| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.04294742008712191, | |
| "eval_steps": 20, | |
| "global_step": 700, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0012270691453463403, | |
| "grad_norm": 0.05003070831298828, | |
| "learning_rate": 0.00019981588314717073, | |
| "loss": 2.6972, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0012270691453463403, | |
| "eval_loss": 2.2967841625213623, | |
| "eval_runtime": 23.2641, | |
| "eval_samples_per_second": 4.298, | |
| "eval_steps_per_second": 0.559, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0024541382906926807, | |
| "grad_norm": 0.07180789858102798, | |
| "learning_rate": 0.00019957039401006504, | |
| "loss": 2.2022, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0024541382906926807, | |
| "eval_loss": 2.068006992340088, | |
| "eval_runtime": 23.5719, | |
| "eval_samples_per_second": 4.242, | |
| "eval_steps_per_second": 0.552, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.003681207436039021, | |
| "grad_norm": 0.08049603551626205, | |
| "learning_rate": 0.00019932490487295938, | |
| "loss": 2.0529, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.003681207436039021, | |
| "eval_loss": 1.9338455200195312, | |
| "eval_runtime": 23.1495, | |
| "eval_samples_per_second": 4.32, | |
| "eval_steps_per_second": 0.562, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.004908276581385361, | |
| "grad_norm": 0.08653070032596588, | |
| "learning_rate": 0.00019907941573585368, | |
| "loss": 1.9395, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.004908276581385361, | |
| "eval_loss": 1.8689138889312744, | |
| "eval_runtime": 23.7275, | |
| "eval_samples_per_second": 4.215, | |
| "eval_steps_per_second": 0.548, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.006135345726731701, | |
| "grad_norm": 0.08481493592262268, | |
| "learning_rate": 0.00019883392659874802, | |
| "loss": 1.8773, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.006135345726731701, | |
| "eval_loss": 1.8180441856384277, | |
| "eval_runtime": 23.5705, | |
| "eval_samples_per_second": 4.243, | |
| "eval_steps_per_second": 0.552, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.007362414872078042, | |
| "grad_norm": 0.11568216979503632, | |
| "learning_rate": 0.00019858843746164233, | |
| "loss": 1.7827, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.007362414872078042, | |
| "eval_loss": 1.774067997932434, | |
| "eval_runtime": 23.9722, | |
| "eval_samples_per_second": 4.172, | |
| "eval_steps_per_second": 0.542, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.008589484017424381, | |
| "grad_norm": 0.10869361460208893, | |
| "learning_rate": 0.00019834294832453666, | |
| "loss": 1.812, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.008589484017424381, | |
| "eval_loss": 1.737804889678955, | |
| "eval_runtime": 23.519, | |
| "eval_samples_per_second": 4.252, | |
| "eval_steps_per_second": 0.553, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.009816553162770723, | |
| "grad_norm": 0.0976206362247467, | |
| "learning_rate": 0.00019809745918743097, | |
| "loss": 1.74, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.009816553162770723, | |
| "eval_loss": 1.700899600982666, | |
| "eval_runtime": 23.2347, | |
| "eval_samples_per_second": 4.304, | |
| "eval_steps_per_second": 0.56, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.011043622308117063, | |
| "grad_norm": 0.1123971939086914, | |
| "learning_rate": 0.00019785197005032528, | |
| "loss": 1.787, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.011043622308117063, | |
| "eval_loss": 1.6765294075012207, | |
| "eval_runtime": 23.6403, | |
| "eval_samples_per_second": 4.23, | |
| "eval_steps_per_second": 0.55, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.012270691453463402, | |
| "grad_norm": 0.10320968925952911, | |
| "learning_rate": 0.0001976064809132196, | |
| "loss": 1.7804, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.012270691453463402, | |
| "eval_loss": 1.6563650369644165, | |
| "eval_runtime": 23.6381, | |
| "eval_samples_per_second": 4.23, | |
| "eval_steps_per_second": 0.55, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.013497760598809742, | |
| "grad_norm": 0.14491896331310272, | |
| "learning_rate": 0.00019736099177611392, | |
| "loss": 1.7043, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.013497760598809742, | |
| "eval_loss": 1.6346065998077393, | |
| "eval_runtime": 23.7121, | |
| "eval_samples_per_second": 4.217, | |
| "eval_steps_per_second": 0.548, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.014724829744156084, | |
| "grad_norm": 0.12502990663051605, | |
| "learning_rate": 0.00019711550263900825, | |
| "loss": 1.7345, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.014724829744156084, | |
| "eval_loss": 1.6147732734680176, | |
| "eval_runtime": 23.5936, | |
| "eval_samples_per_second": 4.238, | |
| "eval_steps_per_second": 0.551, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.015951898889502422, | |
| "grad_norm": 0.1230228915810585, | |
| "learning_rate": 0.00019687001350190256, | |
| "loss": 1.7338, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.015951898889502422, | |
| "eval_loss": 1.5957908630371094, | |
| "eval_runtime": 23.389, | |
| "eval_samples_per_second": 4.276, | |
| "eval_steps_per_second": 0.556, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.017178968034848762, | |
| "grad_norm": 0.12000931799411774, | |
| "learning_rate": 0.00019662452436479687, | |
| "loss": 1.7143, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.017178968034848762, | |
| "eval_loss": 1.585697889328003, | |
| "eval_runtime": 23.566, | |
| "eval_samples_per_second": 4.243, | |
| "eval_steps_per_second": 0.552, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.018406037180195105, | |
| "grad_norm": 0.1442350149154663, | |
| "learning_rate": 0.00019637903522769118, | |
| "loss": 1.6406, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.018406037180195105, | |
| "eval_loss": 1.5710804462432861, | |
| "eval_runtime": 23.5083, | |
| "eval_samples_per_second": 4.254, | |
| "eval_steps_per_second": 0.553, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.019633106325541445, | |
| "grad_norm": 0.09555982798337936, | |
| "learning_rate": 0.00019613354609058549, | |
| "loss": 1.6213, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.019633106325541445, | |
| "eval_loss": 1.5556869506835938, | |
| "eval_runtime": 23.5239, | |
| "eval_samples_per_second": 4.251, | |
| "eval_steps_per_second": 0.553, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.020860175470887785, | |
| "grad_norm": 0.13320715725421906, | |
| "learning_rate": 0.00019588805695347982, | |
| "loss": 1.6956, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.020860175470887785, | |
| "eval_loss": 1.5424914360046387, | |
| "eval_runtime": 23.6064, | |
| "eval_samples_per_second": 4.236, | |
| "eval_steps_per_second": 0.551, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.022087244616234125, | |
| "grad_norm": 0.12061001360416412, | |
| "learning_rate": 0.00019564256781637413, | |
| "loss": 1.6589, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.022087244616234125, | |
| "eval_loss": 1.528477430343628, | |
| "eval_runtime": 23.6796, | |
| "eval_samples_per_second": 4.223, | |
| "eval_steps_per_second": 0.549, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.023314313761580465, | |
| "grad_norm": 0.14327766001224518, | |
| "learning_rate": 0.00019539707867926844, | |
| "loss": 1.5946, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.023314313761580465, | |
| "eval_loss": 1.52202570438385, | |
| "eval_runtime": 23.6756, | |
| "eval_samples_per_second": 4.224, | |
| "eval_steps_per_second": 0.549, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.024541382906926805, | |
| "grad_norm": 0.12291988730430603, | |
| "learning_rate": 0.00019515158954216277, | |
| "loss": 1.5366, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.024541382906926805, | |
| "eval_loss": 1.507960319519043, | |
| "eval_runtime": 23.6216, | |
| "eval_samples_per_second": 4.233, | |
| "eval_steps_per_second": 0.55, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.025768452052273145, | |
| "grad_norm": 0.15288175642490387, | |
| "learning_rate": 0.00019490610040505708, | |
| "loss": 1.5829, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.025768452052273145, | |
| "eval_loss": 1.4994325637817383, | |
| "eval_runtime": 23.6368, | |
| "eval_samples_per_second": 4.231, | |
| "eval_steps_per_second": 0.55, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.026995521197619485, | |
| "grad_norm": 0.13319191336631775, | |
| "learning_rate": 0.0001946606112679514, | |
| "loss": 1.5523, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.026995521197619485, | |
| "eval_loss": 1.4956778287887573, | |
| "eval_runtime": 23.6921, | |
| "eval_samples_per_second": 4.221, | |
| "eval_steps_per_second": 0.549, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.028222590342965825, | |
| "grad_norm": 0.14759239554405212, | |
| "learning_rate": 0.00019441512213084572, | |
| "loss": 1.5735, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.028222590342965825, | |
| "eval_loss": 1.486402988433838, | |
| "eval_runtime": 23.2911, | |
| "eval_samples_per_second": 4.293, | |
| "eval_steps_per_second": 0.558, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.029449659488312168, | |
| "grad_norm": 0.11428073793649673, | |
| "learning_rate": 0.00019416963299374006, | |
| "loss": 1.5788, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.029449659488312168, | |
| "eval_loss": 1.4712104797363281, | |
| "eval_runtime": 23.4851, | |
| "eval_samples_per_second": 4.258, | |
| "eval_steps_per_second": 0.554, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.030676728633658508, | |
| "grad_norm": 0.11649870127439499, | |
| "learning_rate": 0.00019392414385663436, | |
| "loss": 1.5667, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.030676728633658508, | |
| "eval_loss": 1.4620152711868286, | |
| "eval_runtime": 23.5455, | |
| "eval_samples_per_second": 4.247, | |
| "eval_steps_per_second": 0.552, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.031903797779004844, | |
| "grad_norm": 0.16019868850708008, | |
| "learning_rate": 0.00019367865471952867, | |
| "loss": 1.4778, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.031903797779004844, | |
| "eval_loss": 1.4597880840301514, | |
| "eval_runtime": 23.624, | |
| "eval_samples_per_second": 4.233, | |
| "eval_steps_per_second": 0.55, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.03313086692435119, | |
| "grad_norm": 0.1370091289281845, | |
| "learning_rate": 0.00019343316558242298, | |
| "loss": 1.5531, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.03313086692435119, | |
| "eval_loss": 1.443243384361267, | |
| "eval_runtime": 23.5537, | |
| "eval_samples_per_second": 4.246, | |
| "eval_steps_per_second": 0.552, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.034357936069697524, | |
| "grad_norm": 0.1211417093873024, | |
| "learning_rate": 0.0001931876764453173, | |
| "loss": 1.5879, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.034357936069697524, | |
| "eval_loss": 1.4466437101364136, | |
| "eval_runtime": 23.8508, | |
| "eval_samples_per_second": 4.193, | |
| "eval_steps_per_second": 0.545, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.03558500521504387, | |
| "grad_norm": 0.14397528767585754, | |
| "learning_rate": 0.00019294218730821162, | |
| "loss": 1.5352, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.03558500521504387, | |
| "eval_loss": 1.4339115619659424, | |
| "eval_runtime": 23.649, | |
| "eval_samples_per_second": 4.229, | |
| "eval_steps_per_second": 0.55, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.03681207436039021, | |
| "grad_norm": 0.12468410283327103, | |
| "learning_rate": 0.00019269669817110593, | |
| "loss": 1.5045, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.03681207436039021, | |
| "eval_loss": 1.4277862310409546, | |
| "eval_runtime": 23.647, | |
| "eval_samples_per_second": 4.229, | |
| "eval_steps_per_second": 0.55, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.03803914350573655, | |
| "grad_norm": 0.1577584445476532, | |
| "learning_rate": 0.00019245120903400024, | |
| "loss": 1.5497, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.03803914350573655, | |
| "eval_loss": 1.4203659296035767, | |
| "eval_runtime": 23.8622, | |
| "eval_samples_per_second": 4.191, | |
| "eval_steps_per_second": 0.545, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.03926621265108289, | |
| "grad_norm": 0.12410438805818558, | |
| "learning_rate": 0.00019220571989689457, | |
| "loss": 1.503, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.03926621265108289, | |
| "eval_loss": 1.4154139757156372, | |
| "eval_runtime": 23.4706, | |
| "eval_samples_per_second": 4.261, | |
| "eval_steps_per_second": 0.554, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.04049328179642923, | |
| "grad_norm": 0.13563913106918335, | |
| "learning_rate": 0.00019196023075978888, | |
| "loss": 1.4851, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.04049328179642923, | |
| "eval_loss": 1.414802074432373, | |
| "eval_runtime": 23.3961, | |
| "eval_samples_per_second": 4.274, | |
| "eval_steps_per_second": 0.556, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.04172035094177557, | |
| "grad_norm": 0.13915061950683594, | |
| "learning_rate": 0.00019171474162268321, | |
| "loss": 1.4847, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.04172035094177557, | |
| "eval_loss": 1.4029760360717773, | |
| "eval_runtime": 23.6066, | |
| "eval_samples_per_second": 4.236, | |
| "eval_steps_per_second": 0.551, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.04294742008712191, | |
| "grad_norm": 0.14418162405490875, | |
| "learning_rate": 0.00019146925248557752, | |
| "loss": 1.4724, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.04294742008712191, | |
| "eval_loss": 1.4029196500778198, | |
| "eval_runtime": 23.5244, | |
| "eval_samples_per_second": 4.251, | |
| "eval_steps_per_second": 0.553, | |
| "step": 700 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 16299, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "total_flos": 5.08603168290816e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |