| { | |
| "best_metric": 0.9985501993475897, | |
| "best_model_checkpoint": "Smart_Tour_Luxor_v1.0\\checkpoint-194", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 776, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.1416014432907104, | |
| "learning_rate": 6.41025641025641e-06, | |
| "loss": 2.9726, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0776972770690918, | |
| "learning_rate": 1.282051282051282e-05, | |
| "loss": 2.9053, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0615094900131226, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 2.7576, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.0457894802093506, | |
| "learning_rate": 2.564102564102564e-05, | |
| "loss": 2.5866, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.1404309272766113, | |
| "learning_rate": 3.205128205128206e-05, | |
| "loss": 2.3687, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.1770880222320557, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 2.1344, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.2791804075241089, | |
| "learning_rate": 4.4871794871794874e-05, | |
| "loss": 1.8869, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.2911182641983032, | |
| "learning_rate": 4.98567335243553e-05, | |
| "loss": 1.5984, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.3716331720352173, | |
| "learning_rate": 4.914040114613181e-05, | |
| "loss": 1.3587, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.3185632228851318, | |
| "learning_rate": 4.842406876790831e-05, | |
| "loss": 1.1518, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.4818017482757568, | |
| "learning_rate": 4.7707736389684815e-05, | |
| "loss": 0.9694, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.235606074333191, | |
| "learning_rate": 4.699140401146132e-05, | |
| "loss": 0.8449, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 1.0605517625808716, | |
| "learning_rate": 4.627507163323783e-05, | |
| "loss": 0.7504, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.0399278402328491, | |
| "learning_rate": 4.555873925501433e-05, | |
| "loss": 0.6532, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.2140040397644043, | |
| "learning_rate": 4.4842406876790833e-05, | |
| "loss": 0.5724, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.8598456382751465, | |
| "learning_rate": 4.412607449856734e-05, | |
| "loss": 0.5135, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.7887383103370667, | |
| "learning_rate": 4.3409742120343846e-05, | |
| "loss": 0.4758, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.7831109762191772, | |
| "learning_rate": 4.2693409742120346e-05, | |
| "loss": 0.4344, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.6504438519477844, | |
| "learning_rate": 4.197707736389685e-05, | |
| "loss": 0.4048, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9985501993475897, | |
| "eval_loss": 0.3498220443725586, | |
| "eval_runtime": 330.0181, | |
| "eval_samples_per_second": 8.36, | |
| "eval_steps_per_second": 0.264, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.8069166541099548, | |
| "learning_rate": 4.126074498567336e-05, | |
| "loss": 0.3732, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.8006480932235718, | |
| "learning_rate": 4.054441260744986e-05, | |
| "loss": 0.3454, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.695347011089325, | |
| "learning_rate": 3.982808022922636e-05, | |
| "loss": 0.3199, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.5613417029380798, | |
| "learning_rate": 3.9111747851002864e-05, | |
| "loss": 0.2974, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.48088908195495605, | |
| "learning_rate": 3.839541547277937e-05, | |
| "loss": 0.3018, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.43887242674827576, | |
| "learning_rate": 3.767908309455588e-05, | |
| "loss": 0.2825, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.630588173866272, | |
| "learning_rate": 3.6962750716332376e-05, | |
| "loss": 0.2695, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.4198044538497925, | |
| "learning_rate": 3.624641833810888e-05, | |
| "loss": 0.2483, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.9144806265830994, | |
| "learning_rate": 3.553008595988539e-05, | |
| "loss": 0.2452, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.8064547181129456, | |
| "learning_rate": 3.4813753581661896e-05, | |
| "loss": 0.2453, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.6866409182548523, | |
| "learning_rate": 3.4097421203438395e-05, | |
| "loss": 0.2348, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5553216934204102, | |
| "learning_rate": 3.33810888252149e-05, | |
| "loss": 0.2193, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.4838345944881439, | |
| "learning_rate": 3.266475644699141e-05, | |
| "loss": 0.2216, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.6289733648300171, | |
| "learning_rate": 3.1948424068767914e-05, | |
| "loss": 0.202, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.7053207159042358, | |
| "learning_rate": 3.1232091690544414e-05, | |
| "loss": 0.1981, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.30541422963142395, | |
| "learning_rate": 3.0515759312320917e-05, | |
| "loss": 0.202, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.3764614760875702, | |
| "learning_rate": 2.9799426934097423e-05, | |
| "loss": 0.191, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.3376723825931549, | |
| "learning_rate": 2.9083094555873923e-05, | |
| "loss": 0.1861, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.33649709820747375, | |
| "learning_rate": 2.836676217765043e-05, | |
| "loss": 0.1711, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9978252990213845, | |
| "eval_loss": 0.16818779706954956, | |
| "eval_runtime": 316.359, | |
| "eval_samples_per_second": 8.721, | |
| "eval_steps_per_second": 0.275, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.4789816737174988, | |
| "learning_rate": 2.7650429799426936e-05, | |
| "loss": 0.1692, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 0.37141987681388855, | |
| "learning_rate": 2.6934097421203442e-05, | |
| "loss": 0.168, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 0.5811685919761658, | |
| "learning_rate": 2.6217765042979942e-05, | |
| "loss": 0.1614, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.5884858965873718, | |
| "learning_rate": 2.5501432664756448e-05, | |
| "loss": 0.169, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.7159774899482727, | |
| "learning_rate": 2.4785100286532954e-05, | |
| "loss": 0.1584, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.32133427262306213, | |
| "learning_rate": 2.4068767908309457e-05, | |
| "loss": 0.1605, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 1.0729812383651733, | |
| "learning_rate": 2.335243553008596e-05, | |
| "loss": 0.1539, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 0.3316652178764343, | |
| "learning_rate": 2.2636103151862463e-05, | |
| "loss": 0.1433, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 0.6638622879981995, | |
| "learning_rate": 2.191977077363897e-05, | |
| "loss": 0.1494, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 0.7581809163093567, | |
| "learning_rate": 2.1203438395415473e-05, | |
| "loss": 0.1536, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 0.2173503339290619, | |
| "learning_rate": 2.048710601719198e-05, | |
| "loss": 0.1483, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.8919398188591003, | |
| "learning_rate": 1.9770773638968482e-05, | |
| "loss": 0.1439, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 0.8781760334968567, | |
| "learning_rate": 1.905444126074499e-05, | |
| "loss": 0.1364, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.26618167757987976, | |
| "learning_rate": 1.833810888252149e-05, | |
| "loss": 0.1307, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 0.31138116121292114, | |
| "learning_rate": 1.7621776504297995e-05, | |
| "loss": 0.1308, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.8110724091529846, | |
| "learning_rate": 1.6905444126074498e-05, | |
| "loss": 0.1405, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.2559654712677002, | |
| "learning_rate": 1.6189111747851004e-05, | |
| "loss": 0.1315, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 0.2344350814819336, | |
| "learning_rate": 1.5472779369627507e-05, | |
| "loss": 0.1298, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.298566073179245, | |
| "learning_rate": 1.4756446991404013e-05, | |
| "loss": 0.1316, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 0.25082945823669434, | |
| "learning_rate": 1.4040114613180516e-05, | |
| "loss": 0.1307, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9985501993475897, | |
| "eval_loss": 0.11476617306470871, | |
| "eval_runtime": 376.2622, | |
| "eval_samples_per_second": 7.333, | |
| "eval_steps_per_second": 0.231, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.21889334917068481, | |
| "learning_rate": 1.3323782234957023e-05, | |
| "loss": 0.1195, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 0.3546507954597473, | |
| "learning_rate": 1.2607449856733524e-05, | |
| "loss": 0.1333, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 0.2458142638206482, | |
| "learning_rate": 1.1891117478510029e-05, | |
| "loss": 0.1176, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.45963388681411743, | |
| "learning_rate": 1.1174785100286533e-05, | |
| "loss": 0.1342, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.2312440127134323, | |
| "learning_rate": 1.0458452722063038e-05, | |
| "loss": 0.1207, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.21415020525455475, | |
| "learning_rate": 9.742120343839543e-06, | |
| "loss": 0.1294, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.1821221262216568, | |
| "learning_rate": 9.025787965616046e-06, | |
| "loss": 0.1262, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.18438111245632172, | |
| "learning_rate": 8.30945558739255e-06, | |
| "loss": 0.108, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.2131635844707489, | |
| "learning_rate": 7.593123209169055e-06, | |
| "loss": 0.1138, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 1.0909886360168457, | |
| "learning_rate": 6.876790830945559e-06, | |
| "loss": 0.1179, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 0.6753919124603271, | |
| "learning_rate": 6.160458452722064e-06, | |
| "loss": 0.1118, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 0.20348551869392395, | |
| "learning_rate": 5.4441260744985674e-06, | |
| "loss": 0.1019, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 0.22448669373989105, | |
| "learning_rate": 4.727793696275072e-06, | |
| "loss": 0.1093, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 0.6886144280433655, | |
| "learning_rate": 4.011461318051577e-06, | |
| "loss": 0.1067, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.19005030393600464, | |
| "learning_rate": 3.2951289398280802e-06, | |
| "loss": 0.1237, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 0.2179213911294937, | |
| "learning_rate": 2.578796561604585e-06, | |
| "loss": 0.1187, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 0.47610709071159363, | |
| "learning_rate": 1.862464183381089e-06, | |
| "loss": 0.1068, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 0.34112074971199036, | |
| "learning_rate": 1.1461318051575932e-06, | |
| "loss": 0.1204, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 0.24934335052967072, | |
| "learning_rate": 4.2979942693409743e-07, | |
| "loss": 0.115, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9985501993475897, | |
| "eval_loss": 0.10186277329921722, | |
| "eval_runtime": 304.0711, | |
| "eval_samples_per_second": 9.074, | |
| "eval_steps_per_second": 0.286, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 776, | |
| "total_flos": 7.69643703046162e+18, | |
| "train_loss": 0.4810273270938814, | |
| "train_runtime": 53143.3355, | |
| "train_samples_per_second": 1.869, | |
| "train_steps_per_second": 0.015 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 776, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 7.69643703046162e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |