| { | |
| "best_metric": 1.3822078704833984, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.07392804337111877, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0004928536224741252, | |
| "grad_norm": 16.37380599975586, | |
| "learning_rate": 1e-05, | |
| "loss": 6.995, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0004928536224741252, | |
| "eval_loss": 1.744011402130127, | |
| "eval_runtime": 259.1339, | |
| "eval_samples_per_second": 13.186, | |
| "eval_steps_per_second": 3.299, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0009857072449482504, | |
| "grad_norm": 16.33266830444336, | |
| "learning_rate": 2e-05, | |
| "loss": 7.1083, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0014785608674223755, | |
| "grad_norm": 15.080232620239258, | |
| "learning_rate": 3e-05, | |
| "loss": 6.7867, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.001971414489896501, | |
| "grad_norm": 10.568120956420898, | |
| "learning_rate": 4e-05, | |
| "loss": 6.3311, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0024642681123706258, | |
| "grad_norm": 10.395201683044434, | |
| "learning_rate": 5e-05, | |
| "loss": 6.2016, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.002957121734844751, | |
| "grad_norm": 11.448528289794922, | |
| "learning_rate": 6e-05, | |
| "loss": 6.5881, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0034499753573188764, | |
| "grad_norm": 8.285737037658691, | |
| "learning_rate": 7e-05, | |
| "loss": 6.178, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.003942828979793002, | |
| "grad_norm": 6.383825302124023, | |
| "learning_rate": 8e-05, | |
| "loss": 5.9585, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.004435682602267127, | |
| "grad_norm": 7.123228073120117, | |
| "learning_rate": 9e-05, | |
| "loss": 6.3599, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0049285362247412515, | |
| "grad_norm": 6.691042423248291, | |
| "learning_rate": 0.0001, | |
| "loss": 5.9879, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.005421389847215377, | |
| "grad_norm": 6.677425861358643, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 6.0902, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.005914243469689502, | |
| "grad_norm": 6.691093921661377, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 6.13, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.006407097092163627, | |
| "grad_norm": 6.412749290466309, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 5.9985, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.006899950714637753, | |
| "grad_norm": 6.398421287536621, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 6.0254, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.007392804337111878, | |
| "grad_norm": 6.084516525268555, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 6.1664, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.007885657959586003, | |
| "grad_norm": 5.95521354675293, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 5.8991, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.008378511582060127, | |
| "grad_norm": 5.627572536468506, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 6.0787, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.008871365204534253, | |
| "grad_norm": 5.7911248207092285, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 6.2035, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.009364218827008379, | |
| "grad_norm": 6.241641044616699, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 6.0726, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.009857072449482503, | |
| "grad_norm": 5.327273845672607, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 6.2332, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.010349926071956629, | |
| "grad_norm": 5.308168411254883, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 6.0171, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.010842779694430755, | |
| "grad_norm": 5.032196998596191, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 5.8416, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.011335633316904879, | |
| "grad_norm": 4.981048583984375, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 5.5985, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.011828486939379004, | |
| "grad_norm": 5.130867004394531, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 6.1288, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.01232134056185313, | |
| "grad_norm": 4.848458766937256, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 5.8318, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.012814194184327254, | |
| "grad_norm": 5.0001397132873535, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 5.9925, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.01330704780680138, | |
| "grad_norm": 5.232586860656738, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 6.2745, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.013799901429275506, | |
| "grad_norm": 5.037934303283691, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 5.8795, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.01429275505174963, | |
| "grad_norm": 5.175755500793457, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 6.1939, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.014785608674223755, | |
| "grad_norm": 5.293409824371338, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 5.9304, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.015278462296697881, | |
| "grad_norm": 4.89537239074707, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 5.7866, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.015771315919172007, | |
| "grad_norm": 5.026119232177734, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 5.9492, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.016264169541646133, | |
| "grad_norm": 5.800846576690674, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 5.5961, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.016757023164120255, | |
| "grad_norm": 4.967105388641357, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 5.9876, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.01724987678659438, | |
| "grad_norm": 4.971159934997559, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 5.9039, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.017742730409068506, | |
| "grad_norm": 5.209476470947266, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 5.9231, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.018235584031542632, | |
| "grad_norm": 5.320252418518066, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 5.8169, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.018728437654016758, | |
| "grad_norm": 5.2738142013549805, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 6.002, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.019221291276490884, | |
| "grad_norm": 5.119503021240234, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 5.931, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.019714144898965006, | |
| "grad_norm": 5.318321704864502, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 5.9097, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.020206998521439132, | |
| "grad_norm": 5.050459861755371, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 6.1257, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.020699852143913258, | |
| "grad_norm": 5.593949794769287, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 6.2968, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.021192705766387383, | |
| "grad_norm": 5.589847087860107, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 6.233, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.02168555938886151, | |
| "grad_norm": 5.743232727050781, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 5.9828, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.022178413011335635, | |
| "grad_norm": 5.6376848220825195, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 6.1437, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.022671266633809757, | |
| "grad_norm": 11.450542449951172, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 6.2194, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.023164120256283883, | |
| "grad_norm": 6.774994373321533, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 6.1023, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.02365697387875801, | |
| "grad_norm": 5.768123626708984, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 6.2419, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.024149827501232134, | |
| "grad_norm": 7.388202667236328, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 6.5133, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.02464268112370626, | |
| "grad_norm": 7.236172676086426, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 6.2176, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02464268112370626, | |
| "eval_loss": 1.732652187347412, | |
| "eval_runtime": 260.9809, | |
| "eval_samples_per_second": 13.093, | |
| "eval_steps_per_second": 3.276, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.025135534746180386, | |
| "grad_norm": 35.06370544433594, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 7.3535, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.025628388368654508, | |
| "grad_norm": 7.263309001922607, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 6.6987, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.026121241991128634, | |
| "grad_norm": 4.144477844238281, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 5.9729, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.02661409561360276, | |
| "grad_norm": 3.6691770553588867, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 6.0169, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.027106949236076885, | |
| "grad_norm": 3.6410415172576904, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 5.9244, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.02759980285855101, | |
| "grad_norm": 3.8702380657196045, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 6.2104, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.028092656481025137, | |
| "grad_norm": 3.817110776901245, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 5.8548, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.02858551010349926, | |
| "grad_norm": 7.407201290130615, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 5.5015, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.029078363725973385, | |
| "grad_norm": 5.8004984855651855, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 5.5889, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.02957121734844751, | |
| "grad_norm": 4.282701015472412, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 5.8837, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.030064070970921637, | |
| "grad_norm": 4.238832950592041, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 5.9101, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.030556924593395762, | |
| "grad_norm": 4.113567352294922, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 5.9194, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.031049778215869888, | |
| "grad_norm": 4.071282386779785, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 5.705, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.031542631838344014, | |
| "grad_norm": 3.9348578453063965, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 5.9225, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.032035485460818136, | |
| "grad_norm": 6.435613632202148, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 5.9364, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.032528339083292265, | |
| "grad_norm": 5.112850666046143, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 5.9065, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.03302119270576639, | |
| "grad_norm": 4.052131175994873, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 5.6007, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.03351404632824051, | |
| "grad_norm": 3.9013149738311768, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 5.6739, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.03400689995071464, | |
| "grad_norm": 4.061436653137207, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 5.5508, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.03449975357318876, | |
| "grad_norm": 4.149789333343506, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 5.8792, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03499260719566289, | |
| "grad_norm": 3.755802869796753, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 5.2883, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.03548546081813701, | |
| "grad_norm": 3.9959685802459717, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 5.8289, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.035978314440611135, | |
| "grad_norm": 4.5482072830200195, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 5.5595, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.036471168063085264, | |
| "grad_norm": 4.273440361022949, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 5.7877, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.03696402168555939, | |
| "grad_norm": 4.279988765716553, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 5.7015, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.037456875308033516, | |
| "grad_norm": 3.8421711921691895, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 5.6129, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.03794972893050764, | |
| "grad_norm": 4.161666393280029, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 5.6966, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.03844258255298177, | |
| "grad_norm": 4.275053977966309, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 5.7704, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.03893543617545589, | |
| "grad_norm": 4.343554496765137, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 5.9121, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.03942828979793001, | |
| "grad_norm": 4.524814128875732, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 5.6133, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03992114342040414, | |
| "grad_norm": 4.053195953369141, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 5.8539, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.040413997042878264, | |
| "grad_norm": 4.116884231567383, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 5.6884, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.04090685066535239, | |
| "grad_norm": 4.242743492126465, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 5.6097, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.041399704287826515, | |
| "grad_norm": 4.237097263336182, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 5.7156, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.04189255791030064, | |
| "grad_norm": 4.75575590133667, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 5.6946, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.04238541153277477, | |
| "grad_norm": 4.682205677032471, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 5.7993, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.04287826515524889, | |
| "grad_norm": 4.896819591522217, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 6.1002, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.04337111877772302, | |
| "grad_norm": 4.500687599182129, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 5.5547, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.04386397240019714, | |
| "grad_norm": 4.486171722412109, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 5.5062, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.04435682602267127, | |
| "grad_norm": 4.3436737060546875, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 5.5787, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.04484967964514539, | |
| "grad_norm": 4.709849834442139, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 5.8669, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.045342533267619514, | |
| "grad_norm": 4.942005634307861, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 5.7457, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.04583538689009364, | |
| "grad_norm": 4.71359920501709, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 5.8917, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.046328240512567766, | |
| "grad_norm": 4.906127452850342, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 6.1047, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.046821094135041895, | |
| "grad_norm": 4.9009504318237305, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 5.6463, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.04731394775751602, | |
| "grad_norm": 5.0316853523254395, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 5.982, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.04780680137999014, | |
| "grad_norm": 5.0484299659729, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 5.7041, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.04829965500246427, | |
| "grad_norm": 5.215059280395508, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 5.9011, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.04879250862493839, | |
| "grad_norm": 5.8646955490112305, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 6.112, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.04928536224741252, | |
| "grad_norm": 6.156351089477539, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 5.9004, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04928536224741252, | |
| "eval_loss": 1.431005597114563, | |
| "eval_runtime": 261.684, | |
| "eval_samples_per_second": 13.058, | |
| "eval_steps_per_second": 3.267, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04977821586988664, | |
| "grad_norm": 4.2226433753967285, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 5.885, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.05027106949236077, | |
| "grad_norm": 4.225105285644531, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 6.0265, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.050763923114834894, | |
| "grad_norm": 3.3986520767211914, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 5.7708, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.051256776737309016, | |
| "grad_norm": 3.3197200298309326, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 5.5425, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.051749630359783146, | |
| "grad_norm": 3.359449863433838, | |
| "learning_rate": 5e-05, | |
| "loss": 5.6861, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.05224248398225727, | |
| "grad_norm": 3.6861770153045654, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 5.8957, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.0527353376047314, | |
| "grad_norm": 3.3729681968688965, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 5.6209, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.05322819122720552, | |
| "grad_norm": 3.5027880668640137, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 5.3333, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.05372104484967965, | |
| "grad_norm": 3.4611856937408447, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 5.7229, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.05421389847215377, | |
| "grad_norm": 3.641321897506714, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 6.0581, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05470675209462789, | |
| "grad_norm": 3.8468551635742188, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 5.5633, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.05519960571710202, | |
| "grad_norm": 3.6477551460266113, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 5.3808, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.055692459339576145, | |
| "grad_norm": 3.6040382385253906, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 5.1802, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.056185312962050274, | |
| "grad_norm": 3.5388293266296387, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 5.441, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.056678166584524396, | |
| "grad_norm": 3.9764630794525146, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 5.9069, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.05717102020699852, | |
| "grad_norm": 3.8530142307281494, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 5.4141, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.05766387382947265, | |
| "grad_norm": 3.6980135440826416, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 5.2554, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.05815672745194677, | |
| "grad_norm": 3.621671676635742, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 5.2911, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.0586495810744209, | |
| "grad_norm": 3.8117148876190186, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 5.4398, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.05914243469689502, | |
| "grad_norm": 3.9566729068756104, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 5.6742, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.05963528831936915, | |
| "grad_norm": 3.656689405441284, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 5.3545, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.06012814194184327, | |
| "grad_norm": 3.7543258666992188, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 5.2922, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.060620995564317395, | |
| "grad_norm": 3.839445114135742, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 5.4221, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.061113849186791525, | |
| "grad_norm": 4.027908802032471, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 5.3428, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.06160670280926565, | |
| "grad_norm": 3.913569927215576, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 5.5121, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.062099556431739776, | |
| "grad_norm": 3.884404420852661, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 5.4622, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.0625924100542139, | |
| "grad_norm": 4.357018947601318, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 5.842, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.06308526367668803, | |
| "grad_norm": 4.205898761749268, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 5.4984, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.06357811729916214, | |
| "grad_norm": 3.913008213043213, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 5.5454, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.06407097092163627, | |
| "grad_norm": 4.038344383239746, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 5.681, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0645638245441104, | |
| "grad_norm": 3.8500421047210693, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 5.316, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.06505667816658453, | |
| "grad_norm": 4.233459949493408, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 5.6574, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.06554953178905865, | |
| "grad_norm": 4.028703689575195, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 5.219, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.06604238541153278, | |
| "grad_norm": 4.3093109130859375, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 5.5462, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0665352390340069, | |
| "grad_norm": 4.2733378410339355, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 5.4858, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.06702809265648102, | |
| "grad_norm": 4.1830878257751465, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 5.1703, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.06752094627895515, | |
| "grad_norm": 4.2492170333862305, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 5.4769, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.06801379990142928, | |
| "grad_norm": 4.3830060958862305, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 5.5719, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.06850665352390341, | |
| "grad_norm": 4.348109722137451, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 5.3788, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.06899950714637752, | |
| "grad_norm": 4.3019232749938965, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 5.5993, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06949236076885165, | |
| "grad_norm": 4.5521559715271, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 5.8097, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.06998521439132578, | |
| "grad_norm": 4.503178119659424, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 5.5559, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.0704780680137999, | |
| "grad_norm": 4.683341979980469, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 5.5462, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.07097092163627403, | |
| "grad_norm": 4.591628074645996, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 5.4505, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.07146377525874816, | |
| "grad_norm": 4.827720642089844, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 5.8953, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.07195662888122227, | |
| "grad_norm": 4.654827117919922, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 5.7021, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.0724494825036964, | |
| "grad_norm": 5.047876358032227, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 5.9535, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.07294233612617053, | |
| "grad_norm": 5.1422014236450195, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 5.7735, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.07343518974864466, | |
| "grad_norm": 5.423142433166504, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 5.9309, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.07392804337111877, | |
| "grad_norm": 5.620732307434082, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 5.6295, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07392804337111877, | |
| "eval_loss": 1.3822078704833984, | |
| "eval_runtime": 261.5296, | |
| "eval_samples_per_second": 13.065, | |
| "eval_steps_per_second": 3.269, | |
| "step": 150 | |
| } | |
| ], | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.16080942559658e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |