| { | |
| "best_metric": 5.598287582397461, | |
| "best_model_checkpoint": "tiny_bert_km_5_v2/checkpoint-50000", | |
| "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": 3.0139718055725098, | |
| "learning_rate": 5e-06, | |
| "loss": 11.3627, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 2.0012552738189697, | |
| "learning_rate": 1e-05, | |
| "loss": 9.6678, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 5.729489326477051, | |
| "learning_rate": 1.5e-05, | |
| "loss": 8.2767, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 4.382988452911377, | |
| "learning_rate": 2e-05, | |
| "loss": 7.6313, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 4.852724075317383, | |
| "learning_rate": 2.5e-05, | |
| "loss": 7.3777, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 3.679936647415161, | |
| "learning_rate": 3e-05, | |
| "loss": 7.2139, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 2.925384283065796, | |
| "learning_rate": 3.5e-05, | |
| "loss": 7.1037, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 2.3004095554351807, | |
| "learning_rate": 4e-05, | |
| "loss": 7.0136, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 2.5559029579162598, | |
| "learning_rate": 4.5e-05, | |
| "loss": 6.9261, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 2.738003730773926, | |
| "learning_rate": 5e-05, | |
| "loss": 6.8278, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 2.964071035385132, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 6.7575, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 1.9598089456558228, | |
| "learning_rate": 6e-05, | |
| "loss": 6.7182, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 1.9295579195022583, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 6.6741, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 2.3092055320739746, | |
| "learning_rate": 7e-05, | |
| "loss": 6.6438, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 3.3601136207580566, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 6.5719, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 1.7021485567092896, | |
| "learning_rate": 8e-05, | |
| "loss": 6.5494, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 1.4094020128250122, | |
| "learning_rate": 8.5e-05, | |
| "loss": 6.5401, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 2.275599241256714, | |
| "learning_rate": 9e-05, | |
| "loss": 6.516, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 2.0243468284606934, | |
| "learning_rate": 9.5e-05, | |
| "loss": 6.4951, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 1.5493866205215454, | |
| "learning_rate": 0.0001, | |
| "loss": 6.435, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.14886335179053395, | |
| "eval_loss": 6.381496429443359, | |
| "eval_runtime": 0.7333, | |
| "eval_samples_per_second": 653.216, | |
| "eval_steps_per_second": 6.819, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 1.515457034111023, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 6.4271, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 3.5068676471710205, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 6.416, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 2.1910295486450195, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 6.4121, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 2.603455066680908, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 6.3899, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 1.6394344568252563, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 6.3414, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 1.9101860523223877, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 6.3353, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 2.6885523796081543, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 6.3369, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 1.3316004276275635, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 6.3213, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.1802358627319336, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 6.3005, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 1.990135908126831, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 6.2748, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 1.1196672916412354, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 6.2672, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 1.4957544803619385, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 6.261, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 0.9325321316719055, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 6.2557, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 1.4595731496810913, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 6.2272, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 0.8780417442321777, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 6.2107, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 1.325482726097107, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 6.2142, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 1.4712358713150024, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 6.2082, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 1.402755856513977, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 6.2109, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 1.0420176982879639, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 6.175, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 2.0893967151641846, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 6.1783, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15323101097112712, | |
| "eval_loss": 6.305487155914307, | |
| "eval_runtime": 0.6853, | |
| "eval_samples_per_second": 698.993, | |
| "eval_steps_per_second": 7.296, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 2.9806649684906006, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 6.1695, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 1.1048426628112793, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 6.1706, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 1.0069621801376343, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 6.1619, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 1.1569310426712036, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 6.1402, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 1.9851467609405518, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 6.1361, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 1.2456252574920654, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 6.1329, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 1.4847053289413452, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 6.1305, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 1.0716360807418823, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 6.1173, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 1.3109575510025024, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 6.1109, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 1.7119407653808594, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 6.1038, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 1.7951136827468872, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 6.1013, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 1.3359527587890625, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 6.1053, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 2.057919979095459, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 6.0847, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 1.6525475978851318, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 6.0789, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 0.8150427937507629, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 6.0746, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 2.0109965801239014, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 6.0698, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 2.6631064414978027, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 6.0705, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 1.6978521347045898, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 6.0559, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 0.768791139125824, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 6.0483, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 0.981519341468811, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 6.0524, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15465763774239583, | |
| "eval_loss": 6.281867027282715, | |
| "eval_runtime": 0.7174, | |
| "eval_samples_per_second": 667.704, | |
| "eval_steps_per_second": 6.97, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 2.2384562492370605, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 6.048, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 0.7136291861534119, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 6.0477, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 0.9357853531837463, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 6.0348, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 0.7753567695617676, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 6.0282, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 1.545584797859192, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 6.0294, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 1.2191523313522339, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 6.0265, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 1.6021625995635986, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 6.0062, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 0.7979269623756409, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 5.9809, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 1.283188819885254, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 5.9574, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 0.820725679397583, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 5.9365, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 1.4742634296417236, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 5.9152, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 2.1960642337799072, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 5.8915, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 2.1507370471954346, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 5.8741, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 1.5441607236862183, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 5.8519, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 2.054232358932495, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 5.8375, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 2.694913148880005, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 5.8212, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 1.1486177444458008, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 5.8038, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 2.5899977684020996, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 5.7805, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 1.3233723640441895, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 5.7692, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 1.563281536102295, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 5.7467, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.18153526970954356, | |
| "eval_loss": 5.912224292755127, | |
| "eval_runtime": 0.7119, | |
| "eval_samples_per_second": 672.807, | |
| "eval_steps_per_second": 7.023, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 1.3055953979492188, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 5.7354, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 1.5275450944900513, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 5.7105, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 1.7186174392700195, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 5.6924, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 1.5413143634796143, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 5.6785, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 2.001120090484619, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 5.6642, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 2.4105334281921387, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 5.6442, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 2.5755810737609863, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 5.6233, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 1.599410057067871, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 5.6078, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 2.6714165210723877, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 5.5937, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 2.9554543495178223, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 5.5844, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 1.761340618133545, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 5.5632, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 1.5560534000396729, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 5.5421, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 3.103821277618408, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 5.5294, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 1.7085633277893066, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 5.5164, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 2.625190019607544, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 5.4986, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 1.636446237564087, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 5.4841, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 2.6742091178894043, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 5.4702, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 1.8936998844146729, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 5.4526, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 2.109135150909424, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 5.4402, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 1.9321963787078857, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 5.4323, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.22502593433706228, | |
| "eval_loss": 5.598287582397461, | |
| "eval_runtime": 0.6965, | |
| "eval_samples_per_second": 687.705, | |
| "eval_steps_per_second": 7.179, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 3.1609930992126465, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 5.4138, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 3.3288419246673584, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 5.4071, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 2.991629123687744, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 5.3932, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 2.063535213470459, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 5.3818, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 2.3984062671661377, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 5.3729, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 3.5371387004852295, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 5.3548, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 2.4277217388153076, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 5.3483, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 2.5748581886291504, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 5.3455, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 4.068253040313721, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 5.3339, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 2.623516321182251, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 5.3232, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 2.2249653339385986, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 5.3166, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 2.3571181297302246, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 5.3114, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 2.788194179534912, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 5.3017, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 2.578573226928711, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 5.3007, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 2.3986618518829346, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 5.2908, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 2.319240093231201, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 5.2927, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 4.31321907043457, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 5.2833, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 2.2810864448547363, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 5.2811, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 2.1196603775024414, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 5.2797, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 3.049796905526016e+17, | |
| "train_loss": 6.086415274593632, | |
| "train_runtime": 11636.5101, | |
| "train_samples_per_second": 491.21, | |
| "train_steps_per_second": 5.118 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "total_flos": 3.049796905526016e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |