| { |
| "model": "mistralai/Mistral-7B-Instruct-v0.3", |
| "seed": 1, |
| "epochs_requested": 2, |
| "epoch_reached": 1.2962962962962963, |
| "global_step": 140, |
| "max_steps": 216, |
| "early_stopped": true, |
| "best_eval_loss": 1.654082179069519, |
| "best_checkpoint": "finetuned_models/mistralai-Mistral-7B-Instruct-v0.3_seed1/checkpoint-110", |
| "final_metrics": { |
| "eval_loss": 1.654082179069519, |
| "eval_runtime": 13.1871, |
| "eval_samples_per_second": 32.759, |
| "eval_steps_per_second": 8.19, |
| "epoch": 1.2962962962962963 |
| }, |
| "learning_rate": 2.675263789416278e-05, |
| "lora": { |
| "r": 64, |
| "alpha": 128, |
| "dropout": 0.05 |
| }, |
| "batch": { |
| "per_device": 1, |
| "grad_accum": 8, |
| "world_size": 4 |
| }, |
| "max_length": 2048, |
| "eval_steps": 10, |
| "early_stopping_patience": 3, |
| "log_history": [ |
| { |
| "eval_loss": 1.8488812446594238, |
| "eval_runtime": 9.8489, |
| "eval_samples_per_second": 43.863, |
| "eval_steps_per_second": 10.966, |
| "epoch": 0.09259259259259259, |
| "step": 10 |
| }, |
| { |
| "eval_loss": 1.764305591583252, |
| "eval_runtime": 9.8335, |
| "eval_samples_per_second": 43.931, |
| "eval_steps_per_second": 10.983, |
| "epoch": 0.18518518518518517, |
| "step": 20 |
| }, |
| { |
| "loss": 2.0483358764648436, |
| "grad_norm": 2.1267218589782715, |
| "learning_rate": 2.6318280611836455e-05, |
| "epoch": 0.23148148148148148, |
| "step": 25 |
| }, |
| { |
| "eval_loss": 1.728833556175232, |
| "eval_runtime": 9.9208, |
| "eval_samples_per_second": 43.545, |
| "eval_steps_per_second": 10.886, |
| "epoch": 0.2777777777777778, |
| "step": 30 |
| }, |
| { |
| "eval_loss": 1.7081001996994019, |
| "eval_runtime": 9.9092, |
| "eval_samples_per_second": 43.596, |
| "eval_steps_per_second": 10.899, |
| "epoch": 0.37037037037037035, |
| "step": 40 |
| }, |
| { |
| "loss": 1.7389724731445313, |
| "grad_norm": 2.0094456672668457, |
| "learning_rate": 2.4174302619519622e-05, |
| "epoch": 0.46296296296296297, |
| "step": 50 |
| }, |
| { |
| "eval_loss": 1.6942864656448364, |
| "eval_runtime": 9.9717, |
| "eval_samples_per_second": 43.323, |
| "eval_steps_per_second": 10.831, |
| "epoch": 0.46296296296296297, |
| "step": 50 |
| }, |
| { |
| "eval_loss": 1.6831718683242798, |
| "eval_runtime": 10.0379, |
| "eval_samples_per_second": 43.037, |
| "eval_steps_per_second": 10.759, |
| "epoch": 0.5555555555555556, |
| "step": 60 |
| }, |
| { |
| "eval_loss": 1.6765761375427246, |
| "eval_runtime": 10.0079, |
| "eval_samples_per_second": 43.166, |
| "eval_steps_per_second": 10.791, |
| "epoch": 0.6481481481481481, |
| "step": 70 |
| }, |
| { |
| "loss": 1.7271536254882813, |
| "grad_norm": 1.6946066617965698, |
| "learning_rate": 2.0523325458099818e-05, |
| "epoch": 0.6944444444444444, |
| "step": 75 |
| }, |
| { |
| "eval_loss": 1.6678999662399292, |
| "eval_runtime": 10.8076, |
| "eval_samples_per_second": 39.972, |
| "eval_steps_per_second": 9.993, |
| "epoch": 0.7407407407407407, |
| "step": 80 |
| }, |
| { |
| "eval_loss": 1.6625776290893555, |
| "eval_runtime": 10.0228, |
| "eval_samples_per_second": 43.102, |
| "eval_steps_per_second": 10.775, |
| "epoch": 0.8333333333333334, |
| "step": 90 |
| }, |
| { |
| "loss": 1.6924752807617187, |
| "grad_norm": 1.8865303993225098, |
| "learning_rate": 1.587489051192989e-05, |
| "epoch": 0.9259259259259259, |
| "step": 100 |
| }, |
| { |
| "eval_loss": 1.6575430631637573, |
| "eval_runtime": 9.9455, |
| "eval_samples_per_second": 43.437, |
| "eval_steps_per_second": 10.859, |
| "epoch": 0.9259259259259259, |
| "step": 100 |
| }, |
| { |
| "eval_loss": 1.654082179069519, |
| "eval_runtime": 9.9907, |
| "eval_samples_per_second": 43.24, |
| "eval_steps_per_second": 10.81, |
| "epoch": 1.0185185185185186, |
| "step": 110 |
| }, |
| { |
| "eval_loss": 1.6632981300354004, |
| "eval_runtime": 9.9243, |
| "eval_samples_per_second": 43.53, |
| "eval_steps_per_second": 10.882, |
| "epoch": 1.1111111111111112, |
| "step": 120 |
| }, |
| { |
| "loss": 1.5846710205078125, |
| "grad_norm": 1.8140192031860352, |
| "learning_rate": 1.0877747382232885e-05, |
| "epoch": 1.1574074074074074, |
| "step": 125 |
| }, |
| { |
| "eval_loss": 1.6580889225006104, |
| "eval_runtime": 9.8547, |
| "eval_samples_per_second": 43.837, |
| "eval_steps_per_second": 10.959, |
| "epoch": 1.2037037037037037, |
| "step": 130 |
| }, |
| { |
| "eval_loss": 1.6590867042541504, |
| "eval_runtime": 9.9468, |
| "eval_samples_per_second": 43.431, |
| "eval_steps_per_second": 10.858, |
| "epoch": 1.2962962962962963, |
| "step": 140 |
| }, |
| { |
| "train_runtime": 454.6607, |
| "train_samples_per_second": 15.198, |
| "train_steps_per_second": 0.475, |
| "total_flos": 3.648040061632512e+16, |
| "train_loss": 1.7275653975350516, |
| "epoch": 1.2962962962962963, |
| "step": 140 |
| }, |
| { |
| "eval_loss": 1.654082179069519, |
| "eval_runtime": 13.1871, |
| "eval_samples_per_second": 32.759, |
| "eval_steps_per_second": 8.19, |
| "epoch": 1.2962962962962963, |
| "step": 140 |
| } |
| ] |
| } |
