Lukeam/aa_book

Enhanced training parameters for better fine-tuning

Updated gitignore and added enhanced Q&A dataset

Updated Q&A dataset with enhanced questions

Fix token authentication method

Update .gitignore and add test_model.py

Fix import error and rename test file

Add combined training and testing interface

Fix dataset processing and add proper data collator

Fix dataset formatting and tokenization

Fix type conversion in config validation

Update config for A10G hardware

Remove duplicate files from Documents folder

Update training configuration and add train.py implementation

Fix tokenizer padding configuration

Add training data file and update gitignore

Update to load local JSONL file directly

Add PEFT and LoRA for efficient fine-tuning

Add 8-bit quantization for memory efficiency

Add accelerate requirement

Fix training arguments configuration

Add error handling and debugging info

Add error handling and debugging info

Fix Gradio interface parameters

Update interface button text

Add training files

Add requirements file

Update gitignore to allow requirements.txt

Merge branch 'main' of https://huggingface.co/spaces/Lukeam/aa_book

Add training application file

Add training application file

initial commit