65 lines
3.7 KiB
Markdown
65 lines
3.7 KiB
Markdown
# Reproduction guide
|
|
|
|
This directory contains the necessary information and assets to reproduce the results obtained during this Heretic run.
|
|
|
|
## Models
|
|
|
|
- **Base model:** [Qwen/Qwen2.5-Coder-3B](https://huggingface.co/Qwen/Qwen2.5-Coder-3B) (Commit: [`09d9bc5`](https://huggingface.co/Qwen/Qwen2.5-Coder-3B/commit/09d9bc5d376b0cfa0100a0694ea7de7232525803))
|
|
|
|
## Datasets
|
|
|
|
- **Good prompts:** [mlabonne/harmless_alpaca](https://huggingface.co/datasets/mlabonne/harmless_alpaca) (Commit: [`02c6a92`](https://huggingface.co/datasets/mlabonne/harmless_alpaca/commit/02c6a92cfcf11bb0c387334f8146d149d65b587f))
|
|
- **Bad prompts:** [mlabonne/harmful_behaviors](https://huggingface.co/datasets/mlabonne/harmful_behaviors) (Commit: [`01cead0`](https://huggingface.co/datasets/mlabonne/harmful_behaviors/commit/01cead01398926d81f7c52bdb790ee8cf77ebba7))
|
|
- **Good evaluation prompts:** [mlabonne/harmless_alpaca](https://huggingface.co/datasets/mlabonne/harmless_alpaca) (Commit: [`02c6a92`](https://huggingface.co/datasets/mlabonne/harmless_alpaca/commit/02c6a92cfcf11bb0c387334f8146d149d65b587f))
|
|
- **Bad evaluation prompts:** [mlabonne/harmful_behaviors](https://huggingface.co/datasets/mlabonne/harmful_behaviors) (Commit: [`01cead0`](https://huggingface.co/datasets/mlabonne/harmful_behaviors/commit/01cead01398926d81f7c52bdb790ee8cf77ebba7))
|
|
|
|
## Selected trial
|
|
|
|
- **Trial number:** 136
|
|
- **KL divergence:** 0.062553
|
|
- **Refusals:** 4/100
|
|
|
|
## System
|
|
|
|
- **Python:** 3.12.11 (CPython, GCC 11.2.0) [Conda]
|
|
- **Operating system:** Linux-6.11.0-1016-nvidia-x86_64-with-glibc2.39 (x86_64)
|
|
- **CPU:** Intel(R) Xeon(R) Platinum 8468
|
|
|
|
### Accelerators
|
|
|
|
- **CUDA:** Detected 1 device(s) (139.80 GB total VRAM)
|
|
- **CUDA Version:** 12.8
|
|
- **Driver Version:** 580.126.09
|
|
- **Devices:**
|
|
- **CUDA 0:** NVIDIA H200 (139.80 GB)
|
|
|
|
## Environment
|
|
|
|
- **Heretic:** v1.3.0 (Origin: PyPI)
|
|
- **PyTorch:** 2.8.0+cu128
|
|
- **Other dependencies:** See [`requirements.txt`](requirements.txt).
|
|
|
|
## Contents of this directory
|
|
|
|
- [`requirements.txt`](requirements.txt): The exact versions of all Python packages.
|
|
- [`config.toml`](config.toml): The exact configuration used, including the RNG seed.
|
|
- [`Qwen--Qwen2--5-Coder-3B.jsonl`](Qwen--Qwen2--5-Coder-3B.jsonl): The Optuna study journal containing the history of all trials.
|
|
- [`SHA256SUMS`](SHA256SUMS): Cryptographic hashes for all weight files.
|
|
- [`reproduce.json`](reproduce.json): A machine-readable file containing all reproducibility information.
|
|
|
|
## How to reproduce
|
|
|
|
1. Ensure your system matches the specifications in the **System** section above. Exact reproducibility is only guaranteed if all aspects of your system are identical to the one the model was originally generated on.
|
|
1. Install the exact version of Heretic indicated in the **Environment** section above, from its original source.
|
|
1. Install the packages listed in `requirements.txt`: `pip install -r requirements.txt`
|
|
1. Install the correct version of PyTorch: `pip install torch==2.8.0+cu128 --index-url https://download.pytorch.org/whl/cu128`
|
|
1. Place the provided `config.toml` in your working directory.
|
|
1. Run Heretic without any additional arguments: `heretic`
|
|
1. Wait for the run to finish, then select trial **136** and export the model.
|
|
1. Verify that the weight files have been exactly reproduced by comparing their SHA-256 hashes against those in `SHA256SUMS`: `sha256sum -c SHA256SUMS` (or look at the hashes online if you uploaded to Hugging Face)
|
|
|
|
> [!TIP]
|
|
> To use the included Optuna study journal `Qwen--Qwen2--5-Coder-3B.jsonl`, place it in the checkpoints directory (usually `checkpoints/`) before running Heretic.
|
|
>
|
|
> This allows you to export other models from the Pareto front, or to run additional trials without having to re-run the stored trials.
|