{"loss": 4.099612808227539, "grad_norm": 3.3916518688201904, "learning_rate": 7.82608695652174e-05, "entropy": 1.0828558593988418, "mean_token_accuracy": 0.44957960266619923, "num_tokens": 24658.0, "epoch": 0.13245033112582782, "step": 10}
{"eval_loss": 0.5499903559684753, "eval_runtime": 9.3737, "eval_samples_per_second": 16.002, "eval_steps_per_second": 4.054, "eval_entropy": 0.5737218025483584, "eval_mean_token_accuracy": 0.7952850923726433, "eval_num_tokens": 24658.0, "epoch": 0.13245033112582782, "step": 10}
{"loss": 0.4331226825714111, "grad_norm": 3.621782064437866, "learning_rate": 0.00016521739130434784, "entropy": 0.4598923083394766, "mean_token_accuracy": 0.8161924198269844, "num_tokens": 48994.0, "epoch": 0.26490066225165565, "step": 20}
{"eval_loss": 0.5726626515388489, "eval_runtime": 9.3836, "eval_samples_per_second": 15.985, "eval_steps_per_second": 4.05, "eval_entropy": 0.3104277225701432, "eval_mean_token_accuracy": 0.8032894793309664, "eval_num_tokens": 48994.0, "epoch": 0.26490066225165565, "step": 20}
{"loss": 0.37956321239471436, "grad_norm": 3.6681301593780518, "learning_rate": 0.00019414634146341464, "entropy": 0.35528335012495516, "mean_token_accuracy": 0.8529162302613258, "num_tokens": 73216.0, "epoch": 0.3973509933774834, "step": 30}
{"eval_loss": 0.4902452528476715, "eval_runtime": 9.6515, "eval_samples_per_second": 15.542, "eval_steps_per_second": 3.937, "eval_entropy": 0.3591178828164151, "eval_mean_token_accuracy": 0.8244517586733165, "eval_num_tokens": 73216.0, "epoch": 0.3973509933774834, "step": 30}
{"loss": 0.3535017728805542, "grad_norm": 1.529909372329712, "learning_rate": 0.00018439024390243903, "entropy": 0.31258961334824564, "mean_token_accuracy": 0.8600528433918952, "num_tokens": 97165.0, "epoch": 0.5298013245033113, "step": 40}
{"eval_loss": 0.5773032903671265, "eval_runtime": 8.8699, "eval_samples_per_second": 16.911, "eval_steps_per_second": 4.284, "eval_entropy": 0.40491867006609317, "eval_mean_token_accuracy": 0.7638627878929439, "eval_num_tokens": 97165.0, "epoch": 0.5298013245033113, "step": 40}
{"loss": 0.29483020305633545, "grad_norm": 1.4113677740097046, "learning_rate": 0.00017463414634146342, "entropy": 0.3084935720078647, "mean_token_accuracy": 0.8778989180922508, "num_tokens": 121376.0, "epoch": 0.6622516556291391, "step": 50}
{"eval_loss": 0.5696188807487488, "eval_runtime": 8.921, "eval_samples_per_second": 16.814, "eval_steps_per_second": 4.26, "eval_entropy": 0.29807096403582317, "eval_mean_token_accuracy": 0.8039630376978925, "eval_num_tokens": 121376.0, "epoch": 0.6622516556291391, "step": 50}
{"loss": 0.26492722034454347, "grad_norm": 2.0661537647247314, "learning_rate": 0.00016487804878048782, "entropy": 0.268452774733305, "mean_token_accuracy": 0.8974904179573059, "num_tokens": 145557.0, "epoch": 0.7947019867549668, "step": 60}
{"eval_loss": 0.8559898138046265, "eval_runtime": 8.9149, "eval_samples_per_second": 16.826, "eval_steps_per_second": 4.263, "eval_entropy": 0.2022110644458352, "eval_mean_token_accuracy": 0.8000626611082178, "eval_num_tokens": 145557.0, "epoch": 0.7947019867549668, "step": 60}
{"loss": 0.18335200548171998, "grad_norm": 3.771090269088745, "learning_rate": 0.0001551219512195122, "entropy": 0.12369259365368634, "mean_token_accuracy": 0.9281293749809265, "num_tokens": 170605.0, "epoch": 0.9271523178807947, "step": 70}
{"eval_loss": 0.8269464373588562, "eval_runtime": 8.912, "eval_samples_per_second": 16.831, "eval_steps_per_second": 4.264, "eval_entropy": 0.23079313014290834, "eval_mean_token_accuracy": 0.7807487523869464, "eval_num_tokens": 170605.0, "epoch": 0.9271523178807947, "step": 70}
{"loss": 0.18180986642837524, "grad_norm": 2.6767499446868896, "learning_rate": 0.0001453658536585366, "entropy": 0.16967114948324466, "mean_token_accuracy": 0.9279846925484506, "num_tokens": 193719.0, "epoch": 1.0529801324503312, "step": 80}
{"eval_loss": 0.8880820870399475, "eval_runtime": 8.8086, "eval_samples_per_second": 17.029, "eval_steps_per_second": 4.314, "eval_entropy": 0.2405679833119441, "eval_mean_token_accuracy": 0.7817669220660862, "eval_num_tokens": 193719.0, "epoch": 1.0529801324503312, "step": 80}
{"loss": 0.11491576433181763, "grad_norm": 0.69949871301651, "learning_rate": 0.000135609756097561, "entropy": 0.14066557474434377, "mean_token_accuracy": 0.9485534891486168, "num_tokens": 218542.0, "epoch": 1.185430463576159, "step": 90}
{"eval_loss": 0.7975587844848633, "eval_runtime": 8.8612, "eval_samples_per_second": 16.928, "eval_steps_per_second": 4.288, "eval_entropy": 0.19119705641826026, "eval_mean_token_accuracy": 0.7988721861651069, "eval_num_tokens": 218542.0, "epoch": 1.185430463576159, "step": 90}
{"loss": 0.08289253115653991, "grad_norm": 2.014707565307617, "learning_rate": 0.00012585365853658536, "entropy": 0.05604892138508148, "mean_token_accuracy": 0.9736187428236007, "num_tokens": 243302.0, "epoch": 1.3178807947019868, "step": 100}
{"eval_loss": 1.1870805025100708, "eval_runtime": 8.8416, "eval_samples_per_second": 16.965, "eval_steps_per_second": 4.298, "eval_entropy": 0.114802976807512, "eval_mean_token_accuracy": 0.8206453684129213, "eval_num_tokens": 243302.0, "epoch": 1.3178807947019868, "step": 100}
{"loss": 0.06358585357666016, "grad_norm": 1.5396313667297363, "learning_rate": 0.00011609756097560975, "entropy": 0.04575981950329151, "mean_token_accuracy": 0.9785756185650826, "num_tokens": 267729.0, "epoch": 1.4503311258278146, "step": 110}
{"eval_loss": 0.8791955709457397, "eval_runtime": 8.8863, "eval_samples_per_second": 16.88, "eval_steps_per_second": 4.276, "eval_entropy": 0.17355985742515737, "eval_mean_token_accuracy": 0.8070958703756332, "eval_num_tokens": 267729.0, "epoch": 1.4503311258278146, "step": 110}
{"loss": 0.049860316514968875, "grad_norm": 0.678949773311615, "learning_rate": 0.00010634146341463416, "entropy": 0.07773116882308387, "mean_token_accuracy": 0.9833469331264496, "num_tokens": 292279.0, "epoch": 1.5827814569536423, "step": 120}
{"eval_loss": 1.1025115251541138, "eval_runtime": 8.9449, "eval_samples_per_second": 16.769, "eval_steps_per_second": 4.248, "eval_entropy": 0.1594012568683339, "eval_mean_token_accuracy": 0.7980889779956717, "eval_num_tokens": 292279.0, "epoch": 1.5827814569536423, "step": 120}
{"loss": 0.07376678586006165, "grad_norm": 2.6976561546325684, "learning_rate": 9.658536585365854e-05, "entropy": 0.0441209072858328, "mean_token_accuracy": 0.9765854984521866, "num_tokens": 316471.0, "epoch": 1.7152317880794703, "step": 130}
{"eval_loss": 1.2516794204711914, "eval_runtime": 8.9032, "eval_samples_per_second": 16.848, "eval_steps_per_second": 4.268, "eval_entropy": 0.12382908019782535, "eval_mean_token_accuracy": 0.770441735261365, "eval_num_tokens": 316471.0, "epoch": 1.7152317880794703, "step": 130}
{"loss": 0.02282872200012207, "grad_norm": 0.8916032314300537, "learning_rate": 8.682926829268293e-05, "entropy": 0.0408805181941716, "mean_token_accuracy": 0.9929233521223069, "num_tokens": 340856.0, "epoch": 1.847682119205298, "step": 140}
{"eval_loss": 1.273831844329834, "eval_runtime": 8.873, "eval_samples_per_second": 16.905, "eval_steps_per_second": 4.283, "eval_entropy": 0.11421615798501175, "eval_mean_token_accuracy": 0.7660557698262366, "eval_num_tokens": 340856.0, "epoch": 1.847682119205298, "step": 140}
{"loss": 0.020523886382579803, "grad_norm": 2.7703590393066406, "learning_rate": 7.707317073170732e-05, "entropy": 0.028159727145248326, "mean_token_accuracy": 0.9912698417901993, "num_tokens": 364928.0, "epoch": 1.980132450331126, "step": 150}
{"eval_loss": 1.5218627452850342, "eval_runtime": 8.8524, "eval_samples_per_second": 16.945, "eval_steps_per_second": 4.293, "eval_entropy": 0.09264942740800937, "eval_mean_token_accuracy": 0.7557487519163835, "eval_num_tokens": 364928.0, "epoch": 1.980132450331126, "step": 150}
{"loss": 0.015184570848941804, "grad_norm": 0.04658397287130356, "learning_rate": 6.731707317073171e-05, "entropy": 0.016722216987737307, "mean_token_accuracy": 0.9981203000796469, "num_tokens": 387571.0, "epoch": 2.1059602649006623, "step": 160}
{"eval_loss": 1.6605678796768188, "eval_runtime": 8.8568, "eval_samples_per_second": 16.936, "eval_steps_per_second": 4.29, "eval_entropy": 0.07781053681920687, "eval_mean_token_accuracy": 0.770441735261365, "eval_num_tokens": 387571.0, "epoch": 2.1059602649006623, "step": 160}
{"loss": 0.006011705845594406, "grad_norm": 0.013418244197964668, "learning_rate": 5.756097560975609e-05, "entropy": 0.007514824125610176, "mean_token_accuracy": 0.9968627452850342, "num_tokens": 412476.0, "epoch": 2.23841059602649, "step": 170}
{"eval_loss": 1.763867974281311, "eval_runtime": 8.8357, "eval_samples_per_second": 16.977, "eval_steps_per_second": 4.301, "eval_entropy": 0.07184799665475639, "eval_mean_token_accuracy": 0.7737312089455756, "eval_num_tokens": 412476.0, "epoch": 2.23841059602649, "step": 170}
{"loss": 0.004858419671654702, "grad_norm": 0.14898061752319336, "learning_rate": 4.7804878048780485e-05, "entropy": 0.004944811350287637, "mean_token_accuracy": 0.9980769231915474, "num_tokens": 436969.0, "epoch": 2.370860927152318, "step": 180}
{"eval_loss": 1.827174425125122, "eval_runtime": 8.8372, "eval_samples_per_second": 16.974, "eval_steps_per_second": 4.3, "eval_entropy": 0.06969891959229962, "eval_mean_token_accuracy": 0.7720864721034703, "eval_num_tokens": 436969.0, "epoch": 2.370860927152318, "step": 180}
{"loss": 0.005421416833996773, "grad_norm": 0.08094702661037445, "learning_rate": 3.804878048780488e-05, "entropy": 0.004511333805567119, "mean_token_accuracy": 0.9966666668653488, "num_tokens": 461400.0, "epoch": 2.5033112582781456, "step": 190}
{"eval_loss": 1.859080195426941, "eval_runtime": 8.9117, "eval_samples_per_second": 16.832, "eval_steps_per_second": 4.264, "eval_entropy": 0.06829966303944514, "eval_mean_token_accuracy": 0.7753759457876808, "eval_num_tokens": 461400.0, "epoch": 2.5033112582781456, "step": 190}
{"loss": 0.005886765569448471, "grad_norm": 0.007371403276920319, "learning_rate": 2.8292682926829267e-05, "entropy": 0.006164468544739066, "mean_token_accuracy": 0.9951255336403847, "num_tokens": 485443.0, "epoch": 2.6357615894039736, "step": 200}
{"eval_loss": 1.8880680799484253, "eval_runtime": 8.8924, "eval_samples_per_second": 16.868, "eval_steps_per_second": 4.273, "eval_entropy": 0.06679329828732175, "eval_mean_token_accuracy": 0.7753759457876808, "eval_num_tokens": 485443.0, "epoch": 2.6357615894039736, "step": 200}
{"loss": 0.0009710748679935932, "grad_norm": 0.0006506519857794046, "learning_rate": 1.853658536585366e-05, "entropy": 0.002073692464909982, "mean_token_accuracy": 1.0, "num_tokens": 509990.0, "epoch": 2.7682119205298013, "step": 210}
{"eval_loss": 1.9050503969192505, "eval_runtime": 8.8469, "eval_samples_per_second": 16.955, "eval_steps_per_second": 4.295, "eval_entropy": 0.06654788765100468, "eval_mean_token_accuracy": 0.7737312089455756, "eval_num_tokens": 509990.0, "epoch": 2.7682119205298013, "step": 210}
{"loss": 0.01160859763622284, "grad_norm": 0.004122971091419458, "learning_rate": 8.780487804878048e-06, "entropy": 0.00959118189784931, "mean_token_accuracy": 0.9908364906907081, "num_tokens": 534335.0, "epoch": 2.9006622516556293, "step": 220}
{"eval_loss": 1.9109818935394287, "eval_runtime": 8.8226, "eval_samples_per_second": 17.002, "eval_steps_per_second": 4.307, "eval_entropy": 0.06608304418458781, "eval_mean_token_accuracy": 0.7737312089455756, "eval_num_tokens": 534335.0, "epoch": 2.9006622516556293, "step": 220}
{"eval_loss": 1.9071061611175537, "eval_runtime": 8.9501, "eval_samples_per_second": 16.76, "eval_steps_per_second": 4.246, "eval_entropy": 0.06653332771269274, "eval_mean_token_accuracy": 0.7737312089455756, "eval_num_tokens": 552465.0, "epoch": 3.0, "step": 228}
{"train_runtime": 1102.1217, "train_samples_per_second": 3.283, "train_steps_per_second": 0.207, "total_flos": 3.178328110793933e+16, "train_loss": 0.2925478722695915, "epoch": 3.0, "step": 228}