previously, the OpenLLM leaderboard V1 is hard to reproduce as it did not included a easy commend to run, so we created a customized eval.py to load a folder of...

wukaixingxp

This update enhances the script to not only check Python and Shell files but also include handling of all text files by using a more inclusive file extension pattern. This...

mandlinsarah

New metrics for benchmark

2

### 🚀 The feature, motivation and pitch Very nice work, Llama team. Llama is the most popular open-source LLM project and has been adopted for different platforms. It will be...

mathmax12