| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.6, | |
| "eval_steps": 50, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 6.656943321228027, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.6339, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "eval_loss": 0.9032842516899109, | |
| "eval_runtime": 1.8285, | |
| "eval_samples_per_second": 62.347, | |
| "eval_steps_per_second": 3.281, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.039907693862915, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.2834, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_loss": 0.8288049101829529, | |
| "eval_runtime": 1.8279, | |
| "eval_samples_per_second": 62.366, | |
| "eval_steps_per_second": 3.282, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.98620343208313, | |
| "learning_rate": 3.6e-06, | |
| "loss": 1.1369, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "eval_loss": 0.8595764636993408, | |
| "eval_runtime": 1.8209, | |
| "eval_samples_per_second": 62.607, | |
| "eval_steps_per_second": 3.295, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.9106926918029785, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 1.1002, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "eval_loss": 0.8750876784324646, | |
| "eval_runtime": 1.8177, | |
| "eval_samples_per_second": 62.716, | |
| "eval_steps_per_second": 3.301, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.5830564498901367, | |
| "learning_rate": 6e-06, | |
| "loss": 1.0671, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.8878025412559509, | |
| "eval_runtime": 1.8312, | |
| "eval_samples_per_second": 62.255, | |
| "eval_steps_per_second": 3.277, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 3.3819241523742676, | |
| "learning_rate": 7.2e-06, | |
| "loss": 1.0332, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "eval_loss": 0.8920303583145142, | |
| "eval_runtime": 1.8258, | |
| "eval_samples_per_second": 62.439, | |
| "eval_steps_per_second": 3.286, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.588942050933838, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.0199, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "eval_loss": 0.8735177516937256, | |
| "eval_runtime": 1.8258, | |
| "eval_samples_per_second": 62.438, | |
| "eval_steps_per_second": 3.286, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.964590311050415, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 1.0218, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_loss": 0.8841342329978943, | |
| "eval_runtime": 1.8275, | |
| "eval_samples_per_second": 62.38, | |
| "eval_steps_per_second": 3.283, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 4.193049907684326, | |
| "learning_rate": 1.08e-05, | |
| "loss": 0.995, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_loss": 0.8781883120536804, | |
| "eval_runtime": 1.8206, | |
| "eval_samples_per_second": 62.616, | |
| "eval_steps_per_second": 3.296, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.1733250617980957, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.9752, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.8925071358680725, | |
| "eval_runtime": 1.8212, | |
| "eval_samples_per_second": 62.598, | |
| "eval_steps_per_second": 3.295, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 4.2617034912109375, | |
| "learning_rate": 1.32e-05, | |
| "loss": 1.004, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "eval_loss": 0.8857673406600952, | |
| "eval_runtime": 1.8247, | |
| "eval_samples_per_second": 62.475, | |
| "eval_steps_per_second": 3.288, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.422171115875244, | |
| "learning_rate": 1.44e-05, | |
| "loss": 1.0094, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_loss": 0.885766327381134, | |
| "eval_runtime": 1.8226, | |
| "eval_samples_per_second": 62.547, | |
| "eval_steps_per_second": 3.292, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 3.6416265964508057, | |
| "learning_rate": 1.56e-05, | |
| "loss": 0.9922, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_loss": 0.9024901986122131, | |
| "eval_runtime": 1.8283, | |
| "eval_samples_per_second": 62.353, | |
| "eval_steps_per_second": 3.282, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.2401561737060547, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 0.9848, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "eval_loss": 0.8976176977157593, | |
| "eval_runtime": 1.8213, | |
| "eval_samples_per_second": 62.593, | |
| "eval_steps_per_second": 3.294, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.61985182762146, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.9916, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_loss": 0.873446524143219, | |
| "eval_runtime": 1.8236, | |
| "eval_samples_per_second": 62.515, | |
| "eval_steps_per_second": 3.29, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 3.0905354022979736, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 0.9961, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_loss": 0.8803545236587524, | |
| "eval_runtime": 1.8244, | |
| "eval_samples_per_second": 62.488, | |
| "eval_steps_per_second": 3.289, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 3.7789053916931152, | |
| "learning_rate": 2.04e-05, | |
| "loss": 1.02, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "eval_loss": 0.8905940651893616, | |
| "eval_runtime": 1.827, | |
| "eval_samples_per_second": 62.399, | |
| "eval_steps_per_second": 3.284, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.842257261276245, | |
| "learning_rate": 2.16e-05, | |
| "loss": 1.001, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "eval_loss": 0.8835523128509521, | |
| "eval_runtime": 1.8313, | |
| "eval_samples_per_second": 62.251, | |
| "eval_steps_per_second": 3.276, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 3.3730099201202393, | |
| "learning_rate": 2.2800000000000002e-05, | |
| "loss": 1.015, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_loss": 0.953080952167511, | |
| "eval_runtime": 1.8303, | |
| "eval_samples_per_second": 62.285, | |
| "eval_steps_per_second": 3.278, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.283881664276123, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.0089, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.9305523037910461, | |
| "eval_runtime": 1.8222, | |
| "eval_samples_per_second": 62.56, | |
| "eval_steps_per_second": 3.293, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 3.0285823345184326, | |
| "learning_rate": 2.52e-05, | |
| "loss": 1.0215, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_loss": 0.9584159255027771, | |
| "eval_runtime": 1.8216, | |
| "eval_samples_per_second": 62.584, | |
| "eval_steps_per_second": 3.294, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.6454532146453857, | |
| "learning_rate": 2.64e-05, | |
| "loss": 1.0535, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_loss": 0.9347971677780151, | |
| "eval_runtime": 1.8265, | |
| "eval_samples_per_second": 62.414, | |
| "eval_steps_per_second": 3.285, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 2.3046398162841797, | |
| "learning_rate": 2.7600000000000003e-05, | |
| "loss": 1.036, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_loss": 0.9546946883201599, | |
| "eval_runtime": 1.8227, | |
| "eval_samples_per_second": 62.545, | |
| "eval_steps_per_second": 3.292, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 3.419645309448242, | |
| "learning_rate": 2.88e-05, | |
| "loss": 1.0732, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_loss": 0.9314932823181152, | |
| "eval_runtime": 1.8207, | |
| "eval_samples_per_second": 62.613, | |
| "eval_steps_per_second": 3.295, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.927753210067749, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0643, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.9249414801597595, | |
| "eval_runtime": 1.8261, | |
| "eval_samples_per_second": 62.427, | |
| "eval_steps_per_second": 3.286, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.8154520988464355, | |
| "learning_rate": 2.9998537860139564e-05, | |
| "loss": 0.6328, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "eval_loss": 1.008348822593689, | |
| "eval_runtime": 1.8294, | |
| "eval_samples_per_second": 62.316, | |
| "eval_steps_per_second": 3.28, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 6.605989456176758, | |
| "learning_rate": 2.9994151725605313e-05, | |
| "loss": 0.708, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "eval_loss": 0.9453077912330627, | |
| "eval_runtime": 1.8194, | |
| "eval_samples_per_second": 62.657, | |
| "eval_steps_per_second": 3.298, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 2.298051357269287, | |
| "learning_rate": 2.9986842451482876e-05, | |
| "loss": 0.6654, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "eval_loss": 1.035285234451294, | |
| "eval_runtime": 1.8281, | |
| "eval_samples_per_second": 62.359, | |
| "eval_steps_per_second": 3.282, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 3.0220792293548584, | |
| "learning_rate": 2.9976611462729715e-05, | |
| "loss": 0.7063, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "eval_loss": 1.0368359088897705, | |
| "eval_runtime": 1.8316, | |
| "eval_samples_per_second": 62.24, | |
| "eval_steps_per_second": 3.276, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.9698407649993896, | |
| "learning_rate": 2.9963460753897364e-05, | |
| "loss": 0.6812, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 1.0127767324447632, | |
| "eval_runtime": 1.8272, | |
| "eval_samples_per_second": 62.392, | |
| "eval_steps_per_second": 3.284, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 2.5555665493011475, | |
| "learning_rate": 2.9947392888742566e-05, | |
| "loss": 0.7029, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "eval_loss": 1.045531153678894, | |
| "eval_runtime": 1.8258, | |
| "eval_samples_per_second": 62.438, | |
| "eval_steps_per_second": 3.286, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 2.3660807609558105, | |
| "learning_rate": 2.992841099972747e-05, | |
| "loss": 0.6799, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_loss": 1.0579793453216553, | |
| "eval_runtime": 1.8239, | |
| "eval_samples_per_second": 62.504, | |
| "eval_steps_per_second": 3.29, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 2.915484666824341, | |
| "learning_rate": 2.9906518787408948e-05, | |
| "loss": 0.7182, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "eval_loss": 1.061601161956787, | |
| "eval_runtime": 1.8217, | |
| "eval_samples_per_second": 62.581, | |
| "eval_steps_per_second": 3.294, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 2.7276549339294434, | |
| "learning_rate": 2.988172051971717e-05, | |
| "loss": 0.7168, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "eval_loss": 1.0619702339172363, | |
| "eval_runtime": 1.8419, | |
| "eval_samples_per_second": 61.891, | |
| "eval_steps_per_second": 3.257, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.576150894165039, | |
| "learning_rate": 2.9854021031123555e-05, | |
| "loss": 0.729, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_loss": 1.0389442443847656, | |
| "eval_runtime": 1.8255, | |
| "eval_samples_per_second": 62.447, | |
| "eval_steps_per_second": 3.287, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 2.996159553527832, | |
| "learning_rate": 2.9823425721698293e-05, | |
| "loss": 0.7319, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "eval_loss": 1.0429059267044067, | |
| "eval_runtime": 1.8338, | |
| "eval_samples_per_second": 62.167, | |
| "eval_steps_per_second": 3.272, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 3.9739811420440674, | |
| "learning_rate": 2.9789940556057574e-05, | |
| "loss": 0.7281, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_loss": 1.0586906671524048, | |
| "eval_runtime": 1.8323, | |
| "eval_samples_per_second": 62.218, | |
| "eval_steps_per_second": 3.275, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 2.907900810241699, | |
| "learning_rate": 2.975357206220079e-05, | |
| "loss": 0.743, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "eval_loss": 1.0349457263946533, | |
| "eval_runtime": 1.8285, | |
| "eval_samples_per_second": 62.345, | |
| "eval_steps_per_second": 3.281, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 3.018134117126465, | |
| "learning_rate": 2.9714327330237873e-05, | |
| "loss": 0.7251, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "eval_loss": 1.0124412775039673, | |
| "eval_runtime": 1.8281, | |
| "eval_samples_per_second": 62.359, | |
| "eval_steps_per_second": 3.282, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.7769956588745117, | |
| "learning_rate": 2.9672214011007087e-05, | |
| "loss": 0.7403, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 1.0399030447006226, | |
| "eval_runtime": 1.8304, | |
| "eval_samples_per_second": 62.282, | |
| "eval_steps_per_second": 3.278, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 12500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 2000, | |
| "total_flos": 1.081999679773737e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |