{
  "base_model_name": "meta-llama_Llama-3.2-3B",
  "base_model_class": "LlamaForCausalLM",
  "base_loaded_in_4bit": false,
  "base_loaded_in_8bit": false,
  "projections": "q, v",
  "loss": 2.6899,
  "grad_norm": 0.22385260462760925,
  "learning_rate": 0.00015,
  "epoch": 2.0,
  "current_steps": 183,
  "train_runtime": 47.613,
  "train_samples_per_second": 22.053,
  "train_steps_per_second": 0.126,
  "total_flos": 3040582198886400.0,
  "train_loss": 2.67510195573171
}