| { | |
| "best_metric": 2.4741508960723877, | |
| "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": 68.77892303466797, | |
| "learning_rate": 1e-05, | |
| "loss": 12.5759, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0006858710562414266, | |
| "eval_loss": 4.443713188171387, | |
| "eval_runtime": 27.8312, | |
| "eval_samples_per_second": 88.246, | |
| "eval_steps_per_second": 22.062, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0013717421124828531, | |
| "grad_norm": 67.28659057617188, | |
| "learning_rate": 2e-05, | |
| "loss": 13.0793, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00205761316872428, | |
| "grad_norm": 65.68877410888672, | |
| "learning_rate": 3e-05, | |
| "loss": 13.3221, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0027434842249657062, | |
| "grad_norm": 70.37772369384766, | |
| "learning_rate": 4e-05, | |
| "loss": 13.4539, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.003429355281207133, | |
| "grad_norm": 66.16577911376953, | |
| "learning_rate": 5e-05, | |
| "loss": 13.1325, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00411522633744856, | |
| "grad_norm": 62.75472640991211, | |
| "learning_rate": 6e-05, | |
| "loss": 12.2998, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.004801097393689987, | |
| "grad_norm": 63.022159576416016, | |
| "learning_rate": 7e-05, | |
| "loss": 12.9368, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0054869684499314125, | |
| "grad_norm": 64.91657257080078, | |
| "learning_rate": 8e-05, | |
| "loss": 13.1737, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.006172839506172839, | |
| "grad_norm": 59.96748733520508, | |
| "learning_rate": 9e-05, | |
| "loss": 12.9906, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.006858710562414266, | |
| "grad_norm": 54.619380950927734, | |
| "learning_rate": 0.0001, | |
| "loss": 12.1756, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0075445816186556925, | |
| "grad_norm": 63.49129867553711, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 12.136, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.00823045267489712, | |
| "grad_norm": 62.52385330200195, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 12.7785, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.008916323731138546, | |
| "grad_norm": 56.40373229980469, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 12.079, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.009602194787379973, | |
| "grad_norm": 48.15741729736328, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 12.459, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0102880658436214, | |
| "grad_norm": 44.66158676147461, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 12.1763, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010973936899862825, | |
| "grad_norm": 42.26532745361328, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 12.1446, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.011659807956104253, | |
| "grad_norm": 41.23551559448242, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 11.748, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.012345679012345678, | |
| "grad_norm": 47.38997268676758, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 11.9482, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.013031550068587106, | |
| "grad_norm": 47.41958236694336, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 12.2045, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.013717421124828532, | |
| "grad_norm": 57.07180404663086, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 12.6398, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01440329218106996, | |
| "grad_norm": 56.35076141357422, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 11.7696, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.015089163237311385, | |
| "grad_norm": 60.23398971557617, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 12.0006, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01577503429355281, | |
| "grad_norm": 63.09962463378906, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 12.0737, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01646090534979424, | |
| "grad_norm": 75.48200225830078, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 12.5995, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.017146776406035666, | |
| "grad_norm": 85.23303985595703, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 11.8122, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01783264746227709, | |
| "grad_norm": 71.87600708007812, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 11.928, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.018518518518518517, | |
| "grad_norm": 74.48838806152344, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 11.6877, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.019204389574759947, | |
| "grad_norm": 81.24354553222656, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 11.6868, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.019890260631001373, | |
| "grad_norm": 72.92513275146484, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 12.1393, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0205761316872428, | |
| "grad_norm": 92.28253936767578, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 11.7482, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.021262002743484224, | |
| "grad_norm": 86.0582275390625, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 11.6458, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.02194787379972565, | |
| "grad_norm": 84.4745101928711, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 11.2196, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.02263374485596708, | |
| "grad_norm": 96.3499755859375, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 11.9751, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.023319615912208505, | |
| "grad_norm": 83.6324234008789, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 11.439, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.02400548696844993, | |
| "grad_norm": 105.47029876708984, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 11.4096, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.024691358024691357, | |
| "grad_norm": 99.468017578125, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 11.7002, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.025377229080932786, | |
| "grad_norm": 131.2845916748047, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 12.8492, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.02606310013717421, | |
| "grad_norm": 124.80172729492188, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 11.7998, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.026748971193415638, | |
| "grad_norm": 134.95184326171875, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 12.1715, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.027434842249657063, | |
| "grad_norm": 153.27940368652344, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 12.7271, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.028120713305898493, | |
| "grad_norm": 129.91116333007812, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 11.3534, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.02880658436213992, | |
| "grad_norm": 128.22120666503906, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 11.2522, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.029492455418381344, | |
| "grad_norm": 111.56830596923828, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 11.8585, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.03017832647462277, | |
| "grad_norm": 104.10694122314453, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 12.3499, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.030864197530864196, | |
| "grad_norm": 131.64381408691406, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 12.0095, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03155006858710562, | |
| "grad_norm": 118.84125518798828, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 10.7094, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.03223593964334705, | |
| "grad_norm": 142.47061157226562, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 13.2932, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03292181069958848, | |
| "grad_norm": 137.06394958496094, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 12.208, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.033607681755829906, | |
| "grad_norm": 127.30778503417969, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 12.4223, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03429355281207133, | |
| "grad_norm": 143.66258239746094, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 15.7109, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03429355281207133, | |
| "eval_loss": 2.8824493885040283, | |
| "eval_runtime": 27.8541, | |
| "eval_samples_per_second": 88.174, | |
| "eval_steps_per_second": 22.043, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03497942386831276, | |
| "grad_norm": 92.24408721923828, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 11.4115, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.03566529492455418, | |
| "grad_norm": 81.28035736083984, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 11.2028, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.03635116598079561, | |
| "grad_norm": 89.4981918334961, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 11.488, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.037037037037037035, | |
| "grad_norm": 78.65625, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 11.4141, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.03772290809327846, | |
| "grad_norm": 76.7930679321289, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 10.9952, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.038408779149519894, | |
| "grad_norm": 78.28646087646484, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 10.7624, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.03909465020576132, | |
| "grad_norm": 83.59944915771484, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 10.3187, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.039780521262002745, | |
| "grad_norm": 78.39309692382812, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 10.8053, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.04046639231824417, | |
| "grad_norm": 58.5151252746582, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 11.2423, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0411522633744856, | |
| "grad_norm": 65.87265014648438, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 10.3705, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04183813443072702, | |
| "grad_norm": 61.53200149536133, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 10.6142, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.04252400548696845, | |
| "grad_norm": 81.54600524902344, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 10.7285, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.043209876543209874, | |
| "grad_norm": 69.75658416748047, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 11.0367, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.0438957475994513, | |
| "grad_norm": 66.0276870727539, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 11.383, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.04458161865569273, | |
| "grad_norm": 72.12008666992188, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 10.4332, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04526748971193416, | |
| "grad_norm": 83.25178527832031, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 10.285, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.045953360768175584, | |
| "grad_norm": 64.71113586425781, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 10.1593, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.04663923182441701, | |
| "grad_norm": 76.7099609375, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 10.4456, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.047325102880658436, | |
| "grad_norm": 73.85135650634766, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 10.3244, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.04801097393689986, | |
| "grad_norm": 76.21237182617188, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 10.5153, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04869684499314129, | |
| "grad_norm": 61.273319244384766, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 10.2175, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.04938271604938271, | |
| "grad_norm": 60.808902740478516, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 10.7558, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.050068587105624146, | |
| "grad_norm": 62.539634704589844, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 10.6394, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.05075445816186557, | |
| "grad_norm": 102.80731201171875, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 10.2788, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.051440329218107, | |
| "grad_norm": 63.74735641479492, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 11.0018, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.05212620027434842, | |
| "grad_norm": 69.26323699951172, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 10.2744, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.05281207133058985, | |
| "grad_norm": 76.43173217773438, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 10.0894, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.053497942386831275, | |
| "grad_norm": 56.820091247558594, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 10.8731, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0541838134430727, | |
| "grad_norm": 74.794677734375, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 10.7842, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.05486968449931413, | |
| "grad_norm": 69.85340118408203, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 10.3337, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 73.60613250732422, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 10.0417, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.056241426611796985, | |
| "grad_norm": 74.16600799560547, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 11.3293, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.05692729766803841, | |
| "grad_norm": 82.26252746582031, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 10.5422, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.05761316872427984, | |
| "grad_norm": 90.6766357421875, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 11.4532, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.05829903978052126, | |
| "grad_norm": 64.48725128173828, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 10.5487, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.05898491083676269, | |
| "grad_norm": 63.53636169433594, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 11.0885, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.059670781893004114, | |
| "grad_norm": 147.17185974121094, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 9.76, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.06035665294924554, | |
| "grad_norm": 285.7807922363281, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 10.1179, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.061042524005486966, | |
| "grad_norm": 75.27552795410156, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 10.9086, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.06172839506172839, | |
| "grad_norm": 64.7357177734375, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 9.9119, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.062414266117969824, | |
| "grad_norm": 80.8088607788086, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 11.0896, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.06310013717421124, | |
| "grad_norm": 60.98735809326172, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 9.0816, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.06378600823045268, | |
| "grad_norm": 59.03153610229492, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 9.6448, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.0644718792866941, | |
| "grad_norm": 70.60260009765625, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 10.3787, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.06515775034293553, | |
| "grad_norm": 86.890625, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 10.5721, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06584362139917696, | |
| "grad_norm": 81.75714874267578, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 10.3192, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.06652949245541838, | |
| "grad_norm": 73.4039535522461, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 11.0742, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.06721536351165981, | |
| "grad_norm": 80.26651000976562, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 10.8634, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.06790123456790123, | |
| "grad_norm": 81.6557388305664, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 11.6236, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.06858710562414266, | |
| "grad_norm": 113.58912658691406, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 13.2203, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06858710562414266, | |
| "eval_loss": 2.6562275886535645, | |
| "eval_runtime": 27.7794, | |
| "eval_samples_per_second": 88.411, | |
| "eval_steps_per_second": 22.103, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06927297668038408, | |
| "grad_norm": 64.52449035644531, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 10.5687, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.06995884773662552, | |
| "grad_norm": 69.00031280517578, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 10.0724, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.07064471879286695, | |
| "grad_norm": 69.12140655517578, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 10.4581, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.07133058984910837, | |
| "grad_norm": 72.2207260131836, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 10.3929, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0720164609053498, | |
| "grad_norm": 86.40310668945312, | |
| "learning_rate": 5e-05, | |
| "loss": 10.9251, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.07270233196159122, | |
| "grad_norm": 69.9828109741211, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 10.5359, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.07338820301783265, | |
| "grad_norm": 85.34939575195312, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 9.9701, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 62.65793991088867, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 10.7353, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.0747599451303155, | |
| "grad_norm": 76.79057312011719, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 10.5963, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.07544581618655692, | |
| "grad_norm": 73.16259765625, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 10.858, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07613168724279835, | |
| "grad_norm": 75.42756652832031, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 10.1243, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.07681755829903979, | |
| "grad_norm": 77.57795715332031, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 10.494, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0775034293552812, | |
| "grad_norm": 62.20941162109375, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 10.4999, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.07818930041152264, | |
| "grad_norm": 73.04386901855469, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 10.9148, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.07887517146776406, | |
| "grad_norm": 58.56178283691406, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 10.0612, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.07956104252400549, | |
| "grad_norm": 60.28337097167969, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 10.1784, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.08024691358024691, | |
| "grad_norm": 51.978389739990234, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 10.5118, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.08093278463648834, | |
| "grad_norm": 75.6663818359375, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 9.9865, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.08161865569272976, | |
| "grad_norm": 62.4959716796875, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 10.2206, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.0823045267489712, | |
| "grad_norm": 54.797340393066406, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 9.6301, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.08299039780521263, | |
| "grad_norm": 51.61946105957031, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 9.7148, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.08367626886145405, | |
| "grad_norm": 58.46641540527344, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 10.7766, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.08436213991769548, | |
| "grad_norm": 57.30087661743164, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 10.1623, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.0850480109739369, | |
| "grad_norm": 62.32304000854492, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 10.5059, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.08573388203017833, | |
| "grad_norm": 53.862274169921875, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 10.3454, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08641975308641975, | |
| "grad_norm": 60.29814147949219, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 10.0578, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.08710562414266118, | |
| "grad_norm": 60.392398834228516, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 10.0374, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.0877914951989026, | |
| "grad_norm": 60.30309295654297, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 9.8075, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.08847736625514403, | |
| "grad_norm": 56.85651397705078, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 9.8013, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.08916323731138547, | |
| "grad_norm": 77.3592758178711, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 9.2691, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08984910836762688, | |
| "grad_norm": 95.60528564453125, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 10.838, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.09053497942386832, | |
| "grad_norm": 94.47435760498047, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 10.1885, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.09122085048010974, | |
| "grad_norm": 64.9505615234375, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 10.884, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.09190672153635117, | |
| "grad_norm": 80.296142578125, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 10.9472, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.09259259259259259, | |
| "grad_norm": 52.41393280029297, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 10.3691, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.09327846364883402, | |
| "grad_norm": 84.23002624511719, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 9.7432, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.09396433470507544, | |
| "grad_norm": 66.16171264648438, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 9.6973, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.09465020576131687, | |
| "grad_norm": 74.81575775146484, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 11.1423, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.0953360768175583, | |
| "grad_norm": 68.748291015625, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 10.114, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.09602194787379972, | |
| "grad_norm": 65.26239776611328, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 10.8377, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09670781893004116, | |
| "grad_norm": 93.65863037109375, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 9.3308, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.09739368998628258, | |
| "grad_norm": 74.83582305908203, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 8.9685, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.09807956104252401, | |
| "grad_norm": 67.77286529541016, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 9.817, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.09876543209876543, | |
| "grad_norm": 70.82134246826172, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 9.5064, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.09945130315500686, | |
| "grad_norm": 62.630897521972656, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 9.0602, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.10013717421124829, | |
| "grad_norm": 71.15071105957031, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 10.8106, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.10082304526748971, | |
| "grad_norm": 94.39356994628906, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 10.6236, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.10150891632373114, | |
| "grad_norm": 60.53413009643555, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 10.0351, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.10219478737997256, | |
| "grad_norm": 67.53120422363281, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 9.5114, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.102880658436214, | |
| "grad_norm": 69.27748107910156, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 10.9896, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.102880658436214, | |
| "eval_loss": 2.4958090782165527, | |
| "eval_runtime": 27.723, | |
| "eval_samples_per_second": 88.591, | |
| "eval_steps_per_second": 22.148, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.10356652949245541, | |
| "grad_norm": 60.770389556884766, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 10.2229, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.10425240054869685, | |
| "grad_norm": 56.86615753173828, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 10.2429, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.10493827160493827, | |
| "grad_norm": 57.632320404052734, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 9.9653, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.1056241426611797, | |
| "grad_norm": 47.849491119384766, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 9.8793, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.10631001371742113, | |
| "grad_norm": 57.04465866088867, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 10.0272, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.10699588477366255, | |
| "grad_norm": 58.19630432128906, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 10.0255, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.10768175582990398, | |
| "grad_norm": 66.34407806396484, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 10.0586, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.1083676268861454, | |
| "grad_norm": 52.55531692504883, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 9.8339, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.10905349794238683, | |
| "grad_norm": 60.57553482055664, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 9.8565, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.10973936899862825, | |
| "grad_norm": 66.41476440429688, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 9.8901, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.11042524005486969, | |
| "grad_norm": 54.474273681640625, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 9.4769, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 46.790931701660156, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 9.4866, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.11179698216735254, | |
| "grad_norm": 51.67052459716797, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 9.8184, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.11248285322359397, | |
| "grad_norm": 55.770442962646484, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 10.4347, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.11316872427983539, | |
| "grad_norm": 57.43124771118164, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 10.4642, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.11385459533607682, | |
| "grad_norm": 49.7734375, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 10.0576, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.11454046639231824, | |
| "grad_norm": 53.70954132080078, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 10.336, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.11522633744855967, | |
| "grad_norm": 56.110992431640625, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 10.3017, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.11591220850480109, | |
| "grad_norm": 63.508262634277344, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 10.2817, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.11659807956104253, | |
| "grad_norm": 69.57492065429688, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 9.3114, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.11728395061728394, | |
| "grad_norm": 67.96019744873047, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 10.3836, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.11796982167352538, | |
| "grad_norm": 52.86520767211914, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 9.8846, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.11865569272976681, | |
| "grad_norm": 59.025146484375, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 9.5667, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.11934156378600823, | |
| "grad_norm": 78.1929702758789, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 10.2308, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.12002743484224966, | |
| "grad_norm": 53.32894515991211, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 9.7213, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.12071330589849108, | |
| "grad_norm": 53.155670166015625, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 9.7812, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.12139917695473251, | |
| "grad_norm": 59.22550582885742, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 10.5357, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.12208504801097393, | |
| "grad_norm": 59.691349029541016, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 10.2042, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.12277091906721536, | |
| "grad_norm": 58.43455123901367, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 9.2911, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 58.271827697753906, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 9.9495, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.12414266117969822, | |
| "grad_norm": 53.44316482543945, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 10.7796, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.12482853223593965, | |
| "grad_norm": 56.951908111572266, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 10.3049, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.12551440329218108, | |
| "grad_norm": 58.42102813720703, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 10.6577, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.1262002743484225, | |
| "grad_norm": 86.28426361083984, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 9.4439, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.12688614540466392, | |
| "grad_norm": 88.20686340332031, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 10.3754, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.12757201646090535, | |
| "grad_norm": 52.24477767944336, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 9.852, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.12825788751714678, | |
| "grad_norm": 77.35710144042969, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 9.7742, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.1289437585733882, | |
| "grad_norm": 61.559635162353516, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 9.6451, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.12962962962962962, | |
| "grad_norm": 57.520263671875, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 9.5577, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.13031550068587106, | |
| "grad_norm": 66.88455963134766, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 9.2773, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1310013717421125, | |
| "grad_norm": 65.86371612548828, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 9.7137, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.13168724279835392, | |
| "grad_norm": 226.6407928466797, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 9.7101, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.13237311385459533, | |
| "grad_norm": 64.6930923461914, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 9.4937, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.13305898491083676, | |
| "grad_norm": 54.060001373291016, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 9.7317, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.1337448559670782, | |
| "grad_norm": 76.30892181396484, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 9.3322, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.13443072702331962, | |
| "grad_norm": 131.05532836914062, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 8.8468, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.13511659807956103, | |
| "grad_norm": 58.97761535644531, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 9.949, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.13580246913580246, | |
| "grad_norm": 67.570068359375, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 9.4837, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.1364883401920439, | |
| "grad_norm": 59.948848724365234, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 10.3008, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.13717421124828533, | |
| "grad_norm": 76.34727478027344, | |
| "learning_rate": 0.0, | |
| "loss": 15.0173, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13717421124828533, | |
| "eval_loss": 2.4741508960723877, | |
| "eval_runtime": 27.7797, | |
| "eval_samples_per_second": 88.41, | |
| "eval_steps_per_second": 22.102, | |
| "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 | |
| } | |