666-LLAMA3SP-mule-titanium

LLAMA 3 Story Point Estimator - mule - titanium

This model is fine-tuned on issue descriptions from mule and tested on titanium for story point estimation.

Model Details

• Base Model: LLAMA 3.2 1B

• Training Project: mule

• Test Project: titanium

• Task: Story Point Estimation (Regression)

• Architecture: PEFT (LoRA)

• Tokenizer: SP WordPiece

• Input: Issue titles

• Output: Story point estimation (continuous value)

Usage

from transformers import AutoModelForSequenceClassification, BertTokenizer
from peft import PeftConfig, PeftModel

# Load peft config model
config = PeftConfig.from_pretrained("DEVCamiloSepulveda/666-LLAMA3SP-mule-titanium")

# Load tokenizer and model
tokenizer = BertTokenizer('vocab.txt')
base_model = AutoModelForSequenceClassification.from_pretrained(
    config.base_model_name_or_path,
    num_labels=1,
    torch_dtype=torch.float16,
    device_map='auto'
)
model = PeftModel.from_pretrained(base_model, "DEVCamiloSepulveda/666-LLAMA3SP-mule-titanium")

# Prepare input text
text = "Your issue description here"
inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=20, padding="max_length")

# Get prediction
outputs = model(**inputs)
story_points = outputs.logits.item()

Training Details

• Fine-tuning method: LoRA (Low-Rank Adaptation)
• Sequence length: 20 tokens
• Best training epoch: 0 / 20 epochs
• Batch size: 32
• Training time: 18.968 seconds
• Mean Absolute Error (MAE): 3.326
• Median Absolute Error (MdAE): 2.348

Framework versions

• PEFT 0.14.0