{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.0067933301849093615, "eval_steps": 1, "global_step": 187, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 3.6327968903258616e-05, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 1 }, { "epoch": 3.6327968903258616e-05, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 507.0426, "eval_samples_per_second": 66.596, "eval_steps_per_second": 2.775, "step": 1 }, { "epoch": 7.265593780651723e-05, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8984, "step": 2 }, { "epoch": 7.265593780651723e-05, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 504.4591, "eval_samples_per_second": 66.937, "eval_steps_per_second": 2.789, "step": 2 }, { "epoch": 0.00010898390670977586, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8906, "step": 3 }, { "epoch": 0.00010898390670977586, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 506.5358, "eval_samples_per_second": 66.663, "eval_steps_per_second": 2.778, "step": 3 }, { "epoch": 0.00014531187561303446, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 4 }, { "epoch": 0.00014531187561303446, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 507.9364, "eval_samples_per_second": 66.479, "eval_steps_per_second": 2.77, "step": 4 }, { "epoch": 0.0001816398445162931, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 5 }, { "epoch": 0.0001816398445162931, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 504.3114, "eval_samples_per_second": 66.957, "eval_steps_per_second": 2.79, "step": 5 }, { "epoch": 0.00021796781341955172, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 6 }, { "epoch": 0.00021796781341955172, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 503.6918, "eval_samples_per_second": 67.039, "eval_steps_per_second": 2.793, "step": 6 }, { "epoch": 0.00025429578232281035, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8906, "step": 7 }, { "epoch": 0.00025429578232281035, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 503.9607, "eval_samples_per_second": 67.003, "eval_steps_per_second": 2.792, "step": 7 }, { "epoch": 0.0002906237512260689, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 8 }, { "epoch": 0.0002906237512260689, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 505.4484, "eval_samples_per_second": 66.806, "eval_steps_per_second": 2.784, "step": 8 }, { "epoch": 0.00032695172012932756, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.875, "step": 9 }, { "epoch": 0.00032695172012932756, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 503.3281, "eval_samples_per_second": 67.087, "eval_steps_per_second": 2.795, "step": 9 }, { "epoch": 0.0003632796890325862, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8984, "step": 10 }, { "epoch": 0.0003632796890325862, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 502.3143, "eval_samples_per_second": 67.223, "eval_steps_per_second": 2.801, "step": 10 }, { "epoch": 0.0003996076579358448, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 11 }, { "epoch": 0.0003996076579358448, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 504.0173, "eval_samples_per_second": 66.996, "eval_steps_per_second": 2.792, "step": 11 }, { "epoch": 0.00043593562683910344, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8906, "step": 12 }, { "epoch": 0.00043593562683910344, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 502.8909, "eval_samples_per_second": 67.146, "eval_steps_per_second": 2.798, "step": 12 }, { "epoch": 0.000472263595742362, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 13 }, { "epoch": 0.000472263595742362, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 505.6586, "eval_samples_per_second": 66.778, "eval_steps_per_second": 2.783, "step": 13 }, { "epoch": 0.0005085915646456207, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 14 }, { "epoch": 0.0005085915646456207, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 504.0652, "eval_samples_per_second": 66.989, "eval_steps_per_second": 2.791, "step": 14 }, { "epoch": 0.0005449195335488793, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 15 }, { "epoch": 0.0005449195335488793, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 502.8518, "eval_samples_per_second": 67.151, "eval_steps_per_second": 2.798, "step": 15 }, { "epoch": 0.0005812475024521379, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.8828, "step": 16 }, { "epoch": 0.0005812475024521379, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 503.4138, "eval_samples_per_second": 67.076, "eval_steps_per_second": 2.795, "step": 16 }, { "epoch": 0.0006175754713553965, "grad_norm": 0.0, "learning_rate": 1e-05, "loss": 10.875, "step": 17 }, { "epoch": 0.0006175754713553965, "eval_accuracy": 5.786882743483815e-05, "eval_loss": 10.8828125, "eval_runtime": 504.116, "eval_samples_per_second": 66.983, "eval_steps_per_second": 2.791, "step": 17 }, { "epoch": 0.0006539034402586551, "grad_norm": 9.14120864868164, "learning_rate": 9.999996367203111e-06, "loss": 10.8828, "step": 18 }, { "epoch": 0.0006539034402586551, "eval_accuracy": 0.019680148945503458, "eval_loss": 10.6328125, "eval_runtime": 503.1889, "eval_samples_per_second": 67.106, "eval_steps_per_second": 2.796, "step": 18 }, { "epoch": 0.0006902314091619138, "grad_norm": 7.371621131896973, "learning_rate": 9.99999273440622e-06, "loss": 10.6641, "step": 19 }, { "epoch": 0.0006902314091619138, "eval_accuracy": 0.04440203625321373, "eval_loss": 10.484375, "eval_runtime": 502.8755, "eval_samples_per_second": 67.148, "eval_steps_per_second": 2.798, "step": 19 }, { "epoch": 0.0007265593780651724, "grad_norm": 5.806190013885498, "learning_rate": 9.99998910160933e-06, "loss": 10.5078, "step": 20 }, { "epoch": 0.0007265593780651724, "eval_accuracy": 0.049879571177803755, "eval_loss": 10.3828125, "eval_runtime": 503.8208, "eval_samples_per_second": 67.022, "eval_steps_per_second": 2.793, "step": 20 }, { "epoch": 0.0007628873469684309, "grad_norm": 4.848043441772461, "learning_rate": 9.99998546881244e-06, "loss": 10.3984, "step": 21 }, { "epoch": 0.0007628873469684309, "eval_accuracy": 0.05315644057324473, "eval_loss": 10.3125, "eval_runtime": 505.0758, "eval_samples_per_second": 66.855, "eval_steps_per_second": 2.786, "step": 21 }, { "epoch": 0.0007992153158716896, "grad_norm": 3.881626844406128, "learning_rate": 9.99998183601555e-06, "loss": 10.3438, "step": 22 }, { "epoch": 0.0007992153158716896, "eval_accuracy": 0.05502066212418083, "eval_loss": 10.25, "eval_runtime": 508.1599, "eval_samples_per_second": 66.45, "eval_steps_per_second": 2.769, "step": 22 }, { "epoch": 0.0008355432847749482, "grad_norm": 3.3826823234558105, "learning_rate": 9.999978203218658e-06, "loss": 10.2656, "step": 23 }, { "epoch": 0.0008355432847749482, "eval_accuracy": 0.05621471691417821, "eval_loss": 10.203125, "eval_runtime": 508.8692, "eval_samples_per_second": 66.357, "eval_steps_per_second": 2.765, "step": 23 }, { "epoch": 0.0008718712536782069, "grad_norm": 3.2527709007263184, "learning_rate": 9.999974570421768e-06, "loss": 10.25, "step": 24 }, { "epoch": 0.0008718712536782069, "eval_accuracy": 0.05396498301959542, "eval_loss": 10.1640625, "eval_runtime": 504.6518, "eval_samples_per_second": 66.911, "eval_steps_per_second": 2.788, "step": 24 }, { "epoch": 0.0009081992225814655, "grad_norm": 3.0049052238464355, "learning_rate": 9.999970937624878e-06, "loss": 10.1875, "step": 25 }, { "epoch": 0.0009081992225814655, "eval_accuracy": 0.04704188536466089, "eval_loss": 10.1328125, "eval_runtime": 504.5459, "eval_samples_per_second": 66.926, "eval_steps_per_second": 2.789, "step": 25 }, { "epoch": 0.000944527191484724, "grad_norm": 2.725101947784424, "learning_rate": 9.999967304827988e-06, "loss": 10.125, "step": 26 }, { "epoch": 0.000944527191484724, "eval_accuracy": 0.04612232393221085, "eval_loss": 10.109375, "eval_runtime": 505.0881, "eval_samples_per_second": 66.854, "eval_steps_per_second": 2.786, "step": 26 }, { "epoch": 0.0009808551603879827, "grad_norm": 2.5348708629608154, "learning_rate": 9.999963672031098e-06, "loss": 10.125, "step": 27 }, { "epoch": 0.0009808551603879827, "eval_accuracy": 0.04800400166270834, "eval_loss": 10.0859375, "eval_runtime": 506.8917, "eval_samples_per_second": 66.616, "eval_steps_per_second": 2.776, "step": 27 }, { "epoch": 0.0010171831292912414, "grad_norm": 2.571255683898926, "learning_rate": 9.999960039234208e-06, "loss": 10.0938, "step": 28 }, { "epoch": 0.0010171831292912414, "eval_accuracy": 0.04744444860343471, "eval_loss": 10.0703125, "eval_runtime": 506.3522, "eval_samples_per_second": 66.687, "eval_steps_per_second": 2.779, "step": 28 }, { "epoch": 0.0010535110981944999, "grad_norm": 2.22050142288208, "learning_rate": 9.999956406437317e-06, "loss": 10.0625, "step": 29 }, { "epoch": 0.0010535110981944999, "eval_accuracy": 0.0465065914736666, "eval_loss": 10.0546875, "eval_runtime": 504.6118, "eval_samples_per_second": 66.917, "eval_steps_per_second": 2.788, "step": 29 }, { "epoch": 0.0010898390670977586, "grad_norm": 2.1588761806488037, "learning_rate": 9.999952773640427e-06, "loss": 10.0703, "step": 30 }, { "epoch": 0.0010898390670977586, "eval_accuracy": 0.04720460706501668, "eval_loss": 10.0390625, "eval_runtime": 502.6467, "eval_samples_per_second": 67.178, "eval_steps_per_second": 2.799, "step": 30 }, { "epoch": 0.0011261670360010172, "grad_norm": 2.2763571739196777, "learning_rate": 9.999949140843537e-06, "loss": 10.0156, "step": 31 }, { "epoch": 0.0011261670360010172, "eval_accuracy": 0.05153020783188431, "eval_loss": 10.0234375, "eval_runtime": 503.4047, "eval_samples_per_second": 67.077, "eval_steps_per_second": 2.795, "step": 31 }, { "epoch": 0.0011624950049042757, "grad_norm": 2.1108500957489014, "learning_rate": 9.999945508046645e-06, "loss": 10.0859, "step": 32 }, { "epoch": 0.0011624950049042757, "eval_accuracy": 0.05865791622834431, "eval_loss": 10.015625, "eval_runtime": 509.0053, "eval_samples_per_second": 66.339, "eval_steps_per_second": 2.764, "step": 32 }, { "epoch": 0.0011988229738075344, "grad_norm": 2.155444622039795, "learning_rate": 9.999941875249755e-06, "loss": 9.9922, "step": 33 }, { "epoch": 0.0011988229738075344, "eval_accuracy": 0.06134457569194863, "eval_loss": 10.0078125, "eval_runtime": 508.1742, "eval_samples_per_second": 66.448, "eval_steps_per_second": 2.769, "step": 33 }, { "epoch": 0.001235150942710793, "grad_norm": 2.6189229488372803, "learning_rate": 9.999938242452866e-06, "loss": 10.0234, "step": 34 }, { "epoch": 0.001235150942710793, "eval_accuracy": 0.06076637954869899, "eval_loss": 9.9921875, "eval_runtime": 506.0763, "eval_samples_per_second": 66.723, "eval_steps_per_second": 2.78, "step": 34 }, { "epoch": 0.0012714789116140518, "grad_norm": 2.2233242988586426, "learning_rate": 9.999934609655974e-06, "loss": 9.9609, "step": 35 }, { "epoch": 0.0012714789116140518, "eval_accuracy": 0.05996116622448688, "eval_loss": 9.984375, "eval_runtime": 505.1839, "eval_samples_per_second": 66.841, "eval_steps_per_second": 2.785, "step": 35 }, { "epoch": 0.0013078068805173102, "grad_norm": 2.289062261581421, "learning_rate": 9.999930976859084e-06, "loss": 10.0391, "step": 36 }, { "epoch": 0.0013078068805173102, "eval_accuracy": 0.06078849649925438, "eval_loss": 9.9765625, "eval_runtime": 505.7312, "eval_samples_per_second": 66.769, "eval_steps_per_second": 2.782, "step": 36 }, { "epoch": 0.001344134849420569, "grad_norm": 2.283668041229248, "learning_rate": 9.999927344062194e-06, "loss": 9.9922, "step": 37 }, { "epoch": 0.001344134849420569, "eval_accuracy": 0.061857115756847984, "eval_loss": 9.9609375, "eval_runtime": 508.2946, "eval_samples_per_second": 66.432, "eval_steps_per_second": 2.768, "step": 37 }, { "epoch": 0.0013804628183238276, "grad_norm": 2.189363718032837, "learning_rate": 9.999923711265304e-06, "loss": 9.9688, "step": 38 }, { "epoch": 0.0013804628183238276, "eval_accuracy": 0.06226219754889185, "eval_loss": 9.953125, "eval_runtime": 508.653, "eval_samples_per_second": 66.385, "eval_steps_per_second": 2.766, "step": 38 }, { "epoch": 0.001416790787227086, "grad_norm": 2.082035541534424, "learning_rate": 9.999920078468414e-06, "loss": 9.9453, "step": 39 }, { "epoch": 0.001416790787227086, "eval_accuracy": 0.06217419292888089, "eval_loss": 9.9375, "eval_runtime": 507.0173, "eval_samples_per_second": 66.599, "eval_steps_per_second": 2.775, "step": 39 }, { "epoch": 0.0014531187561303447, "grad_norm": 2.1098358631134033, "learning_rate": 9.999916445671524e-06, "loss": 9.9609, "step": 40 }, { "epoch": 0.0014531187561303447, "eval_accuracy": 0.06281911625934278, "eval_loss": 9.9296875, "eval_runtime": 505.2162, "eval_samples_per_second": 66.837, "eval_steps_per_second": 2.785, "step": 40 }, { "epoch": 0.0014894467250336034, "grad_norm": 1.99773108959198, "learning_rate": 9.999912812874633e-06, "loss": 9.9609, "step": 41 }, { "epoch": 0.0014894467250336034, "eval_accuracy": 0.06401496588040619, "eval_loss": 9.9140625, "eval_runtime": 505.419, "eval_samples_per_second": 66.81, "eval_steps_per_second": 2.784, "step": 41 }, { "epoch": 0.0015257746939368619, "grad_norm": 2.0331761837005615, "learning_rate": 9.999909180077743e-06, "loss": 10.0234, "step": 42 }, { "epoch": 0.0015257746939368619, "eval_accuracy": 0.06490028077816117, "eval_loss": 9.8984375, "eval_runtime": 508.1771, "eval_samples_per_second": 66.447, "eval_steps_per_second": 2.769, "step": 42 }, { "epoch": 0.0015621026628401206, "grad_norm": 2.115607500076294, "learning_rate": 9.999905547280853e-06, "loss": 9.9375, "step": 43 }, { "epoch": 0.0015621026628401206, "eval_accuracy": 0.06480990234932096, "eval_loss": 9.890625, "eval_runtime": 509.849, "eval_samples_per_second": 66.229, "eval_steps_per_second": 2.76, "step": 43 }, { "epoch": 0.0015984306317433793, "grad_norm": 2.163503646850586, "learning_rate": 9.999901914483961e-06, "loss": 9.8516, "step": 44 }, { "epoch": 0.0015984306317433793, "eval_accuracy": 0.06442907972555643, "eval_loss": 9.875, "eval_runtime": 509.7023, "eval_samples_per_second": 66.248, "eval_steps_per_second": 2.76, "step": 44 }, { "epoch": 0.001634758600646638, "grad_norm": 2.0586893558502197, "learning_rate": 9.999898281687071e-06, "loss": 9.8672, "step": 45 }, { "epoch": 0.001634758600646638, "eval_accuracy": 0.06429814390440197, "eval_loss": 9.859375, "eval_runtime": 506.8968, "eval_samples_per_second": 66.615, "eval_steps_per_second": 2.776, "step": 45 }, { "epoch": 0.0016710865695498964, "grad_norm": 1.898136854171753, "learning_rate": 9.999894648890181e-06, "loss": 9.8984, "step": 46 }, { "epoch": 0.0016710865695498964, "eval_accuracy": 0.06426870288514172, "eval_loss": 9.84375, "eval_runtime": 505.4907, "eval_samples_per_second": 66.8, "eval_steps_per_second": 2.783, "step": 46 }, { "epoch": 0.001707414538453155, "grad_norm": 1.9898130893707275, "learning_rate": 9.99989101609329e-06, "loss": 9.875, "step": 47 }, { "epoch": 0.001707414538453155, "eval_accuracy": 0.06445736278929022, "eval_loss": 9.8359375, "eval_runtime": 505.8927, "eval_samples_per_second": 66.747, "eval_steps_per_second": 2.781, "step": 47 }, { "epoch": 0.0017437425073564138, "grad_norm": 1.9185965061187744, "learning_rate": 9.9998873832964e-06, "loss": 9.8672, "step": 48 }, { "epoch": 0.0017437425073564138, "eval_accuracy": 0.06462755330279167, "eval_loss": 9.8203125, "eval_runtime": 507.6208, "eval_samples_per_second": 66.52, "eval_steps_per_second": 2.772, "step": 48 }, { "epoch": 0.0017800704762596722, "grad_norm": 1.8137012720108032, "learning_rate": 9.99988375049951e-06, "loss": 9.8984, "step": 49 }, { "epoch": 0.0017800704762596722, "eval_accuracy": 0.06494483311704172, "eval_loss": 9.8125, "eval_runtime": 506.9398, "eval_samples_per_second": 66.609, "eval_steps_per_second": 2.775, "step": 49 }, { "epoch": 0.001816398445162931, "grad_norm": 2.237258195877075, "learning_rate": 9.99988011770262e-06, "loss": 9.7891, "step": 50 }, { "epoch": 0.001816398445162931, "eval_accuracy": 0.06534296717592682, "eval_loss": 9.8046875, "eval_runtime": 504.003, "eval_samples_per_second": 66.998, "eval_steps_per_second": 2.792, "step": 50 }, { "epoch": 0.0018527264140661896, "grad_norm": 2.279499053955078, "learning_rate": 9.99987648490573e-06, "loss": 9.8281, "step": 51 }, { "epoch": 0.0018527264140661896, "eval_accuracy": 0.06550117284972942, "eval_loss": 9.7890625, "eval_runtime": 503.0349, "eval_samples_per_second": 67.127, "eval_steps_per_second": 2.797, "step": 51 }, { "epoch": 0.001889054382969448, "grad_norm": 1.8227834701538086, "learning_rate": 9.99987285210884e-06, "loss": 9.8281, "step": 52 }, { "epoch": 0.001889054382969448, "eval_accuracy": 0.0654096075164746, "eval_loss": 9.78125, "eval_runtime": 505.4701, "eval_samples_per_second": 66.803, "eval_steps_per_second": 2.784, "step": 52 }, { "epoch": 0.0019253823518727068, "grad_norm": 2.207292079925537, "learning_rate": 9.99986921931195e-06, "loss": 9.7969, "step": 53 }, { "epoch": 0.0019253823518727068, "eval_accuracy": 0.06596467349808319, "eval_loss": 9.7734375, "eval_runtime": 506.8878, "eval_samples_per_second": 66.616, "eval_steps_per_second": 2.776, "step": 53 }, { "epoch": 0.0019617103207759654, "grad_norm": 1.9478576183319092, "learning_rate": 9.999865586515059e-06, "loss": 9.7812, "step": 54 }, { "epoch": 0.0019617103207759654, "eval_accuracy": 0.06697281852830743, "eval_loss": 9.765625, "eval_runtime": 506.2987, "eval_samples_per_second": 66.694, "eval_steps_per_second": 2.779, "step": 54 }, { "epoch": 0.001998038289679224, "grad_norm": 2.4436683654785156, "learning_rate": 9.999861953718169e-06, "loss": 9.8047, "step": 55 }, { "epoch": 0.001998038289679224, "eval_accuracy": 0.06823391894328684, "eval_loss": 9.75, "eval_runtime": 505.1115, "eval_samples_per_second": 66.851, "eval_steps_per_second": 2.786, "step": 55 }, { "epoch": 0.002034366258582483, "grad_norm": 1.911065936088562, "learning_rate": 9.999858320921279e-06, "loss": 9.7969, "step": 56 }, { "epoch": 0.002034366258582483, "eval_accuracy": 0.0688041251934039, "eval_loss": 9.7421875, "eval_runtime": 503.4384, "eval_samples_per_second": 67.073, "eval_steps_per_second": 2.795, "step": 56 }, { "epoch": 0.0020706942274857413, "grad_norm": 2.6223678588867188, "learning_rate": 9.999854688124387e-06, "loss": 9.7891, "step": 57 }, { "epoch": 0.0020706942274857413, "eval_accuracy": 0.0691195522788116, "eval_loss": 9.734375, "eval_runtime": 504.3798, "eval_samples_per_second": 66.948, "eval_steps_per_second": 2.79, "step": 57 }, { "epoch": 0.0021070221963889997, "grad_norm": 2.182452917098999, "learning_rate": 9.999851055327497e-06, "loss": 9.6875, "step": 58 }, { "epoch": 0.0021070221963889997, "eval_accuracy": 0.0690351083720445, "eval_loss": 9.7265625, "eval_runtime": 507.3727, "eval_samples_per_second": 66.553, "eval_steps_per_second": 2.773, "step": 58 }, { "epoch": 0.0021433501652922586, "grad_norm": 1.8129832744598389, "learning_rate": 9.999847422530608e-06, "loss": 9.7188, "step": 59 }, { "epoch": 0.0021433501652922586, "eval_accuracy": 0.06855982552620901, "eval_loss": 9.71875, "eval_runtime": 504.2068, "eval_samples_per_second": 66.971, "eval_steps_per_second": 2.791, "step": 59 }, { "epoch": 0.002179678134195517, "grad_norm": 1.9449687004089355, "learning_rate": 9.999843789733716e-06, "loss": 9.7344, "step": 60 }, { "epoch": 0.002179678134195517, "eval_accuracy": 0.06816276257618588, "eval_loss": 9.7109375, "eval_runtime": 506.178, "eval_samples_per_second": 66.71, "eval_steps_per_second": 2.78, "step": 60 }, { "epoch": 0.0022160061030987756, "grad_norm": 2.0629429817199707, "learning_rate": 9.999840156936826e-06, "loss": 9.7344, "step": 61 }, { "epoch": 0.0022160061030987756, "eval_accuracy": 0.06871380466234002, "eval_loss": 9.6953125, "eval_runtime": 504.7991, "eval_samples_per_second": 66.892, "eval_steps_per_second": 2.787, "step": 61 }, { "epoch": 0.0022523340720020345, "grad_norm": 1.7803966999053955, "learning_rate": 9.999836524139936e-06, "loss": 9.7578, "step": 62 }, { "epoch": 0.0022523340720020345, "eval_accuracy": 0.06968408454684902, "eval_loss": 9.6875, "eval_runtime": 503.513, "eval_samples_per_second": 67.063, "eval_steps_per_second": 2.794, "step": 62 }, { "epoch": 0.002288662040905293, "grad_norm": 2.0867655277252197, "learning_rate": 9.999832891343046e-06, "loss": 9.6484, "step": 63 }, { "epoch": 0.002288662040905293, "eval_accuracy": 0.07077357595280706, "eval_loss": 9.671875, "eval_runtime": 505.2322, "eval_samples_per_second": 66.835, "eval_steps_per_second": 2.785, "step": 63 }, { "epoch": 0.0023249900098085514, "grad_norm": 2.1032333374023438, "learning_rate": 9.999829258546155e-06, "loss": 9.6328, "step": 64 }, { "epoch": 0.0023249900098085514, "eval_accuracy": 0.07146232789994546, "eval_loss": 9.6640625, "eval_runtime": 503.6458, "eval_samples_per_second": 67.045, "eval_steps_per_second": 2.794, "step": 64 }, { "epoch": 0.0023613179787118103, "grad_norm": 1.6707836389541626, "learning_rate": 9.999825625749265e-06, "loss": 9.7656, "step": 65 }, { "epoch": 0.0023613179787118103, "eval_accuracy": 0.07207561009564684, "eval_loss": 9.65625, "eval_runtime": 503.2725, "eval_samples_per_second": 67.095, "eval_steps_per_second": 2.796, "step": 65 }, { "epoch": 0.0023976459476150688, "grad_norm": 1.7805981636047363, "learning_rate": 9.999821992952375e-06, "loss": 9.6875, "step": 66 }, { "epoch": 0.0023976459476150688, "eval_accuracy": 0.0724615277237278, "eval_loss": 9.6484375, "eval_runtime": 502.4455, "eval_samples_per_second": 67.205, "eval_steps_per_second": 2.8, "step": 66 }, { "epoch": 0.0024339739165183277, "grad_norm": 2.426743745803833, "learning_rate": 9.999818360155485e-06, "loss": 9.6328, "step": 67 }, { "epoch": 0.0024339739165183277, "eval_accuracy": 0.07273411045465647, "eval_loss": 9.640625, "eval_runtime": 502.4056, "eval_samples_per_second": 67.211, "eval_steps_per_second": 2.801, "step": 67 }, { "epoch": 0.002470301885421586, "grad_norm": 1.7032270431518555, "learning_rate": 9.999814727358595e-06, "loss": 9.6953, "step": 68 }, { "epoch": 0.002470301885421586, "eval_accuracy": 0.07335859587007634, "eval_loss": 9.6328125, "eval_runtime": 505.6904, "eval_samples_per_second": 66.774, "eval_steps_per_second": 2.782, "step": 68 }, { "epoch": 0.0025066298543248446, "grad_norm": 1.7397522926330566, "learning_rate": 9.999811094561705e-06, "loss": 9.7188, "step": 69 }, { "epoch": 0.0025066298543248446, "eval_accuracy": 0.07437455710010418, "eval_loss": 9.625, "eval_runtime": 505.9561, "eval_samples_per_second": 66.739, "eval_steps_per_second": 2.781, "step": 69 }, { "epoch": 0.0025429578232281035, "grad_norm": 1.6626527309417725, "learning_rate": 9.999807461764813e-06, "loss": 9.6875, "step": 70 }, { "epoch": 0.0025429578232281035, "eval_accuracy": 0.075310387807701, "eval_loss": 9.6171875, "eval_runtime": 505.7964, "eval_samples_per_second": 66.76, "eval_steps_per_second": 2.782, "step": 70 }, { "epoch": 0.002579285792131362, "grad_norm": 2.158742666244507, "learning_rate": 9.999803828967923e-06, "loss": 9.625, "step": 71 }, { "epoch": 0.002579285792131362, "eval_accuracy": 0.07634256041509926, "eval_loss": 9.609375, "eval_runtime": 504.0869, "eval_samples_per_second": 66.986, "eval_steps_per_second": 2.791, "step": 71 }, { "epoch": 0.0026156137610346204, "grad_norm": 1.8259432315826416, "learning_rate": 9.999800196171034e-06, "loss": 9.6172, "step": 72 }, { "epoch": 0.0026156137610346204, "eval_accuracy": 0.07690020284775424, "eval_loss": 9.6015625, "eval_runtime": 504.7109, "eval_samples_per_second": 66.904, "eval_steps_per_second": 2.788, "step": 72 }, { "epoch": 0.0026519417299378793, "grad_norm": 1.7585749626159668, "learning_rate": 9.999796563374142e-06, "loss": 9.6016, "step": 73 }, { "epoch": 0.0026519417299378793, "eval_accuracy": 0.07706367721920222, "eval_loss": 9.59375, "eval_runtime": 505.9779, "eval_samples_per_second": 66.736, "eval_steps_per_second": 2.781, "step": 73 }, { "epoch": 0.002688269698841138, "grad_norm": 1.8018831014633179, "learning_rate": 9.999792930577252e-06, "loss": 9.6094, "step": 74 }, { "epoch": 0.002688269698841138, "eval_accuracy": 0.07714539993048214, "eval_loss": 9.5859375, "eval_runtime": 505.2425, "eval_samples_per_second": 66.833, "eval_steps_per_second": 2.785, "step": 74 }, { "epoch": 0.0027245976677443963, "grad_norm": 1.7233467102050781, "learning_rate": 9.999789297780362e-06, "loss": 9.5859, "step": 75 }, { "epoch": 0.0027245976677443963, "eval_accuracy": 0.07710901117806314, "eval_loss": 9.578125, "eval_runtime": 505.5998, "eval_samples_per_second": 66.786, "eval_steps_per_second": 2.783, "step": 75 }, { "epoch": 0.002760925636647655, "grad_norm": 1.9081038236618042, "learning_rate": 9.99978566498347e-06, "loss": 9.5859, "step": 76 }, { "epoch": 0.002760925636647655, "eval_accuracy": 0.07673672847630625, "eval_loss": 9.5703125, "eval_runtime": 503.3457, "eval_samples_per_second": 67.085, "eval_steps_per_second": 2.795, "step": 76 }, { "epoch": 0.0027972536055509136, "grad_norm": 1.7332507371902466, "learning_rate": 9.99978203218658e-06, "loss": 9.5859, "step": 77 }, { "epoch": 0.0027972536055509136, "eval_accuracy": 0.07650699009985658, "eval_loss": 9.5625, "eval_runtime": 502.974, "eval_samples_per_second": 67.135, "eval_steps_per_second": 2.797, "step": 77 }, { "epoch": 0.002833581574454172, "grad_norm": 1.8088840246200562, "learning_rate": 9.99977839938969e-06, "loss": 9.5781, "step": 78 }, { "epoch": 0.002833581574454172, "eval_accuracy": 0.07642793068628753, "eval_loss": 9.5546875, "eval_runtime": 505.8316, "eval_samples_per_second": 66.755, "eval_steps_per_second": 2.782, "step": 78 }, { "epoch": 0.002869909543357431, "grad_norm": 1.6268699169158936, "learning_rate": 9.9997747665928e-06, "loss": 9.6172, "step": 79 }, { "epoch": 0.002869909543357431, "eval_accuracy": 0.07634560004835622, "eval_loss": 9.546875, "eval_runtime": 505.9115, "eval_samples_per_second": 66.745, "eval_steps_per_second": 2.781, "step": 79 }, { "epoch": 0.0029062375122606895, "grad_norm": 1.8675007820129395, "learning_rate": 9.999771133795911e-06, "loss": 9.5859, "step": 80 }, { "epoch": 0.0029062375122606895, "eval_accuracy": 0.07677149609098821, "eval_loss": 9.5390625, "eval_runtime": 505.7348, "eval_samples_per_second": 66.768, "eval_steps_per_second": 2.782, "step": 80 }, { "epoch": 0.002942565481163948, "grad_norm": 1.9098700284957886, "learning_rate": 9.999767500999021e-06, "loss": 9.5859, "step": 81 }, { "epoch": 0.002942565481163948, "eval_accuracy": 0.07698542837450169, "eval_loss": 9.53125, "eval_runtime": 502.794, "eval_samples_per_second": 67.159, "eval_steps_per_second": 2.798, "step": 81 }, { "epoch": 0.002978893450067207, "grad_norm": 1.8791757822036743, "learning_rate": 9.99976386820213e-06, "loss": 9.5391, "step": 82 }, { "epoch": 0.002978893450067207, "eval_accuracy": 0.07695144237980009, "eval_loss": 9.5234375, "eval_runtime": 503.0464, "eval_samples_per_second": 67.125, "eval_steps_per_second": 2.797, "step": 82 }, { "epoch": 0.0030152214189704653, "grad_norm": 1.704999566078186, "learning_rate": 9.99976023540524e-06, "loss": 9.5391, "step": 83 }, { "epoch": 0.0030152214189704653, "eval_accuracy": 0.0764418551014932, "eval_loss": 9.5234375, "eval_runtime": 505.8882, "eval_samples_per_second": 66.748, "eval_steps_per_second": 2.781, "step": 83 }, { "epoch": 0.0030515493878737238, "grad_norm": 1.6351501941680908, "learning_rate": 9.99975660260835e-06, "loss": 9.5312, "step": 84 }, { "epoch": 0.0030515493878737238, "eval_accuracy": 0.07578613383574707, "eval_loss": 9.515625, "eval_runtime": 506.1511, "eval_samples_per_second": 66.713, "eval_steps_per_second": 2.78, "step": 84 }, { "epoch": 0.0030878773567769827, "grad_norm": 1.8447389602661133, "learning_rate": 9.999752969811458e-06, "loss": 9.5547, "step": 85 }, { "epoch": 0.0030878773567769827, "eval_accuracy": 0.07570536643777649, "eval_loss": 9.5078125, "eval_runtime": 504.3987, "eval_samples_per_second": 66.945, "eval_steps_per_second": 2.789, "step": 85 }, { "epoch": 0.003124205325680241, "grad_norm": 1.743525743484497, "learning_rate": 9.999749337014568e-06, "loss": 9.5781, "step": 86 }, { "epoch": 0.003124205325680241, "eval_accuracy": 0.07604039192046953, "eval_loss": 9.5, "eval_runtime": 504.8644, "eval_samples_per_second": 66.883, "eval_steps_per_second": 2.787, "step": 86 }, { "epoch": 0.0031605332945835, "grad_norm": 1.7386131286621094, "learning_rate": 9.999745704217678e-06, "loss": 9.5703, "step": 87 }, { "epoch": 0.0031605332945835, "eval_accuracy": 0.0763665879922733, "eval_loss": 9.4921875, "eval_runtime": 503.9366, "eval_samples_per_second": 67.006, "eval_steps_per_second": 2.792, "step": 87 }, { "epoch": 0.0031968612634867585, "grad_norm": 1.8193249702453613, "learning_rate": 9.999742071420788e-06, "loss": 9.4844, "step": 88 }, { "epoch": 0.0031968612634867585, "eval_accuracy": 0.0764146431466214, "eval_loss": 9.484375, "eval_runtime": 504.69, "eval_samples_per_second": 66.906, "eval_steps_per_second": 2.788, "step": 88 }, { "epoch": 0.003233189232390017, "grad_norm": 1.7229965925216675, "learning_rate": 9.999738438623897e-06, "loss": 9.5312, "step": 89 }, { "epoch": 0.003233189232390017, "eval_accuracy": 0.07648345465378129, "eval_loss": 9.4765625, "eval_runtime": 504.1538, "eval_samples_per_second": 66.978, "eval_steps_per_second": 2.791, "step": 89 }, { "epoch": 0.003269517201293276, "grad_norm": 1.7293239831924438, "learning_rate": 9.999734805827007e-06, "loss": 9.5312, "step": 90 }, { "epoch": 0.003269517201293276, "eval_accuracy": 0.07651958286620683, "eval_loss": 9.46875, "eval_runtime": 504.9175, "eval_samples_per_second": 66.876, "eval_steps_per_second": 2.787, "step": 90 }, { "epoch": 0.0033058451701965343, "grad_norm": 1.6052281856536865, "learning_rate": 9.999731173030117e-06, "loss": 9.5078, "step": 91 }, { "epoch": 0.0033058451701965343, "eval_accuracy": 0.07662689639462152, "eval_loss": 9.46875, "eval_runtime": 505.1868, "eval_samples_per_second": 66.841, "eval_steps_per_second": 2.785, "step": 91 }, { "epoch": 0.003342173139099793, "grad_norm": 1.6207119226455688, "learning_rate": 9.999727540233227e-06, "loss": 9.5, "step": 92 }, { "epoch": 0.003342173139099793, "eval_accuracy": 0.07678099132630518, "eval_loss": 9.4609375, "eval_runtime": 503.2555, "eval_samples_per_second": 67.097, "eval_steps_per_second": 2.796, "step": 92 }, { "epoch": 0.0033785011080030517, "grad_norm": 1.7734899520874023, "learning_rate": 9.999723907436337e-06, "loss": 9.4844, "step": 93 }, { "epoch": 0.0033785011080030517, "eval_accuracy": 0.07692313036717814, "eval_loss": 9.453125, "eval_runtime": 505.6253, "eval_samples_per_second": 66.783, "eval_steps_per_second": 2.783, "step": 93 }, { "epoch": 0.00341482907690631, "grad_norm": 1.8365049362182617, "learning_rate": 9.999720274639447e-06, "loss": 9.4688, "step": 94 }, { "epoch": 0.00341482907690631, "eval_accuracy": 0.0772882916424473, "eval_loss": 9.4453125, "eval_runtime": 504.8802, "eval_samples_per_second": 66.881, "eval_steps_per_second": 2.787, "step": 94 }, { "epoch": 0.0034511570458095686, "grad_norm": 1.5928810834884644, "learning_rate": 9.999716641842555e-06, "loss": 9.5156, "step": 95 }, { "epoch": 0.0034511570458095686, "eval_accuracy": 0.07773769418226642, "eval_loss": 9.4375, "eval_runtime": 508.082, "eval_samples_per_second": 66.46, "eval_steps_per_second": 2.769, "step": 95 }, { "epoch": 0.0034874850147128275, "grad_norm": 1.648245096206665, "learning_rate": 9.999713009045665e-06, "loss": 9.4453, "step": 96 }, { "epoch": 0.0034874850147128275, "eval_accuracy": 0.07832689090301743, "eval_loss": 9.4296875, "eval_runtime": 504.0456, "eval_samples_per_second": 66.992, "eval_steps_per_second": 2.791, "step": 96 }, { "epoch": 0.003523812983616086, "grad_norm": 1.7062604427337646, "learning_rate": 9.999709376248776e-06, "loss": 9.4766, "step": 97 }, { "epoch": 0.003523812983616086, "eval_accuracy": 0.07944866031927555, "eval_loss": 9.421875, "eval_runtime": 504.2362, "eval_samples_per_second": 66.967, "eval_steps_per_second": 2.79, "step": 97 }, { "epoch": 0.0035601409525193445, "grad_norm": 1.7384364604949951, "learning_rate": 9.999705743451884e-06, "loss": 9.4219, "step": 98 }, { "epoch": 0.0035601409525193445, "eval_accuracy": 0.08043025921905568, "eval_loss": 9.421875, "eval_runtime": 505.9376, "eval_samples_per_second": 66.741, "eval_steps_per_second": 2.781, "step": 98 }, { "epoch": 0.0035964689214226034, "grad_norm": 1.6508241891860962, "learning_rate": 9.999702110654994e-06, "loss": 9.4531, "step": 99 }, { "epoch": 0.0035964689214226034, "eval_accuracy": 0.08138432772619424, "eval_loss": 9.4140625, "eval_runtime": 506.1211, "eval_samples_per_second": 66.717, "eval_steps_per_second": 2.78, "step": 99 }, { "epoch": 0.003632796890325862, "grad_norm": 1.690098762512207, "learning_rate": 9.999698477858104e-06, "loss": 9.4141, "step": 100 }, { "epoch": 0.003632796890325862, "eval_accuracy": 0.08187622723383445, "eval_loss": 9.40625, "eval_runtime": 505.2427, "eval_samples_per_second": 66.833, "eval_steps_per_second": 2.785, "step": 100 }, { "epoch": 0.0036691248592291203, "grad_norm": 1.738881230354309, "learning_rate": 9.999694845061213e-06, "loss": 9.375, "step": 101 }, { "epoch": 0.0036691248592291203, "eval_accuracy": 0.08247908782979768, "eval_loss": 9.3984375, "eval_runtime": 504.9005, "eval_samples_per_second": 66.879, "eval_steps_per_second": 2.787, "step": 101 }, { "epoch": 0.003705452828132379, "grad_norm": 1.5808320045471191, "learning_rate": 9.999691212264323e-06, "loss": 9.4219, "step": 102 }, { "epoch": 0.003705452828132379, "eval_accuracy": 0.08282250849005755, "eval_loss": 9.390625, "eval_runtime": 502.7993, "eval_samples_per_second": 67.158, "eval_steps_per_second": 2.798, "step": 102 }, { "epoch": 0.0037417807970356377, "grad_norm": 1.6801838874816895, "learning_rate": 9.999687579467433e-06, "loss": 9.3828, "step": 103 }, { "epoch": 0.0037417807970356377, "eval_accuracy": 0.08283837248077006, "eval_loss": 9.3828125, "eval_runtime": 505.0748, "eval_samples_per_second": 66.855, "eval_steps_per_second": 2.786, "step": 103 }, { "epoch": 0.003778108765938896, "grad_norm": 1.6030634641647339, "learning_rate": 9.999683946670543e-06, "loss": 9.375, "step": 104 }, { "epoch": 0.003778108765938896, "eval_accuracy": 0.08272639818136136, "eval_loss": 9.3828125, "eval_runtime": 505.0381, "eval_samples_per_second": 66.86, "eval_steps_per_second": 2.786, "step": 104 }, { "epoch": 0.003814436734842155, "grad_norm": 2.755161762237549, "learning_rate": 9.999680313873651e-06, "loss": 9.3516, "step": 105 }, { "epoch": 0.003814436734842155, "eval_accuracy": 0.08249729668045126, "eval_loss": 9.375, "eval_runtime": 506.1198, "eval_samples_per_second": 66.717, "eval_steps_per_second": 2.78, "step": 105 }, { "epoch": 0.0038507647037454135, "grad_norm": 1.700968861579895, "learning_rate": 9.999676681076761e-06, "loss": 9.3906, "step": 106 }, { "epoch": 0.0038507647037454135, "eval_accuracy": 0.0824550023548473, "eval_loss": 9.3671875, "eval_runtime": 504.8911, "eval_samples_per_second": 66.88, "eval_steps_per_second": 2.787, "step": 106 }, { "epoch": 0.0038870926726486724, "grad_norm": 1.7681119441986084, "learning_rate": 9.999673048279871e-06, "loss": 9.3672, "step": 107 }, { "epoch": 0.0038870926726486724, "eval_accuracy": 0.08233683299337206, "eval_loss": 9.359375, "eval_runtime": 503.3036, "eval_samples_per_second": 67.091, "eval_steps_per_second": 2.796, "step": 107 }, { "epoch": 0.003923420641551931, "grad_norm": 1.7191691398620605, "learning_rate": 9.999669415482981e-06, "loss": 9.3359, "step": 108 }, { "epoch": 0.003923420641551931, "eval_accuracy": 0.08216516608657437, "eval_loss": 9.3515625, "eval_runtime": 504.9663, "eval_samples_per_second": 66.87, "eval_steps_per_second": 2.786, "step": 108 }, { "epoch": 0.00395974861045519, "grad_norm": 1.5659282207489014, "learning_rate": 9.999665782686092e-06, "loss": 9.4062, "step": 109 }, { "epoch": 0.00395974861045519, "eval_accuracy": 0.08175872369678691, "eval_loss": 9.34375, "eval_runtime": 505.2603, "eval_samples_per_second": 66.831, "eval_steps_per_second": 2.785, "step": 109 }, { "epoch": 0.003996076579358448, "grad_norm": 1.8398549556732178, "learning_rate": 9.999662149889202e-06, "loss": 9.3906, "step": 110 }, { "epoch": 0.003996076579358448, "eval_accuracy": 0.08157759050356041, "eval_loss": 9.34375, "eval_runtime": 506.1037, "eval_samples_per_second": 66.72, "eval_steps_per_second": 2.78, "step": 110 }, { "epoch": 0.004032404548261707, "grad_norm": 2.4529032707214355, "learning_rate": 9.99965851709231e-06, "loss": 9.25, "step": 111 }, { "epoch": 0.004032404548261707, "eval_accuracy": 0.08160373134957025, "eval_loss": 9.3359375, "eval_runtime": 504.6658, "eval_samples_per_second": 66.91, "eval_steps_per_second": 2.788, "step": 111 }, { "epoch": 0.004068732517164966, "grad_norm": 1.7243255376815796, "learning_rate": 9.99965488429542e-06, "loss": 9.3281, "step": 112 }, { "epoch": 0.004068732517164966, "eval_accuracy": 0.08156398452612451, "eval_loss": 9.328125, "eval_runtime": 501.5913, "eval_samples_per_second": 67.32, "eval_steps_per_second": 2.805, "step": 112 }, { "epoch": 0.004105060486068224, "grad_norm": 1.617632508277893, "learning_rate": 9.99965125149853e-06, "loss": 9.375, "step": 113 }, { "epoch": 0.004105060486068224, "eval_accuracy": 0.08125842901157987, "eval_loss": 9.3203125, "eval_runtime": 504.3056, "eval_samples_per_second": 66.957, "eval_steps_per_second": 2.79, "step": 113 }, { "epoch": 0.0041413884549714826, "grad_norm": 1.7473822832107544, "learning_rate": 9.999647618701639e-06, "loss": 9.3906, "step": 114 }, { "epoch": 0.0041413884549714826, "eval_accuracy": 0.0812115607616464, "eval_loss": 9.3203125, "eval_runtime": 507.2626, "eval_samples_per_second": 66.567, "eval_steps_per_second": 2.774, "step": 114 }, { "epoch": 0.0041777164238747415, "grad_norm": 1.7480499744415283, "learning_rate": 9.999643985904749e-06, "loss": 9.3203, "step": 115 }, { "epoch": 0.0041777164238747415, "eval_accuracy": 0.08121138706831743, "eval_loss": 9.3125, "eval_runtime": 505.5579, "eval_samples_per_second": 66.792, "eval_steps_per_second": 2.783, "step": 115 }, { "epoch": 0.0042140443927779995, "grad_norm": 1.645756721496582, "learning_rate": 9.999640353107859e-06, "loss": 9.3125, "step": 116 }, { "epoch": 0.0042140443927779995, "eval_accuracy": 0.08107075336962886, "eval_loss": 9.3046875, "eval_runtime": 504.2295, "eval_samples_per_second": 66.968, "eval_steps_per_second": 2.79, "step": 116 }, { "epoch": 0.004250372361681258, "grad_norm": 1.6185376644134521, "learning_rate": 9.999636720310967e-06, "loss": 9.3359, "step": 117 }, { "epoch": 0.004250372361681258, "eval_accuracy": 0.08092007440674827, "eval_loss": 9.296875, "eval_runtime": 502.016, "eval_samples_per_second": 67.263, "eval_steps_per_second": 2.803, "step": 117 }, { "epoch": 0.004286700330584517, "grad_norm": 1.718162178993225, "learning_rate": 9.999633087514077e-06, "loss": 9.2812, "step": 118 }, { "epoch": 0.004286700330584517, "eval_accuracy": 0.08078224875021137, "eval_loss": 9.296875, "eval_runtime": 503.8149, "eval_samples_per_second": 67.023, "eval_steps_per_second": 2.793, "step": 118 }, { "epoch": 0.004323028299487775, "grad_norm": 2.7064478397369385, "learning_rate": 9.999629454717187e-06, "loss": 9.2031, "step": 119 }, { "epoch": 0.004323028299487775, "eval_accuracy": 0.08067389306182286, "eval_loss": 9.2890625, "eval_runtime": 506.662, "eval_samples_per_second": 66.646, "eval_steps_per_second": 2.777, "step": 119 }, { "epoch": 0.004359356268391034, "grad_norm": 1.9571579694747925, "learning_rate": 9.999625821920297e-06, "loss": 9.2422, "step": 120 }, { "epoch": 0.004359356268391034, "eval_accuracy": 0.08080248402303625, "eval_loss": 9.28125, "eval_runtime": 506.1232, "eval_samples_per_second": 66.717, "eval_steps_per_second": 2.78, "step": 120 }, { "epoch": 0.004395684237294293, "grad_norm": 1.6557737588882446, "learning_rate": 9.999622189123407e-06, "loss": 9.3047, "step": 121 }, { "epoch": 0.004395684237294293, "eval_accuracy": 0.08088458306986226, "eval_loss": 9.28125, "eval_runtime": 503.855, "eval_samples_per_second": 67.017, "eval_steps_per_second": 2.792, "step": 121 }, { "epoch": 0.004432012206197551, "grad_norm": 1.657261848449707, "learning_rate": 9.999618556326518e-06, "loss": 9.2969, "step": 122 }, { "epoch": 0.004432012206197551, "eval_accuracy": 0.08103369879278215, "eval_loss": 9.2734375, "eval_runtime": 503.3156, "eval_samples_per_second": 67.089, "eval_steps_per_second": 2.795, "step": 122 }, { "epoch": 0.00446834017510081, "grad_norm": 1.9978619813919067, "learning_rate": 9.999614923529626e-06, "loss": 9.25, "step": 123 }, { "epoch": 0.00446834017510081, "eval_accuracy": 0.08151164493632851, "eval_loss": 9.265625, "eval_runtime": 503.1303, "eval_samples_per_second": 67.114, "eval_steps_per_second": 2.796, "step": 123 }, { "epoch": 0.004504668144004069, "grad_norm": 1.604614019393921, "learning_rate": 9.999611290732736e-06, "loss": 9.3281, "step": 124 }, { "epoch": 0.004504668144004069, "eval_accuracy": 0.0825292273040934, "eval_loss": 9.2578125, "eval_runtime": 505.0968, "eval_samples_per_second": 66.853, "eval_steps_per_second": 2.786, "step": 124 }, { "epoch": 0.004540996112907327, "grad_norm": 1.707974910736084, "learning_rate": 9.999607657935846e-06, "loss": 9.2656, "step": 125 }, { "epoch": 0.004540996112907327, "eval_accuracy": 0.08359170939739675, "eval_loss": 9.2578125, "eval_runtime": 506.8511, "eval_samples_per_second": 66.621, "eval_steps_per_second": 2.776, "step": 125 }, { "epoch": 0.004577324081810586, "grad_norm": 1.7586125135421753, "learning_rate": 9.999604025138955e-06, "loss": 9.3047, "step": 126 }, { "epoch": 0.004577324081810586, "eval_accuracy": 0.08447563474851999, "eval_loss": 9.25, "eval_runtime": 504.2303, "eval_samples_per_second": 66.967, "eval_steps_per_second": 2.79, "step": 126 }, { "epoch": 0.004613652050713845, "grad_norm": 1.7619037628173828, "learning_rate": 9.999600392342065e-06, "loss": 9.25, "step": 127 }, { "epoch": 0.004613652050713845, "eval_accuracy": 0.08498837745563648, "eval_loss": 9.2421875, "eval_runtime": 502.809, "eval_samples_per_second": 67.157, "eval_steps_per_second": 2.798, "step": 127 }, { "epoch": 0.004649980019617103, "grad_norm": 1.6512857675552368, "learning_rate": 9.999596759545175e-06, "loss": 9.2969, "step": 128 }, { "epoch": 0.004649980019617103, "eval_accuracy": 0.08521718946766497, "eval_loss": 9.234375, "eval_runtime": 504.7926, "eval_samples_per_second": 66.893, "eval_steps_per_second": 2.787, "step": 128 }, { "epoch": 0.004686307988520362, "grad_norm": 1.7101061344146729, "learning_rate": 9.999593126748285e-06, "loss": 9.3203, "step": 129 }, { "epoch": 0.004686307988520362, "eval_accuracy": 0.08527115019519801, "eval_loss": 9.234375, "eval_runtime": 505.6873, "eval_samples_per_second": 66.774, "eval_steps_per_second": 2.782, "step": 129 }, { "epoch": 0.004722635957423621, "grad_norm": 2.3074684143066406, "learning_rate": 9.999589493951393e-06, "loss": 9.25, "step": 130 }, { "epoch": 0.004722635957423621, "eval_accuracy": 0.08533631414244955, "eval_loss": 9.2265625, "eval_runtime": 504.004, "eval_samples_per_second": 66.997, "eval_steps_per_second": 2.792, "step": 130 }, { "epoch": 0.004758963926326879, "grad_norm": 1.86430025100708, "learning_rate": 9.999585861154503e-06, "loss": 9.2422, "step": 131 }, { "epoch": 0.004758963926326879, "eval_accuracy": 0.08537788474584947, "eval_loss": 9.21875, "eval_runtime": 502.5558, "eval_samples_per_second": 67.191, "eval_steps_per_second": 2.8, "step": 131 }, { "epoch": 0.0047952918952301376, "grad_norm": 1.7045161724090576, "learning_rate": 9.999582228357613e-06, "loss": 9.1641, "step": 132 }, { "epoch": 0.0047952918952301376, "eval_accuracy": 0.08549750155173277, "eval_loss": 9.2109375, "eval_runtime": 501.7481, "eval_samples_per_second": 67.299, "eval_steps_per_second": 2.804, "step": 132 }, { "epoch": 0.0048316198641333965, "grad_norm": 1.769091248512268, "learning_rate": 9.999578595560723e-06, "loss": 9.2109, "step": 133 }, { "epoch": 0.0048316198641333965, "eval_accuracy": 0.08575407554750815, "eval_loss": 9.2109375, "eval_runtime": 503.2243, "eval_samples_per_second": 67.101, "eval_steps_per_second": 2.796, "step": 133 }, { "epoch": 0.004867947833036655, "grad_norm": 1.7396882772445679, "learning_rate": 9.999574962763834e-06, "loss": 9.2422, "step": 134 }, { "epoch": 0.004867947833036655, "eval_accuracy": 0.08600703093226333, "eval_loss": 9.203125, "eval_runtime": 505.0001, "eval_samples_per_second": 66.865, "eval_steps_per_second": 2.786, "step": 134 }, { "epoch": 0.004904275801939913, "grad_norm": 1.5990197658538818, "learning_rate": 9.999571329966944e-06, "loss": 9.2188, "step": 135 }, { "epoch": 0.004904275801939913, "eval_accuracy": 0.08614508817990553, "eval_loss": 9.1953125, "eval_runtime": 503.7618, "eval_samples_per_second": 67.03, "eval_steps_per_second": 2.793, "step": 135 }, { "epoch": 0.004940603770843172, "grad_norm": 1.5657048225402832, "learning_rate": 9.999567697170052e-06, "loss": 9.3047, "step": 136 }, { "epoch": 0.004940603770843172, "eval_accuracy": 0.08612786359144943, "eval_loss": 9.1875, "eval_runtime": 504.5155, "eval_samples_per_second": 66.93, "eval_steps_per_second": 2.789, "step": 136 }, { "epoch": 0.004976931739746431, "grad_norm": 1.7186903953552246, "learning_rate": 9.999564064373162e-06, "loss": 9.1641, "step": 137 }, { "epoch": 0.004976931739746431, "eval_accuracy": 0.08613209012912101, "eval_loss": 9.1875, "eval_runtime": 504.0325, "eval_samples_per_second": 66.994, "eval_steps_per_second": 2.791, "step": 137 }, { "epoch": 0.005013259708649689, "grad_norm": 1.6939176321029663, "learning_rate": 9.999560431576272e-06, "loss": 9.2188, "step": 138 }, { "epoch": 0.005013259708649689, "eval_accuracy": 0.08593219805636586, "eval_loss": 9.1796875, "eval_runtime": 502.3491, "eval_samples_per_second": 67.218, "eval_steps_per_second": 2.801, "step": 138 }, { "epoch": 0.005049587677552948, "grad_norm": 1.617654800415039, "learning_rate": 9.99955679877938e-06, "loss": 9.2422, "step": 139 }, { "epoch": 0.005049587677552948, "eval_accuracy": 0.08563775891487524, "eval_loss": 9.171875, "eval_runtime": 503.9529, "eval_samples_per_second": 67.004, "eval_steps_per_second": 2.792, "step": 139 }, { "epoch": 0.005085915646456207, "grad_norm": 1.570475697517395, "learning_rate": 9.99955316598249e-06, "loss": 9.2422, "step": 140 }, { "epoch": 0.005085915646456207, "eval_accuracy": 0.0854915380807715, "eval_loss": 9.171875, "eval_runtime": 502.4421, "eval_samples_per_second": 67.206, "eval_steps_per_second": 2.8, "step": 140 }, { "epoch": 0.005122243615359465, "grad_norm": 1.7714993953704834, "learning_rate": 9.9995495331856e-06, "loss": 9.1484, "step": 141 }, { "epoch": 0.005122243615359465, "eval_accuracy": 0.08518311662629889, "eval_loss": 9.1640625, "eval_runtime": 503.434, "eval_samples_per_second": 67.073, "eval_steps_per_second": 2.795, "step": 141 }, { "epoch": 0.005158571584262724, "grad_norm": 1.7324341535568237, "learning_rate": 9.999545900388709e-06, "loss": 9.2422, "step": 142 }, { "epoch": 0.005158571584262724, "eval_accuracy": 0.08514038806737252, "eval_loss": 9.15625, "eval_runtime": 502.9892, "eval_samples_per_second": 67.133, "eval_steps_per_second": 2.797, "step": 142 }, { "epoch": 0.005194899553165983, "grad_norm": 1.5842994451522827, "learning_rate": 9.99954226759182e-06, "loss": 9.1953, "step": 143 }, { "epoch": 0.005194899553165983, "eval_accuracy": 0.0852237898141658, "eval_loss": 9.1484375, "eval_runtime": 502.9366, "eval_samples_per_second": 67.14, "eval_steps_per_second": 2.798, "step": 143 }, { "epoch": 0.005231227522069241, "grad_norm": 1.8025195598602295, "learning_rate": 9.99953863479493e-06, "loss": 9.1641, "step": 144 }, { "epoch": 0.005231227522069241, "eval_accuracy": 0.08531526830075614, "eval_loss": 9.1484375, "eval_runtime": 503.3947, "eval_samples_per_second": 67.079, "eval_steps_per_second": 2.795, "step": 144 }, { "epoch": 0.0052675554909725, "grad_norm": 1.6925685405731201, "learning_rate": 9.99953500199804e-06, "loss": 9.1875, "step": 145 }, { "epoch": 0.0052675554909725, "eval_accuracy": 0.0853937197876738, "eval_loss": 9.140625, "eval_runtime": 503.2502, "eval_samples_per_second": 67.098, "eval_steps_per_second": 2.796, "step": 145 }, { "epoch": 0.005303883459875759, "grad_norm": 1.6498448848724365, "learning_rate": 9.99953136920115e-06, "loss": 9.1172, "step": 146 }, { "epoch": 0.005303883459875759, "eval_accuracy": 0.0855044782337797, "eval_loss": 9.1328125, "eval_runtime": 501.7641, "eval_samples_per_second": 67.297, "eval_steps_per_second": 2.804, "step": 146 }, { "epoch": 0.005340211428779017, "grad_norm": 1.6799402236938477, "learning_rate": 9.999527736404258e-06, "loss": 9.1094, "step": 147 }, { "epoch": 0.005340211428779017, "eval_accuracy": 0.0855554572258321, "eval_loss": 9.1328125, "eval_runtime": 501.8847, "eval_samples_per_second": 67.28, "eval_steps_per_second": 2.803, "step": 147 }, { "epoch": 0.005376539397682276, "grad_norm": 1.672473669052124, "learning_rate": 9.999524103607368e-06, "loss": 9.1328, "step": 148 }, { "epoch": 0.005376539397682276, "eval_accuracy": 0.0858869798930576, "eval_loss": 9.125, "eval_runtime": 502.0405, "eval_samples_per_second": 67.26, "eval_steps_per_second": 2.803, "step": 148 }, { "epoch": 0.0054128673665855345, "grad_norm": 1.6717029809951782, "learning_rate": 9.999520470810478e-06, "loss": 9.1641, "step": 149 }, { "epoch": 0.0054128673665855345, "eval_accuracy": 0.0863320400996525, "eval_loss": 9.1171875, "eval_runtime": 501.7116, "eval_samples_per_second": 67.304, "eval_steps_per_second": 2.804, "step": 149 }, { "epoch": 0.0054491953354887926, "grad_norm": 1.5716333389282227, "learning_rate": 9.999516838013588e-06, "loss": 9.1641, "step": 150 }, { "epoch": 0.0054491953354887926, "eval_accuracy": 0.08683835615359713, "eval_loss": 9.109375, "eval_runtime": 502.3324, "eval_samples_per_second": 67.22, "eval_steps_per_second": 2.801, "step": 150 }, { "epoch": 0.0054855233043920515, "grad_norm": 1.624820351600647, "learning_rate": 9.999513205216698e-06, "loss": 9.1875, "step": 151 }, { "epoch": 0.0054855233043920515, "eval_accuracy": 0.08733792711660013, "eval_loss": 9.109375, "eval_runtime": 502.804, "eval_samples_per_second": 67.157, "eval_steps_per_second": 2.798, "step": 151 }, { "epoch": 0.00552185127329531, "grad_norm": 1.614936113357544, "learning_rate": 9.999509572419807e-06, "loss": 9.2031, "step": 152 }, { "epoch": 0.00552185127329531, "eval_accuracy": 0.08752846869847912, "eval_loss": 9.1015625, "eval_runtime": 502.8248, "eval_samples_per_second": 67.155, "eval_steps_per_second": 2.798, "step": 152 }, { "epoch": 0.005558179242198568, "grad_norm": 1.8622063398361206, "learning_rate": 9.999505939622917e-06, "loss": 9.0703, "step": 153 }, { "epoch": 0.005558179242198568, "eval_accuracy": 0.08799966975108385, "eval_loss": 9.09375, "eval_runtime": 501.9847, "eval_samples_per_second": 67.267, "eval_steps_per_second": 2.803, "step": 153 }, { "epoch": 0.005594507211101827, "grad_norm": 1.615602731704712, "learning_rate": 9.999502306826027e-06, "loss": 9.1484, "step": 154 }, { "epoch": 0.005594507211101827, "eval_accuracy": 0.08840796486971365, "eval_loss": 9.0859375, "eval_runtime": 501.6895, "eval_samples_per_second": 67.307, "eval_steps_per_second": 2.805, "step": 154 }, { "epoch": 0.005630835180005086, "grad_norm": 1.661203384399414, "learning_rate": 9.999498674029135e-06, "loss": 9.0625, "step": 155 }, { "epoch": 0.005630835180005086, "eval_accuracy": 0.08875231189439468, "eval_loss": 9.0859375, "eval_runtime": 501.9752, "eval_samples_per_second": 67.268, "eval_steps_per_second": 2.803, "step": 155 }, { "epoch": 0.005667163148908344, "grad_norm": 1.7897155284881592, "learning_rate": 9.999495041232245e-06, "loss": 9.0781, "step": 156 }, { "epoch": 0.005667163148908344, "eval_accuracy": 0.08888113444671336, "eval_loss": 9.078125, "eval_runtime": 502.8161, "eval_samples_per_second": 67.156, "eval_steps_per_second": 2.798, "step": 156 }, { "epoch": 0.005703491117811603, "grad_norm": 1.9064967632293701, "learning_rate": 9.999491408435355e-06, "loss": 9.0234, "step": 157 }, { "epoch": 0.005703491117811603, "eval_accuracy": 0.08920223551420071, "eval_loss": 9.0703125, "eval_runtime": 502.0043, "eval_samples_per_second": 67.264, "eval_steps_per_second": 2.803, "step": 157 }, { "epoch": 0.005739819086714862, "grad_norm": 1.635931134223938, "learning_rate": 9.999487775638464e-06, "loss": 9.0781, "step": 158 }, { "epoch": 0.005739819086714862, "eval_accuracy": 0.08936993642332029, "eval_loss": 9.0703125, "eval_runtime": 503.082, "eval_samples_per_second": 67.12, "eval_steps_per_second": 2.797, "step": 158 }, { "epoch": 0.00577614705561812, "grad_norm": 1.833716869354248, "learning_rate": 9.999484142841574e-06, "loss": 9.0, "step": 159 }, { "epoch": 0.00577614705561812, "eval_accuracy": 0.08952518931053041, "eval_loss": 9.0625, "eval_runtime": 501.8316, "eval_samples_per_second": 67.288, "eval_steps_per_second": 2.804, "step": 159 }, { "epoch": 0.005812475024521379, "grad_norm": 1.7018239498138428, "learning_rate": 9.999480510044684e-06, "loss": 9.0312, "step": 160 }, { "epoch": 0.005812475024521379, "eval_accuracy": 0.08957336026043114, "eval_loss": 9.0546875, "eval_runtime": 502.712, "eval_samples_per_second": 67.17, "eval_steps_per_second": 2.799, "step": 160 }, { "epoch": 0.005848802993424638, "grad_norm": 1.6031908988952637, "learning_rate": 9.999476877247794e-06, "loss": 9.0391, "step": 161 }, { "epoch": 0.005848802993424638, "eval_accuracy": 0.08980457503017704, "eval_loss": 9.0546875, "eval_runtime": 504.1644, "eval_samples_per_second": 66.976, "eval_steps_per_second": 2.791, "step": 161 }, { "epoch": 0.005885130962327896, "grad_norm": 1.8437420129776, "learning_rate": 9.999473244450904e-06, "loss": 9.0469, "step": 162 }, { "epoch": 0.005885130962327896, "eval_accuracy": 0.09014976157261477, "eval_loss": 9.046875, "eval_runtime": 504.1854, "eval_samples_per_second": 66.973, "eval_steps_per_second": 2.791, "step": 162 }, { "epoch": 0.005921458931231155, "grad_norm": 1.765324354171753, "learning_rate": 9.999469611654014e-06, "loss": 9.0859, "step": 163 }, { "epoch": 0.005921458931231155, "eval_accuracy": 0.09048160267761005, "eval_loss": 9.0390625, "eval_runtime": 502.1677, "eval_samples_per_second": 67.242, "eval_steps_per_second": 2.802, "step": 163 }, { "epoch": 0.005957786900134414, "grad_norm": 1.6684796810150146, "learning_rate": 9.999465978857123e-06, "loss": 9.0078, "step": 164 }, { "epoch": 0.005957786900134414, "eval_accuracy": 0.0907586435373156, "eval_loss": 9.03125, "eval_runtime": 503.5057, "eval_samples_per_second": 67.064, "eval_steps_per_second": 2.794, "step": 164 }, { "epoch": 0.005994114869037672, "grad_norm": 1.840463399887085, "learning_rate": 9.999462346060233e-06, "loss": 9.0156, "step": 165 }, { "epoch": 0.005994114869037672, "eval_accuracy": 0.09093132365519895, "eval_loss": 9.03125, "eval_runtime": 502.0204, "eval_samples_per_second": 67.262, "eval_steps_per_second": 2.803, "step": 165 }, { "epoch": 0.006030442837940931, "grad_norm": 1.782490849494934, "learning_rate": 9.999458713263343e-06, "loss": 9.0469, "step": 166 }, { "epoch": 0.006030442837940931, "eval_accuracy": 0.09093610022174559, "eval_loss": 9.0234375, "eval_runtime": 502.0137, "eval_samples_per_second": 67.263, "eval_steps_per_second": 2.803, "step": 166 }, { "epoch": 0.0060667708068441895, "grad_norm": 2.1105000972747803, "learning_rate": 9.999455080466451e-06, "loss": 8.9219, "step": 167 }, { "epoch": 0.0060667708068441895, "eval_accuracy": 0.09079144262760257, "eval_loss": 9.0234375, "eval_runtime": 503.1259, "eval_samples_per_second": 67.114, "eval_steps_per_second": 2.797, "step": 167 }, { "epoch": 0.0061030987757474476, "grad_norm": 1.6616843938827515, "learning_rate": 9.999451447669561e-06, "loss": 9.0312, "step": 168 }, { "epoch": 0.0061030987757474476, "eval_accuracy": 0.09065688819542793, "eval_loss": 9.015625, "eval_runtime": 502.4938, "eval_samples_per_second": 67.199, "eval_steps_per_second": 2.8, "step": 168 }, { "epoch": 0.0061394267446507065, "grad_norm": 1.6470061540603638, "learning_rate": 9.999447814872671e-06, "loss": 9.0938, "step": 169 }, { "epoch": 0.0061394267446507065, "eval_accuracy": 0.09056121212005416, "eval_loss": 9.0078125, "eval_runtime": 503.0903, "eval_samples_per_second": 67.119, "eval_steps_per_second": 2.797, "step": 169 }, { "epoch": 0.006175754713553965, "grad_norm": 1.594428539276123, "learning_rate": 9.999444182075781e-06, "loss": 9.0156, "step": 170 }, { "epoch": 0.006175754713553965, "eval_accuracy": 0.09015482762804303, "eval_loss": 9.0, "eval_runtime": 502.1605, "eval_samples_per_second": 67.243, "eval_steps_per_second": 2.802, "step": 170 }, { "epoch": 0.006212082682457223, "grad_norm": 1.5892345905303955, "learning_rate": 9.99944054927889e-06, "loss": 9.0312, "step": 171 }, { "epoch": 0.006212082682457223, "eval_accuracy": 0.08973292653197733, "eval_loss": 9.0, "eval_runtime": 500.8624, "eval_samples_per_second": 67.418, "eval_steps_per_second": 2.809, "step": 171 }, { "epoch": 0.006248410651360482, "grad_norm": 1.5640125274658203, "learning_rate": 9.999436916482e-06, "loss": 9.0625, "step": 172 }, { "epoch": 0.006248410651360482, "eval_accuracy": 0.08926792054143916, "eval_loss": 8.9921875, "eval_runtime": 501.8903, "eval_samples_per_second": 67.28, "eval_steps_per_second": 2.803, "step": 172 }, { "epoch": 0.006284738620263741, "grad_norm": 2.097386121749878, "learning_rate": 9.99943328368511e-06, "loss": 8.9844, "step": 173 }, { "epoch": 0.006284738620263741, "eval_accuracy": 0.08908942169703535, "eval_loss": 8.984375, "eval_runtime": 503.348, "eval_samples_per_second": 67.085, "eval_steps_per_second": 2.795, "step": 173 }, { "epoch": 0.006321066589167, "grad_norm": 1.6901289224624634, "learning_rate": 9.99942965088822e-06, "loss": 9.0703, "step": 174 }, { "epoch": 0.006321066589167, "eval_accuracy": 0.08940612253352216, "eval_loss": 8.984375, "eval_runtime": 506.45, "eval_samples_per_second": 66.674, "eval_steps_per_second": 2.778, "step": 174 }, { "epoch": 0.006357394558070258, "grad_norm": 1.6152195930480957, "learning_rate": 9.99942601809133e-06, "loss": 8.9609, "step": 175 }, { "epoch": 0.006357394558070258, "eval_accuracy": 0.08984990898903795, "eval_loss": 8.9765625, "eval_runtime": 504.771, "eval_samples_per_second": 66.896, "eval_steps_per_second": 2.787, "step": 175 }, { "epoch": 0.006393722526973517, "grad_norm": 1.6358169317245483, "learning_rate": 9.99942238529444e-06, "loss": 8.9922, "step": 176 }, { "epoch": 0.006393722526973517, "eval_accuracy": 0.09047048630455602, "eval_loss": 8.9765625, "eval_runtime": 504.0679, "eval_samples_per_second": 66.989, "eval_steps_per_second": 2.791, "step": 176 }, { "epoch": 0.006430050495876776, "grad_norm": 1.662377953529358, "learning_rate": 9.999418752497549e-06, "loss": 9.0234, "step": 177 }, { "epoch": 0.006430050495876776, "eval_accuracy": 0.09099935354237847, "eval_loss": 8.96875, "eval_runtime": 504.1525, "eval_samples_per_second": 66.978, "eval_steps_per_second": 2.791, "step": 177 }, { "epoch": 0.006466378464780034, "grad_norm": 1.6537806987762451, "learning_rate": 9.999415119700659e-06, "loss": 9.0234, "step": 178 }, { "epoch": 0.006466378464780034, "eval_accuracy": 0.0914514483287966, "eval_loss": 8.9609375, "eval_runtime": 502.7884, "eval_samples_per_second": 67.159, "eval_steps_per_second": 2.798, "step": 178 }, { "epoch": 0.006502706433683293, "grad_norm": 1.7109649181365967, "learning_rate": 9.999411486903769e-06, "loss": 8.9219, "step": 179 }, { "epoch": 0.006502706433683293, "eval_accuracy": 0.09188437895125184, "eval_loss": 8.953125, "eval_runtime": 501.2279, "eval_samples_per_second": 67.369, "eval_steps_per_second": 2.807, "step": 179 }, { "epoch": 0.006539034402586552, "grad_norm": 1.64595365524292, "learning_rate": 9.999407854106877e-06, "loss": 9.0234, "step": 180 }, { "epoch": 0.006539034402586552, "eval_accuracy": 0.09202229145445322, "eval_loss": 8.953125, "eval_runtime": 501.0383, "eval_samples_per_second": 67.394, "eval_steps_per_second": 2.808, "step": 180 }, { "epoch": 0.00657536237148981, "grad_norm": 1.5911909341812134, "learning_rate": 9.999404221309987e-06, "loss": 8.9375, "step": 181 }, { "epoch": 0.00657536237148981, "eval_accuracy": 0.09211403048103702, "eval_loss": 8.9453125, "eval_runtime": 503.6267, "eval_samples_per_second": 67.048, "eval_steps_per_second": 2.794, "step": 181 }, { "epoch": 0.006611690340393069, "grad_norm": 1.4985125064849854, "learning_rate": 9.999400588513097e-06, "loss": 8.9688, "step": 182 }, { "epoch": 0.006611690340393069, "eval_accuracy": 0.09187829968473793, "eval_loss": 8.9375, "eval_runtime": 504.6287, "eval_samples_per_second": 66.915, "eval_steps_per_second": 2.788, "step": 182 }, { "epoch": 0.006648018309296328, "grad_norm": 1.7256035804748535, "learning_rate": 9.999396955716206e-06, "loss": 8.9375, "step": 183 }, { "epoch": 0.006648018309296328, "eval_accuracy": 0.09130482221025861, "eval_loss": 8.9375, "eval_runtime": 502.7855, "eval_samples_per_second": 67.16, "eval_steps_per_second": 2.798, "step": 183 }, { "epoch": 0.006684346278199586, "grad_norm": 1.5093938112258911, "learning_rate": 9.999393322919316e-06, "loss": 9.0, "step": 184 }, { "epoch": 0.006684346278199586, "eval_accuracy": 0.09121345951922091, "eval_loss": 8.9296875, "eval_runtime": 502.1658, "eval_samples_per_second": 67.243, "eval_steps_per_second": 2.802, "step": 184 }, { "epoch": 0.0067206742471028445, "grad_norm": 1.540442943572998, "learning_rate": 9.999389690122426e-06, "loss": 8.9375, "step": 185 }, { "epoch": 0.0067206742471028445, "eval_accuracy": 0.09127399164436661, "eval_loss": 8.921875, "eval_runtime": 502.1504, "eval_samples_per_second": 67.245, "eval_steps_per_second": 2.802, "step": 185 }, { "epoch": 0.006757002216006103, "grad_norm": 1.5126348733901978, "learning_rate": 9.999386057325536e-06, "loss": 8.9609, "step": 186 }, { "epoch": 0.006757002216006103, "eval_accuracy": 0.09126880979338571, "eval_loss": 8.921875, "eval_runtime": 500.7274, "eval_samples_per_second": 67.436, "eval_steps_per_second": 2.81, "step": 186 }, { "epoch": 0.0067933301849093615, "grad_norm": 1.6009377241134644, "learning_rate": 9.999382424528646e-06, "loss": 8.9688, "step": 187 }, { "epoch": 0.0067933301849093615, "eval_accuracy": 0.09167533902983765, "eval_loss": 8.9140625, "eval_runtime": 498.1171, "eval_samples_per_second": 67.789, "eval_steps_per_second": 2.825, "step": 187 }, { "epoch": 0.0067933301849093615, "step": 187, "total_flos": 1043124562427904.0, "train_loss": 9.606408756684491, "train_runtime": 94484.0244, "train_samples_per_second": 699.211, "train_steps_per_second": 29.134 } ], "logging_steps": 1, "max_steps": 2752700, "num_input_tokens_seen": 0, "num_train_epochs": 100, "save_steps": 1000000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": false, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1043124562427904.0, "train_batch_size": 24, "trial_name": null, "trial_params": null }