| { | |
| "best_metric": 0.6883311867713928, | |
| "best_model_checkpoint": "./model_fine-tune/glot/mbert/yor-Latn/checkpoint-16500", | |
| "epoch": 67.34693877551021, | |
| "eval_steps": 500, | |
| "global_step": 16500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.0408163265306123, | |
| "grad_norm": 3.014599323272705, | |
| "learning_rate": 9.95e-05, | |
| "loss": 1.5111, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0408163265306123, | |
| "eval_accuracy": 0.7682061550515823, | |
| "eval_loss": 1.1437931060791016, | |
| "eval_runtime": 16.6138, | |
| "eval_samples_per_second": 117.974, | |
| "eval_steps_per_second": 3.732, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.081632653061225, | |
| "grad_norm": 2.2488036155700684, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 1.111, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.081632653061225, | |
| "eval_accuracy": 0.7948809959721713, | |
| "eval_loss": 0.9893151521682739, | |
| "eval_runtime": 16.5211, | |
| "eval_samples_per_second": 118.636, | |
| "eval_steps_per_second": 3.753, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.122448979591836, | |
| "grad_norm": 2.612577438354492, | |
| "learning_rate": 9.850000000000001e-05, | |
| "loss": 0.9843, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.122448979591836, | |
| "eval_accuracy": 0.8096590077564759, | |
| "eval_loss": 0.9186487793922424, | |
| "eval_runtime": 16.503, | |
| "eval_samples_per_second": 118.766, | |
| "eval_steps_per_second": 3.757, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.16326530612245, | |
| "grad_norm": 2.055933713912964, | |
| "learning_rate": 9.8e-05, | |
| "loss": 0.9099, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.16326530612245, | |
| "eval_accuracy": 0.8180220037796107, | |
| "eval_loss": 0.8695068359375, | |
| "eval_runtime": 16.5184, | |
| "eval_samples_per_second": 118.656, | |
| "eval_steps_per_second": 3.753, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 10.204081632653061, | |
| "grad_norm": 2.095768928527832, | |
| "learning_rate": 9.75e-05, | |
| "loss": 0.8553, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.204081632653061, | |
| "eval_accuracy": 0.8226890203332577, | |
| "eval_loss": 0.853586733341217, | |
| "eval_runtime": 16.0076, | |
| "eval_samples_per_second": 122.442, | |
| "eval_steps_per_second": 3.873, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 12.244897959183673, | |
| "grad_norm": 2.289365530014038, | |
| "learning_rate": 9.7e-05, | |
| "loss": 0.8136, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.244897959183673, | |
| "eval_accuracy": 0.8261254926090332, | |
| "eval_loss": 0.834091067314148, | |
| "eval_runtime": 16.0504, | |
| "eval_samples_per_second": 122.115, | |
| "eval_steps_per_second": 3.863, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "grad_norm": 2.0706839561462402, | |
| "learning_rate": 9.65e-05, | |
| "loss": 0.7732, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "eval_accuracy": 0.8323066793474285, | |
| "eval_loss": 0.7992222309112549, | |
| "eval_runtime": 15.9988, | |
| "eval_samples_per_second": 122.509, | |
| "eval_steps_per_second": 3.875, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 16.3265306122449, | |
| "grad_norm": 2.018846273422241, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.7447, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.3265306122449, | |
| "eval_accuracy": 0.8354064746360119, | |
| "eval_loss": 0.7927262783050537, | |
| "eval_runtime": 16.0248, | |
| "eval_samples_per_second": 122.31, | |
| "eval_steps_per_second": 3.869, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 18.367346938775512, | |
| "grad_norm": 1.8335440158843994, | |
| "learning_rate": 9.55e-05, | |
| "loss": 0.7207, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.367346938775512, | |
| "eval_accuracy": 0.8355885295341857, | |
| "eval_loss": 0.7959266901016235, | |
| "eval_runtime": 16.1763, | |
| "eval_samples_per_second": 121.165, | |
| "eval_steps_per_second": 3.833, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 20.408163265306122, | |
| "grad_norm": 2.1816558837890625, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.6936, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.408163265306122, | |
| "eval_accuracy": 0.8416515194650666, | |
| "eval_loss": 0.7648417353630066, | |
| "eval_runtime": 16.0654, | |
| "eval_samples_per_second": 122.001, | |
| "eval_steps_per_second": 3.859, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 22.448979591836736, | |
| "grad_norm": 1.9500852823257446, | |
| "learning_rate": 9.449999999999999e-05, | |
| "loss": 0.6715, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 22.448979591836736, | |
| "eval_accuracy": 0.8432883182052985, | |
| "eval_loss": 0.7536068558692932, | |
| "eval_runtime": 16.0018, | |
| "eval_samples_per_second": 122.486, | |
| "eval_steps_per_second": 3.875, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 24.489795918367346, | |
| "grad_norm": 1.791885495185852, | |
| "learning_rate": 9.4e-05, | |
| "loss": 0.6501, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 24.489795918367346, | |
| "eval_accuracy": 0.8429277638834043, | |
| "eval_loss": 0.7491132616996765, | |
| "eval_runtime": 16.0055, | |
| "eval_samples_per_second": 122.458, | |
| "eval_steps_per_second": 3.874, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 26.53061224489796, | |
| "grad_norm": 1.9352576732635498, | |
| "learning_rate": 9.350000000000001e-05, | |
| "loss": 0.635, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 26.53061224489796, | |
| "eval_accuracy": 0.8484582078727005, | |
| "eval_loss": 0.7326425909996033, | |
| "eval_runtime": 16.0205, | |
| "eval_samples_per_second": 122.343, | |
| "eval_steps_per_second": 3.87, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "grad_norm": 2.0857200622558594, | |
| "learning_rate": 9.300000000000001e-05, | |
| "loss": 0.6189, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "eval_accuracy": 0.8482937496821441, | |
| "eval_loss": 0.7334189414978027, | |
| "eval_runtime": 16.0058, | |
| "eval_samples_per_second": 122.455, | |
| "eval_steps_per_second": 3.874, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 30.612244897959183, | |
| "grad_norm": 2.3431918621063232, | |
| "learning_rate": 9.250000000000001e-05, | |
| "loss": 0.6035, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 30.612244897959183, | |
| "eval_accuracy": 0.8485175910580918, | |
| "eval_loss": 0.740364670753479, | |
| "eval_runtime": 16.0104, | |
| "eval_samples_per_second": 122.421, | |
| "eval_steps_per_second": 3.872, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 32.6530612244898, | |
| "grad_norm": 1.8592019081115723, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 0.586, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 32.6530612244898, | |
| "eval_accuracy": 0.8498935079850839, | |
| "eval_loss": 0.7310737371444702, | |
| "eval_runtime": 16.0169, | |
| "eval_samples_per_second": 122.371, | |
| "eval_steps_per_second": 3.871, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 34.69387755102041, | |
| "grad_norm": 2.139171838760376, | |
| "learning_rate": 9.15e-05, | |
| "loss": 0.5741, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 34.69387755102041, | |
| "eval_accuracy": 0.8535219936836398, | |
| "eval_loss": 0.7162159085273743, | |
| "eval_runtime": 16.0256, | |
| "eval_samples_per_second": 122.305, | |
| "eval_steps_per_second": 3.869, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 36.734693877551024, | |
| "grad_norm": 2.130950927734375, | |
| "learning_rate": 9.1e-05, | |
| "loss": 0.5613, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 36.734693877551024, | |
| "eval_accuracy": 0.8531061006400094, | |
| "eval_loss": 0.7184491753578186, | |
| "eval_runtime": 16.5555, | |
| "eval_samples_per_second": 118.39, | |
| "eval_steps_per_second": 3.745, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 38.775510204081634, | |
| "grad_norm": 1.8260786533355713, | |
| "learning_rate": 9.05e-05, | |
| "loss": 0.548, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 38.775510204081634, | |
| "eval_accuracy": 0.8545740978201136, | |
| "eval_loss": 0.7058050036430359, | |
| "eval_runtime": 16.5238, | |
| "eval_samples_per_second": 118.617, | |
| "eval_steps_per_second": 3.752, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 40.816326530612244, | |
| "grad_norm": 1.890523076057434, | |
| "learning_rate": 9e-05, | |
| "loss": 0.5395, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 40.816326530612244, | |
| "eval_accuracy": 0.857211986128117, | |
| "eval_loss": 0.710237443447113, | |
| "eval_runtime": 16.486, | |
| "eval_samples_per_second": 118.889, | |
| "eval_steps_per_second": 3.761, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 42.857142857142854, | |
| "grad_norm": 2.4147567749023438, | |
| "learning_rate": 8.950000000000001e-05, | |
| "loss": 0.5266, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 42.857142857142854, | |
| "eval_accuracy": 0.8557428516937341, | |
| "eval_loss": 0.7127901315689087, | |
| "eval_runtime": 16.5108, | |
| "eval_samples_per_second": 118.71, | |
| "eval_steps_per_second": 3.755, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 44.89795918367347, | |
| "grad_norm": 2.065382242202759, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 0.5156, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 44.89795918367347, | |
| "eval_accuracy": 0.8571209622093447, | |
| "eval_loss": 0.7171084880828857, | |
| "eval_runtime": 16.5009, | |
| "eval_samples_per_second": 118.782, | |
| "eval_steps_per_second": 3.757, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 46.93877551020408, | |
| "grad_norm": 2.3719356060028076, | |
| "learning_rate": 8.850000000000001e-05, | |
| "loss": 0.5088, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 46.93877551020408, | |
| "eval_accuracy": 0.8566442539649951, | |
| "eval_loss": 0.7080183029174805, | |
| "eval_runtime": 16.52, | |
| "eval_samples_per_second": 118.644, | |
| "eval_steps_per_second": 3.753, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 48.97959183673469, | |
| "grad_norm": 1.9313708543777466, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.497, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 48.97959183673469, | |
| "eval_accuracy": 0.8599520731972987, | |
| "eval_loss": 0.709360659122467, | |
| "eval_runtime": 16.5272, | |
| "eval_samples_per_second": 118.592, | |
| "eval_steps_per_second": 3.751, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 51.02040816326531, | |
| "grad_norm": 1.9448109865188599, | |
| "learning_rate": 8.75e-05, | |
| "loss": 0.4888, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 51.02040816326531, | |
| "eval_accuracy": 0.8609216073157768, | |
| "eval_loss": 0.706400454044342, | |
| "eval_runtime": 16.504, | |
| "eval_samples_per_second": 118.759, | |
| "eval_steps_per_second": 3.757, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 53.06122448979592, | |
| "grad_norm": 1.7106282711029053, | |
| "learning_rate": 8.7e-05, | |
| "loss": 0.4782, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 53.06122448979592, | |
| "eval_accuracy": 0.8604385547461287, | |
| "eval_loss": 0.7035485506057739, | |
| "eval_runtime": 16.5812, | |
| "eval_samples_per_second": 118.206, | |
| "eval_steps_per_second": 3.739, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 55.10204081632653, | |
| "grad_norm": 2.121896982192993, | |
| "learning_rate": 8.65e-05, | |
| "loss": 0.4719, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 55.10204081632653, | |
| "eval_accuracy": 0.8596197153330952, | |
| "eval_loss": 0.716853678226471, | |
| "eval_runtime": 16.5563, | |
| "eval_samples_per_second": 118.384, | |
| "eval_steps_per_second": 3.745, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 57.142857142857146, | |
| "grad_norm": 1.8903765678405762, | |
| "learning_rate": 8.6e-05, | |
| "loss": 0.4639, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 57.142857142857146, | |
| "eval_accuracy": 0.8602637276126747, | |
| "eval_loss": 0.7129377722740173, | |
| "eval_runtime": 16.5854, | |
| "eval_samples_per_second": 118.176, | |
| "eval_steps_per_second": 3.738, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 59.183673469387756, | |
| "grad_norm": 2.5342397689819336, | |
| "learning_rate": 8.55e-05, | |
| "loss": 0.4554, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 59.183673469387756, | |
| "eval_accuracy": 0.8614343212586901, | |
| "eval_loss": 0.7078380584716797, | |
| "eval_runtime": 16.0738, | |
| "eval_samples_per_second": 121.937, | |
| "eval_steps_per_second": 3.857, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 61.224489795918366, | |
| "grad_norm": 1.810963749885559, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.4437, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 61.224489795918366, | |
| "eval_accuracy": 0.8633158322532457, | |
| "eval_loss": 0.6998348236083984, | |
| "eval_runtime": 16.5802, | |
| "eval_samples_per_second": 118.213, | |
| "eval_steps_per_second": 3.739, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 63.265306122448976, | |
| "grad_norm": 1.8456369638442993, | |
| "learning_rate": 8.450000000000001e-05, | |
| "loss": 0.4424, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 63.265306122448976, | |
| "eval_accuracy": 0.8620835737606731, | |
| "eval_loss": 0.7072030901908875, | |
| "eval_runtime": 16.5746, | |
| "eval_samples_per_second": 118.253, | |
| "eval_steps_per_second": 3.741, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 65.3061224489796, | |
| "grad_norm": 1.9073472023010254, | |
| "learning_rate": 8.4e-05, | |
| "loss": 0.4338, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 65.3061224489796, | |
| "eval_accuracy": 0.8628925222058544, | |
| "eval_loss": 0.6999377012252808, | |
| "eval_runtime": 16.5979, | |
| "eval_samples_per_second": 118.087, | |
| "eval_steps_per_second": 3.735, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 67.34693877551021, | |
| "grad_norm": 1.9411852359771729, | |
| "learning_rate": 8.35e-05, | |
| "loss": 0.4252, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 67.34693877551021, | |
| "eval_accuracy": 0.8659541694954849, | |
| "eval_loss": 0.6883311867713928, | |
| "eval_runtime": 16.5927, | |
| "eval_samples_per_second": 118.124, | |
| "eval_steps_per_second": 3.737, | |
| "step": 16500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 100000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 409, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3911653816991744e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |