| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 87, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03508771929824561, | |
| "grad_norm": 2.0672840185973786, | |
| "learning_rate": 0.0, | |
| "loss": 0.866, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07017543859649122, | |
| "grad_norm": 2.2899858891181903, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 1.043, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 1.910333546900785, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.9429, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.14035087719298245, | |
| "grad_norm": 1.624979132025543, | |
| "learning_rate": 1.2857142857142857e-05, | |
| "loss": 0.985, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.17543859649122806, | |
| "grad_norm": 1.3301367316845518, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 0.9899, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 2.192370335122931, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.9655, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.24561403508771928, | |
| "grad_norm": 2.2285319009652462, | |
| "learning_rate": 2.5714285714285714e-05, | |
| "loss": 0.9456, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2807017543859649, | |
| "grad_norm": 2.541438185031773, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9954, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 2.2177826928862956, | |
| "learning_rate": 2.999581555818041e-05, | |
| "loss": 1.0037, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 1.4333465946127133, | |
| "learning_rate": 2.9983264567328756e-05, | |
| "loss": 0.9675, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.38596491228070173, | |
| "grad_norm": 1.0448310378956214, | |
| "learning_rate": 2.9962354029963835e-05, | |
| "loss": 0.8559, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 1.0781647166558237, | |
| "learning_rate": 2.9933095612609253e-05, | |
| "loss": 0.8733, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.45614035087719296, | |
| "grad_norm": 1.007650519854371, | |
| "learning_rate": 2.989550563928436e-05, | |
| "loss": 0.9403, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.49122807017543857, | |
| "grad_norm": 0.7907009024129646, | |
| "learning_rate": 2.9849605082396678e-05, | |
| "loss": 0.8458, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.7042592030399148, | |
| "learning_rate": 2.9795419551040836e-05, | |
| "loss": 0.8727, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.5614035087719298, | |
| "grad_norm": 0.6419420174929784, | |
| "learning_rate": 2.973297927671063e-05, | |
| "loss": 0.8186, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5964912280701754, | |
| "grad_norm": 0.8076296952096572, | |
| "learning_rate": 2.966231909643208e-05, | |
| "loss": 0.9296, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.6598191860375567, | |
| "learning_rate": 2.958347843332696e-05, | |
| "loss": 0.8284, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.5560706149108875, | |
| "learning_rate": 2.949650127461764e-05, | |
| "loss": 0.8931, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 0.5418452815767641, | |
| "learning_rate": 2.940143614708549e-05, | |
| "loss": 0.8911, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.6442251215106719, | |
| "learning_rate": 2.9298336089996538e-05, | |
| "loss": 0.857, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.7719298245614035, | |
| "grad_norm": 0.5601338870306712, | |
| "learning_rate": 2.9187258625509518e-05, | |
| "loss": 0.8112, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8070175438596491, | |
| "grad_norm": 0.5147949384274566, | |
| "learning_rate": 2.906826572658278e-05, | |
| "loss": 0.8752, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.414227052576231, | |
| "learning_rate": 2.8941423782397987e-05, | |
| "loss": 0.8455, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 0.4724293164215497, | |
| "learning_rate": 2.8806803561319903e-05, | |
| "loss": 0.8332, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9122807017543859, | |
| "grad_norm": 0.5244546146099837, | |
| "learning_rate": 2.866448017141291e-05, | |
| "loss": 0.8992, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.49846452691797277, | |
| "learning_rate": 2.8514533018536286e-05, | |
| "loss": 0.7989, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.9824561403508771, | |
| "grad_norm": 0.5139730942456431, | |
| "learning_rate": 2.835704576204167e-05, | |
| "loss": 0.8281, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5139730942456431, | |
| "learning_rate": 2.8192106268097336e-05, | |
| "loss": 0.7089, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.0350877192982457, | |
| "grad_norm": 0.8522153801917955, | |
| "learning_rate": 2.801980656066545e-05, | |
| "loss": 0.7715, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0701754385964912, | |
| "grad_norm": 0.48621319476048247, | |
| "learning_rate": 2.78402427701595e-05, | |
| "loss": 0.7326, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.4931003602154137, | |
| "learning_rate": 2.7653515079810744e-05, | |
| "loss": 0.7796, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.1403508771929824, | |
| "grad_norm": 0.6820149173810659, | |
| "learning_rate": 2.7459727669773344e-05, | |
| "loss": 0.7558, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.1754385964912282, | |
| "grad_norm": 0.6039546286330506, | |
| "learning_rate": 2.725898865899967e-05, | |
| "loss": 0.7164, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.49413279486249445, | |
| "learning_rate": 2.705141004491792e-05, | |
| "loss": 0.771, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.2456140350877192, | |
| "grad_norm": 0.46540726916290787, | |
| "learning_rate": 2.6837107640945904e-05, | |
| "loss": 0.7745, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.280701754385965, | |
| "grad_norm": 0.7142115362037698, | |
| "learning_rate": 2.6616201011875792e-05, | |
| "loss": 0.7457, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.5357068778365942, | |
| "learning_rate": 2.638881340716583e-05, | |
| "loss": 0.7222, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.3508771929824561, | |
| "grad_norm": 0.46939662640391816, | |
| "learning_rate": 2.6155071692176348e-05, | |
| "loss": 0.6748, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.3859649122807016, | |
| "grad_norm": 0.537500418669669, | |
| "learning_rate": 2.5915106277388293e-05, | |
| "loss": 0.7777, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.59073194769042, | |
| "learning_rate": 2.566905104564393e-05, | |
| "loss": 0.7173, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.456140350877193, | |
| "grad_norm": 0.42637743115128474, | |
| "learning_rate": 2.541704327745013e-05, | |
| "loss": 0.6769, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.4912280701754386, | |
| "grad_norm": 0.4344573872029452, | |
| "learning_rate": 2.5159223574386117e-05, | |
| "loss": 0.6656, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.5018768196524176, | |
| "learning_rate": 2.489573578065821e-05, | |
| "loss": 0.802, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.5614035087719298, | |
| "grad_norm": 0.4529698728985265, | |
| "learning_rate": 2.4626726902845477e-05, | |
| "loss": 0.7586, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.5964912280701755, | |
| "grad_norm": 0.48235341277848787, | |
| "learning_rate": 2.4352347027881003e-05, | |
| "loss": 0.6852, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.3627100867848393, | |
| "learning_rate": 2.4072749239314565e-05, | |
| "loss": 0.6858, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.38664865341383037, | |
| "learning_rate": 2.3788089531903372e-05, | |
| "loss": 0.7365, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.7017543859649122, | |
| "grad_norm": 0.4545364316322588, | |
| "learning_rate": 2.3498526724578637e-05, | |
| "loss": 0.7538, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.518944898034962, | |
| "learning_rate": 2.320422237183641e-05, | |
| "loss": 0.7454, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.7719298245614035, | |
| "grad_norm": 0.4067307322789782, | |
| "learning_rate": 2.2905340673602184e-05, | |
| "loss": 0.7482, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.807017543859649, | |
| "grad_norm": 0.4862538497731571, | |
| "learning_rate": 2.26020483836196e-05, | |
| "loss": 0.7729, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.46087867219067935, | |
| "learning_rate": 2.229451471641422e-05, | |
| "loss": 0.6216, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.8771929824561404, | |
| "grad_norm": 0.5430309866121377, | |
| "learning_rate": 2.198291125288445e-05, | |
| "loss": 0.7631, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.912280701754386, | |
| "grad_norm": 0.5141331795467418, | |
| "learning_rate": 2.166741184457214e-05, | |
| "loss": 0.7101, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.39081537987082116, | |
| "learning_rate": 2.1348192516666376e-05, | |
| "loss": 0.7608, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.9824561403508771, | |
| "grad_norm": 0.4994585530441829, | |
| "learning_rate": 2.1025431369794546e-05, | |
| "loss": 0.7654, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.8215961436633803, | |
| "learning_rate": 2.0699308480655397e-05, | |
| "loss": 0.674, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.0350877192982457, | |
| "grad_norm": 0.7120328127826715, | |
| "learning_rate": 2.03700058015497e-05, | |
| "loss": 0.6013, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.0701754385964914, | |
| "grad_norm": 0.8161065595108032, | |
| "learning_rate": 2.0037707058864343e-05, | |
| "loss": 0.7149, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 1.8952005651095847, | |
| "learning_rate": 1.9702597650566723e-05, | |
| "loss": 0.6234, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.1403508771929824, | |
| "grad_norm": 0.5215133657210498, | |
| "learning_rate": 1.936486454276647e-05, | |
| "loss": 0.6001, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.175438596491228, | |
| "grad_norm": 0.8401793670285913, | |
| "learning_rate": 1.9024696165402272e-05, | |
| "loss": 0.6392, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.7463732624054893, | |
| "learning_rate": 1.8682282307111988e-05, | |
| "loss": 0.6392, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.245614035087719, | |
| "grad_norm": 0.449660878951785, | |
| "learning_rate": 1.8337814009344716e-05, | |
| "loss": 0.5725, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.280701754385965, | |
| "grad_norm": 0.6166851563782507, | |
| "learning_rate": 1.7991483459773887e-05, | |
| "loss": 0.6117, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.4765184980194071, | |
| "learning_rate": 1.7643483885070827e-05, | |
| "loss": 0.5877, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.3508771929824563, | |
| "grad_norm": 0.539140038588759, | |
| "learning_rate": 1.729400944309869e-05, | |
| "loss": 0.6497, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.3859649122807016, | |
| "grad_norm": 0.48435908336158434, | |
| "learning_rate": 1.6943255114586788e-05, | |
| "loss": 0.6004, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 1.0565779360065954, | |
| "learning_rate": 1.659141659434587e-05, | |
| "loss": 0.6898, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.456140350877193, | |
| "grad_norm": 0.44520837075665026, | |
| "learning_rate": 1.623869018208499e-05, | |
| "loss": 0.6189, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.4912280701754383, | |
| "grad_norm": 0.5141213138770342, | |
| "learning_rate": 1.5885272672890842e-05, | |
| "loss": 0.6262, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.5046502197354534, | |
| "learning_rate": 1.553136124743081e-05, | |
| "loss": 0.5617, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.56140350877193, | |
| "grad_norm": 0.47228102896457386, | |
| "learning_rate": 1.517715336194077e-05, | |
| "loss": 0.6288, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.5964912280701755, | |
| "grad_norm": 0.3945749736245955, | |
| "learning_rate": 1.4822846638059234e-05, | |
| "loss": 0.6189, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.47554210947045383, | |
| "learning_rate": 1.4468638752569193e-05, | |
| "loss": 0.6187, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.4168918402865985, | |
| "learning_rate": 1.411472732710916e-05, | |
| "loss": 0.6511, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.7017543859649122, | |
| "grad_norm": 0.3784629074937835, | |
| "learning_rate": 1.3761309817915017e-05, | |
| "loss": 0.6384, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.3766919801894043, | |
| "learning_rate": 1.340858340565413e-05, | |
| "loss": 0.6241, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.7719298245614032, | |
| "grad_norm": 0.3304739692544291, | |
| "learning_rate": 1.3056744885413216e-05, | |
| "loss": 0.5765, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.807017543859649, | |
| "grad_norm": 0.3593006941475982, | |
| "learning_rate": 1.2705990556901311e-05, | |
| "loss": 0.5384, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.3790675720863581, | |
| "learning_rate": 1.2356516114929176e-05, | |
| "loss": 0.6123, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.8771929824561404, | |
| "grad_norm": 0.38906936709954903, | |
| "learning_rate": 1.2008516540226115e-05, | |
| "loss": 0.6196, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.912280701754386, | |
| "grad_norm": 0.38879375158868906, | |
| "learning_rate": 1.1662185990655285e-05, | |
| "loss": 0.6376, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.37164660785811104, | |
| "learning_rate": 1.1317717692888014e-05, | |
| "loss": 0.5652, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.982456140350877, | |
| "grad_norm": 0.3398419044284849, | |
| "learning_rate": 1.0975303834597734e-05, | |
| "loss": 0.6338, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.3398419044284849, | |
| "learning_rate": 1.0635135457233533e-05, | |
| "loss": 0.4196, | |
| "step": 87 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 39139677044736.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |