{
  "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": 0.7126,
  "grad_norm": 0.5774107575416565,
  "learning_rate": 3.821656050955414e-06,
  "epoch": 1.9809264305177112,
  "current_steps": 727,
  "train_runtime": 1007.346,
  "train_samples_per_second": 8.734,
  "train_steps_per_second": 0.181,
  "total_flos": 3.781118530211021e+16,
  "train_loss": 0.8363648201083089
}