| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.9171597633136095, | |
| "eval_steps": 25, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14792899408284024, | |
| "grad_norm": 0.8603515625, | |
| "learning_rate": 0.0001951951951951952, | |
| "loss": 0.939, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14792899408284024, | |
| "eval_loss": 0.6645179390907288, | |
| "eval_runtime": 5.4254, | |
| "eval_samples_per_second": 16.404, | |
| "eval_steps_per_second": 2.212, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2958579881656805, | |
| "grad_norm": 0.67138671875, | |
| "learning_rate": 0.0001901901901901902, | |
| "loss": 0.6144, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2958579881656805, | |
| "eval_loss": 0.6127904653549194, | |
| "eval_runtime": 5.4718, | |
| "eval_samples_per_second": 16.265, | |
| "eval_steps_per_second": 2.193, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4437869822485207, | |
| "grad_norm": 0.7236328125, | |
| "learning_rate": 0.0001851851851851852, | |
| "loss": 0.6042, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4437869822485207, | |
| "eval_loss": 0.6052118539810181, | |
| "eval_runtime": 5.4747, | |
| "eval_samples_per_second": 16.256, | |
| "eval_steps_per_second": 2.192, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.591715976331361, | |
| "grad_norm": 0.65283203125, | |
| "learning_rate": 0.00018018018018018018, | |
| "loss": 0.5929, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.591715976331361, | |
| "eval_loss": 0.5997007489204407, | |
| "eval_runtime": 5.5068, | |
| "eval_samples_per_second": 16.162, | |
| "eval_steps_per_second": 2.179, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7396449704142012, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 0.0001751751751751752, | |
| "loss": 0.5968, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7396449704142012, | |
| "eval_loss": 0.5949457287788391, | |
| "eval_runtime": 5.5046, | |
| "eval_samples_per_second": 16.168, | |
| "eval_steps_per_second": 2.18, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8875739644970414, | |
| "grad_norm": 0.66357421875, | |
| "learning_rate": 0.0001701701701701702, | |
| "loss": 0.6017, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8875739644970414, | |
| "eval_loss": 0.5932657122612, | |
| "eval_runtime": 5.4672, | |
| "eval_samples_per_second": 16.279, | |
| "eval_steps_per_second": 2.195, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0355029585798816, | |
| "grad_norm": 0.59228515625, | |
| "learning_rate": 0.00016516516516516518, | |
| "loss": 0.5471, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.0355029585798816, | |
| "eval_loss": 0.6293187141418457, | |
| "eval_runtime": 5.5245, | |
| "eval_samples_per_second": 16.11, | |
| "eval_steps_per_second": 2.172, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.183431952662722, | |
| "grad_norm": 0.62158203125, | |
| "learning_rate": 0.00016016016016016018, | |
| "loss": 0.4246, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.183431952662722, | |
| "eval_loss": 0.6185322403907776, | |
| "eval_runtime": 5.5491, | |
| "eval_samples_per_second": 16.039, | |
| "eval_steps_per_second": 2.163, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.331360946745562, | |
| "grad_norm": 0.71142578125, | |
| "learning_rate": 0.00015515515515515516, | |
| "loss": 0.4311, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.331360946745562, | |
| "eval_loss": 0.6142793893814087, | |
| "eval_runtime": 5.4779, | |
| "eval_samples_per_second": 16.247, | |
| "eval_steps_per_second": 2.191, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.4792899408284024, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 0.00015015015015015014, | |
| "loss": 0.4175, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.4792899408284024, | |
| "eval_loss": 0.6187874674797058, | |
| "eval_runtime": 5.4791, | |
| "eval_samples_per_second": 16.243, | |
| "eval_steps_per_second": 2.19, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.6272189349112427, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 0.00014514514514514515, | |
| "loss": 0.4303, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.6272189349112427, | |
| "eval_loss": 0.6225253343582153, | |
| "eval_runtime": 5.5347, | |
| "eval_samples_per_second": 16.08, | |
| "eval_steps_per_second": 2.168, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.7751479289940828, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.00014014014014014013, | |
| "loss": 0.4271, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.7751479289940828, | |
| "eval_loss": 0.6251070499420166, | |
| "eval_runtime": 5.4796, | |
| "eval_samples_per_second": 16.242, | |
| "eval_steps_per_second": 2.19, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.72998046875, | |
| "learning_rate": 0.00013513513513513514, | |
| "loss": 0.4248, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "eval_loss": 0.6276720762252808, | |
| "eval_runtime": 5.5272, | |
| "eval_samples_per_second": 16.102, | |
| "eval_steps_per_second": 2.171, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.0710059171597632, | |
| "grad_norm": 0.72705078125, | |
| "learning_rate": 0.00013013013013013014, | |
| "loss": 0.3568, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.0710059171597632, | |
| "eval_loss": 0.6846657395362854, | |
| "eval_runtime": 5.4826, | |
| "eval_samples_per_second": 16.233, | |
| "eval_steps_per_second": 2.189, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.2189349112426036, | |
| "grad_norm": 0.732421875, | |
| "learning_rate": 0.00012512512512512512, | |
| "loss": 0.2759, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.2189349112426036, | |
| "eval_loss": 0.7119464874267578, | |
| "eval_runtime": 5.5426, | |
| "eval_samples_per_second": 16.057, | |
| "eval_steps_per_second": 2.165, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.366863905325444, | |
| "grad_norm": 0.7099609375, | |
| "learning_rate": 0.00012012012012012013, | |
| "loss": 0.2687, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.366863905325444, | |
| "eval_loss": 0.7088969945907593, | |
| "eval_runtime": 5.5287, | |
| "eval_samples_per_second": 16.098, | |
| "eval_steps_per_second": 2.17, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.5147928994082838, | |
| "grad_norm": 0.86767578125, | |
| "learning_rate": 0.00011511511511511512, | |
| "loss": 0.2796, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.5147928994082838, | |
| "eval_loss": 0.716306209564209, | |
| "eval_runtime": 5.5463, | |
| "eval_samples_per_second": 16.047, | |
| "eval_steps_per_second": 2.164, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.662721893491124, | |
| "grad_norm": 0.77587890625, | |
| "learning_rate": 0.00011011011011011012, | |
| "loss": 0.2735, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.662721893491124, | |
| "eval_loss": 0.7141934037208557, | |
| "eval_runtime": 5.4992, | |
| "eval_samples_per_second": 16.184, | |
| "eval_steps_per_second": 2.182, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.8106508875739644, | |
| "grad_norm": 83.5625, | |
| "learning_rate": 0.00010510510510510511, | |
| "loss": 0.284, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.8106508875739644, | |
| "eval_loss": 0.7146495580673218, | |
| "eval_runtime": 5.5227, | |
| "eval_samples_per_second": 16.115, | |
| "eval_steps_per_second": 2.173, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.9585798816568047, | |
| "grad_norm": 0.732421875, | |
| "learning_rate": 0.00010010010010010012, | |
| "loss": 0.2803, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.9585798816568047, | |
| "eval_loss": 0.7089855074882507, | |
| "eval_runtime": 5.499, | |
| "eval_samples_per_second": 16.185, | |
| "eval_steps_per_second": 2.182, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.106508875739645, | |
| "grad_norm": 0.59716796875, | |
| "learning_rate": 9.50950950950951e-05, | |
| "loss": 0.1915, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.106508875739645, | |
| "eval_loss": 0.8113237619400024, | |
| "eval_runtime": 5.5133, | |
| "eval_samples_per_second": 16.143, | |
| "eval_steps_per_second": 2.177, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.2544378698224854, | |
| "grad_norm": 0.76708984375, | |
| "learning_rate": 9.009009009009009e-05, | |
| "loss": 0.16, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.2544378698224854, | |
| "eval_loss": 0.8327358961105347, | |
| "eval_runtime": 5.4786, | |
| "eval_samples_per_second": 16.245, | |
| "eval_steps_per_second": 2.19, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.4023668639053253, | |
| "grad_norm": 0.7197265625, | |
| "learning_rate": 8.50850850850851e-05, | |
| "loss": 0.1621, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.4023668639053253, | |
| "eval_loss": 0.8468723297119141, | |
| "eval_runtime": 5.5083, | |
| "eval_samples_per_second": 16.157, | |
| "eval_steps_per_second": 2.179, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.5502958579881656, | |
| "grad_norm": 0.75146484375, | |
| "learning_rate": 8.008008008008009e-05, | |
| "loss": 0.163, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.5502958579881656, | |
| "eval_loss": 0.8476194143295288, | |
| "eval_runtime": 5.5014, | |
| "eval_samples_per_second": 16.178, | |
| "eval_steps_per_second": 2.181, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.698224852071006, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 7.507507507507507e-05, | |
| "loss": 0.1615, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 3.698224852071006, | |
| "eval_loss": 0.8421955108642578, | |
| "eval_runtime": 5.4752, | |
| "eval_samples_per_second": 16.255, | |
| "eval_steps_per_second": 2.192, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.85107421875, | |
| "learning_rate": 7.007007007007007e-05, | |
| "loss": 0.1737, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "eval_loss": 0.8518087267875671, | |
| "eval_runtime": 5.5254, | |
| "eval_samples_per_second": 16.107, | |
| "eval_steps_per_second": 2.172, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.994082840236686, | |
| "grad_norm": 0.87841796875, | |
| "learning_rate": 6.506506506506507e-05, | |
| "loss": 0.1685, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 3.994082840236686, | |
| "eval_loss": 0.857303261756897, | |
| "eval_runtime": 5.491, | |
| "eval_samples_per_second": 16.208, | |
| "eval_steps_per_second": 2.185, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 4.1420118343195265, | |
| "grad_norm": 0.63134765625, | |
| "learning_rate": 6.0060060060060066e-05, | |
| "loss": 0.0961, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.1420118343195265, | |
| "eval_loss": 0.9935606122016907, | |
| "eval_runtime": 5.4899, | |
| "eval_samples_per_second": 16.212, | |
| "eval_steps_per_second": 2.186, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.289940828402367, | |
| "grad_norm": 0.66650390625, | |
| "learning_rate": 5.505505505505506e-05, | |
| "loss": 0.0874, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 4.289940828402367, | |
| "eval_loss": 1.0188310146331787, | |
| "eval_runtime": 5.4814, | |
| "eval_samples_per_second": 16.237, | |
| "eval_steps_per_second": 2.189, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 4.437869822485207, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 5.005005005005006e-05, | |
| "loss": 0.0891, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.437869822485207, | |
| "eval_loss": 1.0284762382507324, | |
| "eval_runtime": 5.5187, | |
| "eval_samples_per_second": 16.127, | |
| "eval_steps_per_second": 2.174, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.585798816568047, | |
| "grad_norm": 0.6240234375, | |
| "learning_rate": 4.5045045045045046e-05, | |
| "loss": 0.0897, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 4.585798816568047, | |
| "eval_loss": 1.0269498825073242, | |
| "eval_runtime": 5.4812, | |
| "eval_samples_per_second": 16.237, | |
| "eval_steps_per_second": 2.189, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 4.733727810650888, | |
| "grad_norm": 0.5556640625, | |
| "learning_rate": 4.0040040040040046e-05, | |
| "loss": 0.0882, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.733727810650888, | |
| "eval_loss": 1.0333030223846436, | |
| "eval_runtime": 5.5654, | |
| "eval_samples_per_second": 15.992, | |
| "eval_steps_per_second": 2.156, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.881656804733728, | |
| "grad_norm": 0.91455078125, | |
| "learning_rate": 3.503503503503503e-05, | |
| "loss": 0.0889, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 4.881656804733728, | |
| "eval_loss": 1.0527359247207642, | |
| "eval_runtime": 5.4892, | |
| "eval_samples_per_second": 16.214, | |
| "eval_steps_per_second": 2.186, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 5.029585798816568, | |
| "grad_norm": 0.477294921875, | |
| "learning_rate": 3.0030030030030033e-05, | |
| "loss": 0.0826, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.029585798816568, | |
| "eval_loss": 1.0765188932418823, | |
| "eval_runtime": 5.5207, | |
| "eval_samples_per_second": 16.121, | |
| "eval_steps_per_second": 2.174, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.177514792899408, | |
| "grad_norm": 0.56689453125, | |
| "learning_rate": 2.502502502502503e-05, | |
| "loss": 0.0519, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 5.177514792899408, | |
| "eval_loss": 1.1578717231750488, | |
| "eval_runtime": 5.4744, | |
| "eval_samples_per_second": 16.258, | |
| "eval_steps_per_second": 2.192, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 5.325443786982248, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 2.0020020020020023e-05, | |
| "loss": 0.0513, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.325443786982248, | |
| "eval_loss": 1.1684048175811768, | |
| "eval_runtime": 5.482, | |
| "eval_samples_per_second": 16.235, | |
| "eval_steps_per_second": 2.189, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.4733727810650885, | |
| "grad_norm": 0.58251953125, | |
| "learning_rate": 1.5015015015015016e-05, | |
| "loss": 0.0523, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 5.4733727810650885, | |
| "eval_loss": 1.1905882358551025, | |
| "eval_runtime": 5.4845, | |
| "eval_samples_per_second": 16.228, | |
| "eval_steps_per_second": 2.188, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 5.621301775147929, | |
| "grad_norm": 0.53466796875, | |
| "learning_rate": 1.0010010010010011e-05, | |
| "loss": 0.0496, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 5.621301775147929, | |
| "eval_loss": 1.1795778274536133, | |
| "eval_runtime": 5.5354, | |
| "eval_samples_per_second": 16.078, | |
| "eval_steps_per_second": 2.168, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 5.769230769230769, | |
| "grad_norm": 0.432861328125, | |
| "learning_rate": 5.005005005005006e-06, | |
| "loss": 0.0495, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 5.769230769230769, | |
| "eval_loss": 1.1850156784057617, | |
| "eval_runtime": 5.5478, | |
| "eval_samples_per_second": 16.042, | |
| "eval_steps_per_second": 2.163, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 5.9171597633136095, | |
| "grad_norm": 0.495849609375, | |
| "learning_rate": 0.0, | |
| "loss": 0.0479, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.9171597633136095, | |
| "eval_loss": 1.18731689453125, | |
| "eval_runtime": 5.4886, | |
| "eval_samples_per_second": 16.215, | |
| "eval_steps_per_second": 2.186, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.9171597633136095, | |
| "step": 1000, | |
| "total_flos": 1.75885655212032e+17, | |
| "train_loss": 0.27937755072116854, | |
| "train_runtime": 1227.8544, | |
| "train_samples_per_second": 3.258, | |
| "train_steps_per_second": 0.814 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 1000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 25, | |
| "total_flos": 1.75885655212032e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |