| { | |
| "best_metric": 0.032022152096033096, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 3.004424778761062, | |
| "eval_steps": 50, | |
| "global_step": 169, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017699115044247787, | |
| "grad_norm": 1.0721659660339355, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9325, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017699115044247787, | |
| "eval_loss": 1.3935922384262085, | |
| "eval_runtime": 6.5562, | |
| "eval_samples_per_second": 14.49, | |
| "eval_steps_per_second": 3.661, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.035398230088495575, | |
| "grad_norm": 1.2187706232070923, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05309734513274336, | |
| "grad_norm": 1.1939740180969238, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0582, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07079646017699115, | |
| "grad_norm": 1.1176855564117432, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0807, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08849557522123894, | |
| "grad_norm": 1.3982747793197632, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2339, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10619469026548672, | |
| "grad_norm": 1.3325220346450806, | |
| "learning_rate": 6e-05, | |
| "loss": 1.1153, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12389380530973451, | |
| "grad_norm": 1.265844702720642, | |
| "learning_rate": 7e-05, | |
| "loss": 0.958, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1415929203539823, | |
| "grad_norm": 1.1271923780441284, | |
| "learning_rate": 8e-05, | |
| "loss": 0.7678, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1592920353982301, | |
| "grad_norm": 1.0350992679595947, | |
| "learning_rate": 9e-05, | |
| "loss": 0.586, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.17699115044247787, | |
| "grad_norm": 1.1207027435302734, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5873, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19469026548672566, | |
| "grad_norm": 1.0533980131149292, | |
| "learning_rate": 9.999024041442456e-05, | |
| "loss": 0.4849, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.21238938053097345, | |
| "grad_norm": 1.0255162715911865, | |
| "learning_rate": 9.99609654676786e-05, | |
| "loss": 0.4021, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.23008849557522124, | |
| "grad_norm": 1.1924771070480347, | |
| "learning_rate": 9.991218658821608e-05, | |
| "loss": 0.4131, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.24778761061946902, | |
| "grad_norm": 1.1136771440505981, | |
| "learning_rate": 9.984392281850293e-05, | |
| "loss": 0.2954, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.26548672566371684, | |
| "grad_norm": 0.9027830958366394, | |
| "learning_rate": 9.97562008075832e-05, | |
| "loss": 0.2755, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2831858407079646, | |
| "grad_norm": 0.6252666711807251, | |
| "learning_rate": 9.964905480067586e-05, | |
| "loss": 0.1862, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3008849557522124, | |
| "grad_norm": 0.5391494631767273, | |
| "learning_rate": 9.952252662580579e-05, | |
| "loss": 0.1796, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3185840707964602, | |
| "grad_norm": 0.5780326724052429, | |
| "learning_rate": 9.937666567747501e-05, | |
| "loss": 0.1702, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.336283185840708, | |
| "grad_norm": 0.579827070236206, | |
| "learning_rate": 9.921152889737984e-05, | |
| "loss": 0.1355, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.5776816010475159, | |
| "learning_rate": 9.902718075218176e-05, | |
| "loss": 0.1651, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.37168141592920356, | |
| "grad_norm": 0.5652853846549988, | |
| "learning_rate": 9.882369320834069e-05, | |
| "loss": 0.1378, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3893805309734513, | |
| "grad_norm": 0.5807967782020569, | |
| "learning_rate": 9.860114570402054e-05, | |
| "loss": 0.1444, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.40707964601769914, | |
| "grad_norm": 0.46384990215301514, | |
| "learning_rate": 9.835962511807786e-05, | |
| "loss": 0.1163, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4247787610619469, | |
| "grad_norm": 0.5239526033401489, | |
| "learning_rate": 9.809922573614569e-05, | |
| "loss": 0.1363, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4424778761061947, | |
| "grad_norm": 0.5285527110099792, | |
| "learning_rate": 9.782004921382612e-05, | |
| "loss": 0.1366, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.46017699115044247, | |
| "grad_norm": 0.6417925953865051, | |
| "learning_rate": 9.752220453700556e-05, | |
| "loss": 0.1262, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.4778761061946903, | |
| "grad_norm": 0.5849817395210266, | |
| "learning_rate": 9.720580797930845e-05, | |
| "loss": 0.1117, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.49557522123893805, | |
| "grad_norm": 0.6580327749252319, | |
| "learning_rate": 9.687098305670605e-05, | |
| "loss": 0.1371, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5132743362831859, | |
| "grad_norm": 0.43546798825263977, | |
| "learning_rate": 9.651786047929773e-05, | |
| "loss": 0.1155, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5309734513274337, | |
| "grad_norm": 0.42823654413223267, | |
| "learning_rate": 9.614657810028402e-05, | |
| "loss": 0.1186, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5486725663716814, | |
| "grad_norm": 0.39037245512008667, | |
| "learning_rate": 9.575728086215092e-05, | |
| "loss": 0.1127, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5663716814159292, | |
| "grad_norm": 0.3906686305999756, | |
| "learning_rate": 9.535012074008687e-05, | |
| "loss": 0.0985, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.584070796460177, | |
| "grad_norm": 0.32296302914619446, | |
| "learning_rate": 9.492525668265399e-05, | |
| "loss": 0.0729, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6017699115044248, | |
| "grad_norm": 0.3804223835468292, | |
| "learning_rate": 9.448285454973738e-05, | |
| "loss": 0.0887, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6194690265486725, | |
| "grad_norm": 0.528093695640564, | |
| "learning_rate": 9.402308704779599e-05, | |
| "loss": 0.0873, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6371681415929203, | |
| "grad_norm": 0.43243497610092163, | |
| "learning_rate": 9.354613366244108e-05, | |
| "loss": 0.1039, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6548672566371682, | |
| "grad_norm": 0.33139994740486145, | |
| "learning_rate": 9.305218058836778e-05, | |
| "loss": 0.0861, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.672566371681416, | |
| "grad_norm": 0.5364376902580261, | |
| "learning_rate": 9.254142065666801e-05, | |
| "loss": 0.0946, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6902654867256637, | |
| "grad_norm": 0.3789397180080414, | |
| "learning_rate": 9.201405325955221e-05, | |
| "loss": 0.0787, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.445186585187912, | |
| "learning_rate": 9.14702842725101e-05, | |
| "loss": 0.0822, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7256637168141593, | |
| "grad_norm": 0.3613705039024353, | |
| "learning_rate": 9.091032597394012e-05, | |
| "loss": 0.0774, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7433628318584071, | |
| "grad_norm": 0.3099222779273987, | |
| "learning_rate": 9.033439696227965e-05, | |
| "loss": 0.0521, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7610619469026548, | |
| "grad_norm": 0.27817392349243164, | |
| "learning_rate": 8.974272207066767e-05, | |
| "loss": 0.0867, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7787610619469026, | |
| "grad_norm": 0.2684073746204376, | |
| "learning_rate": 8.913553227917367e-05, | |
| "loss": 0.061, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7964601769911505, | |
| "grad_norm": 0.2929267883300781, | |
| "learning_rate": 8.851306462462688e-05, | |
| "loss": 0.0698, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8141592920353983, | |
| "grad_norm": 0.3591836392879486, | |
| "learning_rate": 8.787556210808101e-05, | |
| "loss": 0.0653, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.831858407079646, | |
| "grad_norm": 0.36023128032684326, | |
| "learning_rate": 8.722327359995064e-05, | |
| "loss": 0.0733, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8495575221238938, | |
| "grad_norm": 0.34157848358154297, | |
| "learning_rate": 8.655645374285637e-05, | |
| "loss": 0.0611, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8672566371681416, | |
| "grad_norm": 0.3172074854373932, | |
| "learning_rate": 8.587536285221656e-05, | |
| "loss": 0.0689, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8849557522123894, | |
| "grad_norm": 0.27993178367614746, | |
| "learning_rate": 8.518026681462448e-05, | |
| "loss": 0.0566, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8849557522123894, | |
| "eval_loss": 0.06235470622777939, | |
| "eval_runtime": 6.691, | |
| "eval_samples_per_second": 14.198, | |
| "eval_steps_per_second": 3.587, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9026548672566371, | |
| "grad_norm": 0.21425798535346985, | |
| "learning_rate": 8.44714369840506e-05, | |
| "loss": 0.0481, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9203539823008849, | |
| "grad_norm": 0.2822091579437256, | |
| "learning_rate": 8.374915007591053e-05, | |
| "loss": 0.0539, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9380530973451328, | |
| "grad_norm": 0.3757535517215729, | |
| "learning_rate": 8.301368805903988e-05, | |
| "loss": 0.0578, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.9557522123893806, | |
| "grad_norm": 0.36872968077659607, | |
| "learning_rate": 8.226533804561827e-05, | |
| "loss": 0.0614, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9734513274336283, | |
| "grad_norm": 0.4218049943447113, | |
| "learning_rate": 8.150439217908556e-05, | |
| "loss": 0.0822, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9911504424778761, | |
| "grad_norm": 0.5659615993499756, | |
| "learning_rate": 8.073114752009387e-05, | |
| "loss": 0.0642, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0132743362831858, | |
| "grad_norm": 0.9227009415626526, | |
| "learning_rate": 7.994590593054001e-05, | |
| "loss": 0.127, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.0309734513274336, | |
| "grad_norm": 0.20931123197078705, | |
| "learning_rate": 7.91489739557236e-05, | |
| "loss": 0.0463, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0486725663716814, | |
| "grad_norm": 0.25171875953674316, | |
| "learning_rate": 7.83406627046769e-05, | |
| "loss": 0.0586, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0663716814159292, | |
| "grad_norm": 0.24994538724422455, | |
| "learning_rate": 7.752128772871292e-05, | |
| "loss": 0.0424, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.084070796460177, | |
| "grad_norm": 0.23162053525447845, | |
| "learning_rate": 7.669116889823955e-05, | |
| "loss": 0.0436, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.1017699115044248, | |
| "grad_norm": 0.2777061462402344, | |
| "learning_rate": 7.585063027788731e-05, | |
| "loss": 0.043, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.1194690265486726, | |
| "grad_norm": 0.2399117797613144, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.0475, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1371681415929205, | |
| "grad_norm": 0.2516414225101471, | |
| "learning_rate": 7.413961013653726e-05, | |
| "loss": 0.0434, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.154867256637168, | |
| "grad_norm": 0.15977510809898376, | |
| "learning_rate": 7.326979656943906e-05, | |
| "loss": 0.0282, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1725663716814159, | |
| "grad_norm": 0.23954959213733673, | |
| "learning_rate": 7.239089885950316e-05, | |
| "loss": 0.0324, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1902654867256637, | |
| "grad_norm": 0.15419161319732666, | |
| "learning_rate": 7.150326011382604e-05, | |
| "loss": 0.0235, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.2079646017699115, | |
| "grad_norm": 0.2992177903652191, | |
| "learning_rate": 7.060722685185961e-05, | |
| "loss": 0.0317, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.2256637168141593, | |
| "grad_norm": 0.2721240222454071, | |
| "learning_rate": 6.970314887013584e-05, | |
| "loss": 0.0441, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.2433628318584071, | |
| "grad_norm": 0.19432015717029572, | |
| "learning_rate": 6.879137910571191e-05, | |
| "loss": 0.0235, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.261061946902655, | |
| "grad_norm": 0.2833891212940216, | |
| "learning_rate": 6.787227349838947e-05, | |
| "loss": 0.0408, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.2787610619469025, | |
| "grad_norm": 0.24525463581085205, | |
| "learning_rate": 6.694619085176159e-05, | |
| "loss": 0.0469, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.2964601769911503, | |
| "grad_norm": 0.2467905730009079, | |
| "learning_rate": 6.601349269314188e-05, | |
| "loss": 0.0446, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.3141592920353982, | |
| "grad_norm": 0.18387003242969513, | |
| "learning_rate": 6.507454313243015e-05, | |
| "loss": 0.0334, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.331858407079646, | |
| "grad_norm": 0.22559407353401184, | |
| "learning_rate": 6.412970871996995e-05, | |
| "loss": 0.0292, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.3495575221238938, | |
| "grad_norm": 0.18541377782821655, | |
| "learning_rate": 6.317935830345338e-05, | |
| "loss": 0.0351, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3672566371681416, | |
| "grad_norm": 0.17538294196128845, | |
| "learning_rate": 6.222386288392913e-05, | |
| "loss": 0.0329, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.3849557522123894, | |
| "grad_norm": 0.21896906197071075, | |
| "learning_rate": 6.126359547096975e-05, | |
| "loss": 0.0313, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.4026548672566372, | |
| "grad_norm": 0.1702110767364502, | |
| "learning_rate": 6.029893093705492e-05, | |
| "loss": 0.0168, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.420353982300885, | |
| "grad_norm": 0.279254674911499, | |
| "learning_rate": 5.9330245871227454e-05, | |
| "loss": 0.0371, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4380530973451329, | |
| "grad_norm": 0.20467723906040192, | |
| "learning_rate": 5.835791843207916e-05, | |
| "loss": 0.0251, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.4557522123893805, | |
| "grad_norm": 0.23105274140834808, | |
| "learning_rate": 5.738232820012407e-05, | |
| "loss": 0.0243, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.4734513274336283, | |
| "grad_norm": 0.2554505467414856, | |
| "learning_rate": 5.640385602961634e-05, | |
| "loss": 0.0226, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.491150442477876, | |
| "grad_norm": 0.3279307186603546, | |
| "learning_rate": 5.5422883899871284e-05, | |
| "loss": 0.0181, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.508849557522124, | |
| "grad_norm": 0.23609845340251923, | |
| "learning_rate": 5.4439794766146746e-05, | |
| "loss": 0.03, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.5265486725663717, | |
| "grad_norm": 0.19427117705345154, | |
| "learning_rate": 5.34549724101439e-05, | |
| "loss": 0.0283, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.5442477876106193, | |
| "grad_norm": 0.21057020127773285, | |
| "learning_rate": 5.246880129018516e-05, | |
| "loss": 0.0328, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.5619469026548671, | |
| "grad_norm": 0.157065749168396, | |
| "learning_rate": 5.148166639112799e-05, | |
| "loss": 0.0265, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.579646017699115, | |
| "grad_norm": 0.21837662160396576, | |
| "learning_rate": 5.049395307407329e-05, | |
| "loss": 0.0353, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.5973451327433628, | |
| "grad_norm": 0.2608608305454254, | |
| "learning_rate": 4.950604692592672e-05, | |
| "loss": 0.0445, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.6150442477876106, | |
| "grad_norm": 0.29214149713516235, | |
| "learning_rate": 4.851833360887201e-05, | |
| "loss": 0.0319, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.6327433628318584, | |
| "grad_norm": 0.3026266396045685, | |
| "learning_rate": 4.7531198709814854e-05, | |
| "loss": 0.0387, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.6504424778761062, | |
| "grad_norm": 0.1753360629081726, | |
| "learning_rate": 4.654502758985611e-05, | |
| "loss": 0.0178, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.668141592920354, | |
| "grad_norm": 0.29309213161468506, | |
| "learning_rate": 4.5560205233853266e-05, | |
| "loss": 0.0279, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.6858407079646018, | |
| "grad_norm": 0.1859235018491745, | |
| "learning_rate": 4.4577116100128735e-05, | |
| "loss": 0.0225, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.7035398230088497, | |
| "grad_norm": 0.21251869201660156, | |
| "learning_rate": 4.3596143970383664e-05, | |
| "loss": 0.028, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.7212389380530975, | |
| "grad_norm": 0.2775147259235382, | |
| "learning_rate": 4.2617671799875944e-05, | |
| "loss": 0.0323, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.7389380530973453, | |
| "grad_norm": 0.3155283033847809, | |
| "learning_rate": 4.1642081567920846e-05, | |
| "loss": 0.0194, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.7566371681415929, | |
| "grad_norm": 0.13303042948246002, | |
| "learning_rate": 4.066975412877255e-05, | |
| "loss": 0.0211, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.7743362831858407, | |
| "grad_norm": 0.19871264696121216, | |
| "learning_rate": 3.970106906294509e-05, | |
| "loss": 0.0302, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7743362831858407, | |
| "eval_loss": 0.0379580594599247, | |
| "eval_runtime": 6.6861, | |
| "eval_samples_per_second": 14.209, | |
| "eval_steps_per_second": 3.59, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7920353982300885, | |
| "grad_norm": 0.20991753041744232, | |
| "learning_rate": 3.873640452903026e-05, | |
| "loss": 0.0253, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.8097345132743363, | |
| "grad_norm": 0.3313141167163849, | |
| "learning_rate": 3.777613711607087e-05, | |
| "loss": 0.0289, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.827433628318584, | |
| "grad_norm": 0.19206620752811432, | |
| "learning_rate": 3.682064169654663e-05, | |
| "loss": 0.0265, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.8451327433628317, | |
| "grad_norm": 0.4013370871543884, | |
| "learning_rate": 3.587029128003006e-05, | |
| "loss": 0.0308, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.8628318584070795, | |
| "grad_norm": 0.18024751543998718, | |
| "learning_rate": 3.492545686756986e-05, | |
| "loss": 0.0162, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.8805309734513274, | |
| "grad_norm": 0.2305198758840561, | |
| "learning_rate": 3.3986507306858125e-05, | |
| "loss": 0.0284, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.8982300884955752, | |
| "grad_norm": 0.19050440192222595, | |
| "learning_rate": 3.3053809148238426e-05, | |
| "loss": 0.0227, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.915929203539823, | |
| "grad_norm": 0.15006108582019806, | |
| "learning_rate": 3.212772650161056e-05, | |
| "loss": 0.0191, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.9336283185840708, | |
| "grad_norm": 0.2283119261264801, | |
| "learning_rate": 3.12086208942881e-05, | |
| "loss": 0.0227, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.9513274336283186, | |
| "grad_norm": 0.2260059416294098, | |
| "learning_rate": 3.0296851129864168e-05, | |
| "loss": 0.0253, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.9690265486725664, | |
| "grad_norm": 0.23939450085163116, | |
| "learning_rate": 2.9392773148140408e-05, | |
| "loss": 0.0265, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.9867256637168142, | |
| "grad_norm": 0.1815921664237976, | |
| "learning_rate": 2.8496739886173995e-05, | |
| "loss": 0.0216, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.0088495575221237, | |
| "grad_norm": 0.24775980412960052, | |
| "learning_rate": 2.7609101140496863e-05, | |
| "loss": 0.0232, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.0265486725663715, | |
| "grad_norm": 0.12413739413022995, | |
| "learning_rate": 2.6730203430560947e-05, | |
| "loss": 0.0206, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.0442477876106193, | |
| "grad_norm": 0.1339850276708603, | |
| "learning_rate": 2.5860389863462765e-05, | |
| "loss": 0.0196, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.061946902654867, | |
| "grad_norm": 0.13854354619979858, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.0159, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.079646017699115, | |
| "grad_norm": 0.1154838278889656, | |
| "learning_rate": 2.414936972211272e-05, | |
| "loss": 0.0142, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.0973451327433628, | |
| "grad_norm": 0.10285928845405579, | |
| "learning_rate": 2.3308831101760486e-05, | |
| "loss": 0.01, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.1150442477876106, | |
| "grad_norm": 0.1806727647781372, | |
| "learning_rate": 2.247871227128709e-05, | |
| "loss": 0.0187, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.1327433628318584, | |
| "grad_norm": 0.19010406732559204, | |
| "learning_rate": 2.1659337295323118e-05, | |
| "loss": 0.0235, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.150442477876106, | |
| "grad_norm": 0.1291082799434662, | |
| "learning_rate": 2.0851026044276406e-05, | |
| "loss": 0.0144, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.168141592920354, | |
| "grad_norm": 0.16923344135284424, | |
| "learning_rate": 2.005409406946e-05, | |
| "loss": 0.0138, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.185840707964602, | |
| "grad_norm": 0.14110802114009857, | |
| "learning_rate": 1.9268852479906147e-05, | |
| "loss": 0.0119, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.2035398230088497, | |
| "grad_norm": 0.10906057059764862, | |
| "learning_rate": 1.849560782091445e-05, | |
| "loss": 0.009, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.2212389380530975, | |
| "grad_norm": 0.1177993193268776, | |
| "learning_rate": 1.7734661954381754e-05, | |
| "loss": 0.014, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.2389380530973453, | |
| "grad_norm": 0.10654302686452866, | |
| "learning_rate": 1.6986311940960147e-05, | |
| "loss": 0.0101, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.256637168141593, | |
| "grad_norm": 0.16150303184986115, | |
| "learning_rate": 1.6250849924089484e-05, | |
| "loss": 0.0168, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.274336283185841, | |
| "grad_norm": 0.12198394536972046, | |
| "learning_rate": 1.552856301594942e-05, | |
| "loss": 0.0136, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.2920353982300883, | |
| "grad_norm": 0.16058827936649323, | |
| "learning_rate": 1.4819733185375534e-05, | |
| "loss": 0.0196, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.309734513274336, | |
| "grad_norm": 0.19370707869529724, | |
| "learning_rate": 1.4124637147783432e-05, | |
| "loss": 0.0205, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.327433628318584, | |
| "grad_norm": 0.0963333398103714, | |
| "learning_rate": 1.3443546257143624e-05, | |
| "loss": 0.011, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.3451327433628317, | |
| "grad_norm": 0.1633400171995163, | |
| "learning_rate": 1.277672640004936e-05, | |
| "loss": 0.0208, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.3628318584070795, | |
| "grad_norm": 0.1224859431385994, | |
| "learning_rate": 1.2124437891918993e-05, | |
| "loss": 0.015, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.3805309734513274, | |
| "grad_norm": 0.10458586364984512, | |
| "learning_rate": 1.1486935375373126e-05, | |
| "loss": 0.0101, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.398230088495575, | |
| "grad_norm": 0.15571683645248413, | |
| "learning_rate": 1.0864467720826343e-05, | |
| "loss": 0.0162, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.415929203539823, | |
| "grad_norm": 0.08478286117315292, | |
| "learning_rate": 1.0257277929332332e-05, | |
| "loss": 0.0072, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.433628318584071, | |
| "grad_norm": 0.10365074872970581, | |
| "learning_rate": 9.66560303772035e-06, | |
| "loss": 0.0095, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.4513274336283186, | |
| "grad_norm": 0.07361367344856262, | |
| "learning_rate": 9.08967402605988e-06, | |
| "loss": 0.0076, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.4690265486725664, | |
| "grad_norm": 0.13091503083705902, | |
| "learning_rate": 8.529715727489912e-06, | |
| "loss": 0.0082, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.4867256637168142, | |
| "grad_norm": 0.11134269833564758, | |
| "learning_rate": 7.985946740447791e-06, | |
| "loss": 0.0079, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.504424778761062, | |
| "grad_norm": 0.18341496586799622, | |
| "learning_rate": 7.458579343331995e-06, | |
| "loss": 0.0129, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.52212389380531, | |
| "grad_norm": 0.11337222903966904, | |
| "learning_rate": 6.947819411632223e-06, | |
| "loss": 0.0173, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.5398230088495577, | |
| "grad_norm": 0.20701636373996735, | |
| "learning_rate": 6.45386633755894e-06, | |
| "loss": 0.0175, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.557522123893805, | |
| "grad_norm": 0.20868322253227234, | |
| "learning_rate": 5.976912952204017e-06, | |
| "loss": 0.0147, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.5752212389380533, | |
| "grad_norm": 0.10272178053855896, | |
| "learning_rate": 5.51714545026264e-06, | |
| "loss": 0.0128, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.5929203539823007, | |
| "grad_norm": 0.14340701699256897, | |
| "learning_rate": 5.074743317346009e-06, | |
| "loss": 0.0136, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.6106194690265485, | |
| "grad_norm": 0.12963014841079712, | |
| "learning_rate": 4.649879259913137e-06, | |
| "loss": 0.0093, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.6283185840707963, | |
| "grad_norm": 0.12100816518068314, | |
| "learning_rate": 4.242719137849077e-06, | |
| "loss": 0.0107, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.646017699115044, | |
| "grad_norm": 0.09687142819166183, | |
| "learning_rate": 3.853421899715992e-06, | |
| "loss": 0.0082, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.663716814159292, | |
| "grad_norm": 0.09663711488246918, | |
| "learning_rate": 3.4821395207022766e-06, | |
| "loss": 0.0081, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.663716814159292, | |
| "eval_loss": 0.032022152096033096, | |
| "eval_runtime": 6.6983, | |
| "eval_samples_per_second": 14.183, | |
| "eval_steps_per_second": 3.583, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.6814159292035398, | |
| "grad_norm": 0.12883880734443665, | |
| "learning_rate": 3.1290169432939553e-06, | |
| "loss": 0.0114, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.6991150442477876, | |
| "grad_norm": 0.16089491546154022, | |
| "learning_rate": 2.794192020691544e-06, | |
| "loss": 0.0165, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.7168141592920354, | |
| "grad_norm": 0.25050851702690125, | |
| "learning_rate": 2.4777954629944477e-06, | |
| "loss": 0.0182, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.734513274336283, | |
| "grad_norm": 0.09175170958042145, | |
| "learning_rate": 2.179950786173879e-06, | |
| "loss": 0.0057, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.752212389380531, | |
| "grad_norm": 0.1262611448764801, | |
| "learning_rate": 1.9007742638543102e-06, | |
| "loss": 0.0119, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.769911504424779, | |
| "grad_norm": 0.13375544548034668, | |
| "learning_rate": 1.6403748819221466e-06, | |
| "loss": 0.0125, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.7876106194690267, | |
| "grad_norm": 0.09264933317899704, | |
| "learning_rate": 1.3988542959794627e-06, | |
| "loss": 0.0109, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.8053097345132745, | |
| "grad_norm": 0.10276864469051361, | |
| "learning_rate": 1.1763067916593262e-06, | |
| "loss": 0.0125, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.823008849557522, | |
| "grad_norm": 0.08319110423326492, | |
| "learning_rate": 9.728192478182574e-07, | |
| "loss": 0.0082, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.84070796460177, | |
| "grad_norm": 0.168908029794693, | |
| "learning_rate": 7.884711026201585e-07, | |
| "loss": 0.0144, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.8584070796460175, | |
| "grad_norm": 0.1352635771036148, | |
| "learning_rate": 6.233343225249933e-07, | |
| "loss": 0.0158, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.8761061946902657, | |
| "grad_norm": 0.09505495429039001, | |
| "learning_rate": 4.774733741942206e-07, | |
| "loss": 0.0099, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.893805309734513, | |
| "grad_norm": 0.0929616317152977, | |
| "learning_rate": 3.5094519932415417e-07, | |
| "loss": 0.0078, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.911504424778761, | |
| "grad_norm": 0.10294952243566513, | |
| "learning_rate": 2.437991924167937e-07, | |
| "loss": 0.012, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.9292035398230087, | |
| "grad_norm": 0.125913605093956, | |
| "learning_rate": 1.560771814970885e-07, | |
| "loss": 0.0088, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.9469026548672566, | |
| "grad_norm": 0.14280952513217926, | |
| "learning_rate": 8.781341178393244e-08, | |
| "loss": 0.011, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.9646017699115044, | |
| "grad_norm": 0.1293206512928009, | |
| "learning_rate": 3.9034532321408076e-08, | |
| "loss": 0.0099, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.982300884955752, | |
| "grad_norm": 0.1486680805683136, | |
| "learning_rate": 9.75958557545842e-09, | |
| "loss": 0.0107, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.004424778761062, | |
| "grad_norm": 0.1905655860900879, | |
| "learning_rate": 0.0, | |
| "loss": 0.0132, | |
| "step": 169 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 169, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.248537858501509e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |