| { | |
| "best_metric": 1.4202370643615723, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.018063040009633623, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00012042026673089081, | |
| "grad_norm": 51.0792236328125, | |
| "learning_rate": 1e-05, | |
| "loss": 8.9258, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00012042026673089081, | |
| "eval_loss": 2.869473695755005, | |
| "eval_runtime": 1613.551, | |
| "eval_samples_per_second": 8.668, | |
| "eval_steps_per_second": 2.167, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00024084053346178162, | |
| "grad_norm": 59.88539505004883, | |
| "learning_rate": 2e-05, | |
| "loss": 9.6361, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00036126080019267245, | |
| "grad_norm": 49.332725524902344, | |
| "learning_rate": 3e-05, | |
| "loss": 9.3627, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00048168106692356323, | |
| "grad_norm": 44.69882583618164, | |
| "learning_rate": 4e-05, | |
| "loss": 9.3756, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0006021013336544541, | |
| "grad_norm": 50.62449264526367, | |
| "learning_rate": 5e-05, | |
| "loss": 8.6678, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0007225216003853449, | |
| "grad_norm": 48.275550842285156, | |
| "learning_rate": 6e-05, | |
| "loss": 8.5039, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0008429418671162356, | |
| "grad_norm": 37.728294372558594, | |
| "learning_rate": 7e-05, | |
| "loss": 7.4285, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0009633621338471265, | |
| "grad_norm": 33.27507400512695, | |
| "learning_rate": 8e-05, | |
| "loss": 7.9351, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0010837824005780173, | |
| "grad_norm": 57.52439498901367, | |
| "learning_rate": 9e-05, | |
| "loss": 7.7795, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0012042026673089081, | |
| "grad_norm": 49.76416015625, | |
| "learning_rate": 0.0001, | |
| "loss": 8.1299, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.001324622934039799, | |
| "grad_norm": 30.481006622314453, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 7.527, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0014450432007706898, | |
| "grad_norm": 27.877037048339844, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 7.4031, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0015654634675015804, | |
| "grad_norm": 28.99033546447754, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 7.4073, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0016858837342324713, | |
| "grad_norm": 25.286575317382812, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 7.6599, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.001806304000963362, | |
| "grad_norm": 26.446382522583008, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 5.8333, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.001926724267694253, | |
| "grad_norm": 47.75482177734375, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 6.8325, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0020471445344251438, | |
| "grad_norm": 25.70916748046875, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 6.7639, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0021675648011560346, | |
| "grad_norm": 33.000648498535156, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 7.2915, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0022879850678869254, | |
| "grad_norm": 29.954723358154297, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 6.7451, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0024084053346178163, | |
| "grad_norm": 30.427520751953125, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 7.9425, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.002528825601348707, | |
| "grad_norm": 28.53919792175293, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 7.334, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.002649245868079598, | |
| "grad_norm": 37.44373321533203, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 6.9368, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0027696661348104888, | |
| "grad_norm": 27.007282257080078, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 7.6607, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0028900864015413796, | |
| "grad_norm": 29.33873748779297, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 7.3593, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.00301050666827227, | |
| "grad_norm": 28.406932830810547, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 7.372, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.003130926935003161, | |
| "grad_norm": 48.706180572509766, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 8.2033, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0032513472017340517, | |
| "grad_norm": 36.82452392578125, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 7.1247, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0033717674684649425, | |
| "grad_norm": 71.5036849975586, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 9.2888, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0034921877351958334, | |
| "grad_norm": 33.01215744018555, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 6.2341, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.003612608001926724, | |
| "grad_norm": 23.74918556213379, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 6.6404, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.003733028268657615, | |
| "grad_norm": 21.480255126953125, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 6.034, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.003853448535388506, | |
| "grad_norm": 19.545923233032227, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 5.8033, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.003973868802119396, | |
| "grad_norm": 20.797698974609375, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 5.8928, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0040942890688502875, | |
| "grad_norm": 22.636245727539062, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 5.3007, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.004214709335581178, | |
| "grad_norm": 18.059972763061523, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 4.9009, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.004335129602312069, | |
| "grad_norm": 19.2911319732666, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 5.4462, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.00445554986904296, | |
| "grad_norm": 20.103979110717773, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 5.716, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.004575970135773851, | |
| "grad_norm": 21.43541145324707, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 5.1975, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.004696390402504741, | |
| "grad_norm": 21.74974250793457, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 5.2275, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0048168106692356326, | |
| "grad_norm": 15.888248443603516, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 4.9456, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.004937230935966523, | |
| "grad_norm": 20.183866500854492, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 4.4926, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.005057651202697414, | |
| "grad_norm": 18.041362762451172, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 4.9749, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.005178071469428305, | |
| "grad_norm": 19.268260955810547, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 5.2045, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.005298491736159196, | |
| "grad_norm": 16.374629974365234, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 4.8538, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.005418912002890086, | |
| "grad_norm": 19.21706771850586, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 4.7291, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0055393322696209776, | |
| "grad_norm": 19.981130599975586, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 5.0403, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.005659752536351868, | |
| "grad_norm": 20.54612922668457, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 5.2282, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.005780172803082759, | |
| "grad_norm": 32.53288650512695, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 5.0887, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.00590059306981365, | |
| "grad_norm": 21.54436683654785, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 4.9969, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.00602101333654454, | |
| "grad_norm": 30.681230545043945, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 4.1368, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.00602101333654454, | |
| "eval_loss": 1.7432713508605957, | |
| "eval_runtime": 1626.3643, | |
| "eval_samples_per_second": 8.6, | |
| "eval_steps_per_second": 2.15, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.006141433603275431, | |
| "grad_norm": 48.510963439941406, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 7.0127, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.006261853870006322, | |
| "grad_norm": 34.02531433105469, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 6.9446, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.006382274136737213, | |
| "grad_norm": 20.909475326538086, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 6.2471, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.006502694403468103, | |
| "grad_norm": 30.891647338867188, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 6.6601, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.006623114670198995, | |
| "grad_norm": 20.516788482666016, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 5.7726, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.006743534936929885, | |
| "grad_norm": 15.028369903564453, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 5.8366, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.006863955203660776, | |
| "grad_norm": 21.2078914642334, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 6.0307, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.006984375470391667, | |
| "grad_norm": 19.609695434570312, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 6.0851, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.007104795737122558, | |
| "grad_norm": 19.332590103149414, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 6.0491, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.007225216003853448, | |
| "grad_norm": 22.65894317626953, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 6.6842, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.00734563627058434, | |
| "grad_norm": 16.9778995513916, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 6.5613, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.00746605653731523, | |
| "grad_norm": 18.769847869873047, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 6.2214, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.007586476804046121, | |
| "grad_norm": 18.140674591064453, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 6.2408, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.007706897070777012, | |
| "grad_norm": 23.19234848022461, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 5.6155, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.007827317337507902, | |
| "grad_norm": 20.677770614624023, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 6.368, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.007947737604238793, | |
| "grad_norm": 24.745595932006836, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 6.7105, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.008068157870969685, | |
| "grad_norm": 21.65793800354004, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 6.6292, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.008188578137700575, | |
| "grad_norm": 22.275861740112305, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 6.3876, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.008308998404431465, | |
| "grad_norm": 21.33165740966797, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 6.6753, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.008429418671162356, | |
| "grad_norm": 21.999799728393555, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 6.5288, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.008549838937893248, | |
| "grad_norm": 23.032381057739258, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 6.5278, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.008670259204624138, | |
| "grad_norm": 18.758989334106445, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 6.6759, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.008790679471355029, | |
| "grad_norm": 21.545459747314453, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 7.396, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.00891109973808592, | |
| "grad_norm": 24.817026138305664, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 7.0675, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.009031520004816811, | |
| "grad_norm": 20.182096481323242, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 6.9901, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.009151940271547702, | |
| "grad_norm": 26.779354095458984, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 7.4387, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.009272360538278592, | |
| "grad_norm": 22.794876098632812, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 6.7531, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.009392780805009483, | |
| "grad_norm": 28.068523406982422, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 8.4352, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.009513201071740375, | |
| "grad_norm": 33.00652313232422, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 7.3037, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.009633621338471265, | |
| "grad_norm": 26.842178344726562, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 5.8982, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.009754041605202156, | |
| "grad_norm": 31.74905014038086, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 6.6462, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.009874461871933046, | |
| "grad_norm": 30.011796951293945, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 5.8739, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.009994882138663936, | |
| "grad_norm": 24.284454345703125, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 5.9235, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.010115302405394828, | |
| "grad_norm": 17.595317840576172, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 5.0132, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.010235722672125719, | |
| "grad_norm": 15.925289154052734, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 4.8832, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.01035614293885661, | |
| "grad_norm": 14.465446472167969, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 4.4147, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.0104765632055875, | |
| "grad_norm": 16.447044372558594, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 4.7238, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.010596983472318392, | |
| "grad_norm": 16.312849044799805, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 4.8502, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.010717403739049282, | |
| "grad_norm": 15.325825691223145, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 4.777, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.010837824005780173, | |
| "grad_norm": 15.650074005126953, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 4.7193, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.010958244272511063, | |
| "grad_norm": 17.673521041870117, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 4.3277, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.011078664539241955, | |
| "grad_norm": 14.6486177444458, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 4.9623, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.011199084805972846, | |
| "grad_norm": 16.636259078979492, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 4.8406, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.011319505072703736, | |
| "grad_norm": 14.476512908935547, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 4.1294, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.011439925339434626, | |
| "grad_norm": 21.621837615966797, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 5.1405, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.011560345606165518, | |
| "grad_norm": 15.981372833251953, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 4.4981, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.011680765872896409, | |
| "grad_norm": 16.446104049682617, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 4.3123, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.0118011861396273, | |
| "grad_norm": 17.689773559570312, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 4.2699, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01192160640635819, | |
| "grad_norm": 17.166032791137695, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 4.1504, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.01204202667308908, | |
| "grad_norm": 17.5468807220459, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 4.617, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01204202667308908, | |
| "eval_loss": 1.4824187755584717, | |
| "eval_runtime": 1626.0282, | |
| "eval_samples_per_second": 8.602, | |
| "eval_steps_per_second": 2.151, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.012162446939819972, | |
| "grad_norm": 15.36607837677002, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 5.4322, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.012282867206550863, | |
| "grad_norm": 14.568869590759277, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 5.862, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.012403287473281753, | |
| "grad_norm": 15.210724830627441, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 5.5635, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.012523707740012643, | |
| "grad_norm": 14.810149192810059, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 5.2392, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.012644128006743536, | |
| "grad_norm": 16.995372772216797, | |
| "learning_rate": 5e-05, | |
| "loss": 5.8102, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.012764548273474426, | |
| "grad_norm": 17.859939575195312, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 5.684, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.012884968540205316, | |
| "grad_norm": 15.624403953552246, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 5.9794, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.013005388806936207, | |
| "grad_norm": 15.798469543457031, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 5.8513, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.013125809073667099, | |
| "grad_norm": 19.53667640686035, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 6.372, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.01324622934039799, | |
| "grad_norm": 16.418088912963867, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 6.2454, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.01336664960712888, | |
| "grad_norm": 18.04201316833496, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 6.2441, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.01348706987385977, | |
| "grad_norm": 18.177875518798828, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 6.2149, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.013607490140590662, | |
| "grad_norm": 16.015684127807617, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 6.2159, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.013727910407321553, | |
| "grad_norm": 17.19831657409668, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 6.5553, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.013848330674052443, | |
| "grad_norm": 18.239582061767578, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 6.6621, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.013968750940783333, | |
| "grad_norm": 17.54486656188965, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 6.723, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.014089171207514224, | |
| "grad_norm": 16.560501098632812, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 6.5665, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.014209591474245116, | |
| "grad_norm": 16.162139892578125, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 6.3858, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.014330011740976006, | |
| "grad_norm": 18.116655349731445, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 6.7837, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.014450432007706897, | |
| "grad_norm": 20.372034072875977, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 7.5936, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.014570852274437787, | |
| "grad_norm": 18.741952896118164, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 7.0342, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.01469127254116868, | |
| "grad_norm": 20.603229522705078, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 6.99, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.01481169280789957, | |
| "grad_norm": 19.649383544921875, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 7.2942, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.01493211307463046, | |
| "grad_norm": 22.826797485351562, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 6.736, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.01505253334136135, | |
| "grad_norm": 21.844627380371094, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 7.6197, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.015172953608092243, | |
| "grad_norm": 19.098604202270508, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 7.1382, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.015293373874823133, | |
| "grad_norm": 31.23625373840332, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 9.1712, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.015413794141554023, | |
| "grad_norm": 18.707687377929688, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 5.156, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.015534214408284914, | |
| "grad_norm": 26.03555679321289, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 5.96, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.015654634675015804, | |
| "grad_norm": 18.275354385375977, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 5.7764, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.015775054941746695, | |
| "grad_norm": 17.39764404296875, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 5.5462, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.015895475208477585, | |
| "grad_norm": 15.206930160522461, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 5.3986, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.01601589547520848, | |
| "grad_norm": 15.486876487731934, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 5.3179, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.01613631574193937, | |
| "grad_norm": 14.003046035766602, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 4.9894, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.01625673600867026, | |
| "grad_norm": 17.335914611816406, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 5.3201, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.01637715627540115, | |
| "grad_norm": 16.772851943969727, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 4.6407, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.01649757654213204, | |
| "grad_norm": 15.80712890625, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 4.8431, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.01661799680886293, | |
| "grad_norm": 15.540848731994629, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 5.1288, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.01673841707559382, | |
| "grad_norm": 13.0451078414917, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 4.5057, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.016858837342324712, | |
| "grad_norm": 13.30695629119873, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 4.5, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.016979257609055606, | |
| "grad_norm": 14.346107482910156, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 4.9162, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.017099677875786496, | |
| "grad_norm": 14.035130500793457, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 4.2934, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.017220098142517386, | |
| "grad_norm": 17.546085357666016, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 4.9573, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.017340518409248277, | |
| "grad_norm": 15.321247100830078, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 5.0486, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.017460938675979167, | |
| "grad_norm": 19.09070587158203, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 5.0332, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.017581358942710058, | |
| "grad_norm": 14.127971649169922, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 4.0205, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.017701779209440948, | |
| "grad_norm": 14.474227905273438, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 4.2755, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.01782219947617184, | |
| "grad_norm": 16.422204971313477, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 4.6551, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.01794261974290273, | |
| "grad_norm": 29.208831787109375, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 6.5394, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.018063040009633623, | |
| "grad_norm": 21.595561981201172, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 5.2835, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.018063040009633623, | |
| "eval_loss": 1.4202370643615723, | |
| "eval_runtime": 1626.6636, | |
| "eval_samples_per_second": 8.599, | |
| "eval_steps_per_second": 2.15, | |
| "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": 3.3556941079629005e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |