llama-recipes

llama-recipes copied to clipboard

fp16 support for Single GPU fine tuning

7

### 🚀 The feature, motivation and pitch I am running 3090 with 24GB VRAM and 16GB Shared memory ( Total is 40 GB ) When i am fine tuning 7B...

NiftyliuS

[Question] How to use llama2 for typical NLP tasks?

2

I am wondering how we could adapt the example.py files provided to undertake tasks such as: 1. Identify the extent of positive sentiment and negative sentiment in the following text....

santoshbs

Warning: Asking to pad to max_length but no maximum length is provided and the model has no predefined maximum length.

1

### System Info Cuda 12.1 PyTorch 2.3.0 Python 3.11 ``` Thu May 23 15:30:20 2024 +---------------------------------------------------------------------------------------+ | NVIDIA-SMI 545.23.06 Driver Version: 545.23.06 CUDA Version: 12.3 | |-----------------------------------------+----------------------+----------------------+ | GPU Name...

artkpv

Hi, I was using llama recipe [local inference](https://github.com/meta-llama/llama-recipes/tree/main/recipes/inference/local_inference). But I get the warning: `Token indices sequence length is longer than the specified maximum sequence length for this model (1998 >...

biaoyanf

can't reproduce llama3.1 evaluation results

5

### System Info [pip3] numpy==1.26.3 [pip3] torch==2.3.1+cu121 [pip3] torchaudio==2.3.1+cu121 [pip3] torchvision==0.18.1+cu121 [pip3] triton==2.3.1 [conda] numpy 1.26.3 pypi_0 pypi [conda] torch 2.3.1+cu121 pypi_0 pypi [conda] torchaudio 2.3.1+cu121 pypi_0 pypi [conda] torchvision...

zhuhr925

# What does this PR do? Updated the endpoint to 3.1 support. Also updated Langchain and Gradio support as their framework updated. ## Before submitting - [X] This PR fixes...

WuhanMonkey

Mismatch in Llama Guard 2 prompt between llama.meta.com and code - `begin_text`

2

### System Info This is independent of torch version. ### Information - [X] The official example scripts - [ ] My own modified scripts ### 🐛 Describe the bug The...

BrunoGomesCoelho

Update checkpoint_converter_fsdp_hf.py

5

# What does this PR do? This PR refactors the existing script to improve modularity, readability, and error handling. [src/llama_recipes/inference/checkpoint_converter_fsdp_hf.py] ### Description of the Change - **Modularization**: Introduced `get_model_name_from_yaml` and...

Mattral

LLama 3.1 gets stuck sometimes when producing output

1

### System Info PyTorch version: 2.2.2+cu121 Is debug build: False CUDA used to build PyTorch: 12.1 ROCM used to build PyTorch: N/A OS: Debian GNU/Linux 10 (buster) (x86_64) GCC version:...

Techbhatia

[Recipe] Example featuring built-in tool calling capabilities - Wolfram Alpha, Interpreter, Brave Search

4

# What does this PR do? Meta's latest Llama3.1 models offer unique function calling capabilities. In particular they offer built-in tool calling capabilities for the following 3 external tools: *...

tmoreau89