| { | |
| "best_global_step": 30000, | |
| "best_metric": 6.6873650550842285, | |
| "best_model_checkpoint": "bert_base_km_20_v1/checkpoint-30000", | |
| "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": 3.258617639541626, | |
| "learning_rate": 5e-06, | |
| "loss": 12.1717, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 11.654047012329102, | |
| "learning_rate": 1e-05, | |
| "loss": 10.0315, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 35.01475143432617, | |
| "learning_rate": 1.5e-05, | |
| "loss": 8.522, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 10.415183067321777, | |
| "learning_rate": 2e-05, | |
| "loss": 8.0718, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 23.154544830322266, | |
| "learning_rate": 2.5e-05, | |
| "loss": 7.8358, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 23.617631912231445, | |
| "learning_rate": 3e-05, | |
| "loss": 7.6598, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 7.2276506423950195, | |
| "learning_rate": 3.5e-05, | |
| "loss": 7.5186, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 11.929610252380371, | |
| "learning_rate": 4e-05, | |
| "loss": 7.406, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 8.51960277557373, | |
| "learning_rate": 4.5e-05, | |
| "loss": 7.3016, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 5.851307392120361, | |
| "learning_rate": 5e-05, | |
| "loss": 7.1989, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 5.41884708404541, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 7.1117, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 7.0933685302734375, | |
| "learning_rate": 6e-05, | |
| "loss": 7.0736, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 5.4864349365234375, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 7.0244, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 8.095874786376953, | |
| "learning_rate": 7e-05, | |
| "loss": 6.9969, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 7.268009662628174, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 6.9144, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 5.263991355895996, | |
| "learning_rate": 8e-05, | |
| "loss": 6.8943, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 4.145423412322998, | |
| "learning_rate": 8.5e-05, | |
| "loss": 6.8751, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 5.825566291809082, | |
| "learning_rate": 9e-05, | |
| "loss": 6.8505, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 2.4331114292144775, | |
| "learning_rate": 9.5e-05, | |
| "loss": 6.8354, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 3.7852976322174072, | |
| "learning_rate": 0.0001, | |
| "loss": 6.7411, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15022677314338184, | |
| "eval_loss": 6.849635601043701, | |
| "eval_runtime": 1.0477, | |
| "eval_samples_per_second": 457.175, | |
| "eval_steps_per_second": 4.772, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 3.4165592193603516, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 6.7407, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 2.9745731353759766, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 6.7353, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 3.7354421615600586, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 6.7209, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 3.3837056159973145, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 6.6722, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 3.4129574298858643, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 6.6028, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 2.4542360305786133, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 6.5994, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 4.3186259269714355, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 6.6004, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 2.3866515159606934, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 6.5859, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.493431806564331, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 6.5383, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 2.3935134410858154, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 6.493, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 2.083798885345459, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 6.4959, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 3.8829052448272705, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 6.4908, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 2.9645934104919434, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 6.4852, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 2.5192222595214844, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 6.4163, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 3.0464894771575928, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 6.3987, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 2.7907907962799072, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 6.4096, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 3.183791160583496, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 6.404, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 3.0189096927642822, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 6.4085, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 2.3483211994171143, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 6.3269, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 3.180586338043213, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 6.3276, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.1524172956592497, | |
| "eval_loss": 6.69942569732666, | |
| "eval_runtime": 1.0242, | |
| "eval_samples_per_second": 467.666, | |
| "eval_steps_per_second": 4.882, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 3.1962997913360596, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 6.3298, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 2.635688304901123, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 6.3286, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 1.8444244861602783, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 6.3191, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 3.1084392070770264, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 6.2553, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 2.5472848415374756, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 6.2592, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 6.046384334564209, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 6.2616, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 3.51747465133667, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 6.2656, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 5.2468767166137695, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 6.2321, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 3.305779218673706, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 6.1998, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 2.321261167526245, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 6.2012, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 2.6385366916656494, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 6.2051, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 3.393951654434204, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 6.2093, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 2.333998203277588, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 6.1679, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 3.1242072582244873, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 6.1552, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 2.9733433723449707, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 6.1493, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 4.289365768432617, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 6.1527, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 3.1729888916015625, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 6.1571, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 3.0594305992126465, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 6.121, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 2.6519603729248047, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 6.1069, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 2.6753199100494385, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 6.1116, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15460167333575844, | |
| "eval_loss": 6.6873650550842285, | |
| "eval_runtime": 1.036, | |
| "eval_samples_per_second": 462.364, | |
| "eval_steps_per_second": 4.826, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 3.8139243125915527, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 6.1048, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 4.035758018493652, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 6.1094, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 2.0152852535247803, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 6.0719, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 3.9129137992858887, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 6.0678, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 3.3388304710388184, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 6.0794, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 3.540177822113037, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 6.0766, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 4.535055637359619, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 6.0594, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 1.7694422006607056, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 6.04, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 2.629112720489502, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 6.0451, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 4.2721757888793945, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 6.0445, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 3.537931442260742, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 6.0387, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 0.8599244952201843, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 6.0228, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 3.7414236068725586, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.0206, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 2.032564878463745, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.0151, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 3.004331588745117, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 6.0133, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 3.1127686500549316, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 6.0125, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 1.646429181098938, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 5.997, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 3.3201920986175537, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 5.9885, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 3.53049635887146, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 5.9929, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 3.361233711242676, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 5.9847, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.1527139172367388, | |
| "eval_loss": 7.070965766906738, | |
| "eval_runtime": 1.0216, | |
| "eval_samples_per_second": 468.883, | |
| "eval_steps_per_second": 4.894, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 3.392704725265503, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 5.9884, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 3.117615222930908, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 5.9699, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 1.7372816801071167, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 5.9636, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 3.3655130863189697, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 5.9684, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 2.850968599319458, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 5.9656, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 2.872636318206787, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 5.9582, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 1.0639994144439697, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 5.9489, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 3.5935397148132324, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 5.9461, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 1.488255262374878, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 5.9487, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 4.531470775604248, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 5.9489, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 2.673673152923584, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 5.9409, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 2.250357151031494, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 5.9267, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 2.1105315685272217, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 5.9306, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 2.690736770629883, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 5.9291, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 1.2354042530059814, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 5.9234, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 2.649017572402954, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 5.9185, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 1.3641419410705566, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 5.9142, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 1.5346912145614624, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 5.9084, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 1.198230266571045, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 5.9093, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 3.2783617973327637, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 5.908, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.15221353071466623, | |
| "eval_loss": 7.133647918701172, | |
| "eval_runtime": 1.0107, | |
| "eval_samples_per_second": 473.908, | |
| "eval_steps_per_second": 4.947, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 0.8629825711250305, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 5.8967, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 2.963547945022583, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 5.9004, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 1.3798794746398926, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 5.8967, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 1.1503064632415771, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 5.8965, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 3.0068869590759277, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 5.8939, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 0.8236624002456665, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 5.8806, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 3.124436616897583, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 5.8859, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 0.9516368508338928, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 5.8871, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 2.733294725418091, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 5.8816, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 2.201298713684082, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 5.8786, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 0.8142681121826172, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 5.8737, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 1.012821078300476, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 5.8742, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 1.002347707748413, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 5.8697, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 1.8126763105392456, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 5.8708, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 0.8510775566101074, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 5.8665, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 0.9530029892921448, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 5.868, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 2.775869131088257, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 5.8636, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 0.875149667263031, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 5.8618, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 2.2910497188568115, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 5.8614, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 1.5151126801503744e+18, | |
| "train_loss": 6.377581384820561, | |
| "train_runtime": 24690.555, | |
| "train_samples_per_second": 231.505, | |
| "train_steps_per_second": 2.412 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5151126801503744e+18, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |