| { | |
| "best_metric": 8.520008087158203, | |
| "best_model_checkpoint": "bert_base_lda_50_v1/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": 2.1146457195281982, | |
| "learning_rate": 5e-06, | |
| "loss": 13.0605, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 1.4042017459869385, | |
| "learning_rate": 1e-05, | |
| "loss": 11.3018, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 1.0155179500579834, | |
| "learning_rate": 1.5e-05, | |
| "loss": 10.7875, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 1.093125820159912, | |
| "learning_rate": 2e-05, | |
| "loss": 10.6485, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 1.1090766191482544, | |
| "learning_rate": 2.5e-05, | |
| "loss": 10.5419, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 1.021306037902832, | |
| "learning_rate": 3e-05, | |
| "loss": 10.4448, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 1.1271008253097534, | |
| "learning_rate": 3.5e-05, | |
| "loss": 10.3688, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 0.9003839492797852, | |
| "learning_rate": 4e-05, | |
| "loss": 10.3088, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 1.0701918601989746, | |
| "learning_rate": 4.5e-05, | |
| "loss": 10.2561, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 0.986882746219635, | |
| "learning_rate": 5e-05, | |
| "loss": 10.2069, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 1.1010342836380005, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 10.1661, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 0.9132248163223267, | |
| "learning_rate": 6e-05, | |
| "loss": 10.1313, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 1.0529310703277588, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 10.1, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 0.9321224093437195, | |
| "learning_rate": 7e-05, | |
| "loss": 10.0725, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 0.9696210026741028, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 10.045, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 0.9034942984580994, | |
| "learning_rate": 8e-05, | |
| "loss": 10.0231, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 0.994525671005249, | |
| "learning_rate": 8.5e-05, | |
| "loss": 9.9995, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 1.0354859828948975, | |
| "learning_rate": 9e-05, | |
| "loss": 9.9018, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 1.1252986192703247, | |
| "learning_rate": 9.5e-05, | |
| "loss": 9.7352, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 1.1572073698043823, | |
| "learning_rate": 0.0001, | |
| "loss": 9.5406, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15058849717577005, | |
| "eval_loss": 9.414867401123047, | |
| "eval_runtime": 0.72, | |
| "eval_samples_per_second": 665.324, | |
| "eval_steps_per_second": 6.945, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 1.2121734619140625, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 9.4392, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 1.1941754817962646, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 9.3736, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 1.1061564683914185, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 9.3414, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 1.1628280878067017, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 9.3043, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 1.1327245235443115, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 9.2632, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 1.0951533317565918, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 9.245, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 1.1437737941741943, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 9.2357, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 1.1072016954421997, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 9.2175, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 1.1195781230926514, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 9.1963, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 1.1077725887298584, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 9.1777, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 0.9862968325614929, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 9.1665, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 1.0996145009994507, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 9.1586, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 1.0800743103027344, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 9.1482, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 1.083832025527954, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 9.1272, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 1.0020357370376587, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 9.113, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 1.0792287588119507, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 9.1143, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 1.0629321336746216, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 9.1069, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 1.0538816452026367, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 9.1053, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 1.0253933668136597, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 9.0799, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 1.0982084274291992, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 9.0789, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.1531187743763854, | |
| "eval_loss": 9.082218170166016, | |
| "eval_runtime": 0.7061, | |
| "eval_samples_per_second": 678.367, | |
| "eval_steps_per_second": 7.081, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 1.0530356168746948, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 9.0712, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 0.9998981356620789, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 9.0699, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 1.02177894115448, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 9.0608, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 0.9790524840354919, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 9.0443, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 0.9412717819213867, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 9.0412, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 0.969178318977356, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 9.0392, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 1.0142080783843994, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 9.0336, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 1.0047160387039185, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 9.0196, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 0.971329927444458, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 9.0178, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 0.9855359792709351, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 9.0087, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 0.9914914965629578, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 9.0079, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 0.9419001936912537, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 9.0098, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 0.967641294002533, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 8.9945, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 0.9290738701820374, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 8.9901, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 0.9592576026916504, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 8.984, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 0.9512010216712952, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 8.9806, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 1.004713773727417, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 8.9795, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 0.9433563947677612, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 8.9706, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 0.8946517705917358, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 8.9613, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 0.9454537630081177, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 8.9645, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15524526401208832, | |
| "eval_loss": 9.011049270629883, | |
| "eval_runtime": 0.7111, | |
| "eval_samples_per_second": 673.563, | |
| "eval_steps_per_second": 7.031, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 0.9562968015670776, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 8.9614, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 0.9172261953353882, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 8.9626, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 0.9505202770233154, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 8.9463, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 0.9160086512565613, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 8.9465, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 0.9293331503868103, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 8.9492, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 0.9969606995582581, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 8.9445, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 0.9666557312011719, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 8.9383, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 0.9124544262886047, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 8.9294, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 0.9125744700431824, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 8.9282, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 0.9087144732475281, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 8.9282, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 0.9195201992988586, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 8.9294, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 0.9335217475891113, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 8.9178, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 0.9436230063438416, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 8.9156, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 0.9273329377174377, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 8.9132, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 0.9454130530357361, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 8.9134, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 0.9369338154792786, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 8.912, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 0.9072591662406921, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 8.9064, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 0.882411539554596, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 8.8973, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 0.9195645451545715, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 8.9006, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 0.9187344312667847, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 8.8927, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15327464393854434, | |
| "eval_loss": 8.982623100280762, | |
| "eval_runtime": 0.7115, | |
| "eval_samples_per_second": 673.259, | |
| "eval_steps_per_second": 7.028, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 0.9306628108024597, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 8.8974, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 0.9240749478340149, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 8.8856, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 0.9050202965736389, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 8.883, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 0.8883821368217468, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 8.886, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 0.8940898776054382, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 8.8842, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 0.9101306796073914, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 8.8786, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 0.9014184474945068, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 8.8713, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 0.9065138101577759, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 8.8709, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 0.9120799899101257, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 8.8711, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 0.9068942070007324, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 8.8734, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 0.9323383569717407, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 8.8684, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 0.9209831953048706, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 8.8556, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 0.904669463634491, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 8.8581, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 0.8882988095283508, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 8.8606, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 0.9138925671577454, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 8.8538, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 0.947345495223999, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 8.8409, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 0.9615164399147034, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 8.8114, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 1.0933176279067993, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 8.766, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 1.1006487607955933, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 8.6504, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 1.1190329790115356, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 8.5333, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.20458687301987832, | |
| "eval_loss": 8.520008087158203, | |
| "eval_runtime": 0.7146, | |
| "eval_samples_per_second": 670.296, | |
| "eval_steps_per_second": 6.997, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 1.1501587629318237, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 8.412, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 1.2387570142745972, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 8.3158, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 1.2276296615600586, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 8.2387, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 1.2896901369094849, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 8.1698, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 1.2785300016403198, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 8.1036, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 1.3099133968353271, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 8.0376, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 1.399721622467041, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 7.9941, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 1.377581000328064, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 7.9568, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 1.4110668897628784, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 7.9213, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 1.4229886531829834, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 7.889, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 1.3900176286697388, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 7.8624, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 1.3957171440124512, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 7.8406, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 1.3756403923034668, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 7.8172, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 1.4114515781402588, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 7.8027, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 1.4109306335449219, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 7.789, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 1.410210132598877, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 7.7803, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 1.4097881317138672, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 7.766, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 1.3454077243804932, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 7.7601, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 1.3815090656280518, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 7.7568, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 1.5155177772432384e+18, | |
| "train_loss": 9.060488100348152, | |
| "train_runtime": 17138.1091, | |
| "train_samples_per_second": 333.524, | |
| "train_steps_per_second": 3.475 | |
| } | |
| ], | |
| "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": 1.5155177772432384e+18, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |