| { | |
| "best_metric": 0.4396902620792389, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 3.008695652173913, | |
| "eval_steps": 50, | |
| "global_step": 173, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017391304347826087, | |
| "grad_norm": 8.883275032043457, | |
| "learning_rate": 1e-05, | |
| "loss": 4.2926, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017391304347826087, | |
| "eval_loss": 1.9020391702651978, | |
| "eval_runtime": 9.0598, | |
| "eval_samples_per_second": 10.707, | |
| "eval_steps_per_second": 2.759, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.034782608695652174, | |
| "grad_norm": 18.399316787719727, | |
| "learning_rate": 2e-05, | |
| "loss": 4.2206, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05217391304347826, | |
| "grad_norm": 18.08584213256836, | |
| "learning_rate": 3e-05, | |
| "loss": 4.1941, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06956521739130435, | |
| "grad_norm": 18.791566848754883, | |
| "learning_rate": 4e-05, | |
| "loss": 3.7203, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08695652173913043, | |
| "grad_norm": 14.128814697265625, | |
| "learning_rate": 5e-05, | |
| "loss": 2.8787, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10434782608695652, | |
| "grad_norm": 13.8895263671875, | |
| "learning_rate": 6e-05, | |
| "loss": 2.7352, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12173913043478261, | |
| "grad_norm": 12.573573112487793, | |
| "learning_rate": 7e-05, | |
| "loss": 1.7571, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1391304347826087, | |
| "grad_norm": 17.67091178894043, | |
| "learning_rate": 8e-05, | |
| "loss": 2.9257, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1565217391304348, | |
| "grad_norm": 26.21525001525879, | |
| "learning_rate": 9e-05, | |
| "loss": 3.6193, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 32.96034622192383, | |
| "learning_rate": 0.0001, | |
| "loss": 3.5862, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19130434782608696, | |
| "grad_norm": 34.51445388793945, | |
| "learning_rate": 9.999071352056675e-05, | |
| "loss": 2.0717, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.20869565217391303, | |
| "grad_norm": 20.016292572021484, | |
| "learning_rate": 9.9962857531815e-05, | |
| "loss": 2.6985, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22608695652173913, | |
| "grad_norm": 29.78652572631836, | |
| "learning_rate": 9.99164423811074e-05, | |
| "loss": 2.7999, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.24347826086956523, | |
| "grad_norm": 103.21263122558594, | |
| "learning_rate": 9.985148530977767e-05, | |
| "loss": 4.0946, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 13.88134765625, | |
| "learning_rate": 9.976801044672608e-05, | |
| "loss": 4.4529, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2782608695652174, | |
| "grad_norm": 12.383321762084961, | |
| "learning_rate": 9.966604879945659e-05, | |
| "loss": 4.0322, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2956521739130435, | |
| "grad_norm": 16.159950256347656, | |
| "learning_rate": 9.954563824255878e-05, | |
| "loss": 3.4233, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3130434782608696, | |
| "grad_norm": 10.361807823181152, | |
| "learning_rate": 9.940682350363912e-05, | |
| "loss": 2.4715, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.33043478260869563, | |
| "grad_norm": 11.233977317810059, | |
| "learning_rate": 9.924965614670629e-05, | |
| "loss": 3.0579, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 9.937857627868652, | |
| "learning_rate": 9.907419455301741e-05, | |
| "loss": 1.8839, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3652173913043478, | |
| "grad_norm": 11.618575096130371, | |
| "learning_rate": 9.888050389939172e-05, | |
| "loss": 2.0335, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3826086956521739, | |
| "grad_norm": 12.595673561096191, | |
| "learning_rate": 9.866865613400008e-05, | |
| "loss": 2.1181, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 18.374019622802734, | |
| "learning_rate": 9.843872994963911e-05, | |
| "loss": 4.7754, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.41739130434782606, | |
| "grad_norm": 20.49725914001465, | |
| "learning_rate": 9.819081075450014e-05, | |
| "loss": 1.8529, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 9.583206176757812, | |
| "learning_rate": 9.792499064044342e-05, | |
| "loss": 1.1469, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.45217391304347826, | |
| "grad_norm": 14.794954299926758, | |
| "learning_rate": 9.764136834878986e-05, | |
| "loss": 1.3944, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.46956521739130436, | |
| "grad_norm": 17.454011917114258, | |
| "learning_rate": 9.734004923364257e-05, | |
| "loss": 2.2708, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.48695652173913045, | |
| "grad_norm": 42.63334274291992, | |
| "learning_rate": 9.702114522275216e-05, | |
| "loss": 1.3025, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5043478260869565, | |
| "grad_norm": 8.601645469665527, | |
| "learning_rate": 9.66847747759402e-05, | |
| "loss": 4.1274, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 10.501288414001465, | |
| "learning_rate": 9.63310628410961e-05, | |
| "loss": 3.3343, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5391304347826087, | |
| "grad_norm": 10.755828857421875, | |
| "learning_rate": 9.596014080776423e-05, | |
| "loss": 1.7963, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5565217391304348, | |
| "grad_norm": 7.994204044342041, | |
| "learning_rate": 9.557214645833792e-05, | |
| "loss": 1.995, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5739130434782609, | |
| "grad_norm": 11.35947322845459, | |
| "learning_rate": 9.516722391687902e-05, | |
| "loss": 2.3837, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.591304347826087, | |
| "grad_norm": 9.96140193939209, | |
| "learning_rate": 9.474552359558166e-05, | |
| "loss": 1.6152, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6086956521739131, | |
| "grad_norm": 8.912185668945312, | |
| "learning_rate": 9.43072021389003e-05, | |
| "loss": 2.9803, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6260869565217392, | |
| "grad_norm": 14.464909553527832, | |
| "learning_rate": 9.38524223653626e-05, | |
| "loss": 2.7963, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6434782608695652, | |
| "grad_norm": 11.840982437133789, | |
| "learning_rate": 9.338135320708911e-05, | |
| "loss": 3.135, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6608695652173913, | |
| "grad_norm": 11.059337615966797, | |
| "learning_rate": 9.289416964704185e-05, | |
| "loss": 1.1317, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6782608695652174, | |
| "grad_norm": 10.759260177612305, | |
| "learning_rate": 9.239105265402525e-05, | |
| "loss": 1.2317, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 8.21378231048584, | |
| "learning_rate": 9.187218911546362e-05, | |
| "loss": 0.8413, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7130434782608696, | |
| "grad_norm": 17.352684020996094, | |
| "learning_rate": 9.133777176798013e-05, | |
| "loss": 1.3149, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7304347826086957, | |
| "grad_norm": 26.137418746948242, | |
| "learning_rate": 9.078799912580304e-05, | |
| "loss": 2.0237, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7478260869565218, | |
| "grad_norm": 8.654214859008789, | |
| "learning_rate": 9.022307540702576e-05, | |
| "loss": 3.6969, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7652173913043478, | |
| "grad_norm": 11.014862060546875, | |
| "learning_rate": 8.964321045774807e-05, | |
| "loss": 3.2013, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.782608695652174, | |
| "grad_norm": 8.544939994812012, | |
| "learning_rate": 8.904861967412703e-05, | |
| "loss": 3.5936, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 10.577417373657227, | |
| "learning_rate": 8.843952392236594e-05, | |
| "loss": 2.7795, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8173913043478261, | |
| "grad_norm": 8.548279762268066, | |
| "learning_rate": 8.781614945667169e-05, | |
| "loss": 2.1591, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8347826086956521, | |
| "grad_norm": 9.758591651916504, | |
| "learning_rate": 8.717872783521047e-05, | |
| "loss": 2.1567, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8521739130434782, | |
| "grad_norm": 9.762876510620117, | |
| "learning_rate": 8.65274958340934e-05, | |
| "loss": 2.3352, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 8.696921348571777, | |
| "learning_rate": 8.586269535942385e-05, | |
| "loss": 2.2677, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "eval_loss": 0.4909374713897705, | |
| "eval_runtime": 9.2435, | |
| "eval_samples_per_second": 10.494, | |
| "eval_steps_per_second": 2.705, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8869565217391304, | |
| "grad_norm": 11.873095512390137, | |
| "learning_rate": 8.518457335743926e-05, | |
| "loss": 2.753, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9043478260869565, | |
| "grad_norm": 12.200470924377441, | |
| "learning_rate": 8.449338172278059e-05, | |
| "loss": 2.466, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9217391304347826, | |
| "grad_norm": 26.438901901245117, | |
| "learning_rate": 8.378937720492384e-05, | |
| "loss": 1.0808, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.9391304347826087, | |
| "grad_norm": 22.44983673095703, | |
| "learning_rate": 8.307282131280804e-05, | |
| "loss": 0.7925, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9565217391304348, | |
| "grad_norm": 9.913865089416504, | |
| "learning_rate": 8.23439802176954e-05, | |
| "loss": 1.5618, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9739130434782609, | |
| "grad_norm": 106.08903503417969, | |
| "learning_rate": 8.160312465429952e-05, | |
| "loss": 2.0868, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.991304347826087, | |
| "grad_norm": 10.000048637390137, | |
| "learning_rate": 8.085052982021847e-05, | |
| "loss": 2.5997, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.008695652173913, | |
| "grad_norm": 9.5238676071167, | |
| "learning_rate": 8.008647527371023e-05, | |
| "loss": 2.9929, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0260869565217392, | |
| "grad_norm": 5.974183082580566, | |
| "learning_rate": 7.931124482984802e-05, | |
| "loss": 2.2291, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 7.498763561248779, | |
| "learning_rate": 7.85251264550948e-05, | |
| "loss": 1.6798, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0608695652173914, | |
| "grad_norm": 6.492100715637207, | |
| "learning_rate": 7.772841216033533e-05, | |
| "loss": 1.8157, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.0782608695652174, | |
| "grad_norm": 6.82267951965332, | |
| "learning_rate": 7.692139789240611e-05, | |
| "loss": 1.2322, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0956521739130434, | |
| "grad_norm": 5.267804145812988, | |
| "learning_rate": 7.610438342416319e-05, | |
| "loss": 1.5287, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1130434782608696, | |
| "grad_norm": 5.903561115264893, | |
| "learning_rate": 7.527767224312883e-05, | |
| "loss": 0.9725, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1304347826086956, | |
| "grad_norm": 6.4522624015808105, | |
| "learning_rate": 7.44415714387582e-05, | |
| "loss": 1.1919, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1478260869565218, | |
| "grad_norm": 12.900763511657715, | |
| "learning_rate": 7.359639158836828e-05, | |
| "loss": 1.3117, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1652173913043478, | |
| "grad_norm": 8.239570617675781, | |
| "learning_rate": 7.274244664177097e-05, | |
| "loss": 1.0773, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.182608695652174, | |
| "grad_norm": 6.0005364418029785, | |
| "learning_rate": 7.188005380465364e-05, | |
| "loss": 0.3932, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 8.502946853637695, | |
| "learning_rate": 7.10095334207501e-05, | |
| "loss": 0.9309, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.2173913043478262, | |
| "grad_norm": 13.114483833312988, | |
| "learning_rate": 7.013120885284598e-05, | |
| "loss": 0.6948, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2347826086956522, | |
| "grad_norm": 12.373690605163574, | |
| "learning_rate": 6.924540636266272e-05, | |
| "loss": 1.1723, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.2521739130434781, | |
| "grad_norm": 8.012110710144043, | |
| "learning_rate": 6.835245498966461e-05, | |
| "loss": 2.3979, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.2695652173913043, | |
| "grad_norm": 7.275670528411865, | |
| "learning_rate": 6.745268642883404e-05, | |
| "loss": 2.593, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.2869565217391306, | |
| "grad_norm": 8.32419490814209, | |
| "learning_rate": 6.654643490746042e-05, | |
| "loss": 1.9259, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.3043478260869565, | |
| "grad_norm": 6.298715591430664, | |
| "learning_rate": 6.563403706098833e-05, | |
| "loss": 1.7114, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.3217391304347825, | |
| "grad_norm": 8.741583824157715, | |
| "learning_rate": 6.471583180797121e-05, | |
| "loss": 1.3124, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3391304347826087, | |
| "grad_norm": 6.135732650756836, | |
| "learning_rate": 6.379216022417696e-05, | |
| "loss": 1.1135, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.3565217391304347, | |
| "grad_norm": 7.119026184082031, | |
| "learning_rate": 6.286336541589224e-05, | |
| "loss": 0.8562, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.373913043478261, | |
| "grad_norm": 5.54868221282959, | |
| "learning_rate": 6.192979239247243e-05, | |
| "loss": 0.9345, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 5.854312896728516, | |
| "learning_rate": 6.0991787938184784e-05, | |
| "loss": 0.9421, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4086956521739131, | |
| "grad_norm": 8.708279609680176, | |
| "learning_rate": 6.004970048339226e-05, | |
| "loss": 1.6717, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.4260869565217391, | |
| "grad_norm": 14.216882705688477, | |
| "learning_rate": 5.910387997512573e-05, | |
| "loss": 1.0587, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.4434782608695653, | |
| "grad_norm": 9.43204116821289, | |
| "learning_rate": 5.8154677747093134e-05, | |
| "loss": 0.3196, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.4608695652173913, | |
| "grad_norm": 21.38327980041504, | |
| "learning_rate": 5.7202446389173223e-05, | |
| "loss": 0.6099, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.4782608695652173, | |
| "grad_norm": 12.446172714233398, | |
| "learning_rate": 5.624753961644281e-05, | |
| "loss": 0.5437, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.4956521739130435, | |
| "grad_norm": 19.80995750427246, | |
| "learning_rate": 5.5290312137786146e-05, | |
| "loss": 1.503, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.5130434782608697, | |
| "grad_norm": 5.809496879577637, | |
| "learning_rate": 5.433111952413495e-05, | |
| "loss": 2.0654, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.5304347826086957, | |
| "grad_norm": 7.104956150054932, | |
| "learning_rate": 5.33703180763884e-05, | |
| "loss": 1.8031, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.5478260869565217, | |
| "grad_norm": 8.171567916870117, | |
| "learning_rate": 5.240826469306187e-05, | |
| "loss": 1.3728, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.5652173913043477, | |
| "grad_norm": 7.269556522369385, | |
| "learning_rate": 5.144531673771363e-05, | |
| "loss": 1.2346, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5826086956521739, | |
| "grad_norm": 5.152873992919922, | |
| "learning_rate": 5.048183190619904e-05, | |
| "loss": 0.7145, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 6.587265491485596, | |
| "learning_rate": 4.951816809380097e-05, | |
| "loss": 1.0577, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.617391304347826, | |
| "grad_norm": 7.419044017791748, | |
| "learning_rate": 4.855468326228638e-05, | |
| "loss": 1.4921, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.634782608695652, | |
| "grad_norm": 7.340956687927246, | |
| "learning_rate": 4.759173530693814e-05, | |
| "loss": 1.1331, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.6521739130434783, | |
| "grad_norm": 9.950460433959961, | |
| "learning_rate": 4.6629681923611603e-05, | |
| "loss": 1.1272, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6695652173913045, | |
| "grad_norm": 26.5368595123291, | |
| "learning_rate": 4.566888047586507e-05, | |
| "loss": 0.6709, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.6869565217391305, | |
| "grad_norm": 7.289899826049805, | |
| "learning_rate": 4.4709687862213866e-05, | |
| "loss": 0.5705, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.7043478260869565, | |
| "grad_norm": 30.423858642578125, | |
| "learning_rate": 4.3752460383557195e-05, | |
| "loss": 1.0803, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.7217391304347827, | |
| "grad_norm": 9.387825965881348, | |
| "learning_rate": 4.27975536108268e-05, | |
| "loss": 0.5288, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 15.499227523803711, | |
| "learning_rate": 4.1845322252906864e-05, | |
| "loss": 2.4856, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 0.4396902620792389, | |
| "eval_runtime": 9.2547, | |
| "eval_samples_per_second": 10.481, | |
| "eval_steps_per_second": 2.701, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7565217391304349, | |
| "grad_norm": 9.441740989685059, | |
| "learning_rate": 4.0896120024874286e-05, | |
| "loss": 2.5925, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.7739130434782608, | |
| "grad_norm": 6.262640953063965, | |
| "learning_rate": 3.9950299516607766e-05, | |
| "loss": 1.9076, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.7913043478260868, | |
| "grad_norm": 6.341916084289551, | |
| "learning_rate": 3.900821206181521e-05, | |
| "loss": 1.2718, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.808695652173913, | |
| "grad_norm": 6.276321887969971, | |
| "learning_rate": 3.8070207607527584e-05, | |
| "loss": 1.3664, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.8260869565217392, | |
| "grad_norm": 6.208406925201416, | |
| "learning_rate": 3.713663458410779e-05, | |
| "loss": 0.9109, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.8434782608695652, | |
| "grad_norm": 4.897834300994873, | |
| "learning_rate": 3.620783977582305e-05, | |
| "loss": 0.7668, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.8608695652173912, | |
| "grad_norm": 7.578703880310059, | |
| "learning_rate": 3.528416819202881e-05, | |
| "loss": 1.0439, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.8782608695652174, | |
| "grad_norm": 7.348963737487793, | |
| "learning_rate": 3.43659629390117e-05, | |
| "loss": 1.5878, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.8956521739130436, | |
| "grad_norm": 9.581995964050293, | |
| "learning_rate": 3.345356509253959e-05, | |
| "loss": 1.3644, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "grad_norm": 10.65773868560791, | |
| "learning_rate": 3.254731357116597e-05, | |
| "loss": 0.4552, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.9304347826086956, | |
| "grad_norm": 8.8335542678833, | |
| "learning_rate": 3.16475450103354e-05, | |
| "loss": 0.2004, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.9478260869565216, | |
| "grad_norm": 4.476780891418457, | |
| "learning_rate": 3.0754593637337276e-05, | |
| "loss": 0.3355, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.9652173913043478, | |
| "grad_norm": 6.725592613220215, | |
| "learning_rate": 2.986879114715403e-05, | |
| "loss": 0.4255, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.982608695652174, | |
| "grad_norm": 4.152976989746094, | |
| "learning_rate": 2.899046657924992e-05, | |
| "loss": 1.3199, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 8.147799491882324, | |
| "learning_rate": 2.8119946195346375e-05, | |
| "loss": 1.0268, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.017391304347826, | |
| "grad_norm": 3.997720241546631, | |
| "learning_rate": 2.7257553358229034e-05, | |
| "loss": 2.3721, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.034782608695652, | |
| "grad_norm": 4.236757755279541, | |
| "learning_rate": 2.6403608411631742e-05, | |
| "loss": 1.0922, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.0521739130434784, | |
| "grad_norm": 4.957376956939697, | |
| "learning_rate": 2.555842856124182e-05, | |
| "loss": 0.9086, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.0695652173913044, | |
| "grad_norm": 5.636981964111328, | |
| "learning_rate": 2.472232775687119e-05, | |
| "loss": 0.5955, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 3.7651617527008057, | |
| "learning_rate": 2.389561657583681e-05, | |
| "loss": 0.4258, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.1043478260869564, | |
| "grad_norm": 4.104703903198242, | |
| "learning_rate": 2.30786021075939e-05, | |
| "loss": 0.3615, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.121739130434783, | |
| "grad_norm": 4.564859390258789, | |
| "learning_rate": 2.2271587839664672e-05, | |
| "loss": 0.3972, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.139130434782609, | |
| "grad_norm": 5.188112735748291, | |
| "learning_rate": 2.1474873544905205e-05, | |
| "loss": 0.5548, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.1565217391304348, | |
| "grad_norm": 5.654900550842285, | |
| "learning_rate": 2.0688755170151996e-05, | |
| "loss": 0.6482, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.1739130434782608, | |
| "grad_norm": 6.888501167297363, | |
| "learning_rate": 1.9913524726289784e-05, | |
| "loss": 0.3852, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.1913043478260867, | |
| "grad_norm": 6.729064464569092, | |
| "learning_rate": 1.914947017978153e-05, | |
| "loss": 0.282, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.208695652173913, | |
| "grad_norm": 8.329999923706055, | |
| "learning_rate": 1.8396875345700497e-05, | |
| "loss": 0.1444, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.226086956521739, | |
| "grad_norm": 9.099133491516113, | |
| "learning_rate": 1.76560197823046e-05, | |
| "loss": 0.435, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.243478260869565, | |
| "grad_norm": 3.2181084156036377, | |
| "learning_rate": 1.692717868719195e-05, | |
| "loss": 0.03, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.260869565217391, | |
| "grad_norm": 5.805236339569092, | |
| "learning_rate": 1.621062279507617e-05, | |
| "loss": 1.2943, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.2782608695652176, | |
| "grad_norm": 9.254953384399414, | |
| "learning_rate": 1.550661827721941e-05, | |
| "loss": 1.424, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.2956521739130435, | |
| "grad_norm": 7.753330230712891, | |
| "learning_rate": 1.4815426642560754e-05, | |
| "loss": 0.8185, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.3130434782608695, | |
| "grad_norm": 9.017913818359375, | |
| "learning_rate": 1.413730464057616e-05, | |
| "loss": 0.7997, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.3304347826086955, | |
| "grad_norm": 6.765071868896484, | |
| "learning_rate": 1.3472504165906613e-05, | |
| "loss": 0.4594, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.3478260869565215, | |
| "grad_norm": 4.712010860443115, | |
| "learning_rate": 1.2821272164789544e-05, | |
| "loss": 0.2742, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.365217391304348, | |
| "grad_norm": 4.8088250160217285, | |
| "learning_rate": 1.2183850543328312e-05, | |
| "loss": 0.2534, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.382608695652174, | |
| "grad_norm": 10.322266578674316, | |
| "learning_rate": 1.156047607763407e-05, | |
| "loss": 0.4052, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 15.81982421875, | |
| "learning_rate": 1.0951380325872979e-05, | |
| "loss": 0.5153, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.417391304347826, | |
| "grad_norm": 4.597114562988281, | |
| "learning_rate": 1.0356789542251938e-05, | |
| "loss": 0.1805, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 10.027156829833984, | |
| "learning_rate": 9.776924592974256e-06, | |
| "loss": 0.084, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.4521739130434783, | |
| "grad_norm": 4.2314276695251465, | |
| "learning_rate": 9.212000874196953e-06, | |
| "loss": 0.1051, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.4695652173913043, | |
| "grad_norm": 6.991565704345703, | |
| "learning_rate": 8.662228232019876e-06, | |
| "loss": 0.3733, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.4869565217391303, | |
| "grad_norm": 5.351903915405273, | |
| "learning_rate": 8.127810884536403e-06, | |
| "loss": 0.0471, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.5043478260869563, | |
| "grad_norm": 5.928205966949463, | |
| "learning_rate": 7.60894734597476e-06, | |
| "loss": 2.3398, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.5217391304347827, | |
| "grad_norm": 8.773082733154297, | |
| "learning_rate": 7.105830352958142e-06, | |
| "loss": 1.3465, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.5391304347826087, | |
| "grad_norm": 5.912259578704834, | |
| "learning_rate": 6.618646792910893e-06, | |
| "loss": 0.8377, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.5565217391304347, | |
| "grad_norm": 5.538220405578613, | |
| "learning_rate": 6.147577634637414e-06, | |
| "loss": 0.4977, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.573913043478261, | |
| "grad_norm": 6.276126861572266, | |
| "learning_rate": 5.692797861099719e-06, | |
| "loss": 0.4744, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.591304347826087, | |
| "grad_norm": 5.5036396980285645, | |
| "learning_rate": 5.25447640441834e-06, | |
| "loss": 0.4011, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.608695652173913, | |
| "grad_norm": 4.149447441101074, | |
| "learning_rate": 4.832776083120982e-06, | |
| "loss": 0.22, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.608695652173913, | |
| "eval_loss": 0.5425323247909546, | |
| "eval_runtime": 9.2404, | |
| "eval_samples_per_second": 10.497, | |
| "eval_steps_per_second": 2.706, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.626086956521739, | |
| "grad_norm": 5.254273414611816, | |
| "learning_rate": 4.427853541662091e-06, | |
| "loss": 0.425, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.643478260869565, | |
| "grad_norm": 10.881125450134277, | |
| "learning_rate": 4.039859192235779e-06, | |
| "loss": 0.6446, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.660869565217391, | |
| "grad_norm": 8.573467254638672, | |
| "learning_rate": 3.668937158903901e-06, | |
| "loss": 0.1585, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.6782608695652175, | |
| "grad_norm": 8.719847679138184, | |
| "learning_rate": 3.315225224059809e-06, | |
| "loss": 0.3111, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.6956521739130435, | |
| "grad_norm": 4.025722026824951, | |
| "learning_rate": 2.9788547772478416e-06, | |
| "loss": 0.3628, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.7130434782608694, | |
| "grad_norm": 12.714775085449219, | |
| "learning_rate": 2.6599507663574384e-06, | |
| "loss": 0.2022, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.730434782608696, | |
| "grad_norm": 17.537216186523438, | |
| "learning_rate": 2.3586316512101416e-06, | |
| "loss": 0.1643, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.747826086956522, | |
| "grad_norm": 6.424210071563721, | |
| "learning_rate": 2.0750093595565733e-06, | |
| "loss": 2.195, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.765217391304348, | |
| "grad_norm": 6.1707048416137695, | |
| "learning_rate": 1.8091892454998594e-06, | |
| "loss": 0.915, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 9.390337944030762, | |
| "learning_rate": 1.5612700503608968e-06, | |
| "loss": 0.8495, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 4.434484958648682, | |
| "learning_rate": 1.33134386599994e-06, | |
| "loss": 0.3239, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.8173913043478263, | |
| "grad_norm": 6.292799472808838, | |
| "learning_rate": 1.1194961006082972e-06, | |
| "loss": 0.5492, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.8347826086956522, | |
| "grad_norm": 4.593890190124512, | |
| "learning_rate": 9.258054469825972e-07, | |
| "loss": 0.3086, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.8521739130434782, | |
| "grad_norm": 5.500854969024658, | |
| "learning_rate": 7.503438532937168e-07, | |
| "loss": 0.2752, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.869565217391304, | |
| "grad_norm": 9.617355346679688, | |
| "learning_rate": 5.931764963608866e-07, | |
| "loss": 0.3153, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.8869565217391306, | |
| "grad_norm": 7.330923080444336, | |
| "learning_rate": 4.543617574412184e-07, | |
| "loss": 0.6061, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.9043478260869566, | |
| "grad_norm": 4.938751220703125, | |
| "learning_rate": 3.339512005434309e-07, | |
| "loss": 0.2001, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.9217391304347826, | |
| "grad_norm": 9.237695693969727, | |
| "learning_rate": 2.319895532739369e-07, | |
| "loss": 0.1176, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.9391304347826086, | |
| "grad_norm": 13.457283973693848, | |
| "learning_rate": 1.4851469022234e-07, | |
| "loss": 0.2774, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "grad_norm": 2.380657911300659, | |
| "learning_rate": 8.355761889260461e-08, | |
| "loss": 0.042, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.973913043478261, | |
| "grad_norm": 12.363837242126465, | |
| "learning_rate": 3.7142468185014104e-08, | |
| "loss": 0.1025, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.991304347826087, | |
| "grad_norm": 5.269643783569336, | |
| "learning_rate": 9.286479433257e-09, | |
| "loss": 1.1058, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 3.008695652173913, | |
| "grad_norm": 4.399047374725342, | |
| "learning_rate": 0.0, | |
| "loss": 1.715, | |
| "step": 173 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 173, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.5064824501174272e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |