Phase 4: Qwen Fine-tuning

Fine-tune Qwen2.5-Coder-14B with LoRA on HuggingFace Skills A100 GPU. Runs 3 parallel sweep configs (conservative / standard / aggressive) then full training on top-2.

Status: ready

Key Metrics

MetricValue
model Qwen2.5-Coder-14B-Instruct
hardware A100 80GB (cloud)
sweep configs 3
lora r range 16–64

Technologies

  • HF Skills API
  • PEFT LoRA
  • bitsandbytes
  • A100 80GB

Outputs

  • Fine-tuned LoRA weights on HF Hub

Commands

python -m src.phase4_qwen_finetuning.scripts.launch_validation_sweep --config src/config/v6_config.yaml

python -m src.phase4_qwen_finetuning.scripts.monitor_jobs --config src/config/v6_config.yaml
← Phase 3 Phase 5 →