{ "base_model_name": "Meta-Llama-3-8b", "base_model_class": "LlamaForCausalLM", "base_loaded_in_4bit": true, "base_loaded_in_8bit": false, "projections": "q, v", "loss": 1.0515, "grad_norm": 3.200089693069458, "learning_rate": 7.692307692307693e-09, "epoch": 5.0, "current_steps": 489, "current_steps_adjusted": 489, "epoch_adjusted": 5.0, "train_runtime": 779.2785, "train_samples_per_second": 2.502, "train_steps_per_second": 0.629, "total_flos": 2.2519579410432e+16, "train_loss": 1.0483642434587284 }