{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8072382361844539, "eval_steps": 500, "global_step": 6000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 4.9997567688496474e-05, "loss": 1.6283, "step": 100 }, { "epoch": 0.03, "learning_rate": 4.9990172715142793e-05, "loss": 0.068, "step": 200 }, { "epoch": 0.04, "learning_rate": 4.997781629993153e-05, "loss": 0.0404, "step": 300 }, { "epoch": 0.05, "learning_rate": 4.9960500896052476e-05, "loss": 0.0626, "step": 400 }, { "epoch": 0.07, "learning_rate": 4.993822994123172e-05, "loss": 0.0428, "step": 500 }, { "epoch": 0.08, "learning_rate": 4.9911007857049264e-05, "loss": 0.0265, "step": 600 }, { "epoch": 0.09, "learning_rate": 4.987884004806111e-05, "loss": 0.0263, "step": 700 }, { "epoch": 0.11, "learning_rate": 4.984173290072626e-05, "loss": 0.03, "step": 800 }, { "epoch": 0.12, "learning_rate": 4.979969378213884e-05, "loss": 0.0253, "step": 900 }, { "epoch": 0.13, "learning_rate": 4.975273103856537e-05, "loss": 0.038, "step": 1000 }, { "epoch": 0.15, "learning_rate": 4.970085399378785e-05, "loss": 0.0322, "step": 1100 }, { "epoch": 0.16, "learning_rate": 4.964407294725254e-05, "loss": 0.0296, "step": 1200 }, { "epoch": 0.17, "learning_rate": 4.958239917202523e-05, "loss": 0.0295, "step": 1300 }, { "epoch": 0.19, "learning_rate": 4.9515844912553106e-05, "loss": 0.0171, "step": 1400 }, { "epoch": 0.2, "learning_rate": 4.944442338223378e-05, "loss": 0.0205, "step": 1500 }, { "epoch": 0.22, "learning_rate": 4.9368148760792e-05, "loss": 0.029, "step": 1600 }, { "epoch": 0.23, "learning_rate": 4.9287036191464414e-05, "loss": 0.0172, "step": 1700 }, { "epoch": 0.24, "learning_rate": 4.9201101777993116e-05, "loss": 0.0235, "step": 1800 }, { "epoch": 0.26, "learning_rate": 4.911036258142853e-05, "loss": 0.0168, "step": 1900 }, { "epoch": 0.27, "learning_rate": 4.9014836616742065e-05, "loss": 0.0227, "step": 2000 }, { "epoch": 0.28, "learning_rate": 4.891454284924965e-05, "loss": 0.0227, "step": 2100 }, { "epoch": 0.3, "learning_rate": 4.8810575041794e-05, "loss": 0.0173, "step": 2200 }, { "epoch": 0.31, "learning_rate": 4.8700853511049656e-05, "loss": 0.0227, "step": 2300 }, { "epoch": 0.32, "learning_rate": 4.858642651436035e-05, "loss": 0.0144, "step": 2400 }, { "epoch": 0.34, "learning_rate": 4.846731676957191e-05, "loss": 0.03, "step": 2500 }, { "epoch": 0.35, "learning_rate": 4.834354792422293e-05, "loss": 0.0189, "step": 2600 }, { "epoch": 0.36, "learning_rate": 4.821514455084985e-05, "loss": 0.0111, "step": 2700 }, { "epoch": 0.38, "learning_rate": 4.8082132142108465e-05, "loss": 0.0152, "step": 2800 }, { "epoch": 0.39, "learning_rate": 4.794453710571272e-05, "loss": 0.0182, "step": 2900 }, { "epoch": 0.4, "learning_rate": 4.780238675919182e-05, "loss": 0.0111, "step": 3000 }, { "epoch": 0.42, "learning_rate": 4.765570932446672e-05, "loss": 0.0138, "step": 3100 }, { "epoch": 0.43, "learning_rate": 4.75045339222471e-05, "loss": 0.0256, "step": 3200 }, { "epoch": 0.44, "learning_rate": 4.734889056624983e-05, "loss": 0.0199, "step": 3300 }, { "epoch": 0.46, "learning_rate": 4.718881015724017e-05, "loss": 0.0159, "step": 3400 }, { "epoch": 0.47, "learning_rate": 4.702432447689692e-05, "loss": 0.0121, "step": 3500 }, { "epoch": 0.48, "learning_rate": 4.6855466181502544e-05, "loss": 0.0181, "step": 3600 }, { "epoch": 0.5, "learning_rate": 4.66822687954598e-05, "loss": 0.0148, "step": 3700 }, { "epoch": 0.51, "learning_rate": 4.65047667046359e-05, "loss": 0.0152, "step": 3800 }, { "epoch": 0.52, "learning_rate": 4.632299514953571e-05, "loss": 0.0103, "step": 3900 }, { "epoch": 0.54, "learning_rate": 4.613699021830524e-05, "loss": 0.0217, "step": 4000 }, { "epoch": 0.55, "learning_rate": 4.59467888395669e-05, "loss": 0.0142, "step": 4100 }, { "epoch": 0.57, "learning_rate": 4.575242877508777e-05, "loss": 0.0151, "step": 4200 }, { "epoch": 0.58, "learning_rate": 4.5553948612282607e-05, "loss": 0.0164, "step": 4300 }, { "epoch": 0.59, "learning_rate": 4.5353433433160075e-05, "loss": 0.0095, "step": 4400 }, { "epoch": 0.61, "learning_rate": 4.514687230313633e-05, "loss": 0.015, "step": 4500 }, { "epoch": 0.62, "learning_rate": 4.49363112993749e-05, "loss": 0.015, "step": 4600 }, { "epoch": 0.63, "learning_rate": 4.472179222575561e-05, "loss": 0.0181, "step": 4700 }, { "epoch": 0.65, "learning_rate": 4.4503357671976574e-05, "loss": 0.0175, "step": 4800 }, { "epoch": 0.66, "learning_rate": 4.428105100509852e-05, "loss": 0.0069, "step": 4900 }, { "epoch": 0.67, "learning_rate": 4.4054916360934957e-05, "loss": 0.0108, "step": 5000 }, { "epoch": 0.69, "learning_rate": 4.3824998635289594e-05, "loss": 0.0204, "step": 5100 }, { "epoch": 0.7, "learning_rate": 4.3591343475042946e-05, "loss": 0.0192, "step": 5200 }, { "epoch": 0.71, "learning_rate": 4.3353997269089774e-05, "loss": 0.0086, "step": 5300 }, { "epoch": 0.73, "learning_rate": 4.311300713912926e-05, "loss": 0.0179, "step": 5400 }, { "epoch": 0.74, "learning_rate": 4.286842093030963e-05, "loss": 0.0142, "step": 5500 }, { "epoch": 0.75, "learning_rate": 4.262028720172921e-05, "loss": 0.0187, "step": 5600 }, { "epoch": 0.77, "learning_rate": 4.2368655216795735e-05, "loss": 0.0176, "step": 5700 }, { "epoch": 0.78, "learning_rate": 4.211357493344573e-05, "loss": 0.0145, "step": 5800 }, { "epoch": 0.79, "learning_rate": 4.185509699422615e-05, "loss": 0.0087, "step": 5900 }, { "epoch": 0.81, "learning_rate": 4.1593272716239985e-05, "loss": 0.0136, "step": 6000 } ], "logging_steps": 100, "max_steps": 22296, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 3000, "total_flos": 1.6775846445215908e+18, "train_batch_size": 4, "trial_name": null, "trial_params": null }