diff --git "a/checkpoint-972/trainer_state.json" "b/checkpoint-972/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-972/trainer_state.json" @@ -0,0 +1,7117 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9908256880733946, + "eval_steps": 98, + "global_step": 972, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0010193679918450561, + "grad_norm": 24.08992576599121, + "learning_rate": 0.0, + "loss": 8.5798, + "step": 1 + }, + { + "epoch": 0.0020387359836901123, + "grad_norm": 22.663619995117188, + "learning_rate": 2.2624434389140275e-07, + "loss": 8.4222, + "step": 2 + }, + { + "epoch": 0.0030581039755351682, + "grad_norm": 25.377544403076172, + "learning_rate": 4.524886877828055e-07, + "loss": 9.392, + "step": 3 + }, + { + "epoch": 0.004077471967380225, + "grad_norm": 22.122257232666016, + "learning_rate": 6.787330316742082e-07, + "loss": 8.4193, + "step": 4 + }, + { + "epoch": 0.0050968399592252805, + "grad_norm": Infinity, + "learning_rate": 9.04977375565611e-07, + "loss": 8.808, + "step": 5 + }, + { + "epoch": 0.0061162079510703364, + "grad_norm": 22.832090377807617, + "learning_rate": 9.04977375565611e-07, + "loss": 10.5002, + "step": 6 + }, + { + "epoch": 0.007135575942915392, + "grad_norm": 20.933177947998047, + "learning_rate": 1.1312217194570136e-06, + "loss": 9.5956, + "step": 7 + }, + { + "epoch": 0.00815494393476045, + "grad_norm": 20.44132423400879, + "learning_rate": 1.3574660633484164e-06, + "loss": 8.5526, + "step": 8 + }, + { + "epoch": 0.009174311926605505, + "grad_norm": 22.630067825317383, + "learning_rate": 1.583710407239819e-06, + "loss": 9.8255, + "step": 9 + }, + { + "epoch": 0.010193679918450561, + "grad_norm": 22.625064849853516, + "learning_rate": 1.809954751131222e-06, + "loss": 9.2067, + "step": 10 + }, + { + "epoch": 0.011213047910295617, + "grad_norm": 19.427106857299805, + "learning_rate": 2.0361990950226245e-06, + "loss": 8.667, + "step": 11 + }, + { + "epoch": 0.012232415902140673, + "grad_norm": 24.354387283325195, + "learning_rate": 2.2624434389140273e-06, + "loss": 10.396, + "step": 12 + }, + { + "epoch": 0.013251783893985729, + "grad_norm": 21.99860191345215, + "learning_rate": 2.48868778280543e-06, + "loss": 8.9036, + "step": 13 + }, + { + "epoch": 0.014271151885830785, + "grad_norm": 21.275592803955078, + "learning_rate": 2.7149321266968327e-06, + "loss": 7.9307, + "step": 14 + }, + { + "epoch": 0.01529051987767584, + "grad_norm": 20.020435333251953, + "learning_rate": 2.9411764705882355e-06, + "loss": 9.0655, + "step": 15 + }, + { + "epoch": 0.0163098878695209, + "grad_norm": 20.713603973388672, + "learning_rate": 3.167420814479638e-06, + "loss": 8.6598, + "step": 16 + }, + { + "epoch": 0.017329255861365953, + "grad_norm": 22.857194900512695, + "learning_rate": 3.3936651583710405e-06, + "loss": 9.7562, + "step": 17 + }, + { + "epoch": 0.01834862385321101, + "grad_norm": 19.031551361083984, + "learning_rate": 3.619909502262444e-06, + "loss": 9.2297, + "step": 18 + }, + { + "epoch": 0.019367991845056064, + "grad_norm": 19.30624008178711, + "learning_rate": 3.846153846153847e-06, + "loss": 8.6939, + "step": 19 + }, + { + "epoch": 0.020387359836901122, + "grad_norm": 17.09296417236328, + "learning_rate": 4.072398190045249e-06, + "loss": 8.1317, + "step": 20 + }, + { + "epoch": 0.021406727828746176, + "grad_norm": 19.199600219726562, + "learning_rate": 4.298642533936651e-06, + "loss": 8.3585, + "step": 21 + }, + { + "epoch": 0.022426095820591234, + "grad_norm": 18.50484275817871, + "learning_rate": 4.5248868778280546e-06, + "loss": 8.4533, + "step": 22 + }, + { + "epoch": 0.023445463812436288, + "grad_norm": 19.170618057250977, + "learning_rate": 4.751131221719457e-06, + "loss": 9.3014, + "step": 23 + }, + { + "epoch": 0.024464831804281346, + "grad_norm": 17.692346572875977, + "learning_rate": 4.97737556561086e-06, + "loss": 8.18, + "step": 24 + }, + { + "epoch": 0.0254841997961264, + "grad_norm": 18.87356185913086, + "learning_rate": 5.203619909502263e-06, + "loss": 7.8485, + "step": 25 + }, + { + "epoch": 0.026503567787971458, + "grad_norm": 16.432092666625977, + "learning_rate": 5.4298642533936655e-06, + "loss": 8.9669, + "step": 26 + }, + { + "epoch": 0.027522935779816515, + "grad_norm": 17.064382553100586, + "learning_rate": 5.656108597285068e-06, + "loss": 10.1397, + "step": 27 + }, + { + "epoch": 0.02854230377166157, + "grad_norm": 17.96854591369629, + "learning_rate": 5.882352941176471e-06, + "loss": 10.5216, + "step": 28 + }, + { + "epoch": 0.029561671763506627, + "grad_norm": 16.348352432250977, + "learning_rate": 6.108597285067873e-06, + "loss": 7.4782, + "step": 29 + }, + { + "epoch": 0.03058103975535168, + "grad_norm": 15.834653854370117, + "learning_rate": 6.334841628959276e-06, + "loss": 8.0439, + "step": 30 + }, + { + "epoch": 0.03160040774719674, + "grad_norm": 15.115158081054688, + "learning_rate": 6.5610859728506795e-06, + "loss": 7.1911, + "step": 31 + }, + { + "epoch": 0.0326197757390418, + "grad_norm": 17.570573806762695, + "learning_rate": 6.787330316742081e-06, + "loss": 8.5735, + "step": 32 + }, + { + "epoch": 0.03363914373088685, + "grad_norm": 15.224530220031738, + "learning_rate": 7.013574660633485e-06, + "loss": 8.3855, + "step": 33 + }, + { + "epoch": 0.034658511722731905, + "grad_norm": 16.47282600402832, + "learning_rate": 7.239819004524888e-06, + "loss": 8.4305, + "step": 34 + }, + { + "epoch": 0.03567787971457696, + "grad_norm": 16.739215850830078, + "learning_rate": 7.46606334841629e-06, + "loss": 9.4608, + "step": 35 + }, + { + "epoch": 0.03669724770642202, + "grad_norm": 13.741637229919434, + "learning_rate": 7.692307692307694e-06, + "loss": 8.1572, + "step": 36 + }, + { + "epoch": 0.03771661569826707, + "grad_norm": 14.70285701751709, + "learning_rate": 7.918552036199094e-06, + "loss": 8.1456, + "step": 37 + }, + { + "epoch": 0.03873598369011213, + "grad_norm": 11.470185279846191, + "learning_rate": 8.144796380090498e-06, + "loss": 7.3833, + "step": 38 + }, + { + "epoch": 0.039755351681957186, + "grad_norm": 13.029812812805176, + "learning_rate": 8.3710407239819e-06, + "loss": 8.8539, + "step": 39 + }, + { + "epoch": 0.040774719673802244, + "grad_norm": 12.46716594696045, + "learning_rate": 8.597285067873303e-06, + "loss": 8.9349, + "step": 40 + }, + { + "epoch": 0.0417940876656473, + "grad_norm": 12.875706672668457, + "learning_rate": 8.823529411764707e-06, + "loss": 8.1803, + "step": 41 + }, + { + "epoch": 0.04281345565749235, + "grad_norm": 12.646770477294922, + "learning_rate": 9.049773755656109e-06, + "loss": 6.7532, + "step": 42 + }, + { + "epoch": 0.04383282364933741, + "grad_norm": 13.792744636535645, + "learning_rate": 9.276018099547511e-06, + "loss": 7.127, + "step": 43 + }, + { + "epoch": 0.04485219164118247, + "grad_norm": 11.656695365905762, + "learning_rate": 9.502262443438914e-06, + "loss": 7.5565, + "step": 44 + }, + { + "epoch": 0.045871559633027525, + "grad_norm": 11.562976837158203, + "learning_rate": 9.728506787330318e-06, + "loss": 7.6078, + "step": 45 + }, + { + "epoch": 0.046890927624872576, + "grad_norm": 11.516715049743652, + "learning_rate": 9.95475113122172e-06, + "loss": 8.4153, + "step": 46 + }, + { + "epoch": 0.047910295616717634, + "grad_norm": 11.569866180419922, + "learning_rate": 1.0180995475113122e-05, + "loss": 7.1062, + "step": 47 + }, + { + "epoch": 0.04892966360856269, + "grad_norm": 11.088666915893555, + "learning_rate": 1.0407239819004526e-05, + "loss": 6.8482, + "step": 48 + }, + { + "epoch": 0.04994903160040775, + "grad_norm": 11.396224021911621, + "learning_rate": 1.0633484162895929e-05, + "loss": 7.2262, + "step": 49 + }, + { + "epoch": 0.0509683995922528, + "grad_norm": 11.868388175964355, + "learning_rate": 1.0859728506787331e-05, + "loss": 8.0207, + "step": 50 + }, + { + "epoch": 0.05198776758409786, + "grad_norm": 10.022957801818848, + "learning_rate": 1.1085972850678733e-05, + "loss": 7.6895, + "step": 51 + }, + { + "epoch": 0.053007135575942915, + "grad_norm": 11.007475852966309, + "learning_rate": 1.1312217194570136e-05, + "loss": 7.6185, + "step": 52 + }, + { + "epoch": 0.05402650356778797, + "grad_norm": 10.026458740234375, + "learning_rate": 1.153846153846154e-05, + "loss": 8.8153, + "step": 53 + }, + { + "epoch": 0.05504587155963303, + "grad_norm": 10.358866691589355, + "learning_rate": 1.1764705882352942e-05, + "loss": 7.7666, + "step": 54 + }, + { + "epoch": 0.05606523955147808, + "grad_norm": 10.722491264343262, + "learning_rate": 1.1990950226244344e-05, + "loss": 7.1431, + "step": 55 + }, + { + "epoch": 0.05708460754332314, + "grad_norm": 10.623186111450195, + "learning_rate": 1.2217194570135746e-05, + "loss": 6.3969, + "step": 56 + }, + { + "epoch": 0.0581039755351682, + "grad_norm": 10.13591480255127, + "learning_rate": 1.244343891402715e-05, + "loss": 8.1643, + "step": 57 + }, + { + "epoch": 0.059123343527013254, + "grad_norm": 9.476139068603516, + "learning_rate": 1.2669683257918553e-05, + "loss": 7.1228, + "step": 58 + }, + { + "epoch": 0.060142711518858305, + "grad_norm": 8.608465194702148, + "learning_rate": 1.2895927601809957e-05, + "loss": 6.9228, + "step": 59 + }, + { + "epoch": 0.06116207951070336, + "grad_norm": 10.69497299194336, + "learning_rate": 1.3122171945701359e-05, + "loss": 10.2251, + "step": 60 + }, + { + "epoch": 0.06218144750254842, + "grad_norm": 9.309306144714355, + "learning_rate": 1.3348416289592761e-05, + "loss": 7.1105, + "step": 61 + }, + { + "epoch": 0.06320081549439348, + "grad_norm": 9.268863677978516, + "learning_rate": 1.3574660633484162e-05, + "loss": 7.1156, + "step": 62 + }, + { + "epoch": 0.06422018348623854, + "grad_norm": 10.207130432128906, + "learning_rate": 1.3800904977375568e-05, + "loss": 6.5522, + "step": 63 + }, + { + "epoch": 0.0652395514780836, + "grad_norm": 9.29359245300293, + "learning_rate": 1.402714932126697e-05, + "loss": 6.734, + "step": 64 + }, + { + "epoch": 0.06625891946992865, + "grad_norm": 8.38429069519043, + "learning_rate": 1.425339366515837e-05, + "loss": 8.1303, + "step": 65 + }, + { + "epoch": 0.0672782874617737, + "grad_norm": 9.689257621765137, + "learning_rate": 1.4479638009049776e-05, + "loss": 7.298, + "step": 66 + }, + { + "epoch": 0.06829765545361875, + "grad_norm": 8.886714935302734, + "learning_rate": 1.4705882352941177e-05, + "loss": 6.1227, + "step": 67 + }, + { + "epoch": 0.06931702344546381, + "grad_norm": 9.28791332244873, + "learning_rate": 1.493212669683258e-05, + "loss": 6.7938, + "step": 68 + }, + { + "epoch": 0.07033639143730887, + "grad_norm": 9.196669578552246, + "learning_rate": 1.5158371040723981e-05, + "loss": 6.4562, + "step": 69 + }, + { + "epoch": 0.07135575942915393, + "grad_norm": 10.716215133666992, + "learning_rate": 1.5384615384615387e-05, + "loss": 8.0389, + "step": 70 + }, + { + "epoch": 0.07237512742099898, + "grad_norm": 9.852572441101074, + "learning_rate": 1.5610859728506788e-05, + "loss": 8.7218, + "step": 71 + }, + { + "epoch": 0.07339449541284404, + "grad_norm": 8.59492301940918, + "learning_rate": 1.583710407239819e-05, + "loss": 6.1906, + "step": 72 + }, + { + "epoch": 0.0744138634046891, + "grad_norm": 9.830521583557129, + "learning_rate": 1.6063348416289596e-05, + "loss": 6.7222, + "step": 73 + }, + { + "epoch": 0.07543323139653414, + "grad_norm": 9.12816047668457, + "learning_rate": 1.6289592760180996e-05, + "loss": 7.0611, + "step": 74 + }, + { + "epoch": 0.0764525993883792, + "grad_norm": 10.391504287719727, + "learning_rate": 1.6515837104072397e-05, + "loss": 7.8241, + "step": 75 + }, + { + "epoch": 0.07747196738022426, + "grad_norm": 9.0382719039917, + "learning_rate": 1.67420814479638e-05, + "loss": 6.3791, + "step": 76 + }, + { + "epoch": 0.07849133537206932, + "grad_norm": 11.495955467224121, + "learning_rate": 1.6968325791855205e-05, + "loss": 6.8864, + "step": 77 + }, + { + "epoch": 0.07951070336391437, + "grad_norm": 9.282613754272461, + "learning_rate": 1.7194570135746606e-05, + "loss": 6.8356, + "step": 78 + }, + { + "epoch": 0.08053007135575943, + "grad_norm": 9.06067180633545, + "learning_rate": 1.742081447963801e-05, + "loss": 6.168, + "step": 79 + }, + { + "epoch": 0.08154943934760449, + "grad_norm": 10.343846321105957, + "learning_rate": 1.7647058823529414e-05, + "loss": 8.6845, + "step": 80 + }, + { + "epoch": 0.08256880733944955, + "grad_norm": 10.185526847839355, + "learning_rate": 1.7873303167420814e-05, + "loss": 5.9739, + "step": 81 + }, + { + "epoch": 0.0835881753312946, + "grad_norm": 12.164653778076172, + "learning_rate": 1.8099547511312218e-05, + "loss": 6.2423, + "step": 82 + }, + { + "epoch": 0.08460754332313965, + "grad_norm": 10.543149948120117, + "learning_rate": 1.832579185520362e-05, + "loss": 7.6247, + "step": 83 + }, + { + "epoch": 0.0856269113149847, + "grad_norm": 10.210731506347656, + "learning_rate": 1.8552036199095023e-05, + "loss": 5.8418, + "step": 84 + }, + { + "epoch": 0.08664627930682976, + "grad_norm": 11.613642692565918, + "learning_rate": 1.8778280542986427e-05, + "loss": 7.0948, + "step": 85 + }, + { + "epoch": 0.08766564729867482, + "grad_norm": 12.590648651123047, + "learning_rate": 1.9004524886877827e-05, + "loss": 6.7457, + "step": 86 + }, + { + "epoch": 0.08868501529051988, + "grad_norm": 12.547815322875977, + "learning_rate": 1.923076923076923e-05, + "loss": 5.6837, + "step": 87 + }, + { + "epoch": 0.08970438328236494, + "grad_norm": 14.212437629699707, + "learning_rate": 1.9457013574660635e-05, + "loss": 6.0757, + "step": 88 + }, + { + "epoch": 0.09072375127421, + "grad_norm": 14.821358680725098, + "learning_rate": 1.9683257918552036e-05, + "loss": 6.523, + "step": 89 + }, + { + "epoch": 0.09174311926605505, + "grad_norm": 14.133096694946289, + "learning_rate": 1.990950226244344e-05, + "loss": 6.2917, + "step": 90 + }, + { + "epoch": 0.09276248725790011, + "grad_norm": 14.283154487609863, + "learning_rate": 2.0135746606334844e-05, + "loss": 6.0509, + "step": 91 + }, + { + "epoch": 0.09378185524974515, + "grad_norm": 15.914741516113281, + "learning_rate": 2.0361990950226245e-05, + "loss": 6.8298, + "step": 92 + }, + { + "epoch": 0.09480122324159021, + "grad_norm": 18.067726135253906, + "learning_rate": 2.058823529411765e-05, + "loss": 7.855, + "step": 93 + }, + { + "epoch": 0.09582059123343527, + "grad_norm": 17.288843154907227, + "learning_rate": 2.0814479638009053e-05, + "loss": 6.6372, + "step": 94 + }, + { + "epoch": 0.09683995922528033, + "grad_norm": 22.13617515563965, + "learning_rate": 2.1040723981900453e-05, + "loss": 5.7468, + "step": 95 + }, + { + "epoch": 0.09785932721712538, + "grad_norm": 22.20960235595703, + "learning_rate": 2.1266968325791857e-05, + "loss": 7.5522, + "step": 96 + }, + { + "epoch": 0.09887869520897044, + "grad_norm": 23.28131103515625, + "learning_rate": 2.149321266968326e-05, + "loss": 7.7825, + "step": 97 + }, + { + "epoch": 0.0998980632008155, + "grad_norm": 29.695850372314453, + "learning_rate": 2.1719457013574662e-05, + "loss": 8.7452, + "step": 98 + }, + { + "epoch": 0.0998980632008155, + "eval_Qnli-dev-1024_cosine_accuracy": 0.6458333333333334, + "eval_Qnli-dev-1024_cosine_accuracy_threshold": 0.995652437210083, + "eval_Qnli-dev-1024_cosine_ap": 0.6274798374964984, + "eval_Qnli-dev-1024_cosine_f1": 0.6518518518518519, + "eval_Qnli-dev-1024_cosine_f1_threshold": 0.9515509605407715, + "eval_Qnli-dev-1024_cosine_mcc": 0.1563007361345257, + "eval_Qnli-dev-1024_cosine_precision": 0.4888888888888889, + "eval_Qnli-dev-1024_cosine_recall": 0.9777777777777777, + "eval_Qnli-dev_cosine_accuracy": 0.7395833333333334, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8860945701599121, + "eval_Qnli-dev_cosine_ap": 0.7645314494110582, + "eval_Qnli-dev_cosine_f1": 0.7500000000000001, + "eval_Qnli-dev_cosine_f1_threshold": 0.8442017436027527, + "eval_Qnli-dev_cosine_mcc": 0.48653004754089046, + "eval_Qnli-dev_cosine_precision": 0.6610169491525424, + "eval_Qnli-dev_cosine_recall": 0.8666666666666667, + "eval_allNLI--triplets-1024_cosine_accuracy": 0.7291666865348816, + "eval_allNLI-triplets_cosine_accuracy": 0.96875, + "eval_global_dataset_loss": 2.297825574874878, + "eval_global_dataset_runtime": 104.2196, + "eval_global_dataset_samples_per_second": 7.705, + "eval_global_dataset_steps_per_second": 0.163, + "eval_sequential_score": 0.7291666865348816, + "eval_sts-test-1024_pearson_cosine": 0.470983874633109, + "eval_sts-test-1024_spearman_cosine": 0.7146928621162676, + "eval_sts-test_pearson_cosine": 0.904138891044396, + "eval_sts-test_spearman_cosine": 0.9172742489825538, + "step": 98 + }, + { + "epoch": 0.10091743119266056, + "grad_norm": 21.82425880432129, + "learning_rate": 2.1945701357466062e-05, + "loss": 6.2322, + "step": 99 + }, + { + "epoch": 0.1019367991845056, + "grad_norm": 25.734025955200195, + "learning_rate": 2.2171945701357466e-05, + "loss": 4.8433, + "step": 100 + }, + { + "epoch": 0.10295616717635066, + "grad_norm": 28.17144775390625, + "learning_rate": 2.239819004524887e-05, + "loss": 4.8335, + "step": 101 + }, + { + "epoch": 0.10397553516819572, + "grad_norm": 27.875871658325195, + "learning_rate": 2.262443438914027e-05, + "loss": 5.138, + "step": 102 + }, + { + "epoch": 0.10499490316004077, + "grad_norm": 31.503034591674805, + "learning_rate": 2.2850678733031675e-05, + "loss": 4.8609, + "step": 103 + }, + { + "epoch": 0.10601427115188583, + "grad_norm": 26.674440383911133, + "learning_rate": 2.307692307692308e-05, + "loss": 4.6204, + "step": 104 + }, + { + "epoch": 0.10703363914373089, + "grad_norm": 25.039222717285156, + "learning_rate": 2.330316742081448e-05, + "loss": 4.3809, + "step": 105 + }, + { + "epoch": 0.10805300713557595, + "grad_norm": 26.333913803100586, + "learning_rate": 2.3529411764705884e-05, + "loss": 5.6703, + "step": 106 + }, + { + "epoch": 0.109072375127421, + "grad_norm": 23.51517105102539, + "learning_rate": 2.3755656108597284e-05, + "loss": 5.0237, + "step": 107 + }, + { + "epoch": 0.11009174311926606, + "grad_norm": 18.25855255126953, + "learning_rate": 2.3981900452488688e-05, + "loss": 4.002, + "step": 108 + }, + { + "epoch": 0.1111111111111111, + "grad_norm": 19.852886199951172, + "learning_rate": 2.4208144796380092e-05, + "loss": 5.2532, + "step": 109 + }, + { + "epoch": 0.11213047910295616, + "grad_norm": 17.45444107055664, + "learning_rate": 2.4434389140271493e-05, + "loss": 4.1033, + "step": 110 + }, + { + "epoch": 0.11314984709480122, + "grad_norm": 14.521421432495117, + "learning_rate": 2.4660633484162897e-05, + "loss": 4.0818, + "step": 111 + }, + { + "epoch": 0.11416921508664628, + "grad_norm": 12.525910377502441, + "learning_rate": 2.48868778280543e-05, + "loss": 3.458, + "step": 112 + }, + { + "epoch": 0.11518858307849134, + "grad_norm": 14.503193855285645, + "learning_rate": 2.51131221719457e-05, + "loss": 4.3372, + "step": 113 + }, + { + "epoch": 0.1162079510703364, + "grad_norm": 14.2279634475708, + "learning_rate": 2.5339366515837106e-05, + "loss": 4.9513, + "step": 114 + }, + { + "epoch": 0.11722731906218145, + "grad_norm": 15.238719940185547, + "learning_rate": 2.5565610859728506e-05, + "loss": 5.2602, + "step": 115 + }, + { + "epoch": 0.11824668705402651, + "grad_norm": 11.11528491973877, + "learning_rate": 2.5791855203619913e-05, + "loss": 3.1741, + "step": 116 + }, + { + "epoch": 0.11926605504587157, + "grad_norm": 12.077157974243164, + "learning_rate": 2.6018099547511314e-05, + "loss": 4.1914, + "step": 117 + }, + { + "epoch": 0.12028542303771661, + "grad_norm": 11.872669219970703, + "learning_rate": 2.6244343891402718e-05, + "loss": 2.8383, + "step": 118 + }, + { + "epoch": 0.12130479102956167, + "grad_norm": 9.008302688598633, + "learning_rate": 2.647058823529412e-05, + "loss": 3.4165, + "step": 119 + }, + { + "epoch": 0.12232415902140673, + "grad_norm": 10.702130317687988, + "learning_rate": 2.6696832579185523e-05, + "loss": 3.5085, + "step": 120 + }, + { + "epoch": 0.12334352701325178, + "grad_norm": 10.306276321411133, + "learning_rate": 2.6923076923076923e-05, + "loss": 2.3992, + "step": 121 + }, + { + "epoch": 0.12436289500509684, + "grad_norm": 9.035378456115723, + "learning_rate": 2.7149321266968324e-05, + "loss": 2.4849, + "step": 122 + }, + { + "epoch": 0.12538226299694188, + "grad_norm": 8.996299743652344, + "learning_rate": 2.737556561085973e-05, + "loss": 2.2839, + "step": 123 + }, + { + "epoch": 0.12640163098878696, + "grad_norm": 8.635661125183105, + "learning_rate": 2.7601809954751135e-05, + "loss": 2.567, + "step": 124 + }, + { + "epoch": 0.127420998980632, + "grad_norm": 10.015826225280762, + "learning_rate": 2.7828054298642536e-05, + "loss": 4.5119, + "step": 125 + }, + { + "epoch": 0.12844036697247707, + "grad_norm": 8.679932594299316, + "learning_rate": 2.805429864253394e-05, + "loss": 2.767, + "step": 126 + }, + { + "epoch": 0.12945973496432212, + "grad_norm": 10.05739688873291, + "learning_rate": 2.828054298642534e-05, + "loss": 4.0225, + "step": 127 + }, + { + "epoch": 0.1304791029561672, + "grad_norm": 9.361485481262207, + "learning_rate": 2.850678733031674e-05, + "loss": 1.8294, + "step": 128 + }, + { + "epoch": 0.13149847094801223, + "grad_norm": 9.865928649902344, + "learning_rate": 2.8733031674208145e-05, + "loss": 4.4174, + "step": 129 + }, + { + "epoch": 0.1325178389398573, + "grad_norm": 10.055468559265137, + "learning_rate": 2.8959276018099553e-05, + "loss": 2.0112, + "step": 130 + }, + { + "epoch": 0.13353720693170235, + "grad_norm": 9.528116226196289, + "learning_rate": 2.9185520361990953e-05, + "loss": 1.7772, + "step": 131 + }, + { + "epoch": 0.1345565749235474, + "grad_norm": 9.870166778564453, + "learning_rate": 2.9411764705882354e-05, + "loss": 3.1912, + "step": 132 + }, + { + "epoch": 0.13557594291539246, + "grad_norm": 10.1703462600708, + "learning_rate": 2.9638009049773758e-05, + "loss": 2.4527, + "step": 133 + }, + { + "epoch": 0.1365953109072375, + "grad_norm": 7.443604469299316, + "learning_rate": 2.986425339366516e-05, + "loss": 1.6424, + "step": 134 + }, + { + "epoch": 0.13761467889908258, + "grad_norm": 10.003544807434082, + "learning_rate": 3.0090497737556562e-05, + "loss": 2.6143, + "step": 135 + }, + { + "epoch": 0.13863404689092762, + "grad_norm": 9.352860450744629, + "learning_rate": 3.0316742081447963e-05, + "loss": 2.0498, + "step": 136 + }, + { + "epoch": 0.1396534148827727, + "grad_norm": 7.393095970153809, + "learning_rate": 3.0542986425339374e-05, + "loss": 1.962, + "step": 137 + }, + { + "epoch": 0.14067278287461774, + "grad_norm": 8.278059959411621, + "learning_rate": 3.0769230769230774e-05, + "loss": 1.789, + "step": 138 + }, + { + "epoch": 0.14169215086646278, + "grad_norm": 6.577699184417725, + "learning_rate": 3.0995475113122175e-05, + "loss": 1.459, + "step": 139 + }, + { + "epoch": 0.14271151885830785, + "grad_norm": 8.23404312133789, + "learning_rate": 3.1221719457013576e-05, + "loss": 1.2479, + "step": 140 + }, + { + "epoch": 0.1437308868501529, + "grad_norm": 9.47106647491455, + "learning_rate": 3.1447963800904976e-05, + "loss": 2.5413, + "step": 141 + }, + { + "epoch": 0.14475025484199797, + "grad_norm": 7.330000400543213, + "learning_rate": 3.167420814479638e-05, + "loss": 1.4077, + "step": 142 + }, + { + "epoch": 0.145769622833843, + "grad_norm": 9.64534854888916, + "learning_rate": 3.1900452488687784e-05, + "loss": 2.6988, + "step": 143 + }, + { + "epoch": 0.14678899082568808, + "grad_norm": 8.404465675354004, + "learning_rate": 3.212669683257919e-05, + "loss": 2.9772, + "step": 144 + }, + { + "epoch": 0.14780835881753313, + "grad_norm": 8.019698143005371, + "learning_rate": 3.235294117647059e-05, + "loss": 1.6265, + "step": 145 + }, + { + "epoch": 0.1488277268093782, + "grad_norm": 7.635079860687256, + "learning_rate": 3.257918552036199e-05, + "loss": 1.9404, + "step": 146 + }, + { + "epoch": 0.14984709480122324, + "grad_norm": 7.929011821746826, + "learning_rate": 3.2805429864253393e-05, + "loss": 1.4251, + "step": 147 + }, + { + "epoch": 0.15086646279306828, + "grad_norm": 7.869425296783447, + "learning_rate": 3.3031674208144794e-05, + "loss": 2.6657, + "step": 148 + }, + { + "epoch": 0.15188583078491336, + "grad_norm": 8.369176864624023, + "learning_rate": 3.32579185520362e-05, + "loss": 2.1576, + "step": 149 + }, + { + "epoch": 0.1529051987767584, + "grad_norm": 9.128487586975098, + "learning_rate": 3.34841628959276e-05, + "loss": 1.53, + "step": 150 + }, + { + "epoch": 0.15392456676860347, + "grad_norm": 7.673459529876709, + "learning_rate": 3.371040723981901e-05, + "loss": 1.2642, + "step": 151 + }, + { + "epoch": 0.15494393476044852, + "grad_norm": 9.104422569274902, + "learning_rate": 3.393665158371041e-05, + "loss": 2.4846, + "step": 152 + }, + { + "epoch": 0.1559633027522936, + "grad_norm": 8.658594131469727, + "learning_rate": 3.416289592760181e-05, + "loss": 1.4979, + "step": 153 + }, + { + "epoch": 0.15698267074413863, + "grad_norm": 9.34330940246582, + "learning_rate": 3.438914027149321e-05, + "loss": 1.8149, + "step": 154 + }, + { + "epoch": 0.1580020387359837, + "grad_norm": 9.401769638061523, + "learning_rate": 3.461538461538462e-05, + "loss": 1.4693, + "step": 155 + }, + { + "epoch": 0.15902140672782875, + "grad_norm": 10.389461517333984, + "learning_rate": 3.484162895927602e-05, + "loss": 2.0114, + "step": 156 + }, + { + "epoch": 0.1600407747196738, + "grad_norm": 9.321866989135742, + "learning_rate": 3.506787330316742e-05, + "loss": 1.5511, + "step": 157 + }, + { + "epoch": 0.16106014271151886, + "grad_norm": 10.052262306213379, + "learning_rate": 3.529411764705883e-05, + "loss": 1.621, + "step": 158 + }, + { + "epoch": 0.1620795107033639, + "grad_norm": 7.535787105560303, + "learning_rate": 3.552036199095023e-05, + "loss": 2.1122, + "step": 159 + }, + { + "epoch": 0.16309887869520898, + "grad_norm": 9.70533275604248, + "learning_rate": 3.574660633484163e-05, + "loss": 1.8148, + "step": 160 + }, + { + "epoch": 0.16411824668705402, + "grad_norm": 7.81204080581665, + "learning_rate": 3.5972850678733036e-05, + "loss": 1.9861, + "step": 161 + }, + { + "epoch": 0.1651376146788991, + "grad_norm": 7.583981513977051, + "learning_rate": 3.6199095022624436e-05, + "loss": 1.3943, + "step": 162 + }, + { + "epoch": 0.16615698267074414, + "grad_norm": 8.344895362854004, + "learning_rate": 3.642533936651584e-05, + "loss": 1.7317, + "step": 163 + }, + { + "epoch": 0.1671763506625892, + "grad_norm": 7.1097331047058105, + "learning_rate": 3.665158371040724e-05, + "loss": 1.6, + "step": 164 + }, + { + "epoch": 0.16819571865443425, + "grad_norm": 7.911113739013672, + "learning_rate": 3.6877828054298645e-05, + "loss": 1.2222, + "step": 165 + }, + { + "epoch": 0.1692150866462793, + "grad_norm": 9.282394409179688, + "learning_rate": 3.7104072398190046e-05, + "loss": 1.6152, + "step": 166 + }, + { + "epoch": 0.17023445463812437, + "grad_norm": 7.449146270751953, + "learning_rate": 3.733031674208145e-05, + "loss": 1.0374, + "step": 167 + }, + { + "epoch": 0.1712538226299694, + "grad_norm": 9.164731979370117, + "learning_rate": 3.7556561085972854e-05, + "loss": 1.2844, + "step": 168 + }, + { + "epoch": 0.17227319062181448, + "grad_norm": 6.987304210662842, + "learning_rate": 3.7782805429864254e-05, + "loss": 1.8805, + "step": 169 + }, + { + "epoch": 0.17329255861365953, + "grad_norm": 7.447988033294678, + "learning_rate": 3.8009049773755655e-05, + "loss": 1.0972, + "step": 170 + }, + { + "epoch": 0.1743119266055046, + "grad_norm": 7.7849321365356445, + "learning_rate": 3.8235294117647055e-05, + "loss": 1.7012, + "step": 171 + }, + { + "epoch": 0.17533129459734964, + "grad_norm": 7.341614246368408, + "learning_rate": 3.846153846153846e-05, + "loss": 1.4182, + "step": 172 + }, + { + "epoch": 0.1763506625891947, + "grad_norm": 8.514887809753418, + "learning_rate": 3.868778280542987e-05, + "loss": 2.6053, + "step": 173 + }, + { + "epoch": 0.17737003058103976, + "grad_norm": 7.384711265563965, + "learning_rate": 3.891402714932127e-05, + "loss": 1.4193, + "step": 174 + }, + { + "epoch": 0.1783893985728848, + "grad_norm": 8.553336143493652, + "learning_rate": 3.914027149321267e-05, + "loss": 2.251, + "step": 175 + }, + { + "epoch": 0.17940876656472987, + "grad_norm": 8.517749786376953, + "learning_rate": 3.936651583710407e-05, + "loss": 1.9057, + "step": 176 + }, + { + "epoch": 0.18042813455657492, + "grad_norm": 8.444558143615723, + "learning_rate": 3.959276018099547e-05, + "loss": 1.1228, + "step": 177 + }, + { + "epoch": 0.18144750254842, + "grad_norm": 12.253990173339844, + "learning_rate": 3.981900452488688e-05, + "loss": 4.0905, + "step": 178 + }, + { + "epoch": 0.18246687054026503, + "grad_norm": 5.70052433013916, + "learning_rate": 4.004524886877829e-05, + "loss": 0.9007, + "step": 179 + }, + { + "epoch": 0.1834862385321101, + "grad_norm": 9.525473594665527, + "learning_rate": 4.027149321266969e-05, + "loss": 2.0665, + "step": 180 + }, + { + "epoch": 0.18450560652395515, + "grad_norm": 6.146080493927002, + "learning_rate": 4.049773755656109e-05, + "loss": 1.0946, + "step": 181 + }, + { + "epoch": 0.18552497451580022, + "grad_norm": 7.736543655395508, + "learning_rate": 4.072398190045249e-05, + "loss": 1.7479, + "step": 182 + }, + { + "epoch": 0.18654434250764526, + "grad_norm": 8.404258728027344, + "learning_rate": 4.095022624434389e-05, + "loss": 2.0877, + "step": 183 + }, + { + "epoch": 0.1875637104994903, + "grad_norm": 5.705750942230225, + "learning_rate": 4.11764705882353e-05, + "loss": 0.9239, + "step": 184 + }, + { + "epoch": 0.18858307849133538, + "grad_norm": 7.753995895385742, + "learning_rate": 4.14027149321267e-05, + "loss": 1.7865, + "step": 185 + }, + { + "epoch": 0.18960244648318042, + "grad_norm": 9.15240478515625, + "learning_rate": 4.1628959276018105e-05, + "loss": 2.1053, + "step": 186 + }, + { + "epoch": 0.1906218144750255, + "grad_norm": 7.2251129150390625, + "learning_rate": 4.1855203619909506e-05, + "loss": 1.5273, + "step": 187 + }, + { + "epoch": 0.19164118246687054, + "grad_norm": 6.803040981292725, + "learning_rate": 4.2081447963800907e-05, + "loss": 1.8726, + "step": 188 + }, + { + "epoch": 0.1926605504587156, + "grad_norm": 5.646162509918213, + "learning_rate": 4.230769230769231e-05, + "loss": 1.4663, + "step": 189 + }, + { + "epoch": 0.19367991845056065, + "grad_norm": 7.599930286407471, + "learning_rate": 4.2533936651583714e-05, + "loss": 1.0136, + "step": 190 + }, + { + "epoch": 0.1946992864424057, + "grad_norm": 7.882979393005371, + "learning_rate": 4.2760180995475115e-05, + "loss": 1.121, + "step": 191 + }, + { + "epoch": 0.19571865443425077, + "grad_norm": 8.919268608093262, + "learning_rate": 4.298642533936652e-05, + "loss": 1.6074, + "step": 192 + }, + { + "epoch": 0.1967380224260958, + "grad_norm": 8.914848327636719, + "learning_rate": 4.321266968325792e-05, + "loss": 2.1956, + "step": 193 + }, + { + "epoch": 0.19775739041794088, + "grad_norm": 8.603778839111328, + "learning_rate": 4.3438914027149324e-05, + "loss": 1.5425, + "step": 194 + }, + { + "epoch": 0.19877675840978593, + "grad_norm": 8.500616073608398, + "learning_rate": 4.3665158371040724e-05, + "loss": 1.4552, + "step": 195 + }, + { + "epoch": 0.199796126401631, + "grad_norm": 7.815979957580566, + "learning_rate": 4.3891402714932125e-05, + "loss": 1.2635, + "step": 196 + }, + { + "epoch": 0.199796126401631, + "eval_Qnli-dev-1024_cosine_accuracy": 0.7395833333333334, + "eval_Qnli-dev-1024_cosine_accuracy_threshold": 0.8570283651351929, + "eval_Qnli-dev-1024_cosine_ap": 0.7434694144471753, + "eval_Qnli-dev-1024_cosine_f1": 0.7207207207207208, + "eval_Qnli-dev-1024_cosine_f1_threshold": 0.7906914353370667, + "eval_Qnli-dev-1024_cosine_mcc": 0.4081269865567241, + "eval_Qnli-dev-1024_cosine_precision": 0.6060606060606061, + "eval_Qnli-dev-1024_cosine_recall": 0.8888888888888888, + "eval_Qnli-dev_cosine_accuracy": 0.75, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8287814855575562, + "eval_Qnli-dev_cosine_ap": 0.7646453733471359, + "eval_Qnli-dev_cosine_f1": 0.7378640776699029, + "eval_Qnli-dev_cosine_f1_threshold": 0.7745069265365601, + "eval_Qnli-dev_cosine_mcc": 0.46153029495329345, + "eval_Qnli-dev_cosine_precision": 0.6551724137931034, + "eval_Qnli-dev_cosine_recall": 0.8444444444444444, + "eval_allNLI--triplets-1024_cosine_accuracy": 0.9166666865348816, + "eval_allNLI-triplets_cosine_accuracy": 0.9583333134651184, + "eval_global_dataset_loss": 0.5179261565208435, + "eval_global_dataset_runtime": 104.2216, + "eval_global_dataset_samples_per_second": 7.705, + "eval_global_dataset_steps_per_second": 0.163, + "eval_sequential_score": 0.9166666865348816, + "eval_sts-test-1024_pearson_cosine": 0.8476975008591285, + "eval_sts-test-1024_spearman_cosine": 0.8973182534732806, + "eval_sts-test_pearson_cosine": 0.9039400681490469, + "eval_sts-test_spearman_cosine": 0.9185431775441114, + "step": 196 + }, + { + "epoch": 0.20081549439347604, + "grad_norm": 9.34125804901123, + "learning_rate": 4.411764705882353e-05, + "loss": 1.7222, + "step": 197 + }, + { + "epoch": 0.2018348623853211, + "grad_norm": 10.679852485656738, + "learning_rate": 4.434389140271493e-05, + "loss": 2.377, + "step": 198 + }, + { + "epoch": 0.20285423037716616, + "grad_norm": 7.775190830230713, + "learning_rate": 4.457013574660634e-05, + "loss": 1.5317, + "step": 199 + }, + { + "epoch": 0.2038735983690112, + "grad_norm": 6.390950679779053, + "learning_rate": 4.479638009049774e-05, + "loss": 1.0494, + "step": 200 + }, + { + "epoch": 0.20489296636085627, + "grad_norm": 9.170794486999512, + "learning_rate": 4.502262443438914e-05, + "loss": 1.7392, + "step": 201 + }, + { + "epoch": 0.20591233435270132, + "grad_norm": 7.37787389755249, + "learning_rate": 4.524886877828054e-05, + "loss": 1.2924, + "step": 202 + }, + { + "epoch": 0.2069317023445464, + "grad_norm": 6.836249828338623, + "learning_rate": 4.547511312217195e-05, + "loss": 0.9413, + "step": 203 + }, + { + "epoch": 0.20795107033639143, + "grad_norm": 9.543895721435547, + "learning_rate": 4.570135746606335e-05, + "loss": 2.1448, + "step": 204 + }, + { + "epoch": 0.2089704383282365, + "grad_norm": 7.8430495262146, + "learning_rate": 4.592760180995475e-05, + "loss": 1.0357, + "step": 205 + }, + { + "epoch": 0.20998980632008155, + "grad_norm": 9.558221817016602, + "learning_rate": 4.615384615384616e-05, + "loss": 1.3534, + "step": 206 + }, + { + "epoch": 0.21100917431192662, + "grad_norm": 5.715826034545898, + "learning_rate": 4.638009049773756e-05, + "loss": 1.0564, + "step": 207 + }, + { + "epoch": 0.21202854230377166, + "grad_norm": 8.720932960510254, + "learning_rate": 4.660633484162896e-05, + "loss": 0.9259, + "step": 208 + }, + { + "epoch": 0.2130479102956167, + "grad_norm": 9.008890151977539, + "learning_rate": 4.683257918552037e-05, + "loss": 1.3813, + "step": 209 + }, + { + "epoch": 0.21406727828746178, + "grad_norm": 7.1262006759643555, + "learning_rate": 4.705882352941177e-05, + "loss": 0.8828, + "step": 210 + }, + { + "epoch": 0.21508664627930682, + "grad_norm": 12.986166000366211, + "learning_rate": 4.728506787330317e-05, + "loss": 2.9147, + "step": 211 + }, + { + "epoch": 0.2161060142711519, + "grad_norm": 6.804072380065918, + "learning_rate": 4.751131221719457e-05, + "loss": 0.6539, + "step": 212 + }, + { + "epoch": 0.21712538226299694, + "grad_norm": 9.138653755187988, + "learning_rate": 4.7737556561085976e-05, + "loss": 1.3092, + "step": 213 + }, + { + "epoch": 0.218144750254842, + "grad_norm": 7.303668975830078, + "learning_rate": 4.7963800904977377e-05, + "loss": 1.1562, + "step": 214 + }, + { + "epoch": 0.21916411824668705, + "grad_norm": 7.368769645690918, + "learning_rate": 4.8190045248868784e-05, + "loss": 0.9509, + "step": 215 + }, + { + "epoch": 0.22018348623853212, + "grad_norm": 5.067785263061523, + "learning_rate": 4.8416289592760185e-05, + "loss": 0.6664, + "step": 216 + }, + { + "epoch": 0.22120285423037717, + "grad_norm": 5.643320083618164, + "learning_rate": 4.8642533936651585e-05, + "loss": 1.2315, + "step": 217 + }, + { + "epoch": 0.2222222222222222, + "grad_norm": 6.596173286437988, + "learning_rate": 4.8868778280542986e-05, + "loss": 0.9855, + "step": 218 + }, + { + "epoch": 0.22324159021406728, + "grad_norm": 6.5434770584106445, + "learning_rate": 4.9095022624434386e-05, + "loss": 0.9258, + "step": 219 + }, + { + "epoch": 0.22426095820591233, + "grad_norm": 11.537922859191895, + "learning_rate": 4.9321266968325794e-05, + "loss": 1.6578, + "step": 220 + }, + { + "epoch": 0.2252803261977574, + "grad_norm": 7.364137172698975, + "learning_rate": 4.95475113122172e-05, + "loss": 0.9666, + "step": 221 + }, + { + "epoch": 0.22629969418960244, + "grad_norm": 8.102925300598145, + "learning_rate": 4.97737556561086e-05, + "loss": 0.9808, + "step": 222 + }, + { + "epoch": 0.2273190621814475, + "grad_norm": 10.013775825500488, + "learning_rate": 5e-05, + "loss": 2.4156, + "step": 223 + }, + { + "epoch": 0.22833843017329256, + "grad_norm": 7.974793434143066, + "learning_rate": 5.02262443438914e-05, + "loss": 1.1297, + "step": 224 + }, + { + "epoch": 0.22935779816513763, + "grad_norm": 7.710846424102783, + "learning_rate": 5.0452488687782804e-05, + "loss": 1.3063, + "step": 225 + }, + { + "epoch": 0.23037716615698267, + "grad_norm": 5.633566856384277, + "learning_rate": 5.067873303167421e-05, + "loss": 0.5567, + "step": 226 + }, + { + "epoch": 0.23139653414882771, + "grad_norm": 9.50987720489502, + "learning_rate": 5.090497737556561e-05, + "loss": 1.3551, + "step": 227 + }, + { + "epoch": 0.2324159021406728, + "grad_norm": 10.309268951416016, + "learning_rate": 5.113122171945701e-05, + "loss": 1.4079, + "step": 228 + }, + { + "epoch": 0.23343527013251783, + "grad_norm": 7.812633037567139, + "learning_rate": 5.135746606334841e-05, + "loss": 0.948, + "step": 229 + }, + { + "epoch": 0.2344546381243629, + "grad_norm": 8.013436317443848, + "learning_rate": 5.158371040723983e-05, + "loss": 0.9288, + "step": 230 + }, + { + "epoch": 0.23547400611620795, + "grad_norm": 7.550686359405518, + "learning_rate": 5.180995475113123e-05, + "loss": 1.0077, + "step": 231 + }, + { + "epoch": 0.23649337410805302, + "grad_norm": 7.249583721160889, + "learning_rate": 5.203619909502263e-05, + "loss": 1.0674, + "step": 232 + }, + { + "epoch": 0.23751274209989806, + "grad_norm": 7.766678810119629, + "learning_rate": 5.2262443438914036e-05, + "loss": 1.3354, + "step": 233 + }, + { + "epoch": 0.23853211009174313, + "grad_norm": 7.417704105377197, + "learning_rate": 5.2488687782805436e-05, + "loss": 1.0076, + "step": 234 + }, + { + "epoch": 0.23955147808358818, + "grad_norm": 8.414839744567871, + "learning_rate": 5.271493212669684e-05, + "loss": 0.8814, + "step": 235 + }, + { + "epoch": 0.24057084607543322, + "grad_norm": 9.537981986999512, + "learning_rate": 5.294117647058824e-05, + "loss": 1.7839, + "step": 236 + }, + { + "epoch": 0.2415902140672783, + "grad_norm": 6.3290886878967285, + "learning_rate": 5.316742081447964e-05, + "loss": 1.2614, + "step": 237 + }, + { + "epoch": 0.24260958205912334, + "grad_norm": 8.181835174560547, + "learning_rate": 5.3393665158371045e-05, + "loss": 0.8655, + "step": 238 + }, + { + "epoch": 0.2436289500509684, + "grad_norm": 8.01684856414795, + "learning_rate": 5.3619909502262446e-05, + "loss": 1.8418, + "step": 239 + }, + { + "epoch": 0.24464831804281345, + "grad_norm": 7.891118049621582, + "learning_rate": 5.384615384615385e-05, + "loss": 1.4465, + "step": 240 + }, + { + "epoch": 0.24566768603465852, + "grad_norm": 8.080881118774414, + "learning_rate": 5.407239819004525e-05, + "loss": 0.8695, + "step": 241 + }, + { + "epoch": 0.24668705402650357, + "grad_norm": 6.881638050079346, + "learning_rate": 5.429864253393665e-05, + "loss": 0.9695, + "step": 242 + }, + { + "epoch": 0.24770642201834864, + "grad_norm": 10.03598690032959, + "learning_rate": 5.4524886877828055e-05, + "loss": 1.3374, + "step": 243 + }, + { + "epoch": 0.24872579001019368, + "grad_norm": 7.844127178192139, + "learning_rate": 5.475113122171946e-05, + "loss": 0.6716, + "step": 244 + }, + { + "epoch": 0.24974515800203873, + "grad_norm": 8.654071807861328, + "learning_rate": 5.497737556561087e-05, + "loss": 1.032, + "step": 245 + }, + { + "epoch": 0.25076452599388377, + "grad_norm": 6.731460094451904, + "learning_rate": 5.520361990950227e-05, + "loss": 0.8033, + "step": 246 + }, + { + "epoch": 0.25178389398572887, + "grad_norm": 9.436687469482422, + "learning_rate": 5.542986425339367e-05, + "loss": 0.9257, + "step": 247 + }, + { + "epoch": 0.2528032619775739, + "grad_norm": 7.817379474639893, + "learning_rate": 5.565610859728507e-05, + "loss": 0.8311, + "step": 248 + }, + { + "epoch": 0.25382262996941896, + "grad_norm": 6.328183650970459, + "learning_rate": 5.588235294117647e-05, + "loss": 0.5609, + "step": 249 + }, + { + "epoch": 0.254841997961264, + "grad_norm": 8.576601028442383, + "learning_rate": 5.610859728506788e-05, + "loss": 1.5985, + "step": 250 + }, + { + "epoch": 0.2558613659531091, + "grad_norm": 9.092324256896973, + "learning_rate": 5.633484162895928e-05, + "loss": 1.0025, + "step": 251 + }, + { + "epoch": 0.25688073394495414, + "grad_norm": 11.906094551086426, + "learning_rate": 5.656108597285068e-05, + "loss": 2.0499, + "step": 252 + }, + { + "epoch": 0.2579001019367992, + "grad_norm": 7.968968868255615, + "learning_rate": 5.678733031674208e-05, + "loss": 1.3116, + "step": 253 + }, + { + "epoch": 0.25891946992864423, + "grad_norm": 5.355049133300781, + "learning_rate": 5.701357466063348e-05, + "loss": 0.5969, + "step": 254 + }, + { + "epoch": 0.2599388379204893, + "grad_norm": 8.151896476745605, + "learning_rate": 5.723981900452488e-05, + "loss": 1.1107, + "step": 255 + }, + { + "epoch": 0.2609582059123344, + "grad_norm": 9.651622772216797, + "learning_rate": 5.746606334841629e-05, + "loss": 1.8581, + "step": 256 + }, + { + "epoch": 0.2619775739041794, + "grad_norm": 7.1527533531188965, + "learning_rate": 5.769230769230769e-05, + "loss": 0.572, + "step": 257 + }, + { + "epoch": 0.26299694189602446, + "grad_norm": 6.141374111175537, + "learning_rate": 5.7918552036199105e-05, + "loss": 0.9267, + "step": 258 + }, + { + "epoch": 0.2640163098878695, + "grad_norm": 7.274891376495361, + "learning_rate": 5.8144796380090506e-05, + "loss": 0.6255, + "step": 259 + }, + { + "epoch": 0.2650356778797146, + "grad_norm": 5.81080436706543, + "learning_rate": 5.8371040723981906e-05, + "loss": 0.7615, + "step": 260 + }, + { + "epoch": 0.26605504587155965, + "grad_norm": 6.9981279373168945, + "learning_rate": 5.859728506787331e-05, + "loss": 0.6026, + "step": 261 + }, + { + "epoch": 0.2670744138634047, + "grad_norm": 5.718660831451416, + "learning_rate": 5.882352941176471e-05, + "loss": 0.7263, + "step": 262 + }, + { + "epoch": 0.26809378185524974, + "grad_norm": 5.391998767852783, + "learning_rate": 5.9049773755656115e-05, + "loss": 0.4643, + "step": 263 + }, + { + "epoch": 0.2691131498470948, + "grad_norm": 6.843007564544678, + "learning_rate": 5.9276018099547516e-05, + "loss": 0.5101, + "step": 264 + }, + { + "epoch": 0.2701325178389399, + "grad_norm": 5.087254047393799, + "learning_rate": 5.9502262443438916e-05, + "loss": 0.5562, + "step": 265 + }, + { + "epoch": 0.2711518858307849, + "grad_norm": 7.482615947723389, + "learning_rate": 5.972850678733032e-05, + "loss": 1.256, + "step": 266 + }, + { + "epoch": 0.27217125382262997, + "grad_norm": 6.911371231079102, + "learning_rate": 5.995475113122172e-05, + "loss": 0.6543, + "step": 267 + }, + { + "epoch": 0.273190621814475, + "grad_norm": 7.643139839172363, + "learning_rate": 6.0180995475113125e-05, + "loss": 0.6698, + "step": 268 + }, + { + "epoch": 0.2742099898063201, + "grad_norm": 9.08658504486084, + "learning_rate": 6.0407239819004525e-05, + "loss": 1.3843, + "step": 269 + }, + { + "epoch": 0.27522935779816515, + "grad_norm": 8.890534400939941, + "learning_rate": 6.0633484162895926e-05, + "loss": 1.1421, + "step": 270 + }, + { + "epoch": 0.2762487257900102, + "grad_norm": 9.855698585510254, + "learning_rate": 6.0859728506787327e-05, + "loss": 1.1558, + "step": 271 + }, + { + "epoch": 0.27726809378185524, + "grad_norm": 8.32972526550293, + "learning_rate": 6.108597285067875e-05, + "loss": 1.603, + "step": 272 + }, + { + "epoch": 0.2782874617737003, + "grad_norm": 8.393510818481445, + "learning_rate": 6.131221719457015e-05, + "loss": 0.7985, + "step": 273 + }, + { + "epoch": 0.2793068297655454, + "grad_norm": 7.992040157318115, + "learning_rate": 6.153846153846155e-05, + "loss": 1.3884, + "step": 274 + }, + { + "epoch": 0.2803261977573904, + "grad_norm": 8.646651268005371, + "learning_rate": 6.176470588235295e-05, + "loss": 1.0337, + "step": 275 + }, + { + "epoch": 0.28134556574923547, + "grad_norm": 7.3104329109191895, + "learning_rate": 6.199095022624435e-05, + "loss": 1.0917, + "step": 276 + }, + { + "epoch": 0.2823649337410805, + "grad_norm": 12.030378341674805, + "learning_rate": 6.221719457013575e-05, + "loss": 2.4149, + "step": 277 + }, + { + "epoch": 0.28338430173292556, + "grad_norm": 4.781021595001221, + "learning_rate": 6.244343891402715e-05, + "loss": 0.4301, + "step": 278 + }, + { + "epoch": 0.28440366972477066, + "grad_norm": 4.352090358734131, + "learning_rate": 6.266968325791855e-05, + "loss": 0.5084, + "step": 279 + }, + { + "epoch": 0.2854230377166157, + "grad_norm": 5.88839864730835, + "learning_rate": 6.289592760180995e-05, + "loss": 0.7202, + "step": 280 + }, + { + "epoch": 0.28644240570846075, + "grad_norm": 11.228419303894043, + "learning_rate": 6.312217194570135e-05, + "loss": 1.7983, + "step": 281 + }, + { + "epoch": 0.2874617737003058, + "grad_norm": 6.119421005249023, + "learning_rate": 6.334841628959275e-05, + "loss": 0.673, + "step": 282 + }, + { + "epoch": 0.2884811416921509, + "grad_norm": 6.405134677886963, + "learning_rate": 6.357466063348417e-05, + "loss": 0.6655, + "step": 283 + }, + { + "epoch": 0.28950050968399593, + "grad_norm": 6.735506534576416, + "learning_rate": 6.380090497737557e-05, + "loss": 0.9121, + "step": 284 + }, + { + "epoch": 0.290519877675841, + "grad_norm": 11.012415885925293, + "learning_rate": 6.402714932126697e-05, + "loss": 1.5978, + "step": 285 + }, + { + "epoch": 0.291539245667686, + "grad_norm": 13.007187843322754, + "learning_rate": 6.425339366515838e-05, + "loss": 1.8536, + "step": 286 + }, + { + "epoch": 0.29255861365953106, + "grad_norm": 12.273601531982422, + "learning_rate": 6.447963800904978e-05, + "loss": 1.6397, + "step": 287 + }, + { + "epoch": 0.29357798165137616, + "grad_norm": 9.6339750289917, + "learning_rate": 6.470588235294118e-05, + "loss": 0.8275, + "step": 288 + }, + { + "epoch": 0.2945973496432212, + "grad_norm": 6.717658996582031, + "learning_rate": 6.493212669683258e-05, + "loss": 0.6003, + "step": 289 + }, + { + "epoch": 0.29561671763506625, + "grad_norm": 8.443256378173828, + "learning_rate": 6.515837104072399e-05, + "loss": 0.9834, + "step": 290 + }, + { + "epoch": 0.2966360856269113, + "grad_norm": 8.823105812072754, + "learning_rate": 6.538461538461539e-05, + "loss": 0.603, + "step": 291 + }, + { + "epoch": 0.2976554536187564, + "grad_norm": 6.8099141120910645, + "learning_rate": 6.561085972850679e-05, + "loss": 0.6597, + "step": 292 + }, + { + "epoch": 0.29867482161060144, + "grad_norm": 6.705087661743164, + "learning_rate": 6.583710407239819e-05, + "loss": 0.739, + "step": 293 + }, + { + "epoch": 0.2996941896024465, + "grad_norm": 7.209024906158447, + "learning_rate": 6.606334841628959e-05, + "loss": 1.2564, + "step": 294 + }, + { + "epoch": 0.2996941896024465, + "eval_Qnli-dev-1024_cosine_accuracy": 0.7395833333333334, + "eval_Qnli-dev-1024_cosine_accuracy_threshold": 0.8565528392791748, + "eval_Qnli-dev-1024_cosine_ap": 0.7531377591671699, + "eval_Qnli-dev-1024_cosine_f1": 0.7254901960784313, + "eval_Qnli-dev-1024_cosine_f1_threshold": 0.8200148344039917, + "eval_Qnli-dev-1024_cosine_mcc": 0.43697448216965834, + "eval_Qnli-dev-1024_cosine_precision": 0.6491228070175439, + "eval_Qnli-dev-1024_cosine_recall": 0.8222222222222222, + "eval_Qnli-dev_cosine_accuracy": 0.7395833333333334, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7719540596008301, + "eval_Qnli-dev_cosine_ap": 0.7588639733888536, + "eval_Qnli-dev_cosine_f1": 0.7454545454545455, + "eval_Qnli-dev_cosine_f1_threshold": 0.7090869545936584, + "eval_Qnli-dev_cosine_mcc": 0.47013467657639685, + "eval_Qnli-dev_cosine_precision": 0.6307692307692307, + "eval_Qnli-dev_cosine_recall": 0.9111111111111111, + "eval_allNLI--triplets-1024_cosine_accuracy": 0.8854166865348816, + "eval_allNLI-triplets_cosine_accuracy": 0.9583333134651184, + "eval_global_dataset_loss": 0.4855804145336151, + "eval_global_dataset_runtime": 104.3189, + "eval_global_dataset_samples_per_second": 7.698, + "eval_global_dataset_steps_per_second": 0.163, + "eval_sequential_score": 0.8854166865348816, + "eval_sts-test-1024_pearson_cosine": 0.8681461030339531, + "eval_sts-test-1024_spearman_cosine": 0.9053809631987397, + "eval_sts-test_pearson_cosine": 0.9051731986667259, + "eval_sts-test_spearman_cosine": 0.920630429781229, + "step": 294 + }, + { + "epoch": 0.3007135575942915, + "grad_norm": 7.649487495422363, + "learning_rate": 6.6289592760181e-05, + "loss": 1.4442, + "step": 295 + }, + { + "epoch": 0.30173292558613657, + "grad_norm": 7.740142822265625, + "learning_rate": 6.65158371040724e-05, + "loss": 1.4064, + "step": 296 + }, + { + "epoch": 0.30275229357798167, + "grad_norm": 8.009271621704102, + "learning_rate": 6.67420814479638e-05, + "loss": 0.8456, + "step": 297 + }, + { + "epoch": 0.3037716615698267, + "grad_norm": 5.718809604644775, + "learning_rate": 6.69683257918552e-05, + "loss": 0.7772, + "step": 298 + }, + { + "epoch": 0.30479102956167176, + "grad_norm": 7.34658145904541, + "learning_rate": 6.719457013574662e-05, + "loss": 0.7619, + "step": 299 + }, + { + "epoch": 0.3058103975535168, + "grad_norm": 8.556058883666992, + "learning_rate": 6.742081447963802e-05, + "loss": 1.002, + "step": 300 + }, + { + "epoch": 0.3068297655453619, + "grad_norm": 8.995348930358887, + "learning_rate": 6.764705882352942e-05, + "loss": 1.4774, + "step": 301 + }, + { + "epoch": 0.30784913353720694, + "grad_norm": 8.271109580993652, + "learning_rate": 6.787330316742082e-05, + "loss": 0.6814, + "step": 302 + }, + { + "epoch": 0.308868501529052, + "grad_norm": 8.896450996398926, + "learning_rate": 6.809954751131222e-05, + "loss": 1.2286, + "step": 303 + }, + { + "epoch": 0.30988786952089703, + "grad_norm": 9.076520919799805, + "learning_rate": 6.832579185520362e-05, + "loss": 1.8546, + "step": 304 + }, + { + "epoch": 0.3109072375127421, + "grad_norm": 6.780123233795166, + "learning_rate": 6.855203619909502e-05, + "loss": 0.7547, + "step": 305 + }, + { + "epoch": 0.3119266055045872, + "grad_norm": 7.728740692138672, + "learning_rate": 6.877828054298642e-05, + "loss": 1.3638, + "step": 306 + }, + { + "epoch": 0.3129459734964322, + "grad_norm": 8.941544532775879, + "learning_rate": 6.900452488687784e-05, + "loss": 1.2604, + "step": 307 + }, + { + "epoch": 0.31396534148827726, + "grad_norm": 6.70719575881958, + "learning_rate": 6.923076923076924e-05, + "loss": 0.5111, + "step": 308 + }, + { + "epoch": 0.3149847094801223, + "grad_norm": 7.599255561828613, + "learning_rate": 6.945701357466064e-05, + "loss": 0.7153, + "step": 309 + }, + { + "epoch": 0.3160040774719674, + "grad_norm": 7.323727607727051, + "learning_rate": 6.968325791855204e-05, + "loss": 0.8367, + "step": 310 + }, + { + "epoch": 0.31702344546381245, + "grad_norm": 7.314160346984863, + "learning_rate": 6.990950226244344e-05, + "loss": 0.6643, + "step": 311 + }, + { + "epoch": 0.3180428134556575, + "grad_norm": 8.265671730041504, + "learning_rate": 7.013574660633484e-05, + "loss": 1.0404, + "step": 312 + }, + { + "epoch": 0.31906218144750254, + "grad_norm": 10.820046424865723, + "learning_rate": 7.036199095022625e-05, + "loss": 1.122, + "step": 313 + }, + { + "epoch": 0.3200815494393476, + "grad_norm": 7.194378852844238, + "learning_rate": 7.058823529411765e-05, + "loss": 0.7091, + "step": 314 + }, + { + "epoch": 0.3211009174311927, + "grad_norm": 7.764474868774414, + "learning_rate": 7.081447963800906e-05, + "loss": 1.07, + "step": 315 + }, + { + "epoch": 0.3221202854230377, + "grad_norm": 7.757960796356201, + "learning_rate": 7.104072398190046e-05, + "loss": 0.7246, + "step": 316 + }, + { + "epoch": 0.32313965341488277, + "grad_norm": 9.631681442260742, + "learning_rate": 7.126696832579186e-05, + "loss": 0.821, + "step": 317 + }, + { + "epoch": 0.3241590214067278, + "grad_norm": 6.478396892547607, + "learning_rate": 7.149321266968326e-05, + "loss": 0.9556, + "step": 318 + }, + { + "epoch": 0.3251783893985729, + "grad_norm": 8.858171463012695, + "learning_rate": 7.171945701357467e-05, + "loss": 1.2634, + "step": 319 + }, + { + "epoch": 0.32619775739041795, + "grad_norm": 8.02340030670166, + "learning_rate": 7.194570135746607e-05, + "loss": 0.8621, + "step": 320 + }, + { + "epoch": 0.327217125382263, + "grad_norm": 8.634239196777344, + "learning_rate": 7.217194570135747e-05, + "loss": 1.4215, + "step": 321 + }, + { + "epoch": 0.32823649337410804, + "grad_norm": 8.96740436553955, + "learning_rate": 7.239819004524887e-05, + "loss": 0.8894, + "step": 322 + }, + { + "epoch": 0.3292558613659531, + "grad_norm": 4.730165958404541, + "learning_rate": 7.262443438914027e-05, + "loss": 0.4134, + "step": 323 + }, + { + "epoch": 0.3302752293577982, + "grad_norm": 6.1243181228637695, + "learning_rate": 7.285067873303167e-05, + "loss": 0.4147, + "step": 324 + }, + { + "epoch": 0.3312945973496432, + "grad_norm": 7.8853607177734375, + "learning_rate": 7.307692307692307e-05, + "loss": 0.5721, + "step": 325 + }, + { + "epoch": 0.33231396534148827, + "grad_norm": 9.193514823913574, + "learning_rate": 7.330316742081448e-05, + "loss": 0.8541, + "step": 326 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 12.314509391784668, + "learning_rate": 7.352941176470589e-05, + "loss": 2.2959, + "step": 327 + }, + { + "epoch": 0.3343527013251784, + "grad_norm": 4.384552955627441, + "learning_rate": 7.375565610859729e-05, + "loss": 0.4452, + "step": 328 + }, + { + "epoch": 0.33537206931702346, + "grad_norm": 5.677075386047363, + "learning_rate": 7.398190045248869e-05, + "loss": 0.5008, + "step": 329 + }, + { + "epoch": 0.3363914373088685, + "grad_norm": 6.752626419067383, + "learning_rate": 7.420814479638009e-05, + "loss": 0.4106, + "step": 330 + }, + { + "epoch": 0.33741080530071355, + "grad_norm": 10.971478462219238, + "learning_rate": 7.44343891402715e-05, + "loss": 0.9237, + "step": 331 + }, + { + "epoch": 0.3384301732925586, + "grad_norm": 7.574080944061279, + "learning_rate": 7.46606334841629e-05, + "loss": 0.6275, + "step": 332 + }, + { + "epoch": 0.3394495412844037, + "grad_norm": 9.538507461547852, + "learning_rate": 7.48868778280543e-05, + "loss": 1.5184, + "step": 333 + }, + { + "epoch": 0.34046890927624873, + "grad_norm": 9.139626502990723, + "learning_rate": 7.511312217194571e-05, + "loss": 1.7865, + "step": 334 + }, + { + "epoch": 0.3414882772680938, + "grad_norm": 9.275596618652344, + "learning_rate": 7.533936651583711e-05, + "loss": 1.5947, + "step": 335 + }, + { + "epoch": 0.3425076452599388, + "grad_norm": 9.375283241271973, + "learning_rate": 7.556561085972851e-05, + "loss": 1.0249, + "step": 336 + }, + { + "epoch": 0.3435270132517839, + "grad_norm": 7.951083660125732, + "learning_rate": 7.579185520361991e-05, + "loss": 1.0227, + "step": 337 + }, + { + "epoch": 0.34454638124362896, + "grad_norm": 9.579297065734863, + "learning_rate": 7.601809954751131e-05, + "loss": 1.28, + "step": 338 + }, + { + "epoch": 0.345565749235474, + "grad_norm": 5.935997486114502, + "learning_rate": 7.624434389140271e-05, + "loss": 0.798, + "step": 339 + }, + { + "epoch": 0.34658511722731905, + "grad_norm": 7.16936731338501, + "learning_rate": 7.647058823529411e-05, + "loss": 1.0408, + "step": 340 + }, + { + "epoch": 0.3476044852191641, + "grad_norm": 9.448662757873535, + "learning_rate": 7.669683257918553e-05, + "loss": 0.9732, + "step": 341 + }, + { + "epoch": 0.3486238532110092, + "grad_norm": 7.747692584991455, + "learning_rate": 7.692307692307693e-05, + "loss": 0.7588, + "step": 342 + }, + { + "epoch": 0.34964322120285424, + "grad_norm": 10.198869705200195, + "learning_rate": 7.714932126696833e-05, + "loss": 0.9615, + "step": 343 + }, + { + "epoch": 0.3506625891946993, + "grad_norm": 8.069470405578613, + "learning_rate": 7.737556561085974e-05, + "loss": 0.9895, + "step": 344 + }, + { + "epoch": 0.3516819571865443, + "grad_norm": 10.662049293518066, + "learning_rate": 7.760180995475114e-05, + "loss": 1.923, + "step": 345 + }, + { + "epoch": 0.3527013251783894, + "grad_norm": 6.53238582611084, + "learning_rate": 7.782805429864254e-05, + "loss": 0.615, + "step": 346 + }, + { + "epoch": 0.35372069317023447, + "grad_norm": 11.10132122039795, + "learning_rate": 7.805429864253394e-05, + "loss": 1.4572, + "step": 347 + }, + { + "epoch": 0.3547400611620795, + "grad_norm": 7.372711181640625, + "learning_rate": 7.828054298642534e-05, + "loss": 1.0083, + "step": 348 + }, + { + "epoch": 0.35575942915392456, + "grad_norm": 7.358077526092529, + "learning_rate": 7.850678733031674e-05, + "loss": 0.922, + "step": 349 + }, + { + "epoch": 0.3567787971457696, + "grad_norm": 8.45017147064209, + "learning_rate": 7.873303167420814e-05, + "loss": 1.3767, + "step": 350 + }, + { + "epoch": 0.3577981651376147, + "grad_norm": 4.858506679534912, + "learning_rate": 7.895927601809954e-05, + "loss": 0.6378, + "step": 351 + }, + { + "epoch": 0.35881753312945974, + "grad_norm": 5.764273643493652, + "learning_rate": 7.918552036199095e-05, + "loss": 0.4063, + "step": 352 + }, + { + "epoch": 0.3598369011213048, + "grad_norm": 8.656686782836914, + "learning_rate": 7.941176470588235e-05, + "loss": 1.0834, + "step": 353 + }, + { + "epoch": 0.36085626911314983, + "grad_norm": 5.824944496154785, + "learning_rate": 7.963800904977376e-05, + "loss": 0.807, + "step": 354 + }, + { + "epoch": 0.36187563710499493, + "grad_norm": 6.73368501663208, + "learning_rate": 7.986425339366516e-05, + "loss": 1.0293, + "step": 355 + }, + { + "epoch": 0.36289500509684, + "grad_norm": 5.860096454620361, + "learning_rate": 8.009049773755657e-05, + "loss": 0.4371, + "step": 356 + }, + { + "epoch": 0.363914373088685, + "grad_norm": 5.65436315536499, + "learning_rate": 8.031674208144798e-05, + "loss": 0.4334, + "step": 357 + }, + { + "epoch": 0.36493374108053006, + "grad_norm": 7.566843509674072, + "learning_rate": 8.054298642533938e-05, + "loss": 0.949, + "step": 358 + }, + { + "epoch": 0.3659531090723751, + "grad_norm": 6.286118984222412, + "learning_rate": 8.076923076923078e-05, + "loss": 0.5788, + "step": 359 + }, + { + "epoch": 0.3669724770642202, + "grad_norm": 10.212640762329102, + "learning_rate": 8.099547511312218e-05, + "loss": 0.8535, + "step": 360 + }, + { + "epoch": 0.36799184505606525, + "grad_norm": 9.267760276794434, + "learning_rate": 8.122171945701358e-05, + "loss": 1.2529, + "step": 361 + }, + { + "epoch": 0.3690112130479103, + "grad_norm": 8.794651985168457, + "learning_rate": 8.144796380090498e-05, + "loss": 0.8974, + "step": 362 + }, + { + "epoch": 0.37003058103975534, + "grad_norm": 13.3441162109375, + "learning_rate": 8.167420814479638e-05, + "loss": 1.9105, + "step": 363 + }, + { + "epoch": 0.37104994903160043, + "grad_norm": 9.258030891418457, + "learning_rate": 8.190045248868778e-05, + "loss": 0.7717, + "step": 364 + }, + { + "epoch": 0.3720693170234455, + "grad_norm": 6.051854610443115, + "learning_rate": 8.212669683257918e-05, + "loss": 1.1052, + "step": 365 + }, + { + "epoch": 0.3730886850152905, + "grad_norm": 9.53382682800293, + "learning_rate": 8.23529411764706e-05, + "loss": 0.7298, + "step": 366 + }, + { + "epoch": 0.37410805300713557, + "grad_norm": 6.723752498626709, + "learning_rate": 8.2579185520362e-05, + "loss": 0.7039, + "step": 367 + }, + { + "epoch": 0.3751274209989806, + "grad_norm": 6.844725608825684, + "learning_rate": 8.28054298642534e-05, + "loss": 0.8536, + "step": 368 + }, + { + "epoch": 0.3761467889908257, + "grad_norm": 5.233691692352295, + "learning_rate": 8.303167420814481e-05, + "loss": 0.4774, + "step": 369 + }, + { + "epoch": 0.37716615698267075, + "grad_norm": 4.231795787811279, + "learning_rate": 8.325791855203621e-05, + "loss": 0.3297, + "step": 370 + }, + { + "epoch": 0.3781855249745158, + "grad_norm": 11.760458946228027, + "learning_rate": 8.348416289592761e-05, + "loss": 1.693, + "step": 371 + }, + { + "epoch": 0.37920489296636084, + "grad_norm": 10.05996036529541, + "learning_rate": 8.371040723981901e-05, + "loss": 0.853, + "step": 372 + }, + { + "epoch": 0.38022426095820594, + "grad_norm": 8.649154663085938, + "learning_rate": 8.393665158371041e-05, + "loss": 0.7242, + "step": 373 + }, + { + "epoch": 0.381243628950051, + "grad_norm": 6.6194748878479, + "learning_rate": 8.416289592760181e-05, + "loss": 0.5019, + "step": 374 + }, + { + "epoch": 0.382262996941896, + "grad_norm": 8.058365821838379, + "learning_rate": 8.438914027149321e-05, + "loss": 0.6206, + "step": 375 + }, + { + "epoch": 0.38328236493374107, + "grad_norm": 6.66504430770874, + "learning_rate": 8.461538461538461e-05, + "loss": 0.4872, + "step": 376 + }, + { + "epoch": 0.3843017329255861, + "grad_norm": 5.8679518699646, + "learning_rate": 8.484162895927601e-05, + "loss": 0.4515, + "step": 377 + }, + { + "epoch": 0.3853211009174312, + "grad_norm": 9.830297470092773, + "learning_rate": 8.506787330316743e-05, + "loss": 1.4657, + "step": 378 + }, + { + "epoch": 0.38634046890927626, + "grad_norm": 8.260361671447754, + "learning_rate": 8.529411764705883e-05, + "loss": 0.8411, + "step": 379 + }, + { + "epoch": 0.3873598369011213, + "grad_norm": 8.48035717010498, + "learning_rate": 8.552036199095023e-05, + "loss": 0.7654, + "step": 380 + }, + { + "epoch": 0.38837920489296635, + "grad_norm": 7.481667518615723, + "learning_rate": 8.574660633484163e-05, + "loss": 0.5413, + "step": 381 + }, + { + "epoch": 0.3893985728848114, + "grad_norm": 5.923032760620117, + "learning_rate": 8.597285067873304e-05, + "loss": 0.4594, + "step": 382 + }, + { + "epoch": 0.3904179408766565, + "grad_norm": 11.383003234863281, + "learning_rate": 8.619909502262445e-05, + "loss": 1.2656, + "step": 383 + }, + { + "epoch": 0.39143730886850153, + "grad_norm": 9.154252052307129, + "learning_rate": 8.642533936651585e-05, + "loss": 0.6881, + "step": 384 + }, + { + "epoch": 0.3924566768603466, + "grad_norm": 8.656584739685059, + "learning_rate": 8.665158371040725e-05, + "loss": 0.8169, + "step": 385 + }, + { + "epoch": 0.3934760448521916, + "grad_norm": 9.6775541305542, + "learning_rate": 8.687782805429865e-05, + "loss": 0.937, + "step": 386 + }, + { + "epoch": 0.3944954128440367, + "grad_norm": 12.836816787719727, + "learning_rate": 8.710407239819005e-05, + "loss": 2.1343, + "step": 387 + }, + { + "epoch": 0.39551478083588176, + "grad_norm": 6.1532487869262695, + "learning_rate": 8.733031674208145e-05, + "loss": 0.3644, + "step": 388 + }, + { + "epoch": 0.3965341488277268, + "grad_norm": 6.3952555656433105, + "learning_rate": 8.755656108597285e-05, + "loss": 0.4406, + "step": 389 + }, + { + "epoch": 0.39755351681957185, + "grad_norm": 7.005934238433838, + "learning_rate": 8.778280542986425e-05, + "loss": 0.5444, + "step": 390 + }, + { + "epoch": 0.3985728848114169, + "grad_norm": 8.97732925415039, + "learning_rate": 8.800904977375566e-05, + "loss": 1.3891, + "step": 391 + }, + { + "epoch": 0.399592252803262, + "grad_norm": 6.8778181076049805, + "learning_rate": 8.823529411764706e-05, + "loss": 0.6287, + "step": 392 + }, + { + "epoch": 0.399592252803262, + "eval_Qnli-dev-1024_cosine_accuracy": 0.7395833333333334, + "eval_Qnli-dev-1024_cosine_accuracy_threshold": 0.8465664982795715, + "eval_Qnli-dev-1024_cosine_ap": 0.7683064400770494, + "eval_Qnli-dev-1024_cosine_f1": 0.6976744186046511, + "eval_Qnli-dev-1024_cosine_f1_threshold": 0.834477424621582, + "eval_Qnli-dev-1024_cosine_mcc": 0.45496263625850347, + "eval_Qnli-dev-1024_cosine_precision": 0.7317073170731707, + "eval_Qnli-dev-1024_cosine_recall": 0.6666666666666666, + "eval_Qnli-dev_cosine_accuracy": 0.75, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.74493807554245, + "eval_Qnli-dev_cosine_ap": 0.7575725381948821, + "eval_Qnli-dev_cosine_f1": 0.7476635514018692, + "eval_Qnli-dev_cosine_f1_threshold": 0.7015562057495117, + "eval_Qnli-dev_cosine_mcc": 0.47737827504723207, + "eval_Qnli-dev_cosine_precision": 0.6451612903225806, + "eval_Qnli-dev_cosine_recall": 0.8888888888888888, + "eval_allNLI--triplets-1024_cosine_accuracy": 0.9270833134651184, + "eval_allNLI-triplets_cosine_accuracy": 0.9479166865348816, + "eval_global_dataset_loss": 0.3703947365283966, + "eval_global_dataset_runtime": 104.3143, + "eval_global_dataset_samples_per_second": 7.698, + "eval_global_dataset_steps_per_second": 0.163, + "eval_sequential_score": 0.9270833134651184, + "eval_sts-test-1024_pearson_cosine": 0.8782079507952609, + "eval_sts-test-1024_spearman_cosine": 0.9080003485202497, + "eval_sts-test_pearson_cosine": 0.9052799671643099, + "eval_sts-test_spearman_cosine": 0.9200953636370672, + "step": 392 + }, + { + "epoch": 0.40061162079510704, + "grad_norm": 7.236085414886475, + "learning_rate": 8.846153846153847e-05, + "loss": 1.066, + "step": 393 + }, + { + "epoch": 0.4016309887869521, + "grad_norm": 7.638827323913574, + "learning_rate": 8.868778280542987e-05, + "loss": 1.0406, + "step": 394 + }, + { + "epoch": 0.4026503567787971, + "grad_norm": 6.2278876304626465, + "learning_rate": 8.891402714932127e-05, + "loss": 0.819, + "step": 395 + }, + { + "epoch": 0.4036697247706422, + "grad_norm": 7.04884147644043, + "learning_rate": 8.914027149321268e-05, + "loss": 0.5826, + "step": 396 + }, + { + "epoch": 0.40468909276248727, + "grad_norm": 8.24869441986084, + "learning_rate": 8.936651583710408e-05, + "loss": 0.6355, + "step": 397 + }, + { + "epoch": 0.4057084607543323, + "grad_norm": 9.9276704788208, + "learning_rate": 8.959276018099548e-05, + "loss": 0.7566, + "step": 398 + }, + { + "epoch": 0.40672782874617736, + "grad_norm": 8.717905044555664, + "learning_rate": 8.981900452488688e-05, + "loss": 0.8174, + "step": 399 + }, + { + "epoch": 0.4077471967380224, + "grad_norm": 8.515538215637207, + "learning_rate": 9.004524886877828e-05, + "loss": 0.8905, + "step": 400 + }, + { + "epoch": 0.4087665647298675, + "grad_norm": 6.506967067718506, + "learning_rate": 9.027149321266968e-05, + "loss": 0.6646, + "step": 401 + }, + { + "epoch": 0.40978593272171254, + "grad_norm": 9.33711051940918, + "learning_rate": 9.049773755656108e-05, + "loss": 0.9056, + "step": 402 + }, + { + "epoch": 0.4108053007135576, + "grad_norm": 5.124199867248535, + "learning_rate": 9.07239819004525e-05, + "loss": 0.3689, + "step": 403 + }, + { + "epoch": 0.41182466870540263, + "grad_norm": 5.597712516784668, + "learning_rate": 9.09502262443439e-05, + "loss": 0.3709, + "step": 404 + }, + { + "epoch": 0.41284403669724773, + "grad_norm": 7.897356033325195, + "learning_rate": 9.11764705882353e-05, + "loss": 0.6708, + "step": 405 + }, + { + "epoch": 0.4138634046890928, + "grad_norm": 8.37096881866455, + "learning_rate": 9.14027149321267e-05, + "loss": 1.0531, + "step": 406 + }, + { + "epoch": 0.4148827726809378, + "grad_norm": 7.530358791351318, + "learning_rate": 9.16289592760181e-05, + "loss": 1.1355, + "step": 407 + }, + { + "epoch": 0.41590214067278286, + "grad_norm": 10.304217338562012, + "learning_rate": 9.18552036199095e-05, + "loss": 0.8042, + "step": 408 + }, + { + "epoch": 0.4169215086646279, + "grad_norm": 7.292766094207764, + "learning_rate": 9.20814479638009e-05, + "loss": 0.3915, + "step": 409 + }, + { + "epoch": 0.417940876656473, + "grad_norm": 10.453197479248047, + "learning_rate": 9.230769230769232e-05, + "loss": 1.9388, + "step": 410 + }, + { + "epoch": 0.41896024464831805, + "grad_norm": 2.7471694946289062, + "learning_rate": 9.253393665158372e-05, + "loss": 0.3044, + "step": 411 + }, + { + "epoch": 0.4199796126401631, + "grad_norm": 5.923367023468018, + "learning_rate": 9.276018099547512e-05, + "loss": 0.6153, + "step": 412 + }, + { + "epoch": 0.42099898063200814, + "grad_norm": 8.176202774047852, + "learning_rate": 9.298642533936652e-05, + "loss": 0.9407, + "step": 413 + }, + { + "epoch": 0.42201834862385323, + "grad_norm": 8.41361141204834, + "learning_rate": 9.321266968325792e-05, + "loss": 0.6876, + "step": 414 + }, + { + "epoch": 0.4230377166156983, + "grad_norm": 9.516852378845215, + "learning_rate": 9.343891402714933e-05, + "loss": 0.9694, + "step": 415 + }, + { + "epoch": 0.4240570846075433, + "grad_norm": 7.201638698577881, + "learning_rate": 9.366515837104073e-05, + "loss": 0.7868, + "step": 416 + }, + { + "epoch": 0.42507645259938837, + "grad_norm": 9.961840629577637, + "learning_rate": 9.389140271493213e-05, + "loss": 0.7735, + "step": 417 + }, + { + "epoch": 0.4260958205912334, + "grad_norm": 10.842241287231445, + "learning_rate": 9.411764705882353e-05, + "loss": 1.1682, + "step": 418 + }, + { + "epoch": 0.4271151885830785, + "grad_norm": 5.817572116851807, + "learning_rate": 9.434389140271494e-05, + "loss": 0.3465, + "step": 419 + }, + { + "epoch": 0.42813455657492355, + "grad_norm": 6.870133399963379, + "learning_rate": 9.457013574660634e-05, + "loss": 0.5699, + "step": 420 + }, + { + "epoch": 0.4291539245667686, + "grad_norm": 6.472342014312744, + "learning_rate": 9.479638009049774e-05, + "loss": 0.6128, + "step": 421 + }, + { + "epoch": 0.43017329255861364, + "grad_norm": 6.5723795890808105, + "learning_rate": 9.502262443438914e-05, + "loss": 0.8886, + "step": 422 + }, + { + "epoch": 0.43119266055045874, + "grad_norm": 6.1384429931640625, + "learning_rate": 9.524886877828054e-05, + "loss": 0.5124, + "step": 423 + }, + { + "epoch": 0.4322120285423038, + "grad_norm": 6.241471290588379, + "learning_rate": 9.547511312217195e-05, + "loss": 0.4409, + "step": 424 + }, + { + "epoch": 0.4332313965341488, + "grad_norm": 9.087861061096191, + "learning_rate": 9.570135746606335e-05, + "loss": 0.6368, + "step": 425 + }, + { + "epoch": 0.43425076452599387, + "grad_norm": 9.653539657592773, + "learning_rate": 9.592760180995475e-05, + "loss": 0.9874, + "step": 426 + }, + { + "epoch": 0.4352701325178389, + "grad_norm": 13.366517066955566, + "learning_rate": 9.615384615384617e-05, + "loss": 1.6544, + "step": 427 + }, + { + "epoch": 0.436289500509684, + "grad_norm": 6.302597522735596, + "learning_rate": 9.638009049773757e-05, + "loss": 0.4561, + "step": 428 + }, + { + "epoch": 0.43730886850152906, + "grad_norm": 7.133030891418457, + "learning_rate": 9.660633484162897e-05, + "loss": 0.5443, + "step": 429 + }, + { + "epoch": 0.4383282364933741, + "grad_norm": 6.341556072235107, + "learning_rate": 9.683257918552037e-05, + "loss": 0.5183, + "step": 430 + }, + { + "epoch": 0.43934760448521915, + "grad_norm": 10.657116889953613, + "learning_rate": 9.705882352941177e-05, + "loss": 1.1585, + "step": 431 + }, + { + "epoch": 0.44036697247706424, + "grad_norm": 7.707142353057861, + "learning_rate": 9.728506787330317e-05, + "loss": 1.4285, + "step": 432 + }, + { + "epoch": 0.4413863404689093, + "grad_norm": 8.27905559539795, + "learning_rate": 9.751131221719457e-05, + "loss": 1.0638, + "step": 433 + }, + { + "epoch": 0.44240570846075433, + "grad_norm": 5.601058483123779, + "learning_rate": 9.773755656108597e-05, + "loss": 0.553, + "step": 434 + }, + { + "epoch": 0.4434250764525994, + "grad_norm": 9.084299087524414, + "learning_rate": 9.796380090497737e-05, + "loss": 1.0009, + "step": 435 + }, + { + "epoch": 0.4444444444444444, + "grad_norm": 5.231532573699951, + "learning_rate": 9.819004524886877e-05, + "loss": 0.5211, + "step": 436 + }, + { + "epoch": 0.4454638124362895, + "grad_norm": 7.0478715896606445, + "learning_rate": 9.841628959276019e-05, + "loss": 0.6483, + "step": 437 + }, + { + "epoch": 0.44648318042813456, + "grad_norm": 8.44166088104248, + "learning_rate": 9.864253393665159e-05, + "loss": 1.2634, + "step": 438 + }, + { + "epoch": 0.4475025484199796, + "grad_norm": 7.2984771728515625, + "learning_rate": 9.8868778280543e-05, + "loss": 0.5242, + "step": 439 + }, + { + "epoch": 0.44852191641182465, + "grad_norm": 9.091867446899414, + "learning_rate": 9.90950226244344e-05, + "loss": 1.3739, + "step": 440 + }, + { + "epoch": 0.44954128440366975, + "grad_norm": 8.1068115234375, + "learning_rate": 9.93212669683258e-05, + "loss": 1.0153, + "step": 441 + }, + { + "epoch": 0.4505606523955148, + "grad_norm": 7.902680397033691, + "learning_rate": 9.95475113122172e-05, + "loss": 0.7174, + "step": 442 + }, + { + "epoch": 0.45158002038735984, + "grad_norm": 8.784537315368652, + "learning_rate": 9.97737556561086e-05, + "loss": 0.8631, + "step": 443 + }, + { + "epoch": 0.4525993883792049, + "grad_norm": 8.205148696899414, + "learning_rate": 0.0001, + "loss": 1.0721, + "step": 444 + }, + { + "epoch": 0.4536187563710499, + "grad_norm": 4.789169788360596, + "learning_rate": 9.999964497873585e-05, + "loss": 0.3682, + "step": 445 + }, + { + "epoch": 0.454638124362895, + "grad_norm": 11.335341453552246, + "learning_rate": 9.999857991998499e-05, + "loss": 1.3278, + "step": 446 + }, + { + "epoch": 0.45565749235474007, + "grad_norm": 8.901962280273438, + "learning_rate": 9.999680483887217e-05, + "loss": 0.665, + "step": 447 + }, + { + "epoch": 0.4566768603465851, + "grad_norm": 6.525248050689697, + "learning_rate": 9.999431976060504e-05, + "loss": 0.77, + "step": 448 + }, + { + "epoch": 0.45769622833843016, + "grad_norm": 7.658937931060791, + "learning_rate": 9.999112472047386e-05, + "loss": 0.9903, + "step": 449 + }, + { + "epoch": 0.45871559633027525, + "grad_norm": 5.406915664672852, + "learning_rate": 9.998721976385087e-05, + "loss": 0.3372, + "step": 450 + }, + { + "epoch": 0.4597349643221203, + "grad_norm": 5.920129299163818, + "learning_rate": 9.998260494618979e-05, + "loss": 0.6911, + "step": 451 + }, + { + "epoch": 0.46075433231396534, + "grad_norm": 7.490262985229492, + "learning_rate": 9.997728033302496e-05, + "loss": 0.505, + "step": 452 + }, + { + "epoch": 0.4617737003058104, + "grad_norm": 8.21649169921875, + "learning_rate": 9.997124599997043e-05, + "loss": 1.3397, + "step": 453 + }, + { + "epoch": 0.46279306829765543, + "grad_norm": 5.116532802581787, + "learning_rate": 9.996450203271886e-05, + "loss": 0.2853, + "step": 454 + }, + { + "epoch": 0.46381243628950053, + "grad_norm": 7.29067325592041, + "learning_rate": 9.995704852704029e-05, + "loss": 1.198, + "step": 455 + }, + { + "epoch": 0.4648318042813456, + "grad_norm": 10.033268928527832, + "learning_rate": 9.994888558878086e-05, + "loss": 1.7965, + "step": 456 + }, + { + "epoch": 0.4658511722731906, + "grad_norm": 5.4102606773376465, + "learning_rate": 9.994001333386125e-05, + "loss": 0.2987, + "step": 457 + }, + { + "epoch": 0.46687054026503566, + "grad_norm": 8.109895706176758, + "learning_rate": 9.993043188827501e-05, + "loss": 0.6864, + "step": 458 + }, + { + "epoch": 0.46788990825688076, + "grad_norm": 9.893292427062988, + "learning_rate": 9.992014138808682e-05, + "loss": 0.9016, + "step": 459 + }, + { + "epoch": 0.4689092762487258, + "grad_norm": 7.73169469833374, + "learning_rate": 9.990914197943053e-05, + "loss": 0.7314, + "step": 460 + }, + { + "epoch": 0.46992864424057085, + "grad_norm": 8.335735321044922, + "learning_rate": 9.989743381850711e-05, + "loss": 0.6633, + "step": 461 + }, + { + "epoch": 0.4709480122324159, + "grad_norm": 8.655631065368652, + "learning_rate": 9.988501707158243e-05, + "loss": 0.9783, + "step": 462 + }, + { + "epoch": 0.47196738022426094, + "grad_norm": 9.166102409362793, + "learning_rate": 9.987189191498479e-05, + "loss": 1.1307, + "step": 463 + }, + { + "epoch": 0.47298674821610603, + "grad_norm": 10.597552299499512, + "learning_rate": 9.985805853510262e-05, + "loss": 1.4662, + "step": 464 + }, + { + "epoch": 0.4740061162079511, + "grad_norm": 10.318975448608398, + "learning_rate": 9.984351712838167e-05, + "loss": 1.4666, + "step": 465 + }, + { + "epoch": 0.4750254841997961, + "grad_norm": 7.259106636047363, + "learning_rate": 9.98282679013223e-05, + "loss": 0.5918, + "step": 466 + }, + { + "epoch": 0.47604485219164117, + "grad_norm": 10.192667961120605, + "learning_rate": 9.981231107047648e-05, + "loss": 1.5836, + "step": 467 + }, + { + "epoch": 0.47706422018348627, + "grad_norm": 6.506603717803955, + "learning_rate": 9.97956468624448e-05, + "loss": 0.5698, + "step": 468 + }, + { + "epoch": 0.4780835881753313, + "grad_norm": 5.789127349853516, + "learning_rate": 9.977827551387318e-05, + "loss": 0.4654, + "step": 469 + }, + { + "epoch": 0.47910295616717635, + "grad_norm": 8.615316390991211, + "learning_rate": 9.976019727144956e-05, + "loss": 0.9522, + "step": 470 + }, + { + "epoch": 0.4801223241590214, + "grad_norm": 4.792436599731445, + "learning_rate": 9.974141239190034e-05, + "loss": 0.4748, + "step": 471 + }, + { + "epoch": 0.48114169215086644, + "grad_norm": 9.958406448364258, + "learning_rate": 9.972192114198677e-05, + "loss": 2.0766, + "step": 472 + }, + { + "epoch": 0.48216106014271154, + "grad_norm": 4.340735912322998, + "learning_rate": 9.970172379850122e-05, + "loss": 0.7071, + "step": 473 + }, + { + "epoch": 0.4831804281345566, + "grad_norm": 7.170680999755859, + "learning_rate": 9.968082064826314e-05, + "loss": 0.435, + "step": 474 + }, + { + "epoch": 0.4841997961264016, + "grad_norm": 3.927189350128174, + "learning_rate": 9.965921198811501e-05, + "loss": 0.4551, + "step": 475 + }, + { + "epoch": 0.48521916411824667, + "grad_norm": 10.183062553405762, + "learning_rate": 9.96368981249182e-05, + "loss": 1.1758, + "step": 476 + }, + { + "epoch": 0.48623853211009177, + "grad_norm": 9.819293022155762, + "learning_rate": 9.961387937554857e-05, + "loss": 0.9995, + "step": 477 + }, + { + "epoch": 0.4872579001019368, + "grad_norm": 11.188612937927246, + "learning_rate": 9.95901560668919e-05, + "loss": 1.6207, + "step": 478 + }, + { + "epoch": 0.48827726809378186, + "grad_norm": 7.268994331359863, + "learning_rate": 9.95657285358394e-05, + "loss": 0.6978, + "step": 479 + }, + { + "epoch": 0.4892966360856269, + "grad_norm": 5.575627326965332, + "learning_rate": 9.954059712928275e-05, + "loss": 0.4236, + "step": 480 + }, + { + "epoch": 0.49031600407747195, + "grad_norm": 9.621591567993164, + "learning_rate": 9.951476220410929e-05, + "loss": 1.8218, + "step": 481 + }, + { + "epoch": 0.49133537206931704, + "grad_norm": 7.322023391723633, + "learning_rate": 9.948822412719697e-05, + "loss": 0.8749, + "step": 482 + }, + { + "epoch": 0.4923547400611621, + "grad_norm": 8.407424926757812, + "learning_rate": 9.946098327540902e-05, + "loss": 1.1704, + "step": 483 + }, + { + "epoch": 0.49337410805300713, + "grad_norm": 4.8855438232421875, + "learning_rate": 9.943304003558873e-05, + "loss": 0.5327, + "step": 484 + }, + { + "epoch": 0.4943934760448522, + "grad_norm": 8.738515853881836, + "learning_rate": 9.940439480455386e-05, + "loss": 1.2009, + "step": 485 + }, + { + "epoch": 0.4954128440366973, + "grad_norm": 7.554356575012207, + "learning_rate": 9.937504798909106e-05, + "loss": 0.5427, + "step": 486 + }, + { + "epoch": 0.4964322120285423, + "grad_norm": 8.203272819519043, + "learning_rate": 9.934500000595008e-05, + "loss": 0.5893, + "step": 487 + }, + { + "epoch": 0.49745158002038736, + "grad_norm": 8.477286338806152, + "learning_rate": 9.931425128183782e-05, + "loss": 1.061, + "step": 488 + }, + { + "epoch": 0.4984709480122324, + "grad_norm": 7.389923095703125, + "learning_rate": 9.928280225341232e-05, + "loss": 0.5465, + "step": 489 + }, + { + "epoch": 0.49949031600407745, + "grad_norm": 10.051106452941895, + "learning_rate": 9.925065336727654e-05, + "loss": 0.7035, + "step": 490 + }, + { + "epoch": 0.49949031600407745, + "eval_Qnli-dev-1024_cosine_accuracy": 0.7291666666666666, + "eval_Qnli-dev-1024_cosine_accuracy_threshold": 0.817620038986206, + "eval_Qnli-dev-1024_cosine_ap": 0.7443202788050278, + "eval_Qnli-dev-1024_cosine_f1": 0.7291666666666667, + "eval_Qnli-dev-1024_cosine_f1_threshold": 0.817620038986206, + "eval_Qnli-dev-1024_cosine_mcc": 0.46405228758169936, + "eval_Qnli-dev-1024_cosine_precision": 0.6862745098039216, + "eval_Qnli-dev-1024_cosine_recall": 0.7777777777777778, + "eval_Qnli-dev_cosine_accuracy": 0.7395833333333334, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7795548439025879, + "eval_Qnli-dev_cosine_ap": 0.7446338608862075, + "eval_Qnli-dev_cosine_f1": 0.7378640776699029, + "eval_Qnli-dev_cosine_f1_threshold": 0.6985307335853577, + "eval_Qnli-dev_cosine_mcc": 0.46153029495329345, + "eval_Qnli-dev_cosine_precision": 0.6551724137931034, + "eval_Qnli-dev_cosine_recall": 0.8444444444444444, + "eval_allNLI--triplets-1024_cosine_accuracy": 0.90625, + "eval_allNLI-triplets_cosine_accuracy": 0.9375, + "eval_global_dataset_loss": 0.34814590215682983, + "eval_global_dataset_runtime": 104.2751, + "eval_global_dataset_samples_per_second": 7.701, + "eval_global_dataset_steps_per_second": 0.163, + "eval_sequential_score": 0.90625, + "eval_sts-test-1024_pearson_cosine": 0.8574057933500303, + "eval_sts-test-1024_spearman_cosine": 0.8986116241995802, + "eval_sts-test_pearson_cosine": 0.9019111579722014, + "eval_sts-test_spearman_cosine": 0.9181479205822737, + "step": 490 + }, + { + "epoch": 0.5005096839959225, + "grad_norm": 10.88537311553955, + "learning_rate": 9.921780507997202e-05, + "loss": 1.6596, + "step": 491 + }, + { + "epoch": 0.5015290519877675, + "grad_norm": 5.0818891525268555, + "learning_rate": 9.918425785797235e-05, + "loss": 0.4475, + "step": 492 + }, + { + "epoch": 0.5025484199796126, + "grad_norm": 12.540839195251465, + "learning_rate": 9.915001217767663e-05, + "loss": 2.0803, + "step": 493 + }, + { + "epoch": 0.5035677879714577, + "grad_norm": 6.171934604644775, + "learning_rate": 9.911506852540267e-05, + "loss": 0.4296, + "step": 494 + }, + { + "epoch": 0.5045871559633027, + "grad_norm": 9.624109268188477, + "learning_rate": 9.907942739738001e-05, + "loss": 1.5435, + "step": 495 + }, + { + "epoch": 0.5056065239551478, + "grad_norm": 6.020090579986572, + "learning_rate": 9.904308929974302e-05, + "loss": 0.6073, + "step": 496 + }, + { + "epoch": 0.5066258919469928, + "grad_norm": 8.587658882141113, + "learning_rate": 9.900605474852358e-05, + "loss": 1.1774, + "step": 497 + }, + { + "epoch": 0.5076452599388379, + "grad_norm": 6.535181999206543, + "learning_rate": 9.896832426964382e-05, + "loss": 0.8951, + "step": 498 + }, + { + "epoch": 0.508664627930683, + "grad_norm": 5.945138454437256, + "learning_rate": 9.892989839890863e-05, + "loss": 0.3775, + "step": 499 + }, + { + "epoch": 0.509683995922528, + "grad_norm": 7.641120433807373, + "learning_rate": 9.889077768199806e-05, + "loss": 0.8086, + "step": 500 + }, + { + "epoch": 0.5107033639143731, + "grad_norm": 8.10549545288086, + "learning_rate": 9.885096267445957e-05, + "loss": 0.3864, + "step": 501 + }, + { + "epoch": 0.5117227319062182, + "grad_norm": 7.266530990600586, + "learning_rate": 9.881045394170012e-05, + "loss": 0.8865, + "step": 502 + }, + { + "epoch": 0.5127420998980632, + "grad_norm": 9.056779861450195, + "learning_rate": 9.876925205897818e-05, + "loss": 0.567, + "step": 503 + }, + { + "epoch": 0.5137614678899083, + "grad_norm": 7.140566349029541, + "learning_rate": 9.872735761139554e-05, + "loss": 0.9304, + "step": 504 + }, + { + "epoch": 0.5147808358817533, + "grad_norm": 11.422016143798828, + "learning_rate": 9.868477119388896e-05, + "loss": 0.6977, + "step": 505 + }, + { + "epoch": 0.5158002038735984, + "grad_norm": 11.155719757080078, + "learning_rate": 9.864149341122181e-05, + "loss": 1.3174, + "step": 506 + }, + { + "epoch": 0.5168195718654435, + "grad_norm": 8.781103134155273, + "learning_rate": 9.859752487797542e-05, + "loss": 1.2481, + "step": 507 + }, + { + "epoch": 0.5178389398572885, + "grad_norm": 5.503263473510742, + "learning_rate": 9.855286621854034e-05, + "loss": 0.4894, + "step": 508 + }, + { + "epoch": 0.5188583078491336, + "grad_norm": 7.503839015960693, + "learning_rate": 9.850751806710753e-05, + "loss": 0.8095, + "step": 509 + }, + { + "epoch": 0.5198776758409785, + "grad_norm": 5.623706817626953, + "learning_rate": 9.846148106765933e-05, + "loss": 0.388, + "step": 510 + }, + { + "epoch": 0.5208970438328236, + "grad_norm": 5.178555965423584, + "learning_rate": 9.841475587396028e-05, + "loss": 0.6725, + "step": 511 + }, + { + "epoch": 0.5219164118246687, + "grad_norm": 7.296833038330078, + "learning_rate": 9.836734314954785e-05, + "loss": 0.4804, + "step": 512 + }, + { + "epoch": 0.5229357798165137, + "grad_norm": 8.692532539367676, + "learning_rate": 9.831924356772308e-05, + "loss": 1.2414, + "step": 513 + }, + { + "epoch": 0.5239551478083588, + "grad_norm": 9.865914344787598, + "learning_rate": 9.827045781154093e-05, + "loss": 1.0319, + "step": 514 + }, + { + "epoch": 0.5249745158002038, + "grad_norm": 14.857895851135254, + "learning_rate": 9.822098657380065e-05, + "loss": 2.0732, + "step": 515 + }, + { + "epoch": 0.5259938837920489, + "grad_norm": 6.85409688949585, + "learning_rate": 9.817083055703587e-05, + "loss": 1.1168, + "step": 516 + }, + { + "epoch": 0.527013251783894, + "grad_norm": 10.668725967407227, + "learning_rate": 9.811999047350471e-05, + "loss": 0.8056, + "step": 517 + }, + { + "epoch": 0.528032619775739, + "grad_norm": 6.76224946975708, + "learning_rate": 9.806846704517957e-05, + "loss": 0.5322, + "step": 518 + }, + { + "epoch": 0.5290519877675841, + "grad_norm": 4.4465789794921875, + "learning_rate": 9.801626100373699e-05, + "loss": 0.4348, + "step": 519 + }, + { + "epoch": 0.5300713557594292, + "grad_norm": 8.388195991516113, + "learning_rate": 9.796337309054717e-05, + "loss": 0.6316, + "step": 520 + }, + { + "epoch": 0.5310907237512742, + "grad_norm": 5.859539031982422, + "learning_rate": 9.790980405666344e-05, + "loss": 0.3212, + "step": 521 + }, + { + "epoch": 0.5321100917431193, + "grad_norm": 6.299170017242432, + "learning_rate": 9.785555466281162e-05, + "loss": 0.4739, + "step": 522 + }, + { + "epoch": 0.5331294597349643, + "grad_norm": 9.609426498413086, + "learning_rate": 9.780062567937928e-05, + "loss": 1.1692, + "step": 523 + }, + { + "epoch": 0.5341488277268094, + "grad_norm": 9.116230964660645, + "learning_rate": 9.774501788640471e-05, + "loss": 0.941, + "step": 524 + }, + { + "epoch": 0.5351681957186545, + "grad_norm": 5.024673938751221, + "learning_rate": 9.768873207356586e-05, + "loss": 0.3767, + "step": 525 + }, + { + "epoch": 0.5361875637104995, + "grad_norm": 7.534763336181641, + "learning_rate": 9.763176904016913e-05, + "loss": 0.5264, + "step": 526 + }, + { + "epoch": 0.5372069317023446, + "grad_norm": 7.897163391113281, + "learning_rate": 9.757412959513807e-05, + "loss": 0.4345, + "step": 527 + }, + { + "epoch": 0.5382262996941896, + "grad_norm": 8.391239166259766, + "learning_rate": 9.751581455700181e-05, + "loss": 1.0352, + "step": 528 + }, + { + "epoch": 0.5392456676860347, + "grad_norm": 6.951046466827393, + "learning_rate": 9.745682475388348e-05, + "loss": 1.1014, + "step": 529 + }, + { + "epoch": 0.5402650356778798, + "grad_norm": 6.4283671379089355, + "learning_rate": 9.73971610234885e-05, + "loss": 0.7368, + "step": 530 + }, + { + "epoch": 0.5412844036697247, + "grad_norm": 7.643414497375488, + "learning_rate": 9.733682421309256e-05, + "loss": 0.5324, + "step": 531 + }, + { + "epoch": 0.5423037716615698, + "grad_norm": 7.95609188079834, + "learning_rate": 9.727581517952969e-05, + "loss": 0.5351, + "step": 532 + }, + { + "epoch": 0.5433231396534148, + "grad_norm": 11.28146743774414, + "learning_rate": 9.721413478918007e-05, + "loss": 1.6815, + "step": 533 + }, + { + "epoch": 0.5443425076452599, + "grad_norm": 8.346885681152344, + "learning_rate": 9.715178391795769e-05, + "loss": 0.8125, + "step": 534 + }, + { + "epoch": 0.545361875637105, + "grad_norm": 8.147517204284668, + "learning_rate": 9.708876345129797e-05, + "loss": 0.8629, + "step": 535 + }, + { + "epoch": 0.54638124362895, + "grad_norm": 10.061439514160156, + "learning_rate": 9.702507428414513e-05, + "loss": 1.3161, + "step": 536 + }, + { + "epoch": 0.5474006116207951, + "grad_norm": 8.882964134216309, + "learning_rate": 9.696071732093952e-05, + "loss": 1.0465, + "step": 537 + }, + { + "epoch": 0.5484199796126402, + "grad_norm": 5.954410076141357, + "learning_rate": 9.689569347560475e-05, + "loss": 0.4531, + "step": 538 + }, + { + "epoch": 0.5494393476044852, + "grad_norm": 10.33085823059082, + "learning_rate": 9.683000367153474e-05, + "loss": 0.5567, + "step": 539 + }, + { + "epoch": 0.5504587155963303, + "grad_norm": 5.265343189239502, + "learning_rate": 9.676364884158058e-05, + "loss": 0.7093, + "step": 540 + }, + { + "epoch": 0.5514780835881753, + "grad_norm": 10.214452743530273, + "learning_rate": 9.66966299280373e-05, + "loss": 1.9339, + "step": 541 + }, + { + "epoch": 0.5524974515800204, + "grad_norm": 7.001688480377197, + "learning_rate": 9.662894788263044e-05, + "loss": 0.3659, + "step": 542 + }, + { + "epoch": 0.5535168195718655, + "grad_norm": 6.640339374542236, + "learning_rate": 9.656060366650267e-05, + "loss": 1.0505, + "step": 543 + }, + { + "epoch": 0.5545361875637105, + "grad_norm": 9.303877830505371, + "learning_rate": 9.649159825019996e-05, + "loss": 0.8766, + "step": 544 + }, + { + "epoch": 0.5555555555555556, + "grad_norm": 8.21275806427002, + "learning_rate": 9.642193261365791e-05, + "loss": 0.6526, + "step": 545 + }, + { + "epoch": 0.5565749235474006, + "grad_norm": 6.97646427154541, + "learning_rate": 9.635160774618782e-05, + "loss": 0.5529, + "step": 546 + }, + { + "epoch": 0.5575942915392457, + "grad_norm": 6.77686071395874, + "learning_rate": 9.628062464646264e-05, + "loss": 0.4817, + "step": 547 + }, + { + "epoch": 0.5586136595310908, + "grad_norm": 3.5217092037200928, + "learning_rate": 9.620898432250272e-05, + "loss": 0.4804, + "step": 548 + }, + { + "epoch": 0.5596330275229358, + "grad_norm": 5.6369476318359375, + "learning_rate": 9.613668779166165e-05, + "loss": 0.4508, + "step": 549 + }, + { + "epoch": 0.5606523955147809, + "grad_norm": 5.534257888793945, + "learning_rate": 9.606373608061162e-05, + "loss": 0.4339, + "step": 550 + }, + { + "epoch": 0.5616717635066258, + "grad_norm": 10.922380447387695, + "learning_rate": 9.5990130225329e-05, + "loss": 0.712, + "step": 551 + }, + { + "epoch": 0.5626911314984709, + "grad_norm": 6.2288360595703125, + "learning_rate": 9.59158712710795e-05, + "loss": 0.3974, + "step": 552 + }, + { + "epoch": 0.563710499490316, + "grad_norm": 11.958196640014648, + "learning_rate": 9.58409602724035e-05, + "loss": 1.0016, + "step": 553 + }, + { + "epoch": 0.564729867482161, + "grad_norm": 8.267114639282227, + "learning_rate": 9.576539829310085e-05, + "loss": 0.5751, + "step": 554 + }, + { + "epoch": 0.5657492354740061, + "grad_norm": 11.533574104309082, + "learning_rate": 9.568918640621594e-05, + "loss": 1.111, + "step": 555 + }, + { + "epoch": 0.5667686034658511, + "grad_norm": 6.519062519073486, + "learning_rate": 9.561232569402239e-05, + "loss": 0.4202, + "step": 556 + }, + { + "epoch": 0.5677879714576962, + "grad_norm": 9.009593963623047, + "learning_rate": 9.553481724800768e-05, + "loss": 0.7822, + "step": 557 + }, + { + "epoch": 0.5688073394495413, + "grad_norm": 12.121257781982422, + "learning_rate": 9.545666216885767e-05, + "loss": 1.3844, + "step": 558 + }, + { + "epoch": 0.5698267074413863, + "grad_norm": 5.953427314758301, + "learning_rate": 9.537786156644097e-05, + "loss": 0.3881, + "step": 559 + }, + { + "epoch": 0.5708460754332314, + "grad_norm": 7.334780216217041, + "learning_rate": 9.529841655979315e-05, + "loss": 0.6317, + "step": 560 + }, + { + "epoch": 0.5718654434250765, + "grad_norm": 5.987368583679199, + "learning_rate": 9.521832827710088e-05, + "loss": 0.4976, + "step": 561 + }, + { + "epoch": 0.5728848114169215, + "grad_norm": 3.9462735652923584, + "learning_rate": 9.51375978556859e-05, + "loss": 0.2741, + "step": 562 + }, + { + "epoch": 0.5739041794087666, + "grad_norm": 6.374652862548828, + "learning_rate": 9.505622644198885e-05, + "loss": 0.6232, + "step": 563 + }, + { + "epoch": 0.5749235474006116, + "grad_norm": 3.525486707687378, + "learning_rate": 9.497421519155303e-05, + "loss": 0.2083, + "step": 564 + }, + { + "epoch": 0.5759429153924567, + "grad_norm": 9.60029125213623, + "learning_rate": 9.489156526900795e-05, + "loss": 1.0605, + "step": 565 + }, + { + "epoch": 0.5769622833843018, + "grad_norm": 12.22358226776123, + "learning_rate": 9.480827784805278e-05, + "loss": 1.2086, + "step": 566 + }, + { + "epoch": 0.5779816513761468, + "grad_norm": 4.388841152191162, + "learning_rate": 9.472435411143978e-05, + "loss": 0.2217, + "step": 567 + }, + { + "epoch": 0.5790010193679919, + "grad_norm": 5.581283092498779, + "learning_rate": 9.463979525095738e-05, + "loss": 0.4215, + "step": 568 + }, + { + "epoch": 0.5800203873598369, + "grad_norm": 7.996876239776611, + "learning_rate": 9.455460246741331e-05, + "loss": 0.663, + "step": 569 + }, + { + "epoch": 0.581039755351682, + "grad_norm": 9.21956729888916, + "learning_rate": 9.446877697061757e-05, + "loss": 0.653, + "step": 570 + }, + { + "epoch": 0.582059123343527, + "grad_norm": 8.46827220916748, + "learning_rate": 9.43823199793652e-05, + "loss": 0.6895, + "step": 571 + }, + { + "epoch": 0.583078491335372, + "grad_norm": 9.72203540802002, + "learning_rate": 9.429523272141903e-05, + "loss": 1.1101, + "step": 572 + }, + { + "epoch": 0.5840978593272171, + "grad_norm": 8.79525089263916, + "learning_rate": 9.420751643349219e-05, + "loss": 1.2991, + "step": 573 + }, + { + "epoch": 0.5851172273190621, + "grad_norm": 6.719937801361084, + "learning_rate": 9.411917236123059e-05, + "loss": 0.4072, + "step": 574 + }, + { + "epoch": 0.5861365953109072, + "grad_norm": 8.360040664672852, + "learning_rate": 9.403020175919517e-05, + "loss": 1.169, + "step": 575 + }, + { + "epoch": 0.5871559633027523, + "grad_norm": 5.402820587158203, + "learning_rate": 9.394060589084417e-05, + "loss": 0.3374, + "step": 576 + }, + { + "epoch": 0.5881753312945973, + "grad_norm": 9.037818908691406, + "learning_rate": 9.385038602851515e-05, + "loss": 0.6785, + "step": 577 + }, + { + "epoch": 0.5891946992864424, + "grad_norm": 9.151761054992676, + "learning_rate": 9.375954345340685e-05, + "loss": 1.2757, + "step": 578 + }, + { + "epoch": 0.5902140672782875, + "grad_norm": 5.834461212158203, + "learning_rate": 9.366807945556113e-05, + "loss": 0.5899, + "step": 579 + }, + { + "epoch": 0.5912334352701325, + "grad_norm": 5.722581386566162, + "learning_rate": 9.357599533384453e-05, + "loss": 0.3389, + "step": 580 + }, + { + "epoch": 0.5922528032619776, + "grad_norm": 10.132628440856934, + "learning_rate": 9.348329239592995e-05, + "loss": 1.631, + "step": 581 + }, + { + "epoch": 0.5932721712538226, + "grad_norm": 9.922087669372559, + "learning_rate": 9.338997195827792e-05, + "loss": 1.3975, + "step": 582 + }, + { + "epoch": 0.5942915392456677, + "grad_norm": 8.382550239562988, + "learning_rate": 9.329603534611806e-05, + "loss": 0.4654, + "step": 583 + }, + { + "epoch": 0.5953109072375128, + "grad_norm": 8.080007553100586, + "learning_rate": 9.32014838934301e-05, + "loss": 0.56, + "step": 584 + }, + { + "epoch": 0.5963302752293578, + "grad_norm": 5.616114616394043, + "learning_rate": 9.310631894292518e-05, + "loss": 0.2282, + "step": 585 + }, + { + "epoch": 0.5973496432212029, + "grad_norm": 10.813580513000488, + "learning_rate": 9.301054184602647e-05, + "loss": 1.0754, + "step": 586 + }, + { + "epoch": 0.5983690112130479, + "grad_norm": 8.062788963317871, + "learning_rate": 9.291415396285024e-05, + "loss": 0.4411, + "step": 587 + }, + { + "epoch": 0.599388379204893, + "grad_norm": 8.6395845413208, + "learning_rate": 9.281715666218643e-05, + "loss": 0.9243, + "step": 588 + }, + { + "epoch": 0.599388379204893, + "eval_Qnli-dev-1024_cosine_accuracy": 0.71875, + "eval_Qnli-dev-1024_cosine_accuracy_threshold": 0.8436912298202515, + "eval_Qnli-dev-1024_cosine_ap": 0.7587494204458187, + "eval_Qnli-dev-1024_cosine_f1": 0.6875, + "eval_Qnli-dev-1024_cosine_f1_threshold": 0.8028630018234253, + "eval_Qnli-dev-1024_cosine_mcc": 0.3803921568627451, + "eval_Qnli-dev-1024_cosine_precision": 0.6470588235294118, + "eval_Qnli-dev-1024_cosine_recall": 0.7333333333333333, + "eval_Qnli-dev_cosine_accuracy": 0.71875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8078321218490601, + "eval_Qnli-dev_cosine_ap": 0.7321739553695406, + "eval_Qnli-dev_cosine_f1": 0.7339449541284404, + "eval_Qnli-dev_cosine_f1_threshold": 0.6781572699546814, + "eval_Qnli-dev_cosine_mcc": 0.4428074427700477, + "eval_Qnli-dev_cosine_precision": 0.625, + "eval_Qnli-dev_cosine_recall": 0.8888888888888888, + "eval_allNLI--triplets-1024_cosine_accuracy": 0.90625, + "eval_allNLI-triplets_cosine_accuracy": 0.9375, + "eval_global_dataset_loss": 0.36118289828300476, + "eval_global_dataset_runtime": 104.3983, + "eval_global_dataset_samples_per_second": 7.692, + "eval_global_dataset_steps_per_second": 0.163, + "eval_sequential_score": 0.90625, + "eval_sts-test-1024_pearson_cosine": 0.8631921152381832, + "eval_sts-test-1024_spearman_cosine": 0.9009700758334896, + "eval_sts-test_pearson_cosine": 0.9009355736320144, + "eval_sts-test_spearman_cosine": 0.9171725695772274, + "step": 588 + }, + { + "epoch": 0.6004077471967381, + "grad_norm": 6.184821128845215, + "learning_rate": 9.271955132147916e-05, + "loss": 0.3572, + "step": 589 + }, + { + "epoch": 0.601427115188583, + "grad_norm": 8.318941116333008, + "learning_rate": 9.262133932680733e-05, + "loss": 0.6761, + "step": 590 + }, + { + "epoch": 0.6024464831804281, + "grad_norm": 7.54533052444458, + "learning_rate": 9.252252207286479e-05, + "loss": 0.5754, + "step": 591 + }, + { + "epoch": 0.6034658511722731, + "grad_norm": 4.341547012329102, + "learning_rate": 9.24231009629406e-05, + "loss": 0.3664, + "step": 592 + }, + { + "epoch": 0.6044852191641182, + "grad_norm": 7.616749286651611, + "learning_rate": 9.232307740889909e-05, + "loss": 0.5391, + "step": 593 + }, + { + "epoch": 0.6055045871559633, + "grad_norm": 4.843873977661133, + "learning_rate": 9.222245283115979e-05, + "loss": 0.518, + "step": 594 + }, + { + "epoch": 0.6065239551478083, + "grad_norm": 8.295080184936523, + "learning_rate": 9.21212286586773e-05, + "loss": 0.5263, + "step": 595 + }, + { + "epoch": 0.6075433231396534, + "grad_norm": 3.873260736465454, + "learning_rate": 9.201940632892096e-05, + "loss": 0.2995, + "step": 596 + }, + { + "epoch": 0.6085626911314985, + "grad_norm": 4.403683185577393, + "learning_rate": 9.191698728785448e-05, + "loss": 0.4181, + "step": 597 + }, + { + "epoch": 0.6095820591233435, + "grad_norm": 7.282264709472656, + "learning_rate": 9.181397298991532e-05, + "loss": 0.5087, + "step": 598 + }, + { + "epoch": 0.6106014271151886, + "grad_norm": 5.132986068725586, + "learning_rate": 9.171036489799416e-05, + "loss": 0.4344, + "step": 599 + }, + { + "epoch": 0.6116207951070336, + "grad_norm": 11.096871376037598, + "learning_rate": 9.160616448341403e-05, + "loss": 1.6529, + "step": 600 + }, + { + "epoch": 0.6126401630988787, + "grad_norm": 4.306335926055908, + "learning_rate": 9.150137322590944e-05, + "loss": 0.4079, + "step": 601 + }, + { + "epoch": 0.6136595310907238, + "grad_norm": 5.622674942016602, + "learning_rate": 9.139599261360537e-05, + "loss": 0.3123, + "step": 602 + }, + { + "epoch": 0.6146788990825688, + "grad_norm": 10.172139167785645, + "learning_rate": 9.129002414299617e-05, + "loss": 1.4398, + "step": 603 + }, + { + "epoch": 0.6156982670744139, + "grad_norm": 10.175543785095215, + "learning_rate": 9.118346931892423e-05, + "loss": 1.5553, + "step": 604 + }, + { + "epoch": 0.6167176350662589, + "grad_norm": 7.616044521331787, + "learning_rate": 9.10763296545587e-05, + "loss": 0.7958, + "step": 605 + }, + { + "epoch": 0.617737003058104, + "grad_norm": 5.390756607055664, + "learning_rate": 9.096860667137394e-05, + "loss": 0.3815, + "step": 606 + }, + { + "epoch": 0.6187563710499491, + "grad_norm": 6.750911235809326, + "learning_rate": 9.086030189912794e-05, + "loss": 0.787, + "step": 607 + }, + { + "epoch": 0.6197757390417941, + "grad_norm": 5.695408344268799, + "learning_rate": 9.075141687584057e-05, + "loss": 0.2352, + "step": 608 + }, + { + "epoch": 0.6207951070336392, + "grad_norm": 12.017024040222168, + "learning_rate": 9.06419531477718e-05, + "loss": 0.6469, + "step": 609 + }, + { + "epoch": 0.6218144750254841, + "grad_norm": 9.70870304107666, + "learning_rate": 9.053191226939965e-05, + "loss": 1.0997, + "step": 610 + }, + { + "epoch": 0.6228338430173292, + "grad_norm": 8.183333396911621, + "learning_rate": 9.042129580339822e-05, + "loss": 0.8762, + "step": 611 + }, + { + "epoch": 0.6238532110091743, + "grad_norm": 8.237792015075684, + "learning_rate": 9.031010532061538e-05, + "loss": 0.6259, + "step": 612 + }, + { + "epoch": 0.6248725790010193, + "grad_norm": 7.553733825683594, + "learning_rate": 9.019834240005058e-05, + "loss": 0.8133, + "step": 613 + }, + { + "epoch": 0.6258919469928644, + "grad_norm": 8.876506805419922, + "learning_rate": 9.008600862883235e-05, + "loss": 1.2119, + "step": 614 + }, + { + "epoch": 0.6269113149847095, + "grad_norm": 6.738461971282959, + "learning_rate": 8.997310560219578e-05, + "loss": 0.5068, + "step": 615 + }, + { + "epoch": 0.6279306829765545, + "grad_norm": 9.400090217590332, + "learning_rate": 8.985963492345991e-05, + "loss": 0.7723, + "step": 616 + }, + { + "epoch": 0.6289500509683996, + "grad_norm": 8.690120697021484, + "learning_rate": 8.974559820400486e-05, + "loss": 1.0791, + "step": 617 + }, + { + "epoch": 0.6299694189602446, + "grad_norm": 5.441365718841553, + "learning_rate": 8.963099706324904e-05, + "loss": 0.348, + "step": 618 + }, + { + "epoch": 0.6309887869520897, + "grad_norm": 8.373964309692383, + "learning_rate": 8.951583312862616e-05, + "loss": 0.6523, + "step": 619 + }, + { + "epoch": 0.6320081549439348, + "grad_norm": 8.361169815063477, + "learning_rate": 8.9400108035562e-05, + "loss": 0.6241, + "step": 620 + }, + { + "epoch": 0.6330275229357798, + "grad_norm": 8.095520973205566, + "learning_rate": 8.928382342745137e-05, + "loss": 0.5039, + "step": 621 + }, + { + "epoch": 0.6340468909276249, + "grad_norm": 9.879805564880371, + "learning_rate": 8.916698095563453e-05, + "loss": 1.0113, + "step": 622 + }, + { + "epoch": 0.6350662589194699, + "grad_norm": 11.630424499511719, + "learning_rate": 8.904958227937406e-05, + "loss": 1.0527, + "step": 623 + }, + { + "epoch": 0.636085626911315, + "grad_norm": 9.939377784729004, + "learning_rate": 8.893162906583094e-05, + "loss": 1.3893, + "step": 624 + }, + { + "epoch": 0.6371049949031601, + "grad_norm": 7.852113723754883, + "learning_rate": 8.881312299004117e-05, + "loss": 0.7191, + "step": 625 + }, + { + "epoch": 0.6381243628950051, + "grad_norm": 7.134123802185059, + "learning_rate": 8.86940657348918e-05, + "loss": 0.3591, + "step": 626 + }, + { + "epoch": 0.6391437308868502, + "grad_norm": 5.795046806335449, + "learning_rate": 8.857445899109715e-05, + "loss": 0.9856, + "step": 627 + }, + { + "epoch": 0.6401630988786952, + "grad_norm": 7.4533610343933105, + "learning_rate": 8.845430445717469e-05, + "loss": 0.7603, + "step": 628 + }, + { + "epoch": 0.6411824668705403, + "grad_norm": 9.926379203796387, + "learning_rate": 8.8333603839421e-05, + "loss": 1.1553, + "step": 629 + }, + { + "epoch": 0.6422018348623854, + "grad_norm": 7.032261371612549, + "learning_rate": 8.821235885188754e-05, + "loss": 0.5608, + "step": 630 + }, + { + "epoch": 0.6432212028542303, + "grad_norm": 6.283802509307861, + "learning_rate": 8.809057121635624e-05, + "loss": 0.4338, + "step": 631 + }, + { + "epoch": 0.6442405708460754, + "grad_norm": 2.8640384674072266, + "learning_rate": 8.796824266231511e-05, + "loss": 0.1376, + "step": 632 + }, + { + "epoch": 0.6452599388379205, + "grad_norm": 7.722833633422852, + "learning_rate": 8.784537492693368e-05, + "loss": 0.6539, + "step": 633 + }, + { + "epoch": 0.6462793068297655, + "grad_norm": 7.714670658111572, + "learning_rate": 8.772196975503828e-05, + "loss": 0.5017, + "step": 634 + }, + { + "epoch": 0.6472986748216106, + "grad_norm": 4.0773091316223145, + "learning_rate": 8.759802889908733e-05, + "loss": 0.1888, + "step": 635 + }, + { + "epoch": 0.6483180428134556, + "grad_norm": 12.99943733215332, + "learning_rate": 8.747355411914642e-05, + "loss": 1.6077, + "step": 636 + }, + { + "epoch": 0.6493374108053007, + "grad_norm": 10.86596393585205, + "learning_rate": 8.734854718286327e-05, + "loss": 0.9635, + "step": 637 + }, + { + "epoch": 0.6503567787971458, + "grad_norm": 9.243484497070312, + "learning_rate": 8.722300986544272e-05, + "loss": 0.9786, + "step": 638 + }, + { + "epoch": 0.6513761467889908, + "grad_norm": 10.92319393157959, + "learning_rate": 8.709694394962142e-05, + "loss": 0.6728, + "step": 639 + }, + { + "epoch": 0.6523955147808359, + "grad_norm": 11.628253936767578, + "learning_rate": 8.697035122564266e-05, + "loss": 0.8592, + "step": 640 + }, + { + "epoch": 0.6534148827726809, + "grad_norm": 5.602497100830078, + "learning_rate": 8.684323349123075e-05, + "loss": 0.3945, + "step": 641 + }, + { + "epoch": 0.654434250764526, + "grad_norm": 7.681665420532227, + "learning_rate": 8.671559255156567e-05, + "loss": 0.7486, + "step": 642 + }, + { + "epoch": 0.6554536187563711, + "grad_norm": 9.017338752746582, + "learning_rate": 8.658743021925733e-05, + "loss": 0.7793, + "step": 643 + }, + { + "epoch": 0.6564729867482161, + "grad_norm": 5.24987268447876, + "learning_rate": 8.645874831431982e-05, + "loss": 0.4401, + "step": 644 + }, + { + "epoch": 0.6574923547400612, + "grad_norm": 10.270877838134766, + "learning_rate": 8.632954866414567e-05, + "loss": 0.6189, + "step": 645 + }, + { + "epoch": 0.6585117227319062, + "grad_norm": 8.378297805786133, + "learning_rate": 8.619983310347982e-05, + "loss": 0.7339, + "step": 646 + }, + { + "epoch": 0.6595310907237513, + "grad_norm": 6.045844554901123, + "learning_rate": 8.606960347439355e-05, + "loss": 0.4089, + "step": 647 + }, + { + "epoch": 0.6605504587155964, + "grad_norm": 10.432483673095703, + "learning_rate": 8.593886162625835e-05, + "loss": 1.1412, + "step": 648 + }, + { + "epoch": 0.6615698267074414, + "grad_norm": 5.939512729644775, + "learning_rate": 8.580760941571967e-05, + "loss": 0.798, + "step": 649 + }, + { + "epoch": 0.6625891946992865, + "grad_norm": 12.093332290649414, + "learning_rate": 8.567584870667056e-05, + "loss": 1.0588, + "step": 650 + }, + { + "epoch": 0.6636085626911316, + "grad_norm": 8.624043464660645, + "learning_rate": 8.554358137022513e-05, + "loss": 0.9044, + "step": 651 + }, + { + "epoch": 0.6646279306829765, + "grad_norm": 7.735975742340088, + "learning_rate": 8.54108092846921e-05, + "loss": 0.4464, + "step": 652 + }, + { + "epoch": 0.6656472986748216, + "grad_norm": 3.8205575942993164, + "learning_rate": 8.527753433554797e-05, + "loss": 0.2756, + "step": 653 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 10.537273406982422, + "learning_rate": 8.51437584154104e-05, + "loss": 1.123, + "step": 654 + }, + { + "epoch": 0.6676860346585117, + "grad_norm": 6.052632808685303, + "learning_rate": 8.500948342401124e-05, + "loss": 0.5377, + "step": 655 + }, + { + "epoch": 0.6687054026503568, + "grad_norm": 7.787528991699219, + "learning_rate": 8.48747112681696e-05, + "loss": 0.5164, + "step": 656 + }, + { + "epoch": 0.6697247706422018, + "grad_norm": 10.115964889526367, + "learning_rate": 8.473944386176469e-05, + "loss": 0.7155, + "step": 657 + }, + { + "epoch": 0.6707441386340469, + "grad_norm": 6.880122184753418, + "learning_rate": 8.460368312570873e-05, + "loss": 0.4512, + "step": 658 + }, + { + "epoch": 0.6717635066258919, + "grad_norm": 8.106338500976562, + "learning_rate": 8.446743098791969e-05, + "loss": 0.6199, + "step": 659 + }, + { + "epoch": 0.672782874617737, + "grad_norm": 11.035154342651367, + "learning_rate": 8.433068938329376e-05, + "loss": 0.6673, + "step": 660 + }, + { + "epoch": 0.6738022426095821, + "grad_norm": 4.484703540802002, + "learning_rate": 8.419346025367809e-05, + "loss": 0.5934, + "step": 661 + }, + { + "epoch": 0.6748216106014271, + "grad_norm": 6.977105140686035, + "learning_rate": 8.4055745547843e-05, + "loss": 0.5034, + "step": 662 + }, + { + "epoch": 0.6758409785932722, + "grad_norm": 5.447470664978027, + "learning_rate": 8.391754722145449e-05, + "loss": 0.4161, + "step": 663 + }, + { + "epoch": 0.6768603465851172, + "grad_norm": 13.200489044189453, + "learning_rate": 8.37788672370463e-05, + "loss": 0.9848, + "step": 664 + }, + { + "epoch": 0.6778797145769623, + "grad_norm": 6.03376579284668, + "learning_rate": 8.36397075639922e-05, + "loss": 0.356, + "step": 665 + }, + { + "epoch": 0.6788990825688074, + "grad_norm": 6.075347900390625, + "learning_rate": 8.350007017847788e-05, + "loss": 0.3031, + "step": 666 + }, + { + "epoch": 0.6799184505606524, + "grad_norm": 5.790109157562256, + "learning_rate": 8.335995706347299e-05, + "loss": 0.254, + "step": 667 + }, + { + "epoch": 0.6809378185524975, + "grad_norm": 11.979147911071777, + "learning_rate": 8.321937020870296e-05, + "loss": 0.8646, + "step": 668 + }, + { + "epoch": 0.6819571865443425, + "grad_norm": 9.445723533630371, + "learning_rate": 8.30783116106207e-05, + "loss": 0.7303, + "step": 669 + }, + { + "epoch": 0.6829765545361876, + "grad_norm": 8.001054763793945, + "learning_rate": 8.293678327237827e-05, + "loss": 0.4105, + "step": 670 + }, + { + "epoch": 0.6839959225280327, + "grad_norm": 4.437264919281006, + "learning_rate": 8.279478720379845e-05, + "loss": 0.2874, + "step": 671 + }, + { + "epoch": 0.6850152905198776, + "grad_norm": 4.547714710235596, + "learning_rate": 8.265232542134622e-05, + "loss": 0.2112, + "step": 672 + }, + { + "epoch": 0.6860346585117227, + "grad_norm": 7.875749588012695, + "learning_rate": 8.250939994810003e-05, + "loss": 1.0919, + "step": 673 + }, + { + "epoch": 0.6870540265035678, + "grad_norm": 5.349310874938965, + "learning_rate": 8.236601281372319e-05, + "loss": 0.5927, + "step": 674 + }, + { + "epoch": 0.6880733944954128, + "grad_norm": 11.490046501159668, + "learning_rate": 8.222216605443496e-05, + "loss": 1.011, + "step": 675 + }, + { + "epoch": 0.6890927624872579, + "grad_norm": 7.11298942565918, + "learning_rate": 8.207786171298166e-05, + "loss": 0.5656, + "step": 676 + }, + { + "epoch": 0.6901121304791029, + "grad_norm": 10.48589038848877, + "learning_rate": 8.193310183860771e-05, + "loss": 0.7199, + "step": 677 + }, + { + "epoch": 0.691131498470948, + "grad_norm": 9.364179611206055, + "learning_rate": 8.178788848702643e-05, + "loss": 0.7506, + "step": 678 + }, + { + "epoch": 0.6921508664627931, + "grad_norm": 6.678390026092529, + "learning_rate": 8.164222372039092e-05, + "loss": 0.5386, + "step": 679 + }, + { + "epoch": 0.6931702344546381, + "grad_norm": 6.151979446411133, + "learning_rate": 8.149610960726479e-05, + "loss": 0.6678, + "step": 680 + }, + { + "epoch": 0.6941896024464832, + "grad_norm": 6.415065765380859, + "learning_rate": 8.134954822259271e-05, + "loss": 0.4834, + "step": 681 + }, + { + "epoch": 0.6952089704383282, + "grad_norm": 4.4640326499938965, + "learning_rate": 8.120254164767101e-05, + "loss": 0.2411, + "step": 682 + }, + { + "epoch": 0.6962283384301733, + "grad_norm": 6.626987457275391, + "learning_rate": 8.105509197011807e-05, + "loss": 0.5011, + "step": 683 + }, + { + "epoch": 0.6972477064220184, + "grad_norm": 7.628388404846191, + "learning_rate": 8.090720128384475e-05, + "loss": 0.6573, + "step": 684 + }, + { + "epoch": 0.6982670744138634, + "grad_norm": 3.4043076038360596, + "learning_rate": 8.075887168902459e-05, + "loss": 0.2798, + "step": 685 + }, + { + "epoch": 0.6992864424057085, + "grad_norm": 5.682481288909912, + "learning_rate": 8.061010529206398e-05, + "loss": 0.5887, + "step": 686 + }, + { + "epoch": 0.6992864424057085, + "eval_Qnli-dev-1024_cosine_accuracy": 0.7291666666666666, + "eval_Qnli-dev-1024_cosine_accuracy_threshold": 0.8446075320243835, + "eval_Qnli-dev-1024_cosine_ap": 0.7501532568375827, + "eval_Qnli-dev-1024_cosine_f1": 0.7207207207207208, + "eval_Qnli-dev-1024_cosine_f1_threshold": 0.756614089012146, + "eval_Qnli-dev-1024_cosine_mcc": 0.4081269865567241, + "eval_Qnli-dev-1024_cosine_precision": 0.6060606060606061, + "eval_Qnli-dev-1024_cosine_recall": 0.8888888888888888, + "eval_Qnli-dev_cosine_accuracy": 0.71875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7564685344696045, + "eval_Qnli-dev_cosine_ap": 0.731843650475666, + "eval_Qnli-dev_cosine_f1": 0.7378640776699029, + "eval_Qnli-dev_cosine_f1_threshold": 0.6987220048904419, + "eval_Qnli-dev_cosine_mcc": 0.46153029495329345, + "eval_Qnli-dev_cosine_precision": 0.6551724137931034, + "eval_Qnli-dev_cosine_recall": 0.8444444444444444, + "eval_allNLI--triplets-1024_cosine_accuracy": 0.9270833134651184, + "eval_allNLI-triplets_cosine_accuracy": 0.9583333134651184, + "eval_global_dataset_loss": 0.29955434799194336, + "eval_global_dataset_runtime": 104.3655, + "eval_global_dataset_samples_per_second": 7.694, + "eval_global_dataset_steps_per_second": 0.163, + "eval_sequential_score": 0.9270833134651184, + "eval_sts-test-1024_pearson_cosine": 0.8628190908797548, + "eval_sts-test-1024_spearman_cosine": 0.9062196010289961, + "eval_sts-test_pearson_cosine": 0.9012940791829644, + "eval_sts-test_spearman_cosine": 0.9179642241352577, + "step": 686 + }, + { + "epoch": 0.7003058103975535, + "grad_norm": 5.198816299438477, + "learning_rate": 8.046090420557231e-05, + "loss": 0.6328, + "step": 687 + }, + { + "epoch": 0.7013251783893986, + "grad_norm": 7.044593811035156, + "learning_rate": 8.031127054833189e-05, + "loss": 0.4322, + "step": 688 + }, + { + "epoch": 0.7023445463812437, + "grad_norm": 13.891091346740723, + "learning_rate": 8.016120644526797e-05, + "loss": 1.1473, + "step": 689 + }, + { + "epoch": 0.7033639143730887, + "grad_norm": 9.329078674316406, + "learning_rate": 8.001071402741842e-05, + "loss": 1.0135, + "step": 690 + }, + { + "epoch": 0.7043832823649337, + "grad_norm": 4.1149210929870605, + "learning_rate": 7.985979543190358e-05, + "loss": 0.2655, + "step": 691 + }, + { + "epoch": 0.7054026503567788, + "grad_norm": 7.722234725952148, + "learning_rate": 7.970845280189586e-05, + "loss": 0.5053, + "step": 692 + }, + { + "epoch": 0.7064220183486238, + "grad_norm": 6.9180216789245605, + "learning_rate": 7.955668828658937e-05, + "loss": 0.8647, + "step": 693 + }, + { + "epoch": 0.7074413863404689, + "grad_norm": 5.709589004516602, + "learning_rate": 7.940450404116928e-05, + "loss": 0.4423, + "step": 694 + }, + { + "epoch": 0.7084607543323139, + "grad_norm": 4.812499523162842, + "learning_rate": 7.925190222678133e-05, + "loss": 0.3673, + "step": 695 + }, + { + "epoch": 0.709480122324159, + "grad_norm": 11.944628715515137, + "learning_rate": 7.909888501050109e-05, + "loss": 1.1714, + "step": 696 + }, + { + "epoch": 0.7104994903160041, + "grad_norm": 7.61957311630249, + "learning_rate": 7.894545456530316e-05, + "loss": 0.8142, + "step": 697 + }, + { + "epoch": 0.7115188583078491, + "grad_norm": 9.580735206604004, + "learning_rate": 7.879161307003038e-05, + "loss": 0.8027, + "step": 698 + }, + { + "epoch": 0.7125382262996942, + "grad_norm": 7.831961154937744, + "learning_rate": 7.863736270936284e-05, + "loss": 0.4514, + "step": 699 + }, + { + "epoch": 0.7135575942915392, + "grad_norm": 9.805893898010254, + "learning_rate": 7.848270567378686e-05, + "loss": 0.8798, + "step": 700 + }, + { + "epoch": 0.7145769622833843, + "grad_norm": 8.573545455932617, + "learning_rate": 7.832764415956389e-05, + "loss": 0.7718, + "step": 701 + }, + { + "epoch": 0.7155963302752294, + "grad_norm": 6.185779571533203, + "learning_rate": 7.817218036869932e-05, + "loss": 0.4094, + "step": 702 + }, + { + "epoch": 0.7166156982670744, + "grad_norm": 9.415246963500977, + "learning_rate": 7.80163165089112e-05, + "loss": 0.5358, + "step": 703 + }, + { + "epoch": 0.7176350662589195, + "grad_norm": 5.7925543785095215, + "learning_rate": 7.78600547935989e-05, + "loss": 0.5728, + "step": 704 + }, + { + "epoch": 0.7186544342507645, + "grad_norm": 8.365612983703613, + "learning_rate": 7.770339744181175e-05, + "loss": 0.4349, + "step": 705 + }, + { + "epoch": 0.7196738022426096, + "grad_norm": 11.040353775024414, + "learning_rate": 7.754634667821734e-05, + "loss": 1.0107, + "step": 706 + }, + { + "epoch": 0.7206931702344547, + "grad_norm": 10.400522232055664, + "learning_rate": 7.73889047330701e-05, + "loss": 1.3393, + "step": 707 + }, + { + "epoch": 0.7217125382262997, + "grad_norm": 6.314993381500244, + "learning_rate": 7.723107384217958e-05, + "loss": 0.5175, + "step": 708 + }, + { + "epoch": 0.7227319062181448, + "grad_norm": 7.7337541580200195, + "learning_rate": 7.70728562468787e-05, + "loss": 0.3906, + "step": 709 + }, + { + "epoch": 0.7237512742099899, + "grad_norm": 8.559732437133789, + "learning_rate": 7.691425419399183e-05, + "loss": 0.726, + "step": 710 + }, + { + "epoch": 0.7247706422018348, + "grad_norm": 5.824985504150391, + "learning_rate": 7.675526993580306e-05, + "loss": 0.4299, + "step": 711 + }, + { + "epoch": 0.72579001019368, + "grad_norm": 9.804418563842773, + "learning_rate": 7.659590573002407e-05, + "loss": 0.7486, + "step": 712 + }, + { + "epoch": 0.7268093781855249, + "grad_norm": 5.5835957527160645, + "learning_rate": 7.643616383976214e-05, + "loss": 0.3316, + "step": 713 + }, + { + "epoch": 0.72782874617737, + "grad_norm": 8.719099044799805, + "learning_rate": 7.627604653348796e-05, + "loss": 0.5444, + "step": 714 + }, + { + "epoch": 0.7288481141692151, + "grad_norm": 7.16873025894165, + "learning_rate": 7.611555608500351e-05, + "loss": 0.4717, + "step": 715 + }, + { + "epoch": 0.7298674821610601, + "grad_norm": 8.529095649719238, + "learning_rate": 7.595469477340965e-05, + "loss": 0.5413, + "step": 716 + }, + { + "epoch": 0.7308868501529052, + "grad_norm": 4.7856245040893555, + "learning_rate": 7.579346488307379e-05, + "loss": 0.2207, + "step": 717 + }, + { + "epoch": 0.7319062181447502, + "grad_norm": 8.381448745727539, + "learning_rate": 7.563186870359758e-05, + "loss": 0.8042, + "step": 718 + }, + { + "epoch": 0.7329255861365953, + "grad_norm": 6.099252700805664, + "learning_rate": 7.546990852978415e-05, + "loss": 0.3666, + "step": 719 + }, + { + "epoch": 0.7339449541284404, + "grad_norm": 6.979067802429199, + "learning_rate": 7.530758666160577e-05, + "loss": 0.4511, + "step": 720 + }, + { + "epoch": 0.7349643221202854, + "grad_norm": 8.355476379394531, + "learning_rate": 7.514490540417103e-05, + "loss": 0.782, + "step": 721 + }, + { + "epoch": 0.7359836901121305, + "grad_norm": 3.4271693229675293, + "learning_rate": 7.498186706769213e-05, + "loss": 0.2947, + "step": 722 + }, + { + "epoch": 0.7370030581039755, + "grad_norm": 10.753888130187988, + "learning_rate": 7.481847396745215e-05, + "loss": 1.6206, + "step": 723 + }, + { + "epoch": 0.7380224260958206, + "grad_norm": 10.323583602905273, + "learning_rate": 7.465472842377206e-05, + "loss": 0.6216, + "step": 724 + }, + { + "epoch": 0.7390417940876657, + "grad_norm": 5.268289566040039, + "learning_rate": 7.449063276197789e-05, + "loss": 0.2819, + "step": 725 + }, + { + "epoch": 0.7400611620795107, + "grad_norm": 8.948394775390625, + "learning_rate": 7.432618931236759e-05, + "loss": 0.4797, + "step": 726 + }, + { + "epoch": 0.7410805300713558, + "grad_norm": 4.78109884262085, + "learning_rate": 7.416140041017802e-05, + "loss": 0.3875, + "step": 727 + }, + { + "epoch": 0.7420998980632009, + "grad_norm": 7.643434047698975, + "learning_rate": 7.399626839555176e-05, + "loss": 0.6888, + "step": 728 + }, + { + "epoch": 0.7431192660550459, + "grad_norm": 4.128391742706299, + "learning_rate": 7.383079561350386e-05, + "loss": 0.3023, + "step": 729 + }, + { + "epoch": 0.744138634046891, + "grad_norm": 8.254578590393066, + "learning_rate": 7.36649844138886e-05, + "loss": 0.6654, + "step": 730 + }, + { + "epoch": 0.745158002038736, + "grad_norm": 10.747797966003418, + "learning_rate": 7.3498837151366e-05, + "loss": 0.6517, + "step": 731 + }, + { + "epoch": 0.746177370030581, + "grad_norm": 6.274332046508789, + "learning_rate": 7.333235618536856e-05, + "loss": 0.4537, + "step": 732 + }, + { + "epoch": 0.7471967380224261, + "grad_norm": 8.256685256958008, + "learning_rate": 7.316554388006756e-05, + "loss": 0.7224, + "step": 733 + }, + { + "epoch": 0.7482161060142711, + "grad_norm": 7.657110214233398, + "learning_rate": 7.299840260433965e-05, + "loss": 0.4447, + "step": 734 + }, + { + "epoch": 0.7492354740061162, + "grad_norm": 6.170997142791748, + "learning_rate": 7.283093473173307e-05, + "loss": 0.4127, + "step": 735 + }, + { + "epoch": 0.7502548419979612, + "grad_norm": 5.84876823425293, + "learning_rate": 7.26631426404341e-05, + "loss": 0.3297, + "step": 736 + }, + { + "epoch": 0.7512742099898063, + "grad_norm": 5.986436367034912, + "learning_rate": 7.249502871323314e-05, + "loss": 0.3664, + "step": 737 + }, + { + "epoch": 0.7522935779816514, + "grad_norm": 9.613632202148438, + "learning_rate": 7.232659533749092e-05, + "loss": 0.7934, + "step": 738 + }, + { + "epoch": 0.7533129459734964, + "grad_norm": 5.5741286277771, + "learning_rate": 7.215784490510468e-05, + "loss": 0.4214, + "step": 739 + }, + { + "epoch": 0.7543323139653415, + "grad_norm": 8.343430519104004, + "learning_rate": 7.198877981247406e-05, + "loss": 0.6174, + "step": 740 + }, + { + "epoch": 0.7553516819571865, + "grad_norm": 11.505045890808105, + "learning_rate": 7.18194024604672e-05, + "loss": 0.7011, + "step": 741 + }, + { + "epoch": 0.7563710499490316, + "grad_norm": 9.192388534545898, + "learning_rate": 7.164971525438657e-05, + "loss": 0.6472, + "step": 742 + }, + { + "epoch": 0.7573904179408767, + "grad_norm": 10.685009002685547, + "learning_rate": 7.147972060393478e-05, + "loss": 0.9555, + "step": 743 + }, + { + "epoch": 0.7584097859327217, + "grad_norm": 9.81982421875, + "learning_rate": 7.130942092318051e-05, + "loss": 1.1771, + "step": 744 + }, + { + "epoch": 0.7594291539245668, + "grad_norm": 7.654698848724365, + "learning_rate": 7.113881863052407e-05, + "loss": 0.6876, + "step": 745 + }, + { + "epoch": 0.7604485219164119, + "grad_norm": 10.608144760131836, + "learning_rate": 7.096791614866309e-05, + "loss": 0.6737, + "step": 746 + }, + { + "epoch": 0.7614678899082569, + "grad_norm": 8.949767112731934, + "learning_rate": 7.079671590455821e-05, + "loss": 0.9648, + "step": 747 + }, + { + "epoch": 0.762487257900102, + "grad_norm": 5.873875141143799, + "learning_rate": 7.06252203293985e-05, + "loss": 0.3267, + "step": 748 + }, + { + "epoch": 0.763506625891947, + "grad_norm": 3.814371347427368, + "learning_rate": 7.045343185856701e-05, + "loss": 0.2244, + "step": 749 + }, + { + "epoch": 0.764525993883792, + "grad_norm": 5.834865570068359, + "learning_rate": 7.028135293160611e-05, + "loss": 0.305, + "step": 750 + }, + { + "epoch": 0.7655453618756372, + "grad_norm": 8.765941619873047, + "learning_rate": 7.010898599218296e-05, + "loss": 0.5588, + "step": 751 + }, + { + "epoch": 0.7665647298674821, + "grad_norm": 8.091228485107422, + "learning_rate": 6.99363334880547e-05, + "loss": 1.0974, + "step": 752 + }, + { + "epoch": 0.7675840978593272, + "grad_norm": 7.041286468505859, + "learning_rate": 6.976339787103373e-05, + "loss": 0.603, + "step": 753 + }, + { + "epoch": 0.7686034658511722, + "grad_norm": 6.676450729370117, + "learning_rate": 6.959018159695293e-05, + "loss": 0.6972, + "step": 754 + }, + { + "epoch": 0.7696228338430173, + "grad_norm": 9.935379981994629, + "learning_rate": 6.94166871256307e-05, + "loss": 0.958, + "step": 755 + }, + { + "epoch": 0.7706422018348624, + "grad_norm": 6.536661624908447, + "learning_rate": 6.92429169208361e-05, + "loss": 0.2937, + "step": 756 + }, + { + "epoch": 0.7716615698267074, + "grad_norm": 5.736427307128906, + "learning_rate": 6.906887345025385e-05, + "loss": 0.3384, + "step": 757 + }, + { + "epoch": 0.7726809378185525, + "grad_norm": 5.628017425537109, + "learning_rate": 6.88945591854493e-05, + "loss": 0.3321, + "step": 758 + }, + { + "epoch": 0.7737003058103975, + "grad_norm": 9.1480712890625, + "learning_rate": 6.87199766018332e-05, + "loss": 0.8029, + "step": 759 + }, + { + "epoch": 0.7747196738022426, + "grad_norm": 7.8731770515441895, + "learning_rate": 6.85451281786268e-05, + "loss": 0.7043, + "step": 760 + }, + { + "epoch": 0.7757390417940877, + "grad_norm": 13.733153343200684, + "learning_rate": 6.837001639882641e-05, + "loss": 1.6068, + "step": 761 + }, + { + "epoch": 0.7767584097859327, + "grad_norm": 9.02813720703125, + "learning_rate": 6.819464374916823e-05, + "loss": 1.1273, + "step": 762 + }, + { + "epoch": 0.7777777777777778, + "grad_norm": 8.211151123046875, + "learning_rate": 6.801901272009307e-05, + "loss": 0.5429, + "step": 763 + }, + { + "epoch": 0.7787971457696228, + "grad_norm": 5.243499755859375, + "learning_rate": 6.784312580571091e-05, + "loss": 0.2976, + "step": 764 + }, + { + "epoch": 0.7798165137614679, + "grad_norm": 11.219100952148438, + "learning_rate": 6.766698550376556e-05, + "loss": 0.9183, + "step": 765 + }, + { + "epoch": 0.780835881753313, + "grad_norm": 7.10944938659668, + "learning_rate": 6.749059431559913e-05, + "loss": 0.4734, + "step": 766 + }, + { + "epoch": 0.781855249745158, + "grad_norm": 7.810965061187744, + "learning_rate": 6.731395474611649e-05, + "loss": 0.5437, + "step": 767 + }, + { + "epoch": 0.7828746177370031, + "grad_norm": 6.063333034515381, + "learning_rate": 6.71370693037498e-05, + "loss": 0.3382, + "step": 768 + }, + { + "epoch": 0.7838939857288482, + "grad_norm": 5.784426689147949, + "learning_rate": 6.695994050042277e-05, + "loss": 0.3925, + "step": 769 + }, + { + "epoch": 0.7849133537206932, + "grad_norm": 7.640711784362793, + "learning_rate": 6.678257085151509e-05, + "loss": 0.4345, + "step": 770 + }, + { + "epoch": 0.7859327217125383, + "grad_norm": 9.467418670654297, + "learning_rate": 6.660496287582667e-05, + "loss": 0.9237, + "step": 771 + }, + { + "epoch": 0.7869520897043832, + "grad_norm": 4.449363708496094, + "learning_rate": 6.642711909554174e-05, + "loss": 0.3875, + "step": 772 + }, + { + "epoch": 0.7879714576962283, + "grad_norm": 7.483307838439941, + "learning_rate": 6.624904203619333e-05, + "loss": 0.533, + "step": 773 + }, + { + "epoch": 0.7889908256880734, + "grad_norm": 4.827091693878174, + "learning_rate": 6.607073422662711e-05, + "loss": 0.4211, + "step": 774 + }, + { + "epoch": 0.7900101936799184, + "grad_norm": 6.135465621948242, + "learning_rate": 6.589219819896565e-05, + "loss": 0.5421, + "step": 775 + }, + { + "epoch": 0.7910295616717635, + "grad_norm": 9.622929573059082, + "learning_rate": 6.571343648857242e-05, + "loss": 0.8904, + "step": 776 + }, + { + "epoch": 0.7920489296636085, + "grad_norm": 5.664134502410889, + "learning_rate": 6.553445163401571e-05, + "loss": 0.4604, + "step": 777 + }, + { + "epoch": 0.7930682976554536, + "grad_norm": 9.634468078613281, + "learning_rate": 6.535524617703273e-05, + "loss": 0.7431, + "step": 778 + }, + { + "epoch": 0.7940876656472987, + "grad_norm": 10.855483055114746, + "learning_rate": 6.517582266249336e-05, + "loss": 1.0159, + "step": 779 + }, + { + "epoch": 0.7951070336391437, + "grad_norm": 9.945262908935547, + "learning_rate": 6.499618363836417e-05, + "loss": 0.6554, + "step": 780 + }, + { + "epoch": 0.7961264016309888, + "grad_norm": 7.224388599395752, + "learning_rate": 6.481633165567207e-05, + "loss": 0.8539, + "step": 781 + }, + { + "epoch": 0.7971457696228338, + "grad_norm": 8.917383193969727, + "learning_rate": 6.463626926846817e-05, + "loss": 0.4543, + "step": 782 + }, + { + "epoch": 0.7981651376146789, + "grad_norm": 4.411260604858398, + "learning_rate": 6.445599903379154e-05, + "loss": 0.2281, + "step": 783 + }, + { + "epoch": 0.799184505606524, + "grad_norm": 8.85741138458252, + "learning_rate": 6.427552351163286e-05, + "loss": 1.0334, + "step": 784 + }, + { + "epoch": 0.799184505606524, + "eval_Qnli-dev-1024_cosine_accuracy": 0.7395833333333334, + "eval_Qnli-dev-1024_cosine_accuracy_threshold": 0.8376107215881348, + "eval_Qnli-dev-1024_cosine_ap": 0.7815698422458957, + "eval_Qnli-dev-1024_cosine_f1": 0.7222222222222222, + "eval_Qnli-dev-1024_cosine_f1_threshold": 0.7544304132461548, + "eval_Qnli-dev-1024_cosine_mcc": 0.41614558708189836, + "eval_Qnli-dev-1024_cosine_precision": 0.6190476190476191, + "eval_Qnli-dev-1024_cosine_recall": 0.8666666666666667, + "eval_Qnli-dev_cosine_accuracy": 0.7291666666666666, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7388297319412231, + "eval_Qnli-dev_cosine_ap": 0.7636341718424307, + "eval_Qnli-dev_cosine_f1": 0.7450980392156862, + "eval_Qnli-dev_cosine_f1_threshold": 0.695953369140625, + "eval_Qnli-dev_cosine_mcc": 0.4794765594627558, + "eval_Qnli-dev_cosine_precision": 0.6666666666666666, + "eval_Qnli-dev_cosine_recall": 0.8444444444444444, + "eval_allNLI--triplets-1024_cosine_accuracy": 0.9583333134651184, + "eval_allNLI-triplets_cosine_accuracy": 0.9583333134651184, + "eval_global_dataset_loss": 0.3045359253883362, + "eval_global_dataset_runtime": 103.772, + "eval_global_dataset_samples_per_second": 7.738, + "eval_global_dataset_steps_per_second": 0.164, + "eval_sequential_score": 0.9583333134651184, + "eval_sts-test-1024_pearson_cosine": 0.8706480103495355, + "eval_sts-test-1024_spearman_cosine": 0.9094148980677476, + "eval_sts-test_pearson_cosine": 0.9036838203711135, + "eval_sts-test_spearman_cosine": 0.9196077696084266, + "step": 784 + }, + { + "epoch": 0.800203873598369, + "grad_norm": 10.137900352478027, + "learning_rate": 6.409484526489805e-05, + "loss": 0.9697, + "step": 785 + }, + { + "epoch": 0.8012232415902141, + "grad_norm": 8.976780891418457, + "learning_rate": 6.391396685937186e-05, + "loss": 0.7048, + "step": 786 + }, + { + "epoch": 0.8022426095820592, + "grad_norm": 8.672534942626953, + "learning_rate": 6.373289086368151e-05, + "loss": 0.5263, + "step": 787 + }, + { + "epoch": 0.8032619775739042, + "grad_norm": 9.115574836730957, + "learning_rate": 6.355161984926019e-05, + "loss": 0.5056, + "step": 788 + }, + { + "epoch": 0.8042813455657493, + "grad_norm": 5.47214412689209, + "learning_rate": 6.337015639031044e-05, + "loss": 0.3826, + "step": 789 + }, + { + "epoch": 0.8053007135575942, + "grad_norm": 4.726554870605469, + "learning_rate": 6.318850306376777e-05, + "loss": 0.3029, + "step": 790 + }, + { + "epoch": 0.8063200815494393, + "grad_norm": 9.025796890258789, + "learning_rate": 6.300666244926387e-05, + "loss": 0.7712, + "step": 791 + }, + { + "epoch": 0.8073394495412844, + "grad_norm": 8.51115894317627, + "learning_rate": 6.282463712909018e-05, + "loss": 0.5587, + "step": 792 + }, + { + "epoch": 0.8083588175331294, + "grad_norm": 4.170646667480469, + "learning_rate": 6.264242968816106e-05, + "loss": 0.2386, + "step": 793 + }, + { + "epoch": 0.8093781855249745, + "grad_norm": 7.225284576416016, + "learning_rate": 6.246004271397713e-05, + "loss": 0.5662, + "step": 794 + }, + { + "epoch": 0.8103975535168195, + "grad_norm": 8.109657287597656, + "learning_rate": 6.227747879658859e-05, + "loss": 0.5322, + "step": 795 + }, + { + "epoch": 0.8114169215086646, + "grad_norm": 8.729584693908691, + "learning_rate": 6.20947405285583e-05, + "loss": 0.5122, + "step": 796 + }, + { + "epoch": 0.8124362895005097, + "grad_norm": 6.562040328979492, + "learning_rate": 6.191183050492515e-05, + "loss": 0.5094, + "step": 797 + }, + { + "epoch": 0.8134556574923547, + "grad_norm": 8.552765846252441, + "learning_rate": 6.172875132316703e-05, + "loss": 0.8412, + "step": 798 + }, + { + "epoch": 0.8144750254841998, + "grad_norm": 8.517980575561523, + "learning_rate": 6.154550558316405e-05, + "loss": 0.3771, + "step": 799 + }, + { + "epoch": 0.8154943934760448, + "grad_norm": 9.862586975097656, + "learning_rate": 6.136209588716155e-05, + "loss": 0.626, + "step": 800 + }, + { + "epoch": 0.8165137614678899, + "grad_norm": 11.597122192382812, + "learning_rate": 6.117852483973325e-05, + "loss": 0.8902, + "step": 801 + }, + { + "epoch": 0.817533129459735, + "grad_norm": 4.268974781036377, + "learning_rate": 6.0994795047744144e-05, + "loss": 0.2301, + "step": 802 + }, + { + "epoch": 0.81855249745158, + "grad_norm": 2.586038112640381, + "learning_rate": 6.081090912031358e-05, + "loss": 0.16, + "step": 803 + }, + { + "epoch": 0.8195718654434251, + "grad_norm": 6.814731121063232, + "learning_rate": 6.0626869668778085e-05, + "loss": 0.4375, + "step": 804 + }, + { + "epoch": 0.8205912334352702, + "grad_norm": 9.699979782104492, + "learning_rate": 6.044267930665446e-05, + "loss": 0.9554, + "step": 805 + }, + { + "epoch": 0.8216106014271152, + "grad_norm": 7.751320838928223, + "learning_rate": 6.025834064960247e-05, + "loss": 0.4906, + "step": 806 + }, + { + "epoch": 0.8226299694189603, + "grad_norm": 8.852093696594238, + "learning_rate": 6.007385631538787e-05, + "loss": 0.478, + "step": 807 + }, + { + "epoch": 0.8236493374108053, + "grad_norm": 5.510447025299072, + "learning_rate": 5.988922892384513e-05, + "loss": 0.6057, + "step": 808 + }, + { + "epoch": 0.8246687054026504, + "grad_norm": 6.745148658752441, + "learning_rate": 5.9704461096840204e-05, + "loss": 0.5003, + "step": 809 + }, + { + "epoch": 0.8256880733944955, + "grad_norm": 11.509452819824219, + "learning_rate": 5.9519555458233436e-05, + "loss": 1.0844, + "step": 810 + }, + { + "epoch": 0.8267074413863404, + "grad_norm": 9.71648120880127, + "learning_rate": 5.933451463384213e-05, + "loss": 1.0267, + "step": 811 + }, + { + "epoch": 0.8277268093781855, + "grad_norm": 9.810832023620605, + "learning_rate": 5.91493412514034e-05, + "loss": 0.5415, + "step": 812 + }, + { + "epoch": 0.8287461773700305, + "grad_norm": 5.600392818450928, + "learning_rate": 5.896403794053679e-05, + "loss": 0.3295, + "step": 813 + }, + { + "epoch": 0.8297655453618756, + "grad_norm": 7.511580944061279, + "learning_rate": 5.877860733270692e-05, + "loss": 0.5511, + "step": 814 + }, + { + "epoch": 0.8307849133537207, + "grad_norm": 5.374726295471191, + "learning_rate": 5.8593052061186125e-05, + "loss": 0.3234, + "step": 815 + }, + { + "epoch": 0.8318042813455657, + "grad_norm": 4.7778639793396, + "learning_rate": 5.8407374761017105e-05, + "loss": 0.2917, + "step": 816 + }, + { + "epoch": 0.8328236493374108, + "grad_norm": 4.155742645263672, + "learning_rate": 5.822157806897548e-05, + "loss": 0.3865, + "step": 817 + }, + { + "epoch": 0.8338430173292558, + "grad_norm": 5.087594032287598, + "learning_rate": 5.803566462353225e-05, + "loss": 0.2401, + "step": 818 + }, + { + "epoch": 0.8348623853211009, + "grad_norm": 3.707869529724121, + "learning_rate": 5.7849637064816496e-05, + "loss": 0.1582, + "step": 819 + }, + { + "epoch": 0.835881753312946, + "grad_norm": 7.63162899017334, + "learning_rate": 5.76634980345778e-05, + "loss": 0.5475, + "step": 820 + }, + { + "epoch": 0.836901121304791, + "grad_norm": 5.092942237854004, + "learning_rate": 5.747725017614869e-05, + "loss": 0.3291, + "step": 821 + }, + { + "epoch": 0.8379204892966361, + "grad_norm": 6.86021089553833, + "learning_rate": 5.72908961344072e-05, + "loss": 0.6867, + "step": 822 + }, + { + "epoch": 0.8389398572884812, + "grad_norm": 9.336700439453125, + "learning_rate": 5.710443855573919e-05, + "loss": 0.9519, + "step": 823 + }, + { + "epoch": 0.8399592252803262, + "grad_norm": 6.382976055145264, + "learning_rate": 5.6917880088000894e-05, + "loss": 0.4898, + "step": 824 + }, + { + "epoch": 0.8409785932721713, + "grad_norm": 8.171992301940918, + "learning_rate": 5.6731223380481257e-05, + "loss": 0.3361, + "step": 825 + }, + { + "epoch": 0.8419979612640163, + "grad_norm": 11.304964065551758, + "learning_rate": 5.6544471083864245e-05, + "loss": 1.0131, + "step": 826 + }, + { + "epoch": 0.8430173292558614, + "grad_norm": 7.883802890777588, + "learning_rate": 5.635762585019136e-05, + "loss": 0.4988, + "step": 827 + }, + { + "epoch": 0.8440366972477065, + "grad_norm": 5.304625988006592, + "learning_rate": 5.61706903328238e-05, + "loss": 0.2737, + "step": 828 + }, + { + "epoch": 0.8450560652395515, + "grad_norm": 8.170361518859863, + "learning_rate": 5.598366718640494e-05, + "loss": 0.5214, + "step": 829 + }, + { + "epoch": 0.8460754332313966, + "grad_norm": 7.193360805511475, + "learning_rate": 5.579655906682255e-05, + "loss": 0.5261, + "step": 830 + }, + { + "epoch": 0.8470948012232415, + "grad_norm": 5.908787250518799, + "learning_rate": 5.5609368631171035e-05, + "loss": 0.4337, + "step": 831 + }, + { + "epoch": 0.8481141692150866, + "grad_norm": 11.470138549804688, + "learning_rate": 5.5422098537713815e-05, + "loss": 1.0523, + "step": 832 + }, + { + "epoch": 0.8491335372069317, + "grad_norm": 5.7633514404296875, + "learning_rate": 5.52347514458455e-05, + "loss": 0.59, + "step": 833 + }, + { + "epoch": 0.8501529051987767, + "grad_norm": 9.171930313110352, + "learning_rate": 5.5047330016054154e-05, + "loss": 0.9984, + "step": 834 + }, + { + "epoch": 0.8511722731906218, + "grad_norm": 7.584822177886963, + "learning_rate": 5.48598369098835e-05, + "loss": 1.0533, + "step": 835 + }, + { + "epoch": 0.8521916411824668, + "grad_norm": 5.429177761077881, + "learning_rate": 5.4672274789895104e-05, + "loss": 0.266, + "step": 836 + }, + { + "epoch": 0.8532110091743119, + "grad_norm": 7.292309284210205, + "learning_rate": 5.4484646319630636e-05, + "loss": 0.3497, + "step": 837 + }, + { + "epoch": 0.854230377166157, + "grad_norm": 7.126836776733398, + "learning_rate": 5.429695416357392e-05, + "loss": 0.5161, + "step": 838 + }, + { + "epoch": 0.855249745158002, + "grad_norm": 6.357126235961914, + "learning_rate": 5.410920098711323e-05, + "loss": 0.4256, + "step": 839 + }, + { + "epoch": 0.8562691131498471, + "grad_norm": 6.682480335235596, + "learning_rate": 5.392138945650339e-05, + "loss": 0.3334, + "step": 840 + }, + { + "epoch": 0.8572884811416922, + "grad_norm": 6.9180521965026855, + "learning_rate": 5.373352223882787e-05, + "loss": 0.5704, + "step": 841 + }, + { + "epoch": 0.8583078491335372, + "grad_norm": 6.871384620666504, + "learning_rate": 5.354560200196094e-05, + "loss": 0.3803, + "step": 842 + }, + { + "epoch": 0.8593272171253823, + "grad_norm": 9.186737060546875, + "learning_rate": 5.335763141452982e-05, + "loss": 0.7648, + "step": 843 + }, + { + "epoch": 0.8603465851172273, + "grad_norm": 8.700101852416992, + "learning_rate": 5.3169613145876714e-05, + "loss": 0.7548, + "step": 844 + }, + { + "epoch": 0.8613659531090724, + "grad_norm": 7.032200336456299, + "learning_rate": 5.2981549866020975e-05, + "loss": 0.7275, + "step": 845 + }, + { + "epoch": 0.8623853211009175, + "grad_norm": 13.48193359375, + "learning_rate": 5.2793444245621146e-05, + "loss": 1.1788, + "step": 846 + }, + { + "epoch": 0.8634046890927625, + "grad_norm": 9.682479858398438, + "learning_rate": 5.260529895593702e-05, + "loss": 0.7809, + "step": 847 + }, + { + "epoch": 0.8644240570846076, + "grad_norm": 8.730304718017578, + "learning_rate": 5.241711666879172e-05, + "loss": 0.6487, + "step": 848 + }, + { + "epoch": 0.8654434250764526, + "grad_norm": 6.570590972900391, + "learning_rate": 5.2228900056533836e-05, + "loss": 0.561, + "step": 849 + }, + { + "epoch": 0.8664627930682977, + "grad_norm": 8.695535659790039, + "learning_rate": 5.204065179199931e-05, + "loss": 0.5906, + "step": 850 + }, + { + "epoch": 0.8674821610601428, + "grad_norm": 5.353935241699219, + "learning_rate": 5.1852374548473614e-05, + "loss": 0.5192, + "step": 851 + }, + { + "epoch": 0.8685015290519877, + "grad_norm": 10.60522174835205, + "learning_rate": 5.1664070999653766e-05, + "loss": 0.8094, + "step": 852 + }, + { + "epoch": 0.8695208970438328, + "grad_norm": 3.7188539505004883, + "learning_rate": 5.147574381961032e-05, + "loss": 0.2399, + "step": 853 + }, + { + "epoch": 0.8705402650356778, + "grad_norm": 5.648993492126465, + "learning_rate": 5.128739568274944e-05, + "loss": 0.4103, + "step": 854 + }, + { + "epoch": 0.8715596330275229, + "grad_norm": 6.711026668548584, + "learning_rate": 5.109902926377482e-05, + "loss": 0.4969, + "step": 855 + }, + { + "epoch": 0.872579001019368, + "grad_norm": 5.686347961425781, + "learning_rate": 5.091064723764987e-05, + "loss": 0.37, + "step": 856 + }, + { + "epoch": 0.873598369011213, + "grad_norm": 4.857931613922119, + "learning_rate": 5.072225227955959e-05, + "loss": 0.4109, + "step": 857 + }, + { + "epoch": 0.8746177370030581, + "grad_norm": 8.75938606262207, + "learning_rate": 5.053384706487261e-05, + "loss": 0.525, + "step": 858 + }, + { + "epoch": 0.8756371049949032, + "grad_norm": 5.874378204345703, + "learning_rate": 5.034543426910324e-05, + "loss": 0.5958, + "step": 859 + }, + { + "epoch": 0.8766564729867482, + "grad_norm": 5.085257530212402, + "learning_rate": 5.0157016567873424e-05, + "loss": 0.4708, + "step": 860 + }, + { + "epoch": 0.8776758409785933, + "grad_norm": 7.9917707443237305, + "learning_rate": 4.996859663687479e-05, + "loss": 0.6881, + "step": 861 + }, + { + "epoch": 0.8786952089704383, + "grad_norm": 8.1506929397583, + "learning_rate": 4.9780177151830634e-05, + "loss": 0.5545, + "step": 862 + }, + { + "epoch": 0.8797145769622834, + "grad_norm": 9.375650405883789, + "learning_rate": 4.959176078845789e-05, + "loss": 0.645, + "step": 863 + }, + { + "epoch": 0.8807339449541285, + "grad_norm": 4.8143310546875, + "learning_rate": 4.9403350222429184e-05, + "loss": 0.4112, + "step": 864 + }, + { + "epoch": 0.8817533129459735, + "grad_norm": 7.862481594085693, + "learning_rate": 4.92149481293348e-05, + "loss": 0.4178, + "step": 865 + }, + { + "epoch": 0.8827726809378186, + "grad_norm": 5.252464771270752, + "learning_rate": 4.902655718464473e-05, + "loss": 0.2857, + "step": 866 + }, + { + "epoch": 0.8837920489296636, + "grad_norm": 6.06905460357666, + "learning_rate": 4.883818006367062e-05, + "loss": 0.3374, + "step": 867 + }, + { + "epoch": 0.8848114169215087, + "grad_norm": 6.810131072998047, + "learning_rate": 4.86498194415278e-05, + "loss": 0.5303, + "step": 868 + }, + { + "epoch": 0.8858307849133538, + "grad_norm": 7.676322937011719, + "learning_rate": 4.846147799309734e-05, + "loss": 0.7438, + "step": 869 + }, + { + "epoch": 0.8868501529051988, + "grad_norm": 11.570023536682129, + "learning_rate": 4.8273158392987986e-05, + "loss": 1.0872, + "step": 870 + }, + { + "epoch": 0.8878695208970439, + "grad_norm": 6.312341213226318, + "learning_rate": 4.8084863315498234e-05, + "loss": 0.4497, + "step": 871 + }, + { + "epoch": 0.8888888888888888, + "grad_norm": 7.389033794403076, + "learning_rate": 4.7896595434578356e-05, + "loss": 0.4171, + "step": 872 + }, + { + "epoch": 0.8899082568807339, + "grad_norm": 8.600625038146973, + "learning_rate": 4.770835742379239e-05, + "loss": 0.4417, + "step": 873 + }, + { + "epoch": 0.890927624872579, + "grad_norm": 7.350024223327637, + "learning_rate": 4.7520151956280227e-05, + "loss": 0.7023, + "step": 874 + }, + { + "epoch": 0.891946992864424, + "grad_norm": 12.617684364318848, + "learning_rate": 4.733198170471953e-05, + "loss": 1.0547, + "step": 875 + }, + { + "epoch": 0.8929663608562691, + "grad_norm": 5.219171524047852, + "learning_rate": 4.714384934128796e-05, + "loss": 0.3526, + "step": 876 + }, + { + "epoch": 0.8939857288481141, + "grad_norm": 10.923335075378418, + "learning_rate": 4.6955757537625104e-05, + "loss": 0.7315, + "step": 877 + }, + { + "epoch": 0.8950050968399592, + "grad_norm": 4.7785325050354, + "learning_rate": 4.6767708964794526e-05, + "loss": 0.4082, + "step": 878 + }, + { + "epoch": 0.8960244648318043, + "grad_norm": 7.037627696990967, + "learning_rate": 4.6579706293245944e-05, + "loss": 0.8155, + "step": 879 + }, + { + "epoch": 0.8970438328236493, + "grad_norm": 7.149205207824707, + "learning_rate": 4.6391752192777164e-05, + "loss": 0.5083, + "step": 880 + }, + { + "epoch": 0.8980632008154944, + "grad_norm": 5.331564426422119, + "learning_rate": 4.620384933249631e-05, + "loss": 0.655, + "step": 881 + }, + { + "epoch": 0.8990825688073395, + "grad_norm": 10.019486427307129, + "learning_rate": 4.6016000380783805e-05, + "loss": 0.7207, + "step": 882 + }, + { + "epoch": 0.8990825688073395, + "eval_Qnli-dev-1024_cosine_accuracy": 0.7708333333333334, + "eval_Qnli-dev-1024_cosine_accuracy_threshold": 0.8360349535942078, + "eval_Qnli-dev-1024_cosine_ap": 0.8011558872452826, + "eval_Qnli-dev-1024_cosine_f1": 0.7250000000000001, + "eval_Qnli-dev-1024_cosine_f1_threshold": 0.8360349535942078, + "eval_Qnli-dev-1024_cosine_mcc": 0.5461802806126049, + "eval_Qnli-dev-1024_cosine_precision": 0.8285714285714286, + "eval_Qnli-dev-1024_cosine_recall": 0.6444444444444445, + "eval_Qnli-dev_cosine_accuracy": 0.7291666666666666, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7521146535873413, + "eval_Qnli-dev_cosine_ap": 0.7712094779135136, + "eval_Qnli-dev_cosine_f1": 0.7500000000000001, + "eval_Qnli-dev_cosine_f1_threshold": 0.6768573522567749, + "eval_Qnli-dev_cosine_mcc": 0.48653004754089046, + "eval_Qnli-dev_cosine_precision": 0.6610169491525424, + "eval_Qnli-dev_cosine_recall": 0.8666666666666667, + "eval_allNLI--triplets-1024_cosine_accuracy": 0.9479166865348816, + "eval_allNLI-triplets_cosine_accuracy": 0.9583333134651184, + "eval_global_dataset_loss": 0.26387155055999756, + "eval_global_dataset_runtime": 103.9177, + "eval_global_dataset_samples_per_second": 7.727, + "eval_global_dataset_steps_per_second": 0.164, + "eval_sequential_score": 0.9479166865348816, + "eval_sts-test-1024_pearson_cosine": 0.8810824372715894, + "eval_sts-test-1024_spearman_cosine": 0.9117642789427417, + "eval_sts-test_pearson_cosine": 0.9044525796924666, + "eval_sts-test_spearman_cosine": 0.9182572042166309, + "step": 882 + }, + { + "epoch": 0.9001019367991845, + "grad_norm": 7.724752426147461, + "learning_rate": 4.582820800525455e-05, + "loss": 0.4898, + "step": 883 + }, + { + "epoch": 0.9011213047910296, + "grad_norm": 9.442131042480469, + "learning_rate": 4.564047487272001e-05, + "loss": 0.5506, + "step": 884 + }, + { + "epoch": 0.9021406727828746, + "grad_norm": 8.832263946533203, + "learning_rate": 4.5452803649150324e-05, + "loss": 0.6206, + "step": 885 + }, + { + "epoch": 0.9031600407747197, + "grad_norm": 12.987079620361328, + "learning_rate": 4.5265196999636535e-05, + "loss": 1.9669, + "step": 886 + }, + { + "epoch": 0.9041794087665648, + "grad_norm": 9.050309181213379, + "learning_rate": 4.5077657588352595e-05, + "loss": 0.8493, + "step": 887 + }, + { + "epoch": 0.9051987767584098, + "grad_norm": 9.642857551574707, + "learning_rate": 4.489018807851769e-05, + "loss": 0.9698, + "step": 888 + }, + { + "epoch": 0.9062181447502549, + "grad_norm": 7.444589614868164, + "learning_rate": 4.4702791132358314e-05, + "loss": 0.7322, + "step": 889 + }, + { + "epoch": 0.9072375127420998, + "grad_norm": 9.99152946472168, + "learning_rate": 4.451546941107046e-05, + "loss": 0.484, + "step": 890 + }, + { + "epoch": 0.908256880733945, + "grad_norm": 6.232360363006592, + "learning_rate": 4.432822557478194e-05, + "loss": 0.5604, + "step": 891 + }, + { + "epoch": 0.90927624872579, + "grad_norm": 3.1541106700897217, + "learning_rate": 4.414106228251446e-05, + "loss": 0.2633, + "step": 892 + }, + { + "epoch": 0.910295616717635, + "grad_norm": 5.661106109619141, + "learning_rate": 4.3953982192146006e-05, + "loss": 0.2417, + "step": 893 + }, + { + "epoch": 0.9113149847094801, + "grad_norm": 4.497067451477051, + "learning_rate": 4.3766987960372956e-05, + "loss": 0.4481, + "step": 894 + }, + { + "epoch": 0.9123343527013251, + "grad_norm": 8.505694389343262, + "learning_rate": 4.358008224267245e-05, + "loss": 0.7402, + "step": 895 + }, + { + "epoch": 0.9133537206931702, + "grad_norm": 5.820054054260254, + "learning_rate": 4.3393267693264686e-05, + "loss": 0.4897, + "step": 896 + }, + { + "epoch": 0.9143730886850153, + "grad_norm": 7.943095684051514, + "learning_rate": 4.320654696507511e-05, + "loss": 0.5863, + "step": 897 + }, + { + "epoch": 0.9153924566768603, + "grad_norm": 10.6437349319458, + "learning_rate": 4.301992270969692e-05, + "loss": 0.7101, + "step": 898 + }, + { + "epoch": 0.9164118246687054, + "grad_norm": 3.8055593967437744, + "learning_rate": 4.2833397577353284e-05, + "loss": 0.2404, + "step": 899 + }, + { + "epoch": 0.9174311926605505, + "grad_norm": 8.539854049682617, + "learning_rate": 4.26469742168597e-05, + "loss": 0.5594, + "step": 900 + }, + { + "epoch": 0.9184505606523955, + "grad_norm": 5.611748218536377, + "learning_rate": 4.2460655275586494e-05, + "loss": 0.4047, + "step": 901 + }, + { + "epoch": 0.9194699286442406, + "grad_norm": 4.898343086242676, + "learning_rate": 4.227444339942107e-05, + "loss": 0.4865, + "step": 902 + }, + { + "epoch": 0.9204892966360856, + "grad_norm": 8.28711986541748, + "learning_rate": 4.208834123273047e-05, + "loss": 0.3909, + "step": 903 + }, + { + "epoch": 0.9215086646279307, + "grad_norm": 6.98935604095459, + "learning_rate": 4.190235141832375e-05, + "loss": 0.2808, + "step": 904 + }, + { + "epoch": 0.9225280326197758, + "grad_norm": 9.016980171203613, + "learning_rate": 4.171647659741448e-05, + "loss": 0.7509, + "step": 905 + }, + { + "epoch": 0.9235474006116208, + "grad_norm": 5.859550476074219, + "learning_rate": 4.153071940958321e-05, + "loss": 0.325, + "step": 906 + }, + { + "epoch": 0.9245667686034659, + "grad_norm": 7.970040321350098, + "learning_rate": 4.134508249274002e-05, + "loss": 0.5335, + "step": 907 + }, + { + "epoch": 0.9255861365953109, + "grad_norm": 6.2324981689453125, + "learning_rate": 4.1159568483087e-05, + "loss": 0.6193, + "step": 908 + }, + { + "epoch": 0.926605504587156, + "grad_norm": 5.227268218994141, + "learning_rate": 4.0974180015080897e-05, + "loss": 0.2974, + "step": 909 + }, + { + "epoch": 0.9276248725790011, + "grad_norm": 9.293944358825684, + "learning_rate": 4.078891972139564e-05, + "loss": 0.6725, + "step": 910 + }, + { + "epoch": 0.928644240570846, + "grad_norm": 10.003561019897461, + "learning_rate": 4.060379023288495e-05, + "loss": 0.8828, + "step": 911 + }, + { + "epoch": 0.9296636085626911, + "grad_norm": 9.07729721069336, + "learning_rate": 4.0418794178545076e-05, + "loss": 0.8751, + "step": 912 + }, + { + "epoch": 0.9306829765545361, + "grad_norm": 7.200821876525879, + "learning_rate": 4.023393418547732e-05, + "loss": 0.7019, + "step": 913 + }, + { + "epoch": 0.9317023445463812, + "grad_norm": 10.154699325561523, + "learning_rate": 4.0049212878850793e-05, + "loss": 0.7131, + "step": 914 + }, + { + "epoch": 0.9327217125382263, + "grad_norm": 7.271543025970459, + "learning_rate": 3.98646328818652e-05, + "loss": 0.2849, + "step": 915 + }, + { + "epoch": 0.9337410805300713, + "grad_norm": 9.933566093444824, + "learning_rate": 3.96801968157135e-05, + "loss": 0.8097, + "step": 916 + }, + { + "epoch": 0.9347604485219164, + "grad_norm": 5.370792865753174, + "learning_rate": 3.949590729954467e-05, + "loss": 0.3447, + "step": 917 + }, + { + "epoch": 0.9357798165137615, + "grad_norm": 8.846680641174316, + "learning_rate": 3.931176695042664e-05, + "loss": 0.8601, + "step": 918 + }, + { + "epoch": 0.9367991845056065, + "grad_norm": 5.936051368713379, + "learning_rate": 3.912777838330893e-05, + "loss": 0.4467, + "step": 919 + }, + { + "epoch": 0.9378185524974516, + "grad_norm": 10.40077018737793, + "learning_rate": 3.8943944210985735e-05, + "loss": 0.8137, + "step": 920 + }, + { + "epoch": 0.9388379204892966, + "grad_norm": 7.319591999053955, + "learning_rate": 3.876026704405866e-05, + "loss": 0.4527, + "step": 921 + }, + { + "epoch": 0.9398572884811417, + "grad_norm": 8.947883605957031, + "learning_rate": 3.8576749490899686e-05, + "loss": 0.7656, + "step": 922 + }, + { + "epoch": 0.9408766564729868, + "grad_norm": 10.776662826538086, + "learning_rate": 3.839339415761416e-05, + "loss": 1.1218, + "step": 923 + }, + { + "epoch": 0.9418960244648318, + "grad_norm": 2.9248359203338623, + "learning_rate": 3.821020364800379e-05, + "loss": 0.188, + "step": 924 + }, + { + "epoch": 0.9429153924566769, + "grad_norm": 9.73752212524414, + "learning_rate": 3.8027180563529616e-05, + "loss": 0.8454, + "step": 925 + }, + { + "epoch": 0.9439347604485219, + "grad_norm": 6.643280506134033, + "learning_rate": 3.7844327503275136e-05, + "loss": 0.5368, + "step": 926 + }, + { + "epoch": 0.944954128440367, + "grad_norm": 9.299040794372559, + "learning_rate": 3.7661647063909294e-05, + "loss": 0.7602, + "step": 927 + }, + { + "epoch": 0.9459734964322121, + "grad_norm": 6.660792827606201, + "learning_rate": 3.747914183964974e-05, + "loss": 0.4733, + "step": 928 + }, + { + "epoch": 0.9469928644240571, + "grad_norm": 5.206737995147705, + "learning_rate": 3.729681442222587e-05, + "loss": 0.2305, + "step": 929 + }, + { + "epoch": 0.9480122324159022, + "grad_norm": 9.746971130371094, + "learning_rate": 3.711466740084211e-05, + "loss": 0.7775, + "step": 930 + }, + { + "epoch": 0.9490316004077471, + "grad_norm": 9.825338363647461, + "learning_rate": 3.6932703362141084e-05, + "loss": 0.8859, + "step": 931 + }, + { + "epoch": 0.9500509683995922, + "grad_norm": 7.335731506347656, + "learning_rate": 3.6750924890166914e-05, + "loss": 0.3918, + "step": 932 + }, + { + "epoch": 0.9510703363914373, + "grad_norm": 6.4724931716918945, + "learning_rate": 3.656933456632853e-05, + "loss": 0.3842, + "step": 933 + }, + { + "epoch": 0.9520897043832823, + "grad_norm": 4.886312484741211, + "learning_rate": 3.638793496936296e-05, + "loss": 0.3719, + "step": 934 + }, + { + "epoch": 0.9531090723751274, + "grad_norm": 8.522834777832031, + "learning_rate": 3.620672867529878e-05, + "loss": 0.8043, + "step": 935 + }, + { + "epoch": 0.9541284403669725, + "grad_norm": 9.507696151733398, + "learning_rate": 3.602571825741953e-05, + "loss": 0.8282, + "step": 936 + }, + { + "epoch": 0.9551478083588175, + "grad_norm": 4.895750045776367, + "learning_rate": 3.584490628622705e-05, + "loss": 0.4599, + "step": 937 + }, + { + "epoch": 0.9561671763506626, + "grad_norm": 7.197470664978027, + "learning_rate": 3.566429532940518e-05, + "loss": 0.649, + "step": 938 + }, + { + "epoch": 0.9571865443425076, + "grad_norm": 6.60915470123291, + "learning_rate": 3.548388795178307e-05, + "loss": 0.4325, + "step": 939 + }, + { + "epoch": 0.9582059123343527, + "grad_norm": 10.626359939575195, + "learning_rate": 3.5303686715298955e-05, + "loss": 1.3108, + "step": 940 + }, + { + "epoch": 0.9592252803261978, + "grad_norm": 6.316555023193359, + "learning_rate": 3.51236941789637e-05, + "loss": 0.3018, + "step": 941 + }, + { + "epoch": 0.9602446483180428, + "grad_norm": 7.12025785446167, + "learning_rate": 3.494391289882435e-05, + "loss": 0.6258, + "step": 942 + }, + { + "epoch": 0.9612640163098879, + "grad_norm": 10.008544921875, + "learning_rate": 3.476434542792805e-05, + "loss": 1.2266, + "step": 943 + }, + { + "epoch": 0.9622833843017329, + "grad_norm": 8.917716979980469, + "learning_rate": 3.4584994316285604e-05, + "loss": 0.6593, + "step": 944 + }, + { + "epoch": 0.963302752293578, + "grad_norm": 5.837446689605713, + "learning_rate": 3.4405862110835364e-05, + "loss": 0.3096, + "step": 945 + }, + { + "epoch": 0.9643221202854231, + "grad_norm": 4.312796115875244, + "learning_rate": 3.422695135540697e-05, + "loss": 0.3436, + "step": 946 + }, + { + "epoch": 0.9653414882772681, + "grad_norm": 4.772927284240723, + "learning_rate": 3.404826459068536e-05, + "loss": 0.2497, + "step": 947 + }, + { + "epoch": 0.9663608562691132, + "grad_norm": 3.3676137924194336, + "learning_rate": 3.386980435417457e-05, + "loss": 0.1653, + "step": 948 + }, + { + "epoch": 0.9673802242609582, + "grad_norm": 6.203863143920898, + "learning_rate": 3.369157318016176e-05, + "loss": 0.469, + "step": 949 + }, + { + "epoch": 0.9683995922528033, + "grad_norm": 7.628493309020996, + "learning_rate": 3.351357359968117e-05, + "loss": 0.4919, + "step": 950 + }, + { + "epoch": 0.9694189602446484, + "grad_norm": 7.940287113189697, + "learning_rate": 3.333580814047826e-05, + "loss": 0.4788, + "step": 951 + }, + { + "epoch": 0.9704383282364933, + "grad_norm": 6.046499729156494, + "learning_rate": 3.3158279326973766e-05, + "loss": 0.3041, + "step": 952 + }, + { + "epoch": 0.9714576962283384, + "grad_norm": 4.314492225646973, + "learning_rate": 3.298098968022782e-05, + "loss": 0.3138, + "step": 953 + }, + { + "epoch": 0.9724770642201835, + "grad_norm": 8.91407585144043, + "learning_rate": 3.2803941717904216e-05, + "loss": 0.7758, + "step": 954 + }, + { + "epoch": 0.9734964322120285, + "grad_norm": 11.913896560668945, + "learning_rate": 3.26271379542346e-05, + "loss": 0.6974, + "step": 955 + }, + { + "epoch": 0.9745158002038736, + "grad_norm": 4.831221580505371, + "learning_rate": 3.2450580899982795e-05, + "loss": 0.2964, + "step": 956 + }, + { + "epoch": 0.9755351681957186, + "grad_norm": 6.116502285003662, + "learning_rate": 3.2274273062409154e-05, + "loss": 0.3473, + "step": 957 + }, + { + "epoch": 0.9765545361875637, + "grad_norm": 11.75236988067627, + "learning_rate": 3.2098216945234946e-05, + "loss": 0.8905, + "step": 958 + }, + { + "epoch": 0.9775739041794088, + "grad_norm": 3.468975067138672, + "learning_rate": 3.192241504860675e-05, + "loss": 0.2521, + "step": 959 + }, + { + "epoch": 0.9785932721712538, + "grad_norm": 7.624709606170654, + "learning_rate": 3.1746869869061063e-05, + "loss": 0.4462, + "step": 960 + }, + { + "epoch": 0.9796126401630989, + "grad_norm": 9.019265174865723, + "learning_rate": 3.157158389948871e-05, + "loss": 0.7842, + "step": 961 + }, + { + "epoch": 0.9806320081549439, + "grad_norm": 4.77131986618042, + "learning_rate": 3.1396559629099574e-05, + "loss": 0.2973, + "step": 962 + }, + { + "epoch": 0.981651376146789, + "grad_norm": 8.40596866607666, + "learning_rate": 3.122179954338716e-05, + "loss": 0.6026, + "step": 963 + }, + { + "epoch": 0.9826707441386341, + "grad_norm": 6.705322265625, + "learning_rate": 3.1047306124093335e-05, + "loss": 0.4026, + "step": 964 + }, + { + "epoch": 0.9836901121304791, + "grad_norm": 10.35732364654541, + "learning_rate": 3.087308184917308e-05, + "loss": 0.9181, + "step": 965 + }, + { + "epoch": 0.9847094801223242, + "grad_norm": 6.806704998016357, + "learning_rate": 3.069912919275926e-05, + "loss": 0.473, + "step": 966 + }, + { + "epoch": 0.9857288481141692, + "grad_norm": 10.28345012664795, + "learning_rate": 3.0525450625127575e-05, + "loss": 0.7152, + "step": 967 + }, + { + "epoch": 0.9867482161060143, + "grad_norm": 11.785171508789062, + "learning_rate": 3.0352048612661416e-05, + "loss": 0.9519, + "step": 968 + }, + { + "epoch": 0.9877675840978594, + "grad_norm": 8.55274772644043, + "learning_rate": 3.017892561781682e-05, + "loss": 0.5322, + "step": 969 + }, + { + "epoch": 0.9887869520897044, + "grad_norm": 8.597644805908203, + "learning_rate": 3.0006084099087595e-05, + "loss": 0.8257, + "step": 970 + }, + { + "epoch": 0.9898063200815495, + "grad_norm": 6.743808746337891, + "learning_rate": 2.983352651097031e-05, + "loss": 0.5648, + "step": 971 + }, + { + "epoch": 0.9908256880733946, + "grad_norm": 10.981080055236816, + "learning_rate": 2.9661255303929486e-05, + "loss": 0.909, + "step": 972 + } + ], + "logging_steps": 1, + "max_steps": 2943, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 972, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}