| { | |
| "best_metric": 0.530606210231781, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-300", | |
| "epoch": 0.068575347162695, | |
| "eval_steps": 50, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0002285844905423167, | |
| "grad_norm": 12.226268768310547, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 3.9387, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0002285844905423167, | |
| "eval_loss": 4.918857097625732, | |
| "eval_runtime": 1326.574, | |
| "eval_samples_per_second": 5.554, | |
| "eval_steps_per_second": 2.777, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0004571689810846334, | |
| "grad_norm": 16.508821487426758, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 4.288, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0006857534716269502, | |
| "grad_norm": 19.46694564819336, | |
| "learning_rate": 1e-05, | |
| "loss": 4.2996, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0009143379621692668, | |
| "grad_norm": 16.385190963745117, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 3.5999, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0011429224527115835, | |
| "grad_norm": 19.315223693847656, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 3.9415, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0013715069432539003, | |
| "grad_norm": 20.83527946472168, | |
| "learning_rate": 2e-05, | |
| "loss": 3.7644, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.001600091433796217, | |
| "grad_norm": 50.4894905090332, | |
| "learning_rate": 2.3333333333333336e-05, | |
| "loss": 5.543, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0018286759243385335, | |
| "grad_norm": 109.4534912109375, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 9.1326, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00205726041488085, | |
| "grad_norm": 209.9546356201172, | |
| "learning_rate": 3e-05, | |
| "loss": 11.0199, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.002285844905423167, | |
| "grad_norm": 126.74362182617188, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 6.5991, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.002514429395965484, | |
| "grad_norm": 62.07261657714844, | |
| "learning_rate": 3.6666666666666666e-05, | |
| "loss": 4.5192, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0027430138865078006, | |
| "grad_norm": 57.047672271728516, | |
| "learning_rate": 4e-05, | |
| "loss": 2.792, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.002971598377050117, | |
| "grad_norm": 40.239585876464844, | |
| "learning_rate": 4.3333333333333334e-05, | |
| "loss": 1.6243, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.003200182867592434, | |
| "grad_norm": 147.44810485839844, | |
| "learning_rate": 4.666666666666667e-05, | |
| "loss": 7.7768, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0034287673581347507, | |
| "grad_norm": 129.61868286132812, | |
| "learning_rate": 5e-05, | |
| "loss": 3.3601, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.003657351848677067, | |
| "grad_norm": 67.8729476928711, | |
| "learning_rate": 5.333333333333333e-05, | |
| "loss": 2.3788, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.003885936339219384, | |
| "grad_norm": 53.81801986694336, | |
| "learning_rate": 5.666666666666667e-05, | |
| "loss": 1.5561, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0041145208297617, | |
| "grad_norm": 108.63700866699219, | |
| "learning_rate": 6e-05, | |
| "loss": 3.4077, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.004343105320304018, | |
| "grad_norm": 92.20858764648438, | |
| "learning_rate": 6.333333333333333e-05, | |
| "loss": 5.118, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.004571689810846334, | |
| "grad_norm": 107.2628402709961, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 4.0284, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.00480027430138865, | |
| "grad_norm": 84.7047348022461, | |
| "learning_rate": 7e-05, | |
| "loss": 2.3014, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.005028858791930968, | |
| "grad_norm": 80.38473510742188, | |
| "learning_rate": 7.333333333333333e-05, | |
| "loss": 2.9712, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.005257443282473284, | |
| "grad_norm": 58.196502685546875, | |
| "learning_rate": 7.666666666666667e-05, | |
| "loss": 3.1879, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.005486027773015601, | |
| "grad_norm": 217.900146484375, | |
| "learning_rate": 8e-05, | |
| "loss": 10.9929, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.005714612263557918, | |
| "grad_norm": 137.6026153564453, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 5.2346, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.005943196754100234, | |
| "grad_norm": 67.36244201660156, | |
| "learning_rate": 8.666666666666667e-05, | |
| "loss": 1.8588, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.006171781244642551, | |
| "grad_norm": 53.583919525146484, | |
| "learning_rate": 9e-05, | |
| "loss": 1.208, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.006400365735184868, | |
| "grad_norm": 247.46914672851562, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 23.0121, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.006628950225727184, | |
| "grad_norm": 309.0799255371094, | |
| "learning_rate": 9.666666666666667e-05, | |
| "loss": 11.6811, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.006857534716269501, | |
| "grad_norm": 88.57025146484375, | |
| "learning_rate": 0.0001, | |
| "loss": 6.6625, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.007086119206811818, | |
| "grad_norm": 37.52093505859375, | |
| "learning_rate": 9.999661540018812e-05, | |
| "loss": 5.0423, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.007314703697354134, | |
| "grad_norm": 59.48794937133789, | |
| "learning_rate": 9.998646205897309e-05, | |
| "loss": 4.7137, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.007543288187896451, | |
| "grad_norm": 31.38121223449707, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 2.6695, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.007771872678438768, | |
| "grad_norm": 45.99576950073242, | |
| "learning_rate": 9.994585556692624e-05, | |
| "loss": 3.0682, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.008000457168981085, | |
| "grad_norm": 26.696224212646484, | |
| "learning_rate": 9.991540791356342e-05, | |
| "loss": 3.2612, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0082290416595234, | |
| "grad_norm": 46.0620002746582, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 3.0265, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.008457626150065718, | |
| "grad_norm": 28.95342254638672, | |
| "learning_rate": 9.983424440222531e-05, | |
| "loss": 3.1935, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.008686210640608035, | |
| "grad_norm": 20.682689666748047, | |
| "learning_rate": 9.978353953249022e-05, | |
| "loss": 2.1278, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.00891479513115035, | |
| "grad_norm": 95.37089538574219, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 4.2482, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.009143379621692668, | |
| "grad_norm": 36.080413818359375, | |
| "learning_rate": 9.966191788709716e-05, | |
| "loss": 3.0929, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.009371964112234985, | |
| "grad_norm": 42.082088470458984, | |
| "learning_rate": 9.959101757706308e-05, | |
| "loss": 2.736, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.0096005486027773, | |
| "grad_norm": 22.620975494384766, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 3.1542, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.009829133093319618, | |
| "grad_norm": 31.406814575195312, | |
| "learning_rate": 9.942908597485558e-05, | |
| "loss": 3.6269, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.010057717583861935, | |
| "grad_norm": 25.586286544799805, | |
| "learning_rate": 9.933807660562898e-05, | |
| "loss": 6.2693, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.010286302074404252, | |
| "grad_norm": 29.92267608642578, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 7.0683, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.010514886564946568, | |
| "grad_norm": 20.818403244018555, | |
| "learning_rate": 9.913603233532067e-05, | |
| "loss": 6.0885, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.010743471055488885, | |
| "grad_norm": 20.533842086791992, | |
| "learning_rate": 9.902502478779896e-05, | |
| "loss": 5.4988, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.010972055546031203, | |
| "grad_norm": 16.80043601989746, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 4.4107, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.011200640036573518, | |
| "grad_norm": 13.855122566223145, | |
| "learning_rate": 9.878311400921072e-05, | |
| "loss": 4.2749, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.011429224527115835, | |
| "grad_norm": 21.536596298217773, | |
| "learning_rate": 9.865224352899119e-05, | |
| "loss": 5.1643, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.011429224527115835, | |
| "eval_loss": 2.806919813156128, | |
| "eval_runtime": 1332.9908, | |
| "eval_samples_per_second": 5.527, | |
| "eval_steps_per_second": 2.764, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.011657809017658153, | |
| "grad_norm": 15.51828670501709, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 4.3807, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.011886393508200468, | |
| "grad_norm": 11.271062850952148, | |
| "learning_rate": 9.837076097314319e-05, | |
| "loss": 4.1682, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.012114977998742785, | |
| "grad_norm": 8.433566093444824, | |
| "learning_rate": 9.822018700574695e-05, | |
| "loss": 3.3631, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.012343562489285103, | |
| "grad_norm": 8.091526985168457, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 2.9425, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.012572146979827418, | |
| "grad_norm": 8.342974662780762, | |
| "learning_rate": 9.789947561577445e-05, | |
| "loss": 2.9333, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.012800731470369735, | |
| "grad_norm": 17.649492263793945, | |
| "learning_rate": 9.77293816123866e-05, | |
| "loss": 3.3319, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.013029315960912053, | |
| "grad_norm": 28.43054962158203, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 3.0355, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.013257900451454368, | |
| "grad_norm": 28.488826751708984, | |
| "learning_rate": 9.736983212571646e-05, | |
| "loss": 2.6905, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.013486484941996685, | |
| "grad_norm": 68.5689926147461, | |
| "learning_rate": 9.718042531967918e-05, | |
| "loss": 2.8557, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.013715069432539003, | |
| "grad_norm": 7.864617347717285, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.1518, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.013943653923081318, | |
| "grad_norm": 4.028836727142334, | |
| "learning_rate": 9.678247579197657e-05, | |
| "loss": 0.0393, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.014172238413623636, | |
| "grad_norm": 78.4482192993164, | |
| "learning_rate": 9.657398694630712e-05, | |
| "loss": 0.2168, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.014400822904165953, | |
| "grad_norm": 212.4618682861328, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 3.4277, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.014629407394708268, | |
| "grad_norm": 142.9414825439453, | |
| "learning_rate": 9.613812221777212e-05, | |
| "loss": 11.2092, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.014857991885250586, | |
| "grad_norm": 106.03768157958984, | |
| "learning_rate": 9.591080534401371e-05, | |
| "loss": 3.2856, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.015086576375792903, | |
| "grad_norm": 130.8965606689453, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 3.508, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.015315160866335218, | |
| "grad_norm": 86.60301208496094, | |
| "learning_rate": 9.543755644867822e-05, | |
| "loss": 2.7746, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.015543745356877536, | |
| "grad_norm": 62.65277862548828, | |
| "learning_rate": 9.519168849742604e-05, | |
| "loss": 1.4813, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.01577232984741985, | |
| "grad_norm": 44.15681076049805, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 3.2382, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.01600091433796217, | |
| "grad_norm": 54.2675666809082, | |
| "learning_rate": 9.468163201617062e-05, | |
| "loss": 3.0064, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.016229498828504486, | |
| "grad_norm": 20.503276824951172, | |
| "learning_rate": 9.441751253965021e-05, | |
| "loss": 1.6639, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.0164580833190468, | |
| "grad_norm": 42.72779846191406, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 1.9632, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.01668666780958912, | |
| "grad_norm": 94.91614532470703, | |
| "learning_rate": 9.38712698977291e-05, | |
| "loss": 7.9977, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.016915252300131436, | |
| "grad_norm": 120.97164154052734, | |
| "learning_rate": 9.358922068483812e-05, | |
| "loss": 6.0946, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.01714383679067375, | |
| "grad_norm": 65.04187774658203, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 2.1872, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.01737242128121607, | |
| "grad_norm": 102.3792495727539, | |
| "learning_rate": 9.300745739476829e-05, | |
| "loss": 1.955, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.017601005771758386, | |
| "grad_norm": 78.6114501953125, | |
| "learning_rate": 9.270782207902629e-05, | |
| "loss": 1.5882, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.0178295902623007, | |
| "grad_norm": 112.37207794189453, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 9.1085, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.01805817475284302, | |
| "grad_norm": 123.03736114501953, | |
| "learning_rate": 9.209124692976287e-05, | |
| "loss": 12.134, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.018286759243385336, | |
| "grad_norm": 64.75898742675781, | |
| "learning_rate": 9.177439057064683e-05, | |
| "loss": 8.9277, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01851534373392765, | |
| "grad_norm": 35.20903015136719, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 5.1302, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.01874392822446997, | |
| "grad_norm": 62.34042739868164, | |
| "learning_rate": 9.112375476403312e-05, | |
| "loss": 5.2516, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.018972512715012286, | |
| "grad_norm": 45.33003616333008, | |
| "learning_rate": 9.079006340220862e-05, | |
| "loss": 4.3035, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.0192010972055546, | |
| "grad_norm": 71.29220581054688, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 4.2392, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.01942968169609692, | |
| "grad_norm": 19.70421600341797, | |
| "learning_rate": 9.01061596377522e-05, | |
| "loss": 3.9666, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.019658266186639236, | |
| "grad_norm": 14.824080467224121, | |
| "learning_rate": 8.97560398247424e-05, | |
| "loss": 2.4165, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.019886850677181555, | |
| "grad_norm": 14.691969871520996, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 2.522, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.02011543516772387, | |
| "grad_norm": 20.41330337524414, | |
| "learning_rate": 8.903970133383297e-05, | |
| "loss": 2.6812, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.020344019658266186, | |
| "grad_norm": 16.38035774230957, | |
| "learning_rate": 8.86735796366982e-05, | |
| "loss": 1.7804, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.020572604148808505, | |
| "grad_norm": 14.719985008239746, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 2.0308, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02080118863935082, | |
| "grad_norm": 21.014904022216797, | |
| "learning_rate": 8.792567916744346e-05, | |
| "loss": 2.2666, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.021029773129893136, | |
| "grad_norm": 27.58352279663086, | |
| "learning_rate": 8.754400164907497e-05, | |
| "loss": 3.785, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.021258357620435455, | |
| "grad_norm": 24.515430450439453, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 5.0201, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.02148694211097777, | |
| "grad_norm": 28.423784255981445, | |
| "learning_rate": 8.676545040299145e-05, | |
| "loss": 6.3986, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.021715526601520086, | |
| "grad_norm": 19.08306312561035, | |
| "learning_rate": 8.636868207865244e-05, | |
| "loss": 6.0112, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.021944111092062405, | |
| "grad_norm": 18.091100692749023, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 4.6697, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.02217269558260472, | |
| "grad_norm": 16.675926208496094, | |
| "learning_rate": 8.556042860050687e-05, | |
| "loss": 5.068, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.022401280073147036, | |
| "grad_norm": 13.843223571777344, | |
| "learning_rate": 8.51490528712831e-05, | |
| "loss": 4.7174, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.022629864563689355, | |
| "grad_norm": 11.593822479248047, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 3.1727, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.02285844905423167, | |
| "grad_norm": 14.757932662963867, | |
| "learning_rate": 8.43120818934367e-05, | |
| "loss": 4.6322, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02285844905423167, | |
| "eval_loss": 2.1855103969573975, | |
| "eval_runtime": 1333.7195, | |
| "eval_samples_per_second": 5.524, | |
| "eval_steps_per_second": 2.762, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.023087033544773986, | |
| "grad_norm": 14.086247444152832, | |
| "learning_rate": 8.388659995728663e-05, | |
| "loss": 4.5384, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.023315618035316305, | |
| "grad_norm": 10.055537223815918, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 4.1292, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.02354420252585862, | |
| "grad_norm": 7.018960952758789, | |
| "learning_rate": 8.302193119995039e-05, | |
| "loss": 3.673, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.023772787016400936, | |
| "grad_norm": 7.652724266052246, | |
| "learning_rate": 8.258286144107276e-05, | |
| "loss": 3.1104, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.024001371506943255, | |
| "grad_norm": 9.536213874816895, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 2.8206, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.02422995599748557, | |
| "grad_norm": 9.198083877563477, | |
| "learning_rate": 8.169154836993551e-05, | |
| "loss": 2.8642, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.024458540488027886, | |
| "grad_norm": 14.823854446411133, | |
| "learning_rate": 8.1239425727198e-05, | |
| "loss": 2.7631, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.024687124978570205, | |
| "grad_norm": 44.76399612426758, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 2.9486, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.02491570946911252, | |
| "grad_norm": 29.669414520263672, | |
| "learning_rate": 8.032255426994069e-05, | |
| "loss": 2.7715, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.025144293959654836, | |
| "grad_norm": 60.78176498413086, | |
| "learning_rate": 7.985792958513931e-05, | |
| "loss": 2.6962, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.025372878450197155, | |
| "grad_norm": 17.208572387695312, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.353, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.02560146294073947, | |
| "grad_norm": 3.3832650184631348, | |
| "learning_rate": 7.891661680839932e-05, | |
| "loss": 0.0653, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.025830047431281786, | |
| "grad_norm": 7.614874839782715, | |
| "learning_rate": 7.844005615514259e-05, | |
| "loss": 0.0529, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.026058631921824105, | |
| "grad_norm": 56.453670501708984, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 3.9549, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.02628721641236642, | |
| "grad_norm": 91.68329620361328, | |
| "learning_rate": 7.74754489035403e-05, | |
| "loss": 4.0682, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.026515800902908736, | |
| "grad_norm": 57.930580139160156, | |
| "learning_rate": 7.698753289757565e-05, | |
| "loss": 1.4162, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.026744385393451055, | |
| "grad_norm": 112.63847351074219, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 1.2008, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.02697296988399337, | |
| "grad_norm": 149.06593322753906, | |
| "learning_rate": 7.600080639646077e-05, | |
| "loss": 1.9499, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.027201554374535687, | |
| "grad_norm": 399.54156494140625, | |
| "learning_rate": 7.550212948828377e-05, | |
| "loss": 2.0733, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.027430138865078006, | |
| "grad_norm": 148.1618194580078, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 4.0831, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.02765872335562032, | |
| "grad_norm": 76.7452621459961, | |
| "learning_rate": 7.449448591190435e-05, | |
| "loss": 3.1251, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.027887307846162637, | |
| "grad_norm": 71.67847442626953, | |
| "learning_rate": 7.398565566251232e-05, | |
| "loss": 2.4827, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.028115892336704956, | |
| "grad_norm": 23.501968383789062, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 1.5631, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.02834447682724727, | |
| "grad_norm": 136.9478302001953, | |
| "learning_rate": 7.295832266935059e-05, | |
| "loss": 7.535, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.028573061317789587, | |
| "grad_norm": 91.296142578125, | |
| "learning_rate": 7.243995901002312e-05, | |
| "loss": 3.9653, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.028801645808331906, | |
| "grad_norm": 48.54848861694336, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 2.1426, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.02903023029887422, | |
| "grad_norm": 202.33116149902344, | |
| "learning_rate": 7.139418824708272e-05, | |
| "loss": 2.8557, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.029258814789416537, | |
| "grad_norm": 1095.19873046875, | |
| "learning_rate": 7.08669227240909e-05, | |
| "loss": 4.8724, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.029487399279958856, | |
| "grad_norm": 1942.80859375, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 20.7307, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.02971598377050117, | |
| "grad_norm": 439.0259094238281, | |
| "learning_rate": 6.980398830195785e-05, | |
| "loss": 15.7319, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.029944568261043487, | |
| "grad_norm": 145.61013793945312, | |
| "learning_rate": 6.926846330712242e-05, | |
| "loss": 9.0764, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.030173152751585806, | |
| "grad_norm": 65.11675262451172, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 7.8135, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.03040173724212812, | |
| "grad_norm": 371.9282531738281, | |
| "learning_rate": 6.818966024765758e-05, | |
| "loss": 7.8735, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.030630321732670437, | |
| "grad_norm": 113.37095642089844, | |
| "learning_rate": 6.764652823569344e-05, | |
| "loss": 5.4435, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.030858906223212756, | |
| "grad_norm": 152.44851684570312, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 4.9697, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.03108749071375507, | |
| "grad_norm": 29.180082321166992, | |
| "learning_rate": 6.65531708942479e-05, | |
| "loss": 3.2838, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.03131607520429739, | |
| "grad_norm": 26.18000030517578, | |
| "learning_rate": 6.600309358784857e-05, | |
| "loss": 3.4387, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.0315446596948397, | |
| "grad_norm": 39.24956512451172, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.3899, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.03177324418538202, | |
| "grad_norm": 27.828189849853516, | |
| "learning_rate": 6.48965140519241e-05, | |
| "loss": 2.435, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.03200182867592434, | |
| "grad_norm": 28.83315086364746, | |
| "learning_rate": 6.434016163555452e-05, | |
| "loss": 3.0311, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03223041316646665, | |
| "grad_norm": 23.669389724731445, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 2.4059, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.03245899765700897, | |
| "grad_norm": 13.909722328186035, | |
| "learning_rate": 6.322170810186012e-05, | |
| "loss": 1.7042, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.03268758214755129, | |
| "grad_norm": 14.179429054260254, | |
| "learning_rate": 6.26597584052401e-05, | |
| "loss": 3.3267, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.0329161666380936, | |
| "grad_norm": 17.85869026184082, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 5.1052, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.03314475112863592, | |
| "grad_norm": 23.01960563659668, | |
| "learning_rate": 6.153079353712201e-05, | |
| "loss": 6.2258, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.03337333561917824, | |
| "grad_norm": 17.689586639404297, | |
| "learning_rate": 6.096393120939516e-05, | |
| "loss": 5.5296, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.03360192010972055, | |
| "grad_norm": 15.267470359802246, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 5.1839, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.03383050460026287, | |
| "grad_norm": 13.088833808898926, | |
| "learning_rate": 5.982583047664151e-05, | |
| "loss": 4.1851, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.03405908909080519, | |
| "grad_norm": 13.637064933776855, | |
| "learning_rate": 5.925474615223573e-05, | |
| "loss": 4.2342, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.0342876735813475, | |
| "grad_norm": 11.020079612731934, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 4.5138, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0342876735813475, | |
| "eval_loss": 1.670865774154663, | |
| "eval_runtime": 1333.0768, | |
| "eval_samples_per_second": 5.527, | |
| "eval_steps_per_second": 2.764, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.03451625807188982, | |
| "grad_norm": 15.293791770935059, | |
| "learning_rate": 5.810889615527838e-05, | |
| "loss": 4.6994, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.03474484256243214, | |
| "grad_norm": 9.839681625366211, | |
| "learning_rate": 5.753428561247416e-05, | |
| "loss": 3.8388, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.03497342705297445, | |
| "grad_norm": 7.527024745941162, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 3.2164, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.03520201154351677, | |
| "grad_norm": 8.136163711547852, | |
| "learning_rate": 5.6382082393029746e-05, | |
| "loss": 3.0072, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.03543059603405909, | |
| "grad_norm": 7.6138691902160645, | |
| "learning_rate": 5.5804645706261514e-05, | |
| "loss": 2.8556, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.0356591805246014, | |
| "grad_norm": 23.783597946166992, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 2.5473, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.03588776501514372, | |
| "grad_norm": 20.85479164123535, | |
| "learning_rate": 5.464749304646962e-05, | |
| "loss": 2.6145, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.03611634950568604, | |
| "grad_norm": 24.58341407775879, | |
| "learning_rate": 5.4067933733392915e-05, | |
| "loss": 2.7502, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.03634493399622835, | |
| "grad_norm": 30.31068992614746, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 1.2587, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.03657351848677067, | |
| "grad_norm": 9.1329984664917, | |
| "learning_rate": 5.290724144552379e-05, | |
| "loss": 0.1588, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.03680210297731299, | |
| "grad_norm": 30.12510108947754, | |
| "learning_rate": 5.232626560988735e-05, | |
| "loss": 0.4367, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.0370306874678553, | |
| "grad_norm": 21.02296257019043, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.4107, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.03725927195839762, | |
| "grad_norm": 39.78958511352539, | |
| "learning_rate": 5.116344781870281e-05, | |
| "loss": 2.3939, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.03748785644893994, | |
| "grad_norm": 89.31826782226562, | |
| "learning_rate": 5.0581763290069865e-05, | |
| "loss": 6.3699, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.03771644093948225, | |
| "grad_norm": 50.574974060058594, | |
| "learning_rate": 5e-05, | |
| "loss": 3.5648, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.03794502543002457, | |
| "grad_norm": 30.325014114379883, | |
| "learning_rate": 4.941823670993016e-05, | |
| "loss": 2.0516, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.03817360992056689, | |
| "grad_norm": 19.043434143066406, | |
| "learning_rate": 4.883655218129719e-05, | |
| "loss": 1.2083, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.0384021944111092, | |
| "grad_norm": 29.482452392578125, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 2.1119, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.03863077890165152, | |
| "grad_norm": 37.51374816894531, | |
| "learning_rate": 4.767373439011267e-05, | |
| "loss": 2.4501, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.03885936339219384, | |
| "grad_norm": 23.2747745513916, | |
| "learning_rate": 4.709275855447621e-05, | |
| "loss": 2.0485, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.03908794788273615, | |
| "grad_norm": 15.753173828125, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 1.2318, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.03931653237327847, | |
| "grad_norm": 26.802000045776367, | |
| "learning_rate": 4.593206626660709e-05, | |
| "loss": 1.7534, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.03954511686382079, | |
| "grad_norm": 94.17913818359375, | |
| "learning_rate": 4.535250695353039e-05, | |
| "loss": 6.9058, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.03977370135436311, | |
| "grad_norm": 129.48439025878906, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 7.7345, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.04000228584490542, | |
| "grad_norm": 83.24310302734375, | |
| "learning_rate": 4.4195354293738484e-05, | |
| "loss": 5.7297, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.04023087033544774, | |
| "grad_norm": 116.52914428710938, | |
| "learning_rate": 4.361791760697027e-05, | |
| "loss": 4.7677, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.04045945482599006, | |
| "grad_norm": 39.86652374267578, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 1.596, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.04068803931653237, | |
| "grad_norm": 14.975143432617188, | |
| "learning_rate": 4.246571438752585e-05, | |
| "loss": 1.2558, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.04091662380707469, | |
| "grad_norm": 86.05208587646484, | |
| "learning_rate": 4.1891103844721636e-05, | |
| "loss": 8.7519, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.04114520829761701, | |
| "grad_norm": 40.95692825317383, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 6.2332, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.04137379278815932, | |
| "grad_norm": 82.19147491455078, | |
| "learning_rate": 4.074525384776428e-05, | |
| "loss": 7.8078, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.04160237727870164, | |
| "grad_norm": 46.78995132446289, | |
| "learning_rate": 4.017416952335849e-05, | |
| "loss": 5.3358, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.04183096176924396, | |
| "grad_norm": 23.36475372314453, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 4.3148, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.04205954625978627, | |
| "grad_norm": 25.982982635498047, | |
| "learning_rate": 3.903606879060483e-05, | |
| "loss": 2.5694, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.04228813075032859, | |
| "grad_norm": 16.722654342651367, | |
| "learning_rate": 3.846920646287799e-05, | |
| "loss": 2.6081, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.04251671524087091, | |
| "grad_norm": 18.733665466308594, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 2.1414, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.04274529973141322, | |
| "grad_norm": 22.727380752563477, | |
| "learning_rate": 3.7340241594759916e-05, | |
| "loss": 3.6351, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.04297388422195554, | |
| "grad_norm": 16.99645233154297, | |
| "learning_rate": 3.67782918981399e-05, | |
| "loss": 2.4938, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.04320246871249786, | |
| "grad_norm": 27.48732566833496, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 2.0775, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.04343105320304017, | |
| "grad_norm": 15.861635208129883, | |
| "learning_rate": 3.5659838364445505e-05, | |
| "loss": 2.5506, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.04365963769358249, | |
| "grad_norm": 25.20842933654785, | |
| "learning_rate": 3.51034859480759e-05, | |
| "loss": 2.6434, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.04388822218412481, | |
| "grad_norm": 28.7867374420166, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 4.0947, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.04411680667466712, | |
| "grad_norm": 16.924789428710938, | |
| "learning_rate": 3.399690641215142e-05, | |
| "loss": 4.607, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.04434539116520944, | |
| "grad_norm": 19.492359161376953, | |
| "learning_rate": 3.34468291057521e-05, | |
| "loss": 5.7136, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.04457397565575176, | |
| "grad_norm": 20.93442726135254, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 5.2217, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.04480256014629407, | |
| "grad_norm": 15.588977813720703, | |
| "learning_rate": 3.235347176430656e-05, | |
| "loss": 5.381, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.04503114463683639, | |
| "grad_norm": 13.808337211608887, | |
| "learning_rate": 3.1810339752342446e-05, | |
| "loss": 5.4432, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.04525972912737871, | |
| "grad_norm": 14.99952507019043, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 4.5804, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.04548831361792102, | |
| "grad_norm": 13.91661548614502, | |
| "learning_rate": 3.073153669287759e-05, | |
| "loss": 4.1034, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.04571689810846334, | |
| "grad_norm": 13.113777160644531, | |
| "learning_rate": 3.019601169804216e-05, | |
| "loss": 4.2222, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.04571689810846334, | |
| "eval_loss": 1.62479567527771, | |
| "eval_runtime": 1331.6429, | |
| "eval_samples_per_second": 5.533, | |
| "eval_steps_per_second": 2.767, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.04594548259900566, | |
| "grad_norm": 10.6249418258667, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 3.5265, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.04617406708954797, | |
| "grad_norm": 8.47287368774414, | |
| "learning_rate": 2.9133077275909108e-05, | |
| "loss": 3.4879, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.04640265158009029, | |
| "grad_norm": 7.781945705413818, | |
| "learning_rate": 2.86058117529173e-05, | |
| "loss": 3.0601, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.04663123607063261, | |
| "grad_norm": 7.176991939544678, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 2.6794, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.04685982056117492, | |
| "grad_norm": 6.956686496734619, | |
| "learning_rate": 2.7560040989976892e-05, | |
| "loss": 2.3741, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.04708840505171724, | |
| "grad_norm": 7.541036605834961, | |
| "learning_rate": 2.7041677330649407e-05, | |
| "loss": 2.4602, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.04731698954225956, | |
| "grad_norm": 26.17896270751953, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 1.7172, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.04754557403280187, | |
| "grad_norm": 44.718421936035156, | |
| "learning_rate": 2.6014344337487707e-05, | |
| "loss": 1.8271, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.04777415852334419, | |
| "grad_norm": 23.585582733154297, | |
| "learning_rate": 2.5505514088095655e-05, | |
| "loss": 1.1668, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.04800274301388651, | |
| "grad_norm": 18.206127166748047, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.7469, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.04823132750442882, | |
| "grad_norm": 4.393265724182129, | |
| "learning_rate": 2.4497870511716235e-05, | |
| "loss": 0.1077, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.04845991199497114, | |
| "grad_norm": 4.106040954589844, | |
| "learning_rate": 2.399919360353923e-05, | |
| "loss": 0.2907, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.04868849648551346, | |
| "grad_norm": 13.828383445739746, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 0.8737, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.04891708097605577, | |
| "grad_norm": 59.581298828125, | |
| "learning_rate": 2.3012467102424373e-05, | |
| "loss": 4.8994, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.04914566546659809, | |
| "grad_norm": 52.219642639160156, | |
| "learning_rate": 2.25245510964597e-05, | |
| "loss": 4.3086, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.04937424995714041, | |
| "grad_norm": 57.01164627075195, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 3.0226, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.04960283444768272, | |
| "grad_norm": 35.1441764831543, | |
| "learning_rate": 2.1559943844857422e-05, | |
| "loss": 1.8326, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.04983141893822504, | |
| "grad_norm": 23.194089889526367, | |
| "learning_rate": 2.1083383191600674e-05, | |
| "loss": 0.9515, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.05006000342876736, | |
| "grad_norm": 20.370725631713867, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.2338, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.05028858791930967, | |
| "grad_norm": 31.7033634185791, | |
| "learning_rate": 2.0142070414860704e-05, | |
| "loss": 2.838, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.05051717240985199, | |
| "grad_norm": 30.212299346923828, | |
| "learning_rate": 1.9677445730059346e-05, | |
| "loss": 3.5437, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.05074575690039431, | |
| "grad_norm": 24.48371124267578, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 2.1516, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.05097434139093662, | |
| "grad_norm": 26.28635597229004, | |
| "learning_rate": 1.8760574272802e-05, | |
| "loss": 1.8107, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.05120292588147894, | |
| "grad_norm": 64.8018569946289, | |
| "learning_rate": 1.8308451630064484e-05, | |
| "loss": 3.8766, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.05143151037202126, | |
| "grad_norm": 79.22872161865234, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 5.5396, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.05166009486256357, | |
| "grad_norm": 81.44451904296875, | |
| "learning_rate": 1.7417138558927244e-05, | |
| "loss": 5.3564, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.05188867935310589, | |
| "grad_norm": 74.99747467041016, | |
| "learning_rate": 1.6978068800049624e-05, | |
| "loss": 4.2999, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.05211726384364821, | |
| "grad_norm": 30.039758682250977, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 4.5715, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.05234584833419052, | |
| "grad_norm": 25.890417098999023, | |
| "learning_rate": 1.611340004271339e-05, | |
| "loss": 4.2147, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.05257443282473284, | |
| "grad_norm": 24.455820083618164, | |
| "learning_rate": 1.5687918106563326e-05, | |
| "loss": 3.6963, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.05280301731527516, | |
| "grad_norm": 13.776949882507324, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 2.5523, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.05303160180581747, | |
| "grad_norm": 14.489748001098633, | |
| "learning_rate": 1.4850947128716913e-05, | |
| "loss": 2.8481, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.05326018629635979, | |
| "grad_norm": 14.851402282714844, | |
| "learning_rate": 1.4439571399493146e-05, | |
| "loss": 2.9429, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.05348877078690211, | |
| "grad_norm": 14.959819793701172, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 2.7784, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.05371735527744442, | |
| "grad_norm": 10.266538619995117, | |
| "learning_rate": 1.3631317921347563e-05, | |
| "loss": 2.1112, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.05394593976798674, | |
| "grad_norm": 13.498571395874023, | |
| "learning_rate": 1.3234549597008571e-05, | |
| "loss": 2.61, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.05417452425852906, | |
| "grad_norm": 13.794647216796875, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 2.655, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.05440310874907137, | |
| "grad_norm": 12.700047492980957, | |
| "learning_rate": 1.245599835092504e-05, | |
| "loss": 2.1982, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.05463169323961369, | |
| "grad_norm": 11.33856201171875, | |
| "learning_rate": 1.2074320832556556e-05, | |
| "loss": 1.8852, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.05486027773015601, | |
| "grad_norm": 12.628767013549805, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 2.6147, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.05508886222069832, | |
| "grad_norm": 12.457136154174805, | |
| "learning_rate": 1.132642036330181e-05, | |
| "loss": 2.2797, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.05531744671124064, | |
| "grad_norm": 16.78705406188965, | |
| "learning_rate": 1.096029866616704e-05, | |
| "loss": 2.1772, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.05554603120178296, | |
| "grad_norm": 20.40379524230957, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 4.4865, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.05577461569232527, | |
| "grad_norm": 47.940460205078125, | |
| "learning_rate": 1.0243960175257606e-05, | |
| "loss": 4.1248, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.05600320018286759, | |
| "grad_norm": 20.962236404418945, | |
| "learning_rate": 9.893840362247809e-06, | |
| "loss": 5.0609, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.05623178467340991, | |
| "grad_norm": 17.944730758666992, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 5.8087, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.05646036916395222, | |
| "grad_norm": 14.370197296142578, | |
| "learning_rate": 9.209936597791407e-06, | |
| "loss": 4.9449, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.05668895365449454, | |
| "grad_norm": 13.268957138061523, | |
| "learning_rate": 8.876245235966885e-06, | |
| "loss": 4.4975, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.05691753814503686, | |
| "grad_norm": 13.039083480834961, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 3.907, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.05714612263557917, | |
| "grad_norm": 14.2072172164917, | |
| "learning_rate": 8.225609429353187e-06, | |
| "loss": 3.7406, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.05714612263557917, | |
| "eval_loss": 0.5557529926300049, | |
| "eval_runtime": 1331.7964, | |
| "eval_samples_per_second": 5.532, | |
| "eval_steps_per_second": 2.766, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.05737470712612149, | |
| "grad_norm": 31.539043426513672, | |
| "learning_rate": 7.908753070237123e-06, | |
| "loss": 4.1344, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.05760329161666381, | |
| "grad_norm": 10.60650634765625, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 3.037, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.05783187610720612, | |
| "grad_norm": 5.5060715675354, | |
| "learning_rate": 7.292177920973725e-06, | |
| "loss": 2.9713, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.05806046059774844, | |
| "grad_norm": 5.154425144195557, | |
| "learning_rate": 6.992542605231739e-06, | |
| "loss": 2.2515, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.05828904508829076, | |
| "grad_norm": 5.6335129737854, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 2.3608, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.05851762957883307, | |
| "grad_norm": 6.264554023742676, | |
| "learning_rate": 6.410779315161886e-06, | |
| "loss": 2.2597, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.05874621406937539, | |
| "grad_norm": 9.82129955291748, | |
| "learning_rate": 6.128730102270897e-06, | |
| "loss": 1.9233, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.05897479855991771, | |
| "grad_norm": 12.5000581741333, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 2.1487, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.05920338305046002, | |
| "grad_norm": 17.345787048339844, | |
| "learning_rate": 5.582487460349805e-06, | |
| "loss": 2.1851, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.05943196754100234, | |
| "grad_norm": 15.88868236541748, | |
| "learning_rate": 5.318367983829392e-06, | |
| "loss": 0.7703, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.05966055203154466, | |
| "grad_norm": 18.099214553833008, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 0.5869, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.05988913652208697, | |
| "grad_norm": 11.634132385253906, | |
| "learning_rate": 4.8083115025739756e-06, | |
| "loss": 0.375, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.06011772101262929, | |
| "grad_norm": 12.680456161499023, | |
| "learning_rate": 4.562443551321788e-06, | |
| "loss": 0.6281, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.06034630550317161, | |
| "grad_norm": 10.923900604248047, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 0.5998, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.06057488999371392, | |
| "grad_norm": 33.76637268066406, | |
| "learning_rate": 4.089194655986306e-06, | |
| "loss": 2.3875, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.06080347448425624, | |
| "grad_norm": 29.129016876220703, | |
| "learning_rate": 3.861877782227885e-06, | |
| "loss": 2.1657, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.06103205897479856, | |
| "grad_norm": 25.44561767578125, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 2.1342, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.061260643465340874, | |
| "grad_norm": 29.16153335571289, | |
| "learning_rate": 3.426013053692878e-06, | |
| "loss": 1.9011, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.06148922795588319, | |
| "grad_norm": 27.153125762939453, | |
| "learning_rate": 3.2175242080234313e-06, | |
| "loss": 1.852, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.06171781244642551, | |
| "grad_norm": 23.434755325317383, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 2.194, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.061946396936967824, | |
| "grad_norm": 31.765398025512695, | |
| "learning_rate": 2.8195746803208244e-06, | |
| "loss": 2.348, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.06217498142751014, | |
| "grad_norm": 20.402849197387695, | |
| "learning_rate": 2.63016787428354e-06, | |
| "loss": 2.2186, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.06240356591805246, | |
| "grad_norm": 43.326622009277344, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.9633, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.06263215040859478, | |
| "grad_norm": 55.71512222290039, | |
| "learning_rate": 2.2706183876134045e-06, | |
| "loss": 2.5231, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.0628607348991371, | |
| "grad_norm": 34.0222282409668, | |
| "learning_rate": 2.100524384225555e-06, | |
| "loss": 2.5434, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.0630893193896794, | |
| "grad_norm": 30.28105354309082, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 2.2744, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.06331790388022172, | |
| "grad_norm": 70.08159637451172, | |
| "learning_rate": 1.7798129942530551e-06, | |
| "loss": 2.8638, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.06354648837076404, | |
| "grad_norm": 30.30040740966797, | |
| "learning_rate": 1.6292390268568104e-06, | |
| "loss": 2.3145, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.06377507286130636, | |
| "grad_norm": 12.503889083862305, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 1.9287, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.06400365735184868, | |
| "grad_norm": 10.321149826049805, | |
| "learning_rate": 1.3477564710088098e-06, | |
| "loss": 1.7086, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.064232241842391, | |
| "grad_norm": 13.305763244628906, | |
| "learning_rate": 1.2168859907892904e-06, | |
| "loss": 2.1641, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.0644608263329333, | |
| "grad_norm": 11.012710571289062, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 2.1994, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.06468941082347562, | |
| "grad_norm": 11.165496826171875, | |
| "learning_rate": 9.749752122010346e-07, | |
| "loss": 1.8865, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.06491799531401794, | |
| "grad_norm": 8.645009994506836, | |
| "learning_rate": 8.639676646793382e-07, | |
| "loss": 1.4106, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.06514657980456026, | |
| "grad_norm": 12.307551383972168, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 1.9273, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.06537516429510258, | |
| "grad_norm": 11.830700874328613, | |
| "learning_rate": 6.61923394371039e-07, | |
| "loss": 2.3673, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.0656037487856449, | |
| "grad_norm": 10.462738037109375, | |
| "learning_rate": 5.7091402514442e-07, | |
| "loss": 1.6951, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.0658323332761872, | |
| "grad_norm": 11.972997665405273, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 2.0444, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.06606091776672952, | |
| "grad_norm": 18.008352279663086, | |
| "learning_rate": 4.089824229369155e-07, | |
| "loss": 2.4141, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.06628950225727184, | |
| "grad_norm": 11.071226119995117, | |
| "learning_rate": 3.380821129028489e-07, | |
| "loss": 1.8623, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.06651808674781416, | |
| "grad_norm": 15.886430740356445, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 2.2248, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.06674667123835648, | |
| "grad_norm": 14.11369800567627, | |
| "learning_rate": 2.1646046750978254e-07, | |
| "loss": 3.1426, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.0669752557288988, | |
| "grad_norm": 17.461822509765625, | |
| "learning_rate": 1.657555977746972e-07, | |
| "loss": 3.738, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.0672038402194411, | |
| "grad_norm": 9.49361801147461, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 3.9952, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.06743242470998342, | |
| "grad_norm": 11.396438598632812, | |
| "learning_rate": 8.459208643659122e-08, | |
| "loss": 4.4705, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.06766100920052574, | |
| "grad_norm": 10.36643123626709, | |
| "learning_rate": 5.4144433073771707e-08, | |
| "loss": 4.337, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.06788959369106806, | |
| "grad_norm": 11.684181213378906, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 4.1834, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.06811817818161038, | |
| "grad_norm": 8.283185958862305, | |
| "learning_rate": 1.3537941026914303e-08, | |
| "loss": 2.6821, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.0683467626721527, | |
| "grad_norm": 10.205315589904785, | |
| "learning_rate": 3.384599811889766e-09, | |
| "loss": 3.7379, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.068575347162695, | |
| "grad_norm": 11.798584938049316, | |
| "learning_rate": 0.0, | |
| "loss": 3.3263, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.068575347162695, | |
| "eval_loss": 0.530606210231781, | |
| "eval_runtime": 1331.9933, | |
| "eval_samples_per_second": 5.532, | |
| "eval_steps_per_second": 2.766, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.686338196994785e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |