| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.0045333998231974065, | |
| "eval_steps": 4, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 9.066799646394814e-05, | |
| "grad_norm": 1.305655837059021, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9466, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 9.066799646394814e-05, | |
| "eval_loss": 1.1514371633529663, | |
| "eval_runtime": 281.757, | |
| "eval_samples_per_second": 8.241, | |
| "eval_steps_per_second": 8.241, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0001813359929278963, | |
| "grad_norm": 1.7540494203567505, | |
| "learning_rate": 4e-05, | |
| "loss": 1.2689, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00027200398939184443, | |
| "grad_norm": 1.3806679248809814, | |
| "learning_rate": 6e-05, | |
| "loss": 0.9866, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0003626719858557926, | |
| "grad_norm": 1.6540546417236328, | |
| "learning_rate": 8e-05, | |
| "loss": 0.886, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0003626719858557926, | |
| "eval_loss": 1.063585877418518, | |
| "eval_runtime": 281.615, | |
| "eval_samples_per_second": 8.245, | |
| "eval_steps_per_second": 8.245, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.00045333998231974066, | |
| "grad_norm": 1.351820945739746, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0148, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0005440079787836889, | |
| "grad_norm": 1.6633617877960205, | |
| "learning_rate": 0.00012, | |
| "loss": 1.1922, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.000634675975247637, | |
| "grad_norm": 1.216033697128296, | |
| "learning_rate": 0.00014, | |
| "loss": 0.7838, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0007253439717115851, | |
| "grad_norm": 1.0886664390563965, | |
| "learning_rate": 0.00016, | |
| "loss": 0.6905, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0007253439717115851, | |
| "eval_loss": 0.6156781911849976, | |
| "eval_runtime": 282.2824, | |
| "eval_samples_per_second": 8.226, | |
| "eval_steps_per_second": 8.226, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0008160119681755333, | |
| "grad_norm": 1.2936314344406128, | |
| "learning_rate": 0.00018, | |
| "loss": 0.6603, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0009066799646394813, | |
| "grad_norm": 1.429738163948059, | |
| "learning_rate": 0.0002, | |
| "loss": 0.518, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0009973479611034295, | |
| "grad_norm": 1.501217007637024, | |
| "learning_rate": 0.0001996917333733128, | |
| "loss": 0.475, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0010880159575673777, | |
| "grad_norm": 1.0113286972045898, | |
| "learning_rate": 0.00019876883405951377, | |
| "loss": 0.3233, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0010880159575673777, | |
| "eval_loss": 0.29599475860595703, | |
| "eval_runtime": 282.4723, | |
| "eval_samples_per_second": 8.22, | |
| "eval_steps_per_second": 8.22, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0011786839540313258, | |
| "grad_norm": 1.045129656791687, | |
| "learning_rate": 0.00019723699203976766, | |
| "loss": 0.305, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.001269351950495274, | |
| "grad_norm": 0.7782332897186279, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 0.2367, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.001360019946959222, | |
| "grad_norm": 1.1300699710845947, | |
| "learning_rate": 0.0001923879532511287, | |
| "loss": 0.1923, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0014506879434231703, | |
| "grad_norm": 0.6594939231872559, | |
| "learning_rate": 0.0001891006524188368, | |
| "loss": 0.1823, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0014506879434231703, | |
| "eval_loss": 0.2610986828804016, | |
| "eval_runtime": 282.2669, | |
| "eval_samples_per_second": 8.226, | |
| "eval_steps_per_second": 8.226, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0015413559398871183, | |
| "grad_norm": 1.0713460445404053, | |
| "learning_rate": 0.00018526401643540922, | |
| "loss": 0.2457, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0016320239363510666, | |
| "grad_norm": 1.2344204187393188, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 0.272, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0017226919328150146, | |
| "grad_norm": 0.7098881006240845, | |
| "learning_rate": 0.0001760405965600031, | |
| "loss": 0.225, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0018133599292789627, | |
| "grad_norm": 0.8920848965644836, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.1479, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0018133599292789627, | |
| "eval_loss": 0.23455704748630524, | |
| "eval_runtime": 282.1688, | |
| "eval_samples_per_second": 8.229, | |
| "eval_steps_per_second": 8.229, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.001904027925742911, | |
| "grad_norm": 0.5211899876594543, | |
| "learning_rate": 0.00016494480483301836, | |
| "loss": 0.1664, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.001994695922206859, | |
| "grad_norm": 0.6904815435409546, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 0.2361, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.002085363918670807, | |
| "grad_norm": 0.6779294013977051, | |
| "learning_rate": 0.0001522498564715949, | |
| "loss": 0.1939, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0021760319151347554, | |
| "grad_norm": 0.8565766215324402, | |
| "learning_rate": 0.00014539904997395468, | |
| "loss": 0.2695, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0021760319151347554, | |
| "eval_loss": 0.22548052668571472, | |
| "eval_runtime": 282.2744, | |
| "eval_samples_per_second": 8.226, | |
| "eval_steps_per_second": 8.226, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0022666999115987033, | |
| "grad_norm": 0.7228334546089172, | |
| "learning_rate": 0.000138268343236509, | |
| "loss": 0.2259, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0023573679080626515, | |
| "grad_norm": 0.9707162976264954, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.2173, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0024480359045265998, | |
| "grad_norm": 0.5939818024635315, | |
| "learning_rate": 0.00012334453638559057, | |
| "loss": 0.2012, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.002538703900990548, | |
| "grad_norm": 0.3913775682449341, | |
| "learning_rate": 0.0001156434465040231, | |
| "loss": 0.1645, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.002538703900990548, | |
| "eval_loss": 0.21409040689468384, | |
| "eval_runtime": 282.2725, | |
| "eval_samples_per_second": 8.226, | |
| "eval_steps_per_second": 8.226, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.002629371897454496, | |
| "grad_norm": 0.652338981628418, | |
| "learning_rate": 0.0001078459095727845, | |
| "loss": 0.1954, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.002720039893918444, | |
| "grad_norm": 0.5130963325500488, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1994, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0028107078903823923, | |
| "grad_norm": 0.38596802949905396, | |
| "learning_rate": 9.215409042721552e-05, | |
| "loss": 0.1804, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0029013758868463406, | |
| "grad_norm": 0.481663316488266, | |
| "learning_rate": 8.435655349597689e-05, | |
| "loss": 0.1804, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0029013758868463406, | |
| "eval_loss": 0.20881743729114532, | |
| "eval_runtime": 281.3146, | |
| "eval_samples_per_second": 8.254, | |
| "eval_steps_per_second": 8.254, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0029920438833102884, | |
| "grad_norm": 0.3412702679634094, | |
| "learning_rate": 7.66554636144095e-05, | |
| "loss": 0.2066, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0030827118797742367, | |
| "grad_norm": 0.5542587041854858, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.1895, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.003173379876238185, | |
| "grad_norm": 0.5084295868873596, | |
| "learning_rate": 6.173165676349103e-05, | |
| "loss": 0.2111, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.003264047872702133, | |
| "grad_norm": 0.6083145141601562, | |
| "learning_rate": 5.4600950026045326e-05, | |
| "loss": 0.2318, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.003264047872702133, | |
| "eval_loss": 0.20627427101135254, | |
| "eval_runtime": 282.4785, | |
| "eval_samples_per_second": 8.22, | |
| "eval_steps_per_second": 8.22, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.003354715869166081, | |
| "grad_norm": 0.4692092835903168, | |
| "learning_rate": 4.7750143528405126e-05, | |
| "loss": 0.1626, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0034453838656300292, | |
| "grad_norm": 0.4230363965034485, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 0.1752, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0035360518620939775, | |
| "grad_norm": 0.5133092403411865, | |
| "learning_rate": 3.5055195166981645e-05, | |
| "loss": 0.209, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0036267198585579253, | |
| "grad_norm": 0.3762070834636688, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.1563, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0036267198585579253, | |
| "eval_loss": 0.20195676386356354, | |
| "eval_runtime": 282.3618, | |
| "eval_samples_per_second": 8.223, | |
| "eval_steps_per_second": 8.223, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0037173878550218736, | |
| "grad_norm": 0.3519875407218933, | |
| "learning_rate": 2.3959403439996907e-05, | |
| "loss": 0.1516, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.003808055851485822, | |
| "grad_norm": 0.41183072328567505, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 0.1585, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.00389872384794977, | |
| "grad_norm": 0.613614022731781, | |
| "learning_rate": 1.4735983564590783e-05, | |
| "loss": 0.1563, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.003989391844413718, | |
| "grad_norm": 0.4138711988925934, | |
| "learning_rate": 1.0899347581163221e-05, | |
| "loss": 0.1512, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.003989391844413718, | |
| "eval_loss": 0.20044909417629242, | |
| "eval_runtime": 282.2051, | |
| "eval_samples_per_second": 8.228, | |
| "eval_steps_per_second": 8.228, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.004080059840877667, | |
| "grad_norm": 0.4004136025905609, | |
| "learning_rate": 7.612046748871327e-06, | |
| "loss": 0.1641, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.004170727837341614, | |
| "grad_norm": 0.4379856586456299, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 0.1492, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.004261395833805562, | |
| "grad_norm": 0.32673540711402893, | |
| "learning_rate": 2.7630079602323442e-06, | |
| "loss": 0.1826, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.004352063830269511, | |
| "grad_norm": 0.2924938499927521, | |
| "learning_rate": 1.231165940486234e-06, | |
| "loss": 0.1529, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.004352063830269511, | |
| "eval_loss": 0.20006079971790314, | |
| "eval_runtime": 281.9182, | |
| "eval_samples_per_second": 8.236, | |
| "eval_steps_per_second": 8.236, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.004442731826733459, | |
| "grad_norm": 0.44641199707984924, | |
| "learning_rate": 3.0826662668720364e-07, | |
| "loss": 0.1428, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.0045333998231974065, | |
| "grad_norm": 0.2606041431427002, | |
| "learning_rate": 0.0, | |
| "loss": 0.1358, | |
| "step": 50 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 50, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 4, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.8825716774731776e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |