| { | |
| "best_metric": 1.3909202814102173, | |
| "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": 16.314523696899414, | |
| "learning_rate": 1e-05, | |
| "loss": 6.995, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0004928536224741252, | |
| "eval_loss": 1.744011402130127, | |
| "eval_runtime": 259.7735, | |
| "eval_samples_per_second": 13.154, | |
| "eval_steps_per_second": 3.291, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0009857072449482504, | |
| "grad_norm": 16.28382110595703, | |
| "learning_rate": 2e-05, | |
| "loss": 7.1083, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0014785608674223755, | |
| "grad_norm": 15.063352584838867, | |
| "learning_rate": 3e-05, | |
| "loss": 6.7865, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.001971414489896501, | |
| "grad_norm": 10.722319602966309, | |
| "learning_rate": 4e-05, | |
| "loss": 6.3322, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0024642681123706258, | |
| "grad_norm": 10.376184463500977, | |
| "learning_rate": 5e-05, | |
| "loss": 6.2006, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.002957121734844751, | |
| "grad_norm": 11.510865211486816, | |
| "learning_rate": 6e-05, | |
| "loss": 6.5907, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0034499753573188764, | |
| "grad_norm": 8.317137718200684, | |
| "learning_rate": 7e-05, | |
| "loss": 6.1843, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.003942828979793002, | |
| "grad_norm": 6.411678314208984, | |
| "learning_rate": 8e-05, | |
| "loss": 5.9647, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.004435682602267127, | |
| "grad_norm": 7.069495677947998, | |
| "learning_rate": 9e-05, | |
| "loss": 6.3579, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0049285362247412515, | |
| "grad_norm": 6.966650485992432, | |
| "learning_rate": 0.0001, | |
| "loss": 5.9963, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.005421389847215377, | |
| "grad_norm": 6.528267860412598, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 6.0897, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.005914243469689502, | |
| "grad_norm": 6.692461967468262, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 6.1342, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.006407097092163627, | |
| "grad_norm": 6.611419200897217, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 6.0027, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.006899950714637753, | |
| "grad_norm": 6.110144138336182, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 6.0202, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.007392804337111878, | |
| "grad_norm": 6.099429607391357, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 6.1673, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.007885657959586003, | |
| "grad_norm": 5.863210201263428, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 5.8905, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.008378511582060127, | |
| "grad_norm": 5.425501823425293, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 6.0835, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.008871365204534253, | |
| "grad_norm": 5.532186031341553, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 6.1977, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.009364218827008379, | |
| "grad_norm": 6.280749320983887, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 6.0634, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.009857072449482503, | |
| "grad_norm": 5.083516597747803, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 6.2222, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.010349926071956629, | |
| "grad_norm": 5.291787147521973, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 6.0114, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.010842779694430755, | |
| "grad_norm": 4.903357982635498, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 5.8388, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.011335633316904879, | |
| "grad_norm": 5.019505977630615, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 5.5975, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.011828486939379004, | |
| "grad_norm": 4.963260650634766, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 6.1214, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.01232134056185313, | |
| "grad_norm": 4.857863426208496, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 5.8345, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.012814194184327254, | |
| "grad_norm": 4.881758213043213, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 5.9828, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.01330704780680138, | |
| "grad_norm": 5.054996490478516, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 6.2734, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.013799901429275506, | |
| "grad_norm": 5.001477241516113, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 5.8811, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.01429275505174963, | |
| "grad_norm": 5.1075968742370605, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 6.1891, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.014785608674223755, | |
| "grad_norm": 5.3841376304626465, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 5.9307, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.015278462296697881, | |
| "grad_norm": 4.67518949508667, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 5.7908, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.015771315919172007, | |
| "grad_norm": 5.1068854331970215, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 5.9453, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.016264169541646133, | |
| "grad_norm": 5.1424078941345215, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 5.5909, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.016757023164120255, | |
| "grad_norm": 5.124903678894043, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 5.9921, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.01724987678659438, | |
| "grad_norm": 4.945921421051025, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 5.8984, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.017742730409068506, | |
| "grad_norm": 5.158941268920898, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 5.9213, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.018235584031542632, | |
| "grad_norm": 6.201539039611816, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 5.8184, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.018728437654016758, | |
| "grad_norm": 5.76201057434082, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 6.0167, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.019221291276490884, | |
| "grad_norm": 4.935492992401123, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 5.9262, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.019714144898965006, | |
| "grad_norm": 5.548451900482178, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 5.9082, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.020206998521439132, | |
| "grad_norm": 4.932051658630371, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 6.1315, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.020699852143913258, | |
| "grad_norm": 5.4284234046936035, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 6.2949, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.021192705766387383, | |
| "grad_norm": 5.175426006317139, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 6.2249, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.02168555938886151, | |
| "grad_norm": 5.790852069854736, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 5.9886, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.022178413011335635, | |
| "grad_norm": 5.603730201721191, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 6.1273, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.022671266633809757, | |
| "grad_norm": 12.599672317504883, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 6.2093, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.023164120256283883, | |
| "grad_norm": 5.739752292633057, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 6.1065, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.02365697387875801, | |
| "grad_norm": 5.95276403427124, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 6.2482, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.024149827501232134, | |
| "grad_norm": 6.856138706207275, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 6.5065, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.02464268112370626, | |
| "grad_norm": 7.758269786834717, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 6.2105, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02464268112370626, | |
| "eval_loss": 1.7778630256652832, | |
| "eval_runtime": 262.1729, | |
| "eval_samples_per_second": 13.033, | |
| "eval_steps_per_second": 3.261, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.025135534746180386, | |
| "grad_norm": 43.93449783325195, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 7.5127, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.025628388368654508, | |
| "grad_norm": 7.137115001678467, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 6.6552, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.026121241991128634, | |
| "grad_norm": 3.746063232421875, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 5.9609, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.02661409561360276, | |
| "grad_norm": 3.687379837036133, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 6.0231, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.027106949236076885, | |
| "grad_norm": 3.6193816661834717, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 5.8947, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.02759980285855101, | |
| "grad_norm": 3.8151955604553223, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 6.1616, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.028092656481025137, | |
| "grad_norm": 3.686230182647705, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 5.807, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.02858551010349926, | |
| "grad_norm": 8.27771282196045, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 5.6026, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.029078363725973385, | |
| "grad_norm": 7.051781177520752, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 5.6583, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.02957121734844751, | |
| "grad_norm": 4.224041938781738, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 5.8663, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.030064070970921637, | |
| "grad_norm": 4.13751745223999, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 5.9037, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.030556924593395762, | |
| "grad_norm": 4.013219833374023, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 5.9175, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.031049778215869888, | |
| "grad_norm": 3.9895718097686768, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 5.7139, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.031542631838344014, | |
| "grad_norm": 3.8255083560943604, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 5.9406, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.032035485460818136, | |
| "grad_norm": 3.870253801345825, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 5.8729, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.032528339083292265, | |
| "grad_norm": 4.286971092224121, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 5.8873, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.03302119270576639, | |
| "grad_norm": 4.536320686340332, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 5.6097, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.03351404632824051, | |
| "grad_norm": 3.7853376865386963, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 5.6259, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.03400689995071464, | |
| "grad_norm": 3.8190784454345703, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 5.5137, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.03449975357318876, | |
| "grad_norm": 3.990157127380371, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 5.8526, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03499260719566289, | |
| "grad_norm": 4.003045558929443, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 5.2885, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.03548546081813701, | |
| "grad_norm": 4.0029802322387695, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 5.815, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.035978314440611135, | |
| "grad_norm": 3.875368118286133, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 5.523, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.036471168063085264, | |
| "grad_norm": 4.117485523223877, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 5.7657, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.03696402168555939, | |
| "grad_norm": 4.247128009796143, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 5.6921, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.037456875308033516, | |
| "grad_norm": 3.862671136856079, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 5.5907, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.03794972893050764, | |
| "grad_norm": 3.982457160949707, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 5.6725, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.03844258255298177, | |
| "grad_norm": 4.2674241065979, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 5.7546, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.03893543617545589, | |
| "grad_norm": 4.177145004272461, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 5.8932, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.03942828979793001, | |
| "grad_norm": 4.265317916870117, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 5.5848, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03992114342040414, | |
| "grad_norm": 4.019691467285156, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 5.8258, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.040413997042878264, | |
| "grad_norm": 4.079697132110596, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 5.6694, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.04090685066535239, | |
| "grad_norm": 4.25301456451416, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 5.594, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.041399704287826515, | |
| "grad_norm": 4.196613788604736, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 5.694, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.04189255791030064, | |
| "grad_norm": 4.6560163497924805, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 5.6949, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.04238541153277477, | |
| "grad_norm": 4.8911967277526855, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 5.7983, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.04287826515524889, | |
| "grad_norm": 4.876672267913818, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 6.0781, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.04337111877772302, | |
| "grad_norm": 4.463366508483887, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 5.534, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.04386397240019714, | |
| "grad_norm": 4.509284019470215, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 5.508, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.04435682602267127, | |
| "grad_norm": 5.039193153381348, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 5.5805, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.04484967964514539, | |
| "grad_norm": 4.586795806884766, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 5.8581, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.045342533267619514, | |
| "grad_norm": 4.905638217926025, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 5.7439, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.04583538689009364, | |
| "grad_norm": 4.709536552429199, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 5.8878, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.046328240512567766, | |
| "grad_norm": 4.950308322906494, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 6.1136, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.046821094135041895, | |
| "grad_norm": 4.843292713165283, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 5.6441, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.04731394775751602, | |
| "grad_norm": 6.348719120025635, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 5.9914, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.04780680137999014, | |
| "grad_norm": 4.939000129699707, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 5.6883, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.04829965500246427, | |
| "grad_norm": 5.359467506408691, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 5.9026, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.04879250862493839, | |
| "grad_norm": 5.869283199310303, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 6.1058, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.04928536224741252, | |
| "grad_norm": 9.018366813659668, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 5.928, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04928536224741252, | |
| "eval_loss": 1.443994164466858, | |
| "eval_runtime": 262.8429, | |
| "eval_samples_per_second": 13.0, | |
| "eval_steps_per_second": 3.253, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04977821586988664, | |
| "grad_norm": 4.107466220855713, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 5.9238, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.05027106949236077, | |
| "grad_norm": 4.2340288162231445, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 6.0714, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.050763923114834894, | |
| "grad_norm": 6.402919769287109, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 5.7883, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.051256776737309016, | |
| "grad_norm": 3.29791259765625, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 5.5427, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.051749630359783146, | |
| "grad_norm": 3.453167200088501, | |
| "learning_rate": 5e-05, | |
| "loss": 5.6945, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.05224248398225727, | |
| "grad_norm": 3.6456615924835205, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 5.8859, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.0527353376047314, | |
| "grad_norm": 3.4060540199279785, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 5.6115, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.05322819122720552, | |
| "grad_norm": 3.5317554473876953, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 5.3244, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.05372104484967965, | |
| "grad_norm": 3.5276272296905518, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 5.7114, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.05421389847215377, | |
| "grad_norm": 3.6883718967437744, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 6.0391, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05470675209462789, | |
| "grad_norm": 3.7652981281280518, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 5.5622, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.05519960571710202, | |
| "grad_norm": 3.6254518032073975, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 5.3765, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.055692459339576145, | |
| "grad_norm": 4.109588146209717, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 5.1925, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.056185312962050274, | |
| "grad_norm": 3.6142637729644775, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 5.445, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.056678166584524396, | |
| "grad_norm": 3.8312244415283203, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 5.8935, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.05717102020699852, | |
| "grad_norm": 3.614104747772217, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 5.4059, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.05766387382947265, | |
| "grad_norm": 3.5888030529022217, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 5.2579, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.05815672745194677, | |
| "grad_norm": 3.5380568504333496, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 5.2866, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.0586495810744209, | |
| "grad_norm": 3.732438087463379, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 5.4266, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.05914243469689502, | |
| "grad_norm": 3.9729933738708496, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 5.6585, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.05963528831936915, | |
| "grad_norm": 3.5968165397644043, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 5.3438, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.06012814194184327, | |
| "grad_norm": 3.687399387359619, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 5.2816, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.060620995564317395, | |
| "grad_norm": 3.8859143257141113, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 5.4255, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.061113849186791525, | |
| "grad_norm": 4.007688045501709, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 5.3268, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.06160670280926565, | |
| "grad_norm": 3.9625210762023926, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 5.5169, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.062099556431739776, | |
| "grad_norm": 3.9683034420013428, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 5.4601, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.0625924100542139, | |
| "grad_norm": 4.349674701690674, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 5.8421, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.06308526367668803, | |
| "grad_norm": 4.083832263946533, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 5.4894, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.06357811729916214, | |
| "grad_norm": 3.790072202682495, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 5.5454, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.06407097092163627, | |
| "grad_norm": 4.148959636688232, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 5.6739, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0645638245441104, | |
| "grad_norm": 3.849386215209961, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 5.3132, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.06505667816658453, | |
| "grad_norm": 4.21781587600708, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 5.6404, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.06554953178905865, | |
| "grad_norm": 3.986280679702759, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 5.2083, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.06604238541153278, | |
| "grad_norm": 4.222084045410156, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 5.5415, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0665352390340069, | |
| "grad_norm": 4.2696309089660645, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 5.4758, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.06702809265648102, | |
| "grad_norm": 4.158928394317627, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 5.1622, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.06752094627895515, | |
| "grad_norm": 4.044380187988281, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 5.4663, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.06801379990142928, | |
| "grad_norm": 4.391135215759277, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 5.5693, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.06850665352390341, | |
| "grad_norm": 4.313879489898682, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 5.3598, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.06899950714637752, | |
| "grad_norm": 4.336035251617432, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 5.5988, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06949236076885165, | |
| "grad_norm": 4.5523505210876465, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 5.8046, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.06998521439132578, | |
| "grad_norm": 4.409508228302002, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 5.552, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.0704780680137999, | |
| "grad_norm": 4.644127368927002, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 5.5421, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.07097092163627403, | |
| "grad_norm": 4.544384479522705, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 5.4519, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.07146377525874816, | |
| "grad_norm": 4.849193096160889, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 5.8996, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.07195662888122227, | |
| "grad_norm": 4.661250591278076, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 5.7048, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.0724494825036964, | |
| "grad_norm": 5.036535263061523, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 5.9565, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.07294233612617053, | |
| "grad_norm": 5.321414470672607, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 5.7657, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.07343518974864466, | |
| "grad_norm": 5.340336799621582, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 5.9288, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.07392804337111877, | |
| "grad_norm": 5.687753677368164, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 5.6259, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07392804337111877, | |
| "eval_loss": 1.3909202814102173, | |
| "eval_runtime": 262.5133, | |
| "eval_samples_per_second": 13.016, | |
| "eval_steps_per_second": 3.257, | |
| "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 | |
| } | |