| { | |
| "best_metric": 0.9860224723815918, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-50", | |
| "epoch": 3.0127659574468084, | |
| "eval_steps": 50, | |
| "global_step": 177, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01702127659574468, | |
| "grad_norm": 2.3956034183502197, | |
| "learning_rate": 1e-05, | |
| "loss": 1.5021, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01702127659574468, | |
| "eval_loss": 1.6579022407531738, | |
| "eval_runtime": 4.4292, | |
| "eval_samples_per_second": 22.352, | |
| "eval_steps_per_second": 5.644, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03404255319148936, | |
| "grad_norm": 2.8143370151519775, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4469, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05106382978723404, | |
| "grad_norm": 2.780684232711792, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5057, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06808510638297872, | |
| "grad_norm": 2.466012716293335, | |
| "learning_rate": 4e-05, | |
| "loss": 1.5455, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0851063829787234, | |
| "grad_norm": 2.0236899852752686, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3597, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10212765957446808, | |
| "grad_norm": 1.7500122785568237, | |
| "learning_rate": 6e-05, | |
| "loss": 1.3609, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.11914893617021277, | |
| "grad_norm": 2.0319266319274902, | |
| "learning_rate": 7e-05, | |
| "loss": 1.325, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.13617021276595745, | |
| "grad_norm": 1.9214025735855103, | |
| "learning_rate": 8e-05, | |
| "loss": 1.2035, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15319148936170213, | |
| "grad_norm": 1.8180416822433472, | |
| "learning_rate": 9e-05, | |
| "loss": 1.1698, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 1.722793459892273, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0561, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.18723404255319148, | |
| "grad_norm": 1.8683631420135498, | |
| "learning_rate": 9.999115304121457e-05, | |
| "loss": 1.1662, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.20425531914893616, | |
| "grad_norm": 2.0223588943481445, | |
| "learning_rate": 9.996461529560553e-05, | |
| "loss": 1.1261, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22127659574468084, | |
| "grad_norm": 1.9186768531799316, | |
| "learning_rate": 9.992039615430648e-05, | |
| "loss": 0.9619, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.23829787234042554, | |
| "grad_norm": 2.4184582233428955, | |
| "learning_rate": 9.985851126551428e-05, | |
| "loss": 1.1466, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2553191489361702, | |
| "grad_norm": 1.746878743171692, | |
| "learning_rate": 9.977898252895134e-05, | |
| "loss": 1.0855, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2723404255319149, | |
| "grad_norm": 1.4888535737991333, | |
| "learning_rate": 9.968183808811586e-05, | |
| "loss": 1.0404, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.28936170212765955, | |
| "grad_norm": 1.3668509721755981, | |
| "learning_rate": 9.95671123203224e-05, | |
| "loss": 1.1569, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.30638297872340425, | |
| "grad_norm": 1.2972848415374756, | |
| "learning_rate": 9.943484582453653e-05, | |
| "loss": 1.0638, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.32340425531914896, | |
| "grad_norm": 1.6263564825057983, | |
| "learning_rate": 9.928508540700774e-05, | |
| "loss": 1.1049, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3404255319148936, | |
| "grad_norm": 1.4598687887191772, | |
| "learning_rate": 9.911788406470569e-05, | |
| "loss": 1.0198, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3574468085106383, | |
| "grad_norm": 1.431780457496643, | |
| "learning_rate": 9.893330096656574e-05, | |
| "loss": 1.0425, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.37446808510638296, | |
| "grad_norm": 1.275739073753357, | |
| "learning_rate": 9.873140143255036e-05, | |
| "loss": 1.0508, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.39148936170212767, | |
| "grad_norm": 1.3290923833847046, | |
| "learning_rate": 9.85122569105338e-05, | |
| "loss": 1.0344, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4085106382978723, | |
| "grad_norm": 1.385582685470581, | |
| "learning_rate": 9.827594495101823e-05, | |
| "loss": 0.9878, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.425531914893617, | |
| "grad_norm": 1.3875707387924194, | |
| "learning_rate": 9.802254917969032e-05, | |
| "loss": 1.0025, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4425531914893617, | |
| "grad_norm": 1.3922779560089111, | |
| "learning_rate": 9.775215926782788e-05, | |
| "loss": 0.9845, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.4595744680851064, | |
| "grad_norm": 1.487823247909546, | |
| "learning_rate": 9.746487090056713e-05, | |
| "loss": 0.8108, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4765957446808511, | |
| "grad_norm": 2.1065125465393066, | |
| "learning_rate": 9.716078574304189e-05, | |
| "loss": 1.0918, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.49361702127659574, | |
| "grad_norm": 1.2851176261901855, | |
| "learning_rate": 9.684001140440639e-05, | |
| "loss": 1.1534, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5106382978723404, | |
| "grad_norm": 1.350062370300293, | |
| "learning_rate": 9.650266139975474e-05, | |
| "loss": 1.0002, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5276595744680851, | |
| "grad_norm": 1.263066291809082, | |
| "learning_rate": 9.614885510995047e-05, | |
| "loss": 1.0785, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5446808510638298, | |
| "grad_norm": 1.1791244745254517, | |
| "learning_rate": 9.577871773938011e-05, | |
| "loss": 1.0402, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5617021276595745, | |
| "grad_norm": 1.1864782571792603, | |
| "learning_rate": 9.539238027164619e-05, | |
| "loss": 1.091, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5787234042553191, | |
| "grad_norm": 1.2298381328582764, | |
| "learning_rate": 9.498997942321483e-05, | |
| "loss": 0.998, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5957446808510638, | |
| "grad_norm": 1.2237919569015503, | |
| "learning_rate": 9.457165759503493e-05, | |
| "loss": 0.9893, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6127659574468085, | |
| "grad_norm": 1.3423452377319336, | |
| "learning_rate": 9.413756282214537e-05, | |
| "loss": 0.955, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6297872340425532, | |
| "grad_norm": 1.3232065439224243, | |
| "learning_rate": 9.368784872128878e-05, | |
| "loss": 1.0207, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6468085106382979, | |
| "grad_norm": 1.2950204610824585, | |
| "learning_rate": 9.322267443654972e-05, | |
| "loss": 0.8997, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6638297872340425, | |
| "grad_norm": 1.3665379285812378, | |
| "learning_rate": 9.274220458303727e-05, | |
| "loss": 0.9627, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6808510638297872, | |
| "grad_norm": 1.5041331052780151, | |
| "learning_rate": 9.224660918863104e-05, | |
| "loss": 0.9353, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6978723404255319, | |
| "grad_norm": 1.484763264656067, | |
| "learning_rate": 9.173606363381219e-05, | |
| "loss": 0.9178, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7148936170212766, | |
| "grad_norm": 2.1842877864837646, | |
| "learning_rate": 9.121074858959997e-05, | |
| "loss": 0.912, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7319148936170212, | |
| "grad_norm": 1.146817922592163, | |
| "learning_rate": 9.067084995361623e-05, | |
| "loss": 1.0537, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7489361702127659, | |
| "grad_norm": 1.1362913846969604, | |
| "learning_rate": 9.011655878430019e-05, | |
| "loss": 0.9867, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7659574468085106, | |
| "grad_norm": 1.1847161054611206, | |
| "learning_rate": 8.954807123329704e-05, | |
| "loss": 1.0172, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7829787234042553, | |
| "grad_norm": 1.1899924278259277, | |
| "learning_rate": 8.896558847604414e-05, | |
| "loss": 1.0202, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.22841477394104, | |
| "learning_rate": 8.836931664057935e-05, | |
| "loss": 1.0795, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8170212765957446, | |
| "grad_norm": 1.1131497621536255, | |
| "learning_rate": 8.775946673459681e-05, | |
| "loss": 0.951, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8340425531914893, | |
| "grad_norm": 1.2665643692016602, | |
| "learning_rate": 8.713625457077585e-05, | |
| "loss": 0.936, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 1.2944315671920776, | |
| "learning_rate": 8.649990069040961e-05, | |
| "loss": 1.098, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "eval_loss": 0.9860224723815918, | |
| "eval_runtime": 4.4613, | |
| "eval_samples_per_second": 22.191, | |
| "eval_steps_per_second": 5.604, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8680851063829788, | |
| "grad_norm": 1.3318464756011963, | |
| "learning_rate": 8.585063028536016e-05, | |
| "loss": 0.9964, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8851063829787233, | |
| "grad_norm": 1.2892343997955322, | |
| "learning_rate": 8.518867311836808e-05, | |
| "loss": 0.8719, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.902127659574468, | |
| "grad_norm": 1.4797875881195068, | |
| "learning_rate": 8.451426344174433e-05, | |
| "loss": 1.0485, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.9191489361702128, | |
| "grad_norm": 1.351209044456482, | |
| "learning_rate": 8.382763991447344e-05, | |
| "loss": 0.8408, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9361702127659575, | |
| "grad_norm": 1.4917362928390503, | |
| "learning_rate": 8.312904551775731e-05, | |
| "loss": 0.9081, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9531914893617022, | |
| "grad_norm": 1.636767864227295, | |
| "learning_rate": 8.241872746902935e-05, | |
| "loss": 0.7751, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.9702127659574468, | |
| "grad_norm": 1.0549850463867188, | |
| "learning_rate": 8.169693713446959e-05, | |
| "loss": 0.9314, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9872340425531915, | |
| "grad_norm": 1.174147367477417, | |
| "learning_rate": 8.096392994005177e-05, | |
| "loss": 0.839, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.004255319148936, | |
| "grad_norm": 1.7790963649749756, | |
| "learning_rate": 8.021996528115335e-05, | |
| "loss": 1.0854, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0212765957446808, | |
| "grad_norm": 0.8404403924942017, | |
| "learning_rate": 7.946530643076138e-05, | |
| "loss": 0.791, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0382978723404255, | |
| "grad_norm": 0.9521711468696594, | |
| "learning_rate": 7.870022044630569e-05, | |
| "loss": 0.6483, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.0553191489361702, | |
| "grad_norm": 0.9962537288665771, | |
| "learning_rate": 7.792497807515317e-05, | |
| "loss": 0.7106, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0723404255319149, | |
| "grad_norm": 0.9780325293540955, | |
| "learning_rate": 7.713985365879606e-05, | |
| "loss": 0.6272, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.0893617021276596, | |
| "grad_norm": 0.9337742328643799, | |
| "learning_rate": 7.63451250357685e-05, | |
| "loss": 0.6399, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1063829787234043, | |
| "grad_norm": 0.9755467176437378, | |
| "learning_rate": 7.55410734433254e-05, | |
| "loss": 0.5694, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.123404255319149, | |
| "grad_norm": 1.0322693586349487, | |
| "learning_rate": 7.472798341791877e-05, | |
| "loss": 0.5253, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1404255319148937, | |
| "grad_norm": 1.1534591913223267, | |
| "learning_rate": 7.390614269450634e-05, | |
| "loss": 0.6757, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.1574468085106382, | |
| "grad_norm": 1.1756784915924072, | |
| "learning_rate": 7.307584210472844e-05, | |
| "loss": 0.5744, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.174468085106383, | |
| "grad_norm": 1.227107048034668, | |
| "learning_rate": 7.223737547398898e-05, | |
| "loss": 0.471, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.1914893617021276, | |
| "grad_norm": 1.4421412944793701, | |
| "learning_rate": 7.139103951747695e-05, | |
| "loss": 0.5671, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2085106382978723, | |
| "grad_norm": 1.5038673877716064, | |
| "learning_rate": 7.053713373516538e-05, | |
| "loss": 0.5429, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.225531914893617, | |
| "grad_norm": 1.6070609092712402, | |
| "learning_rate": 6.967596030582478e-05, | |
| "loss": 0.3801, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.2425531914893617, | |
| "grad_norm": 1.7779772281646729, | |
| "learning_rate": 6.880782398008862e-05, | |
| "loss": 0.6141, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.2595744680851064, | |
| "grad_norm": 1.2830954790115356, | |
| "learning_rate": 6.793303197260864e-05, | |
| "loss": 0.667, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.2765957446808511, | |
| "grad_norm": 1.3784350156784058, | |
| "learning_rate": 6.70518938533383e-05, | |
| "loss": 0.6144, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.2936170212765958, | |
| "grad_norm": 1.772112488746643, | |
| "learning_rate": 6.616472143798261e-05, | |
| "loss": 0.8042, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3106382978723405, | |
| "grad_norm": 1.6115304231643677, | |
| "learning_rate": 6.527182867765332e-05, | |
| "loss": 0.6911, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.327659574468085, | |
| "grad_norm": 1.4860713481903076, | |
| "learning_rate": 6.437353154776849e-05, | |
| "loss": 0.668, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.3446808510638297, | |
| "grad_norm": 1.3420848846435547, | |
| "learning_rate": 6.347014793623547e-05, | |
| "loss": 0.5545, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.3617021276595744, | |
| "grad_norm": 1.3645442724227905, | |
| "learning_rate": 6.256199753095745e-05, | |
| "loss": 0.6041, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3787234042553191, | |
| "grad_norm": 1.2304599285125732, | |
| "learning_rate": 6.164940170670266e-05, | |
| "loss": 0.4798, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.3957446808510638, | |
| "grad_norm": 1.1637616157531738, | |
| "learning_rate": 6.0732683411376935e-05, | |
| "loss": 0.4231, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.4127659574468086, | |
| "grad_norm": 1.2594172954559326, | |
| "learning_rate": 5.98121670517393e-05, | |
| "loss": 0.4704, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.4297872340425533, | |
| "grad_norm": 1.3521147966384888, | |
| "learning_rate": 5.8888178378601565e-05, | |
| "loss": 0.4922, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.4468085106382977, | |
| "grad_norm": 1.2671799659729004, | |
| "learning_rate": 5.796104437155213e-05, | |
| "loss": 0.3777, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.4638297872340424, | |
| "grad_norm": 1.4365724325180054, | |
| "learning_rate": 5.7031093123244925e-05, | |
| "loss": 0.4116, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.4808510638297872, | |
| "grad_norm": 1.3327573537826538, | |
| "learning_rate": 5.6098653723294604e-05, | |
| "loss": 0.4738, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.4978723404255319, | |
| "grad_norm": 1.1536500453948975, | |
| "learning_rate": 5.516405614181883e-05, | |
| "loss": 0.6458, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.5148936170212766, | |
| "grad_norm": 1.2099850177764893, | |
| "learning_rate": 5.4227631112668955e-05, | |
| "loss": 0.628, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.5319148936170213, | |
| "grad_norm": 1.3336706161499023, | |
| "learning_rate": 5.3289710016390535e-05, | |
| "loss": 0.6473, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.548936170212766, | |
| "grad_norm": 1.279130220413208, | |
| "learning_rate": 5.2350624762954884e-05, | |
| "loss": 0.6072, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.5659574468085107, | |
| "grad_norm": 1.2410911321640015, | |
| "learning_rate": 5.14107076743033e-05, | |
| "loss": 0.5237, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.5829787234042554, | |
| "grad_norm": 1.3553860187530518, | |
| "learning_rate": 5.047029136674563e-05, | |
| "loss": 0.6179, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.236146092414856, | |
| "learning_rate": 4.95297086332544e-05, | |
| "loss": 0.4814, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.6170212765957448, | |
| "grad_norm": 1.4804574251174927, | |
| "learning_rate": 4.858929232569671e-05, | |
| "loss": 0.5876, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6340425531914895, | |
| "grad_norm": 1.3778893947601318, | |
| "learning_rate": 4.7649375237045135e-05, | |
| "loss": 0.4875, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.6510638297872342, | |
| "grad_norm": 1.3673796653747559, | |
| "learning_rate": 4.671028998360947e-05, | |
| "loss": 0.4232, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.6680851063829787, | |
| "grad_norm": 1.4746342897415161, | |
| "learning_rate": 4.577236888733105e-05, | |
| "loss": 0.4221, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.6851063829787234, | |
| "grad_norm": 1.4111493825912476, | |
| "learning_rate": 4.483594385818118e-05, | |
| "loss": 0.4132, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.702127659574468, | |
| "grad_norm": 1.6582450866699219, | |
| "learning_rate": 4.39013462767054e-05, | |
| "loss": 0.3586, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.702127659574468, | |
| "eval_loss": 1.0457570552825928, | |
| "eval_runtime": 4.456, | |
| "eval_samples_per_second": 22.217, | |
| "eval_steps_per_second": 5.61, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7191489361702128, | |
| "grad_norm": 1.557518720626831, | |
| "learning_rate": 4.29689068767551e-05, | |
| "loss": 0.4448, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.7361702127659573, | |
| "grad_norm": 1.330199122428894, | |
| "learning_rate": 4.203895562844789e-05, | |
| "loss": 0.6906, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.753191489361702, | |
| "grad_norm": 1.3650050163269043, | |
| "learning_rate": 4.1111821621398446e-05, | |
| "loss": 0.572, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.7702127659574467, | |
| "grad_norm": 1.286249041557312, | |
| "learning_rate": 4.0187832948260705e-05, | |
| "loss": 0.5293, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.7872340425531914, | |
| "grad_norm": 1.4000760316848755, | |
| "learning_rate": 3.926731658862307e-05, | |
| "loss": 0.5973, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.804255319148936, | |
| "grad_norm": 1.450547456741333, | |
| "learning_rate": 3.835059829329735e-05, | |
| "loss": 0.5751, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.8212765957446808, | |
| "grad_norm": 1.4450291395187378, | |
| "learning_rate": 3.7438002469042565e-05, | |
| "loss": 0.49, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.8382978723404255, | |
| "grad_norm": 1.4160994291305542, | |
| "learning_rate": 3.6529852063764545e-05, | |
| "loss": 0.5344, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.8553191489361702, | |
| "grad_norm": 1.4834502935409546, | |
| "learning_rate": 3.562646845223153e-05, | |
| "loss": 0.5433, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.872340425531915, | |
| "grad_norm": 1.3811678886413574, | |
| "learning_rate": 3.4728171322346694e-05, | |
| "loss": 0.4107, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8893617021276596, | |
| "grad_norm": 1.3840794563293457, | |
| "learning_rate": 3.38352785620174e-05, | |
| "loss": 0.4713, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.9063829787234043, | |
| "grad_norm": 1.5129518508911133, | |
| "learning_rate": 3.29481061466617e-05, | |
| "loss": 0.4653, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.923404255319149, | |
| "grad_norm": 1.569916009902954, | |
| "learning_rate": 3.2066968027391374e-05, | |
| "loss": 0.3947, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.9404255319148938, | |
| "grad_norm": 1.7366918325424194, | |
| "learning_rate": 3.119217601991139e-05, | |
| "loss": 0.3459, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.9574468085106385, | |
| "grad_norm": 1.6125799417495728, | |
| "learning_rate": 3.0324039694175233e-05, | |
| "loss": 0.3711, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.974468085106383, | |
| "grad_norm": 1.3394767045974731, | |
| "learning_rate": 2.946286626483463e-05, | |
| "loss": 0.6046, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.9914893617021276, | |
| "grad_norm": 1.6290675401687622, | |
| "learning_rate": 2.8608960482523056e-05, | |
| "loss": 0.5723, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.008510638297872, | |
| "grad_norm": 1.4972840547561646, | |
| "learning_rate": 2.7762624526011038e-05, | |
| "loss": 0.5839, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.025531914893617, | |
| "grad_norm": 0.92827969789505, | |
| "learning_rate": 2.6924157895271563e-05, | |
| "loss": 0.414, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.0425531914893615, | |
| "grad_norm": 0.9881418943405151, | |
| "learning_rate": 2.6093857305493664e-05, | |
| "loss": 0.338, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0595744680851062, | |
| "grad_norm": 1.0140328407287598, | |
| "learning_rate": 2.5272016582081236e-05, | |
| "loss": 0.3692, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.076595744680851, | |
| "grad_norm": 1.0715774297714233, | |
| "learning_rate": 2.4458926556674615e-05, | |
| "loss": 0.3516, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0936170212765957, | |
| "grad_norm": 0.9538404941558838, | |
| "learning_rate": 2.3654874964231518e-05, | |
| "loss": 0.2303, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.1106382978723404, | |
| "grad_norm": 1.1230480670928955, | |
| "learning_rate": 2.2860146341203937e-05, | |
| "loss": 0.293, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "grad_norm": 1.1339483261108398, | |
| "learning_rate": 2.207502192484685e-05, | |
| "loss": 0.3407, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.1446808510638298, | |
| "grad_norm": 1.0703402757644653, | |
| "learning_rate": 2.1299779553694323e-05, | |
| "loss": 0.2584, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.1617021276595745, | |
| "grad_norm": 1.0961896181106567, | |
| "learning_rate": 2.053469356923865e-05, | |
| "loss": 0.2334, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.178723404255319, | |
| "grad_norm": 1.2202945947647095, | |
| "learning_rate": 1.978003471884665e-05, | |
| "loss": 0.2705, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.195744680851064, | |
| "grad_norm": 1.281968593597412, | |
| "learning_rate": 1.9036070059948252e-05, | |
| "loss": 0.2222, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.2127659574468086, | |
| "grad_norm": 1.530350685119629, | |
| "learning_rate": 1.8303062865530406e-05, | |
| "loss": 0.1816, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.2297872340425533, | |
| "grad_norm": 1.5826750993728638, | |
| "learning_rate": 1.7581272530970667e-05, | |
| "loss": 0.2032, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.246808510638298, | |
| "grad_norm": 1.333558440208435, | |
| "learning_rate": 1.6870954482242707e-05, | |
| "loss": 0.4571, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.2638297872340427, | |
| "grad_norm": 1.2566792964935303, | |
| "learning_rate": 1.6172360085526565e-05, | |
| "loss": 0.394, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.2808510638297874, | |
| "grad_norm": 1.2111644744873047, | |
| "learning_rate": 1.5485736558255697e-05, | |
| "loss": 0.3292, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.297872340425532, | |
| "grad_norm": 1.3459227085113525, | |
| "learning_rate": 1.4811326881631937e-05, | |
| "loss": 0.3263, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.3148936170212764, | |
| "grad_norm": 1.3436534404754639, | |
| "learning_rate": 1.4149369714639853e-05, | |
| "loss": 0.3349, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.331914893617021, | |
| "grad_norm": 1.4782233238220215, | |
| "learning_rate": 1.3500099309590397e-05, | |
| "loss": 0.3526, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.348936170212766, | |
| "grad_norm": 1.3280771970748901, | |
| "learning_rate": 1.2863745429224144e-05, | |
| "loss": 0.2997, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.3659574468085105, | |
| "grad_norm": 1.3815573453903198, | |
| "learning_rate": 1.2240533265403198e-05, | |
| "loss": 0.2855, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.382978723404255, | |
| "grad_norm": 1.4903278350830078, | |
| "learning_rate": 1.1630683359420652e-05, | |
| "loss": 0.2904, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.3255101442337036, | |
| "learning_rate": 1.103441152395588e-05, | |
| "loss": 0.1762, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.4170212765957446, | |
| "grad_norm": 1.3879741430282593, | |
| "learning_rate": 1.0451928766702979e-05, | |
| "loss": 0.2153, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.4340425531914893, | |
| "grad_norm": 1.2487590312957764, | |
| "learning_rate": 9.883441215699823e-06, | |
| "loss": 0.1699, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.451063829787234, | |
| "grad_norm": 1.142252802848816, | |
| "learning_rate": 9.329150046383772e-06, | |
| "loss": 0.1227, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.4680851063829787, | |
| "grad_norm": 1.8175076246261597, | |
| "learning_rate": 8.789251410400023e-06, | |
| "loss": 0.1298, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.4851063829787234, | |
| "grad_norm": 1.2175217866897583, | |
| "learning_rate": 8.263936366187824e-06, | |
| "loss": 0.3301, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.502127659574468, | |
| "grad_norm": 1.314565896987915, | |
| "learning_rate": 7.753390811368971e-06, | |
| "loss": 0.3832, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.519148936170213, | |
| "grad_norm": 1.393023133277893, | |
| "learning_rate": 7.257795416962753e-06, | |
| "loss": 0.331, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.5361702127659576, | |
| "grad_norm": 1.4550484418869019, | |
| "learning_rate": 6.777325563450282e-06, | |
| "loss": 0.3105, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.5531914893617023, | |
| "grad_norm": 1.555659294128418, | |
| "learning_rate": 6.312151278711237e-06, | |
| "loss": 0.3881, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.5531914893617023, | |
| "eval_loss": 1.250982403755188, | |
| "eval_runtime": 4.4645, | |
| "eval_samples_per_second": 22.175, | |
| "eval_steps_per_second": 5.6, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.570212765957447, | |
| "grad_norm": 1.4013265371322632, | |
| "learning_rate": 5.86243717785463e-06, | |
| "loss": 0.3459, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.5872340425531917, | |
| "grad_norm": 1.444057822227478, | |
| "learning_rate": 5.428342404965076e-06, | |
| "loss": 0.277, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.604255319148936, | |
| "grad_norm": 1.328897476196289, | |
| "learning_rate": 5.010020576785174e-06, | |
| "loss": 0.2324, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.621276595744681, | |
| "grad_norm": 1.4151270389556885, | |
| "learning_rate": 4.607619728353818e-06, | |
| "loss": 0.2435, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.6382978723404253, | |
| "grad_norm": 1.3859100341796875, | |
| "learning_rate": 4.221282260619891e-06, | |
| "loss": 0.2068, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.65531914893617, | |
| "grad_norm": 1.1932073831558228, | |
| "learning_rate": 3.851144890049535e-06, | |
| "loss": 0.156, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.6723404255319148, | |
| "grad_norm": 1.4522919654846191, | |
| "learning_rate": 3.4973386002452535e-06, | |
| "loss": 0.2544, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.6893617021276595, | |
| "grad_norm": 1.2598811388015747, | |
| "learning_rate": 3.159988595593616e-06, | |
| "loss": 0.1501, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.706382978723404, | |
| "grad_norm": 1.3632055521011353, | |
| "learning_rate": 2.839214256958106e-06, | |
| "loss": 0.1397, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.723404255319149, | |
| "grad_norm": 1.353302240371704, | |
| "learning_rate": 2.53512909943287e-06, | |
| "loss": 0.305, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.7404255319148936, | |
| "grad_norm": 1.339839220046997, | |
| "learning_rate": 2.2478407321721296e-06, | |
| "loss": 0.366, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.7574468085106383, | |
| "grad_norm": 1.3590214252471924, | |
| "learning_rate": 1.977450820309684e-06, | |
| "loss": 0.3311, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.774468085106383, | |
| "grad_norm": 1.4828448295593262, | |
| "learning_rate": 1.7240550489817653e-06, | |
| "loss": 0.3547, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.7914893617021277, | |
| "grad_norm": 1.3542424440383911, | |
| "learning_rate": 1.4877430894662036e-06, | |
| "loss": 0.2935, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.8085106382978724, | |
| "grad_norm": 1.304952621459961, | |
| "learning_rate": 1.268598567449647e-06, | |
| "loss": 0.2388, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.825531914893617, | |
| "grad_norm": 1.453139066696167, | |
| "learning_rate": 1.0666990334342707e-06, | |
| "loss": 0.2971, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.842553191489362, | |
| "grad_norm": 1.335195541381836, | |
| "learning_rate": 8.821159352943143e-07, | |
| "loss": 0.2004, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.8595744680851065, | |
| "grad_norm": 1.482730746269226, | |
| "learning_rate": 7.149145929922607e-07, | |
| "loss": 0.2955, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.876595744680851, | |
| "grad_norm": 1.3346716165542603, | |
| "learning_rate": 5.651541754634726e-07, | |
| "loss": 0.2225, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.8936170212765955, | |
| "grad_norm": 1.362627625465393, | |
| "learning_rate": 4.3288767967760715e-07, | |
| "loss": 0.1993, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.9106382978723406, | |
| "grad_norm": 1.4417704343795776, | |
| "learning_rate": 3.1816191188415166e-07, | |
| "loss": 0.2222, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.927659574468085, | |
| "grad_norm": 1.2695422172546387, | |
| "learning_rate": 2.2101747104866788e-07, | |
| "loss": 0.1648, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.94468085106383, | |
| "grad_norm": 1.325673222541809, | |
| "learning_rate": 1.4148873448573408e-07, | |
| "loss": 0.1262, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.9617021276595743, | |
| "grad_norm": 1.330980896949768, | |
| "learning_rate": 7.960384569353219e-08, | |
| "loss": 0.2392, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.978723404255319, | |
| "grad_norm": 1.4023689031600952, | |
| "learning_rate": 3.538470439448105e-08, | |
| "loss": 0.2823, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.9957446808510637, | |
| "grad_norm": 1.465209722518921, | |
| "learning_rate": 8.846958785418968e-09, | |
| "loss": 0.192, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 3.0127659574468084, | |
| "grad_norm": 1.1684350967407227, | |
| "learning_rate": 0.0, | |
| "loss": 0.345, | |
| "step": 177 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 177, | |
| "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": 2 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.334314061384909e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |