metadata
language:
- en
license: mit
library_name: peft
base_model: Qwen/Qwen2.5-Coder-3B-Instruct
tags:
- conventional-commits
- qwen2.5-coder
- text-generation
- code-llm
- fine-tuned
- lora
- qlora
Qwen Commit LoRA - Conventional Commit Message Generator
Generates conventional commit messages from git diffs using a fine-tuned Qwen2.5-Coder-3B model with QLoRA adapters.
Model Details
- Base Model: Qwen/Qwen2.5-Coder-3B-Instruct
- Fine-tuning Method: QLoRA (4-bit quantized, rank=8, alpha=16)
- Training Data: 210 real conventional commits from open-source repositories
- Target Modules: q_proj, k_proj, v_proj, o_proj
Usage
from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer
# Load base model
base_model = AutoModelForCausalLM.from_pretrained(
"Qwen/Qwen2.5-Coder-3B-Instruct",
load_in_4bit=True,
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-Coder-3B-Instruct")
# Load LoRA adapters
model = PeftModel.from_pretrained(base_model, "Pavloffm/qwen-commit-lora")
# Generate commit message
diff = """diff --git a/src/main.py b/src/main.py
index 1234567..abcdefg 100644
--- a/src/main.py
+++ b/src/main.py
@@ -1,3 +1,5 @@
+def new_feature():
+ pass
"""
messages = [{"role": "user", "content": f"Generate a conventional commit message for this diff:\n{diff}"}]
inputs = tokenizer.apply_chat_template(messages, return_tensors="pt").to(model.device)
outputs = model.generate(inputs, max_new_tokens=100)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
Training
- 2 epochs
- Learning rate: 1.5e-4
- LoRA rank: 8, alpha: 16
- 210 training examples
License
MIT License