Modelpublic

dhruvanmurthy/qwen3-8b-multi-tool-use-lora-sft

sourceHugging Facemitupdated 6mo agoView on Hugging Face

0likes

Model Card

dhruvanmurthy/qwen3-8b-multi-tool-use-lora-sft

LoRA adapter for Qwen3-8B tool-use fine-tuning (SFT stage).

Model Details

—Base Model: Qwen/Qwen3-8B
—Training Stage: SFT (Supervised Fine-Tuning)
—Task: Tool-use instruction following
—LoRA Rank: 64
—Training Steps: 5064

Training Configuration

json

{
  "stage": "sft",
  "base_model": "Qwen/Qwen3-8B",
  "lora_rank": 64,
  "learning_rate": 0.0002,
  "batch_size": 8,
  "num_epochs": 3,
  "max_seq_length": 2048,
  "total_steps": 5064
}

License

MIT