llama-recipes

llama-recipes copied to clipboard

Configs like `train_config` should be named using PascalCase

2

### 🚀 The feature, motivation and pitch Config dataclasses like `llama_recipes.configs.training.train_config` should be named using PascalCase, in this case `TrainConfig`. The current naming violates a widely accepted Python convention. It...

vvolhejn

Added a Gradio UI for multi-modal inferencing using Llama 3.2 Vision/

2

What does this PR do? This PR introduces multi-modal inference using the Gradio UI for Llama 3.2 vision models. The Gradio UI allows users to upload images and generate descriptive...

himanshushukla12

Initial Crusoe examples to 3p_integrations recipes

2

# What does this PR do? This PR adds a new section in 3p_integrations for deploying Llama solutions on [Crusoe Cloud](https://crusoe.ai/). ## Feature/Issue validation/testing For easy reproducibility, Crusoe recipes use...

ethxnp

Recipe for fine-tuning with lower resolution images

8

### 🚀 The feature, motivation and pitch Is fine-tuning the Vision models at a lower resolution supported? If so, can you please add a recipe for that(or add a note...

avaradarajanfigma

add support for llama vision model conversion

2

# What does this PR do? Updated the script to support converting finetuned llama 3.2 vision model to HF format, so it works with multimodal inference. ## Feature/Issue validation/testing Tested...

roywei

Learning rate scheduler

3

### 🚀 The feature, motivation and pitch I don't see any option to set up a learning scheduler in the fine-tuning input arguments. Is there a way to implement it?...

NicoZenith

Llama 3.2-11B-vision fully fine-tuned model file question

2

During the use of LoRA fine-tuning, everything was normal, but the following issue arose during full-scale fine-tuning. I use the following script for full fine-tuning : ```shell #!/bin/bash NNODES=1 NPROC_PER_NODE=4...

Kidand

add support for ingesting content from websites, audio files, YouTube, etc.

3

This update include adding support for multiple input formats, improving documentation, and updating dependencies. ### Major Enhancements: #### Support for Multiple Input Formats: 1. **Ingestion Module**: Added an ingestion module...

evanjcosgrove

Found two forward recomputation exist in a single backward when using FSDP with activation checkpointing

1

### System Info PyTorch version: 2.2.0 Is debug build: False CUDA used to build PyTorch: 11.8 GPU: A100 PCIe * 4 transformers: 4.45.2 ### Information - [X] The official example...

mingyuanw-mt

Project license?

1

Hey, first of all thanks for your wonderful project! I've noticed this repo lacks a LICENSE file, which is the usual way of specifying the project's license. I also see...

notpushkin