{ "best_metric": 0.4699208438396454, "best_model_checkpoint": "v11/checkpoint-474", "epoch": 3.0, "eval_steps": 500, "global_step": 474, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.02531645569620253, "grad_norm": 71.55428314208984, "learning_rate": 2.1097046413502108e-07, "loss": 3.1896, "step": 4 }, { "epoch": 0.05063291139240506, "grad_norm": Infinity, "learning_rate": 8.438818565400843e-07, "loss": 2.6405, "step": 8 }, { "epoch": 0.0759493670886076, "grad_norm": 37.44911575317383, "learning_rate": 1.6877637130801686e-06, "loss": 3.0762, "step": 12 }, { "epoch": 0.10126582278481013, "grad_norm": 46.11091995239258, "learning_rate": 2.531645569620253e-06, "loss": 2.5418, "step": 16 }, { "epoch": 0.12658227848101267, "grad_norm": 13.867141723632812, "learning_rate": 3.3755274261603373e-06, "loss": 2.7083, "step": 20 }, { "epoch": 0.1518987341772152, "grad_norm": 44.78251647949219, "learning_rate": 4.219409282700422e-06, "loss": 1.6333, "step": 24 }, { "epoch": 0.17721518987341772, "grad_norm": 63.29035949707031, "learning_rate": 5.063291139240506e-06, "loss": 2.092, "step": 28 }, { "epoch": 0.20253164556962025, "grad_norm": 29.540374755859375, "learning_rate": 5.907172995780591e-06, "loss": 2.0659, "step": 32 }, { "epoch": 0.22784810126582278, "grad_norm": 32.71324157714844, "learning_rate": 6.7510548523206746e-06, "loss": 2.0142, "step": 36 }, { "epoch": 0.25316455696202533, "grad_norm": 45.9168701171875, "learning_rate": 7.5949367088607605e-06, "loss": 1.9482, "step": 40 }, { "epoch": 0.27848101265822783, "grad_norm": 29.264678955078125, "learning_rate": 8.438818565400844e-06, "loss": 1.7309, "step": 44 }, { "epoch": 0.3037974683544304, "grad_norm": 35.710845947265625, "learning_rate": 9.28270042194093e-06, "loss": 1.0276, "step": 48 }, { "epoch": 0.3291139240506329, "grad_norm": 43.064239501953125, "learning_rate": 1.0126582278481012e-05, "loss": 1.3045, "step": 52 }, { "epoch": 0.35443037974683544, "grad_norm": 9.842476844787598, "learning_rate": 1.0970464135021098e-05, "loss": 1.3589, "step": 56 }, { "epoch": 0.379746835443038, "grad_norm": 14.65708065032959, "learning_rate": 1.1814345991561182e-05, "loss": 0.9091, "step": 60 }, { "epoch": 0.4050632911392405, "grad_norm": 11.407841682434082, "learning_rate": 1.2658227848101267e-05, "loss": 0.8581, "step": 64 }, { "epoch": 0.43037974683544306, "grad_norm": 35.808719635009766, "learning_rate": 1.3502109704641349e-05, "loss": 0.9538, "step": 68 }, { "epoch": 0.45569620253164556, "grad_norm": 11.020061492919922, "learning_rate": 1.4345991561181435e-05, "loss": 0.9462, "step": 72 }, { "epoch": 0.4810126582278481, "grad_norm": 6.515781402587891, "learning_rate": 1.5189873417721521e-05, "loss": 0.809, "step": 76 }, { "epoch": 0.5063291139240507, "grad_norm": 9.359097480773926, "learning_rate": 1.6033755274261607e-05, "loss": 0.6867, "step": 80 }, { "epoch": 0.5316455696202531, "grad_norm": 20.292892456054688, "learning_rate": 1.6877637130801688e-05, "loss": 0.6582, "step": 84 }, { "epoch": 0.5569620253164557, "grad_norm": 17.644256591796875, "learning_rate": 1.7721518987341772e-05, "loss": 0.7018, "step": 88 }, { "epoch": 0.5822784810126582, "grad_norm": 16.383758544921875, "learning_rate": 1.856540084388186e-05, "loss": 0.5546, "step": 92 }, { "epoch": 0.6075949367088608, "grad_norm": 13.477869033813477, "learning_rate": 1.940928270042194e-05, "loss": 0.7188, "step": 96 }, { "epoch": 0.6329113924050633, "grad_norm": 10.13427448272705, "learning_rate": 2.0253164556962025e-05, "loss": 0.7572, "step": 100 }, { "epoch": 0.6582278481012658, "grad_norm": 14.624322891235352, "learning_rate": 2.1097046413502112e-05, "loss": 0.6465, "step": 104 }, { "epoch": 0.6835443037974683, "grad_norm": 6.965182781219482, "learning_rate": 2.1940928270042196e-05, "loss": 0.5644, "step": 108 }, { "epoch": 0.7088607594936709, "grad_norm": 14.612415313720703, "learning_rate": 2.278481012658228e-05, "loss": 0.6489, "step": 112 }, { "epoch": 0.7341772151898734, "grad_norm": 10.756507873535156, "learning_rate": 2.3628691983122365e-05, "loss": 0.6872, "step": 116 }, { "epoch": 0.759493670886076, "grad_norm": 10.897459030151367, "learning_rate": 2.447257383966245e-05, "loss": 0.5815, "step": 120 }, { "epoch": 0.7848101265822784, "grad_norm": 5.7723188400268555, "learning_rate": 2.5316455696202533e-05, "loss": 0.6707, "step": 124 }, { "epoch": 0.810126582278481, "grad_norm": 6.329953193664551, "learning_rate": 2.616033755274262e-05, "loss": 0.8948, "step": 128 }, { "epoch": 0.8354430379746836, "grad_norm": 81.23648071289062, "learning_rate": 2.7004219409282698e-05, "loss": 0.7619, "step": 132 }, { "epoch": 0.8607594936708861, "grad_norm": 27.8104248046875, "learning_rate": 2.7848101265822786e-05, "loss": 0.6075, "step": 136 }, { "epoch": 0.8860759493670886, "grad_norm": 8.194925308227539, "learning_rate": 2.869198312236287e-05, "loss": 0.82, "step": 140 }, { "epoch": 0.9113924050632911, "grad_norm": 9.09708309173584, "learning_rate": 2.9535864978902954e-05, "loss": 0.6168, "step": 144 }, { "epoch": 0.9367088607594937, "grad_norm": 13.87863540649414, "learning_rate": 3.0379746835443042e-05, "loss": 0.5575, "step": 148 }, { "epoch": 0.9620253164556962, "grad_norm": 20.190404891967773, "learning_rate": 3.1223628691983126e-05, "loss": 0.9551, "step": 152 }, { "epoch": 0.9873417721518988, "grad_norm": 6.65585470199585, "learning_rate": 3.2067510548523214e-05, "loss": 0.7284, "step": 156 }, { "epoch": 1.0, "eval_accuracy": 0.7611111111111111, "eval_f1_macro": 0.8171827313856298, "eval_f1_micro": 0.7611111111111111, "eval_f1_weighted": 0.7610008434935971, "eval_loss": 0.5506798028945923, "eval_precision_macro": 0.8070455512984248, "eval_precision_micro": 0.7611111111111111, "eval_precision_weighted": 0.7688892436497802, "eval_recall_macro": 0.8340425531914893, "eval_recall_micro": 0.7611111111111111, "eval_recall_weighted": 0.7611111111111111, "eval_runtime": 0.4991, "eval_samples_per_second": 360.643, "eval_steps_per_second": 24.043, "step": 158 }, { "epoch": 1.0126582278481013, "grad_norm": 6.7371063232421875, "learning_rate": 3.291139240506329e-05, "loss": 0.5544, "step": 160 }, { "epoch": 1.0379746835443038, "grad_norm": 7.718777656555176, "learning_rate": 3.3755274261603375e-05, "loss": 0.5192, "step": 164 }, { "epoch": 1.0632911392405062, "grad_norm": 18.91424560546875, "learning_rate": 3.459915611814346e-05, "loss": 0.6722, "step": 168 }, { "epoch": 1.0886075949367089, "grad_norm": 14.480657577514648, "learning_rate": 3.5443037974683544e-05, "loss": 0.6675, "step": 172 }, { "epoch": 1.1139240506329113, "grad_norm": 15.364533424377441, "learning_rate": 3.628691983122363e-05, "loss": 0.5339, "step": 176 }, { "epoch": 1.139240506329114, "grad_norm": 11.355390548706055, "learning_rate": 3.713080168776372e-05, "loss": 0.4978, "step": 180 }, { "epoch": 1.1645569620253164, "grad_norm": 11.57992935180664, "learning_rate": 3.79746835443038e-05, "loss": 0.5707, "step": 184 }, { "epoch": 1.189873417721519, "grad_norm": 13.356510162353516, "learning_rate": 3.881856540084388e-05, "loss": 0.5443, "step": 188 }, { "epoch": 1.2151898734177216, "grad_norm": 6.806561470031738, "learning_rate": 3.966244725738397e-05, "loss": 0.4811, "step": 192 }, { "epoch": 1.240506329113924, "grad_norm": 49.87335205078125, "learning_rate": 4.050632911392405e-05, "loss": 0.5528, "step": 196 }, { "epoch": 1.2658227848101267, "grad_norm": 29.326234817504883, "learning_rate": 4.135021097046414e-05, "loss": 0.6233, "step": 200 }, { "epoch": 1.2911392405063291, "grad_norm": 23.68698501586914, "learning_rate": 4.2194092827004224e-05, "loss": 0.6821, "step": 204 }, { "epoch": 1.3164556962025316, "grad_norm": 21.932170867919922, "learning_rate": 4.3037974683544305e-05, "loss": 0.8705, "step": 208 }, { "epoch": 1.3417721518987342, "grad_norm": 16.624055862426758, "learning_rate": 4.388185654008439e-05, "loss": 0.4985, "step": 212 }, { "epoch": 1.3670886075949367, "grad_norm": 18.823989868164062, "learning_rate": 4.4725738396624474e-05, "loss": 0.7091, "step": 216 }, { "epoch": 1.3924050632911391, "grad_norm": 5.886724948883057, "learning_rate": 4.556962025316456e-05, "loss": 0.4788, "step": 220 }, { "epoch": 1.4177215189873418, "grad_norm": 19.74236297607422, "learning_rate": 4.641350210970464e-05, "loss": 0.681, "step": 224 }, { "epoch": 1.4430379746835442, "grad_norm": 5.201137542724609, "learning_rate": 4.725738396624473e-05, "loss": 0.6167, "step": 228 }, { "epoch": 1.4683544303797469, "grad_norm": 13.234941482543945, "learning_rate": 4.810126582278481e-05, "loss": 0.5769, "step": 232 }, { "epoch": 1.4936708860759493, "grad_norm": 17.14457893371582, "learning_rate": 4.89451476793249e-05, "loss": 0.5358, "step": 236 }, { "epoch": 1.518987341772152, "grad_norm": 18.709863662719727, "learning_rate": 4.9789029535864986e-05, "loss": 0.7973, "step": 240 }, { "epoch": 1.5443037974683544, "grad_norm": 12.674654006958008, "learning_rate": 4.9929676511955e-05, "loss": 0.6866, "step": 244 }, { "epoch": 1.5696202531645569, "grad_norm": 28.57341194152832, "learning_rate": 4.983591186122832e-05, "loss": 0.6526, "step": 248 }, { "epoch": 1.5949367088607596, "grad_norm": 10.77186393737793, "learning_rate": 4.974214721050164e-05, "loss": 0.5577, "step": 252 }, { "epoch": 1.620253164556962, "grad_norm": 8.296313285827637, "learning_rate": 4.964838255977497e-05, "loss": 0.6331, "step": 256 }, { "epoch": 1.6455696202531644, "grad_norm": 10.310996055603027, "learning_rate": 4.955461790904829e-05, "loss": 0.569, "step": 260 }, { "epoch": 1.6708860759493671, "grad_norm": 13.412028312683105, "learning_rate": 4.9460853258321614e-05, "loss": 0.6135, "step": 264 }, { "epoch": 1.6962025316455698, "grad_norm": 12.334749221801758, "learning_rate": 4.936708860759494e-05, "loss": 0.4672, "step": 268 }, { "epoch": 1.721518987341772, "grad_norm": 10.7319974899292, "learning_rate": 4.927332395686826e-05, "loss": 0.5693, "step": 272 }, { "epoch": 1.7468354430379747, "grad_norm": 8.58609676361084, "learning_rate": 4.917955930614159e-05, "loss": 0.4975, "step": 276 }, { "epoch": 1.7721518987341773, "grad_norm": 10.97795581817627, "learning_rate": 4.908579465541491e-05, "loss": 0.6971, "step": 280 }, { "epoch": 1.7974683544303798, "grad_norm": 4.033533573150635, "learning_rate": 4.8992030004688236e-05, "loss": 0.3827, "step": 284 }, { "epoch": 1.8227848101265822, "grad_norm": 8.546496391296387, "learning_rate": 4.889826535396156e-05, "loss": 0.4725, "step": 288 }, { "epoch": 1.8481012658227849, "grad_norm": 16.97895050048828, "learning_rate": 4.8804500703234885e-05, "loss": 0.6739, "step": 292 }, { "epoch": 1.8734177215189873, "grad_norm": 50.98750305175781, "learning_rate": 4.87107360525082e-05, "loss": 0.6749, "step": 296 }, { "epoch": 1.8987341772151898, "grad_norm": 6.033726215362549, "learning_rate": 4.861697140178153e-05, "loss": 0.6029, "step": 300 }, { "epoch": 1.9240506329113924, "grad_norm": 70.4456558227539, "learning_rate": 4.852320675105486e-05, "loss": 0.6886, "step": 304 }, { "epoch": 1.9493670886075949, "grad_norm": 15.961381912231445, "learning_rate": 4.8429442100328175e-05, "loss": 0.8721, "step": 308 }, { "epoch": 1.9746835443037973, "grad_norm": 7.286694526672363, "learning_rate": 4.83356774496015e-05, "loss": 0.6044, "step": 312 }, { "epoch": 2.0, "grad_norm": 17.835411071777344, "learning_rate": 4.824191279887483e-05, "loss": 0.6555, "step": 316 }, { "epoch": 2.0, "eval_accuracy": 0.7833333333333333, "eval_f1_macro": 0.7629566553128967, "eval_f1_micro": 0.7833333333333333, "eval_f1_weighted": 0.7813522674146798, "eval_loss": 0.6177344918251038, "eval_precision_macro": 0.7341277483549993, "eval_precision_micro": 0.7833333333333333, "eval_precision_weighted": 0.8182069491253411, "eval_recall_macro": 0.8563120567375887, "eval_recall_micro": 0.7833333333333333, "eval_recall_weighted": 0.7833333333333333, "eval_runtime": 0.4807, "eval_samples_per_second": 374.457, "eval_steps_per_second": 24.964, "step": 316 }, { "epoch": 2.0253164556962027, "grad_norm": 8.706924438476562, "learning_rate": 4.814814814814815e-05, "loss": 0.6555, "step": 320 }, { "epoch": 2.050632911392405, "grad_norm": 8.192380905151367, "learning_rate": 4.805438349742147e-05, "loss": 0.5678, "step": 324 }, { "epoch": 2.0759493670886076, "grad_norm": 26.89942169189453, "learning_rate": 4.79606188466948e-05, "loss": 0.3685, "step": 328 }, { "epoch": 2.1012658227848102, "grad_norm": 20.788925170898438, "learning_rate": 4.786685419596813e-05, "loss": 0.4393, "step": 332 }, { "epoch": 2.1265822784810124, "grad_norm": 17.45987319946289, "learning_rate": 4.7773089545241446e-05, "loss": 0.3558, "step": 336 }, { "epoch": 2.151898734177215, "grad_norm": 17.20940589904785, "learning_rate": 4.767932489451477e-05, "loss": 0.3101, "step": 340 }, { "epoch": 2.1772151898734178, "grad_norm": 13.34177017211914, "learning_rate": 4.7585560243788094e-05, "loss": 0.6948, "step": 344 }, { "epoch": 2.2025316455696204, "grad_norm": 8.63040542602539, "learning_rate": 4.749179559306142e-05, "loss": 0.4689, "step": 348 }, { "epoch": 2.2278481012658227, "grad_norm": 41.929141998291016, "learning_rate": 4.739803094233474e-05, "loss": 0.4344, "step": 352 }, { "epoch": 2.2531645569620253, "grad_norm": 26.385360717773438, "learning_rate": 4.730426629160807e-05, "loss": 0.6043, "step": 356 }, { "epoch": 2.278481012658228, "grad_norm": 17.394569396972656, "learning_rate": 4.7210501640881385e-05, "loss": 0.3698, "step": 360 }, { "epoch": 2.3037974683544302, "grad_norm": 1.6962029933929443, "learning_rate": 4.7116736990154716e-05, "loss": 0.2685, "step": 364 }, { "epoch": 2.329113924050633, "grad_norm": 19.53297233581543, "learning_rate": 4.702297233942804e-05, "loss": 0.3477, "step": 368 }, { "epoch": 2.3544303797468356, "grad_norm": 5.328404903411865, "learning_rate": 4.692920768870136e-05, "loss": 0.2694, "step": 372 }, { "epoch": 2.379746835443038, "grad_norm": 2.5393640995025635, "learning_rate": 4.683544303797468e-05, "loss": 0.1929, "step": 376 }, { "epoch": 2.4050632911392404, "grad_norm": 14.026851654052734, "learning_rate": 4.6741678387248013e-05, "loss": 0.4045, "step": 380 }, { "epoch": 2.430379746835443, "grad_norm": 36.21650695800781, "learning_rate": 4.664791373652133e-05, "loss": 0.1961, "step": 384 }, { "epoch": 2.4556962025316453, "grad_norm": 38.489322662353516, "learning_rate": 4.6554149085794655e-05, "loss": 0.5762, "step": 388 }, { "epoch": 2.481012658227848, "grad_norm": 56.10824966430664, "learning_rate": 4.646038443506798e-05, "loss": 1.1768, "step": 392 }, { "epoch": 2.5063291139240507, "grad_norm": 2.1792027950286865, "learning_rate": 4.6366619784341304e-05, "loss": 0.3245, "step": 396 }, { "epoch": 2.5316455696202533, "grad_norm": 25.301136016845703, "learning_rate": 4.627285513361463e-05, "loss": 0.4881, "step": 400 }, { "epoch": 2.5569620253164556, "grad_norm": 23.114078521728516, "learning_rate": 4.617909048288795e-05, "loss": 0.6298, "step": 404 }, { "epoch": 2.5822784810126582, "grad_norm": 6.997201442718506, "learning_rate": 4.608532583216128e-05, "loss": 0.3664, "step": 408 }, { "epoch": 2.607594936708861, "grad_norm": 20.35528564453125, "learning_rate": 4.59915611814346e-05, "loss": 0.3108, "step": 412 }, { "epoch": 2.632911392405063, "grad_norm": 24.20511245727539, "learning_rate": 4.5897796530707926e-05, "loss": 0.3533, "step": 416 }, { "epoch": 2.6582278481012658, "grad_norm": 2.4780113697052, "learning_rate": 4.580403187998125e-05, "loss": 0.4303, "step": 420 }, { "epoch": 2.6835443037974684, "grad_norm": 49.66173553466797, "learning_rate": 4.5710267229254575e-05, "loss": 0.4879, "step": 424 }, { "epoch": 2.708860759493671, "grad_norm": 22.363910675048828, "learning_rate": 4.56165025785279e-05, "loss": 0.4178, "step": 428 }, { "epoch": 2.7341772151898733, "grad_norm": 16.556734085083008, "learning_rate": 4.552273792780122e-05, "loss": 0.2627, "step": 432 }, { "epoch": 2.759493670886076, "grad_norm": 14.872969627380371, "learning_rate": 4.542897327707454e-05, "loss": 0.3911, "step": 436 }, { "epoch": 2.7848101265822782, "grad_norm": 15.42312240600586, "learning_rate": 4.533520862634787e-05, "loss": 0.4543, "step": 440 }, { "epoch": 2.810126582278481, "grad_norm": 19.359683990478516, "learning_rate": 4.5241443975621196e-05, "loss": 0.3946, "step": 444 }, { "epoch": 2.8354430379746836, "grad_norm": 20.327611923217773, "learning_rate": 4.5147679324894514e-05, "loss": 0.3452, "step": 448 }, { "epoch": 2.8607594936708862, "grad_norm": 24.550987243652344, "learning_rate": 4.505391467416784e-05, "loss": 0.7032, "step": 452 }, { "epoch": 2.8860759493670884, "grad_norm": 13.352927207946777, "learning_rate": 4.496015002344117e-05, "loss": 0.4979, "step": 456 }, { "epoch": 2.911392405063291, "grad_norm": 20.34762954711914, "learning_rate": 4.486638537271449e-05, "loss": 0.5669, "step": 460 }, { "epoch": 2.9367088607594938, "grad_norm": 2.153109312057495, "learning_rate": 4.477262072198781e-05, "loss": 0.2976, "step": 464 }, { "epoch": 2.962025316455696, "grad_norm": 21.230073928833008, "learning_rate": 4.4678856071261136e-05, "loss": 0.421, "step": 468 }, { "epoch": 2.9873417721518987, "grad_norm": 12.1097412109375, "learning_rate": 4.458509142053446e-05, "loss": 0.2361, "step": 472 }, { "epoch": 3.0, "eval_accuracy": 0.8277777777777777, "eval_f1_macro": 0.8648394526320947, "eval_f1_micro": 0.8277777777777777, "eval_f1_weighted": 0.827145991318232, "eval_loss": 0.4699208438396454, "eval_precision_macro": 0.8595340501792115, "eval_precision_micro": 0.8277777777777777, "eval_precision_weighted": 0.8456027479091995, "eval_recall_macro": 0.8846808510638299, "eval_recall_micro": 0.8277777777777777, "eval_recall_weighted": 0.8277777777777777, "eval_runtime": 0.4998, "eval_samples_per_second": 360.158, "eval_steps_per_second": 24.011, "step": 474 } ], "logging_steps": 4, "max_steps": 2370, "num_input_tokens_seen": 0, "num_train_epochs": 15, "save_steps": 500, "total_flos": 248642179752960.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }