| { | |
| "best_metric": 0.0266929492354393, | |
| "best_model_checkpoint": "deberta-v3-base-zyda-2-transformed-readability/checkpoint-40767", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 40767, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03679446611229671, | |
| "grad_norm": 2.2327511310577393, | |
| "learning_rate": 4.9386758898128385e-05, | |
| "loss": 0.0956, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.07358893222459342, | |
| "grad_norm": 0.9045169353485107, | |
| "learning_rate": 4.877351779625678e-05, | |
| "loss": 0.0524, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.11038339833689013, | |
| "grad_norm": 0.6377315521240234, | |
| "learning_rate": 4.8160276694385164e-05, | |
| "loss": 0.0439, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.14717786444918685, | |
| "grad_norm": 1.2090165615081787, | |
| "learning_rate": 4.754703559251355e-05, | |
| "loss": 0.0395, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.18397233056148354, | |
| "grad_norm": 1.0190351009368896, | |
| "learning_rate": 4.693379449064194e-05, | |
| "loss": 0.0387, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.22076679667378027, | |
| "grad_norm": 0.6785000562667847, | |
| "learning_rate": 4.632055338877033e-05, | |
| "loss": 0.0364, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.257561262786077, | |
| "grad_norm": 0.37076541781425476, | |
| "learning_rate": 4.570731228689872e-05, | |
| "loss": 0.0427, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.2943557288983737, | |
| "grad_norm": 0.2937301695346832, | |
| "learning_rate": 4.509407118502711e-05, | |
| "loss": 0.0359, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.3311501950106704, | |
| "grad_norm": 0.3934372663497925, | |
| "learning_rate": 4.448083008315549e-05, | |
| "loss": 0.0354, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.3679446611229671, | |
| "grad_norm": 0.8720031380653381, | |
| "learning_rate": 4.386758898128389e-05, | |
| "loss": 0.0343, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.40473912723526384, | |
| "grad_norm": 0.2753826677799225, | |
| "learning_rate": 4.325434787941227e-05, | |
| "loss": 0.0333, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.44153359334756054, | |
| "grad_norm": 0.9069143533706665, | |
| "learning_rate": 4.264110677754066e-05, | |
| "loss": 0.0319, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.47832805945985724, | |
| "grad_norm": 1.0423845052719116, | |
| "learning_rate": 4.202786567566905e-05, | |
| "loss": 0.0319, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.515122525572154, | |
| "grad_norm": 0.5137051939964294, | |
| "learning_rate": 4.141462457379743e-05, | |
| "loss": 0.0321, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.5519169916844506, | |
| "grad_norm": 0.34184473752975464, | |
| "learning_rate": 4.080138347192582e-05, | |
| "loss": 0.0316, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.5887114577967474, | |
| "grad_norm": 0.5334771275520325, | |
| "learning_rate": 4.018814237005421e-05, | |
| "loss": 0.0318, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.625505923909044, | |
| "grad_norm": 0.27346959710121155, | |
| "learning_rate": 3.95749012681826e-05, | |
| "loss": 0.0314, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.6623003900213408, | |
| "grad_norm": 0.46926313638687134, | |
| "learning_rate": 3.896166016631099e-05, | |
| "loss": 0.0308, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.6990948561336375, | |
| "grad_norm": 0.661072850227356, | |
| "learning_rate": 3.834841906443938e-05, | |
| "loss": 0.0309, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.7358893222459342, | |
| "grad_norm": 0.3445192575454712, | |
| "learning_rate": 3.773517796256776e-05, | |
| "loss": 0.03, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.7726837883582309, | |
| "grad_norm": 0.5244751572608948, | |
| "learning_rate": 3.712193686069616e-05, | |
| "loss": 0.0299, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.8094782544705277, | |
| "grad_norm": 0.401460200548172, | |
| "learning_rate": 3.650869575882454e-05, | |
| "loss": 0.0294, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.8462727205828243, | |
| "grad_norm": 0.23478317260742188, | |
| "learning_rate": 3.589545465695293e-05, | |
| "loss": 0.029, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.8830671866951211, | |
| "grad_norm": 0.4309717118740082, | |
| "learning_rate": 3.528221355508132e-05, | |
| "loss": 0.029, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.9198616528074177, | |
| "grad_norm": 0.3477807641029358, | |
| "learning_rate": 3.466897245320971e-05, | |
| "loss": 0.0295, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.9566561189197145, | |
| "grad_norm": 0.21652667224407196, | |
| "learning_rate": 3.405573135133809e-05, | |
| "loss": 0.0293, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.9934505850320112, | |
| "grad_norm": 0.46980977058410645, | |
| "learning_rate": 3.344249024946648e-05, | |
| "loss": 0.0288, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.028563737869262695, | |
| "eval_mse": 0.028563737035006864, | |
| "eval_runtime": 98.1597, | |
| "eval_samples_per_second": 509.374, | |
| "eval_steps_per_second": 63.672, | |
| "step": 13589 | |
| }, | |
| { | |
| "epoch": 1.030245051144308, | |
| "grad_norm": 0.32912561297416687, | |
| "learning_rate": 3.282924914759487e-05, | |
| "loss": 0.0258, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.0670395172566045, | |
| "grad_norm": 0.2001865804195404, | |
| "learning_rate": 3.221600804572326e-05, | |
| "loss": 0.0251, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.1038339833689013, | |
| "grad_norm": 0.4719059467315674, | |
| "learning_rate": 3.160276694385165e-05, | |
| "loss": 0.0243, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.140628449481198, | |
| "grad_norm": 0.401038259267807, | |
| "learning_rate": 3.098952584198003e-05, | |
| "loss": 0.0246, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.1774229155934948, | |
| "grad_norm": 0.24117255210876465, | |
| "learning_rate": 3.0376284740108423e-05, | |
| "loss": 0.0248, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.2142173817057915, | |
| "grad_norm": 0.24041427671909332, | |
| "learning_rate": 2.976304363823681e-05, | |
| "loss": 0.025, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.2510118478180883, | |
| "grad_norm": 0.3739044666290283, | |
| "learning_rate": 2.91498025363652e-05, | |
| "loss": 0.0257, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.2878063139303848, | |
| "grad_norm": 0.4344153106212616, | |
| "learning_rate": 2.8536561434493587e-05, | |
| "loss": 0.0244, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.3246007800426816, | |
| "grad_norm": 0.6248531341552734, | |
| "learning_rate": 2.7923320332621977e-05, | |
| "loss": 0.0249, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.3613952461549783, | |
| "grad_norm": 0.34284424781799316, | |
| "learning_rate": 2.7310079230750363e-05, | |
| "loss": 0.0247, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.398189712267275, | |
| "grad_norm": 0.33926498889923096, | |
| "learning_rate": 2.6696838128878755e-05, | |
| "loss": 0.0248, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.4349841783795716, | |
| "grad_norm": 0.2008136361837387, | |
| "learning_rate": 2.6083597027007138e-05, | |
| "loss": 0.0248, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.4717786444918683, | |
| "grad_norm": 0.5362450480461121, | |
| "learning_rate": 2.5470355925135524e-05, | |
| "loss": 0.0246, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.508573110604165, | |
| "grad_norm": 0.4919290840625763, | |
| "learning_rate": 2.4857114823263916e-05, | |
| "loss": 0.0238, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.5453675767164619, | |
| "grad_norm": 0.3778747618198395, | |
| "learning_rate": 2.4243873721392306e-05, | |
| "loss": 0.0248, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.5821620428287586, | |
| "grad_norm": 0.2485371083021164, | |
| "learning_rate": 2.3630632619520692e-05, | |
| "loss": 0.0237, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.6189565089410554, | |
| "grad_norm": 0.2995116412639618, | |
| "learning_rate": 2.301739151764908e-05, | |
| "loss": 0.0232, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.6557509750533521, | |
| "grad_norm": 0.3775917887687683, | |
| "learning_rate": 2.2404150415777467e-05, | |
| "loss": 0.0239, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.6925454411656486, | |
| "grad_norm": 0.25416481494903564, | |
| "learning_rate": 2.1790909313905856e-05, | |
| "loss": 0.0242, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.7293399072779454, | |
| "grad_norm": 0.5196259617805481, | |
| "learning_rate": 2.1177668212034242e-05, | |
| "loss": 0.0238, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 1.7661343733902422, | |
| "grad_norm": 3.1327126026153564, | |
| "learning_rate": 2.056442711016263e-05, | |
| "loss": 0.0308, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.8029288395025387, | |
| "grad_norm": 1.1925427913665771, | |
| "learning_rate": 1.995118600829102e-05, | |
| "loss": 0.0268, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 1.8397233056148354, | |
| "grad_norm": 0.5257470011711121, | |
| "learning_rate": 1.933794490641941e-05, | |
| "loss": 0.0249, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.8765177717271322, | |
| "grad_norm": 0.4024732708930969, | |
| "learning_rate": 1.8724703804547796e-05, | |
| "loss": 0.0237, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 1.913312237839429, | |
| "grad_norm": 0.5063018798828125, | |
| "learning_rate": 1.8111462702676185e-05, | |
| "loss": 0.0231, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.9501067039517257, | |
| "grad_norm": 0.264139860868454, | |
| "learning_rate": 1.7498221600804575e-05, | |
| "loss": 0.0261, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 1.9869011700640224, | |
| "grad_norm": 0.19682620465755463, | |
| "learning_rate": 1.688498049893296e-05, | |
| "loss": 0.023, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.027232788503170013, | |
| "eval_mse": 0.02723278669797115, | |
| "eval_runtime": 77.4011, | |
| "eval_samples_per_second": 645.986, | |
| "eval_steps_per_second": 80.748, | |
| "step": 27178 | |
| }, | |
| { | |
| "epoch": 2.023695636176319, | |
| "grad_norm": 0.3318944275379181, | |
| "learning_rate": 1.627173939706135e-05, | |
| "loss": 0.0208, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.060490102288616, | |
| "grad_norm": 0.20372678339481354, | |
| "learning_rate": 1.565849829518974e-05, | |
| "loss": 0.0199, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.0972845684009127, | |
| "grad_norm": 0.4932423233985901, | |
| "learning_rate": 1.5045257193318127e-05, | |
| "loss": 0.0201, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 2.134079034513209, | |
| "grad_norm": 0.24097684025764465, | |
| "learning_rate": 1.4432016091446513e-05, | |
| "loss": 0.0199, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.1708735006255058, | |
| "grad_norm": 0.46340519189834595, | |
| "learning_rate": 1.38187749895749e-05, | |
| "loss": 0.0194, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 2.2076679667378025, | |
| "grad_norm": 0.17476551234722137, | |
| "learning_rate": 1.320553388770329e-05, | |
| "loss": 0.0199, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.2444624328500993, | |
| "grad_norm": 0.7477974891662598, | |
| "learning_rate": 1.2592292785831677e-05, | |
| "loss": 0.0202, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 2.281256898962396, | |
| "grad_norm": 0.21329531073570251, | |
| "learning_rate": 1.1979051683960066e-05, | |
| "loss": 0.0196, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.318051365074693, | |
| "grad_norm": 0.39124202728271484, | |
| "learning_rate": 1.1365810582088454e-05, | |
| "loss": 0.0198, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 2.3548458311869895, | |
| "grad_norm": 0.30534929037094116, | |
| "learning_rate": 1.0752569480216842e-05, | |
| "loss": 0.0196, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.3916402972992863, | |
| "grad_norm": 0.2803601324558258, | |
| "learning_rate": 1.0139328378345231e-05, | |
| "loss": 0.0197, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.428434763411583, | |
| "grad_norm": 0.30882009863853455, | |
| "learning_rate": 9.526087276473619e-06, | |
| "loss": 0.0194, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.46522922952388, | |
| "grad_norm": 0.310523122549057, | |
| "learning_rate": 8.912846174602008e-06, | |
| "loss": 0.0193, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 2.5020236956361765, | |
| "grad_norm": 0.3990231156349182, | |
| "learning_rate": 8.299605072730394e-06, | |
| "loss": 0.0198, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.5388181617484733, | |
| "grad_norm": 0.41601407527923584, | |
| "learning_rate": 7.686363970858783e-06, | |
| "loss": 0.0192, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 2.5756126278607696, | |
| "grad_norm": 0.2621209919452667, | |
| "learning_rate": 7.073122868987171e-06, | |
| "loss": 0.0198, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.6124070939730664, | |
| "grad_norm": 0.231684148311615, | |
| "learning_rate": 6.459881767115559e-06, | |
| "loss": 0.0192, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 2.649201560085363, | |
| "grad_norm": 0.23939248919487, | |
| "learning_rate": 5.846640665243948e-06, | |
| "loss": 0.0191, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.68599602619766, | |
| "grad_norm": 0.22479325532913208, | |
| "learning_rate": 5.233399563372335e-06, | |
| "loss": 0.0192, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 2.7227904923099566, | |
| "grad_norm": 0.27915650606155396, | |
| "learning_rate": 4.620158461500724e-06, | |
| "loss": 0.0193, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 2.7595849584222534, | |
| "grad_norm": 0.19762490689754486, | |
| "learning_rate": 4.006917359629112e-06, | |
| "loss": 0.0191, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 2.79637942453455, | |
| "grad_norm": 0.42420724034309387, | |
| "learning_rate": 3.3936762577575e-06, | |
| "loss": 0.0193, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 2.8331738906468464, | |
| "grad_norm": 0.34259703755378723, | |
| "learning_rate": 2.7804351558858883e-06, | |
| "loss": 0.0188, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 2.869968356759143, | |
| "grad_norm": 0.2734413743019104, | |
| "learning_rate": 2.1671940540142763e-06, | |
| "loss": 0.019, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.90676282287144, | |
| "grad_norm": 0.16011129319667816, | |
| "learning_rate": 1.5539529521426646e-06, | |
| "loss": 0.0186, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 2.9435572889837367, | |
| "grad_norm": 0.4719178080558777, | |
| "learning_rate": 9.407118502710525e-07, | |
| "loss": 0.0188, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.9803517550960335, | |
| "grad_norm": 0.22022365033626556, | |
| "learning_rate": 3.2747074839944075e-07, | |
| "loss": 0.0189, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.0266929492354393, | |
| "eval_mse": 0.026692949063357767, | |
| "eval_runtime": 78.6017, | |
| "eval_samples_per_second": 636.118, | |
| "eval_steps_per_second": 79.515, | |
| "step": 40767 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 40767, | |
| "total_flos": 1.7161499914378214e+17, | |
| "train_loss": 0.026733717791019525, | |
| "train_runtime": 12981.5942, | |
| "train_samples_per_second": 200.976, | |
| "train_steps_per_second": 3.14 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 40767, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.7161499914378214e+17, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |