|
Download examples/EditScore-train/README.md from chibifire/EditScore-code: direct link, hf CLI and curl.
- Browser
- Download file 2.82 kB
-
https://huggingface.co/chibifire/EditScore-code/resolve/main/examples/EditScore-train/README.md
- Command line
-
hf download hf://chibifire/EditScore-code/examples/EditScore-train/README.md
-
curl -L -o README.md https://huggingface.co/chibifire/EditScore-code/resolve/main/examples/EditScore-train/README.md
2.82 kB
| # EditScore Reward Model Training Guide | |
| This guide explains how to train EditScore reward models using LLaMA-Factory. | |
| ## 1. Environment Setup | |
| ### Clone LLaMA-Factory and Configure Virtual Environment | |
| ```bash | |
| git clone --depth 1 https://github.com/hiyouga/LLaMA-Factory.git | |
| cd LLaMA-Factory | |
| conda create -n llama-factory python=3.10 | |
| conda activate llama-factory | |
| pip install -e ".[torch,metrics]" --no-build-isolation | |
| ``` | |
| ## 2. Directory Structure Configuration | |
| Create necessary folders and files in the LLaMA-Factory root directory: | |
| ```bash | |
| # Create log and output directories | |
| mkdir -p logs | |
| mkdir -p output | |
| # Create training configuration directory | |
| mkdir -p examples/train_editscore | |
| # Copy training configuration files | |
| cp EditScore/examples/EditScore-train/config/*.yaml examples/train_editscore/ | |
| # Copy training script | |
| cp EditScore/examples/EditScore-train/train.sh . | |
| ``` | |
| ## 3. Dataset Registration | |
| Register the EditScore-Reward-Data dataset in `LLaMA-Factory/data/dataset_info.json`: | |
| ```json | |
| "EditScore-Reward-Data": { | |
| "file_name": "/path/to/your/reward.json", | |
| "formatting": "sharegpt", | |
| "columns": { | |
| "messages": "conversations", | |
| "images": "images" | |
| } | |
| } | |
| ``` | |
| ## 4. Training Configuration Description | |
| ### Single-Machine Training Configuration | |
| - `editscore_7B.yaml` - Train EditScore-7B model (single machine) | |
| - `editscore_qwen3_vl_4B_instruct.yaml` - Train EditScore_Qwen3_Vl_4B_Instruct model (single machine) | |
| - `editscore_qwen3_vl_8B_instruct.yaml` - Train EditScore_Qwen3_Vl_8B_Instruct model (single machine) | |
| ### Multi-Machine Training Configuration | |
| - `editscore_32B.yaml` - Train EditScore-32B model (two machines) | |
| - `editscore_72B.yaml` - Train EditScore-72B model (two machines) | |
| ## 5. Start Training | |
| ### Single-Machine Training | |
| ```bash | |
| # Modify experiment_name in train.sh to the corresponding configuration file name | |
| # For example: name=editscore_7B | |
| bash train.sh | |
| ``` | |
| ### Multi-Machine Training | |
| **Master node (rank=0):** | |
| ```bash | |
| bash train.sh --rank=0 --world_size=2 --master_addr=MASTER_NODE_IP --master_port=29500 | |
| ``` | |
| **Worker node (rank=1):** | |
| ```bash | |
| bash train.sh --rank=1 --world_size=2 --master_addr=MASTER_NODE_IP --master_port=29500 | |
| ``` | |
| ## 6. Parameter Configuration | |
| Users can modify the following parameters in the YAML configuration files as needed: | |
| - `per_device_train_batch_size`: Batch size per device | |
| - `gradient_accumulation_steps`: Gradient accumulation steps | |
| - `learning_rate`: Learning rate | |
| - `num_train_epochs`: Number of training epochs | |
| - `max_samples`: Maximum number of samples | |
| - `output_dir`: Output directory | |
| ## 7. Output Files | |
| After training completion, model files will be saved in the corresponding output directories: | |
| - Single-machine training: `LLaMA-Factory/output/model_name/` | |
| - Log files: `LLaMA-Factory/logs/experiment_name_rank.log` | |