| { |
| "best_metric": 10.21642017364502, |
| "best_model_checkpoint": "tiny_bert_rand_50_v2/checkpoint-10000", |
| "epoch": 25.0, |
| "eval_steps": 10000, |
| "global_step": 59550, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.20990764063811923, |
| "grad_norm": 2.5449929237365723, |
| "learning_rate": 5e-06, |
| "loss": 13.7115, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.41981528127623846, |
| "grad_norm": 1.733461618423462, |
| "learning_rate": 1e-05, |
| "loss": 12.1301, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.6297229219143576, |
| "grad_norm": 1.4017963409423828, |
| "learning_rate": 1.5e-05, |
| "loss": 11.1362, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.8396305625524769, |
| "grad_norm": 1.5763887166976929, |
| "learning_rate": 2e-05, |
| "loss": 10.9361, |
| "step": 2000 |
| }, |
| { |
| "epoch": 1.0495382031905962, |
| "grad_norm": 1.620408296585083, |
| "learning_rate": 2.5e-05, |
| "loss": 10.8202, |
| "step": 2500 |
| }, |
| { |
| "epoch": 1.2594458438287153, |
| "grad_norm": 1.6337733268737793, |
| "learning_rate": 3e-05, |
| "loss": 10.713, |
| "step": 3000 |
| }, |
| { |
| "epoch": 1.4693534844668346, |
| "grad_norm": 1.746281623840332, |
| "learning_rate": 3.5e-05, |
| "loss": 10.6258, |
| "step": 3500 |
| }, |
| { |
| "epoch": 1.6792611251049538, |
| "grad_norm": 1.722418189048767, |
| "learning_rate": 4e-05, |
| "loss": 10.5436, |
| "step": 4000 |
| }, |
| { |
| "epoch": 1.8891687657430731, |
| "grad_norm": 1.5402530431747437, |
| "learning_rate": 4.5e-05, |
| "loss": 10.47, |
| "step": 4500 |
| }, |
| { |
| "epoch": 2.0990764063811924, |
| "grad_norm": 1.6758112907409668, |
| "learning_rate": 5e-05, |
| "loss": 10.4045, |
| "step": 5000 |
| }, |
| { |
| "epoch": 2.3089840470193117, |
| "grad_norm": 1.7747454643249512, |
| "learning_rate": 5.500000000000001e-05, |
| "loss": 10.357, |
| "step": 5500 |
| }, |
| { |
| "epoch": 2.5188916876574305, |
| "grad_norm": 1.7545584440231323, |
| "learning_rate": 6e-05, |
| "loss": 10.3254, |
| "step": 6000 |
| }, |
| { |
| "epoch": 2.72879932829555, |
| "grad_norm": 1.6928805112838745, |
| "learning_rate": 6.500000000000001e-05, |
| "loss": 10.2987, |
| "step": 6500 |
| }, |
| { |
| "epoch": 2.938706968933669, |
| "grad_norm": 1.963725209236145, |
| "learning_rate": 7e-05, |
| "loss": 10.275, |
| "step": 7000 |
| }, |
| { |
| "epoch": 3.1486146095717884, |
| "grad_norm": 1.743053913116455, |
| "learning_rate": 7.500000000000001e-05, |
| "loss": 10.2171, |
| "step": 7500 |
| }, |
| { |
| "epoch": 3.3585222502099077, |
| "grad_norm": 2.0638396739959717, |
| "learning_rate": 8e-05, |
| "loss": 10.1824, |
| "step": 8000 |
| }, |
| { |
| "epoch": 3.568429890848027, |
| "grad_norm": 1.9554697275161743, |
| "learning_rate": 8.5e-05, |
| "loss": 10.1727, |
| "step": 8500 |
| }, |
| { |
| "epoch": 3.7783375314861463, |
| "grad_norm": 2.269812822341919, |
| "learning_rate": 9e-05, |
| "loss": 10.149, |
| "step": 9000 |
| }, |
| { |
| "epoch": 3.988245172124265, |
| "grad_norm": 1.9684901237487793, |
| "learning_rate": 9.5e-05, |
| "loss": 10.1363, |
| "step": 9500 |
| }, |
| { |
| "epoch": 4.198152812762385, |
| "grad_norm": 2.655996084213257, |
| "learning_rate": 0.0001, |
| "loss": 9.9984, |
| "step": 10000 |
| }, |
| { |
| "epoch": 4.198152812762385, |
| "eval_accuracy": 0.14992069896213028, |
| "eval_loss": 10.21642017364502, |
| "eval_runtime": 0.7154, |
| "eval_samples_per_second": 669.559, |
| "eval_steps_per_second": 6.989, |
| "step": 10000 |
| }, |
| { |
| "epoch": 4.408060453400504, |
| "grad_norm": 2.357985734939575, |
| "learning_rate": 9.899091826437942e-05, |
| "loss": 9.9935, |
| "step": 10500 |
| }, |
| { |
| "epoch": 4.617968094038623, |
| "grad_norm": 2.9752063751220703, |
| "learning_rate": 9.798183652875883e-05, |
| "loss": 9.9867, |
| "step": 11000 |
| }, |
| { |
| "epoch": 4.827875734676742, |
| "grad_norm": 2.696370840072632, |
| "learning_rate": 9.697275479313825e-05, |
| "loss": 9.9912, |
| "step": 11500 |
| }, |
| { |
| "epoch": 5.037783375314861, |
| "grad_norm": 2.8532896041870117, |
| "learning_rate": 9.596367305751767e-05, |
| "loss": 9.9497, |
| "step": 12000 |
| }, |
| { |
| "epoch": 5.247691015952981, |
| "grad_norm": 2.9102783203125, |
| "learning_rate": 9.495459132189707e-05, |
| "loss": 9.8021, |
| "step": 12500 |
| }, |
| { |
| "epoch": 5.4575986565911, |
| "grad_norm": 2.8862690925598145, |
| "learning_rate": 9.394550958627649e-05, |
| "loss": 9.8164, |
| "step": 13000 |
| }, |
| { |
| "epoch": 5.667506297229219, |
| "grad_norm": 2.739621639251709, |
| "learning_rate": 9.293642785065591e-05, |
| "loss": 9.8376, |
| "step": 13500 |
| }, |
| { |
| "epoch": 5.877413937867338, |
| "grad_norm": 2.729461669921875, |
| "learning_rate": 9.192734611503532e-05, |
| "loss": 9.8429, |
| "step": 14000 |
| }, |
| { |
| "epoch": 6.087321578505458, |
| "grad_norm": 2.7407426834106445, |
| "learning_rate": 9.091826437941474e-05, |
| "loss": 9.7661, |
| "step": 14500 |
| }, |
| { |
| "epoch": 6.297229219143577, |
| "grad_norm": 2.794342517852783, |
| "learning_rate": 8.990918264379415e-05, |
| "loss": 9.684, |
| "step": 15000 |
| }, |
| { |
| "epoch": 6.507136859781696, |
| "grad_norm": 2.880253314971924, |
| "learning_rate": 8.890010090817356e-05, |
| "loss": 9.706, |
| "step": 15500 |
| }, |
| { |
| "epoch": 6.717044500419815, |
| "grad_norm": 2.842390537261963, |
| "learning_rate": 8.789101917255298e-05, |
| "loss": 9.7289, |
| "step": 16000 |
| }, |
| { |
| "epoch": 6.926952141057934, |
| "grad_norm": 3.207631826400757, |
| "learning_rate": 8.68819374369324e-05, |
| "loss": 9.7433, |
| "step": 16500 |
| }, |
| { |
| "epoch": 7.136859781696054, |
| "grad_norm": 2.994023084640503, |
| "learning_rate": 8.58728557013118e-05, |
| "loss": 9.6217, |
| "step": 17000 |
| }, |
| { |
| "epoch": 7.346767422334173, |
| "grad_norm": 3.4115116596221924, |
| "learning_rate": 8.486377396569122e-05, |
| "loss": 9.5814, |
| "step": 17500 |
| }, |
| { |
| "epoch": 7.5566750629722925, |
| "grad_norm": 2.95430588722229, |
| "learning_rate": 8.385469223007064e-05, |
| "loss": 9.6118, |
| "step": 18000 |
| }, |
| { |
| "epoch": 7.766582703610411, |
| "grad_norm": 2.8346710205078125, |
| "learning_rate": 8.284561049445006e-05, |
| "loss": 9.642, |
| "step": 18500 |
| }, |
| { |
| "epoch": 7.97649034424853, |
| "grad_norm": 2.9392590522766113, |
| "learning_rate": 8.183652875882947e-05, |
| "loss": 9.6546, |
| "step": 19000 |
| }, |
| { |
| "epoch": 8.18639798488665, |
| "grad_norm": 3.0181374549865723, |
| "learning_rate": 8.082744702320889e-05, |
| "loss": 9.4845, |
| "step": 19500 |
| }, |
| { |
| "epoch": 8.39630562552477, |
| "grad_norm": 3.2455642223358154, |
| "learning_rate": 7.98183652875883e-05, |
| "loss": 9.4873, |
| "step": 20000 |
| }, |
| { |
| "epoch": 8.39630562552477, |
| "eval_accuracy": 0.15309071522769999, |
| "eval_loss": 10.371352195739746, |
| "eval_runtime": 0.7168, |
| "eval_samples_per_second": 668.219, |
| "eval_steps_per_second": 6.975, |
| "step": 20000 |
| }, |
| { |
| "epoch": 8.606213266162888, |
| "grad_norm": 3.4421379566192627, |
| "learning_rate": 7.880928355196771e-05, |
| "loss": 9.5151, |
| "step": 20500 |
| }, |
| { |
| "epoch": 8.816120906801007, |
| "grad_norm": 3.415717124938965, |
| "learning_rate": 7.780020181634713e-05, |
| "loss": 9.5479, |
| "step": 21000 |
| }, |
| { |
| "epoch": 9.026028547439127, |
| "grad_norm": 3.7510669231414795, |
| "learning_rate": 7.679112008072655e-05, |
| "loss": 9.5205, |
| "step": 21500 |
| }, |
| { |
| "epoch": 9.235936188077247, |
| "grad_norm": 3.8339271545410156, |
| "learning_rate": 7.578203834510595e-05, |
| "loss": 9.3369, |
| "step": 22000 |
| }, |
| { |
| "epoch": 9.445843828715365, |
| "grad_norm": 3.8937506675720215, |
| "learning_rate": 7.477295660948537e-05, |
| "loss": 9.3643, |
| "step": 22500 |
| }, |
| { |
| "epoch": 9.655751469353484, |
| "grad_norm": 3.6239004135131836, |
| "learning_rate": 7.376387487386479e-05, |
| "loss": 9.4036, |
| "step": 23000 |
| }, |
| { |
| "epoch": 9.865659109991604, |
| "grad_norm": 4.1243085861206055, |
| "learning_rate": 7.27547931382442e-05, |
| "loss": 9.4103, |
| "step": 23500 |
| }, |
| { |
| "epoch": 10.075566750629722, |
| "grad_norm": 4.432008743286133, |
| "learning_rate": 7.174571140262362e-05, |
| "loss": 9.3275, |
| "step": 24000 |
| }, |
| { |
| "epoch": 10.285474391267842, |
| "grad_norm": 4.084072113037109, |
| "learning_rate": 7.073662966700304e-05, |
| "loss": 9.1847, |
| "step": 24500 |
| }, |
| { |
| "epoch": 10.495382031905962, |
| "grad_norm": 4.277645587921143, |
| "learning_rate": 6.972754793138244e-05, |
| "loss": 9.2019, |
| "step": 25000 |
| }, |
| { |
| "epoch": 10.705289672544081, |
| "grad_norm": 4.283336162567139, |
| "learning_rate": 6.871846619576186e-05, |
| "loss": 9.2246, |
| "step": 25500 |
| }, |
| { |
| "epoch": 10.9151973131822, |
| "grad_norm": 4.644393444061279, |
| "learning_rate": 6.770938446014128e-05, |
| "loss": 9.2518, |
| "step": 26000 |
| }, |
| { |
| "epoch": 11.125104953820319, |
| "grad_norm": 4.699460983276367, |
| "learning_rate": 6.670030272452068e-05, |
| "loss": 9.0844, |
| "step": 26500 |
| }, |
| { |
| "epoch": 11.335012594458439, |
| "grad_norm": 5.40618896484375, |
| "learning_rate": 6.56912209889001e-05, |
| "loss": 8.9917, |
| "step": 27000 |
| }, |
| { |
| "epoch": 11.544920235096557, |
| "grad_norm": 4.965311527252197, |
| "learning_rate": 6.468213925327952e-05, |
| "loss": 9.0226, |
| "step": 27500 |
| }, |
| { |
| "epoch": 11.754827875734676, |
| "grad_norm": 4.865447998046875, |
| "learning_rate": 6.367305751765893e-05, |
| "loss": 9.0319, |
| "step": 28000 |
| }, |
| { |
| "epoch": 11.964735516372796, |
| "grad_norm": 5.462146282196045, |
| "learning_rate": 6.266397578203835e-05, |
| "loss": 9.063, |
| "step": 28500 |
| }, |
| { |
| "epoch": 12.174643157010916, |
| "grad_norm": 5.475606918334961, |
| "learning_rate": 6.165489404641777e-05, |
| "loss": 8.8053, |
| "step": 29000 |
| }, |
| { |
| "epoch": 12.384550797649034, |
| "grad_norm": 5.42305850982666, |
| "learning_rate": 6.064581231079718e-05, |
| "loss": 8.7714, |
| "step": 29500 |
| }, |
| { |
| "epoch": 12.594458438287154, |
| "grad_norm": 5.619285583496094, |
| "learning_rate": 5.963673057517659e-05, |
| "loss": 8.8082, |
| "step": 30000 |
| }, |
| { |
| "epoch": 12.594458438287154, |
| "eval_accuracy": 0.15314659876318662, |
| "eval_loss": 11.017194747924805, |
| "eval_runtime": 0.7112, |
| "eval_samples_per_second": 673.544, |
| "eval_steps_per_second": 7.031, |
| "step": 30000 |
| }, |
| { |
| "epoch": 12.804366078925273, |
| "grad_norm": 5.700458526611328, |
| "learning_rate": 5.8627648839556e-05, |
| "loss": 8.8299, |
| "step": 30500 |
| }, |
| { |
| "epoch": 13.014273719563391, |
| "grad_norm": 6.359976768493652, |
| "learning_rate": 5.7618567103935416e-05, |
| "loss": 8.824, |
| "step": 31000 |
| }, |
| { |
| "epoch": 13.224181360201511, |
| "grad_norm": 5.780282497406006, |
| "learning_rate": 5.660948536831484e-05, |
| "loss": 8.5267, |
| "step": 31500 |
| }, |
| { |
| "epoch": 13.43408900083963, |
| "grad_norm": 6.6382737159729, |
| "learning_rate": 5.5600403632694254e-05, |
| "loss": 8.5649, |
| "step": 32000 |
| }, |
| { |
| "epoch": 13.64399664147775, |
| "grad_norm": 7.126103401184082, |
| "learning_rate": 5.4591321897073666e-05, |
| "loss": 8.5992, |
| "step": 32500 |
| }, |
| { |
| "epoch": 13.853904282115868, |
| "grad_norm": 6.167848587036133, |
| "learning_rate": 5.358224016145308e-05, |
| "loss": 8.6177, |
| "step": 33000 |
| }, |
| { |
| "epoch": 14.063811922753988, |
| "grad_norm": 6.446425914764404, |
| "learning_rate": 5.257315842583249e-05, |
| "loss": 8.5221, |
| "step": 33500 |
| }, |
| { |
| "epoch": 14.273719563392108, |
| "grad_norm": 7.162424087524414, |
| "learning_rate": 5.15640766902119e-05, |
| "loss": 8.3115, |
| "step": 34000 |
| }, |
| { |
| "epoch": 14.483627204030226, |
| "grad_norm": 6.552924633026123, |
| "learning_rate": 5.055499495459133e-05, |
| "loss": 8.3552, |
| "step": 34500 |
| }, |
| { |
| "epoch": 14.693534844668346, |
| "grad_norm": 7.154749870300293, |
| "learning_rate": 4.954591321897074e-05, |
| "loss": 8.3941, |
| "step": 35000 |
| }, |
| { |
| "epoch": 14.903442485306465, |
| "grad_norm": 7.518148422241211, |
| "learning_rate": 4.853683148335015e-05, |
| "loss": 8.398, |
| "step": 35500 |
| }, |
| { |
| "epoch": 15.113350125944585, |
| "grad_norm": 7.958575248718262, |
| "learning_rate": 4.7527749747729566e-05, |
| "loss": 8.2273, |
| "step": 36000 |
| }, |
| { |
| "epoch": 15.323257766582703, |
| "grad_norm": 7.436439514160156, |
| "learning_rate": 4.6518668012108985e-05, |
| "loss": 8.1238, |
| "step": 36500 |
| }, |
| { |
| "epoch": 15.533165407220823, |
| "grad_norm": 7.364752292633057, |
| "learning_rate": 4.55095862764884e-05, |
| "loss": 8.1421, |
| "step": 37000 |
| }, |
| { |
| "epoch": 15.743073047858942, |
| "grad_norm": 7.644779682159424, |
| "learning_rate": 4.450050454086781e-05, |
| "loss": 8.1782, |
| "step": 37500 |
| }, |
| { |
| "epoch": 15.95298068849706, |
| "grad_norm": 7.473808288574219, |
| "learning_rate": 4.349142280524723e-05, |
| "loss": 8.2065, |
| "step": 38000 |
| }, |
| { |
| "epoch": 16.162888329135182, |
| "grad_norm": 8.74830436706543, |
| "learning_rate": 4.248234106962664e-05, |
| "loss": 7.9658, |
| "step": 38500 |
| }, |
| { |
| "epoch": 16.3727959697733, |
| "grad_norm": 8.685850143432617, |
| "learning_rate": 4.147325933400605e-05, |
| "loss": 7.9293, |
| "step": 39000 |
| }, |
| { |
| "epoch": 16.582703610411418, |
| "grad_norm": 8.217513084411621, |
| "learning_rate": 4.046417759838547e-05, |
| "loss": 7.9615, |
| "step": 39500 |
| }, |
| { |
| "epoch": 16.79261125104954, |
| "grad_norm": 8.556501388549805, |
| "learning_rate": 3.9455095862764884e-05, |
| "loss": 7.984, |
| "step": 40000 |
| }, |
| { |
| "epoch": 16.79261125104954, |
| "eval_accuracy": 0.15134013681731523, |
| "eval_loss": 11.922234535217285, |
| "eval_runtime": 0.7099, |
| "eval_samples_per_second": 674.731, |
| "eval_steps_per_second": 7.043, |
| "step": 40000 |
| }, |
| { |
| "epoch": 17.002518891687657, |
| "grad_norm": 7.801002025604248, |
| "learning_rate": 3.84460141271443e-05, |
| "loss": 8.0097, |
| "step": 40500 |
| }, |
| { |
| "epoch": 17.212426532325775, |
| "grad_norm": 9.762965202331543, |
| "learning_rate": 3.7436932391523715e-05, |
| "loss": 7.7151, |
| "step": 41000 |
| }, |
| { |
| "epoch": 17.422334172963897, |
| "grad_norm": 9.19284725189209, |
| "learning_rate": 3.642785065590313e-05, |
| "loss": 7.7477, |
| "step": 41500 |
| }, |
| { |
| "epoch": 17.632241813602015, |
| "grad_norm": 10.254321098327637, |
| "learning_rate": 3.541876892028255e-05, |
| "loss": 7.7802, |
| "step": 42000 |
| }, |
| { |
| "epoch": 17.842149454240133, |
| "grad_norm": 8.671323776245117, |
| "learning_rate": 3.440968718466196e-05, |
| "loss": 7.8088, |
| "step": 42500 |
| }, |
| { |
| "epoch": 18.052057094878254, |
| "grad_norm": 9.769850730895996, |
| "learning_rate": 3.340060544904137e-05, |
| "loss": 7.7406, |
| "step": 43000 |
| }, |
| { |
| "epoch": 18.261964735516372, |
| "grad_norm": 8.743345260620117, |
| "learning_rate": 3.239152371342079e-05, |
| "loss": 7.5693, |
| "step": 43500 |
| }, |
| { |
| "epoch": 18.471872376154494, |
| "grad_norm": 9.809800148010254, |
| "learning_rate": 3.13824419778002e-05, |
| "loss": 7.5856, |
| "step": 44000 |
| }, |
| { |
| "epoch": 18.68178001679261, |
| "grad_norm": 9.659799575805664, |
| "learning_rate": 3.0373360242179615e-05, |
| "loss": 7.6268, |
| "step": 44500 |
| }, |
| { |
| "epoch": 18.89168765743073, |
| "grad_norm": 9.788838386535645, |
| "learning_rate": 2.9364278506559034e-05, |
| "loss": 7.6434, |
| "step": 45000 |
| }, |
| { |
| "epoch": 19.10159529806885, |
| "grad_norm": 10.07334041595459, |
| "learning_rate": 2.8355196770938446e-05, |
| "loss": 7.5285, |
| "step": 45500 |
| }, |
| { |
| "epoch": 19.31150293870697, |
| "grad_norm": 11.367331504821777, |
| "learning_rate": 2.734611503531786e-05, |
| "loss": 7.4214, |
| "step": 46000 |
| }, |
| { |
| "epoch": 19.521410579345087, |
| "grad_norm": 10.628447532653809, |
| "learning_rate": 2.6337033299697278e-05, |
| "loss": 7.445, |
| "step": 46500 |
| }, |
| { |
| "epoch": 19.73131821998321, |
| "grad_norm": 11.207023620605469, |
| "learning_rate": 2.532795156407669e-05, |
| "loss": 7.459, |
| "step": 47000 |
| }, |
| { |
| "epoch": 19.941225860621326, |
| "grad_norm": 10.480270385742188, |
| "learning_rate": 2.4318869828456106e-05, |
| "loss": 7.473, |
| "step": 47500 |
| }, |
| { |
| "epoch": 20.151133501259444, |
| "grad_norm": 10.361275672912598, |
| "learning_rate": 2.330978809283552e-05, |
| "loss": 7.3493, |
| "step": 48000 |
| }, |
| { |
| "epoch": 20.361041141897566, |
| "grad_norm": 11.062283515930176, |
| "learning_rate": 2.2300706357214934e-05, |
| "loss": 7.291, |
| "step": 48500 |
| }, |
| { |
| "epoch": 20.570948782535684, |
| "grad_norm": 9.788945198059082, |
| "learning_rate": 2.129162462159435e-05, |
| "loss": 7.3206, |
| "step": 49000 |
| }, |
| { |
| "epoch": 20.7808564231738, |
| "grad_norm": 10.5712308883667, |
| "learning_rate": 2.0282542885973765e-05, |
| "loss": 7.3331, |
| "step": 49500 |
| }, |
| { |
| "epoch": 20.990764063811923, |
| "grad_norm": 9.677511215209961, |
| "learning_rate": 1.927346115035318e-05, |
| "loss": 7.3546, |
| "step": 50000 |
| }, |
| { |
| "epoch": 20.990764063811923, |
| "eval_accuracy": 0.1517649367762918, |
| "eval_loss": 12.674510955810547, |
| "eval_runtime": 0.6948, |
| "eval_samples_per_second": 689.38, |
| "eval_steps_per_second": 7.196, |
| "step": 50000 |
| }, |
| { |
| "epoch": 21.20067170445004, |
| "grad_norm": 9.962433815002441, |
| "learning_rate": 1.8264379414732593e-05, |
| "loss": 7.1723, |
| "step": 50500 |
| }, |
| { |
| "epoch": 21.410579345088163, |
| "grad_norm": 9.719462394714355, |
| "learning_rate": 1.725529767911201e-05, |
| "loss": 7.1932, |
| "step": 51000 |
| }, |
| { |
| "epoch": 21.62048698572628, |
| "grad_norm": 11.04371166229248, |
| "learning_rate": 1.6246215943491424e-05, |
| "loss": 7.2175, |
| "step": 51500 |
| }, |
| { |
| "epoch": 21.8303946263644, |
| "grad_norm": 10.860447883605957, |
| "learning_rate": 1.5237134207870837e-05, |
| "loss": 7.2211, |
| "step": 52000 |
| }, |
| { |
| "epoch": 22.04030226700252, |
| "grad_norm": 11.007928848266602, |
| "learning_rate": 1.4228052472250252e-05, |
| "loss": 7.1991, |
| "step": 52500 |
| }, |
| { |
| "epoch": 22.250209907640638, |
| "grad_norm": 8.327049255371094, |
| "learning_rate": 1.3218970736629668e-05, |
| "loss": 7.0927, |
| "step": 53000 |
| }, |
| { |
| "epoch": 22.460117548278756, |
| "grad_norm": 10.348093032836914, |
| "learning_rate": 1.2209889001009082e-05, |
| "loss": 7.0983, |
| "step": 53500 |
| }, |
| { |
| "epoch": 22.670025188916878, |
| "grad_norm": 11.09190559387207, |
| "learning_rate": 1.1200807265388496e-05, |
| "loss": 7.1179, |
| "step": 54000 |
| }, |
| { |
| "epoch": 22.879932829554996, |
| "grad_norm": 9.300477027893066, |
| "learning_rate": 1.0191725529767912e-05, |
| "loss": 7.1152, |
| "step": 54500 |
| }, |
| { |
| "epoch": 23.089840470193113, |
| "grad_norm": 9.715547561645508, |
| "learning_rate": 9.182643794147326e-06, |
| "loss": 7.084, |
| "step": 55000 |
| }, |
| { |
| "epoch": 23.299748110831235, |
| "grad_norm": 11.691274642944336, |
| "learning_rate": 8.173562058526741e-06, |
| "loss": 7.0148, |
| "step": 55500 |
| }, |
| { |
| "epoch": 23.509655751469353, |
| "grad_norm": 8.484163284301758, |
| "learning_rate": 7.164480322906156e-06, |
| "loss": 7.0431, |
| "step": 56000 |
| }, |
| { |
| "epoch": 23.719563392107474, |
| "grad_norm": 10.150425910949707, |
| "learning_rate": 6.155398587285571e-06, |
| "loss": 7.0263, |
| "step": 56500 |
| }, |
| { |
| "epoch": 23.929471032745592, |
| "grad_norm": 11.66572093963623, |
| "learning_rate": 5.146316851664986e-06, |
| "loss": 7.0382, |
| "step": 57000 |
| }, |
| { |
| "epoch": 24.13937867338371, |
| "grad_norm": 10.117345809936523, |
| "learning_rate": 4.1372351160444e-06, |
| "loss": 7.0063, |
| "step": 57500 |
| }, |
| { |
| "epoch": 24.349286314021832, |
| "grad_norm": 10.38186264038086, |
| "learning_rate": 3.1281533804238145e-06, |
| "loss": 6.9944, |
| "step": 58000 |
| }, |
| { |
| "epoch": 24.55919395465995, |
| "grad_norm": 10.669214248657227, |
| "learning_rate": 2.1190716448032293e-06, |
| "loss": 6.9936, |
| "step": 58500 |
| }, |
| { |
| "epoch": 24.769101595298068, |
| "grad_norm": 9.23267936706543, |
| "learning_rate": 1.109989909182644e-06, |
| "loss": 6.9786, |
| "step": 59000 |
| }, |
| { |
| "epoch": 24.97900923593619, |
| "grad_norm": 10.472164154052734, |
| "learning_rate": 1.0090817356205853e-07, |
| "loss": 6.9906, |
| "step": 59500 |
| }, |
| { |
| "epoch": 25.0, |
| "step": 59550, |
| "total_flos": 3.053850510375936e+17, |
| "train_loss": 8.778843151640231, |
| "train_runtime": 11935.3116, |
| "train_samples_per_second": 478.913, |
| "train_steps_per_second": 4.989 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 59550, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 25, |
| "save_steps": 10000, |
| "total_flos": 3.053850510375936e+17, |
| "train_batch_size": 96, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
