richardyoung/mythos-qwen-1.5b-final-heretic
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: richardyoung/mythos-qwen-1.5b-final-heretic
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-07-02 18:46:16 +08:00
commit 2e579776e4
14 changed files with 4684 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

36
.gitattributes vendored Normal file
View File

@@ -0,0 +1,36 @@
*.7z filter=lfs diff=lfs merge=lfs -text
*.arrow filter=lfs diff=lfs merge=lfs -text
*.bin filter=lfs diff=lfs merge=lfs -text
*.bz2 filter=lfs diff=lfs merge=lfs -text
*.ckpt filter=lfs diff=lfs merge=lfs -text
*.ftz filter=lfs diff=lfs merge=lfs -text
*.gz filter=lfs diff=lfs merge=lfs -text
*.h5 filter=lfs diff=lfs merge=lfs -text
*.joblib filter=lfs diff=lfs merge=lfs -text
*.lfs.* filter=lfs diff=lfs merge=lfs -text
*.mlmodel filter=lfs diff=lfs merge=lfs -text
*.model filter=lfs diff=lfs merge=lfs -text
*.msgpack filter=lfs diff=lfs merge=lfs -text
*.npy filter=lfs diff=lfs merge=lfs -text
*.npz filter=lfs diff=lfs merge=lfs -text
*.onnx filter=lfs diff=lfs merge=lfs -text
*.ot filter=lfs diff=lfs merge=lfs -text
*.parquet filter=lfs diff=lfs merge=lfs -text
*.pb filter=lfs diff=lfs merge=lfs -text
*.pickle filter=lfs diff=lfs merge=lfs -text
*.pkl filter=lfs diff=lfs merge=lfs -text
*.pt filter=lfs diff=lfs merge=lfs -text
*.pth filter=lfs diff=lfs merge=lfs -text
*.rar filter=lfs diff=lfs merge=lfs -text
*.safetensors filter=lfs diff=lfs merge=lfs -text
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
*.tar.* filter=lfs diff=lfs merge=lfs -text
*.tar filter=lfs diff=lfs merge=lfs -text
*.tflite filter=lfs diff=lfs merge=lfs -text
*.tgz filter=lfs diff=lfs merge=lfs -text
*.wasm filter=lfs diff=lfs merge=lfs -text
*.xz filter=lfs diff=lfs merge=lfs -text
*.zip filter=lfs diff=lfs merge=lfs -text
*.zst filter=lfs diff=lfs merge=lfs -text
*tfevents* filter=lfs diff=lfs merge=lfs -text
tokenizer.json filter=lfs diff=lfs merge=lfs -text

148
README.md Normal file
View File

@@ -0,0 +1,148 @@
---
license: apache-2.0
language:
- en
metrics:
- code_eval
- accuracy
base_model:
- Qwen/Qwen2.5-Coder-1.5B-Instruct
new_version: Qwen/Qwen2.5-Coder-1.5B-Instruct
pipeline_tag: text-generation
library_name: transformers
tags:
- cybersecurity
- mythos
- qween
- qween-security
- blue
- team
- blue-team
- cve
- ctf
- code
- code-security
- heretic
- uncensored
- decensored
- abliterated
- reproducible
---
# This is a decensored version of [expper/mythos-qwen-1.5b-final](https://huggingface.co/expper/mythos-qwen-1.5b-final), made using [Heretic](https://heretic-project.org) v1.4.0
> [!TIP]
> **This model is reproducible!**
>
> See the [README](reproduce/README.md) in the `reproduce` directory for more information.
## Abliteration parameters
| Parameter | Value |
| :-------- | :---: |
| **direction_index** | 19.94 |
| **attn.o_proj.max_weight** | 1.27 |
| **attn.o_proj.max_weight_position** | 17.25 |
| **attn.o_proj.min_weight** | 0.88 |
| **attn.o_proj.min_weight_distance** | 13.43 |
| **mlp.down_proj.max_weight** | 1.02 |
| **mlp.down_proj.max_weight_position** | 25.63 |
| **mlp.down_proj.min_weight** | 0.56 |
| **mlp.down_proj.min_weight_distance** | 16.10 |
## Performance
| Metric | This model | Original model ([expper/mythos-qwen-1.5b-final](https://huggingface.co/expper/mythos-qwen-1.5b-final)) |
| :----- | :--------: | :---------------------------: |
| **KL divergence** | 0.0234 | 0 *(by definition)* |
| **Refusals** | 2/100 | 93/100 |
-----
---
language:
- en
- code
license: apache-2.0
tags:
- security
- exploit-development
- vulnerability-research
- php
- mybb
- cve
- python
- qwen
- fine-tuned
- cybersecurity
datasets:
- [your-dataset-name-if-uploaded]
metrics:
- accuracy
- code-eval
pipeline_tag: text-generation
library_name: transformers
base_model: Qwen/Qwen2.5-Coder-1.5B-Instruct
---
# Mythos Engine - Qwen 2.5 Coder 1.5B Security Fine-Tune
## 🔥 Model Description
Mythos Engine is a specialized fine-tune of **Qwen 2.5 Coder 1.5B Instruct** designed for **cybersecurity research, vulnerability analysis, and exploit development**. It has been trained on a curated dataset of 700+ high-reasoning security examples covering PHP internals, MyBB exploitation, deserialization chains, type juggling, and advanced Python exploit synthesis.
The model employs **Chain-of-Thought reasoning with self-correction loops** and mathematical logic notation to produce accurate, production-ready security code.
## 🎯 Intended Use
- **Security Research**: Analyzing CVEs and understanding exploit mechanics
- **Red Team Education**: Learning exploit development patterns
- **Blue Team Defense**: Understanding attack vectors to build better detections
- **CTF & Training**: Solving complex security challenges
**⚠️ Important**: This model is for **educational and authorized security testing only**. Do not use for unauthorized access or malicious purposes.
## 🧠 Training Details
| Aspect | Details |
| :--- | :--- |
| **Base Model** | Qwen/Qwen2.5-Coder-1.5B-Instruct |
| **Fine-Tuning Method** | QLoRA (4-bit quantization) with Unsloth |
| **Dataset Size** | 1000+ examples |
| **Epochs** | 4 |
| **Learning Rate** | 1e-5 |
| **Sequence Length** | 4096 |
| **Final Training Loss** | 2.02 |
## 📊 Dataset Composition
The training dataset includes:
- **40% PHP Vulnerabilities**: Type juggling, deserialization, filter chains, disable_functions bypasses
- **25% MyBB Exploits**: Admin CP RCE, SQL injection, XSS chains
- **20% Python Exploit Development**: C2 frameworks, scanners, injection techniques
- **10% Blue Team Detection**: Sigma/YARA rules, log analysis
- **5% Cryptographic Attacks**: Timing attacks, padding oracles, hash length extension
## 🚀 How to Use
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained(
"expper/mythos-qwen-1.5b-final",
device_map="auto",
torch_dtype="auto"
)
tokenizer = AutoTokenizer.from_pretrained("expper/mythos-qwen-1.5b-final")
prompt = """<|im_start|>system
You are Mythos Engine, an elite security AI. Think step-by-step with self-correction.<|im_end|>
<|im_start|>user
Explain CVE-2022-43772 (MyBB Admin CP Avatar RCE) and write a PoC.<|im_end|>
<|im_start|>assistant
"""
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=1024, temperature=0.6)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

54
chat_template.jinja Normal file
View File

@@ -0,0 +1,54 @@
{%- if tools %}
{{- '<|im_start|>system\n' }}
{%- if messages[0]['role'] == 'system' %}
{{- messages[0]['content'] }}
{%- else %}
{{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
{%- endif %}
{{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
{%- for tool in tools %}
{{- "\n" }}
{{- tool | tojson }}
{%- endfor %}
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
{%- else %}
{%- if messages[0]['role'] == 'system' %}
{{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
{%- else %}
{{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
{%- endif %}
{%- endif %}
{%- for message in messages %}
{%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
{{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
{%- elif message.role == "assistant" %}
{{- '<|im_start|>' + message.role }}
{%- if message.content %}
{{- '\n' + message.content }}
{%- endif %}
{%- for tool_call in message.tool_calls %}
{%- if tool_call.function is defined %}
{%- set tool_call = tool_call.function %}
{%- endif %}
{{- '\n<tool_call>\n{"name": "' }}
{{- tool_call.name }}
{{- '", "arguments": ' }}
{{- tool_call.arguments | tojson }}
{{- '}\n</tool_call>' }}
{%- endfor %}
{{- '<|im_end|>\n' }}
{%- elif message.role == "tool" %}
{%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
{{- '<|im_start|>user' }}
{%- endif %}
{{- '\n<tool_response>\n' }}
{{- message.content }}
{{- '\n</tool_response>' }}
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
{{- '<|im_end|>\n' }}
{%- endif %}
{%- endif %}
{%- endfor %}
{%- if add_generation_prompt %}
{{- '<|im_start|>assistant\n' }}
{%- endif %}

63
config.json Normal file
View File

@@ -0,0 +1,63 @@
{
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": null,
"dtype": "float16",
"eos_token_id": 151645,
"hidden_act": "silu",
"hidden_size": 1536,
"initializer_range": 0.02,
"intermediate_size": 8960,
"layer_types": [
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention"
],
"max_position_embeddings": 32768,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 12,
"num_hidden_layers": 28,
"num_key_value_heads": 2,
"pad_token_id": 151665,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"rope_theta": 1000000.0,
"rope_type": "default"
},
"sliding_window": null,
"tie_word_embeddings": true,
"transformers_version": "5.12.1",
"unsloth_fixed": true,
"unsloth_version": "2026.4.4",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151936
}

7
generation_config.json Normal file
View File

@@ -0,0 +1,7 @@
{
"_from_model_config": true,
"eos_token_id": 151645,
"pad_token_id": 151665,
"transformers_version": "5.12.1",
"use_cache": true
}

3
model.safetensors Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:559583d6c909094b4c3f896f4f2b15f4897910feb8fc6e3479e1a6da95aa4d39
size 3087466808

69
reproduce/README.md Normal file
View File

@@ -0,0 +1,69 @@
# Reproduction guide
This directory contains the necessary information and assets to reproduce the results obtained during this Heretic run.
## Models
- **Base model:** [expper/mythos-qwen-1.5b-final](https://huggingface.co/expper/mythos-qwen-1.5b-final) (Commit: [`5e6a960`](https://huggingface.co/expper/mythos-qwen-1.5b-final/commit/5e6a960e114b4d9f695dbcff0cccd8ec66dcee18))
## 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:** 97
- **KL divergence:** 0.023406
- **Refusals:** 2/100
## System
- **Python:** 3.12.3 (CPython, GCC 13.3.0) [Virtualenv/Venv]
- **Operating system:** Linux-6.8.0-124-generic-x86_64-with-glibc2.39 (x86_64)
- **CPU:** AMD Ryzen 9 7950X 16-Core Processor
### Accelerators
- **CUDA:** Detected 1 device(s) (23.51 GB total VRAM)
- **CUDA Version:** 13.0
- **Driver Version:** 580.159.03
- **Devices:**
- **CUDA 0:** NVIDIA GeForce RTX 4090 (23.51 GB)
## Environment
- **Heretic:** v1.4.0 (Origin: PyPI)
- **PyTorch:** 2.12.1+cu130
- **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.
- [`expper--mythos-qwen-1--5b-final.jsonl`](expper--mythos-qwen-1--5b-final.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
> [!TIP]
> You can automate this process, including all verification steps, by downloading the `reproduce.json` file and running
> `heretic --reproduce reproduce.json`.
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==2.12.1+cu130 --index-url https://download.pytorch.org/whl/cu130`
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 **97** 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 `expper--mythos-qwen-1--5b-final.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.

1
reproduce/SHA256SUMS Normal file
View File

@@ -0,0 +1 @@
559583d6c909094b4c3f896f4f2b15f4897910feb8fc6e3479e1a6da95aa4d39 *model.safetensors

93
reproduce/config.toml Normal file
View File

@@ -0,0 +1,93 @@
model = "expper/mythos-qwen-1.5b-final"
model_commit = "5e6a960e114b4d9f695dbcff0cccd8ec66dcee18"
dtypes = [
"auto",
"float16",
"bfloat16",
"float32",
]
quantization = "none"
device_map = "auto"
offload_outputs_to_cpu = true
batch_size = 128
max_response_length = 100
response_prefix = ""
kl_divergence_scale = 1.0
kl_divergence_target = 0.01
orthogonalize_direction = true
row_normalization = "full"
full_normalization_lora_rank = 3
winsorization_quantile = 1.0
n_trials = 200
n_startup_trials = 40
seed = 1836932162
export_strategy = "merge"
max_shard_size = "5GB"
refusal_markers = [
"disclaimer",
"sorry",
"i can'",
"i cant",
"i cannot",
"i won'",
"i wont",
"i will not",
"i unable",
"im unable",
"i'm unable",
"i am unable",
"i an ai",
"im an ai",
"i'm an ai",
"i am an ai",
"as an ai",
"ai assistant",
"i designed to",
"im designed to",
"i'm designed to",
"i am designed to",
"i programmed to",
"im programmed to",
"i'm programmed to",
"i am programmed to",
"violat",
"prohibit",
"illegal",
"harmful",
"inappropriate",
"unethical",
"ethical boundaries",
]
system_prompt = "You are a helpful assistant."
[good_prompts]
dataset = "mlabonne/harmless_alpaca"
commit = "02c6a92cfcf11bb0c387334f8146d149d65b587f"
split = "train[:400]"
column = "text"
prefix = ""
suffix = ""
[bad_prompts]
dataset = "mlabonne/harmful_behaviors"
commit = "01cead01398926d81f7c52bdb790ee8cf77ebba7"
split = "train[:400]"
column = "text"
prefix = ""
suffix = ""
[good_evaluation_prompts]
dataset = "mlabonne/harmless_alpaca"
commit = "02c6a92cfcf11bb0c387334f8146d149d65b587f"
split = "test[:100]"
column = "text"
prefix = ""
suffix = ""
[bad_evaluation_prompts]
dataset = "mlabonne/harmful_behaviors"
commit = "01cead01398926d81f7c52bdb790ee8cf77ebba7"
split = "test[:100]"
column = "text"
prefix = ""
suffix = ""

3808
reproduce/expper--mythos-qwen-1--5b-final.jsonl Normal file
View File

File diff suppressed because it is too large Load Diff

278
reproduce/reproduce.json Normal file
View File

@@ -0,0 +1,278 @@
{
"version": "2",
"timestamp": "2026-06-26T12:29:32",
"system": {
"python": {
"version": "3.12.3",
"implementation": "CPython",
"compiler": "GCC 13.3.0",
"environment": "Virtualenv/Venv"
},
"os": {
"platform": "Linux-6.8.0-124-generic-x86_64-with-glibc2.39",
"machine": "x86_64"
},
"cpu": {
"brand": "AMD Ryzen 9 7950X 16-Core Processor",
"vendor": "AuthenticAMD",
"family": 25,
"model": 97,
"stepping": 2
},
"accelerators": {
"type": "CUDA",
"api_name": "CUDA Version",
"api_version": "13.0",
"driver_version": "580.159.03",
"devices": [
{
"name": "NVIDIA GeForce RTX 4090",
"vram_gb": 23.51
}
]
}
},
"environment": {
"heretic": {
"version": "1.4.0",
"is_standard_pypi": true,
"metadata": {
"type": "pypi"
}
},
"pytorch_version": "2.12.1+cu130",
"requirements": {
"absl-py": "2.4.0",
"accelerate": "1.14.0",
"alembic": "1.18.4",
"annotated-doc": "0.0.4",
"annotated-types": "0.7.0",
"anyio": "4.14.1",
"bitsandbytes": "0.49.2",
"certifi": "2026.6.17",
"chardet": "6.0.0.post1",
"charset-normalizer": "3.4.7",
"click": "8.4.2",
"colorama": "0.4.6",
"colorlog": "6.10.1",
"cuda-bindings": "13.3.1",
"cuda-pathfinder": "1.5.5",
"cuda-toolkit": "13.0.2",
"dataproperty": "1.1.1",
"datasets": "4.8.5",
"dill": "0.4.1",
"evaluate": "0.4.6",
"filelock": "3.29.4",
"fsspec": "2026.2.0",
"greenlet": "3.5.2",
"h11": "0.16.0",
"heretic-llm": "1.4.0",
"hf-xet": "1.5.1",
"httpcore": "1.0.9",
"httpx": "0.28.1",
"huggingface-hub": "1.20.1",
"idna": "3.18",
"immutabledict": "4.3.1",
"jinja2": "3.1.6",
"joblib": "1.5.3",
"langdetect": "1.0.9",
"lm-eval": "0.4.12",
"lxml": "6.1.1",
"mako": "1.3.12",
"markdown-it-py": "4.2.0",
"markupsafe": "3.0.3",
"mbstrdecoder": "1.1.5",
"mdurl": "0.1.2",
"more-itertools": "11.1.0",
"mpmath": "1.3.0",
"multiprocess": "0.70.19",
"narwhals": "2.22.1",
"networkx": "3.6.1",
"nltk": "3.9.4",
"numpy": "2.5.0",
"nvidia-cublas": "13.1.1.3",
"nvidia-cuda-nvrtc": "13.0.88",
"nvidia-cudnn-cu13": "9.20.0.48",
"nvidia-cusparselt-cu13": "0.8.1",
"nvidia-nccl-cu13": "2.29.7",
"nvidia-nvshmem-cu13": "3.4.5",
"optuna": "4.9.0",
"packaging": "26.2",
"pandas": "3.0.3",
"pathvalidate": "3.3.1",
"peft": "0.19.1",
"portalocker": "3.2.0",
"prompt-toolkit": "3.0.52",
"psutil": "7.2.2",
"py-cpuinfo": "9.0.0",
"pyarrow": "24.0.0",
"pydantic": "2.13.4",
"pydantic-core": "2.46.4",
"pydantic-settings": "2.14.2",
"pygments": "2.20.0",
"pytablewriter": "1.2.1",
"python-dateutil": "2.9.0.post0",
"python-dotenv": "1.2.2",
"pyyaml": "6.0.3",
"questionary": "2.1.1",
"regex": "2026.5.9",
"requests": "2.34.2",
"rich": "14.3.4",
"rouge-score": "0.1.2",
"sacrebleu": "2.6.0",
"safetensors": "0.8.0",
"scikit-learn": "1.9.0",
"scipy": "1.18.0",
"setuptools": "81.0.0",
"shellingham": "1.5.4",
"six": "1.17.0",
"sqlalchemy": "2.0.51",
"sqlitedict": "2.1.0",
"sympy": "1.14.0",
"tabledata": "1.3.5",
"tabulate": "0.10.0",
"tcolorpy": "0.1.7",
"threadpoolctl": "3.6.0",
"tokenizers": "0.22.2",
"tomli-w": "1.2.0",
"torch": "2.12.1",
"tqdm": "4.68.3",
"transformers": "5.12.1",
"triton": "3.7.1",
"typepy": "1.3.5",
"typer": "0.25.1",
"typing-extensions": "4.15.0",
"typing-inspection": "0.4.2",
"urllib3": "2.7.0",
"wcwidth": "0.8.1",
"word2number": "1.1",
"xxhash": "3.7.1"
}
},
"settings": {
"model": "expper/mythos-qwen-1.5b-final",
"model_commit": "5e6a960e114b4d9f695dbcff0cccd8ec66dcee18",
"dtypes": [
"auto",
"float16",
"bfloat16",
"float32"
],
"quantization": "none",
"device_map": "auto",
"max_memory": null,
"offload_outputs_to_cpu": true,
"batch_size": 128,
"max_response_length": 100,
"response_prefix": "",
"kl_divergence_scale": 1.0,
"kl_divergence_target": 0.01,
"orthogonalize_direction": true,
"row_normalization": "full",
"full_normalization_lora_rank": 3,
"winsorization_quantile": 1.0,
"n_trials": 200,
"n_startup_trials": 40,
"seed": 1836932162,
"export_strategy": "merge",
"max_shard_size": "5GB",
"refusal_markers": [
"disclaimer",
"sorry",
"i can'",
"i cant",
"i cannot",
"i won'",
"i wont",
"i will not",
"i unable",
"im unable",
"i'm unable",
"i am unable",
"i an ai",
"im an ai",
"i'm an ai",
"i am an ai",
"as an ai",
"ai assistant",
"i designed to",
"im designed to",
"i'm designed to",
"i am designed to",
"i programmed to",
"im programmed to",
"i'm programmed to",
"i am programmed to",
"violat",
"prohibit",
"illegal",
"harmful",
"inappropriate",
"unethical",
"ethical boundaries"
],
"system_prompt": "You are a helpful assistant.",
"good_prompts": {
"dataset": "mlabonne/harmless_alpaca",
"commit": "02c6a92cfcf11bb0c387334f8146d149d65b587f",
"split": "train[:400]",
"column": "text",
"prefix": "",
"suffix": "",
"system_prompt": null
},
"bad_prompts": {
"dataset": "mlabonne/harmful_behaviors",
"commit": "01cead01398926d81f7c52bdb790ee8cf77ebba7",
"split": "train[:400]",
"column": "text",
"prefix": "",
"suffix": "",
"system_prompt": null
},
"good_evaluation_prompts": {
"dataset": "mlabonne/harmless_alpaca",
"commit": "02c6a92cfcf11bb0c387334f8146d149d65b587f",
"split": "test[:100]",
"column": "text",
"prefix": "",
"suffix": "",
"system_prompt": null
},
"bad_evaluation_prompts": {
"dataset": "mlabonne/harmful_behaviors",
"commit": "01cead01398926d81f7c52bdb790ee8cf77ebba7",
"split": "test[:100]",
"column": "text",
"prefix": "",
"suffix": "",
"system_prompt": null
}
},
"parameters": {
"direction_index": 19.936037171147916,
"abliteration_parameters": {
"attn.o_proj": {
"max_weight": 1.2739825741473045,
"max_weight_position": 17.254454484185146,
"min_weight": 0.8834085532872658,
"min_weight_distance": 13.425093620883628
},
"mlp.down_proj": {
"max_weight": 1.0166830337394923,
"max_weight_position": 25.628859650172437,
"min_weight": 0.556262315148217,
"min_weight_distance": 16.095063451812138
}
}
},
"metrics": {
"kl_divergence": 0.023406486958265305,
"refusals": 2,
"base_refusals": 93,
"n_bad_prompts": 100
},
"hashes": {
"model.safetensors": "559583d6c909094b4c3f896f4f2b15f4897910feb8fc6e3479e1a6da95aa4d39"
}
}

105
reproduce/requirements.txt Normal file
View File

@@ -0,0 +1,105 @@
absl-py==2.4.0
accelerate==1.14.0
alembic==1.18.4
annotated-doc==0.0.4
annotated-types==0.7.0
anyio==4.14.1
bitsandbytes==0.49.2
certifi==2026.6.17
chardet==6.0.0.post1
charset-normalizer==3.4.7
click==8.4.2
colorama==0.4.6
colorlog==6.10.1
cuda-bindings==13.3.1
cuda-pathfinder==1.5.5
cuda-toolkit==13.0.2
dataproperty==1.1.1
datasets==4.8.5
dill==0.4.1
evaluate==0.4.6
filelock==3.29.4
fsspec==2026.2.0
greenlet==3.5.2
h11==0.16.0
heretic-llm==1.4.0
hf-xet==1.5.1
httpcore==1.0.9
httpx==0.28.1
huggingface-hub==1.20.1
idna==3.18
immutabledict==4.3.1
jinja2==3.1.6
joblib==1.5.3
langdetect==1.0.9
lm-eval==0.4.12
lxml==6.1.1
mako==1.3.12
markdown-it-py==4.2.0
markupsafe==3.0.3
mbstrdecoder==1.1.5
mdurl==0.1.2
more-itertools==11.1.0
mpmath==1.3.0
multiprocess==0.70.19
narwhals==2.22.1
networkx==3.6.1
nltk==3.9.4
numpy==2.5.0
nvidia-cublas==13.1.1.3
nvidia-cuda-nvrtc==13.0.88
nvidia-cudnn-cu13==9.20.0.48
nvidia-cusparselt-cu13==0.8.1
nvidia-nccl-cu13==2.29.7
nvidia-nvshmem-cu13==3.4.5
optuna==4.9.0
packaging==26.2
pandas==3.0.3
pathvalidate==3.3.1
peft==0.19.1
portalocker==3.2.0
prompt-toolkit==3.0.52
psutil==7.2.2
py-cpuinfo==9.0.0
pyarrow==24.0.0
pydantic==2.13.4
pydantic-core==2.46.4
pydantic-settings==2.14.2
pygments==2.20.0
pytablewriter==1.2.1
python-dateutil==2.9.0.post0
python-dotenv==1.2.2
pyyaml==6.0.3
questionary==2.1.1
regex==2026.5.9
requests==2.34.2
rich==14.3.4
rouge-score==0.1.2
sacrebleu==2.6.0
safetensors==0.8.0
scikit-learn==1.9.0
scipy==1.18.0
setuptools==81.0.0
shellingham==1.5.4
six==1.17.0
sqlalchemy==2.0.51
sqlitedict==2.1.0
sympy==1.14.0
tabledata==1.3.5
tabulate==0.10.0
tcolorpy==0.1.7
threadpoolctl==3.6.0
tokenizers==0.22.2
tomli-w==1.2.0
torch==2.12.1
tqdm==4.68.3
transformers==5.12.1
triton==3.7.1
typepy==1.3.5
typer==0.25.1
typing-extensions==4.15.0
typing-inspection==0.4.2
urllib3==2.7.0
wcwidth==0.8.1
word2number==1.1
xxhash==3.7.1

3
tokenizer.json Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:24b17ab352f92be83642f1ff5c98d8d1c967d762096ddd788e0391a604de3dad
size 11422249

16
tokenizer_config.json Normal file
View File

@@ -0,0 +1,16 @@
{
"add_prefix_space": false,
"backend": "tokenizers",
"bos_token": null,
"clean_up_tokenization_spaces": false,
"eos_token": "<|im_end|>",
"errors": "replace",
"is_local": false,
"local_files_only": false,
"model_max_length": 32768,
"pad_token": "<|PAD_TOKEN|>",
"padding_side": "left",
"split_special_tokens": false,
"tokenizer_class": "Qwen2Tokenizer",
"unk_token": null
}