data_config.json

{
  "source": "/Users/sameersaraf/repositories/digital_employee/agentic-factory/src_slm/data/output/instruction_tuning.jsonl",
  "total_samples": 648338,
  "splits": {
    "train": 583504,
    "valid": 32417,
    "test": 32417
  },
  "train_fraction": 0.9,
  "seed": 42,
  "format": "chat_completions",
  "model_target": "Qwen/Qwen2.5-3B-Instruct"
}