| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "global_step": 1334, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0014992503748125937, | |
| "grad_norm": 14.650150621211907, | |
| "learning_rate": 9.999986134743949e-06, | |
| "loss": 1.2612, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0029985007496251873, | |
| "grad_norm": 14.718272159490025, | |
| "learning_rate": 9.99994453905269e-06, | |
| "loss": 1.2805, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.004497751124437781, | |
| "grad_norm": 2.7348748215022742, | |
| "learning_rate": 9.999875213156919e-06, | |
| "loss": 0.709, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.005997001499250375, | |
| "grad_norm": 2.8331166974967634, | |
| "learning_rate": 9.999778157441126e-06, | |
| "loss": 0.74, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0074962518740629685, | |
| "grad_norm": 1.9531736277871472, | |
| "learning_rate": 9.99965337244359e-06, | |
| "loss": 0.6366, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.008995502248875561, | |
| "grad_norm": 1.8664483197130723, | |
| "learning_rate": 9.999500858856382e-06, | |
| "loss": 0.6812, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.010494752623688156, | |
| "grad_norm": 1.9824005258592772, | |
| "learning_rate": 9.999320617525356e-06, | |
| "loss": 0.6675, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01199400299850075, | |
| "grad_norm": 1.5481376464156096, | |
| "learning_rate": 9.999112649450154e-06, | |
| "loss": 0.6443, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.013493253373313344, | |
| "grad_norm": 1.2720285490512377, | |
| "learning_rate": 9.998876955784183e-06, | |
| "loss": 0.4539, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.014992503748125937, | |
| "grad_norm": 1.3001321942677102, | |
| "learning_rate": 9.998613537834625e-06, | |
| "loss": 0.5368, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01649175412293853, | |
| "grad_norm": 1.5727537551116937, | |
| "learning_rate": 9.998322397062426e-06, | |
| "loss": 0.5702, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.017991004497751123, | |
| "grad_norm": 1.3246173194776407, | |
| "learning_rate": 9.99800353508228e-06, | |
| "loss": 0.4873, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.019490254872563718, | |
| "grad_norm": 1.2463284229299652, | |
| "learning_rate": 9.997656953662627e-06, | |
| "loss": 0.4179, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.020989505247376312, | |
| "grad_norm": 1.5223723976676033, | |
| "learning_rate": 9.997282654725645e-06, | |
| "loss": 0.5615, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.022488755622188907, | |
| "grad_norm": 1.2959934838448572, | |
| "learning_rate": 9.996880640347234e-06, | |
| "loss": 0.4785, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0239880059970015, | |
| "grad_norm": 1.3008567496719285, | |
| "learning_rate": 9.99645091275701e-06, | |
| "loss": 0.4061, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.025487256371814093, | |
| "grad_norm": 1.3847729451207946, | |
| "learning_rate": 9.99599347433828e-06, | |
| "loss": 0.5056, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.026986506746626688, | |
| "grad_norm": 1.3101936719851954, | |
| "learning_rate": 9.99550832762805e-06, | |
| "loss": 0.4843, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.02848575712143928, | |
| "grad_norm": 1.2077163692481527, | |
| "learning_rate": 9.99499547531699e-06, | |
| "loss": 0.3956, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.029985007496251874, | |
| "grad_norm": 1.5434403938270078, | |
| "learning_rate": 9.994454920249433e-06, | |
| "loss": 0.4781, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.031484257871064465, | |
| "grad_norm": 1.4242913656620275, | |
| "learning_rate": 9.993886665423348e-06, | |
| "loss": 0.4932, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03298350824587706, | |
| "grad_norm": 1.1759890716066088, | |
| "learning_rate": 9.993290713990343e-06, | |
| "loss": 0.458, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.034482758620689655, | |
| "grad_norm": 1.0835044343912597, | |
| "learning_rate": 9.99266706925562e-06, | |
| "loss": 0.451, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.035982008995502246, | |
| "grad_norm": 1.3837330036676712, | |
| "learning_rate": 9.992015734677979e-06, | |
| "loss": 0.4129, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.037481259370314844, | |
| "grad_norm": 1.1734482811662261, | |
| "learning_rate": 9.991336713869785e-06, | |
| "loss": 0.4348, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.038980509745127435, | |
| "grad_norm": 0.9554580906720362, | |
| "learning_rate": 9.99063001059696e-06, | |
| "loss": 0.363, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04047976011994003, | |
| "grad_norm": 1.1475370450417492, | |
| "learning_rate": 9.989895628778952e-06, | |
| "loss": 0.4313, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.041979010494752625, | |
| "grad_norm": 1.1509292693175759, | |
| "learning_rate": 9.989133572488716e-06, | |
| "loss": 0.3571, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.043478260869565216, | |
| "grad_norm": 1.1976936721418472, | |
| "learning_rate": 9.988343845952697e-06, | |
| "loss": 0.3824, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.044977511244377814, | |
| "grad_norm": 1.0169234473249682, | |
| "learning_rate": 9.987526453550798e-06, | |
| "loss": 0.3566, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.046476761619190406, | |
| "grad_norm": 1.0643621154655873, | |
| "learning_rate": 9.98668139981636e-06, | |
| "loss": 0.3471, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.047976011994003, | |
| "grad_norm": 1.2615161993208366, | |
| "learning_rate": 9.98580868943614e-06, | |
| "loss": 0.4406, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.049475262368815595, | |
| "grad_norm": 0.9741805517568367, | |
| "learning_rate": 9.984908327250278e-06, | |
| "loss": 0.3263, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.050974512743628186, | |
| "grad_norm": 0.9140520627947049, | |
| "learning_rate": 9.983980318252274e-06, | |
| "loss": 0.326, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.05247376311844078, | |
| "grad_norm": 0.910715926688195, | |
| "learning_rate": 9.983024667588961e-06, | |
| "loss": 0.3915, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.053973013493253376, | |
| "grad_norm": 0.959111339786902, | |
| "learning_rate": 9.982041380560476e-06, | |
| "loss": 0.3637, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.05547226386806597, | |
| "grad_norm": 1.0644027738284538, | |
| "learning_rate": 9.98103046262023e-06, | |
| "loss": 0.3602, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.05697151424287856, | |
| "grad_norm": 1.2170210623405626, | |
| "learning_rate": 9.979991919374877e-06, | |
| "loss": 0.4039, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.05847076461769116, | |
| "grad_norm": 1.2193673016294835, | |
| "learning_rate": 9.978925756584284e-06, | |
| "loss": 0.4343, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.05997001499250375, | |
| "grad_norm": 1.0695509968353465, | |
| "learning_rate": 9.9778319801615e-06, | |
| "loss": 0.3138, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06146926536731634, | |
| "grad_norm": 1.0932796494729822, | |
| "learning_rate": 9.976710596172721e-06, | |
| "loss": 0.4047, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.06296851574212893, | |
| "grad_norm": 1.09254100051343, | |
| "learning_rate": 9.975561610837254e-06, | |
| "loss": 0.3395, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.06446776611694154, | |
| "grad_norm": 1.0054184914430366, | |
| "learning_rate": 9.974385030527496e-06, | |
| "loss": 0.3351, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.06596701649175413, | |
| "grad_norm": 0.9874309966991488, | |
| "learning_rate": 9.973180861768874e-06, | |
| "loss": 0.3205, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.06746626686656672, | |
| "grad_norm": 1.186611625008614, | |
| "learning_rate": 9.971949111239838e-06, | |
| "loss": 0.3991, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06896551724137931, | |
| "grad_norm": 0.9983986056512062, | |
| "learning_rate": 9.970689785771798e-06, | |
| "loss": 0.3923, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0704647676161919, | |
| "grad_norm": 0.9370497182929065, | |
| "learning_rate": 9.969402892349105e-06, | |
| "loss": 0.3659, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.07196401799100449, | |
| "grad_norm": 0.9720626651169325, | |
| "learning_rate": 9.968088438109002e-06, | |
| "loss": 0.2819, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0734632683658171, | |
| "grad_norm": 1.1938762167990808, | |
| "learning_rate": 9.966746430341584e-06, | |
| "loss": 0.3822, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.07496251874062969, | |
| "grad_norm": 1.3992127298889347, | |
| "learning_rate": 9.965376876489765e-06, | |
| "loss": 0.4045, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07646176911544228, | |
| "grad_norm": 0.9969362514387262, | |
| "learning_rate": 9.963979784149232e-06, | |
| "loss": 0.3956, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.07796101949025487, | |
| "grad_norm": 1.1132128368826562, | |
| "learning_rate": 9.962555161068401e-06, | |
| "loss": 0.3731, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.07946026986506746, | |
| "grad_norm": 1.0745222254116487, | |
| "learning_rate": 9.961103015148376e-06, | |
| "loss": 0.3621, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.08095952023988005, | |
| "grad_norm": 0.8841466092235978, | |
| "learning_rate": 9.95962335444291e-06, | |
| "loss": 0.3026, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.08245877061469266, | |
| "grad_norm": 1.1328649437888036, | |
| "learning_rate": 9.958116187158351e-06, | |
| "loss": 0.3424, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08395802098950525, | |
| "grad_norm": 1.0618470758210738, | |
| "learning_rate": 9.956581521653604e-06, | |
| "loss": 0.3984, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.08545727136431784, | |
| "grad_norm": 1.2358730728487723, | |
| "learning_rate": 9.955019366440082e-06, | |
| "loss": 0.3771, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.08695652173913043, | |
| "grad_norm": 1.0593316519905622, | |
| "learning_rate": 9.953429730181653e-06, | |
| "loss": 0.4193, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.08845577211394302, | |
| "grad_norm": 1.4417482958246086, | |
| "learning_rate": 9.95181262169461e-06, | |
| "loss": 0.2696, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.08995502248875563, | |
| "grad_norm": 0.9734052099108674, | |
| "learning_rate": 9.950168049947597e-06, | |
| "loss": 0.3271, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09145427286356822, | |
| "grad_norm": 1.032447358575922, | |
| "learning_rate": 9.948496024061577e-06, | |
| "loss": 0.3554, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.09295352323838081, | |
| "grad_norm": 1.1034624475806982, | |
| "learning_rate": 9.94679655330978e-06, | |
| "loss": 0.4201, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.0944527736131934, | |
| "grad_norm": 1.1232282720994993, | |
| "learning_rate": 9.945069647117645e-06, | |
| "loss": 0.3397, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.095952023988006, | |
| "grad_norm": 1.192492167067343, | |
| "learning_rate": 9.943315315062766e-06, | |
| "loss": 0.4092, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.09745127436281859, | |
| "grad_norm": 0.7959940976444196, | |
| "learning_rate": 9.941533566874852e-06, | |
| "loss": 0.2383, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.09895052473763119, | |
| "grad_norm": 1.0320691157670445, | |
| "learning_rate": 9.939724412435661e-06, | |
| "loss": 0.3969, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.10044977511244378, | |
| "grad_norm": 0.7501563155876843, | |
| "learning_rate": 9.937887861778947e-06, | |
| "loss": 0.2627, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.10194902548725637, | |
| "grad_norm": 1.0436000060265447, | |
| "learning_rate": 9.93602392509041e-06, | |
| "loss": 0.3481, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.10344827586206896, | |
| "grad_norm": 1.1938122370500999, | |
| "learning_rate": 9.934132612707631e-06, | |
| "loss": 0.3421, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.10494752623688156, | |
| "grad_norm": 1.0612328923151821, | |
| "learning_rate": 9.932213935120025e-06, | |
| "loss": 0.3256, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10644677661169415, | |
| "grad_norm": 0.8552762526957014, | |
| "learning_rate": 9.930267902968774e-06, | |
| "loss": 0.2896, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.10794602698650675, | |
| "grad_norm": 1.295880099179343, | |
| "learning_rate": 9.928294527046771e-06, | |
| "loss": 0.4506, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.10944527736131934, | |
| "grad_norm": 1.1551067849789292, | |
| "learning_rate": 9.92629381829856e-06, | |
| "loss": 0.3629, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.11094452773613193, | |
| "grad_norm": 1.2491412084812963, | |
| "learning_rate": 9.924265787820279e-06, | |
| "loss": 0.4711, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.11244377811094453, | |
| "grad_norm": 1.0741251287333227, | |
| "learning_rate": 9.92221044685959e-06, | |
| "loss": 0.3532, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11394302848575712, | |
| "grad_norm": 1.0633613566015792, | |
| "learning_rate": 9.920127806815627e-06, | |
| "loss": 0.3599, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.11544227886056972, | |
| "grad_norm": 0.8518202053744102, | |
| "learning_rate": 9.918017879238922e-06, | |
| "loss": 0.342, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.11694152923538231, | |
| "grad_norm": 0.9162981187033958, | |
| "learning_rate": 9.915880675831352e-06, | |
| "loss": 0.3037, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.1184407796101949, | |
| "grad_norm": 1.1882593724719046, | |
| "learning_rate": 9.913716208446067e-06, | |
| "loss": 0.4347, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1199400299850075, | |
| "grad_norm": 1.0705128944814364, | |
| "learning_rate": 9.91152448908742e-06, | |
| "loss": 0.3368, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12143928035982009, | |
| "grad_norm": 1.0048418197640052, | |
| "learning_rate": 9.909305529910917e-06, | |
| "loss": 0.299, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.12293853073463268, | |
| "grad_norm": 1.1043592063512007, | |
| "learning_rate": 9.907059343223129e-06, | |
| "loss": 0.3786, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.12443778110944528, | |
| "grad_norm": 0.8166202902829717, | |
| "learning_rate": 9.904785941481638e-06, | |
| "loss": 0.2794, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.12593703148425786, | |
| "grad_norm": 0.9738568384008331, | |
| "learning_rate": 9.902485337294965e-06, | |
| "loss": 0.2969, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.12743628185907047, | |
| "grad_norm": 1.0022675842663873, | |
| "learning_rate": 9.900157543422493e-06, | |
| "loss": 0.2785, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.12893553223388307, | |
| "grad_norm": 1.0779162380203724, | |
| "learning_rate": 9.897802572774407e-06, | |
| "loss": 0.3362, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.13043478260869565, | |
| "grad_norm": 0.9479511856705523, | |
| "learning_rate": 9.895420438411616e-06, | |
| "loss": 0.3077, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.13193403298350825, | |
| "grad_norm": 0.8745186676362705, | |
| "learning_rate": 9.893011153545679e-06, | |
| "loss": 0.2721, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.13343328335832083, | |
| "grad_norm": 0.9628356951865044, | |
| "learning_rate": 9.89057473153874e-06, | |
| "loss": 0.3385, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.13493253373313344, | |
| "grad_norm": 1.0483952643024521, | |
| "learning_rate": 9.888111185903442e-06, | |
| "loss": 0.3586, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.136431784107946, | |
| "grad_norm": 1.1983399037897031, | |
| "learning_rate": 9.885620530302865e-06, | |
| "loss": 0.3702, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 1.053597655103566, | |
| "learning_rate": 9.883102778550434e-06, | |
| "loss": 0.3422, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.13943028485757122, | |
| "grad_norm": 0.9907573436968898, | |
| "learning_rate": 9.880557944609863e-06, | |
| "loss": 0.2819, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.1409295352323838, | |
| "grad_norm": 1.0824573725559639, | |
| "learning_rate": 9.877986042595062e-06, | |
| "loss": 0.3107, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.1424287856071964, | |
| "grad_norm": 1.1417152041563616, | |
| "learning_rate": 9.87538708677006e-06, | |
| "loss": 0.3463, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.14392803598200898, | |
| "grad_norm": 1.164406567105473, | |
| "learning_rate": 9.872761091548933e-06, | |
| "loss": 0.4104, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.1454272863568216, | |
| "grad_norm": 1.0734730522391762, | |
| "learning_rate": 9.870108071495721e-06, | |
| "loss": 0.3745, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.1469265367316342, | |
| "grad_norm": 1.0080181511400739, | |
| "learning_rate": 9.867428041324345e-06, | |
| "loss": 0.3595, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.14842578710644677, | |
| "grad_norm": 1.0945855564299365, | |
| "learning_rate": 9.864721015898524e-06, | |
| "loss": 0.2901, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.14992503748125938, | |
| "grad_norm": 1.0595765882776278, | |
| "learning_rate": 9.861987010231701e-06, | |
| "loss": 0.3976, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15142428785607195, | |
| "grad_norm": 1.1058748373076772, | |
| "learning_rate": 9.85922603948695e-06, | |
| "loss": 0.3245, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.15292353823088456, | |
| "grad_norm": 1.0156542328293112, | |
| "learning_rate": 9.856438118976899e-06, | |
| "loss": 0.3423, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.15442278860569716, | |
| "grad_norm": 0.9033818220286804, | |
| "learning_rate": 9.853623264163638e-06, | |
| "loss": 0.2694, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.15592203898050974, | |
| "grad_norm": 0.9515363877732105, | |
| "learning_rate": 9.850781490658643e-06, | |
| "loss": 0.3171, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.15742128935532235, | |
| "grad_norm": 0.9743784586244435, | |
| "learning_rate": 9.84791281422268e-06, | |
| "loss": 0.3191, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.15892053973013492, | |
| "grad_norm": 1.029788588687942, | |
| "learning_rate": 9.845017250765721e-06, | |
| "loss": 0.3811, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.16041979010494753, | |
| "grad_norm": 0.918379969634054, | |
| "learning_rate": 9.84209481634686e-06, | |
| "loss": 0.3368, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1619190404797601, | |
| "grad_norm": 0.8458428594653454, | |
| "learning_rate": 9.839145527174216e-06, | |
| "loss": 0.2859, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.1634182908545727, | |
| "grad_norm": 0.9265237133977762, | |
| "learning_rate": 9.836169399604846e-06, | |
| "loss": 0.2859, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.16491754122938532, | |
| "grad_norm": 1.1492518804004745, | |
| "learning_rate": 9.833166450144665e-06, | |
| "loss": 0.2985, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1664167916041979, | |
| "grad_norm": 1.0447258527528247, | |
| "learning_rate": 9.830136695448334e-06, | |
| "loss": 0.3018, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.1679160419790105, | |
| "grad_norm": 1.0273861664255484, | |
| "learning_rate": 9.827080152319182e-06, | |
| "loss": 0.3494, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.16941529235382308, | |
| "grad_norm": 1.1706520108140446, | |
| "learning_rate": 9.823996837709114e-06, | |
| "loss": 0.3295, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.17091454272863568, | |
| "grad_norm": 0.9853551969840236, | |
| "learning_rate": 9.820886768718503e-06, | |
| "loss": 0.347, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 0.9597749168175927, | |
| "learning_rate": 9.817749962596115e-06, | |
| "loss": 0.3477, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 1.022962888935491, | |
| "learning_rate": 9.814586436738998e-06, | |
| "loss": 0.3442, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.17541229385307347, | |
| "grad_norm": 1.1247773153414995, | |
| "learning_rate": 9.811396208692387e-06, | |
| "loss": 0.336, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.17691154422788605, | |
| "grad_norm": 1.071737078895673, | |
| "learning_rate": 9.808179296149616e-06, | |
| "loss": 0.3123, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.17841079460269865, | |
| "grad_norm": 1.013520489017317, | |
| "learning_rate": 9.804935716952011e-06, | |
| "loss": 0.3407, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.17991004497751126, | |
| "grad_norm": 1.015007610034657, | |
| "learning_rate": 9.801665489088795e-06, | |
| "loss": 0.347, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.18140929535232383, | |
| "grad_norm": 1.1667443746175639, | |
| "learning_rate": 9.798368630696984e-06, | |
| "loss": 0.3419, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.18290854572713644, | |
| "grad_norm": 1.0132978289367176, | |
| "learning_rate": 9.795045160061295e-06, | |
| "loss": 0.2994, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.18440779610194902, | |
| "grad_norm": 1.0161497144949958, | |
| "learning_rate": 9.791695095614036e-06, | |
| "loss": 0.333, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.18590704647676162, | |
| "grad_norm": 0.9491420230436132, | |
| "learning_rate": 9.788318455935008e-06, | |
| "loss": 0.2994, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.1874062968515742, | |
| "grad_norm": 1.043875734679698, | |
| "learning_rate": 9.7849152597514e-06, | |
| "loss": 0.2965, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1889055472263868, | |
| "grad_norm": 0.9835918851905867, | |
| "learning_rate": 9.781485525937683e-06, | |
| "loss": 0.2738, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.1904047976011994, | |
| "grad_norm": 1.3590065547877699, | |
| "learning_rate": 9.778029273515519e-06, | |
| "loss": 0.2848, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.191904047976012, | |
| "grad_norm": 0.9566489047307987, | |
| "learning_rate": 9.774546521653633e-06, | |
| "loss": 0.3811, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.1934032983508246, | |
| "grad_norm": 0.96342609319082, | |
| "learning_rate": 9.771037289667726e-06, | |
| "loss": 0.2858, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.19490254872563717, | |
| "grad_norm": 1.2347212623614718, | |
| "learning_rate": 9.767501597020357e-06, | |
| "loss": 0.3256, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.19640179910044978, | |
| "grad_norm": 1.099566121895579, | |
| "learning_rate": 9.76393946332084e-06, | |
| "loss": 0.3307, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.19790104947526238, | |
| "grad_norm": 1.155840447729321, | |
| "learning_rate": 9.760350908325131e-06, | |
| "loss": 0.3176, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.19940029985007496, | |
| "grad_norm": 0.9553386305176575, | |
| "learning_rate": 9.756735951935725e-06, | |
| "loss": 0.2966, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.20089955022488756, | |
| "grad_norm": 1.0910432721038266, | |
| "learning_rate": 9.753094614201542e-06, | |
| "loss": 0.2794, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.20239880059970014, | |
| "grad_norm": 1.0001802619655082, | |
| "learning_rate": 9.749426915317812e-06, | |
| "loss": 0.3198, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.20389805097451275, | |
| "grad_norm": 1.0078055050561296, | |
| "learning_rate": 9.74573287562597e-06, | |
| "loss": 0.3909, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.20539730134932535, | |
| "grad_norm": 0.8684548555291881, | |
| "learning_rate": 9.742012515613536e-06, | |
| "loss": 0.2787, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.20689655172413793, | |
| "grad_norm": 0.8075094897018598, | |
| "learning_rate": 9.738265855914014e-06, | |
| "loss": 0.2955, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.20839580209895053, | |
| "grad_norm": 0.9931978175670576, | |
| "learning_rate": 9.734492917306754e-06, | |
| "loss": 0.3678, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.2098950524737631, | |
| "grad_norm": 0.8586308883928482, | |
| "learning_rate": 9.730693720716866e-06, | |
| "loss": 0.3252, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21139430284857572, | |
| "grad_norm": 1.0276652792727465, | |
| "learning_rate": 9.72686828721508e-06, | |
| "loss": 0.3291, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2128935532233883, | |
| "grad_norm": 1.0300000160559564, | |
| "learning_rate": 9.723016638017644e-06, | |
| "loss": 0.3581, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.2143928035982009, | |
| "grad_norm": 1.0785623395472599, | |
| "learning_rate": 9.719138794486198e-06, | |
| "loss": 0.3079, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2158920539730135, | |
| "grad_norm": 0.9635318716477408, | |
| "learning_rate": 9.715234778127658e-06, | |
| "loss": 0.2425, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.21739130434782608, | |
| "grad_norm": 1.0613644952003296, | |
| "learning_rate": 9.711304610594104e-06, | |
| "loss": 0.3058, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.21889055472263869, | |
| "grad_norm": 1.1109237147553006, | |
| "learning_rate": 9.70734831368264e-06, | |
| "loss": 0.3783, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.22038980509745126, | |
| "grad_norm": 0.9779909553289717, | |
| "learning_rate": 9.7033659093353e-06, | |
| "loss": 0.2669, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.22188905547226387, | |
| "grad_norm": 1.408549060190886, | |
| "learning_rate": 9.699357419638904e-06, | |
| "loss": 0.4111, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.22338830584707647, | |
| "grad_norm": 0.926451706820931, | |
| "learning_rate": 9.695322866824948e-06, | |
| "loss": 0.2713, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.22488755622188905, | |
| "grad_norm": 1.1812022466562178, | |
| "learning_rate": 9.691262273269472e-06, | |
| "loss": 0.3404, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.22638680659670166, | |
| "grad_norm": 1.0011358403593225, | |
| "learning_rate": 9.687175661492944e-06, | |
| "loss": 0.3079, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.22788605697151423, | |
| "grad_norm": 0.8013136115197694, | |
| "learning_rate": 9.683063054160136e-06, | |
| "loss": 0.2101, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.22938530734632684, | |
| "grad_norm": 1.0946385218820338, | |
| "learning_rate": 9.678924474079986e-06, | |
| "loss": 0.3363, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.23088455772113944, | |
| "grad_norm": 1.0073689784900812, | |
| "learning_rate": 9.67475994420548e-06, | |
| "loss": 0.3215, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.23238380809595202, | |
| "grad_norm": 1.1625959346555566, | |
| "learning_rate": 9.670569487633534e-06, | |
| "loss": 0.3284, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.23388305847076463, | |
| "grad_norm": 1.0014988973127323, | |
| "learning_rate": 9.666353127604845e-06, | |
| "loss": 0.3772, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2353823088455772, | |
| "grad_norm": 0.9374125532313022, | |
| "learning_rate": 9.66211088750378e-06, | |
| "loss": 0.292, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.2368815592203898, | |
| "grad_norm": 1.0435222620064046, | |
| "learning_rate": 9.657842790858235e-06, | |
| "loss": 0.32, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.2383808095952024, | |
| "grad_norm": 1.387060801781273, | |
| "learning_rate": 9.65354886133951e-06, | |
| "loss": 0.412, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.239880059970015, | |
| "grad_norm": 1.0009404680533003, | |
| "learning_rate": 9.64922912276218e-06, | |
| "loss": 0.3087, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2413793103448276, | |
| "grad_norm": 1.053627433229224, | |
| "learning_rate": 9.644883599083959e-06, | |
| "loss": 0.3145, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.24287856071964017, | |
| "grad_norm": 1.0542621843077424, | |
| "learning_rate": 9.640512314405563e-06, | |
| "loss": 0.2977, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.24437781109445278, | |
| "grad_norm": 1.0854305161828657, | |
| "learning_rate": 9.636115292970587e-06, | |
| "loss": 0.3614, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.24587706146926536, | |
| "grad_norm": 1.1278786141979982, | |
| "learning_rate": 9.63169255916536e-06, | |
| "loss": 0.3033, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.24737631184407796, | |
| "grad_norm": 0.9521502445551386, | |
| "learning_rate": 9.627244137518821e-06, | |
| "loss": 0.2638, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.24887556221889057, | |
| "grad_norm": 1.0519786565062144, | |
| "learning_rate": 9.622770052702366e-06, | |
| "loss": 0.3058, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.25037481259370314, | |
| "grad_norm": 0.927140826803852, | |
| "learning_rate": 9.618270329529734e-06, | |
| "loss": 0.2616, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.2518740629685157, | |
| "grad_norm": 0.933752312753363, | |
| "learning_rate": 9.613744992956844e-06, | |
| "loss": 0.2722, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.25337331334332835, | |
| "grad_norm": 0.8576493685641376, | |
| "learning_rate": 9.609194068081682e-06, | |
| "loss": 0.2667, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.25487256371814093, | |
| "grad_norm": 1.1651436510547246, | |
| "learning_rate": 9.60461758014414e-06, | |
| "loss": 0.3329, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2563718140929535, | |
| "grad_norm": 0.9907409171288238, | |
| "learning_rate": 9.60001555452589e-06, | |
| "loss": 0.3278, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.25787106446776614, | |
| "grad_norm": 1.1007424842842195, | |
| "learning_rate": 9.595388016750236e-06, | |
| "loss": 0.306, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.2593703148425787, | |
| "grad_norm": 1.1051260763018076, | |
| "learning_rate": 9.590734992481978e-06, | |
| "loss": 0.2672, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 0.945367920439993, | |
| "learning_rate": 9.586056507527266e-06, | |
| "loss": 0.354, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2623688155922039, | |
| "grad_norm": 1.1623459551411868, | |
| "learning_rate": 9.581352587833455e-06, | |
| "loss": 0.3161, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2638680659670165, | |
| "grad_norm": 1.2801475675828409, | |
| "learning_rate": 9.576623259488966e-06, | |
| "loss": 0.4023, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.2653673163418291, | |
| "grad_norm": 1.260513630358247, | |
| "learning_rate": 9.571868548723137e-06, | |
| "loss": 0.4823, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.26686656671664166, | |
| "grad_norm": 0.9763576108576292, | |
| "learning_rate": 9.567088481906084e-06, | |
| "loss": 0.2868, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.2683658170914543, | |
| "grad_norm": 0.8318841432176839, | |
| "learning_rate": 9.562283085548546e-06, | |
| "loss": 0.2306, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.2698650674662669, | |
| "grad_norm": 1.236132515309932, | |
| "learning_rate": 9.55745238630174e-06, | |
| "loss": 0.3682, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.27136431784107945, | |
| "grad_norm": 1.0296533261851089, | |
| "learning_rate": 9.552596410957224e-06, | |
| "loss": 0.3127, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.272863568215892, | |
| "grad_norm": 1.0118791903302382, | |
| "learning_rate": 9.547715186446732e-06, | |
| "loss": 0.2833, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.27436281859070466, | |
| "grad_norm": 1.0370082613699223, | |
| "learning_rate": 9.542808739842034e-06, | |
| "loss": 0.3372, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 1.0835469829364954, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.3075, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.2773613193403298, | |
| "grad_norm": 1.1054660875818354, | |
| "learning_rate": 9.532920289336378e-06, | |
| "loss": 0.3063, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.27886056971514245, | |
| "grad_norm": 0.9159523641726186, | |
| "learning_rate": 9.52793834027778e-06, | |
| "loss": 0.3148, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.280359820089955, | |
| "grad_norm": 0.9652813820134847, | |
| "learning_rate": 9.522931278809393e-06, | |
| "loss": 0.3551, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.2818590704647676, | |
| "grad_norm": 1.0410950266453554, | |
| "learning_rate": 9.517899132700889e-06, | |
| "loss": 0.3022, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.28335832083958024, | |
| "grad_norm": 0.885911419723158, | |
| "learning_rate": 9.512841929861069e-06, | |
| "loss": 0.2464, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.2848575712143928, | |
| "grad_norm": 1.0531270596401427, | |
| "learning_rate": 9.507759698337698e-06, | |
| "loss": 0.3804, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2863568215892054, | |
| "grad_norm": 0.9867400326109136, | |
| "learning_rate": 9.50265246631735e-06, | |
| "loss": 0.3373, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.28785607196401797, | |
| "grad_norm": 0.8489434068725049, | |
| "learning_rate": 9.49752026212526e-06, | |
| "loss": 0.304, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.2893553223388306, | |
| "grad_norm": 1.1227437844262784, | |
| "learning_rate": 9.492363114225156e-06, | |
| "loss": 0.4224, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2908545727136432, | |
| "grad_norm": 0.9068188456630527, | |
| "learning_rate": 9.487181051219107e-06, | |
| "loss": 0.2755, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.29235382308845576, | |
| "grad_norm": 1.1024035587304113, | |
| "learning_rate": 9.481974101847371e-06, | |
| "loss": 0.3191, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2938530734632684, | |
| "grad_norm": 1.1384337762946877, | |
| "learning_rate": 9.476742294988214e-06, | |
| "loss": 0.264, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.29535232383808097, | |
| "grad_norm": 1.1614857450712157, | |
| "learning_rate": 9.471485659657782e-06, | |
| "loss": 0.3162, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.29685157421289354, | |
| "grad_norm": 1.157683023427669, | |
| "learning_rate": 9.466204225009905e-06, | |
| "loss": 0.2919, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2983508245877061, | |
| "grad_norm": 0.8952187179177139, | |
| "learning_rate": 9.460898020335964e-06, | |
| "loss": 0.3106, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.29985007496251875, | |
| "grad_norm": 0.8968450017164568, | |
| "learning_rate": 9.455567075064715e-06, | |
| "loss": 0.2734, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.29985007496251875, | |
| "eval_loss": 0.31869983673095703, | |
| "eval_runtime": 9.5143, | |
| "eval_samples_per_second": 5.676, | |
| "eval_steps_per_second": 1.471, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30134932533733133, | |
| "grad_norm": 0.8480235125315602, | |
| "learning_rate": 9.450211418762123e-06, | |
| "loss": 0.2494, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.3028485757121439, | |
| "grad_norm": 1.1233348565193118, | |
| "learning_rate": 9.444831081131209e-06, | |
| "loss": 0.3299, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.30434782608695654, | |
| "grad_norm": 1.0905246954218262, | |
| "learning_rate": 9.439426092011877e-06, | |
| "loss": 0.2661, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.3058470764617691, | |
| "grad_norm": 0.961390023433098, | |
| "learning_rate": 9.433996481380747e-06, | |
| "loss": 0.247, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.3073463268365817, | |
| "grad_norm": 1.1208797915173023, | |
| "learning_rate": 9.428542279351e-06, | |
| "loss": 0.3165, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.30884557721139433, | |
| "grad_norm": 1.0124640331930874, | |
| "learning_rate": 9.423063516172195e-06, | |
| "loss": 0.3063, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.3103448275862069, | |
| "grad_norm": 0.9867173146536662, | |
| "learning_rate": 9.417560222230115e-06, | |
| "loss": 0.3145, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3118440779610195, | |
| "grad_norm": 1.2925204620576731, | |
| "learning_rate": 9.412032428046594e-06, | |
| "loss": 0.3723, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.31334332833583206, | |
| "grad_norm": 1.0368115611593605, | |
| "learning_rate": 9.40648016427934e-06, | |
| "loss": 0.3181, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.3148425787106447, | |
| "grad_norm": 1.0050923453796818, | |
| "learning_rate": 9.400903461721783e-06, | |
| "loss": 0.3259, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.31634182908545727, | |
| "grad_norm": 1.02379608686639, | |
| "learning_rate": 9.395302351302881e-06, | |
| "loss": 0.3039, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.31784107946026985, | |
| "grad_norm": 1.0415218515865996, | |
| "learning_rate": 9.38967686408697e-06, | |
| "loss": 0.2956, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.3193403298350825, | |
| "grad_norm": 1.2475178813377137, | |
| "learning_rate": 9.384027031273575e-06, | |
| "loss": 0.3694, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.32083958020989506, | |
| "grad_norm": 1.0873796471991162, | |
| "learning_rate": 9.37835288419725e-06, | |
| "loss": 0.3226, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.32233883058470764, | |
| "grad_norm": 0.9799704748901195, | |
| "learning_rate": 9.372654454327394e-06, | |
| "loss": 0.284, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.3238380809595202, | |
| "grad_norm": 1.0896196226918418, | |
| "learning_rate": 9.366931773268083e-06, | |
| "loss": 0.3346, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.32533733133433285, | |
| "grad_norm": 0.9063587210153955, | |
| "learning_rate": 9.361184872757894e-06, | |
| "loss": 0.2978, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3268365817091454, | |
| "grad_norm": 1.1336743666854052, | |
| "learning_rate": 9.355413784669722e-06, | |
| "loss": 0.3091, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.328335832083958, | |
| "grad_norm": 0.9881894028379479, | |
| "learning_rate": 9.349618541010616e-06, | |
| "loss": 0.2787, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.32983508245877063, | |
| "grad_norm": 1.1506314349563826, | |
| "learning_rate": 9.343799173921591e-06, | |
| "loss": 0.3291, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3313343328335832, | |
| "grad_norm": 1.0868397563169947, | |
| "learning_rate": 9.337955715677452e-06, | |
| "loss": 0.3133, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.3328335832083958, | |
| "grad_norm": 1.0278359691744634, | |
| "learning_rate": 9.332088198686618e-06, | |
| "loss": 0.3366, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.3343328335832084, | |
| "grad_norm": 0.8726520467431229, | |
| "learning_rate": 9.326196655490935e-06, | |
| "loss": 0.2936, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.335832083958021, | |
| "grad_norm": 1.1459514552149352, | |
| "learning_rate": 9.32028111876551e-06, | |
| "loss": 0.3035, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.3373313343328336, | |
| "grad_norm": 0.935188707816523, | |
| "learning_rate": 9.314341621318512e-06, | |
| "loss": 0.295, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.33883058470764615, | |
| "grad_norm": 1.142087227793732, | |
| "learning_rate": 9.308378196091006e-06, | |
| "loss": 0.3896, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.3403298350824588, | |
| "grad_norm": 1.0100547561716957, | |
| "learning_rate": 9.302390876156756e-06, | |
| "loss": 0.2791, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.34182908545727136, | |
| "grad_norm": 0.916292348845555, | |
| "learning_rate": 9.296379694722051e-06, | |
| "loss": 0.2708, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.34332833583208394, | |
| "grad_norm": 0.823903635334227, | |
| "learning_rate": 9.29034468512552e-06, | |
| "loss": 0.2627, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 0.9158027819379534, | |
| "learning_rate": 9.284285880837947e-06, | |
| "loss": 0.301, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.34632683658170915, | |
| "grad_norm": 1.0070175568137585, | |
| "learning_rate": 9.278203315462078e-06, | |
| "loss": 0.2663, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.9612837362063901, | |
| "learning_rate": 9.272097022732444e-06, | |
| "loss": 0.2654, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.3493253373313343, | |
| "grad_norm": 0.9769397316259165, | |
| "learning_rate": 9.26596703651517e-06, | |
| "loss": 0.2936, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.35082458770614694, | |
| "grad_norm": 1.0997796439009568, | |
| "learning_rate": 9.259813390807788e-06, | |
| "loss": 0.3625, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.3523238380809595, | |
| "grad_norm": 1.0445827808394417, | |
| "learning_rate": 9.253636119739046e-06, | |
| "loss": 0.2903, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.3538230884557721, | |
| "grad_norm": 1.003243248019039, | |
| "learning_rate": 9.247435257568724e-06, | |
| "loss": 0.307, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.3553223388305847, | |
| "grad_norm": 1.0400112568895221, | |
| "learning_rate": 9.241210838687438e-06, | |
| "loss": 0.3319, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3568215892053973, | |
| "grad_norm": 1.0866708452078757, | |
| "learning_rate": 9.23496289761645e-06, | |
| "loss": 0.3353, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.3583208395802099, | |
| "grad_norm": 1.1117381299175488, | |
| "learning_rate": 9.228691469007487e-06, | |
| "loss": 0.3505, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.3598200899550225, | |
| "grad_norm": 1.208318363755207, | |
| "learning_rate": 9.222396587642528e-06, | |
| "loss": 0.3425, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3613193403298351, | |
| "grad_norm": 1.0176690641127804, | |
| "learning_rate": 9.216078288433632e-06, | |
| "loss": 0.3147, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.36281859070464767, | |
| "grad_norm": 1.2159000663637003, | |
| "learning_rate": 9.209736606422736e-06, | |
| "loss": 0.3116, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.36431784107946025, | |
| "grad_norm": 1.122447414910652, | |
| "learning_rate": 9.203371576781457e-06, | |
| "loss": 0.3231, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.3658170914542729, | |
| "grad_norm": 1.1055483702777753, | |
| "learning_rate": 9.1969832348109e-06, | |
| "loss": 0.2927, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.36731634182908546, | |
| "grad_norm": 0.9208033228849616, | |
| "learning_rate": 9.190571615941462e-06, | |
| "loss": 0.2995, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.36881559220389803, | |
| "grad_norm": 1.0188220361660114, | |
| "learning_rate": 9.18413675573264e-06, | |
| "loss": 0.3036, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.37031484257871067, | |
| "grad_norm": 1.0874804672536833, | |
| "learning_rate": 9.177678689872831e-06, | |
| "loss": 0.3819, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.37181409295352325, | |
| "grad_norm": 0.9161139949581087, | |
| "learning_rate": 9.171197454179124e-06, | |
| "loss": 0.3008, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.3733133433283358, | |
| "grad_norm": 1.0563168005147139, | |
| "learning_rate": 9.16469308459712e-06, | |
| "loss": 0.3199, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.3748125937031484, | |
| "grad_norm": 1.0950011538291917, | |
| "learning_rate": 9.158165617200717e-06, | |
| "loss": 0.3135, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.37631184407796103, | |
| "grad_norm": 0.9432584968868432, | |
| "learning_rate": 9.151615088191918e-06, | |
| "loss": 0.2826, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.3778110944527736, | |
| "grad_norm": 0.8988437642485999, | |
| "learning_rate": 9.14504153390063e-06, | |
| "loss": 0.2792, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.3793103448275862, | |
| "grad_norm": 1.026473133644252, | |
| "learning_rate": 9.138444990784455e-06, | |
| "loss": 0.3597, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.3808095952023988, | |
| "grad_norm": 1.100392331418073, | |
| "learning_rate": 9.131825495428496e-06, | |
| "loss": 0.316, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.3823088455772114, | |
| "grad_norm": 0.9839108819280205, | |
| "learning_rate": 9.125183084545158e-06, | |
| "loss": 0.3144, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.383808095952024, | |
| "grad_norm": 1.193147882406105, | |
| "learning_rate": 9.118517794973925e-06, | |
| "loss": 0.3399, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.3853073463268366, | |
| "grad_norm": 0.8783552350627609, | |
| "learning_rate": 9.111829663681182e-06, | |
| "loss": 0.2929, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.3868065967016492, | |
| "grad_norm": 0.966124200647951, | |
| "learning_rate": 9.105118727759984e-06, | |
| "loss": 0.3225, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.38830584707646176, | |
| "grad_norm": 1.2709550085127772, | |
| "learning_rate": 9.098385024429875e-06, | |
| "loss": 0.3149, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.38980509745127434, | |
| "grad_norm": 0.950976592907919, | |
| "learning_rate": 9.09162859103666e-06, | |
| "loss": 0.3206, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.391304347826087, | |
| "grad_norm": 1.2779271729217527, | |
| "learning_rate": 9.08484946505221e-06, | |
| "loss": 0.3719, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.39280359820089955, | |
| "grad_norm": 1.079501979886789, | |
| "learning_rate": 9.078047684074254e-06, | |
| "loss": 0.3463, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.39430284857571213, | |
| "grad_norm": 0.9629488364217332, | |
| "learning_rate": 9.071223285826166e-06, | |
| "loss": 0.2554, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.39580209895052476, | |
| "grad_norm": 1.207628299858363, | |
| "learning_rate": 9.064376308156754e-06, | |
| "loss": 0.357, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.39730134932533734, | |
| "grad_norm": 1.1714560996761474, | |
| "learning_rate": 9.057506789040063e-06, | |
| "loss": 0.3248, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.3988005997001499, | |
| "grad_norm": 1.0809050778287788, | |
| "learning_rate": 9.050614766575147e-06, | |
| "loss": 0.3761, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.4002998500749625, | |
| "grad_norm": 0.9700226820304788, | |
| "learning_rate": 9.043700278985867e-06, | |
| "loss": 0.277, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.4017991004497751, | |
| "grad_norm": 0.9994596920093544, | |
| "learning_rate": 9.03676336462068e-06, | |
| "loss": 0.3085, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.4032983508245877, | |
| "grad_norm": 0.9587320500996325, | |
| "learning_rate": 9.029804061952426e-06, | |
| "loss": 0.2744, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.4047976011994003, | |
| "grad_norm": 0.9448939896455943, | |
| "learning_rate": 9.022822409578106e-06, | |
| "loss": 0.282, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4062968515742129, | |
| "grad_norm": 0.9259228086820999, | |
| "learning_rate": 9.015818446218683e-06, | |
| "loss": 0.2672, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.4077961019490255, | |
| "grad_norm": 1.084983188448112, | |
| "learning_rate": 9.008792210718854e-06, | |
| "loss": 0.3067, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.40929535232383807, | |
| "grad_norm": 1.1759225275823966, | |
| "learning_rate": 9.00174374204684e-06, | |
| "loss": 0.3859, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.4107946026986507, | |
| "grad_norm": 1.1029647931455526, | |
| "learning_rate": 8.994673079294171e-06, | |
| "loss": 0.279, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.4122938530734633, | |
| "grad_norm": 1.1396826609832766, | |
| "learning_rate": 8.987580261675466e-06, | |
| "loss": 0.3024, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.41379310344827586, | |
| "grad_norm": 1.2219689691404456, | |
| "learning_rate": 8.98046532852822e-06, | |
| "loss": 0.3517, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.41529235382308843, | |
| "grad_norm": 1.1987269205403046, | |
| "learning_rate": 8.973328319312577e-06, | |
| "loss": 0.3676, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.41679160419790107, | |
| "grad_norm": 0.8197601234835827, | |
| "learning_rate": 8.966169273611125e-06, | |
| "loss": 0.2303, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.41829085457271364, | |
| "grad_norm": 1.2065670867402163, | |
| "learning_rate": 8.958988231128665e-06, | |
| "loss": 0.3414, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.4197901049475262, | |
| "grad_norm": 1.0171683320335805, | |
| "learning_rate": 8.95178523169199e-06, | |
| "loss": 0.2959, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.42128935532233885, | |
| "grad_norm": 1.0318396386786353, | |
| "learning_rate": 8.944560315249676e-06, | |
| "loss": 0.2676, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.42278860569715143, | |
| "grad_norm": 0.8332344882614791, | |
| "learning_rate": 8.937313521871846e-06, | |
| "loss": 0.2686, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.424287856071964, | |
| "grad_norm": 1.0217493363671384, | |
| "learning_rate": 8.930044891749962e-06, | |
| "loss": 0.3098, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.4257871064467766, | |
| "grad_norm": 1.0887663653945965, | |
| "learning_rate": 8.922754465196591e-06, | |
| "loss": 0.2741, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.4272863568215892, | |
| "grad_norm": 0.9167218135294277, | |
| "learning_rate": 8.915442282645183e-06, | |
| "loss": 0.2932, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.4287856071964018, | |
| "grad_norm": 0.937509863460441, | |
| "learning_rate": 8.908108384649856e-06, | |
| "loss": 0.2792, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.4302848575712144, | |
| "grad_norm": 1.0225725825184255, | |
| "learning_rate": 8.900752811885152e-06, | |
| "loss": 0.3024, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.431784107946027, | |
| "grad_norm": 0.9440449911878376, | |
| "learning_rate": 8.893375605145837e-06, | |
| "loss": 0.2753, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.4332833583208396, | |
| "grad_norm": 0.9187579224085831, | |
| "learning_rate": 8.885976805346651e-06, | |
| "loss": 0.3467, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 1.0621519774761492, | |
| "learning_rate": 8.8785564535221e-06, | |
| "loss": 0.3422, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4362818590704648, | |
| "grad_norm": 1.450220623057443, | |
| "learning_rate": 8.871114590826211e-06, | |
| "loss": 0.3236, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.43778110944527737, | |
| "grad_norm": 1.3758988492922357, | |
| "learning_rate": 8.86365125853232e-06, | |
| "loss": 0.2963, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.43928035982008995, | |
| "grad_norm": 0.9710368260403254, | |
| "learning_rate": 8.85616649803283e-06, | |
| "loss": 0.3271, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.4407796101949025, | |
| "grad_norm": 0.8809469474942717, | |
| "learning_rate": 8.84866035083899e-06, | |
| "loss": 0.2302, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.44227886056971516, | |
| "grad_norm": 0.8919894898867272, | |
| "learning_rate": 8.841132858580661e-06, | |
| "loss": 0.2743, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.44377811094452774, | |
| "grad_norm": 1.0570473669211173, | |
| "learning_rate": 8.833584063006088e-06, | |
| "loss": 0.299, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.4452773613193403, | |
| "grad_norm": 1.291180693184136, | |
| "learning_rate": 8.826014005981662e-06, | |
| "loss": 0.3704, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.44677661169415295, | |
| "grad_norm": 1.0426741605221057, | |
| "learning_rate": 8.818422729491693e-06, | |
| "loss": 0.246, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.4482758620689655, | |
| "grad_norm": 1.065529245036711, | |
| "learning_rate": 8.810810275638183e-06, | |
| "loss": 0.2989, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.4497751124437781, | |
| "grad_norm": 0.9297531288460176, | |
| "learning_rate": 8.803176686640577e-06, | |
| "loss": 0.2288, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4512743628185907, | |
| "grad_norm": 1.0115935908212754, | |
| "learning_rate": 8.795522004835543e-06, | |
| "loss": 0.297, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.4527736131934033, | |
| "grad_norm": 0.9868983777055481, | |
| "learning_rate": 8.787846272676728e-06, | |
| "loss": 0.3066, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.4542728635682159, | |
| "grad_norm": 0.9348545508549615, | |
| "learning_rate": 8.780149532734531e-06, | |
| "loss": 0.2856, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.45577211394302847, | |
| "grad_norm": 1.0101707757994063, | |
| "learning_rate": 8.772431827695862e-06, | |
| "loss": 0.3262, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.4572713643178411, | |
| "grad_norm": 1.0392969185349228, | |
| "learning_rate": 8.764693200363897e-06, | |
| "loss": 0.2927, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.4587706146926537, | |
| "grad_norm": 1.1170970253905805, | |
| "learning_rate": 8.756933693657863e-06, | |
| "loss": 0.2851, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.46026986506746626, | |
| "grad_norm": 1.0343800090675301, | |
| "learning_rate": 8.749153350612774e-06, | |
| "loss": 0.2999, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.4617691154422789, | |
| "grad_norm": 0.9580715128189217, | |
| "learning_rate": 8.74135221437921e-06, | |
| "loss": 0.2478, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.46326836581709147, | |
| "grad_norm": 1.1018853721458668, | |
| "learning_rate": 8.733530328223076e-06, | |
| "loss": 0.3087, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.46476761619190404, | |
| "grad_norm": 1.0925703390579977, | |
| "learning_rate": 8.725687735525347e-06, | |
| "loss": 0.2646, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4662668665667166, | |
| "grad_norm": 0.9182992803068464, | |
| "learning_rate": 8.71782447978185e-06, | |
| "loss": 0.2615, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.46776611694152925, | |
| "grad_norm": 1.1867525010582973, | |
| "learning_rate": 8.709940604603006e-06, | |
| "loss": 0.3699, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.46926536731634183, | |
| "grad_norm": 1.0083625115162043, | |
| "learning_rate": 8.702036153713594e-06, | |
| "loss": 0.2731, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.4707646176911544, | |
| "grad_norm": 1.157461253467528, | |
| "learning_rate": 8.694111170952508e-06, | |
| "loss": 0.2844, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.47226386806596704, | |
| "grad_norm": 1.058100167940622, | |
| "learning_rate": 8.686165700272513e-06, | |
| "loss": 0.3583, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.4737631184407796, | |
| "grad_norm": 0.9609097999994505, | |
| "learning_rate": 8.678199785740003e-06, | |
| "loss": 0.288, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.4752623688155922, | |
| "grad_norm": 1.2356615721057218, | |
| "learning_rate": 8.670213471534759e-06, | |
| "loss": 0.3505, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.4767616191904048, | |
| "grad_norm": 0.8643982411259091, | |
| "learning_rate": 8.662206801949694e-06, | |
| "loss": 0.2442, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.4782608695652174, | |
| "grad_norm": 1.0241917774755254, | |
| "learning_rate": 8.65417982139062e-06, | |
| "loss": 0.2783, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.47976011994003, | |
| "grad_norm": 1.265780047847942, | |
| "learning_rate": 8.646132574375994e-06, | |
| "loss": 0.3458, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.48125937031484256, | |
| "grad_norm": 1.061441986652464, | |
| "learning_rate": 8.638065105536669e-06, | |
| "loss": 0.3457, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.4827586206896552, | |
| "grad_norm": 1.127642436863335, | |
| "learning_rate": 8.629977459615655e-06, | |
| "loss": 0.2831, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.48425787106446777, | |
| "grad_norm": 1.0790015228740566, | |
| "learning_rate": 8.621869681467865e-06, | |
| "loss": 0.2997, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.48575712143928035, | |
| "grad_norm": 0.9813936310316222, | |
| "learning_rate": 8.613741816059867e-06, | |
| "loss": 0.2474, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.487256371814093, | |
| "grad_norm": 1.2144571513441706, | |
| "learning_rate": 8.605593908469635e-06, | |
| "loss": 0.3013, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.48875562218890556, | |
| "grad_norm": 0.9858218100657198, | |
| "learning_rate": 8.597426003886295e-06, | |
| "loss": 0.2577, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.49025487256371814, | |
| "grad_norm": 1.078671413677644, | |
| "learning_rate": 8.58923814760989e-06, | |
| "loss": 0.3569, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.4917541229385307, | |
| "grad_norm": 1.068267650020554, | |
| "learning_rate": 8.581030385051105e-06, | |
| "loss": 0.3442, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.49325337331334335, | |
| "grad_norm": 1.0097959427928236, | |
| "learning_rate": 8.572802761731031e-06, | |
| "loss": 0.309, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.4947526236881559, | |
| "grad_norm": 1.0982883161829704, | |
| "learning_rate": 8.564555323280913e-06, | |
| "loss": 0.342, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4962518740629685, | |
| "grad_norm": 1.0846238809649005, | |
| "learning_rate": 8.556288115441887e-06, | |
| "loss": 0.2983, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.49775112443778113, | |
| "grad_norm": 1.0336870722645928, | |
| "learning_rate": 8.548001184064733e-06, | |
| "loss": 0.3042, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.4992503748125937, | |
| "grad_norm": 0.974305397696469, | |
| "learning_rate": 8.539694575109626e-06, | |
| "loss": 0.2794, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.5007496251874063, | |
| "grad_norm": 1.1076226831080926, | |
| "learning_rate": 8.531368334645865e-06, | |
| "loss": 0.2835, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.5022488755622189, | |
| "grad_norm": 0.9963509130416958, | |
| "learning_rate": 8.523022508851634e-06, | |
| "loss": 0.3165, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.5037481259370314, | |
| "grad_norm": 0.9852675575374601, | |
| "learning_rate": 8.514657144013738e-06, | |
| "loss": 0.3055, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.5052473763118441, | |
| "grad_norm": 1.189795843498438, | |
| "learning_rate": 8.506272286527346e-06, | |
| "loss": 0.3923, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.5067466266866567, | |
| "grad_norm": 0.9736788807928332, | |
| "learning_rate": 8.497867982895741e-06, | |
| "loss": 0.2694, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.5082458770614693, | |
| "grad_norm": 0.9472384958915275, | |
| "learning_rate": 8.489444279730046e-06, | |
| "loss": 0.2691, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.5097451274362819, | |
| "grad_norm": 1.2372381856549797, | |
| "learning_rate": 8.481001223748986e-06, | |
| "loss": 0.2695, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5112443778110944, | |
| "grad_norm": 1.2184382217115792, | |
| "learning_rate": 8.47253886177861e-06, | |
| "loss": 0.376, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.512743628185907, | |
| "grad_norm": 1.2620953237730903, | |
| "learning_rate": 8.464057240752046e-06, | |
| "loss": 0.3099, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.5142428785607196, | |
| "grad_norm": 1.0917409905851174, | |
| "learning_rate": 8.455556407709235e-06, | |
| "loss": 0.2423, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.5157421289355323, | |
| "grad_norm": 1.056908600779691, | |
| "learning_rate": 8.447036409796663e-06, | |
| "loss": 0.3181, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 1.0312074162522642, | |
| "learning_rate": 8.438497294267117e-06, | |
| "loss": 0.316, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5187406296851574, | |
| "grad_norm": 1.2658860635128897, | |
| "learning_rate": 8.429939108479403e-06, | |
| "loss": 0.4003, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.52023988005997, | |
| "grad_norm": 1.013360081871048, | |
| "learning_rate": 8.421361899898095e-06, | |
| "loss": 0.2935, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 1.0746280466195366, | |
| "learning_rate": 8.412765716093273e-06, | |
| "loss": 0.2839, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.5232383808095952, | |
| "grad_norm": 0.986023389485262, | |
| "learning_rate": 8.404150604740248e-06, | |
| "loss": 0.3102, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.5247376311844077, | |
| "grad_norm": 1.0770947329987295, | |
| "learning_rate": 8.395516613619315e-06, | |
| "loss": 0.2913, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5262368815592204, | |
| "grad_norm": 0.9977850610477095, | |
| "learning_rate": 8.386863790615472e-06, | |
| "loss": 0.3382, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.527736131934033, | |
| "grad_norm": 1.1589340597849656, | |
| "learning_rate": 8.378192183718158e-06, | |
| "loss": 0.3329, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.5292353823088456, | |
| "grad_norm": 1.1827900090569905, | |
| "learning_rate": 8.369501841021e-06, | |
| "loss": 0.3414, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.5307346326836582, | |
| "grad_norm": 1.0733531154886229, | |
| "learning_rate": 8.360792810721522e-06, | |
| "loss": 0.2786, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.5322338830584707, | |
| "grad_norm": 1.055964869173219, | |
| "learning_rate": 8.352065141120902e-06, | |
| "loss": 0.3048, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5337331334332833, | |
| "grad_norm": 1.0890553542800532, | |
| "learning_rate": 8.343318880623688e-06, | |
| "loss": 0.2799, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.5352323838080959, | |
| "grad_norm": 1.0802743114613247, | |
| "learning_rate": 8.334554077737535e-06, | |
| "loss": 0.3362, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.5367316341829086, | |
| "grad_norm": 1.0021684126909913, | |
| "learning_rate": 8.325770781072939e-06, | |
| "loss": 0.2855, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.5382308845577212, | |
| "grad_norm": 0.8516926263370231, | |
| "learning_rate": 8.316969039342963e-06, | |
| "loss": 0.2951, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.5397301349325337, | |
| "grad_norm": 0.914814803274709, | |
| "learning_rate": 8.30814890136297e-06, | |
| "loss": 0.2215, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5412293853073463, | |
| "grad_norm": 1.2983028064517146, | |
| "learning_rate": 8.299310416050345e-06, | |
| "loss": 0.3727, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.5427286356821589, | |
| "grad_norm": 0.7912480595717224, | |
| "learning_rate": 8.290453632424236e-06, | |
| "loss": 0.2378, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.5442278860569715, | |
| "grad_norm": 1.3828092470286426, | |
| "learning_rate": 8.281578599605269e-06, | |
| "loss": 0.3254, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.545727136431784, | |
| "grad_norm": 1.140299554198238, | |
| "learning_rate": 8.272685366815287e-06, | |
| "loss": 0.3976, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5472263868065967, | |
| "grad_norm": 0.9458988142023489, | |
| "learning_rate": 8.26377398337707e-06, | |
| "loss": 0.3329, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5487256371814093, | |
| "grad_norm": 1.014680990890851, | |
| "learning_rate": 8.254844498714063e-06, | |
| "loss": 0.2847, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.5502248875562219, | |
| "grad_norm": 1.0115931752205278, | |
| "learning_rate": 8.2458969623501e-06, | |
| "loss": 0.3558, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 0.9677315739730885, | |
| "learning_rate": 8.23693142390914e-06, | |
| "loss": 0.3486, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.553223388305847, | |
| "grad_norm": 0.9920654299731347, | |
| "learning_rate": 8.227947933114971e-06, | |
| "loss": 0.2436, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.5547226386806596, | |
| "grad_norm": 1.025033334197306, | |
| "learning_rate": 8.218946539790957e-06, | |
| "loss": 0.3166, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5562218890554723, | |
| "grad_norm": 1.0100064444333272, | |
| "learning_rate": 8.209927293859746e-06, | |
| "loss": 0.2958, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.5577211394302849, | |
| "grad_norm": 1.1332437145603942, | |
| "learning_rate": 8.200890245342999e-06, | |
| "loss": 0.3691, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.5592203898050975, | |
| "grad_norm": 1.1244200799599144, | |
| "learning_rate": 8.191835444361113e-06, | |
| "loss": 0.3887, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.56071964017991, | |
| "grad_norm": 0.8761376024123404, | |
| "learning_rate": 8.182762941132944e-06, | |
| "loss": 0.2556, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.5622188905547226, | |
| "grad_norm": 0.9905645755691745, | |
| "learning_rate": 8.173672785975522e-06, | |
| "loss": 0.3179, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5637181409295352, | |
| "grad_norm": 1.238308733159341, | |
| "learning_rate": 8.16456502930378e-06, | |
| "loss": 0.4096, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.5652173913043478, | |
| "grad_norm": 1.0689468591667568, | |
| "learning_rate": 8.155439721630265e-06, | |
| "loss": 0.3352, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.5667166416791605, | |
| "grad_norm": 0.9261362654765208, | |
| "learning_rate": 8.146296913564872e-06, | |
| "loss": 0.2717, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.568215892053973, | |
| "grad_norm": 1.0482376327524345, | |
| "learning_rate": 8.13713665581455e-06, | |
| "loss": 0.2865, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.5697151424287856, | |
| "grad_norm": 0.9208517013614638, | |
| "learning_rate": 8.127958999183027e-06, | |
| "loss": 0.291, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5712143928035982, | |
| "grad_norm": 0.9648765246687899, | |
| "learning_rate": 8.118763994570528e-06, | |
| "loss": 0.3215, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.5727136431784108, | |
| "grad_norm": 1.1181484676977491, | |
| "learning_rate": 8.109551692973487e-06, | |
| "loss": 0.2813, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.5742128935532234, | |
| "grad_norm": 0.903837073355292, | |
| "learning_rate": 8.100322145484275e-06, | |
| "loss": 0.3259, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.5757121439280359, | |
| "grad_norm": 0.9821027000628693, | |
| "learning_rate": 8.091075403290905e-06, | |
| "loss": 0.3082, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.5772113943028486, | |
| "grad_norm": 0.9555515935613017, | |
| "learning_rate": 8.081811517676759e-06, | |
| "loss": 0.2952, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5787106446776612, | |
| "grad_norm": 0.9123656815937776, | |
| "learning_rate": 8.072530540020294e-06, | |
| "loss": 0.3124, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.5802098950524738, | |
| "grad_norm": 1.0087738105713129, | |
| "learning_rate": 8.063232521794762e-06, | |
| "loss": 0.2744, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.5817091454272864, | |
| "grad_norm": 0.7814886257030517, | |
| "learning_rate": 8.053917514567927e-06, | |
| "loss": 0.2273, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.5832083958020989, | |
| "grad_norm": 1.0000345945933324, | |
| "learning_rate": 8.04458557000177e-06, | |
| "loss": 0.2828, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.5847076461769115, | |
| "grad_norm": 1.0960329702355303, | |
| "learning_rate": 8.035236739852214e-06, | |
| "loss": 0.2944, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5862068965517241, | |
| "grad_norm": 1.2043782799581229, | |
| "learning_rate": 8.025871075968828e-06, | |
| "loss": 0.295, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.5877061469265368, | |
| "grad_norm": 0.9351109027146031, | |
| "learning_rate": 8.016488630294539e-06, | |
| "loss": 0.2739, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.5892053973013494, | |
| "grad_norm": 1.0809177679179123, | |
| "learning_rate": 8.007089454865358e-06, | |
| "loss": 0.3121, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.5907046476761619, | |
| "grad_norm": 1.0583391261322341, | |
| "learning_rate": 7.997673601810071e-06, | |
| "loss": 0.263, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.5922038980509745, | |
| "grad_norm": 0.8327955059474648, | |
| "learning_rate": 7.988241123349965e-06, | |
| "loss": 0.2907, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5937031484257871, | |
| "grad_norm": 1.161122055136904, | |
| "learning_rate": 7.97879207179853e-06, | |
| "loss": 0.2786, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.5952023988005997, | |
| "grad_norm": 1.0912562545793394, | |
| "learning_rate": 7.969326499561173e-06, | |
| "loss": 0.274, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.5967016491754122, | |
| "grad_norm": 1.1247683078486286, | |
| "learning_rate": 7.95984445913493e-06, | |
| "loss": 0.3318, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.5982008995502249, | |
| "grad_norm": 0.941143786313585, | |
| "learning_rate": 7.950346003108167e-06, | |
| "loss": 0.2467, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.5997001499250375, | |
| "grad_norm": 0.8855235148835696, | |
| "learning_rate": 7.940831184160294e-06, | |
| "loss": 0.2738, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5997001499250375, | |
| "eval_loss": 0.2967388331890106, | |
| "eval_runtime": 9.5286, | |
| "eval_samples_per_second": 5.667, | |
| "eval_steps_per_second": 1.469, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6011994002998501, | |
| "grad_norm": 1.0409321185895595, | |
| "learning_rate": 7.93130005506147e-06, | |
| "loss": 0.3129, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.6026986506746627, | |
| "grad_norm": 0.8706975187420958, | |
| "learning_rate": 7.921752668672316e-06, | |
| "loss": 0.2717, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.6041979010494752, | |
| "grad_norm": 0.9946334707475166, | |
| "learning_rate": 7.912189077943613e-06, | |
| "loss": 0.2742, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.6056971514242878, | |
| "grad_norm": 1.0342754977418127, | |
| "learning_rate": 7.902609335916015e-06, | |
| "loss": 0.2453, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.6071964017991005, | |
| "grad_norm": 1.0542033923841698, | |
| "learning_rate": 7.893013495719752e-06, | |
| "loss": 0.2964, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6086956521739131, | |
| "grad_norm": 1.1287048298132587, | |
| "learning_rate": 7.883401610574338e-06, | |
| "loss": 0.3061, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.6101949025487257, | |
| "grad_norm": 1.2724791344499653, | |
| "learning_rate": 7.873773733788268e-06, | |
| "loss": 0.3203, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.6116941529235382, | |
| "grad_norm": 1.14984215601152, | |
| "learning_rate": 7.864129918758738e-06, | |
| "loss": 0.3242, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.6131934032983508, | |
| "grad_norm": 0.9166120830683945, | |
| "learning_rate": 7.854470218971333e-06, | |
| "loss": 0.3093, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.6146926536731634, | |
| "grad_norm": 1.0611366267630755, | |
| "learning_rate": 7.844794687999737e-06, | |
| "loss": 0.2895, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.616191904047976, | |
| "grad_norm": 1.1613615351446742, | |
| "learning_rate": 7.835103379505433e-06, | |
| "loss": 0.3049, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.6176911544227887, | |
| "grad_norm": 1.2293032673395698, | |
| "learning_rate": 7.825396347237413e-06, | |
| "loss": 0.2959, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.6191904047976012, | |
| "grad_norm": 1.1290278839266072, | |
| "learning_rate": 7.815673645031871e-06, | |
| "loss": 0.3108, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.6206896551724138, | |
| "grad_norm": 0.9651265650210589, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.3003, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.6221889055472264, | |
| "grad_norm": 1.1957745857723652, | |
| "learning_rate": 7.796181446587244e-06, | |
| "loss": 0.2623, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.623688155922039, | |
| "grad_norm": 1.034776434724627, | |
| "learning_rate": 7.786412058453886e-06, | |
| "loss": 0.2822, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.6251874062968515, | |
| "grad_norm": 0.961705663339655, | |
| "learning_rate": 7.776627216593863e-06, | |
| "loss": 0.2551, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.6266866566716641, | |
| "grad_norm": 1.0317139358978142, | |
| "learning_rate": 7.766826975274916e-06, | |
| "loss": 0.2499, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.6281859070464768, | |
| "grad_norm": 0.9956181640788834, | |
| "learning_rate": 7.75701138885018e-06, | |
| "loss": 0.2889, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.6296851574212894, | |
| "grad_norm": 1.0773371808514443, | |
| "learning_rate": 7.747180511757908e-06, | |
| "loss": 0.2897, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.631184407796102, | |
| "grad_norm": 0.9727771029291843, | |
| "learning_rate": 7.737334398521149e-06, | |
| "loss": 0.2601, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.6326836581709145, | |
| "grad_norm": 1.1274602592599927, | |
| "learning_rate": 7.727473103747456e-06, | |
| "loss": 0.3451, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.6341829085457271, | |
| "grad_norm": 1.0034524615792924, | |
| "learning_rate": 7.717596682128578e-06, | |
| "loss": 0.2411, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.6356821589205397, | |
| "grad_norm": 1.032601010686511, | |
| "learning_rate": 7.707705188440165e-06, | |
| "loss": 0.2785, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.6371814092953523, | |
| "grad_norm": 1.078532422129021, | |
| "learning_rate": 7.697798677541448e-06, | |
| "loss": 0.2953, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.638680659670165, | |
| "grad_norm": 1.007133000476496, | |
| "learning_rate": 7.687877204374957e-06, | |
| "loss": 0.2366, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.6401799100449775, | |
| "grad_norm": 1.1072154958433897, | |
| "learning_rate": 7.677940823966196e-06, | |
| "loss": 0.2476, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.6416791604197901, | |
| "grad_norm": 0.9767964102059562, | |
| "learning_rate": 7.667989591423349e-06, | |
| "loss": 0.2887, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.6431784107946027, | |
| "grad_norm": 0.8793080070942004, | |
| "learning_rate": 7.658023561936966e-06, | |
| "loss": 0.2205, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.6446776611694153, | |
| "grad_norm": 1.016711998983967, | |
| "learning_rate": 7.648042790779677e-06, | |
| "loss": 0.2917, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6461769115442278, | |
| "grad_norm": 1.1157660793308206, | |
| "learning_rate": 7.638047333305853e-06, | |
| "loss": 0.3154, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.6476761619190404, | |
| "grad_norm": 0.9326603034683526, | |
| "learning_rate": 7.628037244951328e-06, | |
| "loss": 0.3019, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.6491754122938531, | |
| "grad_norm": 1.1559273837703319, | |
| "learning_rate": 7.618012581233076e-06, | |
| "loss": 0.3177, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.6506746626686657, | |
| "grad_norm": 1.0166591209501548, | |
| "learning_rate": 7.607973397748909e-06, | |
| "loss": 0.3044, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 1.0964853693033865, | |
| "learning_rate": 7.597919750177168e-06, | |
| "loss": 0.275, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6536731634182908, | |
| "grad_norm": 1.1053705316832438, | |
| "learning_rate": 7.587851694276412e-06, | |
| "loss": 0.2773, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.6551724137931034, | |
| "grad_norm": 1.0825363577321558, | |
| "learning_rate": 7.57776928588511e-06, | |
| "loss": 0.2814, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.656671664167916, | |
| "grad_norm": 0.9442813024295046, | |
| "learning_rate": 7.56767258092133e-06, | |
| "loss": 0.2887, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.6581709145427287, | |
| "grad_norm": 1.1882145003618723, | |
| "learning_rate": 7.557561635382433e-06, | |
| "loss": 0.347, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.6596701649175413, | |
| "grad_norm": 0.8569254933620801, | |
| "learning_rate": 7.54743650534476e-06, | |
| "loss": 0.2573, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6611694152923538, | |
| "grad_norm": 1.1573476085896066, | |
| "learning_rate": 7.537297246963316e-06, | |
| "loss": 0.3149, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.6626686656671664, | |
| "grad_norm": 1.1322697339095917, | |
| "learning_rate": 7.5271439164714695e-06, | |
| "loss": 0.3134, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.664167916041979, | |
| "grad_norm": 0.97793161331879, | |
| "learning_rate": 7.5169765701806295e-06, | |
| "loss": 0.2555, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.6656671664167916, | |
| "grad_norm": 1.0869435860674974, | |
| "learning_rate": 7.506795264479941e-06, | |
| "loss": 0.2877, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.6671664167916042, | |
| "grad_norm": 1.1464458520602188, | |
| "learning_rate": 7.4966000558359675e-06, | |
| "loss": 0.3032, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6686656671664168, | |
| "grad_norm": 1.1719380303931501, | |
| "learning_rate": 7.486391000792379e-06, | |
| "loss": 0.3384, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.6701649175412294, | |
| "grad_norm": 1.177532411147632, | |
| "learning_rate": 7.476168155969643e-06, | |
| "loss": 0.4072, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.671664167916042, | |
| "grad_norm": 1.04783192835799, | |
| "learning_rate": 7.465931578064703e-06, | |
| "loss": 0.272, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.6731634182908546, | |
| "grad_norm": 0.9531158380930339, | |
| "learning_rate": 7.455681323850669e-06, | |
| "loss": 0.2689, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.6746626686656672, | |
| "grad_norm": 0.9081864322618658, | |
| "learning_rate": 7.4454174501765e-06, | |
| "loss": 0.2917, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6761619190404797, | |
| "grad_norm": 0.9296593819878303, | |
| "learning_rate": 7.4351400139666894e-06, | |
| "loss": 0.251, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.6776611694152923, | |
| "grad_norm": 0.9457250544634667, | |
| "learning_rate": 7.424849072220953e-06, | |
| "loss": 0.3132, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.679160419790105, | |
| "grad_norm": 0.9878097107078544, | |
| "learning_rate": 7.414544682013907e-06, | |
| "loss": 0.2824, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.6806596701649176, | |
| "grad_norm": 1.0902569486781268, | |
| "learning_rate": 7.404226900494753e-06, | |
| "loss": 0.3308, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.6821589205397302, | |
| "grad_norm": 0.8541761919458658, | |
| "learning_rate": 7.3938957848869684e-06, | |
| "loss": 0.2683, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6836581709145427, | |
| "grad_norm": 1.1955941902071734, | |
| "learning_rate": 7.3835513924879755e-06, | |
| "loss": 0.3049, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.6851574212893553, | |
| "grad_norm": 0.8035957109155117, | |
| "learning_rate": 7.373193780668835e-06, | |
| "loss": 0.2348, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.6866566716641679, | |
| "grad_norm": 0.8700137254180897, | |
| "learning_rate": 7.36282300687392e-06, | |
| "loss": 0.2883, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.6881559220389805, | |
| "grad_norm": 1.205671160333747, | |
| "learning_rate": 7.35243912862061e-06, | |
| "loss": 0.3453, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 1.0225706198309454, | |
| "learning_rate": 7.342042203498952e-06, | |
| "loss": 0.3134, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6911544227886057, | |
| "grad_norm": 1.1671396502467404, | |
| "learning_rate": 7.33163228917136e-06, | |
| "loss": 0.3664, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.6926536731634183, | |
| "grad_norm": 0.9129770499883939, | |
| "learning_rate": 7.321209443372284e-06, | |
| "loss": 0.2442, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.6941529235382309, | |
| "grad_norm": 1.045457980802117, | |
| "learning_rate": 7.310773723907895e-06, | |
| "loss": 0.376, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 1.0378683552206422, | |
| "learning_rate": 7.300325188655762e-06, | |
| "loss": 0.2554, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.697151424287856, | |
| "grad_norm": 1.1066414270631715, | |
| "learning_rate": 7.289863895564531e-06, | |
| "loss": 0.3372, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.6986506746626686, | |
| "grad_norm": 1.0038512890646993, | |
| "learning_rate": 7.279389902653606e-06, | |
| "loss": 0.2912, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.7001499250374813, | |
| "grad_norm": 1.2394376615759466, | |
| "learning_rate": 7.268903268012823e-06, | |
| "loss": 0.3434, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.7016491754122939, | |
| "grad_norm": 0.837778722091501, | |
| "learning_rate": 7.258404049802135e-06, | |
| "loss": 0.219, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.7031484257871065, | |
| "grad_norm": 0.8211123296382956, | |
| "learning_rate": 7.247892306251276e-06, | |
| "loss": 0.2171, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.704647676161919, | |
| "grad_norm": 0.9032078860729328, | |
| "learning_rate": 7.237368095659459e-06, | |
| "loss": 0.3007, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7061469265367316, | |
| "grad_norm": 1.1623307376409528, | |
| "learning_rate": 7.226831476395028e-06, | |
| "loss": 0.2892, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.7076461769115442, | |
| "grad_norm": 1.1850522882497698, | |
| "learning_rate": 7.216282506895155e-06, | |
| "loss": 0.2751, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.7091454272863568, | |
| "grad_norm": 1.0755043264832627, | |
| "learning_rate": 7.2057212456655055e-06, | |
| "loss": 0.2975, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.7106446776611695, | |
| "grad_norm": 1.0556959537832395, | |
| "learning_rate": 7.195147751279915e-06, | |
| "loss": 0.2782, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.712143928035982, | |
| "grad_norm": 1.0924905303379417, | |
| "learning_rate": 7.184562082380069e-06, | |
| "loss": 0.3054, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7136431784107946, | |
| "grad_norm": 1.3624044716025572, | |
| "learning_rate": 7.173964297675168e-06, | |
| "loss": 0.3491, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.7151424287856072, | |
| "grad_norm": 1.1117950275673425, | |
| "learning_rate": 7.163354455941614e-06, | |
| "loss": 0.3205, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.7166416791604198, | |
| "grad_norm": 0.9983986364334534, | |
| "learning_rate": 7.152732616022675e-06, | |
| "loss": 0.2615, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.7181409295352323, | |
| "grad_norm": 0.9054215325182234, | |
| "learning_rate": 7.142098836828162e-06, | |
| "loss": 0.3046, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.719640179910045, | |
| "grad_norm": 0.9722836605425593, | |
| "learning_rate": 7.131453177334103e-06, | |
| "loss": 0.2985, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7211394302848576, | |
| "grad_norm": 0.8346217565471921, | |
| "learning_rate": 7.120795696582419e-06, | |
| "loss": 0.2476, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.7226386806596702, | |
| "grad_norm": 1.0994219555638456, | |
| "learning_rate": 7.1101264536805885e-06, | |
| "loss": 0.2678, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.7241379310344828, | |
| "grad_norm": 1.0055481727880695, | |
| "learning_rate": 7.099445507801324e-06, | |
| "loss": 0.2751, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.7256371814092953, | |
| "grad_norm": 0.9853412108796591, | |
| "learning_rate": 7.088752918182247e-06, | |
| "loss": 0.2577, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.7271364317841079, | |
| "grad_norm": 1.0548458745592642, | |
| "learning_rate": 7.078048744125553e-06, | |
| "loss": 0.3143, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7286356821589205, | |
| "grad_norm": 0.7879982733656735, | |
| "learning_rate": 7.067333044997689e-06, | |
| "loss": 0.2446, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.7301349325337332, | |
| "grad_norm": 1.2248918909142623, | |
| "learning_rate": 7.0566058802290196e-06, | |
| "loss": 0.2996, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.7316341829085458, | |
| "grad_norm": 0.9375116375156618, | |
| "learning_rate": 7.045867309313499e-06, | |
| "loss": 0.2973, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.7331334332833583, | |
| "grad_norm": 1.0178579412171902, | |
| "learning_rate": 7.035117391808341e-06, | |
| "loss": 0.2819, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.7346326836581709, | |
| "grad_norm": 1.0058066295527148, | |
| "learning_rate": 7.024356187333692e-06, | |
| "loss": 0.3416, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7361319340329835, | |
| "grad_norm": 0.9442967236862722, | |
| "learning_rate": 7.01358375557229e-06, | |
| "loss": 0.3108, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.7376311844077961, | |
| "grad_norm": 0.9412878929970308, | |
| "learning_rate": 7.0028001562691475e-06, | |
| "loss": 0.2609, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.7391304347826086, | |
| "grad_norm": 0.9076713411204123, | |
| "learning_rate": 6.9920054492312086e-06, | |
| "loss": 0.3102, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.7406296851574213, | |
| "grad_norm": 1.0837835543176282, | |
| "learning_rate": 6.981199694327024e-06, | |
| "loss": 0.3102, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.7421289355322339, | |
| "grad_norm": 0.9169371732698439, | |
| "learning_rate": 6.97038295148642e-06, | |
| "loss": 0.2642, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.7436281859070465, | |
| "grad_norm": 1.0612642906681782, | |
| "learning_rate": 6.959555280700162e-06, | |
| "loss": 0.2886, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.7451274362818591, | |
| "grad_norm": 1.150526665842185, | |
| "learning_rate": 6.948716742019616e-06, | |
| "loss": 0.2829, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.7466266866566716, | |
| "grad_norm": 0.9457216701757661, | |
| "learning_rate": 6.937867395556428e-06, | |
| "loss": 0.3038, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.7481259370314842, | |
| "grad_norm": 1.0340920013178942, | |
| "learning_rate": 6.927007301482187e-06, | |
| "loss": 0.2717, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.7496251874062968, | |
| "grad_norm": 0.8881450123848462, | |
| "learning_rate": 6.916136520028087e-06, | |
| "loss": 0.2857, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7511244377811095, | |
| "grad_norm": 1.1470400307795396, | |
| "learning_rate": 6.905255111484592e-06, | |
| "loss": 0.2863, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.7526236881559221, | |
| "grad_norm": 1.174630547427213, | |
| "learning_rate": 6.894363136201114e-06, | |
| "loss": 0.3559, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.7541229385307346, | |
| "grad_norm": 0.9363971361492817, | |
| "learning_rate": 6.88346065458566e-06, | |
| "loss": 0.3053, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.7556221889055472, | |
| "grad_norm": 0.9671898903458015, | |
| "learning_rate": 6.8725477271045085e-06, | |
| "loss": 0.3009, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.7571214392803598, | |
| "grad_norm": 1.0088148796498786, | |
| "learning_rate": 6.861624414281875e-06, | |
| "loss": 0.2679, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7586206896551724, | |
| "grad_norm": 0.9711527684116306, | |
| "learning_rate": 6.850690776699574e-06, | |
| "loss": 0.2542, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.760119940029985, | |
| "grad_norm": 1.1114718339400917, | |
| "learning_rate": 6.8397468749966735e-06, | |
| "loss": 0.3395, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.7616191904047976, | |
| "grad_norm": 1.0009352533756088, | |
| "learning_rate": 6.8287927698691745e-06, | |
| "loss": 0.3057, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.7631184407796102, | |
| "grad_norm": 0.9402317453266618, | |
| "learning_rate": 6.8178285220696686e-06, | |
| "loss": 0.3153, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.7646176911544228, | |
| "grad_norm": 0.9406762488113258, | |
| "learning_rate": 6.806854192406995e-06, | |
| "loss": 0.2342, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7661169415292354, | |
| "grad_norm": 1.1351742029907934, | |
| "learning_rate": 6.795869841745912e-06, | |
| "loss": 0.3025, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.767616191904048, | |
| "grad_norm": 1.032196821043251, | |
| "learning_rate": 6.784875531006751e-06, | |
| "loss": 0.3294, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.7691154422788605, | |
| "grad_norm": 0.7704191388002084, | |
| "learning_rate": 6.7738713211650885e-06, | |
| "loss": 0.2984, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.7706146926536732, | |
| "grad_norm": 1.0624828773134936, | |
| "learning_rate": 6.762857273251396e-06, | |
| "loss": 0.3134, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.7721139430284858, | |
| "grad_norm": 1.2062397195410002, | |
| "learning_rate": 6.751833448350713e-06, | |
| "loss": 0.2887, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7736131934032984, | |
| "grad_norm": 1.2518566120628374, | |
| "learning_rate": 6.740799907602302e-06, | |
| "loss": 0.2975, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.775112443778111, | |
| "grad_norm": 0.8919482008811496, | |
| "learning_rate": 6.729756712199309e-06, | |
| "loss": 0.2778, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.7766116941529235, | |
| "grad_norm": 1.0563404231488287, | |
| "learning_rate": 6.718703923388427e-06, | |
| "loss": 0.2322, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.7781109445277361, | |
| "grad_norm": 0.9249472647461907, | |
| "learning_rate": 6.707641602469554e-06, | |
| "loss": 0.2484, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.7796101949025487, | |
| "grad_norm": 1.1630248268569159, | |
| "learning_rate": 6.696569810795455e-06, | |
| "loss": 0.3154, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7811094452773614, | |
| "grad_norm": 1.0346170497851708, | |
| "learning_rate": 6.685488609771422e-06, | |
| "loss": 0.2892, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.782608695652174, | |
| "grad_norm": 1.204479314343301, | |
| "learning_rate": 6.674398060854931e-06, | |
| "loss": 0.325, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.7841079460269865, | |
| "grad_norm": 1.0135864684959381, | |
| "learning_rate": 6.6632982255553004e-06, | |
| "loss": 0.2664, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.7856071964017991, | |
| "grad_norm": 1.1922793225632307, | |
| "learning_rate": 6.652189165433356e-06, | |
| "loss": 0.3294, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.7871064467766117, | |
| "grad_norm": 0.9035726530442074, | |
| "learning_rate": 6.64107094210108e-06, | |
| "loss": 0.2547, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.7886056971514243, | |
| "grad_norm": 1.0969074812648842, | |
| "learning_rate": 6.62994361722128e-06, | |
| "loss": 0.3652, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.7901049475262368, | |
| "grad_norm": 0.9191696546741331, | |
| "learning_rate": 6.618807252507238e-06, | |
| "loss": 0.2686, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.7916041979010495, | |
| "grad_norm": 1.1711747620394066, | |
| "learning_rate": 6.6076619097223735e-06, | |
| "loss": 0.3799, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.7931034482758621, | |
| "grad_norm": 1.1957684656113474, | |
| "learning_rate": 6.5965076506799e-06, | |
| "loss": 0.3199, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.7946026986506747, | |
| "grad_norm": 1.153252525336776, | |
| "learning_rate": 6.5853445372424805e-06, | |
| "loss": 0.2982, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7961019490254873, | |
| "grad_norm": 1.1610514466336297, | |
| "learning_rate": 6.574172631321885e-06, | |
| "loss": 0.2919, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.7976011994002998, | |
| "grad_norm": 1.2286346778035668, | |
| "learning_rate": 6.562991994878649e-06, | |
| "loss": 0.3179, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.7991004497751124, | |
| "grad_norm": 1.1032693211440454, | |
| "learning_rate": 6.551802689921726e-06, | |
| "loss": 0.3109, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.800599700149925, | |
| "grad_norm": 0.866223625167243, | |
| "learning_rate": 6.5406047785081485e-06, | |
| "loss": 0.2876, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.8020989505247377, | |
| "grad_norm": 0.8524398206611851, | |
| "learning_rate": 6.529398322742677e-06, | |
| "loss": 0.2078, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8035982008995503, | |
| "grad_norm": 0.9347162108471077, | |
| "learning_rate": 6.518183384777468e-06, | |
| "loss": 0.2277, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.8050974512743628, | |
| "grad_norm": 1.1747947505172665, | |
| "learning_rate": 6.506960026811712e-06, | |
| "loss": 0.3357, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.8065967016491754, | |
| "grad_norm": 1.0052144225532982, | |
| "learning_rate": 6.495728311091303e-06, | |
| "loss": 0.2372, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.808095952023988, | |
| "grad_norm": 0.9519832529140138, | |
| "learning_rate": 6.484488299908487e-06, | |
| "loss": 0.2333, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.8095952023988006, | |
| "grad_norm": 1.2577025741697758, | |
| "learning_rate": 6.473240055601517e-06, | |
| "loss": 0.296, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8110944527736131, | |
| "grad_norm": 0.9415313714585996, | |
| "learning_rate": 6.46198364055431e-06, | |
| "loss": 0.3184, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.8125937031484258, | |
| "grad_norm": 1.0517679946737712, | |
| "learning_rate": 6.450719117196094e-06, | |
| "loss": 0.2778, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.8140929535232384, | |
| "grad_norm": 1.128729392426932, | |
| "learning_rate": 6.439446548001069e-06, | |
| "loss": 0.2776, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.815592203898051, | |
| "grad_norm": 1.2664066158308929, | |
| "learning_rate": 6.4281659954880605e-06, | |
| "loss": 0.3261, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.8170914542728636, | |
| "grad_norm": 1.1270962190283893, | |
| "learning_rate": 6.416877522220167e-06, | |
| "loss": 0.3219, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.8185907046476761, | |
| "grad_norm": 1.1897841192478287, | |
| "learning_rate": 6.405581190804418e-06, | |
| "loss": 0.3072, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.8200899550224887, | |
| "grad_norm": 1.0438751362356706, | |
| "learning_rate": 6.394277063891422e-06, | |
| "loss": 0.2967, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.8215892053973014, | |
| "grad_norm": 0.9965715839476298, | |
| "learning_rate": 6.382965204175027e-06, | |
| "loss": 0.3491, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.823088455772114, | |
| "grad_norm": 1.0445647944304888, | |
| "learning_rate": 6.371645674391967e-06, | |
| "loss": 0.2559, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.8245877061469266, | |
| "grad_norm": 1.0813890674681987, | |
| "learning_rate": 6.3603185373215105e-06, | |
| "loss": 0.2643, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8260869565217391, | |
| "grad_norm": 0.8883596314867054, | |
| "learning_rate": 6.348983855785122e-06, | |
| "loss": 0.2844, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 0.991305956934697, | |
| "learning_rate": 6.337641692646106e-06, | |
| "loss": 0.283, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.8290854572713643, | |
| "grad_norm": 0.7418670337953376, | |
| "learning_rate": 6.326292110809258e-06, | |
| "loss": 0.2494, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.8305847076461769, | |
| "grad_norm": 0.85809194200392, | |
| "learning_rate": 6.314935173220524e-06, | |
| "loss": 0.238, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.8320839580209896, | |
| "grad_norm": 0.9524026144666775, | |
| "learning_rate": 6.303570942866643e-06, | |
| "loss": 0.2945, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.8335832083958021, | |
| "grad_norm": 0.9748146825388692, | |
| "learning_rate": 6.2921994827748e-06, | |
| "loss": 0.2144, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.8350824587706147, | |
| "grad_norm": 1.0183447849879501, | |
| "learning_rate": 6.280820856012277e-06, | |
| "loss": 0.2961, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.8365817091454273, | |
| "grad_norm": 0.8661361222854527, | |
| "learning_rate": 6.269435125686105e-06, | |
| "loss": 0.2562, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.8380809595202399, | |
| "grad_norm": 1.0301329601149423, | |
| "learning_rate": 6.258042354942708e-06, | |
| "loss": 0.2991, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.8395802098950524, | |
| "grad_norm": 0.8394388405658721, | |
| "learning_rate": 6.2466426069675626e-06, | |
| "loss": 0.2302, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.841079460269865, | |
| "grad_norm": 0.9223149546934514, | |
| "learning_rate": 6.235235944984835e-06, | |
| "loss": 0.2397, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.8425787106446777, | |
| "grad_norm": 1.0811978772212658, | |
| "learning_rate": 6.223822432257043e-06, | |
| "loss": 0.305, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.8440779610194903, | |
| "grad_norm": 1.3240959511775114, | |
| "learning_rate": 6.212402132084697e-06, | |
| "loss": 0.3288, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.8455772113943029, | |
| "grad_norm": 1.1662227945725414, | |
| "learning_rate": 6.200975107805951e-06, | |
| "loss": 0.3307, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.8470764617691154, | |
| "grad_norm": 0.9574706807499568, | |
| "learning_rate": 6.189541422796254e-06, | |
| "loss": 0.2801, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.848575712143928, | |
| "grad_norm": 1.055894071591082, | |
| "learning_rate": 6.1781011404679905e-06, | |
| "loss": 0.3423, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.8500749625187406, | |
| "grad_norm": 0.9650583118449237, | |
| "learning_rate": 6.16665432427014e-06, | |
| "loss": 0.2938, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.8515742128935532, | |
| "grad_norm": 0.9803791071322102, | |
| "learning_rate": 6.155201037687917e-06, | |
| "loss": 0.3079, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.8530734632683659, | |
| "grad_norm": 0.8483489993796048, | |
| "learning_rate": 6.1437413442424236e-06, | |
| "loss": 0.2406, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.8545727136431784, | |
| "grad_norm": 0.9112718690173138, | |
| "learning_rate": 6.132275307490291e-06, | |
| "loss": 0.2521, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.856071964017991, | |
| "grad_norm": 1.0856580671000233, | |
| "learning_rate": 6.120802991023334e-06, | |
| "loss": 0.2727, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.8575712143928036, | |
| "grad_norm": 0.9605667003995878, | |
| "learning_rate": 6.109324458468198e-06, | |
| "loss": 0.2794, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.8590704647676162, | |
| "grad_norm": 1.085453793478151, | |
| "learning_rate": 6.097839773485995e-06, | |
| "loss": 0.27, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.8605697151424287, | |
| "grad_norm": 1.046946540706997, | |
| "learning_rate": 6.086348999771967e-06, | |
| "loss": 0.3219, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 0.9883922397993943, | |
| "learning_rate": 6.074852201055121e-06, | |
| "loss": 0.2508, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.863568215892054, | |
| "grad_norm": 0.9172005793691745, | |
| "learning_rate": 6.063349441097881e-06, | |
| "loss": 0.2528, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.8650674662668666, | |
| "grad_norm": 1.0656160209547039, | |
| "learning_rate": 6.051840783695731e-06, | |
| "loss": 0.2437, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.8665667166416792, | |
| "grad_norm": 1.1671320044121045, | |
| "learning_rate": 6.040326292676865e-06, | |
| "loss": 0.2463, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.8680659670164917, | |
| "grad_norm": 1.2655266376268557, | |
| "learning_rate": 6.028806031901829e-06, | |
| "loss": 0.3669, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 0.9719904940425894, | |
| "learning_rate": 6.0172800652631706e-06, | |
| "loss": 0.2728, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8710644677661169, | |
| "grad_norm": 1.0892722055553172, | |
| "learning_rate": 6.005748456685077e-06, | |
| "loss": 0.2376, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.8725637181409296, | |
| "grad_norm": 1.035562821103207, | |
| "learning_rate": 5.994211270123034e-06, | |
| "loss": 0.3403, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.8740629685157422, | |
| "grad_norm": 1.1213662649850735, | |
| "learning_rate": 5.9826685695634575e-06, | |
| "loss": 0.2589, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.8755622188905547, | |
| "grad_norm": 1.0051099052650143, | |
| "learning_rate": 5.971120419023349e-06, | |
| "loss": 0.2929, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.8770614692653673, | |
| "grad_norm": 0.6954031352070463, | |
| "learning_rate": 5.959566882549936e-06, | |
| "loss": 0.2423, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.8785607196401799, | |
| "grad_norm": 1.1147852546784172, | |
| "learning_rate": 5.948008024220311e-06, | |
| "loss": 0.2583, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.8800599700149925, | |
| "grad_norm": 0.9862220866148104, | |
| "learning_rate": 5.936443908141088e-06, | |
| "loss": 0.2483, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.881559220389805, | |
| "grad_norm": 0.9575695836019265, | |
| "learning_rate": 5.924874598448038e-06, | |
| "loss": 0.2765, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.8830584707646177, | |
| "grad_norm": 0.8427218231893465, | |
| "learning_rate": 5.913300159305741e-06, | |
| "loss": 0.2865, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.8845577211394303, | |
| "grad_norm": 0.9177492818808969, | |
| "learning_rate": 5.901720654907217e-06, | |
| "loss": 0.289, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8860569715142429, | |
| "grad_norm": 0.9650773856831488, | |
| "learning_rate": 5.8901361494735874e-06, | |
| "loss": 0.2771, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.8875562218890555, | |
| "grad_norm": 1.3303441589254317, | |
| "learning_rate": 5.878546707253704e-06, | |
| "loss": 0.331, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.889055472263868, | |
| "grad_norm": 1.129456164359237, | |
| "learning_rate": 5.8669523925238e-06, | |
| "loss": 0.3231, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.8905547226386806, | |
| "grad_norm": 0.9307322018834451, | |
| "learning_rate": 5.855353269587134e-06, | |
| "loss": 0.2414, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.8920539730134932, | |
| "grad_norm": 0.977281101652398, | |
| "learning_rate": 5.843749402773629e-06, | |
| "loss": 0.2845, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.8935532233883059, | |
| "grad_norm": 1.0565292867002212, | |
| "learning_rate": 5.8321408564395165e-06, | |
| "loss": 0.2517, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.8950524737631185, | |
| "grad_norm": 1.074899708276676, | |
| "learning_rate": 5.820527694966988e-06, | |
| "loss": 0.278, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.896551724137931, | |
| "grad_norm": 1.0158517491903019, | |
| "learning_rate": 5.808909982763825e-06, | |
| "loss": 0.3239, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.8980509745127436, | |
| "grad_norm": 0.9192638176807242, | |
| "learning_rate": 5.797287784263047e-06, | |
| "loss": 0.2891, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.8995502248875562, | |
| "grad_norm": 1.1151114107566769, | |
| "learning_rate": 5.785661163922558e-06, | |
| "loss": 0.3223, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8995502248875562, | |
| "eval_loss": 0.28777456283569336, | |
| "eval_runtime": 9.4949, | |
| "eval_samples_per_second": 5.687, | |
| "eval_steps_per_second": 1.474, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9010494752623688, | |
| "grad_norm": 1.2254228669170264, | |
| "learning_rate": 5.774030186224786e-06, | |
| "loss": 0.3074, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.9025487256371814, | |
| "grad_norm": 0.8505541761356764, | |
| "learning_rate": 5.762394915676325e-06, | |
| "loss": 0.2336, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.904047976011994, | |
| "grad_norm": 1.0455012039938063, | |
| "learning_rate": 5.750755416807575e-06, | |
| "loss": 0.2709, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.9055472263868066, | |
| "grad_norm": 1.1046394655362906, | |
| "learning_rate": 5.7391117541723914e-06, | |
| "loss": 0.2305, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.9070464767616192, | |
| "grad_norm": 1.152406358782211, | |
| "learning_rate": 5.727463992347719e-06, | |
| "loss": 0.326, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.9085457271364318, | |
| "grad_norm": 0.9182748286736299, | |
| "learning_rate": 5.715812195933238e-06, | |
| "loss": 0.2452, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.9100449775112444, | |
| "grad_norm": 1.1811655823883245, | |
| "learning_rate": 5.704156429551004e-06, | |
| "loss": 0.35, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.9115442278860569, | |
| "grad_norm": 0.8950134493853912, | |
| "learning_rate": 5.692496757845092e-06, | |
| "loss": 0.2086, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.9130434782608695, | |
| "grad_norm": 1.0731918918959729, | |
| "learning_rate": 5.680833245481234e-06, | |
| "loss": 0.2781, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.9145427286356822, | |
| "grad_norm": 1.2636083995385172, | |
| "learning_rate": 5.6691659571464655e-06, | |
| "loss": 0.3695, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9160419790104948, | |
| "grad_norm": 0.928883601034792, | |
| "learning_rate": 5.657494957548761e-06, | |
| "loss": 0.2902, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.9175412293853074, | |
| "grad_norm": 1.1393344052730445, | |
| "learning_rate": 5.645820311416681e-06, | |
| "loss": 0.2214, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.9190404797601199, | |
| "grad_norm": 0.9557046663623131, | |
| "learning_rate": 5.63414208349901e-06, | |
| "loss": 0.2981, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.9205397301349325, | |
| "grad_norm": 1.0170197626712119, | |
| "learning_rate": 5.622460338564393e-06, | |
| "loss": 0.2618, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.9220389805097451, | |
| "grad_norm": 1.0705882882051736, | |
| "learning_rate": 5.610775141400986e-06, | |
| "loss": 0.247, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9235382308845578, | |
| "grad_norm": 1.0394998754597833, | |
| "learning_rate": 5.599086556816089e-06, | |
| "loss": 0.2852, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.9250374812593704, | |
| "grad_norm": 0.9128678257675251, | |
| "learning_rate": 5.587394649635789e-06, | |
| "loss": 0.2904, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.9265367316341829, | |
| "grad_norm": 1.2368076457030965, | |
| "learning_rate": 5.575699484704599e-06, | |
| "loss": 0.207, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.9280359820089955, | |
| "grad_norm": 1.027604231085994, | |
| "learning_rate": 5.564001126885106e-06, | |
| "loss": 0.2532, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.9295352323838081, | |
| "grad_norm": 1.0364130967093776, | |
| "learning_rate": 5.552299641057596e-06, | |
| "loss": 0.2943, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9310344827586207, | |
| "grad_norm": 1.1817202552018278, | |
| "learning_rate": 5.540595092119709e-06, | |
| "loss": 0.2658, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.9325337331334332, | |
| "grad_norm": 1.1252518250208885, | |
| "learning_rate": 5.5288875449860745e-06, | |
| "loss": 0.2684, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.9340329835082459, | |
| "grad_norm": 0.9999737050282641, | |
| "learning_rate": 5.517177064587945e-06, | |
| "loss": 0.235, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.9355322338830585, | |
| "grad_norm": 0.8269034350245899, | |
| "learning_rate": 5.505463715872846e-06, | |
| "loss": 0.2354, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.9370314842578711, | |
| "grad_norm": 1.1075137496530656, | |
| "learning_rate": 5.493747563804211e-06, | |
| "loss": 0.2736, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9385307346326837, | |
| "grad_norm": 0.8073540744562585, | |
| "learning_rate": 5.482028673361015e-06, | |
| "loss": 0.2586, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.9400299850074962, | |
| "grad_norm": 0.9987830124271582, | |
| "learning_rate": 5.470307109537427e-06, | |
| "loss": 0.2835, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.9415292353823088, | |
| "grad_norm": 1.0215954818134905, | |
| "learning_rate": 5.45858293734244e-06, | |
| "loss": 0.2342, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.9430284857571214, | |
| "grad_norm": 1.1840545077768232, | |
| "learning_rate": 5.446856221799515e-06, | |
| "loss": 0.3115, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.9445277361319341, | |
| "grad_norm": 1.1558163016089784, | |
| "learning_rate": 5.435127027946215e-06, | |
| "loss": 0.285, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9460269865067467, | |
| "grad_norm": 0.957684055708224, | |
| "learning_rate": 5.423395420833853e-06, | |
| "loss": 0.2705, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.9475262368815592, | |
| "grad_norm": 1.1707757272927524, | |
| "learning_rate": 5.411661465527123e-06, | |
| "loss": 0.3201, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.9490254872563718, | |
| "grad_norm": 1.045488273709154, | |
| "learning_rate": 5.39992522710374e-06, | |
| "loss": 0.2875, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.9505247376311844, | |
| "grad_norm": 1.140860854539818, | |
| "learning_rate": 5.38818677065409e-06, | |
| "loss": 0.3184, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.952023988005997, | |
| "grad_norm": 0.9920977379605783, | |
| "learning_rate": 5.376446161280851e-06, | |
| "loss": 0.2496, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9535232383808095, | |
| "grad_norm": 0.9812850718918373, | |
| "learning_rate": 5.364703464098645e-06, | |
| "loss": 0.31, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.9550224887556222, | |
| "grad_norm": 1.0487194387940328, | |
| "learning_rate": 5.352958744233673e-06, | |
| "loss": 0.2697, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.9565217391304348, | |
| "grad_norm": 0.9528359479897234, | |
| "learning_rate": 5.341212066823356e-06, | |
| "loss": 0.3022, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.9580209895052474, | |
| "grad_norm": 1.0248593254124527, | |
| "learning_rate": 5.329463497015969e-06, | |
| "loss": 0.2414, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.95952023988006, | |
| "grad_norm": 1.2417760992345999, | |
| "learning_rate": 5.317713099970283e-06, | |
| "loss": 0.2767, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.9610194902548725, | |
| "grad_norm": 0.7303624666250885, | |
| "learning_rate": 5.305960940855205e-06, | |
| "loss": 0.2186, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.9625187406296851, | |
| "grad_norm": 1.0735220587956722, | |
| "learning_rate": 5.294207084849412e-06, | |
| "loss": 0.287, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.9640179910044977, | |
| "grad_norm": 0.9496961389550538, | |
| "learning_rate": 5.282451597140994e-06, | |
| "loss": 0.2663, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.9655172413793104, | |
| "grad_norm": 1.2683004294901852, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.3536, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.967016491754123, | |
| "grad_norm": 1.2463881053905386, | |
| "learning_rate": 5.258935987413524e-06, | |
| "loss": 0.3089, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.9685157421289355, | |
| "grad_norm": 0.8953281287708827, | |
| "learning_rate": 5.247175995814452e-06, | |
| "loss": 0.2783, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.9700149925037481, | |
| "grad_norm": 1.1398548481957569, | |
| "learning_rate": 5.235414633351992e-06, | |
| "loss": 0.297, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.9715142428785607, | |
| "grad_norm": 0.9522912625328807, | |
| "learning_rate": 5.223651965255864e-06, | |
| "loss": 0.2722, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.9730134932533733, | |
| "grad_norm": 1.0049670204208179, | |
| "learning_rate": 5.211888056763029e-06, | |
| "loss": 0.22, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.974512743628186, | |
| "grad_norm": 1.0259712557992315, | |
| "learning_rate": 5.20012297311733e-06, | |
| "loss": 0.2659, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9760119940029985, | |
| "grad_norm": 0.7262636727090428, | |
| "learning_rate": 5.188356779569125e-06, | |
| "loss": 0.2625, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.9775112443778111, | |
| "grad_norm": 1.4512773361514966, | |
| "learning_rate": 5.176589541374929e-06, | |
| "loss": 0.3188, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.9790104947526237, | |
| "grad_norm": 1.1351582744460824, | |
| "learning_rate": 5.164821323797051e-06, | |
| "loss": 0.3221, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.9805097451274363, | |
| "grad_norm": 0.9766011096025039, | |
| "learning_rate": 5.1530521921032305e-06, | |
| "loss": 0.2631, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.9820089955022488, | |
| "grad_norm": 1.0077384368629463, | |
| "learning_rate": 5.141282211566276e-06, | |
| "loss": 0.2692, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.9835082458770614, | |
| "grad_norm": 1.2290746951327147, | |
| "learning_rate": 5.129511447463705e-06, | |
| "loss": 0.2896, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.9850074962518741, | |
| "grad_norm": 1.1369111174772528, | |
| "learning_rate": 5.117739965077382e-06, | |
| "loss": 0.2705, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.9865067466266867, | |
| "grad_norm": 1.145143090138757, | |
| "learning_rate": 5.105967829693155e-06, | |
| "loss": 0.3, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.9880059970014993, | |
| "grad_norm": 1.0815670091882013, | |
| "learning_rate": 5.0941951066004906e-06, | |
| "loss": 0.2953, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.9895052473763118, | |
| "grad_norm": 1.0811618184237886, | |
| "learning_rate": 5.082421861092116e-06, | |
| "loss": 0.3445, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9910044977511244, | |
| "grad_norm": 0.9863756233276578, | |
| "learning_rate": 5.0706481584636605e-06, | |
| "loss": 0.2929, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.992503748125937, | |
| "grad_norm": 0.9073066580362485, | |
| "learning_rate": 5.0588740640132805e-06, | |
| "loss": 0.3171, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.9940029985007496, | |
| "grad_norm": 0.9710643248812654, | |
| "learning_rate": 5.047099643041312e-06, | |
| "loss": 0.2333, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.9955022488755623, | |
| "grad_norm": 0.8405785079417468, | |
| "learning_rate": 5.0353249608499e-06, | |
| "loss": 0.2736, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.9970014992503748, | |
| "grad_norm": 1.1118666345486943, | |
| "learning_rate": 5.023550082742637e-06, | |
| "loss": 0.4096, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.9985007496251874, | |
| "grad_norm": 1.171357402016781, | |
| "learning_rate": 5.011775074024202e-06, | |
| "loss": 0.3399, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.907880575212693, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2158, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.0014992503748126, | |
| "grad_norm": 0.903288235080705, | |
| "learning_rate": 4.988224925975799e-06, | |
| "loss": 0.159, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.0029985007496252, | |
| "grad_norm": 0.8455644147047462, | |
| "learning_rate": 4.976449917257365e-06, | |
| "loss": 0.1324, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.0044977511244377, | |
| "grad_norm": 0.7722192119293851, | |
| "learning_rate": 4.964675039150102e-06, | |
| "loss": 0.1521, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0059970014992503, | |
| "grad_norm": 0.7703151140362237, | |
| "learning_rate": 4.952900356958689e-06, | |
| "loss": 0.1494, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.0074962518740629, | |
| "grad_norm": 1.0121713234206255, | |
| "learning_rate": 4.941125935986721e-06, | |
| "loss": 0.166, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.0089955022488755, | |
| "grad_norm": 0.8169746487453988, | |
| "learning_rate": 4.929351841536342e-06, | |
| "loss": 0.1381, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.0104947526236883, | |
| "grad_norm": 0.7423369594543408, | |
| "learning_rate": 4.917578138907884e-06, | |
| "loss": 0.1605, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.0119940029985008, | |
| "grad_norm": 1.0678067171777428, | |
| "learning_rate": 4.90580489339951e-06, | |
| "loss": 0.1766, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.0134932533733134, | |
| "grad_norm": 0.9947629083272405, | |
| "learning_rate": 4.894032170306846e-06, | |
| "loss": 0.1613, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.014992503748126, | |
| "grad_norm": 0.8053764526381059, | |
| "learning_rate": 4.882260034922618e-06, | |
| "loss": 0.1429, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.0164917541229386, | |
| "grad_norm": 0.8052641264230386, | |
| "learning_rate": 4.870488552536296e-06, | |
| "loss": 0.1293, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.0179910044977512, | |
| "grad_norm": 0.9312084337461259, | |
| "learning_rate": 4.858717788433725e-06, | |
| "loss": 0.1943, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.0194902548725637, | |
| "grad_norm": 0.8127838993145354, | |
| "learning_rate": 4.846947807896771e-06, | |
| "loss": 0.1574, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0209895052473763, | |
| "grad_norm": 1.0237113590827869, | |
| "learning_rate": 4.83517867620295e-06, | |
| "loss": 0.1632, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.0224887556221889, | |
| "grad_norm": 0.6258418604692355, | |
| "learning_rate": 4.823410458625072e-06, | |
| "loss": 0.1044, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.0239880059970015, | |
| "grad_norm": 0.782774697497071, | |
| "learning_rate": 4.811643220430877e-06, | |
| "loss": 0.1286, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.025487256371814, | |
| "grad_norm": 1.0528765846956039, | |
| "learning_rate": 4.7998770268826726e-06, | |
| "loss": 0.1999, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.0269865067466266, | |
| "grad_norm": 0.9693999574177994, | |
| "learning_rate": 4.788111943236973e-06, | |
| "loss": 0.1384, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.0284857571214392, | |
| "grad_norm": 0.806263505006784, | |
| "learning_rate": 4.7763480347441395e-06, | |
| "loss": 0.1635, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.0299850074962518, | |
| "grad_norm": 0.8137390544275231, | |
| "learning_rate": 4.7645853666480104e-06, | |
| "loss": 0.1324, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.0314842578710646, | |
| "grad_norm": 0.9073657173761188, | |
| "learning_rate": 4.752824004185548e-06, | |
| "loss": 0.1665, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.0329835082458771, | |
| "grad_norm": 0.7818222327297528, | |
| "learning_rate": 4.7410640125864785e-06, | |
| "loss": 0.0989, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.0344827586206897, | |
| "grad_norm": 1.006175356949153, | |
| "learning_rate": 4.729305457072913e-06, | |
| "loss": 0.192, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0359820089955023, | |
| "grad_norm": 0.8139309092657723, | |
| "learning_rate": 4.717548402859008e-06, | |
| "loss": 0.1552, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.0374812593703149, | |
| "grad_norm": 0.7967421278417623, | |
| "learning_rate": 4.7057929151505895e-06, | |
| "loss": 0.1742, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.0389805097451275, | |
| "grad_norm": 0.9412389510453285, | |
| "learning_rate": 4.694039059144797e-06, | |
| "loss": 0.1671, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.04047976011994, | |
| "grad_norm": 0.8630351370957148, | |
| "learning_rate": 4.6822869000297185e-06, | |
| "loss": 0.1324, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.0419790104947526, | |
| "grad_norm": 0.9675934541224243, | |
| "learning_rate": 4.670536502984033e-06, | |
| "loss": 0.1257, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 0.8823328849658667, | |
| "learning_rate": 4.6587879331766465e-06, | |
| "loss": 0.1765, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.0449775112443778, | |
| "grad_norm": 0.9065592863648484, | |
| "learning_rate": 4.647041255766329e-06, | |
| "loss": 0.1575, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.0464767616191903, | |
| "grad_norm": 0.862555008922206, | |
| "learning_rate": 4.6352965359013576e-06, | |
| "loss": 0.1607, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.047976011994003, | |
| "grad_norm": 0.8631189398370362, | |
| "learning_rate": 4.623553838719151e-06, | |
| "loss": 0.1474, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.0494752623688155, | |
| "grad_norm": 0.875484397556485, | |
| "learning_rate": 4.611813229345911e-06, | |
| "loss": 0.1322, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.050974512743628, | |
| "grad_norm": 0.8642434645372525, | |
| "learning_rate": 4.6000747728962606e-06, | |
| "loss": 0.1584, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.0524737631184409, | |
| "grad_norm": 0.9084418475787728, | |
| "learning_rate": 4.588338534472878e-06, | |
| "loss": 0.1643, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.0539730134932535, | |
| "grad_norm": 0.9688956367918491, | |
| "learning_rate": 4.576604579166147e-06, | |
| "loss": 0.177, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.055472263868066, | |
| "grad_norm": 0.8985437432083577, | |
| "learning_rate": 4.564872972053786e-06, | |
| "loss": 0.1374, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.0569715142428786, | |
| "grad_norm": 0.9810117230123548, | |
| "learning_rate": 4.553143778200486e-06, | |
| "loss": 0.2164, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.0584707646176912, | |
| "grad_norm": 0.9267853389422427, | |
| "learning_rate": 4.541417062657561e-06, | |
| "loss": 0.1298, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.0599700149925038, | |
| "grad_norm": 1.0064302398901162, | |
| "learning_rate": 4.529692890462574e-06, | |
| "loss": 0.1741, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.0614692653673163, | |
| "grad_norm": 0.8452874310085142, | |
| "learning_rate": 4.5179713266389866e-06, | |
| "loss": 0.1403, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.062968515742129, | |
| "grad_norm": 0.809705149717974, | |
| "learning_rate": 4.50625243619579e-06, | |
| "loss": 0.1134, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.0644677661169415, | |
| "grad_norm": 0.9645694950112623, | |
| "learning_rate": 4.494536284127155e-06, | |
| "loss": 0.1695, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.065967016491754, | |
| "grad_norm": 0.911540879662766, | |
| "learning_rate": 4.4828229354120565e-06, | |
| "loss": 0.1266, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.0674662668665666, | |
| "grad_norm": 1.043231692279107, | |
| "learning_rate": 4.471112455013928e-06, | |
| "loss": 0.146, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.0689655172413792, | |
| "grad_norm": 0.9077689468353303, | |
| "learning_rate": 4.459404907880293e-06, | |
| "loss": 0.1449, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.0704647676161918, | |
| "grad_norm": 1.1799425689034388, | |
| "learning_rate": 4.447700358942407e-06, | |
| "loss": 0.2251, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.0719640179910046, | |
| "grad_norm": 0.9946791183612375, | |
| "learning_rate": 4.435998873114895e-06, | |
| "loss": 0.1424, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.0734632683658172, | |
| "grad_norm": 1.0239758994003485, | |
| "learning_rate": 4.424300515295401e-06, | |
| "loss": 0.1647, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.0749625187406298, | |
| "grad_norm": 1.0084531347013064, | |
| "learning_rate": 4.412605350364213e-06, | |
| "loss": 0.1585, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.0764617691154423, | |
| "grad_norm": 1.019785813305053, | |
| "learning_rate": 4.400913443183913e-06, | |
| "loss": 0.1824, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.077961019490255, | |
| "grad_norm": 1.0841656107993747, | |
| "learning_rate": 4.389224858599015e-06, | |
| "loss": 0.161, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.0794602698650675, | |
| "grad_norm": 0.7392996097131955, | |
| "learning_rate": 4.377539661435608e-06, | |
| "loss": 0.1036, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.08095952023988, | |
| "grad_norm": 0.9602461836338625, | |
| "learning_rate": 4.365857916500991e-06, | |
| "loss": 0.1214, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.0824587706146926, | |
| "grad_norm": 0.9524356165789148, | |
| "learning_rate": 4.35417968858332e-06, | |
| "loss": 0.2164, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.0839580209895052, | |
| "grad_norm": 0.8070383767113144, | |
| "learning_rate": 4.3425050424512405e-06, | |
| "loss": 0.1398, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.0854572713643178, | |
| "grad_norm": 1.0600840461442933, | |
| "learning_rate": 4.330834042853537e-06, | |
| "loss": 0.1669, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.0869565217391304, | |
| "grad_norm": 1.0724650304080148, | |
| "learning_rate": 4.319166754518768e-06, | |
| "loss": 0.1531, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.088455772113943, | |
| "grad_norm": 0.9740329731274233, | |
| "learning_rate": 4.30750324215491e-06, | |
| "loss": 0.1334, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.0899550224887555, | |
| "grad_norm": 1.12248392530442, | |
| "learning_rate": 4.295843570448998e-06, | |
| "loss": 0.1806, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.0914542728635683, | |
| "grad_norm": 0.802680971813453, | |
| "learning_rate": 4.284187804066764e-06, | |
| "loss": 0.1462, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.092953523238381, | |
| "grad_norm": 0.9029062822763915, | |
| "learning_rate": 4.272536007652281e-06, | |
| "loss": 0.1456, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.0944527736131935, | |
| "grad_norm": 0.8723804658338138, | |
| "learning_rate": 4.260888245827608e-06, | |
| "loss": 0.1256, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.095952023988006, | |
| "grad_norm": 0.927697513395748, | |
| "learning_rate": 4.249244583192425e-06, | |
| "loss": 0.1989, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.0974512743628186, | |
| "grad_norm": 0.8762618026652915, | |
| "learning_rate": 4.237605084323676e-06, | |
| "loss": 0.159, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.0989505247376312, | |
| "grad_norm": 1.062849675512425, | |
| "learning_rate": 4.225969813775215e-06, | |
| "loss": 0.1418, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.1004497751124438, | |
| "grad_norm": 0.8732009834129478, | |
| "learning_rate": 4.214338836077444e-06, | |
| "loss": 0.1508, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.1019490254872564, | |
| "grad_norm": 1.0317602135685793, | |
| "learning_rate": 4.202712215736955e-06, | |
| "loss": 0.2059, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.103448275862069, | |
| "grad_norm": 1.0549722224624556, | |
| "learning_rate": 4.191090017236177e-06, | |
| "loss": 0.1853, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.1049475262368815, | |
| "grad_norm": 0.8921453002130978, | |
| "learning_rate": 4.1794723050330125e-06, | |
| "loss": 0.1577, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.106446776611694, | |
| "grad_norm": 0.8125601882549341, | |
| "learning_rate": 4.167859143560484e-06, | |
| "loss": 0.1453, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.1079460269865067, | |
| "grad_norm": 0.8063729152064001, | |
| "learning_rate": 4.1562505972263735e-06, | |
| "loss": 0.1144, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.1094452773613193, | |
| "grad_norm": 0.9285730592952013, | |
| "learning_rate": 4.144646730412868e-06, | |
| "loss": 0.1668, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1109445277361318, | |
| "grad_norm": 1.0633621229588555, | |
| "learning_rate": 4.133047607476202e-06, | |
| "loss": 0.1729, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.1124437781109444, | |
| "grad_norm": 0.7834681787835724, | |
| "learning_rate": 4.121453292746297e-06, | |
| "loss": 0.1783, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.1139430284857572, | |
| "grad_norm": 0.8123535333277304, | |
| "learning_rate": 4.109863850526413e-06, | |
| "loss": 0.1653, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.1154422788605698, | |
| "grad_norm": 0.9830924525678609, | |
| "learning_rate": 4.098279345092783e-06, | |
| "loss": 0.1643, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.1169415292353824, | |
| "grad_norm": 0.8698718364015038, | |
| "learning_rate": 4.086699840694262e-06, | |
| "loss": 0.1187, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.118440779610195, | |
| "grad_norm": 0.8756702713666746, | |
| "learning_rate": 4.075125401551963e-06, | |
| "loss": 0.1624, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.1199400299850075, | |
| "grad_norm": 0.9538752407632045, | |
| "learning_rate": 4.063556091858914e-06, | |
| "loss": 0.1306, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.12143928035982, | |
| "grad_norm": 1.0308038525560326, | |
| "learning_rate": 4.051991975779691e-06, | |
| "loss": 0.1598, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.1229385307346327, | |
| "grad_norm": 1.0713675201781632, | |
| "learning_rate": 4.040433117450066e-06, | |
| "loss": 0.1889, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.1244377811094453, | |
| "grad_norm": 1.0518237687576573, | |
| "learning_rate": 4.0288795809766516e-06, | |
| "loss": 0.1856, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1259370314842578, | |
| "grad_norm": 0.8093721467145427, | |
| "learning_rate": 4.017331430436543e-06, | |
| "loss": 0.1273, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.1274362818590704, | |
| "grad_norm": 1.0273922062183876, | |
| "learning_rate": 4.005788729876968e-06, | |
| "loss": 0.2022, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.128935532233883, | |
| "grad_norm": 0.8375450796557817, | |
| "learning_rate": 3.994251543314925e-06, | |
| "loss": 0.1364, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.1304347826086956, | |
| "grad_norm": 0.8805338907201848, | |
| "learning_rate": 3.982719934736832e-06, | |
| "loss": 0.1449, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.1319340329835081, | |
| "grad_norm": 0.975464378419334, | |
| "learning_rate": 3.971193968098172e-06, | |
| "loss": 0.1885, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.133433283358321, | |
| "grad_norm": 0.8807162429905603, | |
| "learning_rate": 3.959673707323135e-06, | |
| "loss": 0.219, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.1349325337331335, | |
| "grad_norm": 0.820086929154, | |
| "learning_rate": 3.948159216304269e-06, | |
| "loss": 0.1463, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.136431784107946, | |
| "grad_norm": 1.0652754470421382, | |
| "learning_rate": 3.93665055890212e-06, | |
| "loss": 0.1537, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.1379310344827587, | |
| "grad_norm": 1.1229175153818756, | |
| "learning_rate": 3.92514779894488e-06, | |
| "loss": 0.138, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.1394302848575713, | |
| "grad_norm": 0.8960720209614561, | |
| "learning_rate": 3.9136510002280344e-06, | |
| "loss": 0.2064, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1409295352323838, | |
| "grad_norm": 0.9309987651062527, | |
| "learning_rate": 3.902160226514007e-06, | |
| "loss": 0.1696, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.1424287856071964, | |
| "grad_norm": 0.8687713185728801, | |
| "learning_rate": 3.8906755415318045e-06, | |
| "loss": 0.17, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.143928035982009, | |
| "grad_norm": 0.8211547268507953, | |
| "learning_rate": 3.8791970089766665e-06, | |
| "loss": 0.141, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.1454272863568216, | |
| "grad_norm": 0.8489226302459136, | |
| "learning_rate": 3.86772469250971e-06, | |
| "loss": 0.1585, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.1469265367316341, | |
| "grad_norm": 0.9045519199043318, | |
| "learning_rate": 3.856258655757578e-06, | |
| "loss": 0.1043, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.1484257871064467, | |
| "grad_norm": 0.9629268454673666, | |
| "learning_rate": 3.844798962312085e-06, | |
| "loss": 0.1819, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.1499250374812593, | |
| "grad_norm": 0.8687753311561689, | |
| "learning_rate": 3.833345675729863e-06, | |
| "loss": 0.1433, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.1514242878560719, | |
| "grad_norm": 0.9534783014620999, | |
| "learning_rate": 3.821898859532013e-06, | |
| "loss": 0.1357, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.1529235382308847, | |
| "grad_norm": 0.9070860485726895, | |
| "learning_rate": 3.8104585772037493e-06, | |
| "loss": 0.1495, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.1544227886056972, | |
| "grad_norm": 0.9054763456293431, | |
| "learning_rate": 3.7990248921940485e-06, | |
| "loss": 0.166, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.1559220389805098, | |
| "grad_norm": 0.8045228646756142, | |
| "learning_rate": 3.787597867915303e-06, | |
| "loss": 0.1379, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.1574212893553224, | |
| "grad_norm": 1.0123768169928782, | |
| "learning_rate": 3.7761775677429567e-06, | |
| "loss": 0.1616, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.158920539730135, | |
| "grad_norm": 0.8771201233764615, | |
| "learning_rate": 3.7647640550151666e-06, | |
| "loss": 0.1265, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.1604197901049476, | |
| "grad_norm": 0.88971114729701, | |
| "learning_rate": 3.7533573930324395e-06, | |
| "loss": 0.135, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.1619190404797601, | |
| "grad_norm": 0.946750723434636, | |
| "learning_rate": 3.7419576450572924e-06, | |
| "loss": 0.155, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.1634182908545727, | |
| "grad_norm": 0.8730338646552244, | |
| "learning_rate": 3.7305648743138966e-06, | |
| "loss": 0.1601, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.1649175412293853, | |
| "grad_norm": 1.1081553311157504, | |
| "learning_rate": 3.7191791439877236e-06, | |
| "loss": 0.1591, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.1664167916041979, | |
| "grad_norm": 0.8996933242764917, | |
| "learning_rate": 3.7078005172252015e-06, | |
| "loss": 0.1649, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.1679160419790104, | |
| "grad_norm": 0.9354851933993101, | |
| "learning_rate": 3.6964290571333583e-06, | |
| "loss": 0.1467, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.169415292353823, | |
| "grad_norm": 1.0934560459547265, | |
| "learning_rate": 3.6850648267794776e-06, | |
| "loss": 0.1643, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.1709145427286356, | |
| "grad_norm": 0.935069222014449, | |
| "learning_rate": 3.673707889190744e-06, | |
| "loss": 0.1061, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.1724137931034484, | |
| "grad_norm": 0.942971929012004, | |
| "learning_rate": 3.662358307353897e-06, | |
| "loss": 0.133, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.1739130434782608, | |
| "grad_norm": 0.8489133393720603, | |
| "learning_rate": 3.6510161442148783e-06, | |
| "loss": 0.1279, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.1754122938530736, | |
| "grad_norm": 0.8205529892197596, | |
| "learning_rate": 3.63968146267849e-06, | |
| "loss": 0.1563, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.1769115442278861, | |
| "grad_norm": 0.7153571156208284, | |
| "learning_rate": 3.6283543256080334e-06, | |
| "loss": 0.1348, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.1784107946026987, | |
| "grad_norm": 0.9070169382427566, | |
| "learning_rate": 3.6170347958249728e-06, | |
| "loss": 0.1583, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.1799100449775113, | |
| "grad_norm": 0.8592443718739954, | |
| "learning_rate": 3.605722936108579e-06, | |
| "loss": 0.1423, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.1814092953523239, | |
| "grad_norm": 1.0583141897989798, | |
| "learning_rate": 3.5944188091955843e-06, | |
| "loss": 0.1324, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.1829085457271364, | |
| "grad_norm": 0.7864406323167323, | |
| "learning_rate": 3.5831224777798346e-06, | |
| "loss": 0.124, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.184407796101949, | |
| "grad_norm": 0.8784850159977831, | |
| "learning_rate": 3.5718340045119416e-06, | |
| "loss": 0.1584, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.1859070464767616, | |
| "grad_norm": 1.0625611327746805, | |
| "learning_rate": 3.5605534519989327e-06, | |
| "loss": 0.1417, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.1874062968515742, | |
| "grad_norm": 0.9119774046690301, | |
| "learning_rate": 3.5492808828039083e-06, | |
| "loss": 0.1083, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.1889055472263867, | |
| "grad_norm": 0.8420821007340418, | |
| "learning_rate": 3.538016359445692e-06, | |
| "loss": 0.1309, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.1904047976011993, | |
| "grad_norm": 0.8796317342709745, | |
| "learning_rate": 3.5267599443984848e-06, | |
| "loss": 0.1902, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.191904047976012, | |
| "grad_norm": 1.0896283213687823, | |
| "learning_rate": 3.5155117000915153e-06, | |
| "loss": 0.1796, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.1934032983508245, | |
| "grad_norm": 0.7903853753446982, | |
| "learning_rate": 3.5042716889086998e-06, | |
| "loss": 0.1447, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.1949025487256373, | |
| "grad_norm": 0.8935142454203702, | |
| "learning_rate": 3.493039973188289e-06, | |
| "loss": 0.12, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.1964017991004499, | |
| "grad_norm": 0.9493540792674124, | |
| "learning_rate": 3.481816615222533e-06, | |
| "loss": 0.203, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.1979010494752624, | |
| "grad_norm": 0.7522519356032099, | |
| "learning_rate": 3.470601677257323e-06, | |
| "loss": 0.1461, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.199400299850075, | |
| "grad_norm": 0.918504685912817, | |
| "learning_rate": 3.459395221491853e-06, | |
| "loss": 0.1772, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.199400299850075, | |
| "eval_loss": 0.305374413728714, | |
| "eval_runtime": 9.5082, | |
| "eval_samples_per_second": 5.679, | |
| "eval_steps_per_second": 1.472, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2008995502248876, | |
| "grad_norm": 0.9197017215268253, | |
| "learning_rate": 3.4481973100782756e-06, | |
| "loss": 0.1639, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.2023988005997002, | |
| "grad_norm": 1.0131975117483911, | |
| "learning_rate": 3.4370080051213527e-06, | |
| "loss": 0.2364, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.2038980509745127, | |
| "grad_norm": 0.909959203644668, | |
| "learning_rate": 3.4258273686781156e-06, | |
| "loss": 0.1297, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.2053973013493253, | |
| "grad_norm": 0.9120558754100766, | |
| "learning_rate": 3.4146554627575207e-06, | |
| "loss": 0.1637, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.206896551724138, | |
| "grad_norm": 1.0176446407586068, | |
| "learning_rate": 3.403492349320101e-06, | |
| "loss": 0.1764, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.2083958020989505, | |
| "grad_norm": 0.8212976307002104, | |
| "learning_rate": 3.392338090277628e-06, | |
| "loss": 0.1063, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.209895052473763, | |
| "grad_norm": 1.0046735405513456, | |
| "learning_rate": 3.3811927474927644e-06, | |
| "loss": 0.164, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.2113943028485756, | |
| "grad_norm": 0.9594136895444884, | |
| "learning_rate": 3.3700563827787224e-06, | |
| "loss": 0.1459, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.2128935532233882, | |
| "grad_norm": 0.7567955363468315, | |
| "learning_rate": 3.358929057898922e-06, | |
| "loss": 0.1275, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.214392803598201, | |
| "grad_norm": 0.7700659891115205, | |
| "learning_rate": 3.3478108345666456e-06, | |
| "loss": 0.1508, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2158920539730136, | |
| "grad_norm": 0.7541247281844553, | |
| "learning_rate": 3.3367017744446995e-06, | |
| "loss": 0.1114, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.2173913043478262, | |
| "grad_norm": 0.9253165849150354, | |
| "learning_rate": 3.3256019391450696e-06, | |
| "loss": 0.1939, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.2188905547226387, | |
| "grad_norm": 0.7568656805836542, | |
| "learning_rate": 3.314511390228578e-06, | |
| "loss": 0.1123, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.2203898050974513, | |
| "grad_norm": 0.9812333530337655, | |
| "learning_rate": 3.303430189204545e-06, | |
| "loss": 0.1302, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.221889055472264, | |
| "grad_norm": 0.9677602898937396, | |
| "learning_rate": 3.2923583975304474e-06, | |
| "loss": 0.1937, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.2233883058470765, | |
| "grad_norm": 0.7192177765704818, | |
| "learning_rate": 3.2812960766115747e-06, | |
| "loss": 0.1307, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.224887556221889, | |
| "grad_norm": 1.2433081499995242, | |
| "learning_rate": 3.270243287800693e-06, | |
| "loss": 0.1702, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.2263868065967016, | |
| "grad_norm": 0.8609019933433536, | |
| "learning_rate": 3.2592000923976997e-06, | |
| "loss": 0.173, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.2278860569715142, | |
| "grad_norm": 0.9338580105199386, | |
| "learning_rate": 3.2481665516492876e-06, | |
| "loss": 0.1424, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.2293853073463268, | |
| "grad_norm": 0.8596458266228822, | |
| "learning_rate": 3.2371427267486044e-06, | |
| "loss": 0.1307, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2308845577211394, | |
| "grad_norm": 0.8561516903076785, | |
| "learning_rate": 3.2261286788349127e-06, | |
| "loss": 0.1779, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.232383808095952, | |
| "grad_norm": 0.9350102393490926, | |
| "learning_rate": 3.2151244689932505e-06, | |
| "loss": 0.1552, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.2338830584707647, | |
| "grad_norm": 1.032656253717905, | |
| "learning_rate": 3.2041301582540903e-06, | |
| "loss": 0.1498, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.235382308845577, | |
| "grad_norm": 0.8087670924290683, | |
| "learning_rate": 3.1931458075930046e-06, | |
| "loss": 0.1429, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.23688155922039, | |
| "grad_norm": 0.8345032251877209, | |
| "learning_rate": 3.182171477930332e-06, | |
| "loss": 0.1572, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.2383808095952025, | |
| "grad_norm": 0.8620387506407755, | |
| "learning_rate": 3.171207230130826e-06, | |
| "loss": 0.1416, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.239880059970015, | |
| "grad_norm": 1.0178814037831463, | |
| "learning_rate": 3.1602531250033286e-06, | |
| "loss": 0.1656, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.2413793103448276, | |
| "grad_norm": 1.1614996318405568, | |
| "learning_rate": 3.149309223300428e-06, | |
| "loss": 0.1533, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.2428785607196402, | |
| "grad_norm": 0.921704556779215, | |
| "learning_rate": 3.1383755857181253e-06, | |
| "loss": 0.1136, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.2443778110944528, | |
| "grad_norm": 0.9867233974004074, | |
| "learning_rate": 3.1274522728954928e-06, | |
| "loss": 0.1602, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2458770614692654, | |
| "grad_norm": 0.893031656578016, | |
| "learning_rate": 3.1165393454143423e-06, | |
| "loss": 0.1257, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.247376311844078, | |
| "grad_norm": 0.9398366229391505, | |
| "learning_rate": 3.1056368637988876e-06, | |
| "loss": 0.1492, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.2488755622188905, | |
| "grad_norm": 0.886321430907055, | |
| "learning_rate": 3.0947448885154085e-06, | |
| "loss": 0.1076, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.250374812593703, | |
| "grad_norm": 0.9956891388658676, | |
| "learning_rate": 3.0838634799719157e-06, | |
| "loss": 0.1894, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.2518740629685157, | |
| "grad_norm": 0.7592222026908945, | |
| "learning_rate": 3.072992698517815e-06, | |
| "loss": 0.1258, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.2533733133433285, | |
| "grad_norm": 0.9916964302071175, | |
| "learning_rate": 3.0621326044435738e-06, | |
| "loss": 0.1924, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.2548725637181408, | |
| "grad_norm": 0.9701311917004072, | |
| "learning_rate": 3.0512832579803873e-06, | |
| "loss": 0.1459, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.2563718140929536, | |
| "grad_norm": 1.0210844248483126, | |
| "learning_rate": 3.0404447192998398e-06, | |
| "loss": 0.1817, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.2578710644677662, | |
| "grad_norm": 0.9264595051895576, | |
| "learning_rate": 3.029617048513579e-06, | |
| "loss": 0.1111, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.2593703148425788, | |
| "grad_norm": 0.776439100382626, | |
| "learning_rate": 3.0188003056729752e-06, | |
| "loss": 0.1304, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.2608695652173914, | |
| "grad_norm": 0.958809752543696, | |
| "learning_rate": 3.007994550768793e-06, | |
| "loss": 0.1339, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.262368815592204, | |
| "grad_norm": 0.9460624053988907, | |
| "learning_rate": 2.9971998437308546e-06, | |
| "loss": 0.1623, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.2638680659670165, | |
| "grad_norm": 0.9197207908035859, | |
| "learning_rate": 2.9864162444277118e-06, | |
| "loss": 0.1757, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.265367316341829, | |
| "grad_norm": 1.1058875560600865, | |
| "learning_rate": 2.97564381266631e-06, | |
| "loss": 0.1468, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.2668665667166417, | |
| "grad_norm": 0.7997177378786167, | |
| "learning_rate": 2.964882608191659e-06, | |
| "loss": 0.1152, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.2683658170914542, | |
| "grad_norm": 0.9698040525649909, | |
| "learning_rate": 2.954132690686502e-06, | |
| "loss": 0.1695, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.2698650674662668, | |
| "grad_norm": 0.9483455181922433, | |
| "learning_rate": 2.9433941197709813e-06, | |
| "loss": 0.167, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.2713643178410794, | |
| "grad_norm": 1.0272438408873248, | |
| "learning_rate": 2.9326669550023124e-06, | |
| "loss": 0.1594, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.272863568215892, | |
| "grad_norm": 0.8884247772824763, | |
| "learning_rate": 2.921951255874449e-06, | |
| "loss": 0.1177, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.2743628185907045, | |
| "grad_norm": 0.9130766619026013, | |
| "learning_rate": 2.9112470818177563e-06, | |
| "loss": 0.1434, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.2758620689655173, | |
| "grad_norm": 0.8864393102065208, | |
| "learning_rate": 2.9005544921986774e-06, | |
| "loss": 0.1722, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.2773613193403297, | |
| "grad_norm": 0.9701882806935522, | |
| "learning_rate": 2.8898735463194128e-06, | |
| "loss": 0.1494, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.2788605697151425, | |
| "grad_norm": 0.9313164614982317, | |
| "learning_rate": 2.8792043034175817e-06, | |
| "loss": 0.1785, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.280359820089955, | |
| "grad_norm": 0.9059416653292283, | |
| "learning_rate": 2.8685468226658974e-06, | |
| "loss": 0.1558, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.2818590704647677, | |
| "grad_norm": 0.9230454095036406, | |
| "learning_rate": 2.85790116317184e-06, | |
| "loss": 0.1448, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.2833583208395802, | |
| "grad_norm": 0.7630132744207047, | |
| "learning_rate": 2.8472673839773267e-06, | |
| "loss": 0.1411, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.2848575712143928, | |
| "grad_norm": 1.021216591541522, | |
| "learning_rate": 2.8366455440583874e-06, | |
| "loss": 0.1692, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.2863568215892054, | |
| "grad_norm": 1.0112137599318012, | |
| "learning_rate": 2.8260357023248323e-06, | |
| "loss": 0.1964, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.287856071964018, | |
| "grad_norm": 0.8438794948467353, | |
| "learning_rate": 2.815437917619932e-06, | |
| "loss": 0.1711, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.2893553223388305, | |
| "grad_norm": 0.9410400943874152, | |
| "learning_rate": 2.804852248720085e-06, | |
| "loss": 0.1528, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.2908545727136431, | |
| "grad_norm": 0.7695900524206701, | |
| "learning_rate": 2.7942787543344957e-06, | |
| "loss": 0.1725, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.2923538230884557, | |
| "grad_norm": 0.8767791481118428, | |
| "learning_rate": 2.783717493104846e-06, | |
| "loss": 0.1402, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.2938530734632683, | |
| "grad_norm": 0.8905915146257021, | |
| "learning_rate": 2.7731685236049745e-06, | |
| "loss": 0.2116, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.295352323838081, | |
| "grad_norm": 0.8772107737092024, | |
| "learning_rate": 2.762631904340546e-06, | |
| "loss": 0.134, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.2968515742128934, | |
| "grad_norm": 1.0172657564335914, | |
| "learning_rate": 2.7521076937487248e-06, | |
| "loss": 0.1586, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.2983508245877062, | |
| "grad_norm": 0.8608265568437447, | |
| "learning_rate": 2.7415959501978674e-06, | |
| "loss": 0.1483, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.2998500749625188, | |
| "grad_norm": 0.8268391765991547, | |
| "learning_rate": 2.731096731987177e-06, | |
| "loss": 0.1395, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.3013493253373314, | |
| "grad_norm": 0.914276240441856, | |
| "learning_rate": 2.7206100973463958e-06, | |
| "loss": 0.1261, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.302848575712144, | |
| "grad_norm": 0.836522180514579, | |
| "learning_rate": 2.71013610443547e-06, | |
| "loss": 0.1522, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.3043478260869565, | |
| "grad_norm": 0.8957409405519454, | |
| "learning_rate": 2.6996748113442397e-06, | |
| "loss": 0.1655, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3058470764617691, | |
| "grad_norm": 0.9311665285675249, | |
| "learning_rate": 2.689226276092107e-06, | |
| "loss": 0.1667, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.3073463268365817, | |
| "grad_norm": 0.987005409069299, | |
| "learning_rate": 2.6787905566277185e-06, | |
| "loss": 0.2067, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.3088455772113943, | |
| "grad_norm": 0.892851607871428, | |
| "learning_rate": 2.6683677108286423e-06, | |
| "loss": 0.1631, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.3103448275862069, | |
| "grad_norm": 0.9908722303948867, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.1539, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.3118440779610194, | |
| "grad_norm": 0.854910683339314, | |
| "learning_rate": 2.6475608713793923e-06, | |
| "loss": 0.1232, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.313343328335832, | |
| "grad_norm": 0.9247630401276781, | |
| "learning_rate": 2.6371769931260806e-06, | |
| "loss": 0.1145, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.3148425787106448, | |
| "grad_norm": 0.992213888599991, | |
| "learning_rate": 2.6268062193311672e-06, | |
| "loss": 0.1578, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.3163418290854572, | |
| "grad_norm": 0.8317652749759097, | |
| "learning_rate": 2.6164486075120245e-06, | |
| "loss": 0.1401, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.31784107946027, | |
| "grad_norm": 0.9239602521531679, | |
| "learning_rate": 2.606104215113033e-06, | |
| "loss": 0.1515, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.3193403298350825, | |
| "grad_norm": 0.932377577011233, | |
| "learning_rate": 2.5957730995052477e-06, | |
| "loss": 0.1441, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3208395802098951, | |
| "grad_norm": 0.689822591481137, | |
| "learning_rate": 2.585455317986095e-06, | |
| "loss": 0.1183, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.3223388305847077, | |
| "grad_norm": 0.8152086406538063, | |
| "learning_rate": 2.5751509277790487e-06, | |
| "loss": 0.1917, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.3238380809595203, | |
| "grad_norm": 0.9614386601041492, | |
| "learning_rate": 2.5648599860333122e-06, | |
| "loss": 0.1169, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.3253373313343328, | |
| "grad_norm": 1.0380740383409897, | |
| "learning_rate": 2.554582549823502e-06, | |
| "loss": 0.1458, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.3268365817091454, | |
| "grad_norm": 0.7791964513311427, | |
| "learning_rate": 2.5443186761493327e-06, | |
| "loss": 0.16, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.328335832083958, | |
| "grad_norm": 0.8273532151898028, | |
| "learning_rate": 2.5340684219352977e-06, | |
| "loss": 0.1452, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.3298350824587706, | |
| "grad_norm": 1.075351806591204, | |
| "learning_rate": 2.523831844030358e-06, | |
| "loss": 0.2524, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.3313343328335832, | |
| "grad_norm": 0.8709194882774809, | |
| "learning_rate": 2.513608999207622e-06, | |
| "loss": 0.1353, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.3328335832083957, | |
| "grad_norm": 0.964757590838339, | |
| "learning_rate": 2.503399944164035e-06, | |
| "loss": 0.1667, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.3343328335832085, | |
| "grad_norm": 0.917818772201969, | |
| "learning_rate": 2.4932047355200613e-06, | |
| "loss": 0.1556, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.3358320839580209, | |
| "grad_norm": 0.891821453340637, | |
| "learning_rate": 2.483023429819372e-06, | |
| "loss": 0.1416, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.3373313343328337, | |
| "grad_norm": 0.7973384231712907, | |
| "learning_rate": 2.472856083528531e-06, | |
| "loss": 0.156, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.338830584707646, | |
| "grad_norm": 0.8485835989819746, | |
| "learning_rate": 2.4627027530366836e-06, | |
| "loss": 0.134, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.3403298350824588, | |
| "grad_norm": 0.9061489422532453, | |
| "learning_rate": 2.4525634946552405e-06, | |
| "loss": 0.1666, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.3418290854572714, | |
| "grad_norm": 1.4075171601828986, | |
| "learning_rate": 2.442438364617567e-06, | |
| "loss": 0.2314, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.343328335832084, | |
| "grad_norm": 1.1565498158947582, | |
| "learning_rate": 2.4323274190786703e-06, | |
| "loss": 0.1527, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.3448275862068966, | |
| "grad_norm": 0.827599472301407, | |
| "learning_rate": 2.422230714114891e-06, | |
| "loss": 0.1555, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.3463268365817092, | |
| "grad_norm": 0.7513662255168734, | |
| "learning_rate": 2.4121483057235884e-06, | |
| "loss": 0.177, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.3478260869565217, | |
| "grad_norm": 0.9003524821850603, | |
| "learning_rate": 2.4020802498228333e-06, | |
| "loss": 0.1526, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.3493253373313343, | |
| "grad_norm": 0.6745200666289144, | |
| "learning_rate": 2.392026602251093e-06, | |
| "loss": 0.1163, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.3508245877061469, | |
| "grad_norm": 1.0006065482585125, | |
| "learning_rate": 2.3819874187669266e-06, | |
| "loss": 0.1756, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.3523238380809595, | |
| "grad_norm": 0.915638400356647, | |
| "learning_rate": 2.371962755048675e-06, | |
| "loss": 0.1587, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 1.353823088455772, | |
| "grad_norm": 1.4219216577051486, | |
| "learning_rate": 2.36195266669415e-06, | |
| "loss": 0.1658, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.3553223388305846, | |
| "grad_norm": 0.7722684394224646, | |
| "learning_rate": 2.351957209220326e-06, | |
| "loss": 0.0971, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.3568215892053974, | |
| "grad_norm": 0.8818710203100792, | |
| "learning_rate": 2.341976438063035e-06, | |
| "loss": 0.176, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.3583208395802098, | |
| "grad_norm": 0.8642408961688145, | |
| "learning_rate": 2.332010408576653e-06, | |
| "loss": 0.135, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 1.3598200899550226, | |
| "grad_norm": 0.857322002919173, | |
| "learning_rate": 2.3220591760338046e-06, | |
| "loss": 0.1357, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 1.3613193403298351, | |
| "grad_norm": 0.9286233073602603, | |
| "learning_rate": 2.3121227956250435e-06, | |
| "loss": 0.1427, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.3628185907046477, | |
| "grad_norm": 0.952874425923247, | |
| "learning_rate": 2.302201322458552e-06, | |
| "loss": 0.1374, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 1.3643178410794603, | |
| "grad_norm": 0.8704450911829884, | |
| "learning_rate": 2.292294811559837e-06, | |
| "loss": 0.1434, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.3658170914542729, | |
| "grad_norm": 0.8345741802796142, | |
| "learning_rate": 2.282403317871422e-06, | |
| "loss": 0.14, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 1.3673163418290855, | |
| "grad_norm": 0.8927131443111073, | |
| "learning_rate": 2.2725268962525454e-06, | |
| "loss": 0.1919, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.368815592203898, | |
| "grad_norm": 0.9378749901174825, | |
| "learning_rate": 2.262665601478852e-06, | |
| "loss": 0.1681, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 1.3703148425787106, | |
| "grad_norm": 0.8907537107335353, | |
| "learning_rate": 2.252819488242093e-06, | |
| "loss": 0.1555, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 1.3718140929535232, | |
| "grad_norm": 0.8936572188423556, | |
| "learning_rate": 2.24298861114982e-06, | |
| "loss": 0.1484, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.3733133433283358, | |
| "grad_norm": 0.7316261559655752, | |
| "learning_rate": 2.2331730247250857e-06, | |
| "loss": 0.1036, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.3748125937031483, | |
| "grad_norm": 1.0741197334302268, | |
| "learning_rate": 2.223372783406137e-06, | |
| "loss": 0.1345, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.3763118440779611, | |
| "grad_norm": 0.9348129526165928, | |
| "learning_rate": 2.2135879415461152e-06, | |
| "loss": 0.1328, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 1.3778110944527735, | |
| "grad_norm": 1.1289964905947087, | |
| "learning_rate": 2.203818553412757e-06, | |
| "loss": 0.1545, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 0.913964120580775, | |
| "learning_rate": 2.1940646731880887e-06, | |
| "loss": 0.1649, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.3808095952023989, | |
| "grad_norm": 0.7645558200782704, | |
| "learning_rate": 2.1843263549681287e-06, | |
| "loss": 0.1578, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 1.3823088455772115, | |
| "grad_norm": 0.9069244715340903, | |
| "learning_rate": 2.174603652762588e-06, | |
| "loss": 0.1217, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 1.383808095952024, | |
| "grad_norm": 0.9971644502579307, | |
| "learning_rate": 2.164896620494569e-06, | |
| "loss": 0.1453, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 1.3853073463268366, | |
| "grad_norm": 1.0419416539896398, | |
| "learning_rate": 2.1552053120002655e-06, | |
| "loss": 0.1411, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.3868065967016492, | |
| "grad_norm": 1.0556218928816279, | |
| "learning_rate": 2.145529781028668e-06, | |
| "loss": 0.1374, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.3883058470764618, | |
| "grad_norm": 0.7714546976049913, | |
| "learning_rate": 2.1358700812412625e-06, | |
| "loss": 0.1193, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 1.3898050974512743, | |
| "grad_norm": 0.8040245969029627, | |
| "learning_rate": 2.1262262662117327e-06, | |
| "loss": 0.1412, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.9139753999524988, | |
| "learning_rate": 2.1165983894256647e-06, | |
| "loss": 0.1314, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 1.3928035982008995, | |
| "grad_norm": 1.133801055067359, | |
| "learning_rate": 2.1069865042802502e-06, | |
| "loss": 0.1813, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 1.394302848575712, | |
| "grad_norm": 0.9816013761255433, | |
| "learning_rate": 2.0973906640839867e-06, | |
| "loss": 0.1057, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.3958020989505249, | |
| "grad_norm": 0.686731174841431, | |
| "learning_rate": 2.0878109220563884e-06, | |
| "loss": 0.1084, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.3973013493253372, | |
| "grad_norm": 0.8106897523208525, | |
| "learning_rate": 2.078247331327685e-06, | |
| "loss": 0.1666, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 1.39880059970015, | |
| "grad_norm": 1.0393587333424863, | |
| "learning_rate": 2.0686999449385286e-06, | |
| "loss": 0.1534, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 1.4002998500749624, | |
| "grad_norm": 0.8378186658590169, | |
| "learning_rate": 2.0591688158397054e-06, | |
| "loss": 0.0965, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 1.4017991004497752, | |
| "grad_norm": 0.9429173724886153, | |
| "learning_rate": 2.0496539968918342e-06, | |
| "loss": 0.1314, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.4032983508245878, | |
| "grad_norm": 0.8282593315671637, | |
| "learning_rate": 2.0401555408650714e-06, | |
| "loss": 0.1561, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 1.4047976011994003, | |
| "grad_norm": 0.9713460475660101, | |
| "learning_rate": 2.030673500438828e-06, | |
| "loss": 0.1531, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 1.406296851574213, | |
| "grad_norm": 0.8075293864143902, | |
| "learning_rate": 2.0212079282014725e-06, | |
| "loss": 0.1064, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 1.4077961019490255, | |
| "grad_norm": 0.8388301191356846, | |
| "learning_rate": 2.0117588766500375e-06, | |
| "loss": 0.1741, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 1.409295352323838, | |
| "grad_norm": 1.0222117234758845, | |
| "learning_rate": 2.002326398189931e-06, | |
| "loss": 0.1898, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.4107946026986506, | |
| "grad_norm": 0.9609151534810089, | |
| "learning_rate": 1.9929105451346436e-06, | |
| "loss": 0.1312, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 1.4122938530734632, | |
| "grad_norm": 1.3851428390429072, | |
| "learning_rate": 1.983511369705462e-06, | |
| "loss": 0.1563, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 1.4137931034482758, | |
| "grad_norm": 1.00143726888369, | |
| "learning_rate": 1.9741289240311757e-06, | |
| "loss": 0.1269, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 1.4152923538230884, | |
| "grad_norm": 0.927544224590705, | |
| "learning_rate": 1.9647632601477877e-06, | |
| "loss": 0.1652, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 1.416791604197901, | |
| "grad_norm": 0.93882585284601, | |
| "learning_rate": 1.9554144299982314e-06, | |
| "loss": 0.1605, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.4182908545727138, | |
| "grad_norm": 0.8902875185097399, | |
| "learning_rate": 1.9460824854320755e-06, | |
| "loss": 0.1362, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 1.419790104947526, | |
| "grad_norm": 0.9030559835786072, | |
| "learning_rate": 1.9367674782052376e-06, | |
| "loss": 0.1765, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 1.421289355322339, | |
| "grad_norm": 0.9142906702591052, | |
| "learning_rate": 1.9274694599797067e-06, | |
| "loss": 0.1719, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 1.4227886056971515, | |
| "grad_norm": 0.8967317425522585, | |
| "learning_rate": 1.918188482323242e-06, | |
| "loss": 0.1898, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 1.424287856071964, | |
| "grad_norm": 0.874591250781882, | |
| "learning_rate": 1.9089245967090952e-06, | |
| "loss": 0.1427, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4257871064467766, | |
| "grad_norm": 0.9428513506526265, | |
| "learning_rate": 1.8996778545157263e-06, | |
| "loss": 0.1457, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 1.4272863568215892, | |
| "grad_norm": 0.9001876865744428, | |
| "learning_rate": 1.8904483070265133e-06, | |
| "loss": 0.1669, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.4287856071964018, | |
| "grad_norm": 0.9889286421941543, | |
| "learning_rate": 1.8812360054294725e-06, | |
| "loss": 0.1456, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 1.4302848575712144, | |
| "grad_norm": 1.0214931422156825, | |
| "learning_rate": 1.8720410008169727e-06, | |
| "loss": 0.1755, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 1.431784107946027, | |
| "grad_norm": 1.1335499673591838, | |
| "learning_rate": 1.8628633441854515e-06, | |
| "loss": 0.2711, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.4332833583208395, | |
| "grad_norm": 1.049525915847894, | |
| "learning_rate": 1.8537030864351303e-06, | |
| "loss": 0.1756, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 1.434782608695652, | |
| "grad_norm": 0.7953308767615797, | |
| "learning_rate": 1.8445602783697375e-06, | |
| "loss": 0.1439, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 1.4362818590704647, | |
| "grad_norm": 1.1115770613677376, | |
| "learning_rate": 1.8354349706962243e-06, | |
| "loss": 0.1434, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 1.4377811094452775, | |
| "grad_norm": 0.7713897825374411, | |
| "learning_rate": 1.8263272140244803e-06, | |
| "loss": 0.1513, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 1.4392803598200898, | |
| "grad_norm": 0.8747431483860569, | |
| "learning_rate": 1.8172370588670563e-06, | |
| "loss": 0.1458, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4407796101949026, | |
| "grad_norm": 1.0316699268639047, | |
| "learning_rate": 1.8081645556388866e-06, | |
| "loss": 0.1429, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 1.4422788605697152, | |
| "grad_norm": 0.8824468306097878, | |
| "learning_rate": 1.7991097546570018e-06, | |
| "loss": 0.154, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 1.4437781109445278, | |
| "grad_norm": 0.9609867710768895, | |
| "learning_rate": 1.7900727061402556e-06, | |
| "loss": 0.1755, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 1.4452773613193404, | |
| "grad_norm": 1.0850737880601766, | |
| "learning_rate": 1.7810534602090445e-06, | |
| "loss": 0.1435, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 1.446776611694153, | |
| "grad_norm": 1.1622158566403475, | |
| "learning_rate": 1.77205206688503e-06, | |
| "loss": 0.1556, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.4482758620689655, | |
| "grad_norm": 1.0010439818552093, | |
| "learning_rate": 1.7630685760908623e-06, | |
| "loss": 0.1815, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.449775112443778, | |
| "grad_norm": 0.8502593655178186, | |
| "learning_rate": 1.7541030376499002e-06, | |
| "loss": 0.1391, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 1.4512743628185907, | |
| "grad_norm": 0.936338821826363, | |
| "learning_rate": 1.745155501285939e-06, | |
| "loss": 0.1773, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.4527736131934033, | |
| "grad_norm": 0.7221709768805439, | |
| "learning_rate": 1.736226016622931e-06, | |
| "loss": 0.1356, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 1.4542728635682158, | |
| "grad_norm": 1.0804879395566012, | |
| "learning_rate": 1.727314633184714e-06, | |
| "loss": 0.2075, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.4557721139430284, | |
| "grad_norm": 1.110799311241841, | |
| "learning_rate": 1.718421400394732e-06, | |
| "loss": 0.1632, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 1.4572713643178412, | |
| "grad_norm": 0.8956578962583012, | |
| "learning_rate": 1.7095463675757656e-06, | |
| "loss": 0.1667, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 1.4587706146926536, | |
| "grad_norm": 0.910102291577408, | |
| "learning_rate": 1.7006895839496557e-06, | |
| "loss": 0.1326, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 1.4602698650674664, | |
| "grad_norm": 0.9312393308823834, | |
| "learning_rate": 1.6918510986370312e-06, | |
| "loss": 0.137, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 1.461769115442279, | |
| "grad_norm": 0.972409110322856, | |
| "learning_rate": 1.6830309606570372e-06, | |
| "loss": 0.1464, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.4632683658170915, | |
| "grad_norm": 0.7708259594033834, | |
| "learning_rate": 1.674229218927062e-06, | |
| "loss": 0.0984, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.464767616191904, | |
| "grad_norm": 0.9210002916874346, | |
| "learning_rate": 1.665445922262467e-06, | |
| "loss": 0.1442, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 1.4662668665667167, | |
| "grad_norm": 1.0961294975113978, | |
| "learning_rate": 1.6566811193763149e-06, | |
| "loss": 0.1422, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 1.4677661169415293, | |
| "grad_norm": 0.8338488215942835, | |
| "learning_rate": 1.6479348588791e-06, | |
| "loss": 0.1644, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 1.4692653673163418, | |
| "grad_norm": 0.9463894230668072, | |
| "learning_rate": 1.6392071892784789e-06, | |
| "loss": 0.1705, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.4707646176911544, | |
| "grad_norm": 0.720695334720179, | |
| "learning_rate": 1.6304981589790015e-06, | |
| "loss": 0.1175, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 1.472263868065967, | |
| "grad_norm": 0.9576021607972085, | |
| "learning_rate": 1.6218078162818418e-06, | |
| "loss": 0.181, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 1.4737631184407796, | |
| "grad_norm": 0.7816097865187535, | |
| "learning_rate": 1.6131362093845299e-06, | |
| "loss": 0.1693, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 1.4752623688155921, | |
| "grad_norm": 0.9195720318956135, | |
| "learning_rate": 1.6044833863806864e-06, | |
| "loss": 0.1994, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.4767616191904047, | |
| "grad_norm": 0.7391581542683123, | |
| "learning_rate": 1.5958493952597536e-06, | |
| "loss": 0.1458, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.4782608695652173, | |
| "grad_norm": 0.7620958930797089, | |
| "learning_rate": 1.5872342839067305e-06, | |
| "loss": 0.1474, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 1.47976011994003, | |
| "grad_norm": 1.0302644216377386, | |
| "learning_rate": 1.5786381001019052e-06, | |
| "loss": 0.1521, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.4812593703148424, | |
| "grad_norm": 0.8209744342424903, | |
| "learning_rate": 1.5700608915205978e-06, | |
| "loss": 0.1189, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.4827586206896552, | |
| "grad_norm": 0.873584857363622, | |
| "learning_rate": 1.561502705732883e-06, | |
| "loss": 0.1414, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 1.4842578710644678, | |
| "grad_norm": 0.7515213751446108, | |
| "learning_rate": 1.5529635902033358e-06, | |
| "loss": 0.1206, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.4857571214392804, | |
| "grad_norm": 1.0327016126402075, | |
| "learning_rate": 1.5444435922907669e-06, | |
| "loss": 0.1292, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 1.487256371814093, | |
| "grad_norm": 0.8774881510173159, | |
| "learning_rate": 1.5359427592479553e-06, | |
| "loss": 0.1696, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 1.4887556221889056, | |
| "grad_norm": 1.0031139149529746, | |
| "learning_rate": 1.5274611382213922e-06, | |
| "loss": 0.1551, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 1.4902548725637181, | |
| "grad_norm": 1.0620043652375637, | |
| "learning_rate": 1.5189987762510167e-06, | |
| "loss": 0.1373, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 1.4917541229385307, | |
| "grad_norm": 0.7965667033813011, | |
| "learning_rate": 1.510555720269955e-06, | |
| "loss": 0.1018, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.4932533733133433, | |
| "grad_norm": 0.9200627759124954, | |
| "learning_rate": 1.5021320171042608e-06, | |
| "loss": 0.1844, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 1.4947526236881559, | |
| "grad_norm": 1.0962179320764676, | |
| "learning_rate": 1.4937277134726542e-06, | |
| "loss": 0.155, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 1.4962518740629684, | |
| "grad_norm": 0.823948862092206, | |
| "learning_rate": 1.4853428559862637e-06, | |
| "loss": 0.1467, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 1.497751124437781, | |
| "grad_norm": 0.9250905738064557, | |
| "learning_rate": 1.4769774911483686e-06, | |
| "loss": 0.1757, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 1.4992503748125938, | |
| "grad_norm": 0.9549606123978195, | |
| "learning_rate": 1.4686316653541377e-06, | |
| "loss": 0.1444, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.4992503748125938, | |
| "eval_loss": 0.3005059063434601, | |
| "eval_runtime": 9.5023, | |
| "eval_samples_per_second": 5.683, | |
| "eval_steps_per_second": 1.473, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5007496251874062, | |
| "grad_norm": 0.8846085679232111, | |
| "learning_rate": 1.4603054248903752e-06, | |
| "loss": 0.1976, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 1.502248875562219, | |
| "grad_norm": 0.8765989371577764, | |
| "learning_rate": 1.4519988159352665e-06, | |
| "loss": 0.157, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 1.5037481259370313, | |
| "grad_norm": 0.9733674197904918, | |
| "learning_rate": 1.4437118845581138e-06, | |
| "loss": 0.1459, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 1.5052473763118441, | |
| "grad_norm": 0.9184167693195132, | |
| "learning_rate": 1.4354446767190873e-06, | |
| "loss": 0.1449, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 1.5067466266866567, | |
| "grad_norm": 0.9124177335834135, | |
| "learning_rate": 1.4271972382689685e-06, | |
| "loss": 0.1361, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.5082458770614693, | |
| "grad_norm": 0.7942779600512551, | |
| "learning_rate": 1.4189696149488956e-06, | |
| "loss": 0.1453, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 1.5097451274362819, | |
| "grad_norm": 0.9215398336728725, | |
| "learning_rate": 1.4107618523901101e-06, | |
| "loss": 0.1138, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 1.5112443778110944, | |
| "grad_norm": 0.8370295745843028, | |
| "learning_rate": 1.4025739961137043e-06, | |
| "loss": 0.1553, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 1.512743628185907, | |
| "grad_norm": 0.9906572303202251, | |
| "learning_rate": 1.394406091530367e-06, | |
| "loss": 0.125, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 1.5142428785607196, | |
| "grad_norm": 1.1188215487509212, | |
| "learning_rate": 1.3862581839401346e-06, | |
| "loss": 0.205, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.5157421289355324, | |
| "grad_norm": 0.8514241208849797, | |
| "learning_rate": 1.3781303185321377e-06, | |
| "loss": 0.1006, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 1.5172413793103448, | |
| "grad_norm": 1.010810213838376, | |
| "learning_rate": 1.370022540384347e-06, | |
| "loss": 0.1299, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 1.5187406296851576, | |
| "grad_norm": 0.9747511854643393, | |
| "learning_rate": 1.3619348944633331e-06, | |
| "loss": 0.1445, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 1.52023988005997, | |
| "grad_norm": 0.849733571546591, | |
| "learning_rate": 1.3538674256240087e-06, | |
| "loss": 0.1739, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 1.5217391304347827, | |
| "grad_norm": 1.087014853170662, | |
| "learning_rate": 1.3458201786093795e-06, | |
| "loss": 0.1539, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.523238380809595, | |
| "grad_norm": 1.1901770992446896, | |
| "learning_rate": 1.3377931980503055e-06, | |
| "loss": 0.2126, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 1.5247376311844079, | |
| "grad_norm": 0.6973656378066366, | |
| "learning_rate": 1.3297865284652417e-06, | |
| "loss": 0.1205, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 1.5262368815592204, | |
| "grad_norm": 0.9680310176441411, | |
| "learning_rate": 1.3218002142599973e-06, | |
| "loss": 0.161, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 1.527736131934033, | |
| "grad_norm": 0.8428376447807903, | |
| "learning_rate": 1.3138342997274883e-06, | |
| "loss": 0.1559, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 1.5292353823088456, | |
| "grad_norm": 0.7575333572876146, | |
| "learning_rate": 1.3058888290474937e-06, | |
| "loss": 0.117, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5307346326836582, | |
| "grad_norm": 1.042012809480401, | |
| "learning_rate": 1.2979638462864069e-06, | |
| "loss": 0.1379, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 1.5322338830584707, | |
| "grad_norm": 1.0641507593083732, | |
| "learning_rate": 1.2900593953969947e-06, | |
| "loss": 0.1517, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 1.5337331334332833, | |
| "grad_norm": 0.9352246096557318, | |
| "learning_rate": 1.2821755202181503e-06, | |
| "loss": 0.1338, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 1.535232383808096, | |
| "grad_norm": 0.9395030914121527, | |
| "learning_rate": 1.2743122644746536e-06, | |
| "loss": 0.1793, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 1.5367316341829085, | |
| "grad_norm": 0.9841324464782555, | |
| "learning_rate": 1.266469671776926e-06, | |
| "loss": 0.1307, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.5382308845577213, | |
| "grad_norm": 1.1299491044322119, | |
| "learning_rate": 1.2586477856207902e-06, | |
| "loss": 0.1869, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 1.5397301349325336, | |
| "grad_norm": 0.9403829344681439, | |
| "learning_rate": 1.2508466493872273e-06, | |
| "loss": 0.1585, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 1.5412293853073464, | |
| "grad_norm": 0.9322141461645381, | |
| "learning_rate": 1.2430663063421388e-06, | |
| "loss": 0.1392, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 1.5427286356821588, | |
| "grad_norm": 0.9310800649319809, | |
| "learning_rate": 1.2353067996361034e-06, | |
| "loss": 0.1499, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 1.5442278860569716, | |
| "grad_norm": 0.9661786485951669, | |
| "learning_rate": 1.2275681723041406e-06, | |
| "loss": 0.1335, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.545727136431784, | |
| "grad_norm": 0.8602739759612478, | |
| "learning_rate": 1.2198504672654694e-06, | |
| "loss": 0.1242, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 1.5472263868065967, | |
| "grad_norm": 0.8166969309747116, | |
| "learning_rate": 1.212153727323273e-06, | |
| "loss": 0.133, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 1.5487256371814093, | |
| "grad_norm": 0.7737263294451023, | |
| "learning_rate": 1.2044779951644586e-06, | |
| "loss": 0.1476, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 1.550224887556222, | |
| "grad_norm": 1.1282626749271907, | |
| "learning_rate": 1.1968233133594243e-06, | |
| "loss": 0.1475, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 1.5517241379310345, | |
| "grad_norm": 0.7784102607856898, | |
| "learning_rate": 1.1891897243618184e-06, | |
| "loss": 0.137, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.553223388305847, | |
| "grad_norm": 1.00547755014787, | |
| "learning_rate": 1.1815772705083072e-06, | |
| "loss": 0.1318, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 1.5547226386806596, | |
| "grad_norm": 0.9004256381975071, | |
| "learning_rate": 1.17398599401834e-06, | |
| "loss": 0.1579, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 1.5562218890554722, | |
| "grad_norm": 0.905778825998295, | |
| "learning_rate": 1.1664159369939137e-06, | |
| "loss": 0.148, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 1.557721139430285, | |
| "grad_norm": 0.9998446440085478, | |
| "learning_rate": 1.1588671414193397e-06, | |
| "loss": 0.1667, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 1.5592203898050974, | |
| "grad_norm": 0.964288191281469, | |
| "learning_rate": 1.1513396491610113e-06, | |
| "loss": 0.1274, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.5607196401799102, | |
| "grad_norm": 0.7991337749251353, | |
| "learning_rate": 1.1438335019671715e-06, | |
| "loss": 0.1045, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 1.5622188905547225, | |
| "grad_norm": 0.9760014280654934, | |
| "learning_rate": 1.1363487414676805e-06, | |
| "loss": 0.1353, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 1.5637181409295353, | |
| "grad_norm": 1.0337165441470804, | |
| "learning_rate": 1.128885409173789e-06, | |
| "loss": 0.1607, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 1.5652173913043477, | |
| "grad_norm": 0.8613884084825266, | |
| "learning_rate": 1.1214435464779006e-06, | |
| "loss": 0.161, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 1.5667166416791605, | |
| "grad_norm": 0.8722006310273899, | |
| "learning_rate": 1.1140231946533486e-06, | |
| "loss": 0.1397, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.568215892053973, | |
| "grad_norm": 1.0042463357931943, | |
| "learning_rate": 1.1066243948541638e-06, | |
| "loss": 0.1535, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 1.5697151424287856, | |
| "grad_norm": 0.964820864389745, | |
| "learning_rate": 1.0992471881148497e-06, | |
| "loss": 0.1683, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 1.5712143928035982, | |
| "grad_norm": 0.9574219276059943, | |
| "learning_rate": 1.091891615350147e-06, | |
| "loss": 0.1596, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 1.5727136431784108, | |
| "grad_norm": 0.9596782578773558, | |
| "learning_rate": 1.0845577173548172e-06, | |
| "loss": 0.1548, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 1.5742128935532234, | |
| "grad_norm": 0.8057334797805433, | |
| "learning_rate": 1.07724553480341e-06, | |
| "loss": 0.1328, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.575712143928036, | |
| "grad_norm": 0.8884570023307804, | |
| "learning_rate": 1.0699551082500387e-06, | |
| "loss": 0.1098, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 1.5772113943028487, | |
| "grad_norm": 1.0734906262935926, | |
| "learning_rate": 1.0626864781281553e-06, | |
| "loss": 0.1332, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 1.578710644677661, | |
| "grad_norm": 0.9878104025555289, | |
| "learning_rate": 1.0554396847503272e-06, | |
| "loss": 0.1266, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 1.580209895052474, | |
| "grad_norm": 0.9767316146743106, | |
| "learning_rate": 1.0482147683080125e-06, | |
| "loss": 0.1856, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 1.5817091454272862, | |
| "grad_norm": 0.8472503624448181, | |
| "learning_rate": 1.0410117688713366e-06, | |
| "loss": 0.1612, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.583208395802099, | |
| "grad_norm": 0.9827879949925994, | |
| "learning_rate": 1.0338307263888748e-06, | |
| "loss": 0.1547, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 1.5847076461769114, | |
| "grad_norm": 0.8842184605750372, | |
| "learning_rate": 1.0266716806874227e-06, | |
| "loss": 0.1893, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 1.5862068965517242, | |
| "grad_norm": 0.9889558695237448, | |
| "learning_rate": 1.0195346714717813e-06, | |
| "loss": 0.1425, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 1.5877061469265368, | |
| "grad_norm": 0.9297628179391901, | |
| "learning_rate": 1.0124197383245344e-06, | |
| "loss": 0.1503, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 1.5892053973013494, | |
| "grad_norm": 1.2714213362350966, | |
| "learning_rate": 1.0053269207058298e-06, | |
| "loss": 0.1188, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.590704647676162, | |
| "grad_norm": 0.9099832342809827, | |
| "learning_rate": 9.982562579531607e-07, | |
| "loss": 0.1549, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 1.5922038980509745, | |
| "grad_norm": 0.8903206652475655, | |
| "learning_rate": 9.912077892811473e-07, | |
| "loss": 0.1014, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 1.593703148425787, | |
| "grad_norm": 0.9074018479301309, | |
| "learning_rate": 9.841815537813177e-07, | |
| "loss": 0.1169, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 1.5952023988005997, | |
| "grad_norm": 1.0820626511449882, | |
| "learning_rate": 9.77177590421895e-07, | |
| "loss": 0.1334, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 1.5967016491754122, | |
| "grad_norm": 1.0272487325376969, | |
| "learning_rate": 9.70195938047576e-07, | |
| "loss": 0.2007, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.5982008995502248, | |
| "grad_norm": 0.9447763684217668, | |
| "learning_rate": 9.63236635379321e-07, | |
| "loss": 0.1321, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 1.5997001499250376, | |
| "grad_norm": 1.0446941605033662, | |
| "learning_rate": 9.562997210141355e-07, | |
| "loss": 0.1337, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 1.60119940029985, | |
| "grad_norm": 0.9749824794420779, | |
| "learning_rate": 9.49385233424856e-07, | |
| "loss": 0.1476, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 1.6026986506746628, | |
| "grad_norm": 0.8359827164531823, | |
| "learning_rate": 9.424932109599372e-07, | |
| "loss": 0.1499, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 1.6041979010494751, | |
| "grad_norm": 0.9116771567353064, | |
| "learning_rate": 9.356236918432454e-07, | |
| "loss": 0.1529, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.605697151424288, | |
| "grad_norm": 0.9495103404975435, | |
| "learning_rate": 9.287767141738352e-07, | |
| "loss": 0.1525, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 1.6071964017991005, | |
| "grad_norm": 0.8090516414866303, | |
| "learning_rate": 9.21952315925746e-07, | |
| "loss": 0.1363, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 1.608695652173913, | |
| "grad_norm": 1.1062232929866718, | |
| "learning_rate": 9.151505349477901e-07, | |
| "loss": 0.1543, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 1.6101949025487257, | |
| "grad_norm": 0.839720737541938, | |
| "learning_rate": 9.08371408963341e-07, | |
| "loss": 0.1436, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 1.6116941529235382, | |
| "grad_norm": 0.9789805208676281, | |
| "learning_rate": 9.016149755701259e-07, | |
| "loss": 0.1713, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.6131934032983508, | |
| "grad_norm": 0.8526578282604436, | |
| "learning_rate": 8.948812722400157e-07, | |
| "loss": 0.1394, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 1.6146926536731634, | |
| "grad_norm": 0.8755768022265608, | |
| "learning_rate": 8.881703363188199e-07, | |
| "loss": 0.1363, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 1.616191904047976, | |
| "grad_norm": 0.7447292878968165, | |
| "learning_rate": 8.814822050260758e-07, | |
| "loss": 0.1199, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 1.6176911544227885, | |
| "grad_norm": 0.8761143286746008, | |
| "learning_rate": 8.748169154548448e-07, | |
| "loss": 0.1237, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 1.6191904047976013, | |
| "grad_norm": 0.8585265758297184, | |
| "learning_rate": 8.681745045715045e-07, | |
| "loss": 0.1651, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6206896551724137, | |
| "grad_norm": 0.6968181811389437, | |
| "learning_rate": 8.615550092155478e-07, | |
| "loss": 0.1286, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 1.6221889055472265, | |
| "grad_norm": 1.0900673980057367, | |
| "learning_rate": 8.549584660993726e-07, | |
| "loss": 0.1301, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 1.6236881559220389, | |
| "grad_norm": 1.2587498302669191, | |
| "learning_rate": 8.483849118080828e-07, | |
| "loss": 0.1481, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 1.6251874062968517, | |
| "grad_norm": 0.8952364875686002, | |
| "learning_rate": 8.418343827992842e-07, | |
| "loss": 0.1479, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 1.626686656671664, | |
| "grad_norm": 0.888432257587943, | |
| "learning_rate": 8.353069154028814e-07, | |
| "loss": 0.1293, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.6281859070464768, | |
| "grad_norm": 0.9509113446089996, | |
| "learning_rate": 8.28802545820877e-07, | |
| "loss": 0.1658, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 1.6296851574212894, | |
| "grad_norm": 1.2799889053276086, | |
| "learning_rate": 8.223213101271709e-07, | |
| "loss": 0.1738, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 1.631184407796102, | |
| "grad_norm": 0.9385320067998599, | |
| "learning_rate": 8.158632442673603e-07, | |
| "loss": 0.1599, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 1.6326836581709145, | |
| "grad_norm": 0.9308547294625913, | |
| "learning_rate": 8.094283840585398e-07, | |
| "loss": 0.1759, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 1.6341829085457271, | |
| "grad_norm": 0.7066019848078807, | |
| "learning_rate": 8.03016765189103e-07, | |
| "loss": 0.1131, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6356821589205397, | |
| "grad_norm": 0.8353583939258259, | |
| "learning_rate": 7.966284232185451e-07, | |
| "loss": 0.1185, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 1.6371814092953523, | |
| "grad_norm": 0.9921506436714349, | |
| "learning_rate": 7.902633935772647e-07, | |
| "loss": 0.1675, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 1.638680659670165, | |
| "grad_norm": 1.103206643975485, | |
| "learning_rate": 7.839217115663683e-07, | |
| "loss": 0.1413, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 1.6401799100449774, | |
| "grad_norm": 1.168066328854014, | |
| "learning_rate": 7.776034123574738e-07, | |
| "loss": 0.2059, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 1.6416791604197902, | |
| "grad_norm": 1.0112655313827763, | |
| "learning_rate": 7.713085309925156e-07, | |
| "loss": 0.1627, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.6431784107946026, | |
| "grad_norm": 0.8317845210467163, | |
| "learning_rate": 7.650371023835495e-07, | |
| "loss": 0.1248, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 1.6446776611694154, | |
| "grad_norm": 0.9943734775607934, | |
| "learning_rate": 7.587891613125631e-07, | |
| "loss": 0.1858, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 1.6461769115442277, | |
| "grad_norm": 0.9590283779498894, | |
| "learning_rate": 7.525647424312766e-07, | |
| "loss": 0.1328, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 1.6476761619190405, | |
| "grad_norm": 1.1880906571682928, | |
| "learning_rate": 7.46363880260954e-07, | |
| "loss": 0.1739, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 1.6491754122938531, | |
| "grad_norm": 0.9625079090659409, | |
| "learning_rate": 7.401866091922133e-07, | |
| "loss": 0.1278, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6506746626686657, | |
| "grad_norm": 0.8337068211561395, | |
| "learning_rate": 7.340329634848309e-07, | |
| "loss": 0.177, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 1.6521739130434783, | |
| "grad_norm": 0.8522131528745677, | |
| "learning_rate": 7.279029772675572e-07, | |
| "loss": 0.1667, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 1.6536731634182908, | |
| "grad_norm": 0.7961092325609691, | |
| "learning_rate": 7.217966845379243e-07, | |
| "loss": 0.1161, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 1.6551724137931034, | |
| "grad_norm": 0.9436724069943285, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.1756, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 1.656671664167916, | |
| "grad_norm": 0.856993102695668, | |
| "learning_rate": 7.096553148744806e-07, | |
| "loss": 0.1402, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.6581709145427288, | |
| "grad_norm": 0.9786189307421419, | |
| "learning_rate": 7.036203052779506e-07, | |
| "loss": 0.1629, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 1.6596701649175412, | |
| "grad_norm": 0.9682290333870899, | |
| "learning_rate": 6.97609123843247e-07, | |
| "loss": 0.142, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 1.661169415292354, | |
| "grad_norm": 0.9761165504591279, | |
| "learning_rate": 6.916218039089961e-07, | |
| "loss": 0.1418, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 1.6626686656671663, | |
| "grad_norm": 1.0137145132566012, | |
| "learning_rate": 6.856583786814891e-07, | |
| "loss": 0.1711, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 1.6641679160419791, | |
| "grad_norm": 0.9071611271279851, | |
| "learning_rate": 6.797188812344907e-07, | |
| "loss": 0.2139, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.6656671664167915, | |
| "grad_norm": 0.8793609426146525, | |
| "learning_rate": 6.738033445090653e-07, | |
| "loss": 0.1446, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 1.6671664167916043, | |
| "grad_norm": 0.8028659389561664, | |
| "learning_rate": 6.67911801313384e-07, | |
| "loss": 0.1172, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 1.6686656671664168, | |
| "grad_norm": 0.8452517763126202, | |
| "learning_rate": 6.620442843225483e-07, | |
| "loss": 0.1556, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 1.6701649175412294, | |
| "grad_norm": 0.9122874940606722, | |
| "learning_rate": 6.562008260784092e-07, | |
| "loss": 0.0953, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 1.671664167916042, | |
| "grad_norm": 1.049264067212477, | |
| "learning_rate": 6.503814589893836e-07, | |
| "loss": 0.188, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.6731634182908546, | |
| "grad_norm": 0.9752920858910517, | |
| "learning_rate": 6.445862153302784e-07, | |
| "loss": 0.1252, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 1.6746626686656672, | |
| "grad_norm": 0.9020087795670768, | |
| "learning_rate": 6.388151272421078e-07, | |
| "loss": 0.1482, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 1.6761619190404797, | |
| "grad_norm": 0.779431539347626, | |
| "learning_rate": 6.330682267319177e-07, | |
| "loss": 0.1622, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 1.6776611694152923, | |
| "grad_norm": 0.9374379993380763, | |
| "learning_rate": 6.273455456726074e-07, | |
| "loss": 0.1552, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 1.6791604197901049, | |
| "grad_norm": 0.8766162246402146, | |
| "learning_rate": 6.216471158027515e-07, | |
| "loss": 0.1985, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.6806596701649177, | |
| "grad_norm": 0.8314923674269031, | |
| "learning_rate": 6.159729687264254e-07, | |
| "loss": 0.1148, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 1.68215892053973, | |
| "grad_norm": 0.9393699644553948, | |
| "learning_rate": 6.103231359130308e-07, | |
| "loss": 0.1504, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 1.6836581709145428, | |
| "grad_norm": 0.8991202682047384, | |
| "learning_rate": 6.046976486971201e-07, | |
| "loss": 0.1539, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 1.6851574212893552, | |
| "grad_norm": 0.9306398679852055, | |
| "learning_rate": 5.990965382782177e-07, | |
| "loss": 0.1022, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 1.686656671664168, | |
| "grad_norm": 0.9013402257335543, | |
| "learning_rate": 5.935198357206595e-07, | |
| "loss": 0.1485, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.6881559220389803, | |
| "grad_norm": 0.9448943276182934, | |
| "learning_rate": 5.879675719534078e-07, | |
| "loss": 0.173, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 1.6896551724137931, | |
| "grad_norm": 0.8968659921769028, | |
| "learning_rate": 5.824397777698859e-07, | |
| "loss": 0.1358, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 1.6911544227886057, | |
| "grad_norm": 0.9544887013736196, | |
| "learning_rate": 5.769364838278063e-07, | |
| "loss": 0.1089, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 1.6926536731634183, | |
| "grad_norm": 1.0217905299362213, | |
| "learning_rate": 5.714577206490018e-07, | |
| "loss": 0.1241, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 1.6941529235382309, | |
| "grad_norm": 0.9467006136666858, | |
| "learning_rate": 5.660035186192531e-07, | |
| "loss": 0.164, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.6956521739130435, | |
| "grad_norm": 0.9244319408568124, | |
| "learning_rate": 5.60573907988124e-07, | |
| "loss": 0.1479, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 1.697151424287856, | |
| "grad_norm": 0.9009479442383403, | |
| "learning_rate": 5.551689188687909e-07, | |
| "loss": 0.1391, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 1.6986506746626686, | |
| "grad_norm": 0.8184263123240574, | |
| "learning_rate": 5.497885812378772e-07, | |
| "loss": 0.1438, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 1.7001499250374814, | |
| "grad_norm": 0.7799204727201581, | |
| "learning_rate": 5.444329249352859e-07, | |
| "loss": 0.1528, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 1.7016491754122938, | |
| "grad_norm": 1.0461872245157817, | |
| "learning_rate": 5.391019796640362e-07, | |
| "loss": 0.1106, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 1.7031484257871066, | |
| "grad_norm": 0.8856575277474454, | |
| "learning_rate": 5.337957749900958e-07, | |
| "loss": 0.1458, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 1.704647676161919, | |
| "grad_norm": 1.0460372182731463, | |
| "learning_rate": 5.285143403422188e-07, | |
| "loss": 0.1945, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 1.7061469265367317, | |
| "grad_norm": 0.8961198982187754, | |
| "learning_rate": 5.23257705011786e-07, | |
| "loss": 0.1677, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 1.707646176911544, | |
| "grad_norm": 0.8792730024850949, | |
| "learning_rate": 5.18025898152631e-07, | |
| "loss": 0.1482, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 1.7091454272863569, | |
| "grad_norm": 0.7315245558464043, | |
| "learning_rate": 5.128189487808927e-07, | |
| "loss": 0.0832, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7106446776611695, | |
| "grad_norm": 0.8844603291712118, | |
| "learning_rate": 5.076368857748454e-07, | |
| "loss": 0.1587, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 1.712143928035982, | |
| "grad_norm": 0.8237756091045937, | |
| "learning_rate": 5.024797378747414e-07, | |
| "loss": 0.1385, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 1.7136431784107946, | |
| "grad_norm": 1.0104932751488818, | |
| "learning_rate": 4.973475336826506e-07, | |
| "loss": 0.1836, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 1.7151424287856072, | |
| "grad_norm": 0.7544862068818067, | |
| "learning_rate": 4.922403016623034e-07, | |
| "loss": 0.1213, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 1.7166416791604198, | |
| "grad_norm": 0.8776467683421563, | |
| "learning_rate": 4.871580701389316e-07, | |
| "loss": 0.1153, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 1.7181409295352323, | |
| "grad_norm": 0.9203065009713807, | |
| "learning_rate": 4.821008672991118e-07, | |
| "loss": 0.1637, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 1.7196401799100451, | |
| "grad_norm": 1.0276149397924434, | |
| "learning_rate": 4.770687211906089e-07, | |
| "loss": 0.158, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 1.7211394302848575, | |
| "grad_norm": 0.946802751470967, | |
| "learning_rate": 4.720616597222205e-07, | |
| "loss": 0.1276, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 1.7226386806596703, | |
| "grad_norm": 0.8074941399715195, | |
| "learning_rate": 4.6707971066362324e-07, | |
| "loss": 0.1779, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 0.931305857586346, | |
| "learning_rate": 4.6212290164521554e-07, | |
| "loss": 0.1181, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7256371814092955, | |
| "grad_norm": 0.8864084898421931, | |
| "learning_rate": 4.5719126015796757e-07, | |
| "loss": 0.135, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 1.7271364317841078, | |
| "grad_norm": 0.858979394499242, | |
| "learning_rate": 4.522848135532698e-07, | |
| "loss": 0.1354, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 1.7286356821589206, | |
| "grad_norm": 0.9744688184418555, | |
| "learning_rate": 4.474035890427769e-07, | |
| "loss": 0.14, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 1.7301349325337332, | |
| "grad_norm": 0.9633329525184013, | |
| "learning_rate": 4.4254761369825984e-07, | |
| "loss": 0.1615, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 1.7316341829085458, | |
| "grad_norm": 0.7910907105727082, | |
| "learning_rate": 4.377169144514554e-07, | |
| "loss": 0.1305, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 1.7331334332833583, | |
| "grad_norm": 0.7998628656563037, | |
| "learning_rate": 4.329115180939164e-07, | |
| "loss": 0.1477, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 1.734632683658171, | |
| "grad_norm": 0.7243777117866047, | |
| "learning_rate": 4.281314512768625e-07, | |
| "loss": 0.091, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 1.7361319340329835, | |
| "grad_norm": 1.0970903681484496, | |
| "learning_rate": 4.2337674051103504e-07, | |
| "loss": 0.1845, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 1.737631184407796, | |
| "grad_norm": 0.8556577761802462, | |
| "learning_rate": 4.186474121665468e-07, | |
| "loss": 0.1285, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.8260709802146723, | |
| "learning_rate": 4.139434924727359e-07, | |
| "loss": 0.1158, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.7406296851574212, | |
| "grad_norm": 0.8356654462117995, | |
| "learning_rate": 4.092650075180232e-07, | |
| "loss": 0.1561, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 1.742128935532234, | |
| "grad_norm": 0.8126421524641756, | |
| "learning_rate": 4.046119832497658e-07, | |
| "loss": 0.1491, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 1.7436281859070464, | |
| "grad_norm": 0.8781296414625271, | |
| "learning_rate": 3.9998444547411255e-07, | |
| "loss": 0.1321, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 1.7451274362818592, | |
| "grad_norm": 0.8356109666447434, | |
| "learning_rate": 3.9538241985586144e-07, | |
| "loss": 0.0999, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 1.7466266866566715, | |
| "grad_norm": 0.9637992606000515, | |
| "learning_rate": 3.908059319183194e-07, | |
| "loss": 0.1512, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 1.7481259370314843, | |
| "grad_norm": 0.8362787382001656, | |
| "learning_rate": 3.8625500704315645e-07, | |
| "loss": 0.142, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 1.7496251874062967, | |
| "grad_norm": 0.9159940051583131, | |
| "learning_rate": 3.8172967047026834e-07, | |
| "loss": 0.144, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 1.7511244377811095, | |
| "grad_norm": 0.8475482163801797, | |
| "learning_rate": 3.7722994729763427e-07, | |
| "loss": 0.1639, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 1.752623688155922, | |
| "grad_norm": 0.9812227302151535, | |
| "learning_rate": 3.7275586248118114e-07, | |
| "loss": 0.1855, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 1.7541229385307346, | |
| "grad_norm": 0.8596141939276791, | |
| "learning_rate": 3.683074408346404e-07, | |
| "loss": 0.1445, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.7556221889055472, | |
| "grad_norm": 0.998436950488683, | |
| "learning_rate": 3.6388470702941436e-07, | |
| "loss": 0.1567, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 1.7571214392803598, | |
| "grad_norm": 0.9651694890649214, | |
| "learning_rate": 3.594876855944385e-07, | |
| "loss": 0.1406, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 1.7586206896551724, | |
| "grad_norm": 0.9533982927437473, | |
| "learning_rate": 3.5511640091604293e-07, | |
| "loss": 0.1276, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 1.760119940029985, | |
| "grad_norm": 0.803329142384361, | |
| "learning_rate": 3.50770877237821e-07, | |
| "loss": 0.1487, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 1.7616191904047978, | |
| "grad_norm": 0.9435905072918633, | |
| "learning_rate": 3.4645113866049187e-07, | |
| "loss": 0.1312, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.76311844077961, | |
| "grad_norm": 0.8335060961954337, | |
| "learning_rate": 3.42157209141768e-07, | |
| "loss": 0.1349, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 1.764617691154423, | |
| "grad_norm": 0.9191679848955163, | |
| "learning_rate": 3.3788911249622194e-07, | |
| "loss": 0.1231, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 1.7661169415292353, | |
| "grad_norm": 0.9908408191388912, | |
| "learning_rate": 3.336468723951558e-07, | |
| "loss": 0.1251, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 1.767616191904048, | |
| "grad_norm": 0.9803494726809994, | |
| "learning_rate": 3.294305123664665e-07, | |
| "loss": 0.149, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 1.7691154422788604, | |
| "grad_norm": 0.9175761618379262, | |
| "learning_rate": 3.2524005579452014e-07, | |
| "loss": 0.1622, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.7706146926536732, | |
| "grad_norm": 0.9568780304269298, | |
| "learning_rate": 3.2107552592001657e-07, | |
| "loss": 0.1571, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 1.7721139430284858, | |
| "grad_norm": 0.802989444779344, | |
| "learning_rate": 3.169369458398652e-07, | |
| "loss": 0.1148, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 1.7736131934032984, | |
| "grad_norm": 0.7788269197754069, | |
| "learning_rate": 3.128243385070562e-07, | |
| "loss": 0.1605, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 1.775112443778111, | |
| "grad_norm": 0.954694788497693, | |
| "learning_rate": 3.087377267305297e-07, | |
| "loss": 0.134, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 1.7766116941529235, | |
| "grad_norm": 0.8600791787195781, | |
| "learning_rate": 3.0467713317505363e-07, | |
| "loss": 0.1355, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 1.778110944527736, | |
| "grad_norm": 0.8463533841104126, | |
| "learning_rate": 3.006425803610963e-07, | |
| "loss": 0.1636, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 1.7796101949025487, | |
| "grad_norm": 0.8856030180569605, | |
| "learning_rate": 2.9663409066470025e-07, | |
| "loss": 0.1515, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 1.7811094452773615, | |
| "grad_norm": 0.926015901746546, | |
| "learning_rate": 2.9265168631736005e-07, | |
| "loss": 0.2155, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 1.7826086956521738, | |
| "grad_norm": 0.9123203241325265, | |
| "learning_rate": 2.88695389405898e-07, | |
| "loss": 0.1361, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 1.7841079460269866, | |
| "grad_norm": 0.9024817561861025, | |
| "learning_rate": 2.8476522187234177e-07, | |
| "loss": 0.1269, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.785607196401799, | |
| "grad_norm": 0.8705444317912489, | |
| "learning_rate": 2.808612055138038e-07, | |
| "loss": 0.1539, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 1.7871064467766118, | |
| "grad_norm": 0.7785631255599886, | |
| "learning_rate": 2.76983361982357e-07, | |
| "loss": 0.1018, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 1.7886056971514241, | |
| "grad_norm": 0.9638742712120287, | |
| "learning_rate": 2.731317127849209e-07, | |
| "loss": 0.156, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 1.790104947526237, | |
| "grad_norm": 0.8304821991558828, | |
| "learning_rate": 2.693062792831358e-07, | |
| "loss": 0.1534, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 1.7916041979010495, | |
| "grad_norm": 0.8198747386526178, | |
| "learning_rate": 2.655070826932471e-07, | |
| "loss": 0.1496, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 1.793103448275862, | |
| "grad_norm": 1.011190277251662, | |
| "learning_rate": 2.617341440859883e-07, | |
| "loss": 0.1463, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 1.7946026986506747, | |
| "grad_norm": 0.7592481056767786, | |
| "learning_rate": 2.5798748438646326e-07, | |
| "loss": 0.1365, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 1.7961019490254873, | |
| "grad_norm": 0.7629393298555008, | |
| "learning_rate": 2.5426712437403134e-07, | |
| "loss": 0.1665, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 1.7976011994002998, | |
| "grad_norm": 0.9850413992719725, | |
| "learning_rate": 2.5057308468218913e-07, | |
| "loss": 0.1596, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 1.7991004497751124, | |
| "grad_norm": 0.7802136495597287, | |
| "learning_rate": 2.4690538579845933e-07, | |
| "loss": 0.1154, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.7991004497751124, | |
| "eval_loss": 0.2998337745666504, | |
| "eval_runtime": 9.5116, | |
| "eval_samples_per_second": 5.677, | |
| "eval_steps_per_second": 1.472, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.800599700149925, | |
| "grad_norm": 0.8330168321687108, | |
| "learning_rate": 2.432640480642756e-07, | |
| "loss": 0.1615, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 1.8020989505247376, | |
| "grad_norm": 0.8714653729749743, | |
| "learning_rate": 2.396490916748706e-07, | |
| "loss": 0.1429, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 1.8035982008995504, | |
| "grad_norm": 1.206235507534175, | |
| "learning_rate": 2.360605366791624e-07, | |
| "loss": 0.2045, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 1.8050974512743627, | |
| "grad_norm": 0.9405222785930386, | |
| "learning_rate": 2.32498402979644e-07, | |
| "loss": 0.2016, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 1.8065967016491755, | |
| "grad_norm": 0.8351287805305182, | |
| "learning_rate": 2.2896271033227392e-07, | |
| "loss": 0.1521, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 1.8080959520239879, | |
| "grad_norm": 0.8324792616070745, | |
| "learning_rate": 2.2545347834636632e-07, | |
| "loss": 0.1505, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 1.8095952023988007, | |
| "grad_norm": 0.9163178164847272, | |
| "learning_rate": 2.219707264844806e-07, | |
| "loss": 0.147, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 1.811094452773613, | |
| "grad_norm": 0.8019443522090565, | |
| "learning_rate": 2.1851447406231573e-07, | |
| "loss": 0.1128, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 1.8125937031484258, | |
| "grad_norm": 0.8429902295496241, | |
| "learning_rate": 2.1508474024860171e-07, | |
| "loss": 0.1852, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 1.8140929535232384, | |
| "grad_norm": 0.833719686582817, | |
| "learning_rate": 2.1168154406499275e-07, | |
| "loss": 0.1148, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.815592203898051, | |
| "grad_norm": 0.9279072314445156, | |
| "learning_rate": 2.0830490438596418e-07, | |
| "loss": 0.1013, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 1.8170914542728636, | |
| "grad_norm": 0.7868298641094045, | |
| "learning_rate": 2.0495483993870578e-07, | |
| "loss": 0.1232, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 1.8185907046476761, | |
| "grad_norm": 0.8433931890335177, | |
| "learning_rate": 2.0163136930301696e-07, | |
| "loss": 0.1231, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 1.8200899550224887, | |
| "grad_norm": 1.0520694193582327, | |
| "learning_rate": 1.9833451091120727e-07, | |
| "loss": 0.1301, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 1.8215892053973013, | |
| "grad_norm": 0.8936712345075764, | |
| "learning_rate": 1.9506428304799095e-07, | |
| "loss": 0.1662, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 1.823088455772114, | |
| "grad_norm": 0.9910283918691146, | |
| "learning_rate": 1.9182070385038555e-07, | |
| "loss": 0.1728, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 1.8245877061469264, | |
| "grad_norm": 0.768050968083848, | |
| "learning_rate": 1.886037913076144e-07, | |
| "loss": 0.1246, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 1.8260869565217392, | |
| "grad_norm": 0.8463002920970154, | |
| "learning_rate": 1.8541356326100436e-07, | |
| "loss": 0.1492, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 1.8275862068965516, | |
| "grad_norm": 0.9274517912935811, | |
| "learning_rate": 1.8225003740388546e-07, | |
| "loss": 0.138, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 1.8290854572713644, | |
| "grad_norm": 0.8249110672258541, | |
| "learning_rate": 1.791132312814975e-07, | |
| "loss": 0.1462, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.8305847076461768, | |
| "grad_norm": 0.9137919311881847, | |
| "learning_rate": 1.760031622908881e-07, | |
| "loss": 0.1496, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 1.8320839580209896, | |
| "grad_norm": 0.9406341312135775, | |
| "learning_rate": 1.729198476808186e-07, | |
| "loss": 0.1304, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 1.8335832083958021, | |
| "grad_norm": 0.993101311314096, | |
| "learning_rate": 1.6986330455166733e-07, | |
| "loss": 0.1545, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 1.8350824587706147, | |
| "grad_norm": 1.1084991506512774, | |
| "learning_rate": 1.6683354985533583e-07, | |
| "loss": 0.1413, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 1.8365817091454273, | |
| "grad_norm": 1.0594304606648992, | |
| "learning_rate": 1.6383060039515343e-07, | |
| "loss": 0.111, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.8380809595202399, | |
| "grad_norm": 1.1676391904339343, | |
| "learning_rate": 1.6085447282578548e-07, | |
| "loss": 0.1714, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 1.8395802098950524, | |
| "grad_norm": 0.8190865187098476, | |
| "learning_rate": 1.579051836531409e-07, | |
| "loss": 0.1336, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 1.841079460269865, | |
| "grad_norm": 0.8516882614964311, | |
| "learning_rate": 1.5498274923427925e-07, | |
| "loss": 0.1191, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 1.8425787106446778, | |
| "grad_norm": 0.8358670807624862, | |
| "learning_rate": 1.5208718577732096e-07, | |
| "loss": 0.1671, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 1.8440779610194902, | |
| "grad_norm": 0.8144177555935608, | |
| "learning_rate": 1.4921850934135785e-07, | |
| "loss": 0.1538, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.845577211394303, | |
| "grad_norm": 1.072239974077697, | |
| "learning_rate": 1.463767358363627e-07, | |
| "loss": 0.1656, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 1.8470764617691153, | |
| "grad_norm": 1.0616204549554635, | |
| "learning_rate": 1.4356188102310266e-07, | |
| "loss": 0.1565, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 1.8485757121439281, | |
| "grad_norm": 0.8459504824524964, | |
| "learning_rate": 1.4077396051305093e-07, | |
| "loss": 0.1478, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 1.8500749625187405, | |
| "grad_norm": 0.7427095632434784, | |
| "learning_rate": 1.3801298976830025e-07, | |
| "loss": 0.109, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 1.8515742128935533, | |
| "grad_norm": 0.9117468537123162, | |
| "learning_rate": 1.3527898410147677e-07, | |
| "loss": 0.1318, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 1.8530734632683659, | |
| "grad_norm": 0.8753131798034437, | |
| "learning_rate": 1.325719586756563e-07, | |
| "loss": 0.1251, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 1.8545727136431784, | |
| "grad_norm": 0.7745219670537332, | |
| "learning_rate": 1.2989192850427933e-07, | |
| "loss": 0.1375, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 1.856071964017991, | |
| "grad_norm": 0.9015294487213966, | |
| "learning_rate": 1.2723890845106723e-07, | |
| "loss": 0.1498, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 1.8575712143928036, | |
| "grad_norm": 0.9431496225634594, | |
| "learning_rate": 1.2461291322994118e-07, | |
| "loss": 0.1615, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 1.8590704647676162, | |
| "grad_norm": 0.8850778040120311, | |
| "learning_rate": 1.2201395740493948e-07, | |
| "loss": 0.1514, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.8605697151424287, | |
| "grad_norm": 0.9186728601066754, | |
| "learning_rate": 1.1944205539013708e-07, | |
| "loss": 0.1026, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 1.8620689655172413, | |
| "grad_norm": 0.7300440246313782, | |
| "learning_rate": 1.1689722144956672e-07, | |
| "loss": 0.1455, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 1.863568215892054, | |
| "grad_norm": 0.817742436211651, | |
| "learning_rate": 1.1437946969713731e-07, | |
| "loss": 0.1415, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 1.8650674662668667, | |
| "grad_norm": 0.9666741334280409, | |
| "learning_rate": 1.1188881409655849e-07, | |
| "loss": 0.1557, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 1.866566716641679, | |
| "grad_norm": 1.0627952474695799, | |
| "learning_rate": 1.0942526846126122e-07, | |
| "loss": 0.1626, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 1.8680659670164919, | |
| "grad_norm": 0.884073305297456, | |
| "learning_rate": 1.0698884645432117e-07, | |
| "loss": 0.1572, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 1.8695652173913042, | |
| "grad_norm": 0.9065268711371114, | |
| "learning_rate": 1.0457956158838545e-07, | |
| "loss": 0.1019, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 1.871064467766117, | |
| "grad_norm": 0.7938329553641408, | |
| "learning_rate": 1.0219742722559433e-07, | |
| "loss": 0.1408, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 1.8725637181409296, | |
| "grad_norm": 0.8919898576059135, | |
| "learning_rate": 9.984245657750857e-08, | |
| "loss": 0.1404, | |
| "step": 1249 | |
| }, | |
| { | |
| "epoch": 1.8740629685157422, | |
| "grad_norm": 1.3329144954478531, | |
| "learning_rate": 9.751466270503718e-08, | |
| "loss": 0.1721, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.8755622188905547, | |
| "grad_norm": 0.9557734164088499, | |
| "learning_rate": 9.521405851836252e-08, | |
| "loss": 0.1604, | |
| "step": 1251 | |
| }, | |
| { | |
| "epoch": 1.8770614692653673, | |
| "grad_norm": 1.0599713957427297, | |
| "learning_rate": 9.294065677687202e-08, | |
| "loss": 0.1595, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 1.87856071964018, | |
| "grad_norm": 0.9185453501467904, | |
| "learning_rate": 9.069447008908383e-08, | |
| "loss": 0.1824, | |
| "step": 1253 | |
| }, | |
| { | |
| "epoch": 1.8800599700149925, | |
| "grad_norm": 0.840567293621228, | |
| "learning_rate": 8.847551091257956e-08, | |
| "loss": 0.1677, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 1.881559220389805, | |
| "grad_norm": 0.8130940930702549, | |
| "learning_rate": 8.62837915539344e-08, | |
| "loss": 0.1314, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 1.8830584707646176, | |
| "grad_norm": 0.8169942278984823, | |
| "learning_rate": 8.411932416864832e-08, | |
| "loss": 0.1242, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 1.8845577211394304, | |
| "grad_norm": 1.0000625060017108, | |
| "learning_rate": 8.198212076107881e-08, | |
| "loss": 0.1707, | |
| "step": 1257 | |
| }, | |
| { | |
| "epoch": 1.8860569715142428, | |
| "grad_norm": 0.9002945466584514, | |
| "learning_rate": 7.987219318437489e-08, | |
| "loss": 0.1387, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 1.8875562218890556, | |
| "grad_norm": 1.112798249862703, | |
| "learning_rate": 7.778955314041103e-08, | |
| "loss": 0.153, | |
| "step": 1259 | |
| }, | |
| { | |
| "epoch": 1.889055472263868, | |
| "grad_norm": 0.9484459661703071, | |
| "learning_rate": 7.573421217972222e-08, | |
| "loss": 0.1665, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.8905547226386807, | |
| "grad_norm": 0.8247537961006699, | |
| "learning_rate": 7.370618170144062e-08, | |
| "loss": 0.1427, | |
| "step": 1261 | |
| }, | |
| { | |
| "epoch": 1.892053973013493, | |
| "grad_norm": 0.793254457704947, | |
| "learning_rate": 7.170547295323016e-08, | |
| "loss": 0.1121, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 1.893553223388306, | |
| "grad_norm": 0.823156719967554, | |
| "learning_rate": 6.973209703122652e-08, | |
| "loss": 0.116, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 1.8950524737631185, | |
| "grad_norm": 0.9727834715369116, | |
| "learning_rate": 6.778606487997496e-08, | |
| "loss": 0.1798, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 1.896551724137931, | |
| "grad_norm": 0.9705802232278228, | |
| "learning_rate": 6.58673872923693e-08, | |
| "loss": 0.1437, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 1.8980509745127436, | |
| "grad_norm": 0.8807208490826871, | |
| "learning_rate": 6.397607490959134e-08, | |
| "loss": 0.1273, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 1.8995502248875562, | |
| "grad_norm": 1.1232273978697145, | |
| "learning_rate": 6.211213822105378e-08, | |
| "loss": 0.1517, | |
| "step": 1267 | |
| }, | |
| { | |
| "epoch": 1.9010494752623688, | |
| "grad_norm": 1.1076600981182103, | |
| "learning_rate": 6.027558756434015e-08, | |
| "loss": 0.1368, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 1.9025487256371814, | |
| "grad_norm": 0.8755244869701937, | |
| "learning_rate": 5.846643312514888e-08, | |
| "loss": 0.1287, | |
| "step": 1269 | |
| }, | |
| { | |
| "epoch": 1.9040479760119942, | |
| "grad_norm": 0.896373431904789, | |
| "learning_rate": 5.668468493723489e-08, | |
| "loss": 0.1095, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.9055472263868065, | |
| "grad_norm": 0.9248601644001376, | |
| "learning_rate": 5.4930352882357486e-08, | |
| "loss": 0.1281, | |
| "step": 1271 | |
| }, | |
| { | |
| "epoch": 1.9070464767616193, | |
| "grad_norm": 1.036061510911137, | |
| "learning_rate": 5.3203446690220374e-08, | |
| "loss": 0.1827, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 1.9085457271364317, | |
| "grad_norm": 0.8734192888608425, | |
| "learning_rate": 5.1503975938422824e-08, | |
| "loss": 0.1136, | |
| "step": 1273 | |
| }, | |
| { | |
| "epoch": 1.9100449775112445, | |
| "grad_norm": 0.9544728456848607, | |
| "learning_rate": 4.983195005240415e-08, | |
| "loss": 0.1469, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 1.9115442278860568, | |
| "grad_norm": 0.9146638778779617, | |
| "learning_rate": 4.8187378305390994e-08, | |
| "loss": 0.1513, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "grad_norm": 0.8621339594183031, | |
| "learning_rate": 4.657026981834623e-08, | |
| "loss": 0.124, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 1.9145427286356822, | |
| "grad_norm": 0.9560003331202336, | |
| "learning_rate": 4.498063355991955e-08, | |
| "loss": 0.1252, | |
| "step": 1277 | |
| }, | |
| { | |
| "epoch": 1.9160419790104948, | |
| "grad_norm": 0.8327447454075502, | |
| "learning_rate": 4.341847834639645e-08, | |
| "loss": 0.1278, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 1.9175412293853074, | |
| "grad_norm": 1.0736183727588948, | |
| "learning_rate": 4.188381284164933e-08, | |
| "loss": 0.1751, | |
| "step": 1279 | |
| }, | |
| { | |
| "epoch": 1.91904047976012, | |
| "grad_norm": 0.9045013770925274, | |
| "learning_rate": 4.0376645557090864e-08, | |
| "loss": 0.1331, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9205397301349325, | |
| "grad_norm": 0.8130460732903044, | |
| "learning_rate": 3.889698485162463e-08, | |
| "loss": 0.1228, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 1.922038980509745, | |
| "grad_norm": 0.9609946230577785, | |
| "learning_rate": 3.744483893160067e-08, | |
| "loss": 0.192, | |
| "step": 1282 | |
| }, | |
| { | |
| "epoch": 1.9235382308845579, | |
| "grad_norm": 0.8203819521380308, | |
| "learning_rate": 3.602021585076942e-08, | |
| "loss": 0.126, | |
| "step": 1283 | |
| }, | |
| { | |
| "epoch": 1.9250374812593702, | |
| "grad_norm": 1.0264638735223461, | |
| "learning_rate": 3.462312351023567e-08, | |
| "loss": 0.1757, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 1.926536731634183, | |
| "grad_norm": 0.8396507855161882, | |
| "learning_rate": 3.325356965841686e-08, | |
| "loss": 0.1004, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 1.9280359820089954, | |
| "grad_norm": 0.8046220758707456, | |
| "learning_rate": 3.191156189099931e-08, | |
| "loss": 0.1727, | |
| "step": 1286 | |
| }, | |
| { | |
| "epoch": 1.9295352323838082, | |
| "grad_norm": 0.8508323226021619, | |
| "learning_rate": 3.0597107650894855e-08, | |
| "loss": 0.1749, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 1.9310344827586206, | |
| "grad_norm": 1.0923733311622432, | |
| "learning_rate": 2.9310214228202016e-08, | |
| "loss": 0.1533, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 1.9325337331334334, | |
| "grad_norm": 0.9847931766958902, | |
| "learning_rate": 2.8050888760163265e-08, | |
| "loss": 0.1169, | |
| "step": 1289 | |
| }, | |
| { | |
| "epoch": 1.934032983508246, | |
| "grad_norm": 0.9660488127891025, | |
| "learning_rate": 2.6819138231126695e-08, | |
| "loss": 0.1904, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.9355322338830585, | |
| "grad_norm": 0.9306158809065935, | |
| "learning_rate": 2.5614969472506634e-08, | |
| "loss": 0.1673, | |
| "step": 1291 | |
| }, | |
| { | |
| "epoch": 1.937031484257871, | |
| "grad_norm": 0.736883486164082, | |
| "learning_rate": 2.4438389162746434e-08, | |
| "loss": 0.1392, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 1.9385307346326837, | |
| "grad_norm": 0.7586902632324848, | |
| "learning_rate": 2.3289403827281287e-08, | |
| "loss": 0.113, | |
| "step": 1293 | |
| }, | |
| { | |
| "epoch": 1.9400299850074962, | |
| "grad_norm": 0.8786788579863171, | |
| "learning_rate": 2.2168019838501032e-08, | |
| "loss": 0.1674, | |
| "step": 1294 | |
| }, | |
| { | |
| "epoch": 1.9415292353823088, | |
| "grad_norm": 0.8325001530221126, | |
| "learning_rate": 2.1074243415716288e-08, | |
| "loss": 0.1092, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 1.9430284857571214, | |
| "grad_norm": 0.8712679758965902, | |
| "learning_rate": 2.0008080625124048e-08, | |
| "loss": 0.1502, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 1.944527736131934, | |
| "grad_norm": 0.8353702240694473, | |
| "learning_rate": 1.896953737977103e-08, | |
| "loss": 0.1163, | |
| "step": 1297 | |
| }, | |
| { | |
| "epoch": 1.9460269865067468, | |
| "grad_norm": 0.8643735038133271, | |
| "learning_rate": 1.7958619439524817e-08, | |
| "loss": 0.18, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 1.9475262368815591, | |
| "grad_norm": 1.0587700188524507, | |
| "learning_rate": 1.6975332411040547e-08, | |
| "loss": 0.1289, | |
| "step": 1299 | |
| }, | |
| { | |
| "epoch": 1.949025487256372, | |
| "grad_norm": 0.8857430111266853, | |
| "learning_rate": 1.601968174772761e-08, | |
| "loss": 0.1363, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.9505247376311843, | |
| "grad_norm": 0.8076621182206598, | |
| "learning_rate": 1.5091672749723564e-08, | |
| "loss": 0.1283, | |
| "step": 1301 | |
| }, | |
| { | |
| "epoch": 1.952023988005997, | |
| "grad_norm": 0.9945062299554772, | |
| "learning_rate": 1.4191310563860806e-08, | |
| "loss": 0.1514, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 1.9535232383808094, | |
| "grad_norm": 0.786379663689969, | |
| "learning_rate": 1.331860018363995e-08, | |
| "loss": 0.1003, | |
| "step": 1303 | |
| }, | |
| { | |
| "epoch": 1.9550224887556222, | |
| "grad_norm": 0.9191874804366194, | |
| "learning_rate": 1.2473546449203178e-08, | |
| "loss": 0.1723, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 1.9565217391304348, | |
| "grad_norm": 0.7270925271825052, | |
| "learning_rate": 1.1656154047303691e-08, | |
| "loss": 0.1266, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 1.9580209895052474, | |
| "grad_norm": 0.856910876432065, | |
| "learning_rate": 1.0866427511285194e-08, | |
| "loss": 0.1347, | |
| "step": 1306 | |
| }, | |
| { | |
| "epoch": 1.95952023988006, | |
| "grad_norm": 0.8790630395335999, | |
| "learning_rate": 1.0104371221050236e-08, | |
| "loss": 0.1041, | |
| "step": 1307 | |
| }, | |
| { | |
| "epoch": 1.9610194902548725, | |
| "grad_norm": 0.6963655848795207, | |
| "learning_rate": 9.369989403041347e-09, | |
| "loss": 0.102, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 1.9625187406296851, | |
| "grad_norm": 0.9349704271479988, | |
| "learning_rate": 8.663286130216608e-09, | |
| "loss": 0.1838, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 1.9640179910044977, | |
| "grad_norm": 1.0374248195431748, | |
| "learning_rate": 7.984265322023011e-09, | |
| "loss": 0.1446, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.9655172413793105, | |
| "grad_norm": 0.8609759689284708, | |
| "learning_rate": 7.332930744380906e-09, | |
| "loss": 0.1612, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 1.9670164917541229, | |
| "grad_norm": 1.0118266520623416, | |
| "learning_rate": 6.709286009657368e-09, | |
| "loss": 0.1645, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 1.9685157421289357, | |
| "grad_norm": 0.9229674712064004, | |
| "learning_rate": 6.1133345766511975e-09, | |
| "loss": 0.2072, | |
| "step": 1313 | |
| }, | |
| { | |
| "epoch": 1.970014992503748, | |
| "grad_norm": 0.946008051655759, | |
| "learning_rate": 5.5450797505690605e-09, | |
| "loss": 0.1593, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 1.9715142428785608, | |
| "grad_norm": 1.1332968361574405, | |
| "learning_rate": 5.004524683011048e-09, | |
| "loss": 0.1601, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 1.9730134932533732, | |
| "grad_norm": 0.9601945691990238, | |
| "learning_rate": 4.491672371950695e-09, | |
| "loss": 0.1617, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 1.974512743628186, | |
| "grad_norm": 0.8519598676400196, | |
| "learning_rate": 4.0065256617199954e-09, | |
| "loss": 0.1256, | |
| "step": 1317 | |
| }, | |
| { | |
| "epoch": 1.9760119940029985, | |
| "grad_norm": 0.9776166003054295, | |
| "learning_rate": 3.5490872429910784e-09, | |
| "loss": 0.1619, | |
| "step": 1318 | |
| }, | |
| { | |
| "epoch": 1.9775112443778111, | |
| "grad_norm": 0.882571776287137, | |
| "learning_rate": 3.119359652765108e-09, | |
| "loss": 0.1954, | |
| "step": 1319 | |
| }, | |
| { | |
| "epoch": 1.9790104947526237, | |
| "grad_norm": 0.8897040655872479, | |
| "learning_rate": 2.7173452743550767e-09, | |
| "loss": 0.1848, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.9805097451274363, | |
| "grad_norm": 0.9352118790095659, | |
| "learning_rate": 2.343046337374144e-09, | |
| "loss": 0.1435, | |
| "step": 1321 | |
| }, | |
| { | |
| "epoch": 1.9820089955022488, | |
| "grad_norm": 0.7809459407888095, | |
| "learning_rate": 1.9964649177223184e-09, | |
| "loss": 0.1274, | |
| "step": 1322 | |
| }, | |
| { | |
| "epoch": 1.9835082458770614, | |
| "grad_norm": 0.8247185430413772, | |
| "learning_rate": 1.6776029375759062e-09, | |
| "loss": 0.1227, | |
| "step": 1323 | |
| }, | |
| { | |
| "epoch": 1.9850074962518742, | |
| "grad_norm": 1.0942577775703748, | |
| "learning_rate": 1.386462165375857e-09, | |
| "loss": 0.1967, | |
| "step": 1324 | |
| }, | |
| { | |
| "epoch": 1.9865067466266866, | |
| "grad_norm": 0.7531145087395039, | |
| "learning_rate": 1.1230442158188804e-09, | |
| "loss": 0.1353, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.9880059970014994, | |
| "grad_norm": 1.0368429073628775, | |
| "learning_rate": 8.873505498474544e-10, | |
| "loss": 0.1356, | |
| "step": 1326 | |
| }, | |
| { | |
| "epoch": 1.9895052473763117, | |
| "grad_norm": 0.9311934844484981, | |
| "learning_rate": 6.793824746437194e-10, | |
| "loss": 0.1791, | |
| "step": 1327 | |
| }, | |
| { | |
| "epoch": 1.9910044977511245, | |
| "grad_norm": 0.9800148413207548, | |
| "learning_rate": 4.991411436189308e-10, | |
| "loss": 0.1405, | |
| "step": 1328 | |
| }, | |
| { | |
| "epoch": 1.992503748125937, | |
| "grad_norm": 0.8262478018062133, | |
| "learning_rate": 3.466275564101285e-10, | |
| "loss": 0.1285, | |
| "step": 1329 | |
| }, | |
| { | |
| "epoch": 1.9940029985007497, | |
| "grad_norm": 0.9646879650594153, | |
| "learning_rate": 2.2184255887403028e-10, | |
| "loss": 0.1585, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.9955022488755623, | |
| "grad_norm": 0.8166980422406553, | |
| "learning_rate": 1.2478684308037115e-10, | |
| "loss": 0.1302, | |
| "step": 1331 | |
| }, | |
| { | |
| "epoch": 1.9970014992503748, | |
| "grad_norm": 0.8000105351035276, | |
| "learning_rate": 5.5460947310237435e-11, | |
| "loss": 0.1289, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 1.9985007496251874, | |
| "grad_norm": 0.8189664806954898, | |
| "learning_rate": 1.3865256052181252e-11, | |
| "loss": 0.132, | |
| "step": 1333 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7674469143234484, | |
| "learning_rate": 0.0, | |
| "loss": 0.0956, | |
| "step": 1334 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 1334, | |
| "total_flos": 39377831657472.0, | |
| "train_loss": 0.23282673676265053, | |
| "train_runtime": 3822.3446, | |
| "train_samples_per_second": 2.788, | |
| "train_steps_per_second": 0.349 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1334, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 39377831657472.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |