| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 4020, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0024875621890547263, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 4.975124378109453e-07, | |
| "loss": 2.7789, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012437810945273632, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 2.4875621890547264e-06, | |
| "loss": 2.7762, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024875621890547265, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 4.975124378109453e-06, | |
| "loss": 2.7881, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03731343283582089, | |
| "grad_norm": 3.375, | |
| "learning_rate": 7.4626865671641785e-06, | |
| "loss": 2.7896, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04975124378109453, | |
| "grad_norm": 38.25, | |
| "learning_rate": 9.950248756218906e-06, | |
| "loss": 2.7064, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06218905472636816, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 1.2437810945273633e-05, | |
| "loss": 2.6224, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07462686567164178, | |
| "grad_norm": 3.875, | |
| "learning_rate": 1.4925373134328357e-05, | |
| "loss": 2.5409, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08706467661691543, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 1.7412935323383088e-05, | |
| "loss": 2.3895, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09950248756218906, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 1.990049751243781e-05, | |
| "loss": 2.2849, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11194029850746269, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 2.238805970149254e-05, | |
| "loss": 2.1737, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.12437810945273632, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 2.4875621890547266e-05, | |
| "loss": 2.0748, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13681592039800994, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 2.736318407960199e-05, | |
| "loss": 1.9485, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.14925373134328357, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 2.9850746268656714e-05, | |
| "loss": 1.8546, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16169154228855723, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 3.233830845771145e-05, | |
| "loss": 1.7576, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.17412935323383086, | |
| "grad_norm": 0.98828125, | |
| "learning_rate": 3.4825870646766175e-05, | |
| "loss": 1.6891, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1865671641791045, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 3.73134328358209e-05, | |
| "loss": 1.603, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.19900497512437812, | |
| "grad_norm": 0.83203125, | |
| "learning_rate": 3.980099502487562e-05, | |
| "loss": 1.5317, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.21144278606965175, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 4.228855721393035e-05, | |
| "loss": 1.4721, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.22388059701492538, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 4.477611940298508e-05, | |
| "loss": 1.4417, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.236318407960199, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 4.7263681592039805e-05, | |
| "loss": 1.3933, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.24875621890547264, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 4.975124378109453e-05, | |
| "loss": 1.3623, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26119402985074625, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 5.223880597014925e-05, | |
| "loss": 1.3171, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2736318407960199, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 5.472636815920398e-05, | |
| "loss": 1.2842, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2860696517412935, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 5.721393034825871e-05, | |
| "loss": 1.2869, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 5.970149253731343e-05, | |
| "loss": 1.2504, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.31094527363184077, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 6.218905472636816e-05, | |
| "loss": 1.2319, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.32338308457711445, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 6.46766169154229e-05, | |
| "loss": 1.2176, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3358208955223881, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 6.716417910447762e-05, | |
| "loss": 1.2129, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3482587064676617, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 6.965174129353235e-05, | |
| "loss": 1.1856, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.36069651741293535, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 7.213930348258708e-05, | |
| "loss": 1.2091, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.373134328358209, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 7.46268656716418e-05, | |
| "loss": 1.1711, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3855721393034826, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 7.711442786069652e-05, | |
| "loss": 1.1618, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.39800995024875624, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 7.960199004975125e-05, | |
| "loss": 1.1649, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.41044776119402987, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 8.208955223880597e-05, | |
| "loss": 1.1657, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4228855721393035, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 8.45771144278607e-05, | |
| "loss": 1.1389, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.43532338308457713, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 8.706467661691543e-05, | |
| "loss": 1.1364, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.44776119402985076, | |
| "grad_norm": 0.5, | |
| "learning_rate": 8.955223880597016e-05, | |
| "loss": 1.146, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4601990049751244, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 9.203980099502488e-05, | |
| "loss": 1.1374, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.472636815920398, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 9.452736318407961e-05, | |
| "loss": 1.1275, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.48507462686567165, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 9.701492537313434e-05, | |
| "loss": 1.1275, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4975124378109453, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 9.950248756218906e-05, | |
| "loss": 1.131, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5099502487562189, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 0.00010199004975124378, | |
| "loss": 1.1071, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5223880597014925, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.0001044776119402985, | |
| "loss": 1.1169, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5348258706467661, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 0.00010696517412935323, | |
| "loss": 1.1034, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5472636815920398, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 0.00010945273631840796, | |
| "loss": 1.0966, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5597014925373134, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00011194029850746269, | |
| "loss": 1.0964, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.572139303482587, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00011442786069651741, | |
| "loss": 1.0892, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5845771144278606, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00011691542288557213, | |
| "loss": 1.0794, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 0.00011940298507462686, | |
| "loss": 1.087, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6094527363184079, | |
| "grad_norm": 0.75, | |
| "learning_rate": 0.00012189054726368158, | |
| "loss": 1.079, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6218905472636815, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 0.0001243781094527363, | |
| "loss": 1.0825, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6343283582089553, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.00012686567164179105, | |
| "loss": 1.0758, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6467661691542289, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 0.0001293532338308458, | |
| "loss": 1.0802, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6592039800995025, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 0.0001318407960199005, | |
| "loss": 1.0734, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6716417910447762, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00013432835820895525, | |
| "loss": 1.0696, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6840796019900498, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 0.00013681592039800996, | |
| "loss": 1.0704, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6965174129353234, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.0001393034825870647, | |
| "loss": 1.0645, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7089552238805971, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 0.00014179104477611942, | |
| "loss": 1.067, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7213930348258707, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00014427860696517416, | |
| "loss": 1.0581, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7338308457711443, | |
| "grad_norm": 0.75, | |
| "learning_rate": 0.00014676616915422887, | |
| "loss": 1.0562, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.0001492537313432836, | |
| "loss": 1.0526, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7587064676616916, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 0.00015174129353233832, | |
| "loss": 1.0433, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7711442786069652, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00015422885572139304, | |
| "loss": 1.0522, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7835820895522388, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 0.00015671641791044778, | |
| "loss": 1.0384, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7960199004975125, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.0001592039800995025, | |
| "loss": 1.0564, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8084577114427861, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 0.00016169154228855723, | |
| "loss": 1.0552, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8208955223880597, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 0.00016417910447761195, | |
| "loss": 1.0369, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 1.0578, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.845771144278607, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.0001691542288557214, | |
| "loss": 1.0378, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8582089552238806, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00017164179104477614, | |
| "loss": 1.0414, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.8706467661691543, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00017412935323383086, | |
| "loss": 1.0403, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8830845771144279, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 0.0001766169154228856, | |
| "loss": 1.045, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8955223880597015, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.0001791044776119403, | |
| "loss": 1.035, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9079601990049752, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.00018159203980099505, | |
| "loss": 1.0408, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.9203980099502488, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00018407960199004977, | |
| "loss": 1.0245, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9328358208955224, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 0.00018656716417910448, | |
| "loss": 1.0161, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.945273631840796, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 0.00018905472636815922, | |
| "loss": 1.0275, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9577114427860697, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 0.00019154228855721393, | |
| "loss": 1.0326, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.9701492537313433, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 0.00019402985074626867, | |
| "loss": 1.0283, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9825870646766169, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 0.0001965174129353234, | |
| "loss": 1.0265, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9950248756218906, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00019900497512437813, | |
| "loss": 1.0192, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.451380968093872, | |
| "eval_runtime": 0.5369, | |
| "eval_samples_per_second": 18.624, | |
| "eval_steps_per_second": 1.862, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.007462686567164, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00019999966070700876, | |
| "loss": 1.0109, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.0199004975124377, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 0.000199997587258178, | |
| "loss": 1.0148, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0323383084577114, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00019999362889565903, | |
| "loss": 0.9985, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.044776119402985, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 0.0001999877856940653, | |
| "loss": 0.9966, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.0572139303482586, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00019998005776353885, | |
| "loss": 1.0056, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.0696517412935322, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 0.00019997044524974799, | |
| "loss": 1.0063, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.0820895522388059, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.00019995894833388465, | |
| "loss": 1.014, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.0945273631840795, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.00019994556723266103, | |
| "loss": 1.0014, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.1069651741293531, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.0001999303021983054, | |
| "loss": 1.0067, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.1194029850746268, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00019991315351855748, | |
| "loss": 0.9913, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1318407960199006, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.0001998941215166628, | |
| "loss": 1.0055, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.144278606965174, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00019987320655136693, | |
| "loss": 0.9968, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.1567164179104479, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.0001998504090169083, | |
| "loss": 0.9908, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.1691542288557213, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00019982572934301122, | |
| "loss": 0.9925, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.1815920398009951, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.0001997991679948774, | |
| "loss": 0.9979, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.1940298507462686, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.0001997707254731775, | |
| "loss": 0.9786, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.2064676616915424, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00019974040231404133, | |
| "loss": 0.9718, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.2189054726368158, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00019970819908904814, | |
| "loss": 0.9866, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.2313432835820897, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00019967411640521552, | |
| "loss": 0.9954, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.243781094527363, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 0.00019963815490498817, | |
| "loss": 0.9949, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.256218905472637, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00019960031526622572, | |
| "loss": 0.9773, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.2686567164179103, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00019956059820218982, | |
| "loss": 0.9883, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.2810945273631842, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00019951900446153097, | |
| "loss": 0.9913, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.2935323383084576, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00019947553482827418, | |
| "loss": 0.9902, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.3059701492537314, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.0001994301901218043, | |
| "loss": 0.9783, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.3184079601990049, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00019938297119685054, | |
| "loss": 0.9824, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.3308457711442787, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 0.00019933387894347036, | |
| "loss": 0.9778, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.3432835820895521, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 0.00019928291428703262, | |
| "loss": 0.9842, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.355721393034826, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00019923007818820035, | |
| "loss": 0.9869, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.3681592039800994, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00019917537164291244, | |
| "loss": 0.9649, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.3805970149253732, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.0001991187956823649, | |
| "loss": 0.9673, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.3930348258706466, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.0001990603513729915, | |
| "loss": 0.9646, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.4054726368159205, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.0001990000398164436, | |
| "loss": 0.9674, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.417910447761194, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00019893786214956945, | |
| "loss": 0.9789, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.4303482587064678, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00019887381954439263, | |
| "loss": 0.9498, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.4427860696517412, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.0001988079132080901, | |
| "loss": 0.9655, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.455223880597015, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00019874014438296946, | |
| "loss": 0.9689, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.4676616915422884, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.0001986705143464453, | |
| "loss": 0.9712, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.4800995024875623, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.0001985990244110154, | |
| "loss": 0.9733, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.0001985256759242359, | |
| "loss": 0.9574, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5049751243781095, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00019845047026869578, | |
| "loss": 0.9744, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.517412935323383, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00019837340886199096, | |
| "loss": 0.9604, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.5298507462686568, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00019829449315669748, | |
| "loss": 0.9593, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.5422885572139302, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00019821372464034416, | |
| "loss": 0.9551, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.554726368159204, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00019813110483538459, | |
| "loss": 0.9552, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.5671641791044775, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00019804663529916826, | |
| "loss": 0.9643, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.5796019900497513, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00019796031762391153, | |
| "loss": 0.9589, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.5920398009950247, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00019787215343666732, | |
| "loss": 0.9584, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.6044776119402986, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 0.00019778214439929452, | |
| "loss": 0.955, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.616915422885572, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00019769029220842677, | |
| "loss": 0.9535, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.6293532338308458, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00019759659859544036, | |
| "loss": 0.9579, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.6417910447761193, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.0001975010653264216, | |
| "loss": 0.9551, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.654228855721393, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 0.00019740369420213365, | |
| "loss": 0.9512, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 0.9429, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.6791044776119404, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.00019720344576398203, | |
| "loss": 0.9533, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.6915422885572138, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00019710057222471967, | |
| "loss": 0.9549, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.7039800995024876, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00019699586837931956, | |
| "loss": 0.9522, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.716417910447761, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00019688933620140637, | |
| "loss": 0.9519, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.728855721393035, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.0001967809776990682, | |
| "loss": 0.9496, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.7412935323383083, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.0001966707949148186, | |
| "loss": 0.9427, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.7537313432835822, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.000196558789925558, | |
| "loss": 0.9475, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.7661691542288556, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 0.00019644496484253474, | |
| "loss": 0.9551, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.7786069651741294, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00019632932181130514, | |
| "loss": 0.949, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.7910447761194028, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 0.00019621186301169315, | |
| "loss": 0.9548, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.8034825870646767, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00019609259065774914, | |
| "loss": 0.9601, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.81592039800995, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.00019597150699770835, | |
| "loss": 0.9453, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.828358208955224, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00019584861431394822, | |
| "loss": 0.9443, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.8407960199004973, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.0001957239149229458, | |
| "loss": 0.9537, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.8532338308457712, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 0.00019559741117523358, | |
| "loss": 0.9455, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.8656716417910446, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00019546910545535558, | |
| "loss": 0.9447, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.8781094527363185, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00019533900018182222, | |
| "loss": 0.953, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.890547263681592, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00019520709780706486, | |
| "loss": 0.9374, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.9029850746268657, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.0001950734008173893, | |
| "loss": 0.9484, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.9154228855721394, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00019493791173292923, | |
| "loss": 0.9399, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.927860696517413, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.0001948006331075986, | |
| "loss": 0.9505, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.9402985074626866, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00019466156752904343, | |
| "loss": 0.9431, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.9527363184079602, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 0.00019452071761859308, | |
| "loss": 0.9502, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.9651741293532339, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00019437808603121087, | |
| "loss": 0.93, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.9776119402985075, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00019423367545544395, | |
| "loss": 0.9397, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.9900497512437811, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00019408748861337273, | |
| "loss": 0.9424, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.460439682006836, | |
| "eval_runtime": 0.5385, | |
| "eval_samples_per_second": 18.57, | |
| "eval_steps_per_second": 1.857, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 2.0024875621890548, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.0001939395282605595, | |
| "loss": 0.9394, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 2.014925373134328, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00019378979718599645, | |
| "loss": 0.9134, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.027363184079602, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00019363829821205327, | |
| "loss": 0.9059, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 2.0398009950248754, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001934850341944237, | |
| "loss": 0.9106, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.0522388059701493, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.0001933300080220719, | |
| "loss": 0.9131, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.0646766169154227, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00019317322261717794, | |
| "loss": 0.8934, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.0771144278606966, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00019301468093508266, | |
| "loss": 0.912, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.08955223880597, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00019285438596423204, | |
| "loss": 0.9088, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.101990049751244, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00019269234072612082, | |
| "loss": 0.92, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.1144278606965172, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00019252854827523557, | |
| "loss": 0.9265, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.126865671641791, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 0.00019236301169899713, | |
| "loss": 0.9119, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.1393034825870645, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00019219573411770235, | |
| "loss": 0.9087, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.1517412935323383, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00019202671868446533, | |
| "loss": 0.9137, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.1641791044776117, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.000191855968585158, | |
| "loss": 0.9102, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.1766169154228856, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00019168348703834997, | |
| "loss": 0.9104, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.189054726368159, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.000191509277295248, | |
| "loss": 0.9011, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.201492537313433, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001913333426396346, | |
| "loss": 0.9154, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 2.2139303482587063, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00019115568638780622, | |
| "loss": 0.9139, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.22636815920398, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.0001909763118885106, | |
| "loss": 0.9109, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 2.2388059701492535, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00019079522252288386, | |
| "loss": 0.9054, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.2512437810945274, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001906124217043866, | |
| "loss": 0.9115, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 2.2636815920398012, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00019042791287873957, | |
| "loss": 0.9189, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.2761194029850746, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.00019024169952385885, | |
| "loss": 0.9114, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 2.288557213930348, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00019005378514979008, | |
| "loss": 0.9104, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.300995024875622, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00018986417329864254, | |
| "loss": 0.9075, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.3134328358208958, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00018967286754452214, | |
| "loss": 0.9053, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.325870646766169, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.0001894798714934643, | |
| "loss": 0.9162, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.3383084577114426, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.0001892851887833657, | |
| "loss": 0.9048, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.3507462686567164, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00018908882308391598, | |
| "loss": 0.9014, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.3631840796019903, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.0001888907780965284, | |
| "loss": 0.9147, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.3756218905472637, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00018869105755427005, | |
| "loss": 0.9187, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.388059701492537, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00018848966522179168, | |
| "loss": 0.9055, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.400497512437811, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00018828660489525646, | |
| "loss": 0.9199, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.412935323383085, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00018808188040226868, | |
| "loss": 0.9103, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.425373134328358, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.0001878754956018014, | |
| "loss": 0.9107, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.4378109452736316, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00018766745438412384, | |
| "loss": 0.8977, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.4502487562189055, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.000187457760670728, | |
| "loss": 0.9016, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.4626865671641793, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.00018724641841425478, | |
| "loss": 0.9022, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.4751243781094527, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00018703343159841943, | |
| "loss": 0.8932, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.487562189054726, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00018681880423793642, | |
| "loss": 0.9083, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 0.9119, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.512437810945274, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00018638464409642723, | |
| "loss": 0.8991, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.5248756218905473, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00018616511949914245, | |
| "loss": 0.9094, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.5373134328358207, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00018594397072453856, | |
| "loss": 0.9042, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.5497512437810945, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.0001857212019411796, | |
| "loss": 0.9043, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.5621890547263684, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00018549681734816623, | |
| "loss": 0.8897, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.574626865671642, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.0001852708211750564, | |
| "loss": 0.905, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.587064676616915, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.0001850432176817857, | |
| "loss": 0.898, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.599502487562189, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.00018481401115858702, | |
| "loss": 0.9063, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.611940298507463, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00018458320592590975, | |
| "loss": 0.9146, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.6243781094527363, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00018435080633433826, | |
| "loss": 0.9005, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.6368159203980097, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00018411681676450999, | |
| "loss": 0.9003, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.6492537313432836, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00018388124162703275, | |
| "loss": 0.9021, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.6616915422885574, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.0001836440853624017, | |
| "loss": 0.9048, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.674129353233831, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001834053524409156, | |
| "loss": 0.9039, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.6865671641791042, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00018316504736259255, | |
| "loss": 0.8942, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.699004975124378, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00018292317465708514, | |
| "loss": 0.8977, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.711442786069652, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00018267973888359509, | |
| "loss": 0.9015, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.7238805970149254, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00018243474463078735, | |
| "loss": 0.9046, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.7363184079601988, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00018218819651670356, | |
| "loss": 0.9004, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.7487562189054726, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00018194009918867497, | |
| "loss": 0.8979, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.7611940298507465, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00018169045732323492, | |
| "loss": 0.8932, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.77363184079602, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00018143927562603066, | |
| "loss": 0.9046, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.7860696517412933, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00018118655883173456, | |
| "loss": 0.9004, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.798507462686567, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00018093231170395507, | |
| "loss": 0.8965, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.810945273631841, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001806765390351467, | |
| "loss": 0.8965, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.8233830845771144, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00018041924564651983, | |
| "loss": 0.9044, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.835820895522388, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00018016043638794974, | |
| "loss": 0.8979, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.8482587064676617, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00017990011613788531, | |
| "loss": 0.8951, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.8606965174129355, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00017963828980325697, | |
| "loss": 0.8963, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.873134328358209, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.0001793749623193842, | |
| "loss": 0.8894, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.8855721393034823, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00017911013864988252, | |
| "loss": 0.9034, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.898009950248756, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00017884382378657, | |
| "loss": 0.9033, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.91044776119403, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00017857602274937308, | |
| "loss": 0.8922, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.9228855721393034, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00017830674058623194, | |
| "loss": 0.8979, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.935323383084577, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00017803598237300537, | |
| "loss": 0.9076, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.9477611940298507, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00017776375321337521, | |
| "loss": 0.8988, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.9601990049751246, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00017749005823874988, | |
| "loss": 0.896, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.972636815920398, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.00017721490260816798, | |
| "loss": 0.9036, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00017693829150820068, | |
| "loss": 0.8985, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.9975124378109452, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.0001766602301528544, | |
| "loss": 0.8955, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.5064175128936768, | |
| "eval_runtime": 0.5358, | |
| "eval_samples_per_second": 18.664, | |
| "eval_steps_per_second": 1.866, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 3.009950248756219, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00017638072378347203, | |
| "loss": 0.8619, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.0223880597014925, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00017609977766863458, | |
| "loss": 0.8565, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 3.0348258706467663, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001758173971040616, | |
| "loss": 0.8522, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.0472636815920398, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00017553358741251133, | |
| "loss": 0.8667, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 3.0597014925373136, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00017524835394368065, | |
| "loss": 0.859, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.072139303482587, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.000174961702074104, | |
| "loss": 0.8619, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 3.084577114427861, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00017467363720705204, | |
| "loss": 0.8649, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.0970149253731343, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001743841647724299, | |
| "loss": 0.875, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 3.109452736318408, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.0001740932902266747, | |
| "loss": 0.8754, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.1218905472636815, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00017380101905265297, | |
| "loss": 0.8579, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 3.1343283582089554, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00017350735675955697, | |
| "loss": 0.8606, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.146766169154229, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.000173212308882801, | |
| "loss": 0.8697, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 3.1592039800995027, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.000172915880983917, | |
| "loss": 0.8682, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.171641791044776, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00017261807865044993, | |
| "loss": 0.8622, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 3.18407960199005, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.0001723189074958521, | |
| "loss": 0.8543, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 3.1965174129353233, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00017201837315937766, | |
| "loss": 0.8555, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 3.208955223880597, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00017171648130597612, | |
| "loss": 0.8612, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.2213930348258706, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.0001714132376261856, | |
| "loss": 0.8644, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 3.2338308457711444, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.0001711086478360257, | |
| "loss": 0.87, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.246268656716418, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00017080271767688963, | |
| "loss": 0.8591, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 3.2587064676616917, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.0001704954529154359, | |
| "loss": 0.8527, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 3.271144278606965, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001701868593434798, | |
| "loss": 0.8787, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 3.283582089552239, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00016987694277788417, | |
| "loss": 0.863, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 3.2960199004975124, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00016956570906044975, | |
| "loss": 0.8707, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 3.308457711442786, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.000169253164057805, | |
| "loss": 0.863, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 3.3208955223880596, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.00016893931366129562, | |
| "loss": 0.8546, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 0.8623, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 3.345771144278607, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00016830772037498476, | |
| "loss": 0.8629, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 3.3582089552238807, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.00016798998939045895, | |
| "loss": 0.8782, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.370646766169154, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001676709768223951, | |
| "loss": 0.8707, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 3.383084577114428, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00016735068868404998, | |
| "loss": 0.8614, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 3.3955223880597014, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.0001670291310127242, | |
| "loss": 0.8691, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 3.4079601990049753, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001667063098696485, | |
| "loss": 0.8771, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 3.4203980099502487, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.0001663822313398697, | |
| "loss": 0.8536, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 3.4328358208955225, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.0001660569015321357, | |
| "loss": 0.8815, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 3.445273631840796, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00016573032657878057, | |
| "loss": 0.8617, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 3.45771144278607, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00016540251263560878, | |
| "loss": 0.8634, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 3.470149253731343, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00016507346588177935, | |
| "loss": 0.8643, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 3.482587064676617, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00016474319251968923, | |
| "loss": 0.8707, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.4950248756218905, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.0001644116987748564, | |
| "loss": 0.8658, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 3.5074626865671643, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00016407899089580262, | |
| "loss": 0.8724, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.5199004975124377, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00016374507515393556, | |
| "loss": 0.8607, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 3.5323383084577116, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001634099578434306, | |
| "loss": 0.862, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.544776119402985, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.0001630736452811122, | |
| "loss": 0.8544, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 3.557213930348259, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00016273614380633484, | |
| "loss": 0.8595, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.5696517412935322, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00016239745978086353, | |
| "loss": 0.8764, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 3.582089552238806, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.0001620575995887538, | |
| "loss": 0.8787, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.5945273631840795, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00016171656963623156, | |
| "loss": 0.8673, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 3.6069651741293534, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00016137437635157213, | |
| "loss": 0.8595, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.6194029850746268, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00016103102618497922, | |
| "loss": 0.8675, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 3.6318407960199006, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00016068652560846327, | |
| "loss": 0.8521, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.644278606965174, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00016034088111571942, | |
| "loss": 0.8648, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 3.656716417910448, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.0001599940992220053, | |
| "loss": 0.8657, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.6691542288557213, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.000159646186464018, | |
| "loss": 0.8537, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 3.681592039800995, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001592971493997709, | |
| "loss": 0.8603, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.6940298507462686, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00015894699460847016, | |
| "loss": 0.8632, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 3.7064676616915424, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 0.00015859572869039064, | |
| "loss": 0.8675, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.718905472636816, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00015824335826675144, | |
| "loss": 0.8725, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 3.7313432835820897, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00015788988997959114, | |
| "loss": 0.871, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.743781094527363, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.0001575353304916427, | |
| "loss": 0.8678, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 3.756218905472637, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00015717968648620764, | |
| "loss": 0.8633, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.7686567164179103, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00015682296466703024, | |
| "loss": 0.863, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 3.781094527363184, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00015646517175817114, | |
| "loss": 0.8584, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.7935323383084576, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00015610631450388058, | |
| "loss": 0.8671, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 3.8059701492537314, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00015574639966847126, | |
| "loss": 0.859, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.818407960199005, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00015538543403619088, | |
| "loss": 0.8604, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 3.8308457711442787, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00015502342441109422, | |
| "loss": 0.8583, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.843283582089552, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.0001546603776169149, | |
| "loss": 0.8666, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 3.855721393034826, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00015429630049693674, | |
| "loss": 0.863, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.8681592039800994, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0001539311999138648, | |
| "loss": 0.8497, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 3.8805970149253732, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.00015356508274969594, | |
| "loss": 0.8589, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.8930348258706466, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00015319795590558924, | |
| "loss": 0.8689, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 3.9054726368159205, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00015282982630173585, | |
| "loss": 0.8636, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.917910447761194, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00015246070087722845, | |
| "loss": 0.8505, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 3.9303482587064678, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00015209058658993056, | |
| "loss": 0.8645, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.942786069651741, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00015171949041634534, | |
| "loss": 0.8702, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 3.955223880597015, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.0001513474193514842, | |
| "loss": 0.8673, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 3.9676616915422884, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00015097438040873474, | |
| "loss": 0.8536, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 3.9800995024875623, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00015060038061972874, | |
| "loss": 0.8702, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.9925373134328357, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.0001502254270342095, | |
| "loss": 0.8659, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.5305564403533936, | |
| "eval_runtime": 0.5393, | |
| "eval_samples_per_second": 18.543, | |
| "eval_steps_per_second": 1.854, | |
| "step": 1608 | |
| }, | |
| { | |
| "epoch": 4.0049751243781095, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.000149849526719899, | |
| "loss": 0.8534, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 4.017412935323383, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00014947268676236464, | |
| "loss": 0.8268, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 4.029850746268656, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00014909491426488578, | |
| "loss": 0.8281, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 4.04228855721393, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00014871621634831978, | |
| "loss": 0.8173, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 4.054726368159204, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00014833660015096766, | |
| "loss": 0.827, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 4.067164179104478, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.0001479560728284398, | |
| "loss": 0.8284, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 4.079601990049751, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.00014757464155352082, | |
| "loss": 0.8349, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.092039800995025, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.0001471923135160345, | |
| "loss": 0.8305, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 4.104477611940299, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.0001468090959227082, | |
| "loss": 0.8251, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 4.116915422885572, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00014642499599703716, | |
| "loss": 0.83, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 4.129353233830845, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00014604002097914806, | |
| "loss": 0.8371, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 4.141791044776119, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00014565417812566285, | |
| "loss": 0.8292, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 4.154228855721393, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00014526747470956176, | |
| "loss": 0.8341, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 4.166666666666667, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00014487991802004623, | |
| "loss": 0.826, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 4.17910447761194, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00014449151536240166, | |
| "loss": 0.8325, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 4.191542288557214, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00014410227405785953, | |
| "loss": 0.8234, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 4.203980099502488, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 0.00014371220144345954, | |
| "loss": 0.8345, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 4.2164179104477615, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001433213048719111, | |
| "loss": 0.8337, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 4.2288557213930345, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.0001429295917114551, | |
| "loss": 0.835, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.241293532338308, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001425370693457245, | |
| "loss": 0.8275, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 4.253731343283582, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00014214374517360575, | |
| "loss": 0.8282, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 4.266169154228856, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.00014174962660909888, | |
| "loss": 0.8332, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 4.278606965174129, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00014135472108117787, | |
| "loss": 0.8195, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 4.291044776119403, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00014095903603365066, | |
| "loss": 0.836, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 4.303482587064677, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00014056257892501885, | |
| "loss": 0.832, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 4.3159203980099505, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00014016535722833704, | |
| "loss": 0.8306, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 4.3283582089552235, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00013976737843107202, | |
| "loss": 0.8147, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 4.340796019900497, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.0001393686500349616, | |
| "loss": 0.8281, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 4.353233830845771, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.00013896917955587328, | |
| "loss": 0.8313, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.365671641791045, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00013856897452366242, | |
| "loss": 0.8236, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 4.378109452736318, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00013816804248203052, | |
| "loss": 0.8245, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 4.390547263681592, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00013776639098838283, | |
| "loss": 0.8226, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 4.402985074626866, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00013736402761368598, | |
| "loss": 0.8372, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 4.41542288557214, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00013696095994232533, | |
| "loss": 0.8293, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 4.4278606965174125, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00013655719557196185, | |
| "loss": 0.8375, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 4.440298507462686, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00013615274211338909, | |
| "loss": 0.8328, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 4.45273631840796, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.0001357476071903896, | |
| "loss": 0.8338, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 4.465174129353234, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.0001353417984395912, | |
| "loss": 0.8341, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 4.477611940298507, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.0001349353235103232, | |
| "loss": 0.8253, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.490049751243781, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00013452819006447202, | |
| "loss": 0.8337, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 4.502487562189055, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00013412040577633687, | |
| "loss": 0.832, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 4.514925373134329, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00013371197833248507, | |
| "loss": 0.8288, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 4.5273631840796025, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.0001333029154316072, | |
| "loss": 0.8239, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 4.539800995024875, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00013289322478437184, | |
| "loss": 0.8305, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 4.552238805970149, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00013248291411328047, | |
| "loss": 0.8243, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 4.564676616915423, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00013207199115252167, | |
| "loss": 0.8335, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 4.577114427860696, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00013166046364782545, | |
| "loss": 0.8273, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 4.58955223880597, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00013124833935631726, | |
| "loss": 0.8261, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 4.601990049751244, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.0001308356260463717, | |
| "loss": 0.8321, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.614427860696518, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00013042233149746613, | |
| "loss": 0.8362, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 4.6268656716417915, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.0001300084635000341, | |
| "loss": 0.8276, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 4.6393034825870645, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00012959402985531824, | |
| "loss": 0.8356, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 4.651741293532338, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.0001291790383752237, | |
| "loss": 0.835, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 4.664179104477612, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.0001287634968821703, | |
| "loss": 0.8327, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 4.676616915422885, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00012834741320894553, | |
| "loss": 0.8333, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.689054726368159, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00012793079519855675, | |
| "loss": 0.8299, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 4.701492537313433, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00012751365070408333, | |
| "loss": 0.8293, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.713930348258707, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.0001270959875885287, | |
| "loss": 0.831, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 4.726368159203981, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00012667781372467202, | |
| "loss": 0.8322, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.7388059701492535, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00012625913699491984, | |
| "loss": 0.8323, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 4.751243781094527, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00012583996529115762, | |
| "loss": 0.8352, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.763681592039801, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00012542030651460073, | |
| "loss": 0.8414, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 4.776119402985074, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00012500016857564585, | |
| "loss": 0.8282, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.788557213930348, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.0001245795593937215, | |
| "loss": 0.8353, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 4.800995024875622, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00012415848689713903, | |
| "loss": 0.8272, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.813432835820896, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00012373695902294314, | |
| "loss": 0.8346, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 4.82587064676617, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00012331498371676204, | |
| "loss": 0.8274, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.838308457711443, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.00012289256893265803, | |
| "loss": 0.8397, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 4.850746268656716, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001224697226329772, | |
| "loss": 0.8325, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.86318407960199, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00012204645278819973, | |
| "loss": 0.8394, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 4.875621890547263, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00012162276737678933, | |
| "loss": 0.8313, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.888059701492537, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00012119867438504302, | |
| "loss": 0.8309, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 4.900497512437811, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.0001207741818069405, | |
| "loss": 0.8267, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 4.912935323383085, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00012034929764399357, | |
| "loss": 0.8321, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 4.925373134328359, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.00011992402990509515, | |
| "loss": 0.827, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.937810945273632, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00011949838660636857, | |
| "loss": 0.832, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 4.9502487562189055, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00011907237577101611, | |
| "loss": 0.8235, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 4.962686567164179, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00011864600542916813, | |
| "loss": 0.8248, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 4.975124378109452, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00011821928361773147, | |
| "loss": 0.8284, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.987562189054726, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00011779221838023798, | |
| "loss": 0.8266, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.8359, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.570625066757202, | |
| "eval_runtime": 0.5336, | |
| "eval_samples_per_second": 18.742, | |
| "eval_steps_per_second": 1.874, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 5.012437810945274, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00011693708983342371, | |
| "loss": 0.793, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 5.024875621890548, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00011650904264292687, | |
| "loss": 0.8034, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 5.037313432835821, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001160806842637173, | |
| "loss": 0.802, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 5.0497512437810945, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.00011565202277017551, | |
| "loss": 0.8011, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 5.062189054726368, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00011522306624239558, | |
| "loss": 0.7894, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 5.074626865671641, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.000114793822766033, | |
| "loss": 0.7972, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 5.087064676616915, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00011436430043215197, | |
| "loss": 0.8035, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 5.099502487562189, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00011393450733707309, | |
| "loss": 0.7885, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.111940298507463, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00011350445158222074, | |
| "loss": 0.8039, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 5.124378109452737, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00011307414127397027, | |
| "loss": 0.7873, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 5.13681592039801, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00011264358452349528, | |
| "loss": 0.8081, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 5.149253731343284, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00011221278944661473, | |
| "loss": 0.8095, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 5.161691542288557, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00011178176416363992, | |
| "loss": 0.8008, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 5.174129353233831, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00011135051679922141, | |
| "loss": 0.7983, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 5.186567164179104, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00011091905548219597, | |
| "loss": 0.7977, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 5.199004975124378, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00011048738834543319, | |
| "loss": 0.8033, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 5.211442786069652, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00011005552352568238, | |
| "loss": 0.8081, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 5.223880597014926, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00010962346916341903, | |
| "loss": 0.8068, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.236318407960199, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00010919123340269147, | |
| "loss": 0.7944, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 5.248756218905473, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00010875882439096729, | |
| "loss": 0.798, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 5.2611940298507465, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00010832625027897978, | |
| "loss": 0.802, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 5.273631840796019, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00010789351922057435, | |
| "loss": 0.8025, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 5.286069651741293, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00010746063937255473, | |
| "loss": 0.7957, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 5.298507462686567, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.0001070276188945293, | |
| "loss": 0.8078, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 5.310945273631841, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00010659446594875728, | |
| "loss": 0.8031, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 5.323383084577115, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00010616118869999483, | |
| "loss": 0.799, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 5.335820895522388, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00010572779531534112, | |
| "loss": 0.8115, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 5.348258706467662, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00010529429396408452, | |
| "loss": 0.8075, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 5.3606965174129355, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00010486069281754854, | |
| "loss": 0.8052, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 5.373134328358209, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00010442700004893764, | |
| "loss": 0.796, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 5.385572139303482, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.0001039932238331835, | |
| "loss": 0.8113, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 5.398009950248756, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00010355937234679065, | |
| "loss": 0.7993, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 5.41044776119403, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00010312545376768246, | |
| "loss": 0.8087, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 5.422885572139304, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00010269147627504692, | |
| "loss": 0.8035, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 5.435323383084577, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.0001022574480491826, | |
| "loss": 0.8029, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 5.447761194029851, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.0001018233772713443, | |
| "loss": 0.8011, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 5.460199004975125, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00010138927212358902, | |
| "loss": 0.7929, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 5.472636815920398, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00010095514078862147, | |
| "loss": 0.8009, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.485074626865671, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00010052099144964004, | |
| "loss": 0.7995, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 5.497512437810945, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00010008683229018256, | |
| "loss": 0.8006, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 5.509950248756219, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.965267149397182e-05, | |
| "loss": 0.8061, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 5.522388059701493, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.92185172447616e-05, | |
| "loss": 0.8014, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 5.534825870646766, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.878437772618226e-05, | |
| "loss": 0.7934, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 5.54726368159204, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 9.835026112158637e-05, | |
| "loss": 0.8084, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 5.559701492537314, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 9.791617561389475e-05, | |
| "loss": 0.8055, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 5.572139303482587, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.74821293854419e-05, | |
| "loss": 0.7976, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 5.58457711442786, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 9.704813061782202e-05, | |
| "loss": 0.7975, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 5.597014925373134, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 9.661418749173467e-05, | |
| "loss": 0.8062, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 5.609452736318408, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.618030818683062e-05, | |
| "loss": 0.8017, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 5.621890547263682, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.574650088155752e-05, | |
| "loss": 0.7992, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 5.634328358208955, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.531277375300599e-05, | |
| "loss": 0.8055, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 5.646766169154229, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.487913497675536e-05, | |
| "loss": 0.8082, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 5.659203980099503, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.444559272671939e-05, | |
| "loss": 0.796, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 5.6716417910447765, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 9.40121551749925e-05, | |
| "loss": 0.7999, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 5.6840796019900495, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.357883049169559e-05, | |
| "loss": 0.8063, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 5.696517412935323, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 9.314562684482202e-05, | |
| "loss": 0.813, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 5.708955223880597, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.271255240008356e-05, | |
| "loss": 0.8053, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 5.721393034825871, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.227961532075671e-05, | |
| "loss": 0.8058, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.733830845771144, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 9.184682376752864e-05, | |
| "loss": 0.8085, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 5.746268656716418, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 9.141418589834339e-05, | |
| "loss": 0.8022, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 5.758706467661692, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.098170986824818e-05, | |
| "loss": 0.7961, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 5.7711442786069655, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.054940382923953e-05, | |
| "loss": 0.8097, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 5.7835820895522385, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.011727593010981e-05, | |
| "loss": 0.8017, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 5.796019900497512, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 8.96853343162934e-05, | |
| "loss": 0.8002, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 5.808457711442786, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 8.92535871297134e-05, | |
| "loss": 0.7919, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 5.82089552238806, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 8.882204250862796e-05, | |
| "loss": 0.7906, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 5.833333333333333, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 8.839070858747697e-05, | |
| "loss": 0.7957, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.845771144278607, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 8.795959349672878e-05, | |
| "loss": 0.8007, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 5.858208955223881, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 8.752870536272673e-05, | |
| "loss": 0.7986, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 5.870646766169155, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 8.709805230753627e-05, | |
| "loss": 0.8033, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 5.883084577114428, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 8.66676424487916e-05, | |
| "loss": 0.8091, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 5.895522388059701, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 8.623748389954283e-05, | |
| "loss": 0.8039, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 5.907960199004975, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 8.580758476810295e-05, | |
| "loss": 0.8074, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 5.920398009950249, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 8.537795315789509e-05, | |
| "loss": 0.7956, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 5.932835820895522, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 8.494859716729962e-05, | |
| "loss": 0.7965, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 5.945273631840796, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 8.451952488950166e-05, | |
| "loss": 0.8017, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 5.95771144278607, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 8.40907444123385e-05, | |
| "loss": 0.8008, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 5.970149253731344, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 8.366226381814697e-05, | |
| "loss": 0.8091, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 5.982587064676617, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 8.323409118361141e-05, | |
| "loss": 0.8097, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 5.9950248756218905, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 8.280623457961102e-05, | |
| "loss": 0.7986, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.619605779647827, | |
| "eval_runtime": 0.5379, | |
| "eval_samples_per_second": 18.589, | |
| "eval_steps_per_second": 1.859, | |
| "step": 2412 | |
| }, | |
| { | |
| "epoch": 6.007462686567164, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 8.237870207106821e-05, | |
| "loss": 0.7835, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 6.019900497512438, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 8.195150171679608e-05, | |
| "loss": 0.7707, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 6.032338308457711, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 8.152464156934695e-05, | |
| "loss": 0.778, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 6.044776119402985, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 8.109812967486025e-05, | |
| "loss": 0.7807, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 6.057213930348259, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 8.067197407291105e-05, | |
| "loss": 0.7784, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 6.069651741293533, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 8.02461827963585e-05, | |
| "loss": 0.7723, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 6.082089552238806, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 7.982076387119421e-05, | |
| "loss": 0.7758, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 6.0945273631840795, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 7.939572531639128e-05, | |
| "loss": 0.7792, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 6.106965174129353, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 7.897107514375283e-05, | |
| "loss": 0.7772, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 6.119402985074627, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 7.854682135776131e-05, | |
| "loss": 0.7751, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.1318407960199, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 7.812297195542728e-05, | |
| "loss": 0.7726, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 6.144278606965174, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 7.769953492613899e-05, | |
| "loss": 0.7847, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 6.156716417910448, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 7.727651825151145e-05, | |
| "loss": 0.7762, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 6.169154228855722, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 7.685392990523626e-05, | |
| "loss": 0.7675, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 6.181592039800995, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 7.643177785293128e-05, | |
| "loss": 0.7764, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 6.1940298507462686, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 7.601007005199021e-05, | |
| "loss": 0.7794, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 6.206467661691542, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 7.558881445143301e-05, | |
| "loss": 0.7846, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 6.218905472636816, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 7.516801899175565e-05, | |
| "loss": 0.7809, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.231343283582089, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 7.474769160478079e-05, | |
| "loss": 0.7651, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 6.243781094527363, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 7.432784021350796e-05, | |
| "loss": 0.773, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 6.256218905472637, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 7.390847273196455e-05, | |
| "loss": 0.7799, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 6.268656716417911, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 7.348959706505626e-05, | |
| "loss": 0.7782, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 6.281094527363184, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 7.307122110841843e-05, | |
| "loss": 0.7746, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 6.293532338308458, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 7.265335274826704e-05, | |
| "loss": 0.7701, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 6.3059701492537314, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 7.223599986124994e-05, | |
| "loss": 0.7794, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 6.318407960199005, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 7.181917031429874e-05, | |
| "loss": 0.7777, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 6.330845771144278, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 7.140287196448016e-05, | |
| "loss": 0.7797, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 6.343283582089552, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 7.09871126588481e-05, | |
| "loss": 0.7765, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 6.355721393034826, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 7.057190023429569e-05, | |
| "loss": 0.7759, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 6.3681592039801, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 7.015724251740766e-05, | |
| "loss": 0.7805, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 6.380597014925373, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.974314732431258e-05, | |
| "loss": 0.7805, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 6.393034825870647, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 6.932962246053577e-05, | |
| "loss": 0.7768, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 6.4054726368159205, | |
| "grad_norm": 0.375, | |
| "learning_rate": 6.891667572085209e-05, | |
| "loss": 0.774, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 6.417910447761194, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 6.850431488913895e-05, | |
| "loss": 0.7724, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 6.430348258706467, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 6.80925477382297e-05, | |
| "loss": 0.7768, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 6.442786069651741, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 6.76813820297669e-05, | |
| "loss": 0.7893, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 6.455223880597015, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 6.727082551405636e-05, | |
| "loss": 0.7737, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 6.467661691542289, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 6.686088592992067e-05, | |
| "loss": 0.7768, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.480099502487562, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 6.645157100455368e-05, | |
| "loss": 0.7715, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 6.492537313432836, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 6.604288845337453e-05, | |
| "loss": 0.7822, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 6.5049751243781095, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.563484597988248e-05, | |
| "loss": 0.7779, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 6.517412935323383, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 6.522745127551158e-05, | |
| "loss": 0.7774, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 6.529850746268656, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 6.482071201948556e-05, | |
| "loss": 0.775, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 6.54228855721393, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 6.44146358786734e-05, | |
| "loss": 0.79, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 6.554726368159204, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.400923050744449e-05, | |
| "loss": 0.7681, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.567164179104478, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 6.360450354752458e-05, | |
| "loss": 0.7785, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 6.579601990049751, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 6.320046262785156e-05, | |
| "loss": 0.7856, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 6.592039800995025, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 6.279711536443185e-05, | |
| "loss": 0.7767, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 6.604477611940299, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 6.239446936019657e-05, | |
| "loss": 0.7798, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 6.616915422885572, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 6.199253220485856e-05, | |
| "loss": 0.7742, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 6.629353233830845, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 6.159131147476907e-05, | |
| "loss": 0.7827, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 6.641791044776119, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 6.119081473277501e-05, | |
| "loss": 0.7897, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 6.654228855721393, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 6.0791049528076506e-05, | |
| "loss": 0.7689, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 0.772, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 6.67910447761194, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 5.99937438582782e-05, | |
| "loss": 0.7799, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 6.691542288557214, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 5.959621842206474e-05, | |
| "loss": 0.7853, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 6.703980099502488, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 5.919945458063614e-05, | |
| "loss": 0.7758, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 6.7164179104477615, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 5.880345981282876e-05, | |
| "loss": 0.7791, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.7288557213930345, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 5.8408241582982304e-05, | |
| "loss": 0.7793, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 6.741293532338308, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 5.801380734079907e-05, | |
| "loss": 0.7726, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 6.753731343283582, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 5.7620164521203356e-05, | |
| "loss": 0.783, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 6.766169154228856, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 5.722732054420172e-05, | |
| "loss": 0.7808, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 6.778606965174129, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 5.683528281474255e-05, | |
| "loss": 0.7766, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 6.791044776119403, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 5.6444058722577165e-05, | |
| "loss": 0.7773, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 6.803482587064677, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 5.605365564211975e-05, | |
| "loss": 0.7806, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 6.8159203980099505, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 5.566408093230911e-05, | |
| "loss": 0.7776, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 6.8283582089552235, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 5.5275341936469396e-05, | |
| "loss": 0.7886, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 6.840796019900497, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 5.4887445982171906e-05, | |
| "loss": 0.7848, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 6.853233830845771, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 5.450040038109713e-05, | |
| "loss": 0.7815, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 6.865671641791045, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 5.4114212428896424e-05, | |
| "loss": 0.7847, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 6.878109452736318, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 5.3728889405055136e-05, | |
| "loss": 0.7825, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 6.890547263681592, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 5.334443857275487e-05, | |
| "loss": 0.7697, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 6.902985074626866, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 5.296086717873685e-05, | |
| "loss": 0.7836, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 6.91542288557214, | |
| "grad_norm": 0.375, | |
| "learning_rate": 5.257818245316522e-05, | |
| "loss": 0.7832, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 6.927860696517413, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 5.2196391609490766e-05, | |
| "loss": 0.7735, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 6.940298507462686, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.1815501844315105e-05, | |
| "loss": 0.7803, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 6.95273631840796, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 5.143552033725464e-05, | |
| "loss": 0.7799, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 6.965174129353234, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.105645425080572e-05, | |
| "loss": 0.7927, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 6.977611940298507, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 5.0678310730209275e-05, | |
| "loss": 0.7959, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 6.990049751243781, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 5.030109690331625e-05, | |
| "loss": 0.7778, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 2.6582648754119873, | |
| "eval_runtime": 0.5405, | |
| "eval_samples_per_second": 18.501, | |
| "eval_steps_per_second": 1.85, | |
| "step": 2814 | |
| }, | |
| { | |
| "epoch": 7.002487562189055, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 4.9924819880453254e-05, | |
| "loss": 0.7746, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 7.014925373134329, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 4.954948675428853e-05, | |
| "loss": 0.7586, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 7.027363184079602, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.917510459969825e-05, | |
| "loss": 0.7669, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 7.039800995024875, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.880168047363312e-05, | |
| "loss": 0.7616, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 7.052238805970149, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.84292214149855e-05, | |
| "loss": 0.7597, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 7.064676616915423, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.8057734444456536e-05, | |
| "loss": 0.7684, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 7.077114427860696, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 4.768722656442394e-05, | |
| "loss": 0.7541, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 7.08955223880597, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 4.7317704758809946e-05, | |
| "loss": 0.7639, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 7.101990049751244, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.694917599294972e-05, | |
| "loss": 0.7582, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 7.114427860696518, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.658164721345998e-05, | |
| "loss": 0.769, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 7.126865671641791, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 4.6215125348108247e-05, | |
| "loss": 0.7611, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 7.1393034825870645, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 4.584961730568188e-05, | |
| "loss": 0.7584, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.151741293532338, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 4.5485129975858346e-05, | |
| "loss": 0.7626, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 7.164179104477612, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.512167022907494e-05, | |
| "loss": 0.7679, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 7.176616915422885, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.475924491639949e-05, | |
| "loss": 0.7628, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 7.189054726368159, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.439786086940115e-05, | |
| "loss": 0.7575, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 7.201492537313433, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.40375249000216e-05, | |
| "loss": 0.7553, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 7.213930348258707, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 4.3678243800446835e-05, | |
| "loss": 0.7764, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 7.22636815920398, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.332002434297875e-05, | |
| "loss": 0.7777, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 7.2388059701492535, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 4.296287327990797e-05, | |
| "loss": 0.7573, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 7.251243781094527, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.260679734338607e-05, | |
| "loss": 0.7528, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 7.263681592039801, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 4.225180324529917e-05, | |
| "loss": 0.7622, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 7.276119402985074, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 4.189789767714102e-05, | |
| "loss": 0.7622, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 7.288557213930348, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.1545087309887045e-05, | |
| "loss": 0.7629, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 7.300995024875622, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 4.11933787938687e-05, | |
| "loss": 0.7546, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 7.313432835820896, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.084277875864776e-05, | |
| "loss": 0.7624, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 7.32587064676617, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.049329381289184e-05, | |
| "loss": 0.7709, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 7.338308457711443, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 4.014493054424944e-05, | |
| "loss": 0.7623, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 7.350746268656716, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 3.979769551922592e-05, | |
| "loss": 0.7528, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 7.36318407960199, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.945159528305971e-05, | |
| "loss": 0.7588, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 7.375621890547263, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 3.91066363595989e-05, | |
| "loss": 0.7581, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 7.388059701492537, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 3.876282525117847e-05, | |
| "loss": 0.7656, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 7.400497512437811, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.8420168438497276e-05, | |
| "loss": 0.762, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 7.412935323383085, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 3.807867238049642e-05, | |
| "loss": 0.7756, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 7.425373134328359, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 3.7738343514237106e-05, | |
| "loss": 0.7609, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 7.437810945273632, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 3.739918825477953e-05, | |
| "loss": 0.767, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 7.4502487562189055, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 3.706121299506183e-05, | |
| "loss": 0.7622, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 7.462686567164179, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.672442410577965e-05, | |
| "loss": 0.7604, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.475124378109452, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 3.638882793526603e-05, | |
| "loss": 0.7588, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 7.487562189054726, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 3.605443080937172e-05, | |
| "loss": 0.7486, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 0.7688, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 7.512437810945274, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 3.5389258881718e-05, | |
| "loss": 0.7663, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 7.524875621890548, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 3.5058496618177797e-05, | |
| "loss": 0.7583, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 7.537313432835821, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.472895847545905e-05, | |
| "loss": 0.7682, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 7.5497512437810945, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 3.44006506652212e-05, | |
| "loss": 0.7597, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 7.562189054726368, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 3.407357937593237e-05, | |
| "loss": 0.7604, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 7.574626865671641, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.374775077275293e-05, | |
| "loss": 0.7605, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 7.587064676616915, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 3.342317099741886e-05, | |
| "loss": 0.7645, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 7.599502487562189, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.309984616812649e-05, | |
| "loss": 0.7667, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 7.611940298507463, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 3.27777823794168e-05, | |
| "loss": 0.765, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 7.624378109452737, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.2456985702060694e-05, | |
| "loss": 0.7558, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 7.63681592039801, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 3.213746218294455e-05, | |
| "loss": 0.7708, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 7.649253731343284, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 3.1819217844956214e-05, | |
| "loss": 0.7598, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 7.661691542288557, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 3.150225868687161e-05, | |
| "loss": 0.7697, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 7.67412935323383, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.118659068324129e-05, | |
| "loss": 0.7646, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 7.686567164179104, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.0872219784278354e-05, | |
| "loss": 0.7634, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 7.699004975124378, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 3.055915191574582e-05, | |
| "loss": 0.7599, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 7.711442786069652, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 3.02473929788452e-05, | |
| "loss": 0.7518, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 7.723880597014926, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.9936948850105152e-05, | |
| "loss": 0.7671, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 7.736318407960199, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 2.96278253812707e-05, | |
| "loss": 0.7726, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 7.748756218905473, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.93200283991931e-05, | |
| "loss": 0.7541, | |
| "step": 3115 | |
| }, | |
| { | |
| "epoch": 7.7611940298507465, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.901356370571967e-05, | |
| "loss": 0.7545, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 7.773631840796019, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 2.870843707758484e-05, | |
| "loss": 0.7667, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 7.786069651741293, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 2.840465426630091e-05, | |
| "loss": 0.7589, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 7.798507462686567, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 2.8102220998049843e-05, | |
| "loss": 0.7588, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 7.810945273631841, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.7801142973575243e-05, | |
| "loss": 0.7628, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 7.823383084577115, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.7501425868074925e-05, | |
| "loss": 0.7645, | |
| "step": 3145 | |
| }, | |
| { | |
| "epoch": 7.835820895522388, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.7203075331094017e-05, | |
| "loss": 0.7703, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 7.848258706467662, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 2.6906096986418218e-05, | |
| "loss": 0.7627, | |
| "step": 3155 | |
| }, | |
| { | |
| "epoch": 7.8606965174129355, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 2.6610496431968125e-05, | |
| "loss": 0.7569, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 7.8731343283582085, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.6316279239693465e-05, | |
| "loss": 0.7695, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 7.885572139303482, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 2.6023450955468176e-05, | |
| "loss": 0.7631, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 7.898009950248756, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 2.5732017098985863e-05, | |
| "loss": 0.768, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 7.91044776119403, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 2.54419831636557e-05, | |
| "loss": 0.7639, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 7.922885572139304, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.5153354616498924e-05, | |
| "loss": 0.758, | |
| "step": 3185 | |
| }, | |
| { | |
| "epoch": 7.935323383084577, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.4866136898045843e-05, | |
| "loss": 0.7612, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 7.947761194029851, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 2.458033542223316e-05, | |
| "loss": 0.7595, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 7.960199004975125, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.4295955576301965e-05, | |
| "loss": 0.7566, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 7.9726368159203975, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.4013002720696253e-05, | |
| "loss": 0.7721, | |
| "step": 3205 | |
| }, | |
| { | |
| "epoch": 7.985074626865671, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.3731482188961818e-05, | |
| "loss": 0.7598, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 7.997512437810945, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 2.3451399287645726e-05, | |
| "loss": 0.7562, | |
| "step": 3215 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.6845827102661133, | |
| "eval_runtime": 0.5378, | |
| "eval_samples_per_second": 18.593, | |
| "eval_steps_per_second": 1.859, | |
| "step": 3216 | |
| }, | |
| { | |
| "epoch": 8.009950248756219, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 2.317275929619627e-05, | |
| "loss": 0.7565, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 8.022388059701493, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 2.28955674668636e-05, | |
| "loss": 0.7484, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 8.034825870646767, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.261982902460039e-05, | |
| "loss": 0.7625, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 8.04726368159204, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 2.2345549166963796e-05, | |
| "loss": 0.7578, | |
| "step": 3235 | |
| }, | |
| { | |
| "epoch": 8.059701492537313, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 2.2072733064017103e-05, | |
| "loss": 0.7537, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 8.072139303482587, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 2.1801385858232437e-05, | |
| "loss": 0.75, | |
| "step": 3245 | |
| }, | |
| { | |
| "epoch": 8.08457711442786, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 2.1531512664393838e-05, | |
| "loss": 0.7514, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 8.097014925373134, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.1263118569500795e-05, | |
| "loss": 0.7536, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 8.109452736318408, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.0996208632672475e-05, | |
| "loss": 0.7608, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 8.121890547263682, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 2.0730787885052105e-05, | |
| "loss": 0.7569, | |
| "step": 3265 | |
| }, | |
| { | |
| "epoch": 8.134328358208956, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.0466861329712473e-05, | |
| "loss": 0.7481, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 8.14676616915423, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 2.020443394156134e-05, | |
| "loss": 0.762, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 8.159203980099502, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.9943510667247813e-05, | |
| "loss": 0.7551, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.171641791044776, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.968409642506913e-05, | |
| "loss": 0.7572, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 8.18407960199005, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.9426196104877735e-05, | |
| "loss": 0.7559, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 8.196517412935323, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.9169814567989465e-05, | |
| "loss": 0.7529, | |
| "step": 3295 | |
| }, | |
| { | |
| "epoch": 8.208955223880597, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 1.89149566470915e-05, | |
| "loss": 0.7524, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 8.221393034825871, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.866162714615165e-05, | |
| "loss": 0.7589, | |
| "step": 3305 | |
| }, | |
| { | |
| "epoch": 8.233830845771145, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.8409830840327546e-05, | |
| "loss": 0.7517, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 8.246268656716419, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.8159572475876718e-05, | |
| "loss": 0.7495, | |
| "step": 3315 | |
| }, | |
| { | |
| "epoch": 8.25870646766169, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.791085677006722e-05, | |
| "loss": 0.7566, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 8.271144278606965, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.7663688411088442e-05, | |
| "loss": 0.7529, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 8.283582089552239, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.741807205796314e-05, | |
| "loss": 0.7449, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 8.296019900497512, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.717401234045918e-05, | |
| "loss": 0.7498, | |
| "step": 3335 | |
| }, | |
| { | |
| "epoch": 8.308457711442786, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 1.6931513859002635e-05, | |
| "loss": 0.7573, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 8.32089552238806, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.6690581184590858e-05, | |
| "loss": 0.7546, | |
| "step": 3345 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.6451218858706374e-05, | |
| "loss": 0.755, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 8.345771144278608, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.6213431393231283e-05, | |
| "loss": 0.7582, | |
| "step": 3355 | |
| }, | |
| { | |
| "epoch": 8.35820895522388, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.5977223270362196e-05, | |
| "loss": 0.7511, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 8.370646766169154, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.574259894252583e-05, | |
| "loss": 0.7506, | |
| "step": 3365 | |
| }, | |
| { | |
| "epoch": 8.383084577114428, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.5509562832294944e-05, | |
| "loss": 0.7572, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 8.395522388059701, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 1.5278119332305053e-05, | |
| "loss": 0.7653, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 8.407960199004975, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 1.5048272805171615e-05, | |
| "loss": 0.7573, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 8.42039800995025, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 1.4820027583407836e-05, | |
| "loss": 0.7536, | |
| "step": 3385 | |
| }, | |
| { | |
| "epoch": 8.432835820895523, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.459338796934293e-05, | |
| "loss": 0.7562, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 8.445273631840797, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.4368358235041102e-05, | |
| "loss": 0.7566, | |
| "step": 3395 | |
| }, | |
| { | |
| "epoch": 8.457711442786069, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 1.4144942622220902e-05, | |
| "loss": 0.7493, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 8.470149253731343, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.3923145342175482e-05, | |
| "loss": 0.7536, | |
| "step": 3405 | |
| }, | |
| { | |
| "epoch": 8.482587064676617, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.3702970575692975e-05, | |
| "loss": 0.7591, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 8.49502487562189, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 1.3484422472977808e-05, | |
| "loss": 0.75, | |
| "step": 3415 | |
| }, | |
| { | |
| "epoch": 8.507462686567164, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.3267505153572501e-05, | |
| "loss": 0.7542, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 8.519900497512438, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.3052222706279938e-05, | |
| "loss": 0.757, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 8.532338308457712, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.2838579189086353e-05, | |
| "loss": 0.753, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 8.544776119402986, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 1.2626578629084784e-05, | |
| "loss": 0.7491, | |
| "step": 3435 | |
| }, | |
| { | |
| "epoch": 8.557213930348258, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.2416225022399286e-05, | |
| "loss": 0.7622, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 8.569651741293532, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.2207522334109368e-05, | |
| "loss": 0.7497, | |
| "step": 3445 | |
| }, | |
| { | |
| "epoch": 8.582089552238806, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.2000474498175552e-05, | |
| "loss": 0.757, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 8.59452736318408, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 1.179508541736497e-05, | |
| "loss": 0.7557, | |
| "step": 3455 | |
| }, | |
| { | |
| "epoch": 8.606965174129353, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.1591358963177923e-05, | |
| "loss": 0.7582, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 8.619402985074627, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.138929897577493e-05, | |
| "loss": 0.76, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 8.631840796019901, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.118890926390419e-05, | |
| "loss": 0.7678, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 8.644278606965175, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.0990193604830013e-05, | |
| "loss": 0.7659, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 8.656716417910447, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.0793155744261351e-05, | |
| "loss": 0.7586, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 8.66915422885572, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.0597799396281482e-05, | |
| "loss": 0.7583, | |
| "step": 3485 | |
| }, | |
| { | |
| "epoch": 8.681592039800995, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 1.0404128243277777e-05, | |
| "loss": 0.7484, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 8.694029850746269, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 1.0212145935872375e-05, | |
| "loss": 0.7496, | |
| "step": 3495 | |
| }, | |
| { | |
| "epoch": 8.706467661691542, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.0021856092853432e-05, | |
| "loss": 0.7563, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.718905472636816, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.833262301106705e-06, | |
| "loss": 0.765, | |
| "step": 3505 | |
| }, | |
| { | |
| "epoch": 8.73134328358209, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 9.646368115548232e-06, | |
| "loss": 0.7648, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 8.743781094527364, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.461177059057069e-06, | |
| "loss": 0.7466, | |
| "step": 3515 | |
| }, | |
| { | |
| "epoch": 8.756218905472636, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.277692622409018e-06, | |
| "loss": 0.7601, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 8.76865671641791, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.095918264210779e-06, | |
| "loss": 0.7509, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 8.781094527363184, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 8.915857410834794e-06, | |
| "loss": 0.7546, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 8.793532338308458, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 8.737513456354617e-06, | |
| "loss": 0.7589, | |
| "step": 3535 | |
| }, | |
| { | |
| "epoch": 8.805970149253731, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 8.56088976248095e-06, | |
| "loss": 0.7593, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 8.818407960199005, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 8.385989658498339e-06, | |
| "loss": 0.7457, | |
| "step": 3545 | |
| }, | |
| { | |
| "epoch": 8.83084577114428, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 8.212816441202309e-06, | |
| "loss": 0.7467, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 8.843283582089553, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 8.041373374837302e-06, | |
| "loss": 0.7541, | |
| "step": 3555 | |
| }, | |
| { | |
| "epoch": 8.855721393034825, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 7.871663691035103e-06, | |
| "loss": 0.7469, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 8.868159203980099, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 7.703690588753965e-06, | |
| "loss": 0.7576, | |
| "step": 3565 | |
| }, | |
| { | |
| "epoch": 8.880597014925373, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 7.53745723421827e-06, | |
| "loss": 0.7444, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 8.893034825870647, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 7.372966760858891e-06, | |
| "loss": 0.7573, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 8.90547263681592, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 7.2102222692540415e-06, | |
| "loss": 0.7625, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 8.917910447761194, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 7.0492268270709875e-06, | |
| "loss": 0.7466, | |
| "step": 3585 | |
| }, | |
| { | |
| "epoch": 8.930348258706468, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 6.889983469008055e-06, | |
| "loss": 0.7584, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 8.942786069651742, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 6.73249519673752e-06, | |
| "loss": 0.7482, | |
| "step": 3595 | |
| }, | |
| { | |
| "epoch": 8.955223880597014, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 6.576764978849004e-06, | |
| "loss": 0.7509, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 8.967661691542288, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.422795750793509e-06, | |
| "loss": 0.7608, | |
| "step": 3605 | |
| }, | |
| { | |
| "epoch": 8.980099502487562, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 6.27059041482817e-06, | |
| "loss": 0.7519, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 8.992537313432836, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.1201518399613635e-06, | |
| "loss": 0.7563, | |
| "step": 3615 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 2.6926610469818115, | |
| "eval_runtime": 0.5353, | |
| "eval_samples_per_second": 18.68, | |
| "eval_steps_per_second": 1.868, | |
| "step": 3618 | |
| }, | |
| { | |
| "epoch": 9.00497512437811, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 5.971482861898836e-06, | |
| "loss": 0.7501, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 9.017412935323383, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 5.824586282990085e-06, | |
| "loss": 0.7529, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 9.029850746268657, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 5.679464872175666e-06, | |
| "loss": 0.7556, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 9.042288557213931, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 5.5361213649349095e-06, | |
| "loss": 0.7495, | |
| "step": 3635 | |
| }, | |
| { | |
| "epoch": 9.054726368159203, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 5.394558463234378e-06, | |
| "loss": 0.7526, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 9.067164179104477, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 5.254778835477015e-06, | |
| "loss": 0.7554, | |
| "step": 3645 | |
| }, | |
| { | |
| "epoch": 9.07960199004975, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 5.116785116451661e-06, | |
| "loss": 0.7491, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 9.092039800995025, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 4.9805799072836335e-06, | |
| "loss": 0.7585, | |
| "step": 3655 | |
| }, | |
| { | |
| "epoch": 9.104477611940299, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.846165775385459e-06, | |
| "loss": 0.754, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 9.116915422885572, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.71354525440868e-06, | |
| "loss": 0.7496, | |
| "step": 3665 | |
| }, | |
| { | |
| "epoch": 9.129353233830846, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 4.5827208441959424e-06, | |
| "loss": 0.7587, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 9.14179104477612, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 4.4536950107339605e-06, | |
| "loss": 0.752, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 9.154228855721392, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 4.3264701861070345e-06, | |
| "loss": 0.7593, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 9.166666666666666, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 4.20104876845111e-06, | |
| "loss": 0.7581, | |
| "step": 3685 | |
| }, | |
| { | |
| "epoch": 9.17910447761194, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 4.077433121908747e-06, | |
| "loss": 0.757, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 9.191542288557214, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 3.955625576584376e-06, | |
| "loss": 0.76, | |
| "step": 3695 | |
| }, | |
| { | |
| "epoch": 9.203980099502488, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.835628428500515e-06, | |
| "loss": 0.7555, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 9.216417910447761, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 3.717443939554388e-06, | |
| "loss": 0.7512, | |
| "step": 3705 | |
| }, | |
| { | |
| "epoch": 9.228855721393035, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.601074337475352e-06, | |
| "loss": 0.7533, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 9.24129353233831, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.486521815782895e-06, | |
| "loss": 0.7541, | |
| "step": 3715 | |
| }, | |
| { | |
| "epoch": 9.253731343283581, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.3737885337452814e-06, | |
| "loss": 0.7492, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 9.266169154228855, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 3.262876616338839e-06, | |
| "loss": 0.7546, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 9.278606965174129, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 3.153788154207926e-06, | |
| "loss": 0.7509, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 9.291044776119403, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 3.0465252036255165e-06, | |
| "loss": 0.7614, | |
| "step": 3735 | |
| }, | |
| { | |
| "epoch": 9.303482587064677, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 2.9410897864544206e-06, | |
| "loss": 0.7572, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 9.31592039800995, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.8374838901092047e-06, | |
| "loss": 0.7582, | |
| "step": 3745 | |
| }, | |
| { | |
| "epoch": 9.328358208955224, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 2.735709467518699e-06, | |
| "loss": 0.7539, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 9.340796019900498, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.6357684370892256e-06, | |
| "loss": 0.7438, | |
| "step": 3755 | |
| }, | |
| { | |
| "epoch": 9.35323383084577, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 2.5376626826683956e-06, | |
| "loss": 0.7583, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 9.365671641791044, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.4413940535096156e-06, | |
| "loss": 0.7617, | |
| "step": 3765 | |
| }, | |
| { | |
| "epoch": 9.378109452736318, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.3469643642372586e-06, | |
| "loss": 0.7467, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 9.390547263681592, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 2.2543753948123803e-06, | |
| "loss": 0.7604, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 9.402985074626866, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.1636288904992585e-06, | |
| "loss": 0.7458, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 9.41542288557214, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 2.0747265618324406e-06, | |
| "loss": 0.7502, | |
| "step": 3785 | |
| }, | |
| { | |
| "epoch": 9.427860696517413, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 1.9876700845845475e-06, | |
| "loss": 0.7523, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 9.440298507462687, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.902461099734587e-06, | |
| "loss": 0.751, | |
| "step": 3795 | |
| }, | |
| { | |
| "epoch": 9.45273631840796, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.8191012134371577e-06, | |
| "loss": 0.7623, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 9.465174129353233, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.7375919969920717e-06, | |
| "loss": 0.743, | |
| "step": 3805 | |
| }, | |
| { | |
| "epoch": 9.477611940298507, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.6579349868147687e-06, | |
| "loss": 0.7576, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 9.490049751243781, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.580131684407382e-06, | |
| "loss": 0.7462, | |
| "step": 3815 | |
| }, | |
| { | |
| "epoch": 9.502487562189055, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.5041835563303742e-06, | |
| "loss": 0.7554, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 9.514925373134329, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.430092034174979e-06, | |
| "loss": 0.7636, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 9.527363184079602, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 1.3578585145360812e-06, | |
| "loss": 0.7449, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 9.539800995024876, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.2874843589860907e-06, | |
| "loss": 0.7582, | |
| "step": 3835 | |
| }, | |
| { | |
| "epoch": 9.552238805970148, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.2189708940490652e-06, | |
| "loss": 0.7444, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 9.564676616915422, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 1.1523194111758506e-06, | |
| "loss": 0.7498, | |
| "step": 3845 | |
| }, | |
| { | |
| "epoch": 9.577114427860696, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.0875311667196908e-06, | |
| "loss": 0.7495, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 9.58955223880597, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 1.0246073819125345e-06, | |
| "loss": 0.7473, | |
| "step": 3855 | |
| }, | |
| { | |
| "epoch": 9.601990049751244, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.635492428420434e-07, | |
| "loss": 0.7408, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 9.614427860696518, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.043579004291758e-07, | |
| "loss": 0.7482, | |
| "step": 3865 | |
| }, | |
| { | |
| "epoch": 9.626865671641792, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 8.470344704066046e-07, | |
| "loss": 0.752, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 9.639303482587065, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 7.915800332975565e-07, | |
| "loss": 0.7514, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 9.65174129353234, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 7.379956343955386e-07, | |
| "loss": 0.7671, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 9.664179104477611, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.862822837445881e-07, | |
| "loss": 0.7522, | |
| "step": 3885 | |
| }, | |
| { | |
| "epoch": 9.676616915422885, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.364409561202323e-07, | |
| "loss": 0.7669, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 9.689054726368159, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 5.884725910111688e-07, | |
| "loss": 0.7586, | |
| "step": 3895 | |
| }, | |
| { | |
| "epoch": 9.701492537313433, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.42378092601481e-07, | |
| "loss": 0.7607, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 9.713930348258707, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 4.981583297536507e-07, | |
| "loss": 0.7594, | |
| "step": 3905 | |
| }, | |
| { | |
| "epoch": 9.72636815920398, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 4.558141359921386e-07, | |
| "loss": 0.7475, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 9.738805970149254, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 4.153463094877186e-07, | |
| "loss": 0.7511, | |
| "step": 3915 | |
| }, | |
| { | |
| "epoch": 9.751243781094526, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 3.7675561304238994e-07, | |
| "loss": 0.7481, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 9.7636815920398, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.4004277407502226e-07, | |
| "loss": 0.7507, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 9.776119402985074, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 3.0520848460765527e-07, | |
| "loss": 0.7523, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 9.788557213930348, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.72253401252387e-07, | |
| "loss": 0.755, | |
| "step": 3935 | |
| }, | |
| { | |
| "epoch": 9.800995024875622, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 2.4117814519911684e-07, | |
| "loss": 0.745, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 9.813432835820896, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 2.119833022037221e-07, | |
| "loss": 0.7367, | |
| "step": 3945 | |
| }, | |
| { | |
| "epoch": 9.82587064676617, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.846694225770551e-07, | |
| "loss": 0.7577, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 9.838308457711443, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.5923702117461858e-07, | |
| "loss": 0.7581, | |
| "step": 3955 | |
| }, | |
| { | |
| "epoch": 9.850746268656717, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.3568657738678435e-07, | |
| "loss": 0.7489, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 9.86318407960199, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.1401853512980066e-07, | |
| "loss": 0.7615, | |
| "step": 3965 | |
| }, | |
| { | |
| "epoch": 9.875621890547263, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.423330283742093e-08, | |
| "loss": 0.7536, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 9.888059701492537, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 7.633125345317682e-08, | |
| "loss": 0.752, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 9.900497512437811, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 6.031272442341696e-08, | |
| "loss": 0.7586, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 9.912935323383085, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.617801769085661e-08, | |
| "loss": 0.7611, | |
| "step": 3985 | |
| }, | |
| { | |
| "epoch": 9.925373134328359, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.392739968894887e-08, | |
| "loss": 0.7445, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 9.937810945273633, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 2.356110133683309e-08, | |
| "loss": 0.7508, | |
| "step": 3995 | |
| }, | |
| { | |
| "epoch": 9.950248756218905, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.5079318035016164e-08, | |
| "loss": 0.7522, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.962686567164178, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 8.48220966164215e-09, | |
| "loss": 0.7533, | |
| "step": 4005 | |
| }, | |
| { | |
| "epoch": 9.975124378109452, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 3.769900569505769e-09, | |
| "loss": 0.7453, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 9.987562189054726, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.424795837098544e-10, | |
| "loss": 0.7576, | |
| "step": 4015 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0, | |
| "loss": 0.7461, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.6927685737609863, | |
| "eval_runtime": 0.5357, | |
| "eval_samples_per_second": 18.666, | |
| "eval_steps_per_second": 1.867, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 4020, | |
| "total_flos": 2.396116594708185e+18, | |
| "train_loss": 0.875829508648583, | |
| "train_runtime": 25148.6449, | |
| "train_samples_per_second": 7.668, | |
| "train_steps_per_second": 0.16 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 4020, | |
| "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": 2.396116594708185e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |