| { | |
| "best_metric": 0.9359054565429688, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-3_120-samples_config-1_full_auto/checkpoint-77", | |
| "epoch": 14.0, | |
| "eval_steps": 500, | |
| "global_step": 154, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 0.44340184330940247, | |
| "learning_rate": 1.818181818181818e-06, | |
| "loss": 1.2873, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.5295612215995789, | |
| "learning_rate": 3.636363636363636e-06, | |
| "loss": 1.5104, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.41922619938850403, | |
| "learning_rate": 7.272727272727272e-06, | |
| "loss": 1.3424, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.45832565426826477, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 1.3772, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.4448011815547943, | |
| "learning_rate": 1.4545454545454545e-05, | |
| "loss": 1.43, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.40825021266937256, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 1.3429, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.3558740615844727, | |
| "eval_runtime": 7.6804, | |
| "eval_samples_per_second": 3.125, | |
| "eval_steps_per_second": 3.125, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.4454409182071686, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 1.4164, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.3652619421482086, | |
| "learning_rate": 2.5454545454545454e-05, | |
| "loss": 1.3038, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.4184987246990204, | |
| "learning_rate": 2.909090909090909e-05, | |
| "loss": 1.3385, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.35863184928894043, | |
| "learning_rate": 3.272727272727273e-05, | |
| "loss": 1.3393, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.29392459988594055, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 1.2529, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.26279276609420776, | |
| "learning_rate": 4e-05, | |
| "loss": 1.1584, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.2111910581588745, | |
| "eval_runtime": 7.4943, | |
| "eval_samples_per_second": 3.202, | |
| "eval_steps_per_second": 3.202, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.2251596301794052, | |
| "learning_rate": 4.3636363636363636e-05, | |
| "loss": 1.1951, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.27620360255241394, | |
| "learning_rate": 4.7272727272727275e-05, | |
| "loss": 1.1708, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.25171273946762085, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 1.1228, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.2604583203792572, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 1.1486, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.2893660366535187, | |
| "learning_rate": 5.818181818181818e-05, | |
| "loss": 1.0941, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.0876060724258423, | |
| "eval_runtime": 7.5096, | |
| "eval_samples_per_second": 3.196, | |
| "eval_steps_per_second": 3.196, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.26696836948394775, | |
| "learning_rate": 6.181818181818182e-05, | |
| "loss": 1.0925, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.2858625650405884, | |
| "learning_rate": 6.545454545454546e-05, | |
| "loss": 1.0371, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.34232887625694275, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 1.0255, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.2392912358045578, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.9489, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.4012575149536133, | |
| "learning_rate": 7.636363636363637e-05, | |
| "loss": 0.9203, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.22504228353500366, | |
| "learning_rate": 8e-05, | |
| "loss": 1.0134, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.9891138076782227, | |
| "eval_runtime": 7.4936, | |
| "eval_samples_per_second": 3.203, | |
| "eval_steps_per_second": 3.203, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.25609034299850464, | |
| "learning_rate": 8.363636363636364e-05, | |
| "loss": 0.9138, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.2524278461933136, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.9498, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.20994818210601807, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.8959, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.24883906543254852, | |
| "learning_rate": 9.454545454545455e-05, | |
| "loss": 0.9049, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 0.18293549120426178, | |
| "learning_rate": 9.818181818181818e-05, | |
| "loss": 0.9378, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.9540689587593079, | |
| "eval_runtime": 7.5014, | |
| "eval_samples_per_second": 3.199, | |
| "eval_steps_per_second": 3.199, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.28788265585899353, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 0.8829, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.2040722519159317, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.9036, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.2116885632276535, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.8431, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 0.21737386286258698, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 0.8668, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.21680592000484467, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.8547, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.31425392627716064, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.8949, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.9391613602638245, | |
| "eval_runtime": 7.508, | |
| "eval_samples_per_second": 3.197, | |
| "eval_steps_per_second": 3.197, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.25037598609924316, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.8083, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.22885167598724365, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.8632, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.25459662079811096, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.7854, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.25617337226867676, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.8213, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.2213735282421112, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 0.8546, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.9359054565429688, | |
| "eval_runtime": 7.4985, | |
| "eval_samples_per_second": 3.201, | |
| "eval_steps_per_second": 3.201, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.090909090909091, | |
| "grad_norm": 0.22576187551021576, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.8968, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.2886678874492645, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.7526, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.454545454545454, | |
| "grad_norm": 0.27642518281936646, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.7545, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.48505523800849915, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.7889, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 7.818181818181818, | |
| "grad_norm": 0.3221338987350464, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.7757, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.2804253399372101, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.822, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.9410832524299622, | |
| "eval_runtime": 7.4993, | |
| "eval_samples_per_second": 3.2, | |
| "eval_steps_per_second": 3.2, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 0.3164221942424774, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.7738, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.36224448680877686, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.7169, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 8.545454545454545, | |
| "grad_norm": 0.35652151703834534, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.7579, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.4091964066028595, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.6862, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "grad_norm": 0.41904446482658386, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.7437, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.9579834342002869, | |
| "eval_runtime": 7.5118, | |
| "eval_samples_per_second": 3.195, | |
| "eval_steps_per_second": 3.195, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.4185417592525482, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.7455, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.272727272727273, | |
| "grad_norm": 0.4688882529735565, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.6541, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.6384677290916443, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 0.6505, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 9.636363636363637, | |
| "grad_norm": 0.5532212257385254, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.7368, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.5052789449691772, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.6758, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.674145519733429, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.6443, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.0089081525802612, | |
| "eval_runtime": 7.4975, | |
| "eval_samples_per_second": 3.201, | |
| "eval_steps_per_second": 3.201, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.5758641362190247, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.5998, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 10.363636363636363, | |
| "grad_norm": 0.6187158823013306, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.5899, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.6771482825279236, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.5955, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 10.727272727272727, | |
| "grad_norm": 0.6831199526786804, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.5858, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.6889560222625732, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.6397, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 1.0663317441940308, | |
| "eval_runtime": 7.4957, | |
| "eval_samples_per_second": 3.202, | |
| "eval_steps_per_second": 3.202, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 11.090909090909092, | |
| "grad_norm": 0.8115834593772888, | |
| "learning_rate": 9.554728301876526e-05, | |
| "loss": 0.5702, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.9677160382270813, | |
| "learning_rate": 9.528180468815155e-05, | |
| "loss": 0.5163, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 11.454545454545455, | |
| "grad_norm": 0.838153600692749, | |
| "learning_rate": 9.50090306530454e-05, | |
| "loss": 0.5314, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.7266411781311035, | |
| "learning_rate": 9.472900486219769e-05, | |
| "loss": 0.5657, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 11.818181818181818, | |
| "grad_norm": 0.9129046201705933, | |
| "learning_rate": 9.444177243274618e-05, | |
| "loss": 0.5353, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.9295918345451355, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.5092, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.141348958015442, | |
| "eval_runtime": 7.4975, | |
| "eval_samples_per_second": 3.201, | |
| "eval_steps_per_second": 3.201, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.181818181818182, | |
| "grad_norm": 0.8798962235450745, | |
| "learning_rate": 9.384587392471515e-05, | |
| "loss": 0.4856, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 1.078805685043335, | |
| "learning_rate": 9.353730385598887e-05, | |
| "loss": 0.4332, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 12.545454545454545, | |
| "grad_norm": 1.0153404474258423, | |
| "learning_rate": 9.322171915289635e-05, | |
| "loss": 0.4568, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 1.0150026082992554, | |
| "learning_rate": 9.289917066174886e-05, | |
| "loss": 0.4417, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "grad_norm": 1.4472639560699463, | |
| "learning_rate": 9.256971035084785e-05, | |
| "loss": 0.5121, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 1.2276443243026733, | |
| "eval_runtime": 7.512, | |
| "eval_samples_per_second": 3.195, | |
| "eval_steps_per_second": 3.195, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.9721190929412842, | |
| "learning_rate": 9.223339130211192e-05, | |
| "loss": 0.361, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 13.272727272727273, | |
| "grad_norm": 1.4202260971069336, | |
| "learning_rate": 9.189026770252436e-05, | |
| "loss": 0.4109, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 1.0497146844863892, | |
| "learning_rate": 9.154039483540273e-05, | |
| "loss": 0.4459, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 13.636363636363637, | |
| "grad_norm": 1.0282554626464844, | |
| "learning_rate": 9.118382907149165e-05, | |
| "loss": 0.3977, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 1.5992000102996826, | |
| "learning_rate": 9.082062785988049e-05, | |
| "loss": 0.3675, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 1.129107117652893, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.3324, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 1.3167015314102173, | |
| "eval_runtime": 7.4961, | |
| "eval_samples_per_second": 3.202, | |
| "eval_steps_per_second": 3.202, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "step": 154, | |
| "total_flos": 3.334524867130163e+16, | |
| "train_loss": 0.8300043313534229, | |
| "train_runtime": 1413.2295, | |
| "train_samples_per_second": 3.113, | |
| "train_steps_per_second": 0.389 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.334524867130163e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |