| { | |
| "best_metric": 1.3218529224395752, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.09721322099805574, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006480881399870382, | |
| "grad_norm": 33.14695739746094, | |
| "learning_rate": 1e-05, | |
| "loss": 10.2522, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0006480881399870382, | |
| "eval_loss": 3.3373067378997803, | |
| "eval_runtime": 364.3083, | |
| "eval_samples_per_second": 7.134, | |
| "eval_steps_per_second": 1.784, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0012961762799740765, | |
| "grad_norm": 34.84189224243164, | |
| "learning_rate": 2e-05, | |
| "loss": 10.4056, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0019442644199611147, | |
| "grad_norm": 22.47279930114746, | |
| "learning_rate": 3e-05, | |
| "loss": 10.273, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.002592352559948153, | |
| "grad_norm": 17.44116973876953, | |
| "learning_rate": 4e-05, | |
| "loss": 9.8766, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0032404406999351912, | |
| "grad_norm": 20.306068420410156, | |
| "learning_rate": 5e-05, | |
| "loss": 10.3176, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0038885288399222295, | |
| "grad_norm": 20.21626091003418, | |
| "learning_rate": 6e-05, | |
| "loss": 9.7409, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.004536616979909268, | |
| "grad_norm": 17.080589294433594, | |
| "learning_rate": 7e-05, | |
| "loss": 9.4632, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.005184705119896306, | |
| "grad_norm": 15.596521377563477, | |
| "learning_rate": 8e-05, | |
| "loss": 9.37, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.005832793259883344, | |
| "grad_norm": 15.58171558380127, | |
| "learning_rate": 9e-05, | |
| "loss": 8.8766, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0064808813998703824, | |
| "grad_norm": 14.870361328125, | |
| "learning_rate": 0.0001, | |
| "loss": 8.662, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.007128969539857421, | |
| "grad_norm": 14.336933135986328, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 8.3562, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.007777057679844459, | |
| "grad_norm": 13.64636516571045, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 7.846, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.008425145819831496, | |
| "grad_norm": 13.630660057067871, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 7.6262, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.009073233959818535, | |
| "grad_norm": 13.014859199523926, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 7.4459, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.009721322099805573, | |
| "grad_norm": 12.917651176452637, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 7.005, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010369410239792612, | |
| "grad_norm": 13.161778450012207, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 7.3123, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.01101749837977965, | |
| "grad_norm": 11.873573303222656, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 6.974, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.011665586519766688, | |
| "grad_norm": 11.698904991149902, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 6.983, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.012313674659753726, | |
| "grad_norm": 11.819367408752441, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 6.7763, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.012961762799740765, | |
| "grad_norm": 11.071271896362305, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 7.1351, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013609850939727802, | |
| "grad_norm": 10.759208679199219, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 6.8721, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.014257939079714841, | |
| "grad_norm": 11.884669303894043, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 6.8917, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.014906027219701879, | |
| "grad_norm": 11.246404647827148, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 6.2715, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.015554115359688918, | |
| "grad_norm": 11.680570602416992, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 7.0292, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.016202203499675955, | |
| "grad_norm": 11.046855926513672, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 5.9442, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.016850291639662993, | |
| "grad_norm": 12.189892768859863, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 6.3471, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.017498379779650033, | |
| "grad_norm": 13.022201538085938, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 6.5698, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.01814646791963707, | |
| "grad_norm": 12.793331146240234, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 6.3188, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.018794556059624108, | |
| "grad_norm": 11.348553657531738, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 6.3723, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.019442644199611146, | |
| "grad_norm": 10.75669002532959, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 6.1807, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.020090732339598186, | |
| "grad_norm": 11.137311935424805, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 6.1366, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.020738820479585224, | |
| "grad_norm": 12.525829315185547, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 6.3884, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.02138690861957226, | |
| "grad_norm": 11.812135696411133, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 5.9797, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0220349967595593, | |
| "grad_norm": 12.773241996765137, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 6.0294, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.02268308489954634, | |
| "grad_norm": 13.393978118896484, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 6.6003, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.023331173039533377, | |
| "grad_norm": 13.857983589172363, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 6.0449, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.023979261179520414, | |
| "grad_norm": 13.792213439941406, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 6.1337, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.02462734931950745, | |
| "grad_norm": 14.81916618347168, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 6.3006, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.025275437459494492, | |
| "grad_norm": 16.306072235107422, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 6.673, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.02592352559948153, | |
| "grad_norm": 14.562975883483887, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 6.4593, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.026571613739468567, | |
| "grad_norm": 18.16436195373535, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 7.6702, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.027219701879455604, | |
| "grad_norm": 17.879405975341797, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 7.6658, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.027867790019442645, | |
| "grad_norm": 19.32358741760254, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 6.861, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.028515878159429683, | |
| "grad_norm": 32.35950469970703, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 9.8327, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.02916396629941672, | |
| "grad_norm": 18.36507797241211, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 5.6257, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.029812054439403757, | |
| "grad_norm": 13.660843849182129, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 4.965, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0304601425793908, | |
| "grad_norm": 10.093605041503906, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 4.6639, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.031108230719377836, | |
| "grad_norm": 10.09553337097168, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 4.754, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.031756318859364877, | |
| "grad_norm": 12.687553405761719, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 4.5943, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03240440699935191, | |
| "grad_norm": 10.109593391418457, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 3.9154, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03240440699935191, | |
| "eval_loss": 1.5928395986557007, | |
| "eval_runtime": 368.1229, | |
| "eval_samples_per_second": 7.06, | |
| "eval_steps_per_second": 1.766, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03305249513933895, | |
| "grad_norm": 6.980609893798828, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 6.867, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.033700583279325985, | |
| "grad_norm": 6.9170098304748535, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 6.5762, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.034348671419313026, | |
| "grad_norm": 6.112064361572266, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 6.7365, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.03499675955930007, | |
| "grad_norm": 6.1971635818481445, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 6.3913, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0356448476992871, | |
| "grad_norm": 6.6786017417907715, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 6.6206, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03629293583927414, | |
| "grad_norm": 6.822116374969482, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 6.2955, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.03694102397926118, | |
| "grad_norm": 6.868690013885498, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 6.1542, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.037589112119248216, | |
| "grad_norm": 7.442452907562256, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 6.2679, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.03823720025923526, | |
| "grad_norm": 7.3351006507873535, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 5.8176, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.03888528839922229, | |
| "grad_norm": 7.827886581420898, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 5.9771, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03953337653920933, | |
| "grad_norm": 7.912171363830566, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 5.7801, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.04018146467919637, | |
| "grad_norm": 7.832915782928467, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 5.5801, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.04082955281918341, | |
| "grad_norm": 8.083939552307129, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 6.0501, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.04147764095917045, | |
| "grad_norm": 7.929896831512451, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 5.9395, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.04212572909915749, | |
| "grad_norm": 7.770750045776367, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 5.2696, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04277381723914452, | |
| "grad_norm": 8.002516746520996, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 5.3783, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.04342190537913156, | |
| "grad_norm": 8.261298179626465, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 5.5322, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.0440699935191186, | |
| "grad_norm": 8.621538162231445, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 6.0108, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.04471808165910564, | |
| "grad_norm": 8.629108428955078, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 5.5071, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.04536616979909268, | |
| "grad_norm": 8.700213432312012, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 5.6949, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04601425793907971, | |
| "grad_norm": 8.888072967529297, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 5.2336, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.046662346079066754, | |
| "grad_norm": 9.403554916381836, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 5.8668, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.047310434219053794, | |
| "grad_norm": 9.120081901550293, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 6.0044, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.04795852235904083, | |
| "grad_norm": 9.56356143951416, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 5.7729, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.04860661049902787, | |
| "grad_norm": 9.18362045288086, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 5.7522, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.0492546986390149, | |
| "grad_norm": 9.9910249710083, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 5.6483, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.049902786779001944, | |
| "grad_norm": 11.102643966674805, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 5.4694, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.050550874918988985, | |
| "grad_norm": 9.472458839416504, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 5.0966, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.05119896305897602, | |
| "grad_norm": 9.36568832397461, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 5.2008, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.05184705119896306, | |
| "grad_norm": 10.136297225952148, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 5.381, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0524951393389501, | |
| "grad_norm": 9.870697021484375, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 5.6316, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.053143227478937134, | |
| "grad_norm": 8.686858177185059, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 5.0702, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.053791315618924175, | |
| "grad_norm": 11.320849418640137, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 6.1287, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.05443940375891121, | |
| "grad_norm": 10.530974388122559, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 5.5568, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.05508749189889825, | |
| "grad_norm": 9.906576156616211, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 5.4907, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.05573558003888529, | |
| "grad_norm": 11.5390625, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 5.6662, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.056383668178872325, | |
| "grad_norm": 11.390203475952148, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 5.6948, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.057031756318859365, | |
| "grad_norm": 14.094040870666504, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 6.0756, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.057679844458846406, | |
| "grad_norm": 12.248970031738281, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 5.9675, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.05832793259883344, | |
| "grad_norm": 11.868046760559082, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 5.9813, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05897602073882048, | |
| "grad_norm": 12.603425025939941, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 5.7393, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.059624108878807515, | |
| "grad_norm": 13.412906646728516, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 6.1246, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.060272197018794556, | |
| "grad_norm": 16.6093807220459, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 6.923, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.0609202851587816, | |
| "grad_norm": 17.576526641845703, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 5.9134, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.06156837329876863, | |
| "grad_norm": 11.265088081359863, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 4.6002, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06221646143875567, | |
| "grad_norm": 10.883245468139648, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 4.139, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.06286454957874271, | |
| "grad_norm": 9.871200561523438, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 4.9324, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.06351263771872975, | |
| "grad_norm": 7.893340110778809, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 4.3389, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.06416072585871678, | |
| "grad_norm": 9.363809585571289, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 4.5932, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.06480881399870382, | |
| "grad_norm": 9.364603042602539, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 4.3352, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06480881399870382, | |
| "eval_loss": 1.419034719467163, | |
| "eval_runtime": 367.8135, | |
| "eval_samples_per_second": 7.066, | |
| "eval_steps_per_second": 1.767, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06545690213869086, | |
| "grad_norm": 4.814598560333252, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 5.9494, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.0661049902786779, | |
| "grad_norm": 4.578697204589844, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 5.5302, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.06675307841866494, | |
| "grad_norm": 5.0061421394348145, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 5.5884, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.06740116655865197, | |
| "grad_norm": 5.34958553314209, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 5.775, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.06804925469863901, | |
| "grad_norm": 5.827226161956787, | |
| "learning_rate": 5e-05, | |
| "loss": 5.9442, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.06869734283862605, | |
| "grad_norm": 5.971827030181885, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 5.573, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.06934543097861309, | |
| "grad_norm": 6.491287708282471, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 5.302, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.06999351911860013, | |
| "grad_norm": 7.205959796905518, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 5.6267, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.07064160725858717, | |
| "grad_norm": 7.4618330001831055, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 5.4974, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.0712896953985742, | |
| "grad_norm": 7.016729831695557, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 5.6011, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07193778353856124, | |
| "grad_norm": 7.163577079772949, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 5.402, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.07258587167854828, | |
| "grad_norm": 6.554802894592285, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 5.0592, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.07323395981853532, | |
| "grad_norm": 7.247341156005859, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 5.3753, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.07388204795852236, | |
| "grad_norm": 7.453572750091553, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 5.6571, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.07453013609850939, | |
| "grad_norm": 7.110829830169678, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 5.1858, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.07517822423849643, | |
| "grad_norm": 7.949575424194336, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 5.8766, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.07582631237848347, | |
| "grad_norm": 8.235194206237793, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 5.2648, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.07647440051847051, | |
| "grad_norm": 7.894912242889404, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 5.6548, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.07712248865845756, | |
| "grad_norm": 7.492299556732178, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 5.0731, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.07777057679844458, | |
| "grad_norm": 7.998632431030273, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 5.5284, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07841866493843162, | |
| "grad_norm": 8.506697654724121, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 5.0284, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.07906675307841866, | |
| "grad_norm": 8.809338569641113, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 5.5977, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.0797148412184057, | |
| "grad_norm": 8.432890892028809, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 5.4425, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.08036292935839275, | |
| "grad_norm": 8.092004776000977, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 5.1623, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.08101101749837979, | |
| "grad_norm": 8.064069747924805, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 4.67, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08165910563836681, | |
| "grad_norm": 9.591204643249512, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 5.3424, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.08230719377835385, | |
| "grad_norm": 8.736778259277344, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 5.6868, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.0829552819183409, | |
| "grad_norm": 9.050400733947754, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 5.7636, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.08360337005832794, | |
| "grad_norm": 8.817517280578613, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 5.2625, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.08425145819831498, | |
| "grad_norm": 9.457335472106934, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 5.4587, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.084899546338302, | |
| "grad_norm": 8.814817428588867, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 5.1092, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.08554763447828904, | |
| "grad_norm": 9.131630897521973, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 4.4849, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.08619572261827609, | |
| "grad_norm": 10.105237007141113, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 5.6843, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.08684381075826313, | |
| "grad_norm": 8.991281509399414, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 4.762, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.08749189889825017, | |
| "grad_norm": 9.899235725402832, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 4.2467, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.0881399870382372, | |
| "grad_norm": 10.578429222106934, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 5.4836, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.08878807517822424, | |
| "grad_norm": 11.909724235534668, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 5.7676, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.08943616331821128, | |
| "grad_norm": 10.881829261779785, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 4.7597, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.09008425145819832, | |
| "grad_norm": 10.840633392333984, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 4.5822, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.09073233959818536, | |
| "grad_norm": 12.91557502746582, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 5.6312, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09138042773817238, | |
| "grad_norm": 16.27786636352539, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 6.1149, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.09202851587815943, | |
| "grad_norm": 13.686140060424805, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 5.5252, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.09267660401814647, | |
| "grad_norm": 20.249967575073242, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 7.0736, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.09332469215813351, | |
| "grad_norm": 11.816630363464355, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 5.1945, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.09397278029812055, | |
| "grad_norm": 9.47885799407959, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 4.6706, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.09462086843810759, | |
| "grad_norm": 7.617250919342041, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 3.5462, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.09526895657809462, | |
| "grad_norm": 7.987148761749268, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 3.8455, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.09591704471808166, | |
| "grad_norm": 7.913012981414795, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 3.5029, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.0965651328580687, | |
| "grad_norm": 7.453940391540527, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 3.257, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.09721322099805574, | |
| "grad_norm": 8.691618919372559, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 3.9776, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09721322099805574, | |
| "eval_loss": 1.3218529224395752, | |
| "eval_runtime": 368.0191, | |
| "eval_samples_per_second": 7.062, | |
| "eval_steps_per_second": 1.766, | |
| "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": 4.078450525392077e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |