| { | |
| "best_metric": 0.8594623804092407, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.5145797598627787, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003430531732418525, | |
| "grad_norm": 11.61568832397461, | |
| "learning_rate": 5e-06, | |
| "loss": 5.0126, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003430531732418525, | |
| "eval_loss": 1.4074695110321045, | |
| "eval_runtime": 40.4473, | |
| "eval_samples_per_second": 12.139, | |
| "eval_steps_per_second": 6.082, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00686106346483705, | |
| "grad_norm": 13.636929512023926, | |
| "learning_rate": 1e-05, | |
| "loss": 4.982, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010291595197255575, | |
| "grad_norm": 14.98882007598877, | |
| "learning_rate": 1.5e-05, | |
| "loss": 4.99, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0137221269296741, | |
| "grad_norm": 10.213421821594238, | |
| "learning_rate": 2e-05, | |
| "loss": 4.9476, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.017152658662092625, | |
| "grad_norm": 8.086559295654297, | |
| "learning_rate": 2.5e-05, | |
| "loss": 4.9681, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02058319039451115, | |
| "grad_norm": 9.853785514831543, | |
| "learning_rate": 3e-05, | |
| "loss": 4.7214, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.024013722126929673, | |
| "grad_norm": 8.948219299316406, | |
| "learning_rate": 3.5e-05, | |
| "loss": 4.8078, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0274442538593482, | |
| "grad_norm": 7.905782699584961, | |
| "learning_rate": 4e-05, | |
| "loss": 4.7025, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.030874785591766724, | |
| "grad_norm": 7.111548900604248, | |
| "learning_rate": 4.5e-05, | |
| "loss": 4.6991, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03430531732418525, | |
| "grad_norm": 7.046090602874756, | |
| "learning_rate": 5e-05, | |
| "loss": 4.5764, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03773584905660377, | |
| "grad_norm": 6.978516101837158, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 4.5503, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0411663807890223, | |
| "grad_norm": 6.4846320152282715, | |
| "learning_rate": 6e-05, | |
| "loss": 4.2789, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.044596912521440824, | |
| "grad_norm": 6.363194942474365, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 4.2641, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.048027444253859346, | |
| "grad_norm": 7.184693336486816, | |
| "learning_rate": 7e-05, | |
| "loss": 4.4203, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.051457975986277875, | |
| "grad_norm": 6.728307723999023, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 4.5207, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0548885077186964, | |
| "grad_norm": 5.945248126983643, | |
| "learning_rate": 8e-05, | |
| "loss": 4.2506, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.058319039451114926, | |
| "grad_norm": 6.138426303863525, | |
| "learning_rate": 8.5e-05, | |
| "loss": 4.3399, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06174957118353345, | |
| "grad_norm": 6.568054676055908, | |
| "learning_rate": 9e-05, | |
| "loss": 4.3236, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06518010291595197, | |
| "grad_norm": 5.957953453063965, | |
| "learning_rate": 9.5e-05, | |
| "loss": 4.2521, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0686106346483705, | |
| "grad_norm": 6.30268669128418, | |
| "learning_rate": 0.0001, | |
| "loss": 4.3019, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07204116638078903, | |
| "grad_norm": 5.6199493408203125, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 3.9036, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07547169811320754, | |
| "grad_norm": 6.134159564971924, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 4.1892, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07890222984562607, | |
| "grad_norm": 5.793374538421631, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 4.14, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0823327615780446, | |
| "grad_norm": 6.174600124359131, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 3.9926, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08576329331046312, | |
| "grad_norm": 6.1931023597717285, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 4.3605, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08919382504288165, | |
| "grad_norm": 5.781077861785889, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 4.0821, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09262435677530018, | |
| "grad_norm": 5.663956165313721, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 3.8834, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09605488850771869, | |
| "grad_norm": 5.542593479156494, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 4.05, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09948542024013722, | |
| "grad_norm": 5.556862831115723, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 3.7908, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10291595197255575, | |
| "grad_norm": 5.6052093505859375, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 3.9222, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10634648370497427, | |
| "grad_norm": 5.743956565856934, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 4.0824, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.1097770154373928, | |
| "grad_norm": 5.386566162109375, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 3.8401, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.11320754716981132, | |
| "grad_norm": 5.688109874725342, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 3.7746, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.11663807890222985, | |
| "grad_norm": 5.799955368041992, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 3.8266, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.12006861063464837, | |
| "grad_norm": 5.360304832458496, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 3.5762, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1234991423670669, | |
| "grad_norm": 6.067436695098877, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 3.8938, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.1269296740994854, | |
| "grad_norm": 6.037448406219482, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 3.677, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.13036020583190394, | |
| "grad_norm": 6.000154972076416, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 4.0107, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.13379073756432247, | |
| "grad_norm": 6.464019775390625, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 4.0673, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.137221269296741, | |
| "grad_norm": 5.797187805175781, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 3.724, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14065180102915953, | |
| "grad_norm": 6.833194732666016, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 4.3818, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.14408233276157806, | |
| "grad_norm": 5.868514060974121, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 3.9079, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.14751286449399656, | |
| "grad_norm": 6.458984851837158, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 3.9524, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1509433962264151, | |
| "grad_norm": 6.594856262207031, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 3.6577, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.15437392795883362, | |
| "grad_norm": 7.024727821350098, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 3.9407, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15780445969125215, | |
| "grad_norm": 7.5185136795043945, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 4.1704, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.16123499142367068, | |
| "grad_norm": 7.434786319732666, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 3.7691, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1646655231560892, | |
| "grad_norm": 7.97662353515625, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 4.1047, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.1680960548885077, | |
| "grad_norm": 7.469397068023682, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 4.0323, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.17152658662092624, | |
| "grad_norm": 9.097603797912598, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 4.0814, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17152658662092624, | |
| "eval_loss": 1.0552418231964111, | |
| "eval_runtime": 41.0958, | |
| "eval_samples_per_second": 11.948, | |
| "eval_steps_per_second": 5.986, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17495711835334476, | |
| "grad_norm": 7.74439811706543, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 4.8988, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1783876500857633, | |
| "grad_norm": 5.500019073486328, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 4.8596, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 3.838832378387451, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 4.4589, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.18524871355060035, | |
| "grad_norm": 3.795142650604248, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 4.2005, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.18867924528301888, | |
| "grad_norm": 3.751390218734741, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 4.1903, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.19210977701543738, | |
| "grad_norm": 3.823895215988159, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 3.9744, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1955403087478559, | |
| "grad_norm": 3.9442529678344727, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 4.3034, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.19897084048027444, | |
| "grad_norm": 4.070255279541016, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 4.0883, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.20240137221269297, | |
| "grad_norm": 3.7813539505004883, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 3.9695, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.2058319039451115, | |
| "grad_norm": 3.8070478439331055, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 3.7485, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.20926243567753003, | |
| "grad_norm": 3.9176182746887207, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 4.011, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.21269296740994853, | |
| "grad_norm": 4.07338809967041, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 4.0232, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.21612349914236706, | |
| "grad_norm": 4.0383782386779785, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 3.978, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2195540308747856, | |
| "grad_norm": 3.6648452281951904, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 3.4885, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.22298456260720412, | |
| "grad_norm": 3.8878605365753174, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 3.8533, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22641509433962265, | |
| "grad_norm": 4.185667991638184, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 3.6215, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.22984562607204118, | |
| "grad_norm": 4.071247100830078, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 3.531, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2332761578044597, | |
| "grad_norm": 4.344533443450928, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 3.7186, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2367066895368782, | |
| "grad_norm": 4.426985740661621, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 3.6128, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.24013722126929674, | |
| "grad_norm": 4.162020206451416, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 3.3318, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.24356775300171526, | |
| "grad_norm": 4.368765354156494, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 3.4157, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2469982847341338, | |
| "grad_norm": 4.309731960296631, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 3.5808, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2504288164665523, | |
| "grad_norm": 4.442713260650635, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 3.4605, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2538593481989708, | |
| "grad_norm": 4.398868083953857, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 3.6249, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.25728987993138935, | |
| "grad_norm": 4.782496929168701, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 3.7258, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2607204116638079, | |
| "grad_norm": 4.372551918029785, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 3.7134, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.2641509433962264, | |
| "grad_norm": 4.290441989898682, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 3.7541, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.26758147512864494, | |
| "grad_norm": 4.507718086242676, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 3.9408, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.27101200686106347, | |
| "grad_norm": 4.117275714874268, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 3.4982, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.274442538593482, | |
| "grad_norm": 4.557469844818115, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 3.7516, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.27787307032590053, | |
| "grad_norm": 5.15295934677124, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 3.5183, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.28130360205831906, | |
| "grad_norm": 4.726450443267822, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 3.526, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2847341337907376, | |
| "grad_norm": 4.847685813903809, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 3.858, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.2881646655231561, | |
| "grad_norm": 4.568885803222656, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 3.6064, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2915951972555746, | |
| "grad_norm": 4.916790962219238, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 3.383, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2950257289879931, | |
| "grad_norm": 5.058016777038574, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 3.8242, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.29845626072041165, | |
| "grad_norm": 4.883471965789795, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 3.9667, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3018867924528302, | |
| "grad_norm": 4.773396968841553, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 3.5179, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3053173241852487, | |
| "grad_norm": 5.102062225341797, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 3.7781, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.30874785591766724, | |
| "grad_norm": 4.849315643310547, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 3.59, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.31217838765008576, | |
| "grad_norm": 5.05657958984375, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 3.8574, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3156089193825043, | |
| "grad_norm": 5.094900608062744, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.6297, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3190394511149228, | |
| "grad_norm": 5.497709274291992, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 3.4883, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.32246998284734135, | |
| "grad_norm": 5.846022605895996, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 3.5978, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3259005145797599, | |
| "grad_norm": 5.982120990753174, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 3.7544, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3293310463121784, | |
| "grad_norm": 5.516687870025635, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 3.3577, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.33276157804459694, | |
| "grad_norm": 5.789833068847656, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 3.3941, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3361921097770154, | |
| "grad_norm": 5.645137310028076, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 3.3852, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.33962264150943394, | |
| "grad_norm": 7.209895610809326, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 3.7848, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.34305317324185247, | |
| "grad_norm": 9.69320011138916, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 3.4986, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.34305317324185247, | |
| "eval_loss": 0.9467170834541321, | |
| "eval_runtime": 41.1492, | |
| "eval_samples_per_second": 11.932, | |
| "eval_steps_per_second": 5.978, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.346483704974271, | |
| "grad_norm": 6.344611644744873, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 4.4591, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.34991423670668953, | |
| "grad_norm": 5.29261589050293, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 4.3903, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.35334476843910806, | |
| "grad_norm": 3.7538249492645264, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 4.1005, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3567753001715266, | |
| "grad_norm": 3.1446287631988525, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 3.7796, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.3602058319039451, | |
| "grad_norm": 2.960186719894409, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 3.916, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 3.2721056938171387, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 3.9345, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3670668953687822, | |
| "grad_norm": 3.109593629837036, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 3.6805, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.3704974271012007, | |
| "grad_norm": 3.3663551807403564, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 4.1047, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.37392795883361923, | |
| "grad_norm": 3.3839170932769775, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 3.7359, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.37735849056603776, | |
| "grad_norm": 3.2243165969848633, | |
| "learning_rate": 5e-05, | |
| "loss": 3.369, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.38078902229845624, | |
| "grad_norm": 3.484187364578247, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 3.868, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.38421955403087477, | |
| "grad_norm": 3.4150843620300293, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 3.5268, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.3876500857632933, | |
| "grad_norm": 3.528571367263794, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 3.4223, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3910806174957118, | |
| "grad_norm": 3.6814565658569336, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 3.5602, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.39451114922813035, | |
| "grad_norm": 3.4944686889648438, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 3.3631, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3979416809605489, | |
| "grad_norm": 3.5598247051239014, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 3.2197, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.4013722126929674, | |
| "grad_norm": 3.637119770050049, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 3.6207, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.40480274442538594, | |
| "grad_norm": 3.621417760848999, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 3.66, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.40823327615780447, | |
| "grad_norm": 3.71958065032959, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 3.4363, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.411663807890223, | |
| "grad_norm": 3.881575107574463, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 3.6837, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.41509433962264153, | |
| "grad_norm": 3.5961995124816895, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 3.4126, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.41852487135506006, | |
| "grad_norm": 3.635561943054199, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 3.4925, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4219554030874786, | |
| "grad_norm": 3.558687448501587, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 3.1293, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.42538593481989706, | |
| "grad_norm": 3.7333993911743164, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 3.1551, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.4288164665523156, | |
| "grad_norm": 4.045219898223877, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 3.5025, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4322469982847341, | |
| "grad_norm": 4.082742691040039, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 3.7588, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.43567753001715265, | |
| "grad_norm": 4.03236722946167, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 3.6083, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.4391080617495712, | |
| "grad_norm": 4.0244879722595215, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 3.5657, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4425385934819897, | |
| "grad_norm": 3.9571797847747803, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 3.5047, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.44596912521440824, | |
| "grad_norm": 4.1520161628723145, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 3.4308, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.44939965694682676, | |
| "grad_norm": 3.731208086013794, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 3.3306, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4528301886792453, | |
| "grad_norm": 4.370545864105225, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 3.4649, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4562607204116638, | |
| "grad_norm": 4.4801025390625, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 3.3989, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.45969125214408235, | |
| "grad_norm": 4.434890270233154, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 3.3249, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.4631217838765009, | |
| "grad_norm": 4.163226127624512, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 3.2922, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4665523156089194, | |
| "grad_norm": 4.4532151222229, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 3.7554, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4699828473413379, | |
| "grad_norm": 4.65369987487793, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 3.5652, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4734133790737564, | |
| "grad_norm": 4.77457857131958, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 3.6568, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.47684391080617494, | |
| "grad_norm": 4.440027713775635, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 3.459, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.48027444253859347, | |
| "grad_norm": 5.082893371582031, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 3.4999, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.483704974271012, | |
| "grad_norm": 4.5789103507995605, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 3.4001, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.48713550600343053, | |
| "grad_norm": 4.885765552520752, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 3.5424, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.49056603773584906, | |
| "grad_norm": 5.264125347137451, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 3.5362, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.4939965694682676, | |
| "grad_norm": 4.9319539070129395, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 3.3161, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4974271012006861, | |
| "grad_norm": 5.442240238189697, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 3.5072, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5008576329331046, | |
| "grad_norm": 5.463189125061035, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 3.3095, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5042881646655232, | |
| "grad_norm": 5.8514862060546875, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 3.3437, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5077186963979416, | |
| "grad_norm": 6.340907096862793, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 3.4434, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5111492281303602, | |
| "grad_norm": 7.861410140991211, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 4.1433, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5145797598627787, | |
| "grad_norm": 8.327228546142578, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 4.0453, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5145797598627787, | |
| "eval_loss": 0.8594623804092407, | |
| "eval_runtime": 41.113, | |
| "eval_samples_per_second": 11.943, | |
| "eval_steps_per_second": 5.984, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.2144694189083853e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |