llama-recipes
llama-recipes copied to clipboard
Scripts for fine-tuning Meta Llama with composable FSDP & PEFT methods to cover single/multi-node GPUs. Supports default & custom datasets for applications such as summarization and Q&A. Supporting a...
### π The feature, motivation and pitch I am new to llama-recipes. Right now I have finetuned a llama3 model based on "openbookqa" dataset. It store a model for me...
### System Info MacOS x86-64 AMD GPU ### Information - [X] The official example scripts - [ ] My own modified scripts ### π Describe the bug The nightly pip...
### π The feature, motivation and pitch FP8 is pretty main-stream - just ask Jensen - but I have no idea how to do the FP8 quantization using torch.float8_e4m3fn? Could...
LLaMA3 supports an 8K token context length. When continuously pretraining with proprietary data, the majority of the text data is significantly shorter than 8K tokens, resulting in a substantial amount...
Hello Worldβ‘β‘β‘
### π The feature, motivation and pitch Hello! I got the answer "Your request to access this repo has been rejected by the repo's authors" on huggingface,because I write wrong...
I have searched in β[Model Cards and Prompt Formats](https://llama.meta.com/docs/model-cards-and-prompt-formats/llama3_1/)β and in the Github repositories about what is the correct way to reference a file in the prompt to be used...
[PromptGuard](https://github.com/meta-llama/PurpleLlama/tree/main/Prompt-Guard) has been introduced as a system safety tool to be used to check LLM prompts for malicious text. This PR adds this guard to the safety_utils module and adjusts...
In the notebook, under section 4.3.2, LangChain Q&A Retriever, the code block where we test chat history has two commands switched in order, resulting in an error while running. Switching...
### π The feature, motivation and pitch Thank you for this notebook. It would be helpful to add suggested code for storing and retrieving text embeds with Faiss, and also...