How harshh0107/llama_2_ft_10k-guanaco works
This repository is a packaged Llama 2 chat fine-tuning adapter, with adapter weights, tokenizer assets, and saved training metadata. The tree supports describing these deliverables, but provides no source code or README content showing how a user loads or runs them. Accordingly, the graph records the artifact roles without asserting runtime steps or relationships between files.
Components 6
Model adapter
- Adapter weights
Llama-2-7b-chat-finetune-10k-records/adapter_model.bin - Adapter configuration
Llama-2-7b-chat-finetune-10k-records/adapter_config.json
Tokenizer assets
- Tokenizer vocabulary
Llama-2-7b-chat-finetune-10k-records/tokenizer.json - Tokenizer configuration
Llama-2-7b-chat-finetune-10k-records/tokenizer_config.json - Special tokens
Llama-2-7b-chat-finetune-10k-records/special_tokens_map.json
Training metadata
- Training arguments
Llama-2-7b-chat-finetune-10k-records/training_args.bin