| { | |
| "best_metric": 2.4264190196990967, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.13717421124828533, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006858710562414266, | |
| "grad_norm": 67.56613159179688, | |
| "learning_rate": 1e-05, | |
| "loss": 12.5759, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0006858710562414266, | |
| "eval_loss": 4.443713188171387, | |
| "eval_runtime": 27.9177, | |
| "eval_samples_per_second": 87.973, | |
| "eval_steps_per_second": 21.993, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0013717421124828531, | |
| "grad_norm": 67.73968505859375, | |
| "learning_rate": 2e-05, | |
| "loss": 13.0793, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00205761316872428, | |
| "grad_norm": 69.86872100830078, | |
| "learning_rate": 3e-05, | |
| "loss": 13.3411, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0027434842249657062, | |
| "grad_norm": 71.4351577758789, | |
| "learning_rate": 4e-05, | |
| "loss": 13.443, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.003429355281207133, | |
| "grad_norm": 67.84893035888672, | |
| "learning_rate": 5e-05, | |
| "loss": 13.1132, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00411522633744856, | |
| "grad_norm": 70.73761749267578, | |
| "learning_rate": 6e-05, | |
| "loss": 12.334, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.004801097393689987, | |
| "grad_norm": 80.68836975097656, | |
| "learning_rate": 7e-05, | |
| "loss": 12.9704, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0054869684499314125, | |
| "grad_norm": 67.63671112060547, | |
| "learning_rate": 8e-05, | |
| "loss": 13.2997, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.006172839506172839, | |
| "grad_norm": 73.50115966796875, | |
| "learning_rate": 9e-05, | |
| "loss": 13.0838, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.006858710562414266, | |
| "grad_norm": 63.94585037231445, | |
| "learning_rate": 0.0001, | |
| "loss": 12.2977, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0075445816186556925, | |
| "grad_norm": 69.82437133789062, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 12.2031, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.00823045267489712, | |
| "grad_norm": 79.25262451171875, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 12.9035, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.008916323731138546, | |
| "grad_norm": 90.1045913696289, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 12.1417, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.009602194787379973, | |
| "grad_norm": 60.609615325927734, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 12.5545, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0102880658436214, | |
| "grad_norm": 81.66038513183594, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 12.3919, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010973936899862825, | |
| "grad_norm": 77.67892456054688, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 12.3826, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.011659807956104253, | |
| "grad_norm": 66.22179412841797, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 12.0993, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.012345679012345678, | |
| "grad_norm": 79.32089233398438, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 12.3574, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.013031550068587106, | |
| "grad_norm": 73.30381774902344, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 12.669, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.013717421124828532, | |
| "grad_norm": 84.61729431152344, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 12.7761, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01440329218106996, | |
| "grad_norm": 76.04741668701172, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 12.0316, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.015089163237311385, | |
| "grad_norm": 77.5232162475586, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 12.0874, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01577503429355281, | |
| "grad_norm": 89.0643310546875, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 12.4647, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01646090534979424, | |
| "grad_norm": 69.02548217773438, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 12.9106, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.017146776406035666, | |
| "grad_norm": 76.06600952148438, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 12.1283, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01783264746227709, | |
| "grad_norm": 67.31135559082031, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 12.0358, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.018518518518518517, | |
| "grad_norm": 87.58185577392578, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 11.9628, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.019204389574759947, | |
| "grad_norm": 72.38446807861328, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 12.231, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.019890260631001373, | |
| "grad_norm": 110.03680419921875, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 12.7041, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0205761316872428, | |
| "grad_norm": 96.28618621826172, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 12.5188, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.021262002743484224, | |
| "grad_norm": 111.32992553710938, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 12.266, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.02194787379972565, | |
| "grad_norm": 129.9169158935547, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 12.1162, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.02263374485596708, | |
| "grad_norm": 196.08529663085938, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 13.319, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.023319615912208505, | |
| "grad_norm": 148.8671417236328, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 13.3208, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.02400548696844993, | |
| "grad_norm": 247.100341796875, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 12.8171, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.024691358024691357, | |
| "grad_norm": 192.64491271972656, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 12.9477, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.025377229080932786, | |
| "grad_norm": 167.75289916992188, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 14.3512, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.02606310013717421, | |
| "grad_norm": 160.7109375, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 13.2111, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.026748971193415638, | |
| "grad_norm": 193.69012451171875, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 13.3398, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.027434842249657063, | |
| "grad_norm": 251.85670471191406, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 13.9219, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.028120713305898493, | |
| "grad_norm": 166.14584350585938, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 12.5066, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.02880658436213992, | |
| "grad_norm": 149.8164825439453, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 12.4201, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.029492455418381344, | |
| "grad_norm": 180.7564239501953, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 12.7494, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.03017832647462277, | |
| "grad_norm": 178.08023071289062, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 13.5972, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.030864197530864196, | |
| "grad_norm": 186.38841247558594, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 12.9432, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03155006858710562, | |
| "grad_norm": 181.45640563964844, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 11.2809, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.03223593964334705, | |
| "grad_norm": 189.8750762939453, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 14.0784, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03292181069958848, | |
| "grad_norm": 137.40380859375, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 12.4905, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.033607681755829906, | |
| "grad_norm": 263.31329345703125, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 13.5553, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03429355281207133, | |
| "grad_norm": 239.6564178466797, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 15.5905, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03429355281207133, | |
| "eval_loss": 3.1247825622558594, | |
| "eval_runtime": 27.9052, | |
| "eval_samples_per_second": 88.012, | |
| "eval_steps_per_second": 22.003, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03497942386831276, | |
| "grad_norm": 129.30958557128906, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 12.4781, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.03566529492455418, | |
| "grad_norm": 137.0768280029297, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 12.0811, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.03635116598079561, | |
| "grad_norm": 113.71504974365234, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 12.2344, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.037037037037037035, | |
| "grad_norm": 104.11981201171875, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 11.7471, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.03772290809327846, | |
| "grad_norm": 93.32283020019531, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 11.5967, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.038408779149519894, | |
| "grad_norm": 109.50666809082031, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 11.3471, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.03909465020576132, | |
| "grad_norm": 89.5738525390625, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 10.8263, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.039780521262002745, | |
| "grad_norm": 88.97535705566406, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 11.192, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.04046639231824417, | |
| "grad_norm": 86.61604309082031, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 11.6289, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0411522633744856, | |
| "grad_norm": 81.41461944580078, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 10.9449, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04183813443072702, | |
| "grad_norm": 77.4903335571289, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 10.9307, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.04252400548696845, | |
| "grad_norm": 69.21846771240234, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 10.9383, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.043209876543209874, | |
| "grad_norm": 100.75662994384766, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 11.1941, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.0438957475994513, | |
| "grad_norm": 69.18630981445312, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 11.5315, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.04458161865569273, | |
| "grad_norm": 66.34451293945312, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 10.4314, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04526748971193416, | |
| "grad_norm": 66.86237335205078, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 10.4658, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.045953360768175584, | |
| "grad_norm": 83.72396087646484, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 10.0248, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.04663923182441701, | |
| "grad_norm": 72.43347930908203, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 10.5432, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.047325102880658436, | |
| "grad_norm": 60.512451171875, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 10.3905, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.04801097393689986, | |
| "grad_norm": 56.54389190673828, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 10.4991, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04869684499314129, | |
| "grad_norm": 49.95751953125, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 10.473, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.04938271604938271, | |
| "grad_norm": 44.739166259765625, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 10.8978, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.050068587105624146, | |
| "grad_norm": 51.43025207519531, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 10.7114, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.05075445816186557, | |
| "grad_norm": 52.89293670654297, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 10.7065, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.051440329218107, | |
| "grad_norm": 50.14055252075195, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 11.2078, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.05212620027434842, | |
| "grad_norm": 41.636505126953125, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 10.3809, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.05281207133058985, | |
| "grad_norm": 53.60658645629883, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 10.0238, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.053497942386831275, | |
| "grad_norm": 51.80974197387695, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 11.0681, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0541838134430727, | |
| "grad_norm": 52.04350280761719, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 11.0406, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.05486968449931413, | |
| "grad_norm": 60.315467834472656, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 10.433, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 50.76766586303711, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 10.315, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.056241426611796985, | |
| "grad_norm": 65.96510314941406, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 11.7765, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.05692729766803841, | |
| "grad_norm": 77.0702896118164, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 10.7374, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.05761316872427984, | |
| "grad_norm": 84.89005279541016, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 11.6438, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.05829903978052126, | |
| "grad_norm": 69.22117614746094, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 10.9764, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.05898491083676269, | |
| "grad_norm": 77.36370086669922, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 11.4199, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.059670781893004114, | |
| "grad_norm": 68.88736724853516, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 10.1972, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.06035665294924554, | |
| "grad_norm": 69.5047378540039, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 10.2018, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.061042524005486966, | |
| "grad_norm": 68.36778259277344, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 11.2361, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.06172839506172839, | |
| "grad_norm": 78.47174835205078, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 10.0716, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.062414266117969824, | |
| "grad_norm": 69.16641998291016, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 11.3155, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.06310013717421124, | |
| "grad_norm": 98.92090606689453, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 9.1467, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.06378600823045268, | |
| "grad_norm": 63.03489303588867, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 9.7435, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.0644718792866941, | |
| "grad_norm": 73.33062744140625, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 10.6939, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.06515775034293553, | |
| "grad_norm": 60.80948257446289, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 10.4996, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06584362139917696, | |
| "grad_norm": 56.09263610839844, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 10.5311, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.06652949245541838, | |
| "grad_norm": 71.61680603027344, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 11.2077, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.06721536351165981, | |
| "grad_norm": 73.19332122802734, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 11.0189, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.06790123456790123, | |
| "grad_norm": 71.53459167480469, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 11.906, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.06858710562414266, | |
| "grad_norm": 78.68098449707031, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 13.9879, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06858710562414266, | |
| "eval_loss": 2.6943976879119873, | |
| "eval_runtime": 27.9516, | |
| "eval_samples_per_second": 87.866, | |
| "eval_steps_per_second": 21.967, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06927297668038408, | |
| "grad_norm": 46.52141189575195, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 10.6047, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.06995884773662552, | |
| "grad_norm": 46.65863800048828, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 10.15, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.07064471879286695, | |
| "grad_norm": 45.21689224243164, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 10.6968, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.07133058984910837, | |
| "grad_norm": 49.33715057373047, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 10.4853, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0720164609053498, | |
| "grad_norm": 59.95785903930664, | |
| "learning_rate": 5e-05, | |
| "loss": 10.8698, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.07270233196159122, | |
| "grad_norm": 48.05885696411133, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 10.3103, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.07338820301783265, | |
| "grad_norm": 49.50995635986328, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 10.0637, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 44.41816329956055, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 10.6463, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.0747599451303155, | |
| "grad_norm": 40.42985916137695, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 10.5458, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.07544581618655692, | |
| "grad_norm": 46.90631866455078, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 10.553, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07613168724279835, | |
| "grad_norm": 53.724761962890625, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 10.0551, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.07681755829903979, | |
| "grad_norm": 40.973236083984375, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 10.3622, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0775034293552812, | |
| "grad_norm": 37.83047103881836, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 10.2278, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.07818930041152264, | |
| "grad_norm": 36.523193359375, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 10.5923, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.07887517146776406, | |
| "grad_norm": 30.8020076751709, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 9.8918, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.07956104252400549, | |
| "grad_norm": 34.273311614990234, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 10.0503, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.08024691358024691, | |
| "grad_norm": 36.07250213623047, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 10.3518, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.08093278463648834, | |
| "grad_norm": 33.813743591308594, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 9.7466, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.08161865569272976, | |
| "grad_norm": 39.58462142944336, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 10.0692, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.0823045267489712, | |
| "grad_norm": 30.966215133666992, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 9.5037, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.08299039780521263, | |
| "grad_norm": 34.97091293334961, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 9.5057, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.08367626886145405, | |
| "grad_norm": 31.762781143188477, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 10.5041, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.08436213991769548, | |
| "grad_norm": 32.93566131591797, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 9.9268, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.0850480109739369, | |
| "grad_norm": 34.30867004394531, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 10.408, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.08573388203017833, | |
| "grad_norm": 44.63151931762695, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 10.3327, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08641975308641975, | |
| "grad_norm": 35.688411712646484, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 9.7768, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.08710562414266118, | |
| "grad_norm": 40.22312927246094, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 9.7416, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.0877914951989026, | |
| "grad_norm": 31.907785415649414, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 9.6265, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.08847736625514403, | |
| "grad_norm": 40.40849304199219, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 9.7528, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.08916323731138547, | |
| "grad_norm": 38.003379821777344, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 9.1257, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08984910836762688, | |
| "grad_norm": 37.73146057128906, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 10.7232, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.09053497942386832, | |
| "grad_norm": 57.599876403808594, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 10.0192, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.09122085048010974, | |
| "grad_norm": 38.22669219970703, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 10.4424, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.09190672153635117, | |
| "grad_norm": 35.635154724121094, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 10.8573, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.09259259259259259, | |
| "grad_norm": 38.25225067138672, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 10.2126, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.09327846364883402, | |
| "grad_norm": 35.093833923339844, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 9.477, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.09396433470507544, | |
| "grad_norm": 34.92509078979492, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 9.6712, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.09465020576131687, | |
| "grad_norm": 37.29582977294922, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 11.0698, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.0953360768175583, | |
| "grad_norm": 37.750606536865234, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 9.8634, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.09602194787379972, | |
| "grad_norm": 37.50721740722656, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 10.742, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09670781893004116, | |
| "grad_norm": 37.4149284362793, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 9.287, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.09739368998628258, | |
| "grad_norm": 50.917484283447266, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 8.6149, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.09807956104252401, | |
| "grad_norm": 40.897979736328125, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 9.7073, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.09876543209876543, | |
| "grad_norm": 47.89082336425781, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 9.6094, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.09945130315500686, | |
| "grad_norm": 44.85136032104492, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 8.7717, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.10013717421124829, | |
| "grad_norm": 44.114994049072266, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 10.6737, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.10082304526748971, | |
| "grad_norm": 55.8568000793457, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 10.3604, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.10150891632373114, | |
| "grad_norm": 48.95426940917969, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 9.817, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.10219478737997256, | |
| "grad_norm": 50.1852912902832, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 9.3906, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.102880658436214, | |
| "grad_norm": 64.58821105957031, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 10.721, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.102880658436214, | |
| "eval_loss": 2.458784818649292, | |
| "eval_runtime": 27.9472, | |
| "eval_samples_per_second": 87.88, | |
| "eval_steps_per_second": 21.97, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.10356652949245541, | |
| "grad_norm": 32.33224868774414, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 10.0477, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.10425240054869685, | |
| "grad_norm": 32.00208282470703, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 10.1126, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.10493827160493827, | |
| "grad_norm": 33.496917724609375, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 9.8477, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.1056241426611797, | |
| "grad_norm": 26.901329040527344, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 9.7178, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.10631001371742113, | |
| "grad_norm": 29.162796020507812, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 9.952, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.10699588477366255, | |
| "grad_norm": 27.379783630371094, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 9.799, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.10768175582990398, | |
| "grad_norm": 32.06669235229492, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 9.94, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.1083676268861454, | |
| "grad_norm": 30.382644653320312, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 9.7466, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.10905349794238683, | |
| "grad_norm": 28.083467483520508, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 9.6022, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.10973936899862825, | |
| "grad_norm": 29.265764236450195, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 9.6124, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.11042524005486969, | |
| "grad_norm": 32.63068389892578, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 9.381, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 26.65264892578125, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 9.3156, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.11179698216735254, | |
| "grad_norm": 29.897174835205078, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 9.6713, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.11248285322359397, | |
| "grad_norm": 27.99970054626465, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 10.3865, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.11316872427983539, | |
| "grad_norm": 28.752399444580078, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 10.3144, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.11385459533607682, | |
| "grad_norm": 28.662458419799805, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 9.7915, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.11454046639231824, | |
| "grad_norm": 35.952301025390625, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 10.1204, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.11522633744855967, | |
| "grad_norm": 29.5347843170166, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 10.0006, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.11591220850480109, | |
| "grad_norm": 28.053680419921875, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 10.083, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.11659807956104253, | |
| "grad_norm": 30.08479881286621, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 9.0967, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.11728395061728394, | |
| "grad_norm": 32.3925666809082, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 10.1549, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.11796982167352538, | |
| "grad_norm": 36.964935302734375, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 9.668, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.11865569272976681, | |
| "grad_norm": 31.115142822265625, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 9.323, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.11934156378600823, | |
| "grad_norm": 34.504268646240234, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 9.951, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.12002743484224966, | |
| "grad_norm": 30.281644821166992, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 9.4932, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.12071330589849108, | |
| "grad_norm": 31.386791229248047, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 9.5956, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.12139917695473251, | |
| "grad_norm": 33.98624038696289, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 10.3343, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.12208504801097393, | |
| "grad_norm": 36.686302185058594, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 10.0007, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.12277091906721536, | |
| "grad_norm": 35.59688949584961, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 9.2315, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 37.250431060791016, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 9.6842, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.12414266117969822, | |
| "grad_norm": 39.598575592041016, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 10.5317, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.12482853223593965, | |
| "grad_norm": 36.71938705444336, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 10.1804, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.12551440329218108, | |
| "grad_norm": 37.504150390625, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 10.5273, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.1262002743484225, | |
| "grad_norm": 40.19125747680664, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 9.2472, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.12688614540466392, | |
| "grad_norm": 39.66238784790039, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 10.1434, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.12757201646090535, | |
| "grad_norm": 39.80047607421875, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 9.7466, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.12825788751714678, | |
| "grad_norm": 44.574928283691406, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 9.4836, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.1289437585733882, | |
| "grad_norm": 47.09717559814453, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 9.2818, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.12962962962962962, | |
| "grad_norm": 41.736751556396484, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 9.337, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.13031550068587106, | |
| "grad_norm": 48.15314865112305, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 9.3095, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1310013717421125, | |
| "grad_norm": 44.57954025268555, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 9.5468, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.13168724279835392, | |
| "grad_norm": 45.98891830444336, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 9.5177, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.13237311385459533, | |
| "grad_norm": 55.68307876586914, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 9.2744, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.13305898491083676, | |
| "grad_norm": 41.510536193847656, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 9.6411, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.1337448559670782, | |
| "grad_norm": 57.00276184082031, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 9.1863, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.13443072702331962, | |
| "grad_norm": 49.09486770629883, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 8.4998, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.13511659807956103, | |
| "grad_norm": 53.23307418823242, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 9.8922, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.13580246913580246, | |
| "grad_norm": 75.66525268554688, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 9.38, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.1364883401920439, | |
| "grad_norm": 59.829307556152344, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 9.8218, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.13717421124828533, | |
| "grad_norm": 67.54158020019531, | |
| "learning_rate": 0.0, | |
| "loss": 15.1747, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13717421124828533, | |
| "eval_loss": 2.4264190196990967, | |
| "eval_runtime": 27.9343, | |
| "eval_samples_per_second": 87.92, | |
| "eval_steps_per_second": 21.98, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.49024359317504e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |