| { | |
| "best_global_step": 10000, | |
| "best_metric": 10.123960494995117, | |
| "best_model_checkpoint": "distilbert_rand_50_v1/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": 2.6359527111053467, | |
| "learning_rate": 5e-06, | |
| "loss": 13.2342, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 1.8972421884536743, | |
| "learning_rate": 1e-05, | |
| "loss": 11.5226, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 2.0409884452819824, | |
| "learning_rate": 1.5e-05, | |
| "loss": 10.9869, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 2.3096764087677, | |
| "learning_rate": 2e-05, | |
| "loss": 10.8348, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 2.2824838161468506, | |
| "learning_rate": 2.5e-05, | |
| "loss": 10.703, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 2.202655076980591, | |
| "learning_rate": 3e-05, | |
| "loss": 10.5867, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 2.6729981899261475, | |
| "learning_rate": 3.5e-05, | |
| "loss": 10.4911, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 2.458367109298706, | |
| "learning_rate": 4e-05, | |
| "loss": 10.4144, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 2.0687592029571533, | |
| "learning_rate": 4.5e-05, | |
| "loss": 10.3571, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 2.413729667663574, | |
| "learning_rate": 5e-05, | |
| "loss": 10.308, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 2.6078577041625977, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 10.2672, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 2.3658957481384277, | |
| "learning_rate": 6e-05, | |
| "loss": 10.2337, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 2.451045274734497, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 10.2053, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 3.027066946029663, | |
| "learning_rate": 7e-05, | |
| "loss": 10.1799, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 2.489773988723755, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 10.1206, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.913222551345825, | |
| "learning_rate": 8e-05, | |
| "loss": 10.0885, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 2.8502819538116455, | |
| "learning_rate": 8.5e-05, | |
| "loss": 10.0796, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 3.0633347034454346, | |
| "learning_rate": 9e-05, | |
| "loss": 10.0606, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 2.7516486644744873, | |
| "learning_rate": 9.5e-05, | |
| "loss": 10.0502, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 3.562567949295044, | |
| "learning_rate": 0.0001, | |
| "loss": 9.8834, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.1514510698683881, | |
| "eval_loss": 10.123960494995117, | |
| "eval_runtime": 0.8335, | |
| "eval_samples_per_second": 574.701, | |
| "eval_steps_per_second": 5.999, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 3.0441133975982666, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 9.8913, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 3.58797025680542, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 9.8939, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 3.082925319671631, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 9.9055, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 3.3463780879974365, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 9.8539, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 3.3159289360046387, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 9.6777, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 3.5410890579223633, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 9.705, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 3.3903536796569824, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 9.7322, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 3.3420181274414062, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 9.7421, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 3.429137706756592, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 9.6375, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 3.4593329429626465, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 9.5219, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 3.396314859390259, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 9.5556, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 3.636416435241699, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 9.5809, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 3.90594482421875, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 9.6022, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 3.833756446838379, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 9.3931, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 4.425719738006592, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 9.3098, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 4.336478233337402, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 9.3441, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 4.405770778656006, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 9.3796, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 4.051723480224609, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 9.4001, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 4.795436382293701, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 9.0342, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 4.936302661895752, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 9.0198, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15208058587502454, | |
| "eval_loss": 10.622453689575195, | |
| "eval_runtime": 0.8186, | |
| "eval_samples_per_second": 585.178, | |
| "eval_steps_per_second": 6.108, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 5.398108005523682, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 9.0459, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 5.233779430389404, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 9.0872, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 6.226819038391113, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 9.0267, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 5.534121513366699, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 8.6149, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 6.475548267364502, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 8.6501, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 6.23988151550293, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 8.695, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 6.956918239593506, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 8.7107, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 7.125720977783203, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 8.5501, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 6.763567924499512, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 8.2583, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 7.2223944664001465, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 8.2971, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 7.531473159790039, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 8.3315, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 7.295470714569092, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 8.366, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 7.148488521575928, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 8.0824, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 8.001736640930176, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 7.9366, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 6.924074649810791, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 7.974, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 8.327032089233398, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 8.0056, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 8.413107872009277, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 8.044, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 7.6440958976745605, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 7.6487, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 9.179951667785645, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 7.6022, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 8.806946754455566, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 7.6604, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15381817164283515, | |
| "eval_loss": 11.978323936462402, | |
| "eval_runtime": 0.8118, | |
| "eval_samples_per_second": 590.019, | |
| "eval_steps_per_second": 6.159, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 8.8045654296875, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 7.7048, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 8.67996597290039, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 7.6927, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 9.088833808898926, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 7.2767, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 9.68818473815918, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 7.3328, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 8.956292152404785, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 7.3766, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 9.119683265686035, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 7.389, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 7.701849460601807, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 7.28, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 9.369430541992188, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 7.0168, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 9.682147026062012, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 7.0663, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 8.92928695678711, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 7.0967, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 9.656011581420898, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 7.117, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 10.814836502075195, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 6.9506, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 9.774897575378418, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.8193, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 8.79526424407959, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.8664, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 10.250818252563477, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 6.8888, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 12.649961471557617, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 6.9039, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 10.088028907775879, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 6.701, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 11.095525741577148, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 6.6809, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 7.743183612823486, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 6.6966, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 10.943224906921387, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 6.7176, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15117191880677358, | |
| "eval_loss": 13.387857437133789, | |
| "eval_runtime": 0.8112, | |
| "eval_samples_per_second": 590.462, | |
| "eval_steps_per_second": 6.163, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 6.660904884338379, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 6.7364, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 7.744833946228027, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 6.5556, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 10.628016471862793, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 6.5654, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 9.486727714538574, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 6.5833, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 9.917757987976074, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 6.6049, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 8.177474021911621, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 6.5781, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 9.21579647064209, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 6.4823, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 7.01672887802124, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 6.4921, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 10.70284366607666, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 6.5075, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 8.969956398010254, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 6.5238, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 7.561686038970947, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 6.4638, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 8.126607894897461, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 6.43, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 9.955617904663086, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 6.429, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 9.902905464172363, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 6.4332, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 7.163830280303955, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 6.4407, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 8.653093338012695, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 6.4064, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 8.98547077178955, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 6.386, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 6.314413070678711, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 6.3839, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 9.231555938720703, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 6.3824, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 5.8110737800598145, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 6.3921, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1517649367762918, | |
| "eval_loss": 14.1333646774292, | |
| "eval_runtime": 0.8141, | |
| "eval_samples_per_second": 588.369, | |
| "eval_steps_per_second": 6.142, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 8.034831047058105, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 6.343, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 8.232477188110352, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 6.3517, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 8.41584587097168, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 6.3447, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 7.5299763679504395, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 6.3436, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 6.112133979797363, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 6.3425, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 5.476301670074463, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 6.3068, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 9.581059455871582, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 6.3135, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 8.758219718933105, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 6.319, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 8.004023551940918, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 6.3058, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 6.037941932678223, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 6.3003, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 7.76659631729126, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.2793, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 8.344924926757812, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.2935, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 8.129399299621582, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.2754, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 8.312440872192383, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 6.2841, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 5.360718727111816, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.2682, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 6.858188152313232, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.272, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 6.355225563049316, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.271, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 6.370845317840576, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.2594, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 6.191452980041504, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.2535, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 7.68761901614592e+17, | |
| "train_loss": 8.061118055418897, | |
| "train_runtime": 16926.9274, | |
| "train_samples_per_second": 337.685, | |
| "train_steps_per_second": 3.518 | |
| } | |
| ], | |
| "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": 7.68761901614592e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |