| { | |
| "best_metric": 0.6717210412025452, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.026522853859075237, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00017681902572716824, | |
| "grad_norm": 31.273752212524414, | |
| "learning_rate": 1e-05, | |
| "loss": 5.7904, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00017681902572716824, | |
| "eval_loss": 1.5057493448257446, | |
| "eval_runtime": 707.0176, | |
| "eval_samples_per_second": 13.472, | |
| "eval_steps_per_second": 3.369, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0003536380514543365, | |
| "grad_norm": 33.96555709838867, | |
| "learning_rate": 2e-05, | |
| "loss": 6.1229, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0005304570771815047, | |
| "grad_norm": 22.80571937561035, | |
| "learning_rate": 3e-05, | |
| "loss": 5.831, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.000707276102908673, | |
| "grad_norm": 17.077383041381836, | |
| "learning_rate": 4e-05, | |
| "loss": 4.8649, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0008840951286358412, | |
| "grad_norm": 15.064048767089844, | |
| "learning_rate": 5e-05, | |
| "loss": 4.4974, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0010609141543630094, | |
| "grad_norm": 12.101446151733398, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9707, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0012377331800901778, | |
| "grad_norm": 10.376118659973145, | |
| "learning_rate": 7e-05, | |
| "loss": 3.7705, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.001414552205817346, | |
| "grad_norm": 10.354647636413574, | |
| "learning_rate": 8e-05, | |
| "loss": 3.6031, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.001591371231544514, | |
| "grad_norm": 10.469047546386719, | |
| "learning_rate": 9e-05, | |
| "loss": 3.4053, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0017681902572716825, | |
| "grad_norm": 9.278939247131348, | |
| "learning_rate": 0.0001, | |
| "loss": 3.659, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0019450092829988506, | |
| "grad_norm": 15.399981498718262, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 3.6484, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.002121828308726019, | |
| "grad_norm": 10.160966873168945, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 3.5971, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.002298647334453187, | |
| "grad_norm": 7.888012886047363, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 3.4807, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0024754663601803555, | |
| "grad_norm": 7.219862937927246, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 3.3483, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0026522853859075237, | |
| "grad_norm": 6.999874114990234, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 3.27, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.002829104411634692, | |
| "grad_norm": 6.7764692306518555, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 3.2381, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.00300592343736186, | |
| "grad_norm": 7.229355812072754, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 3.1572, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.003182742463089028, | |
| "grad_norm": 7.318708419799805, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 3.2629, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0033595614888161968, | |
| "grad_norm": 7.011357307434082, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 3.1216, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.003536380514543365, | |
| "grad_norm": 6.351388454437256, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 3.1227, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.003713199540270533, | |
| "grad_norm": 6.193704128265381, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 3.0576, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0038900185659977013, | |
| "grad_norm": 6.23767614364624, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 3.1666, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.00406683759172487, | |
| "grad_norm": 5.724069118499756, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 3.0868, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.004243656617452038, | |
| "grad_norm": 6.114058494567871, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 3.1426, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.004420475643179206, | |
| "grad_norm": 6.249318599700928, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 3.1009, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.004597294668906374, | |
| "grad_norm": 5.979328632354736, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 3.0182, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0047741136946335425, | |
| "grad_norm": 5.138391971588135, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 2.7047, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.004950932720360711, | |
| "grad_norm": 5.40130090713501, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 3.1115, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.005127751746087879, | |
| "grad_norm": 5.664342403411865, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.8735, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.005304570771815047, | |
| "grad_norm": 5.740615367889404, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 3.1383, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.005481389797542215, | |
| "grad_norm": 5.822990417480469, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 2.7514, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.005658208823269384, | |
| "grad_norm": 6.541563034057617, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 2.896, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.005835027848996552, | |
| "grad_norm": 5.708773612976074, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 2.9389, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.00601184687472372, | |
| "grad_norm": 5.139838218688965, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 2.85, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.006188665900450889, | |
| "grad_norm": 4.9747796058654785, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 2.9934, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.006365484926178056, | |
| "grad_norm": 4.744929313659668, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 2.893, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.006542303951905225, | |
| "grad_norm": 5.647818088531494, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 3.0005, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0067191229776323936, | |
| "grad_norm": 5.120396614074707, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 2.8626, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.006895942003359561, | |
| "grad_norm": 5.247471809387207, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 2.9666, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.00707276102908673, | |
| "grad_norm": 5.75712251663208, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 2.7657, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.007249580054813898, | |
| "grad_norm": 5.175816535949707, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 2.9367, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.007426399080541066, | |
| "grad_norm": 5.635525703430176, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 2.8576, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.007603218106268235, | |
| "grad_norm": 4.829882621765137, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 2.6487, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0077800371319954025, | |
| "grad_norm": 4.661524295806885, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 2.7636, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.007956856157722571, | |
| "grad_norm": 5.161839008331299, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 2.7386, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.00813367518344974, | |
| "grad_norm": 5.641641616821289, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 2.6801, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.008310494209176908, | |
| "grad_norm": 5.284437656402588, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 2.7854, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.008487313234904075, | |
| "grad_norm": 4.893701076507568, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.6241, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.008664132260631244, | |
| "grad_norm": 5.710663795471191, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 2.6486, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.008840951286358412, | |
| "grad_norm": 5.808746337890625, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 2.8028, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.008840951286358412, | |
| "eval_loss": 0.8329792022705078, | |
| "eval_runtime": 712.2685, | |
| "eval_samples_per_second": 13.373, | |
| "eval_steps_per_second": 3.344, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.009017770312085581, | |
| "grad_norm": 11.850320816040039, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 4.0343, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.009194589337812748, | |
| "grad_norm": 7.7840576171875, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 3.4929, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.009371408363539916, | |
| "grad_norm": 4.552064895629883, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 3.1584, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.009548227389267085, | |
| "grad_norm": 7.188632965087891, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 3.3249, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.009725046414994254, | |
| "grad_norm": 5.568719863891602, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 3.0549, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.009901865440721422, | |
| "grad_norm": 4.3227972984313965, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 3.1477, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.010078684466448589, | |
| "grad_norm": 4.280993938446045, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 3.0837, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.010255503492175758, | |
| "grad_norm": 4.789489269256592, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 3.0091, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.010432322517902926, | |
| "grad_norm": 5.41645622253418, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 3.1189, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.010609141543630095, | |
| "grad_norm": 4.8120269775390625, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 2.991, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.010785960569357263, | |
| "grad_norm": 4.251246929168701, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 2.8324, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.01096277959508443, | |
| "grad_norm": 4.403325080871582, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 3.0057, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.011139598620811599, | |
| "grad_norm": 4.967185974121094, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 3.0389, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.011316417646538767, | |
| "grad_norm": 4.585970401763916, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 2.8278, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.011493236672265936, | |
| "grad_norm": 4.517243385314941, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 3.1275, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.011670055697993105, | |
| "grad_norm": 4.599977970123291, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 2.8354, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.011846874723720272, | |
| "grad_norm": 4.074687480926514, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 2.7571, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.01202369374944744, | |
| "grad_norm": 4.451271057128906, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 2.9244, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.012200512775174609, | |
| "grad_norm": 3.967043161392212, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 2.7281, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.012377331800901777, | |
| "grad_norm": 4.431678295135498, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 2.8933, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.012554150826628946, | |
| "grad_norm": 4.48666524887085, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 2.9408, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.012730969852356113, | |
| "grad_norm": 4.583484172821045, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 2.8614, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.012907788878083281, | |
| "grad_norm": 4.457082271575928, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 3.0324, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.01308460790381045, | |
| "grad_norm": 4.137736797332764, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 2.8703, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.013261426929537619, | |
| "grad_norm": 4.20756721496582, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 2.7438, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.013438245955264787, | |
| "grad_norm": 4.50335168838501, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 2.836, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.013615064980991954, | |
| "grad_norm": 3.9836575984954834, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 2.7682, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.013791884006719123, | |
| "grad_norm": 4.124038219451904, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 2.9161, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.013968703032446291, | |
| "grad_norm": 4.177953243255615, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 2.8599, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01414552205817346, | |
| "grad_norm": 4.370811939239502, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 2.9956, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.014322341083900628, | |
| "grad_norm": 4.351916313171387, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 2.7608, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.014499160109627795, | |
| "grad_norm": 4.02456521987915, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 2.8675, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.014675979135354964, | |
| "grad_norm": 3.745039701461792, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 2.735, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.014852798161082132, | |
| "grad_norm": 4.1861042976379395, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 2.7153, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.015029617186809301, | |
| "grad_norm": 4.138434886932373, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 2.9203, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.01520643621253647, | |
| "grad_norm": 4.018141746520996, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 2.6352, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.015383255238263636, | |
| "grad_norm": 4.0129923820495605, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 2.7209, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.015560074263990805, | |
| "grad_norm": 4.6002888679504395, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 2.7132, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.015736893289717974, | |
| "grad_norm": 4.2588067054748535, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 2.4424, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.015913712315445142, | |
| "grad_norm": 4.202187538146973, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 2.5436, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.01609053134117231, | |
| "grad_norm": 4.206346035003662, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 2.5855, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.01626735036689948, | |
| "grad_norm": 4.269227981567383, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 2.712, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.016444169392626648, | |
| "grad_norm": 3.9102425575256348, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 2.4231, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.016620988418353817, | |
| "grad_norm": 4.122182846069336, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 2.4536, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.01679780744408098, | |
| "grad_norm": 4.083293437957764, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 2.6099, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.01697462646980815, | |
| "grad_norm": 4.000704765319824, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 2.5202, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.01715144549553532, | |
| "grad_norm": 4.2389092445373535, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 2.5216, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.017328264521262488, | |
| "grad_norm": 4.1938605308532715, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 2.4519, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.017505083546989656, | |
| "grad_norm": 4.141265392303467, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 2.425, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.017681902572716825, | |
| "grad_norm": 4.398811340332031, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 2.416, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.017681902572716825, | |
| "eval_loss": 0.7194002270698547, | |
| "eval_runtime": 712.028, | |
| "eval_samples_per_second": 13.377, | |
| "eval_steps_per_second": 3.345, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.017858721598443993, | |
| "grad_norm": 7.377161026000977, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 3.2334, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.018035540624171162, | |
| "grad_norm": 5.567739009857178, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 3.3289, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.01821235964989833, | |
| "grad_norm": 4.169541358947754, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 2.9622, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.018389178675625496, | |
| "grad_norm": 5.395923614501953, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 3.1449, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.018565997701352664, | |
| "grad_norm": 4.531886577606201, | |
| "learning_rate": 5e-05, | |
| "loss": 3.1416, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.018742816727079833, | |
| "grad_norm": 3.8139052391052246, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 2.9983, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.018919635752807, | |
| "grad_norm": 3.7022151947021484, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 3.0179, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.01909645477853417, | |
| "grad_norm": 3.809081792831421, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 2.8028, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.01927327380426134, | |
| "grad_norm": 3.4583377838134766, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 2.7208, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.019450092829988507, | |
| "grad_norm": 3.455575466156006, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 2.8423, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.019626911855715676, | |
| "grad_norm": 3.6886868476867676, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 2.6708, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.019803730881442844, | |
| "grad_norm": 4.032835960388184, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 2.9931, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.019980549907170013, | |
| "grad_norm": 3.7966928482055664, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 2.8558, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.020157368932897178, | |
| "grad_norm": 3.60139799118042, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 2.4373, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.020334187958624347, | |
| "grad_norm": 3.8890609741210938, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 2.758, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.020511006984351515, | |
| "grad_norm": 3.8674120903015137, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 2.8288, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.020687826010078684, | |
| "grad_norm": 3.7245192527770996, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 2.7111, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.020864645035805852, | |
| "grad_norm": 3.8925111293792725, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 2.782, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.02104146406153302, | |
| "grad_norm": 3.7165839672088623, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 2.7444, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.02121828308726019, | |
| "grad_norm": 4.107248783111572, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 2.7172, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.021395102112987358, | |
| "grad_norm": 3.8568592071533203, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 2.5638, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.021571921138714527, | |
| "grad_norm": 3.638155460357666, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 2.6479, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.021748740164441695, | |
| "grad_norm": 3.860167980194092, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 2.6039, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.02192555919016886, | |
| "grad_norm": 3.792959690093994, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.6498, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.02210237821589603, | |
| "grad_norm": 4.04556941986084, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 2.632, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.022279197241623198, | |
| "grad_norm": 6.900913238525391, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 2.5513, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.022456016267350366, | |
| "grad_norm": 3.745027780532837, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 2.6285, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.022632835293077535, | |
| "grad_norm": 3.830739974975586, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 2.7284, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.022809654318804704, | |
| "grad_norm": 3.877413511276245, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 2.5802, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.022986473344531872, | |
| "grad_norm": 3.963332414627075, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 2.8791, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02316329237025904, | |
| "grad_norm": 3.840338706970215, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 2.4996, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.02334011139598621, | |
| "grad_norm": 3.7065539360046387, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 2.5997, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.023516930421713378, | |
| "grad_norm": 3.834458589553833, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 2.4962, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.023693749447440543, | |
| "grad_norm": 3.911560297012329, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 2.5727, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.02387056847316771, | |
| "grad_norm": 3.85789155960083, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 2.4992, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.02404738749889488, | |
| "grad_norm": 4.070404529571533, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 2.7064, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.02422420652462205, | |
| "grad_norm": 3.5851242542266846, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 2.449, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.024401025550349217, | |
| "grad_norm": 3.727949857711792, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 2.6118, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.024577844576076386, | |
| "grad_norm": 3.50991153717041, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 2.431, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.024754663601803555, | |
| "grad_norm": 3.723494529724121, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 2.2647, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.024931482627530723, | |
| "grad_norm": 3.584364175796509, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 2.4095, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.025108301653257892, | |
| "grad_norm": 3.8995087146759033, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 2.5542, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.02528512067898506, | |
| "grad_norm": 3.8994150161743164, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.3609, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.025461939704712225, | |
| "grad_norm": 3.8739383220672607, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 2.4378, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.025638758730439394, | |
| "grad_norm": 3.730926513671875, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 2.3718, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.025815577756166563, | |
| "grad_norm": 4.219221591949463, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 2.5712, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.02599239678189373, | |
| "grad_norm": 4.296256065368652, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 2.6027, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.0261692158076209, | |
| "grad_norm": 3.8536908626556396, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 2.2772, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.02634603483334807, | |
| "grad_norm": 3.92279314994812, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 2.2474, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.026522853859075237, | |
| "grad_norm": 4.397715091705322, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 2.454, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.026522853859075237, | |
| "eval_loss": 0.6717210412025452, | |
| "eval_runtime": 712.25, | |
| "eval_samples_per_second": 13.373, | |
| "eval_steps_per_second": 3.344, | |
| "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": 2.146499429400576e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |