| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.863013698630137, | |
| "eval_steps": 500, | |
| "global_step": 360, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0273972602739726, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 3.0637, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.136986301369863, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 3.056, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 3.0171, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.410958904109589, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 2.8054, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 2.4802, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 0.0001388888888888889, | |
| "loss": 2.2946, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 2.0875, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.958904109589041, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00019444444444444446, | |
| "loss": 1.8806, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.9863013698630136, | |
| "eval_loss": 2.6093997955322266, | |
| "eval_runtime": 0.5633, | |
| "eval_samples_per_second": 17.753, | |
| "eval_steps_per_second": 1.775, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "grad_norm": 44.0, | |
| "learning_rate": 0.00019992479525042303, | |
| "loss": 1.6979, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2328767123287672, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 0.00019961946980917456, | |
| "loss": 1.5925, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.00019908004033648453, | |
| "loss": 1.5152, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5068493150684932, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.00019830777448228603, | |
| "loss": 1.4565, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 1.3866, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.7808219178082192, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00019607253577167205, | |
| "loss": 1.3522, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00019461481568757506, | |
| "loss": 1.3239, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.535839080810547, | |
| "eval_runtime": 0.5617, | |
| "eval_samples_per_second": 17.804, | |
| "eval_steps_per_second": 1.78, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.0547945205479454, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00019293475242268223, | |
| "loss": 1.3035, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.0001910362940966147, | |
| "loss": 1.2854, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.328767123287671, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00018892390205361062, | |
| "loss": 1.2525, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 1.26, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.602739726027397, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.00018407766423091034, | |
| "loss": 1.2323, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00018135520702629675, | |
| "loss": 1.2267, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.8767123287671232, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00017844156649195759, | |
| "loss": 1.2327, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.9863013698630136, | |
| "eval_loss": 2.519155979156494, | |
| "eval_runtime": 0.5663, | |
| "eval_samples_per_second": 17.659, | |
| "eval_steps_per_second": 1.766, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.0136986301369864, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00017534358963276607, | |
| "loss": 1.2242, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.1506849315068495, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00017206855664077147, | |
| "loss": 1.1938, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.287671232876712, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 1.1879, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.4246575342465753, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.00016501850533471836, | |
| "loss": 1.1748, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.5616438356164384, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.0001612600545193203, | |
| "loss": 1.1704, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.6986301369863015, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.0001573576436351046, | |
| "loss": 1.1679, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 3.8356164383561646, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00015332044328016914, | |
| "loss": 1.1663, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.9726027397260273, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00014915794080553707, | |
| "loss": 1.1735, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.520301103591919, | |
| "eval_runtime": 0.5527, | |
| "eval_samples_per_second": 18.093, | |
| "eval_steps_per_second": 1.809, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.109589041095891, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00014487991802004623, | |
| "loss": 1.144, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.2465753424657535, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00014049642820326735, | |
| "loss": 1.1273, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.383561643835616, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00013601777248047105, | |
| "loss": 1.1344, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.52054794520548, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.00013145447561516138, | |
| "loss": 1.1364, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 4.657534246575342, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00012681726127606376, | |
| "loss": 1.1344, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 4.794520547945205, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.00012211702683668878, | |
| "loss": 1.1252, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 4.931506849315069, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 1.1354, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.986301369863014, | |
| "eval_loss": 2.546712875366211, | |
| "eval_runtime": 0.5597, | |
| "eval_samples_per_second": 17.868, | |
| "eval_steps_per_second": 1.787, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.068493150684931, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.00011257180167521629, | |
| "loss": 1.1069, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.205479452054795, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.0001077492420671931, | |
| "loss": 1.0983, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.342465753424658, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00010290847187431113, | |
| "loss": 1.0952, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.47945205479452, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 9.806086682281758e-05, | |
| "loss": 1.1105, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.616438356164384, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 9.321781870075908e-05, | |
| "loss": 1.1041, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 5.7534246575342465, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 8.839070858747697e-05, | |
| "loss": 1.0954, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 5.890410958904109, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 8.35908801082676e-05, | |
| "loss": 1.1015, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.5496411323547363, | |
| "eval_runtime": 0.5525, | |
| "eval_samples_per_second": 18.099, | |
| "eval_steps_per_second": 1.81, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 6.027397260273973, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 7.882961277705895e-05, | |
| "loss": 1.0967, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.164383561643835, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 7.411809548974792e-05, | |
| "loss": 1.0792, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 6.301369863013699, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 6.94674002304887e-05, | |
| "loss": 1.0787, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 6.438356164383562, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 6.488845605272113e-05, | |
| "loss": 1.0699, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 6.575342465753424, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 1.0875, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 6.712328767123288, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 5.5988668799569545e-05, | |
| "loss": 1.0721, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 6.8493150684931505, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 5.168874007033615e-05, | |
| "loss": 1.0769, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 6.986301369863014, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 4.7502341966544e-05, | |
| "loss": 1.0858, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 6.986301369863014, | |
| "eval_loss": 2.568040609359741, | |
| "eval_runtime": 0.5489, | |
| "eval_samples_per_second": 18.22, | |
| "eval_steps_per_second": 1.822, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 7.123287671232877, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.343931245134616e-05, | |
| "loss": 1.0577, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 7.260273972602739, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 3.950919957384582e-05, | |
| "loss": 1.0614, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 7.397260273972603, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 1.0698, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 7.534246575342466, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 3.2084332465620694e-05, | |
| "loss": 1.0695, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 7.671232876712329, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 2.8607026544210114e-05, | |
| "loss": 1.0702, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 7.808219178082192, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 2.529749287590042e-05, | |
| "loss": 1.0632, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 7.945205479452055, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.2163508807583998e-05, | |
| "loss": 1.0624, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.572319746017456, | |
| "eval_runtime": 0.5428, | |
| "eval_samples_per_second": 18.424, | |
| "eval_steps_per_second": 1.842, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 8.082191780821917, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.921243914762889e-05, | |
| "loss": 1.0583, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 8.219178082191782, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 1.6451218858706374e-05, | |
| "loss": 1.0536, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 8.356164383561644, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.388633676074862e-05, | |
| "loss": 1.0613, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 8.493150684931507, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.1523820282334219e-05, | |
| "loss": 1.0589, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 8.63013698630137, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.369221296335006e-06, | |
| "loss": 1.0664, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 8.767123287671232, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 7.427603073110967e-06, | |
| "loss": 1.0656, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 8.904109589041095, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 5.7035283819124155e-06, | |
| "loss": 1.0546, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 8.986301369863014, | |
| "eval_loss": 2.5756430625915527, | |
| "eval_runtime": 0.5591, | |
| "eval_samples_per_second": 17.886, | |
| "eval_steps_per_second": 1.789, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 9.04109589041096, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 4.20104876845111e-06, | |
| "loss": 1.0515, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 9.178082191780822, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 2.9236950338380033e-06, | |
| "loss": 1.0521, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 9.315068493150685, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 1.874468937261531e-06, | |
| "loss": 1.0563, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 9.452054794520548, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 1.055836141905553e-06, | |
| "loss": 1.0535, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 9.58904109589041, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.6972042068341714e-07, | |
| "loss": 1.0521, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 9.726027397260275, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 1.1749913540496371e-07, | |
| "loss": 1.0664, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 9.863013698630137, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.0, | |
| "loss": 1.0623, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 9.863013698630137, | |
| "eval_loss": 2.5758092403411865, | |
| "eval_runtime": 0.5499, | |
| "eval_samples_per_second": 18.186, | |
| "eval_steps_per_second": 1.819, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 9.863013698630137, | |
| "step": 360, | |
| "total_flos": 2.1145440197646746e+17, | |
| "train_loss": 1.288370986117257, | |
| "train_runtime": 1930.7273, | |
| "train_samples_per_second": 9.074, | |
| "train_steps_per_second": 0.186 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 360, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.1145440197646746e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |