{
  "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": 0.9479,
  "grad_norm": 0.46024736762046814,
  "learning_rate": 1.4285714285714284e-05,
  "epoch": 1.9662921348314608,
  "current_steps": 351,
  "train_runtime": 477.8422,
  "train_samples_per_second": 8.907,
  "train_steps_per_second": 0.184,
  "total_flos": 1.8179205044895744e+16,
  "train_loss": 1.1324877928603778
}