| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "global_step": 254, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007874015748031496, | |
| "grad_norm": 0.048636828926147534, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.1544, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.015748031496062992, | |
| "grad_norm": 0.06829153340487745, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.2093, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.023622047244094488, | |
| "grad_norm": 0.14998917767699588, | |
| "learning_rate": 1.153846153846154e-05, | |
| "loss": 0.3021, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.031496062992125984, | |
| "grad_norm": 0.05894881793284517, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.1696, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03937007874015748, | |
| "grad_norm": 0.06797578046068338, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 0.1865, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.047244094488188976, | |
| "grad_norm": 0.04989501389059178, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.1527, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05511811023622047, | |
| "grad_norm": 0.13344186213137674, | |
| "learning_rate": 2.6923076923076923e-05, | |
| "loss": 0.2722, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06299212598425197, | |
| "grad_norm": 0.10816772928150016, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.243, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07086614173228346, | |
| "grad_norm": 0.04182475098336472, | |
| "learning_rate": 3.461538461538462e-05, | |
| "loss": 0.1224, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.07874015748031496, | |
| "grad_norm": 0.1554026910734981, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.2677, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08661417322834646, | |
| "grad_norm": 0.15617841077219685, | |
| "learning_rate": 4.230769230769231e-05, | |
| "loss": 0.2695, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.09448818897637795, | |
| "grad_norm": 0.07385844760092473, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.1463, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.10236220472440945, | |
| "grad_norm": 0.16533108037457117, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2453, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.11023622047244094, | |
| "grad_norm": 0.25293589244560055, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.2824, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.11811023622047244, | |
| "grad_norm": 0.08466230707261538, | |
| "learning_rate": 5.769230769230769e-05, | |
| "loss": 0.133, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.12598425196850394, | |
| "grad_norm": 0.1754852847424689, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.2315, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.13385826771653545, | |
| "grad_norm": 0.11115487227175415, | |
| "learning_rate": 6.538461538461539e-05, | |
| "loss": 0.1498, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.14173228346456693, | |
| "grad_norm": 0.11637176629211, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.1766, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.14960629921259844, | |
| "grad_norm": 0.08400494585959933, | |
| "learning_rate": 7.307692307692307e-05, | |
| "loss": 0.1329, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.15748031496062992, | |
| "grad_norm": 0.06131863575653607, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.1113, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.16535433070866143, | |
| "grad_norm": 0.07242179489824115, | |
| "learning_rate": 8.076923076923078e-05, | |
| "loss": 0.1365, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1732283464566929, | |
| "grad_norm": 0.09210836600277003, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.1275, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.18110236220472442, | |
| "grad_norm": 0.08327248894343327, | |
| "learning_rate": 8.846153846153847e-05, | |
| "loss": 0.1333, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1889763779527559, | |
| "grad_norm": 0.12488058178149539, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.1552, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1968503937007874, | |
| "grad_norm": 0.0715803184402007, | |
| "learning_rate": 9.615384615384617e-05, | |
| "loss": 0.1005, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2047244094488189, | |
| "grad_norm": 0.08165576376897732, | |
| "learning_rate": 0.0001, | |
| "loss": 0.129, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2125984251968504, | |
| "grad_norm": 0.06777023964931363, | |
| "learning_rate": 9.999525361252996e-05, | |
| "loss": 0.0818, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2204724409448819, | |
| "grad_norm": 0.08188897028908457, | |
| "learning_rate": 9.998101535124758e-05, | |
| "loss": 0.1067, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2283464566929134, | |
| "grad_norm": 0.05955056508032511, | |
| "learning_rate": 9.995728791936504e-05, | |
| "loss": 0.0781, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.23622047244094488, | |
| "grad_norm": 0.10052278242607535, | |
| "learning_rate": 9.992407582166581e-05, | |
| "loss": 0.1132, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2440944881889764, | |
| "grad_norm": 0.07776039997533422, | |
| "learning_rate": 9.988138536364922e-05, | |
| "loss": 0.0974, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.25196850393700787, | |
| "grad_norm": 0.08388901010421355, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 0.1009, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.25984251968503935, | |
| "grad_norm": 0.06560022866292156, | |
| "learning_rate": 9.976760358471686e-05, | |
| "loss": 0.1105, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2677165354330709, | |
| "grad_norm": 0.05590070356365914, | |
| "learning_rate": 9.969653386589748e-05, | |
| "loss": 0.1103, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2755905511811024, | |
| "grad_norm": 0.04905786983982704, | |
| "learning_rate": 9.961602898685226e-05, | |
| "loss": 0.0881, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.28346456692913385, | |
| "grad_norm": 0.05490757430787913, | |
| "learning_rate": 9.952610423187516e-05, | |
| "loss": 0.0918, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.29133858267716534, | |
| "grad_norm": 0.043950941798686236, | |
| "learning_rate": 9.942677667367541e-05, | |
| "loss": 0.0881, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2992125984251969, | |
| "grad_norm": 0.05071542251542392, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 0.0729, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.30708661417322836, | |
| "grad_norm": 0.040398473430997464, | |
| "learning_rate": 9.9199990360734e-05, | |
| "loss": 0.0793, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.31496062992125984, | |
| "grad_norm": 0.037634938852860234, | |
| "learning_rate": 9.90725746626209e-05, | |
| "loss": 0.0776, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3228346456692913, | |
| "grad_norm": 0.0479244660692272, | |
| "learning_rate": 9.893584226636772e-05, | |
| "loss": 0.0801, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.33070866141732286, | |
| "grad_norm": 0.05346016466242833, | |
| "learning_rate": 9.878981913137179e-05, | |
| "loss": 0.0817, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.33858267716535434, | |
| "grad_norm": 0.07429962714762295, | |
| "learning_rate": 9.86345329809282e-05, | |
| "loss": 0.0942, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3464566929133858, | |
| "grad_norm": 0.036250514133953575, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 0.0761, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.3543307086614173, | |
| "grad_norm": 0.03881655826068151, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.0751, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.36220472440944884, | |
| "grad_norm": 0.040273401041872256, | |
| "learning_rate": 9.811340001546251e-05, | |
| "loss": 0.0912, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3700787401574803, | |
| "grad_norm": 0.05203429733895773, | |
| "learning_rate": 9.792137412291265e-05, | |
| "loss": 0.0894, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3779527559055118, | |
| "grad_norm": 0.04242863105236421, | |
| "learning_rate": 9.772025009397537e-05, | |
| "loss": 0.0804, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3858267716535433, | |
| "grad_norm": 0.040515570278649306, | |
| "learning_rate": 9.751006611315356e-05, | |
| "loss": 0.0717, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3937007874015748, | |
| "grad_norm": 0.04280194987120334, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 0.0713, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4015748031496063, | |
| "grad_norm": 0.07612965297736185, | |
| "learning_rate": 9.706267962669998e-05, | |
| "loss": 0.0935, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4094488188976378, | |
| "grad_norm": 0.046501972727124995, | |
| "learning_rate": 9.682556205985274e-05, | |
| "loss": 0.0558, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.41732283464566927, | |
| "grad_norm": 0.039381255305059475, | |
| "learning_rate": 9.657955440256395e-05, | |
| "loss": 0.0741, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.4251968503937008, | |
| "grad_norm": 0.040869656588052625, | |
| "learning_rate": 9.632470336074009e-05, | |
| "loss": 0.0693, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4330708661417323, | |
| "grad_norm": 0.034972215143239324, | |
| "learning_rate": 9.606105731925283e-05, | |
| "loss": 0.0654, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.4409448818897638, | |
| "grad_norm": 0.04651590524829242, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 0.0895, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.44881889763779526, | |
| "grad_norm": 0.03790710812187682, | |
| "learning_rate": 9.550758211616684e-05, | |
| "loss": 0.0587, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.4566929133858268, | |
| "grad_norm": 0.06064386306880854, | |
| "learning_rate": 9.521785803487889e-05, | |
| "loss": 0.0704, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4645669291338583, | |
| "grad_norm": 0.04665702311958301, | |
| "learning_rate": 9.491954909459895e-05, | |
| "loss": 0.0785, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.47244094488188976, | |
| "grad_norm": 0.03845050550687295, | |
| "learning_rate": 9.46127119309197e-05, | |
| "loss": 0.0707, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.48031496062992124, | |
| "grad_norm": 0.06770425704069039, | |
| "learning_rate": 9.42974047985639e-05, | |
| "loss": 0.0989, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4881889763779528, | |
| "grad_norm": 0.0493379370118964, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 0.0773, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.49606299212598426, | |
| "grad_norm": 0.05424018637256655, | |
| "learning_rate": 9.364162167569907e-05, | |
| "loss": 0.0862, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5039370078740157, | |
| "grad_norm": 0.035888638018077, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.0696, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5118110236220472, | |
| "grad_norm": 0.05419778194628531, | |
| "learning_rate": 9.295269771849427e-05, | |
| "loss": 0.0655, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5196850393700787, | |
| "grad_norm": 0.042949223060275404, | |
| "learning_rate": 9.259597044191636e-05, | |
| "loss": 0.0607, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5275590551181102, | |
| "grad_norm": 0.04264699828582032, | |
| "learning_rate": 9.223115608612325e-05, | |
| "loss": 0.0647, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5354330708661418, | |
| "grad_norm": 0.04465981374861148, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 0.0721, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5433070866141733, | |
| "grad_norm": 0.15086641781783566, | |
| "learning_rate": 9.147754470716408e-05, | |
| "loss": 0.0588, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5511811023622047, | |
| "grad_norm": 0.03608693394469442, | |
| "learning_rate": 9.108889076126226e-05, | |
| "loss": 0.0598, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5590551181102362, | |
| "grad_norm": 0.04512147014513813, | |
| "learning_rate": 9.069243586350975e-05, | |
| "loss": 0.0683, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5669291338582677, | |
| "grad_norm": 0.034531651539235535, | |
| "learning_rate": 9.028825528304892e-05, | |
| "loss": 0.0534, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5748031496062992, | |
| "grad_norm": 0.04001741026747657, | |
| "learning_rate": 8.987642575578545e-05, | |
| "loss": 0.0679, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5826771653543307, | |
| "grad_norm": 0.0627167723825249, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 0.0843, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5905511811023622, | |
| "grad_norm": 0.041162273807440045, | |
| "learning_rate": 8.903013405060211e-05, | |
| "loss": 0.0769, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5984251968503937, | |
| "grad_norm": 0.06413471441605206, | |
| "learning_rate": 8.859583254581605e-05, | |
| "loss": 0.079, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6062992125984252, | |
| "grad_norm": 0.038139347016296524, | |
| "learning_rate": 8.815420340999033e-05, | |
| "loss": 0.061, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6141732283464567, | |
| "grad_norm": 0.05637898761021731, | |
| "learning_rate": 8.770533048884482e-05, | |
| "loss": 0.0619, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6220472440944882, | |
| "grad_norm": 0.0357824590275947, | |
| "learning_rate": 8.724929900337186e-05, | |
| "loss": 0.0584, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.6299212598425197, | |
| "grad_norm": 0.051212360329821134, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 0.0839, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6377952755905512, | |
| "grad_norm": 0.05049167558999489, | |
| "learning_rate": 8.631610800243926e-05, | |
| "loss": 0.0589, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.6456692913385826, | |
| "grad_norm": 0.03961922655074931, | |
| "learning_rate": 8.583912565842257e-05, | |
| "loss": 0.0657, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.6535433070866141, | |
| "grad_norm": 0.04037566238898647, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.0747, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.6614173228346457, | |
| "grad_norm": 0.039919573702817145, | |
| "learning_rate": 8.486484005469977e-05, | |
| "loss": 0.0719, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.6692913385826772, | |
| "grad_norm": 0.033793800349021845, | |
| "learning_rate": 8.436772176847294e-05, | |
| "loss": 0.0582, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.6771653543307087, | |
| "grad_norm": 0.033700515363253886, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 0.0669, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6850393700787402, | |
| "grad_norm": 0.040605736837914866, | |
| "learning_rate": 8.335400611257067e-05, | |
| "loss": 0.0652, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6929133858267716, | |
| "grad_norm": 0.054522693875205565, | |
| "learning_rate": 8.283760120238672e-05, | |
| "loss": 0.0717, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7007874015748031, | |
| "grad_norm": 0.03702465616892497, | |
| "learning_rate": 8.231496189304704e-05, | |
| "loss": 0.0731, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7086614173228346, | |
| "grad_norm": 0.03285710986831703, | |
| "learning_rate": 8.178618741049842e-05, | |
| "loss": 0.0668, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7165354330708661, | |
| "grad_norm": 0.053063329951814404, | |
| "learning_rate": 8.125137814548393e-05, | |
| "loss": 0.067, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.7244094488188977, | |
| "grad_norm": 0.040180722630613536, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 0.0698, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.7322834645669292, | |
| "grad_norm": 0.0416900575795704, | |
| "learning_rate": 8.016406254043595e-05, | |
| "loss": 0.0725, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.7401574803149606, | |
| "grad_norm": 0.03869655273152511, | |
| "learning_rate": 7.961176263324901e-05, | |
| "loss": 0.0659, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.7480314960629921, | |
| "grad_norm": 0.044825068618721646, | |
| "learning_rate": 7.905384077009693e-05, | |
| "loss": 0.0731, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.7559055118110236, | |
| "grad_norm": 0.044483649059968176, | |
| "learning_rate": 7.849040287551331e-05, | |
| "loss": 0.0634, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.7637795275590551, | |
| "grad_norm": 0.04017053730657027, | |
| "learning_rate": 7.79215559212807e-05, | |
| "loss": 0.078, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.7716535433070866, | |
| "grad_norm": 0.04145460320890759, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 0.0768, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.7795275590551181, | |
| "grad_norm": 0.03633052402965569, | |
| "learning_rate": 7.676806783519304e-05, | |
| "loss": 0.0664, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.7874015748031497, | |
| "grad_norm": 0.04021689685346706, | |
| "learning_rate": 7.618364569939391e-05, | |
| "loss": 0.0674, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7952755905511811, | |
| "grad_norm": 0.060396460565417545, | |
| "learning_rate": 7.559425245448006e-05, | |
| "loss": 0.0694, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.8031496062992126, | |
| "grad_norm": 0.03969555967865163, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.0559, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.8110236220472441, | |
| "grad_norm": 0.038622088305060684, | |
| "learning_rate": 7.440100115804991e-05, | |
| "loss": 0.0678, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.8188976377952756, | |
| "grad_norm": 0.0408599543219439, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 0.0596, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.8267716535433071, | |
| "grad_norm": 0.03612580723785473, | |
| "learning_rate": 7.318922008417203e-05, | |
| "loss": 0.0613, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.8346456692913385, | |
| "grad_norm": 0.051121513111016405, | |
| "learning_rate": 7.257666791554448e-05, | |
| "loss": 0.0681, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.84251968503937, | |
| "grad_norm": 0.053947523356233006, | |
| "learning_rate": 7.195982944236851e-05, | |
| "loss": 0.0642, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.8503937007874016, | |
| "grad_norm": 0.04708378321870759, | |
| "learning_rate": 7.133882177482019e-05, | |
| "loss": 0.0714, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.8582677165354331, | |
| "grad_norm": 0.05157517691265944, | |
| "learning_rate": 7.071376281461994e-05, | |
| "loss": 0.0679, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.8661417322834646, | |
| "grad_norm": 0.043251187130909544, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 0.0638, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.8740157480314961, | |
| "grad_norm": 0.03498275614195765, | |
| "learning_rate": 6.94519664464163e-05, | |
| "loss": 0.0545, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.8818897637795275, | |
| "grad_norm": 0.04509157725839411, | |
| "learning_rate": 6.881546859739179e-05, | |
| "loss": 0.072, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.889763779527559, | |
| "grad_norm": 0.03949179802033183, | |
| "learning_rate": 6.817539852819149e-05, | |
| "loss": 0.0679, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.8976377952755905, | |
| "grad_norm": 0.04824956524429339, | |
| "learning_rate": 6.753187775963773e-05, | |
| "loss": 0.0602, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.905511811023622, | |
| "grad_norm": 0.05052923605617683, | |
| "learning_rate": 6.688502846768696e-05, | |
| "loss": 0.07, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.9133858267716536, | |
| "grad_norm": 0.049650560302814, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 0.0588, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.9212598425196851, | |
| "grad_norm": 0.04693328297700466, | |
| "learning_rate": 6.558183615379707e-05, | |
| "loss": 0.0848, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.9291338582677166, | |
| "grad_norm": 0.05162117635179258, | |
| "learning_rate": 6.492574055008473e-05, | |
| "loss": 0.0779, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.937007874015748, | |
| "grad_norm": 0.04642140808574063, | |
| "learning_rate": 6.426681121245527e-05, | |
| "loss": 0.0711, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.9448818897637795, | |
| "grad_norm": 0.03821222457115445, | |
| "learning_rate": 6.360517324226676e-05, | |
| "loss": 0.0597, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.952755905511811, | |
| "grad_norm": 0.07405595024512149, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 0.0691, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.9606299212598425, | |
| "grad_norm": 0.04965701613939319, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.0697, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.968503937007874, | |
| "grad_norm": 0.053755810398937655, | |
| "learning_rate": 6.16052661204734e-05, | |
| "loss": 0.0713, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.9763779527559056, | |
| "grad_norm": 0.041060956975168636, | |
| "learning_rate": 6.09340545603188e-05, | |
| "loss": 0.0752, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.984251968503937, | |
| "grad_norm": 0.05257209833173453, | |
| "learning_rate": 6.026076710978171e-05, | |
| "loss": 0.0749, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9921259842519685, | |
| "grad_norm": 0.04634248903143434, | |
| "learning_rate": 5.958553159618693e-05, | |
| "loss": 0.0721, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.04595655646478586, | |
| "learning_rate": 5.890847621670966e-05, | |
| "loss": 0.0643, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.0078740157480315, | |
| "grad_norm": 0.0428889003832125, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 0.0592, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.015748031496063, | |
| "grad_norm": 0.06513209718503772, | |
| "learning_rate": 5.7549420351961844e-05, | |
| "loss": 0.0661, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.0236220472440944, | |
| "grad_norm": 0.048508681912102766, | |
| "learning_rate": 5.686767789092041e-05, | |
| "loss": 0.062, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.031496062992126, | |
| "grad_norm": 0.03942306111426942, | |
| "learning_rate": 5.618463156346739e-05, | |
| "loss": 0.06, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.0393700787401574, | |
| "grad_norm": 0.05920893966651, | |
| "learning_rate": 5.550041104970397e-05, | |
| "loss": 0.0574, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.047244094488189, | |
| "grad_norm": 0.06537752227653837, | |
| "learning_rate": 5.481514625265709e-05, | |
| "loss": 0.0789, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.0551181102362204, | |
| "grad_norm": 0.052133215264540934, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 0.0539, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.0629921259842519, | |
| "grad_norm": 0.03961482885456455, | |
| "learning_rate": 5.344200438743489e-05, | |
| "loss": 0.0584, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.0708661417322836, | |
| "grad_norm": 0.04991806278320797, | |
| "learning_rate": 5.2754388017793274e-05, | |
| "loss": 0.0597, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.078740157480315, | |
| "grad_norm": 0.05656237899859479, | |
| "learning_rate": 5.2066248712440656e-05, | |
| "loss": 0.0758, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.0866141732283465, | |
| "grad_norm": 0.055792787158515794, | |
| "learning_rate": 5.1377717118408105e-05, | |
| "loss": 0.0754, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.094488188976378, | |
| "grad_norm": 0.058089701897722286, | |
| "learning_rate": 5.068892395720483e-05, | |
| "loss": 0.0673, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.1023622047244095, | |
| "grad_norm": 0.049023325150838494, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0591, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.110236220472441, | |
| "grad_norm": 0.04294743238196575, | |
| "learning_rate": 4.9311076042795185e-05, | |
| "loss": 0.0573, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.1181102362204725, | |
| "grad_norm": 0.04629796470945678, | |
| "learning_rate": 4.8622282881591906e-05, | |
| "loss": 0.0662, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.125984251968504, | |
| "grad_norm": 0.04706455796489872, | |
| "learning_rate": 4.7933751287559335e-05, | |
| "loss": 0.0705, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.1338582677165354, | |
| "grad_norm": 0.04805424749077547, | |
| "learning_rate": 4.7245611982206724e-05, | |
| "loss": 0.0617, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.141732283464567, | |
| "grad_norm": 0.04537362622805077, | |
| "learning_rate": 4.6557995612565144e-05, | |
| "loss": 0.0563, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.1496062992125984, | |
| "grad_norm": 0.043403479823984015, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 0.0534, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.1574803149606299, | |
| "grad_norm": 0.03764537967703119, | |
| "learning_rate": 4.518485374734292e-05, | |
| "loss": 0.0515, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.1653543307086613, | |
| "grad_norm": 0.056823438107632934, | |
| "learning_rate": 4.449958895029604e-05, | |
| "loss": 0.0742, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.1732283464566928, | |
| "grad_norm": 0.04193687232087755, | |
| "learning_rate": 4.381536843653262e-05, | |
| "loss": 0.0542, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.1811023622047245, | |
| "grad_norm": 0.046536924814026416, | |
| "learning_rate": 4.3132322109079596e-05, | |
| "loss": 0.0855, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.188976377952756, | |
| "grad_norm": 0.05457156199530363, | |
| "learning_rate": 4.2450579648038154e-05, | |
| "loss": 0.0553, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.1968503937007875, | |
| "grad_norm": 0.04033132170380821, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 0.0538, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.204724409448819, | |
| "grad_norm": 0.049656351067623145, | |
| "learning_rate": 4.109152378329036e-05, | |
| "loss": 0.0542, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.2125984251968505, | |
| "grad_norm": 0.03998164718927446, | |
| "learning_rate": 4.0414468403813095e-05, | |
| "loss": 0.0521, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.220472440944882, | |
| "grad_norm": 0.05745317845238628, | |
| "learning_rate": 3.973923289021829e-05, | |
| "loss": 0.0609, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.2283464566929134, | |
| "grad_norm": 0.042323711634250194, | |
| "learning_rate": 3.9065945439681214e-05, | |
| "loss": 0.0626, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.236220472440945, | |
| "grad_norm": 0.05723447684288661, | |
| "learning_rate": 3.839473387952662e-05, | |
| "loss": 0.0643, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.2440944881889764, | |
| "grad_norm": 0.05609057194605591, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 0.0446, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.2519685039370079, | |
| "grad_norm": 0.04383475827201075, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 0.054, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.2598425196850394, | |
| "grad_norm": 0.04721149501353491, | |
| "learning_rate": 3.639482675773324e-05, | |
| "loss": 0.0768, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.2677165354330708, | |
| "grad_norm": 0.044660845340955664, | |
| "learning_rate": 3.5733188787544745e-05, | |
| "loss": 0.0433, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.2755905511811023, | |
| "grad_norm": 0.05604911676781777, | |
| "learning_rate": 3.5074259449915284e-05, | |
| "loss": 0.0621, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.2834645669291338, | |
| "grad_norm": 0.049412392554855175, | |
| "learning_rate": 3.4418163846202944e-05, | |
| "loss": 0.067, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.2913385826771653, | |
| "grad_norm": 0.05304697118137444, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 0.06, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.2992125984251968, | |
| "grad_norm": 0.05294702902408293, | |
| "learning_rate": 3.3114971532313056e-05, | |
| "loss": 0.067, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.3070866141732282, | |
| "grad_norm": 0.041258642954892746, | |
| "learning_rate": 3.2468122240362284e-05, | |
| "loss": 0.0472, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.3149606299212597, | |
| "grad_norm": 0.053885127357226115, | |
| "learning_rate": 3.18246014718085e-05, | |
| "loss": 0.0679, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.3228346456692912, | |
| "grad_norm": 0.06157836754578656, | |
| "learning_rate": 3.118453140260823e-05, | |
| "loss": 0.0735, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.330708661417323, | |
| "grad_norm": 0.05459458021353128, | |
| "learning_rate": 3.0548033553583705e-05, | |
| "loss": 0.0618, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.3385826771653544, | |
| "grad_norm": 0.06138229722884799, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 0.06, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.3464566929133859, | |
| "grad_norm": 0.04607977533183901, | |
| "learning_rate": 2.928623718538006e-05, | |
| "loss": 0.056, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.3543307086614174, | |
| "grad_norm": 0.04457994569263475, | |
| "learning_rate": 2.866117822517982e-05, | |
| "loss": 0.0505, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.3622047244094488, | |
| "grad_norm": 0.05347674759185391, | |
| "learning_rate": 2.804017055763149e-05, | |
| "loss": 0.0701, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.3700787401574803, | |
| "grad_norm": 0.04603095018943048, | |
| "learning_rate": 2.7423332084455544e-05, | |
| "loss": 0.0661, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.3779527559055118, | |
| "grad_norm": 0.04535204168167475, | |
| "learning_rate": 2.681077991582797e-05, | |
| "loss": 0.0638, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.3858267716535433, | |
| "grad_norm": 0.044079347580669566, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 0.0483, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.3937007874015748, | |
| "grad_norm": 0.04787312823352356, | |
| "learning_rate": 2.5598998841950107e-05, | |
| "loss": 0.0485, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.4015748031496063, | |
| "grad_norm": 0.053556935205416635, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.0598, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.4094488188976377, | |
| "grad_norm": 0.048892583117746094, | |
| "learning_rate": 2.4405747545519963e-05, | |
| "loss": 0.0671, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.4173228346456692, | |
| "grad_norm": 0.05573731864284115, | |
| "learning_rate": 2.381635430060611e-05, | |
| "loss": 0.0519, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.425196850393701, | |
| "grad_norm": 0.04104144198286972, | |
| "learning_rate": 2.323193216480698e-05, | |
| "loss": 0.0619, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.4330708661417324, | |
| "grad_norm": 0.04902777007420262, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 0.0587, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.4409448818897639, | |
| "grad_norm": 0.04484354432801945, | |
| "learning_rate": 2.207844407871929e-05, | |
| "loss": 0.0554, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.4488188976377954, | |
| "grad_norm": 0.06590936087340947, | |
| "learning_rate": 2.150959712448669e-05, | |
| "loss": 0.0811, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.4566929133858268, | |
| "grad_norm": 0.04102430626416167, | |
| "learning_rate": 2.094615922990309e-05, | |
| "loss": 0.0428, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.4645669291338583, | |
| "grad_norm": 0.041794329984435635, | |
| "learning_rate": 2.0388237366751006e-05, | |
| "loss": 0.0605, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.4724409448818898, | |
| "grad_norm": 0.05045021005842732, | |
| "learning_rate": 1.9835937459564064e-05, | |
| "loss": 0.0536, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.4803149606299213, | |
| "grad_norm": 0.04618763538981042, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 0.0556, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.4881889763779528, | |
| "grad_norm": 0.05106802741492108, | |
| "learning_rate": 1.874862185451608e-05, | |
| "loss": 0.0709, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.4960629921259843, | |
| "grad_norm": 0.05597419656508567, | |
| "learning_rate": 1.821381258950161e-05, | |
| "loss": 0.0489, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.5039370078740157, | |
| "grad_norm": 0.06222180812872232, | |
| "learning_rate": 1.768503810695295e-05, | |
| "loss": 0.0578, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.5118110236220472, | |
| "grad_norm": 0.04964108845251158, | |
| "learning_rate": 1.7162398797613282e-05, | |
| "loss": 0.0555, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.5196850393700787, | |
| "grad_norm": 0.05088603506896626, | |
| "learning_rate": 1.6645993887429345e-05, | |
| "loss": 0.0622, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.5275590551181102, | |
| "grad_norm": 0.046495091752318314, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 0.0595, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.5354330708661417, | |
| "grad_norm": 0.053917463079799346, | |
| "learning_rate": 1.563227823152708e-05, | |
| "loss": 0.075, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.5433070866141732, | |
| "grad_norm": 0.08260510658354386, | |
| "learning_rate": 1.5135159945300231e-05, | |
| "loss": 0.0716, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.5511811023622046, | |
| "grad_norm": 0.04738267952758889, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0587, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.5590551181102361, | |
| "grad_norm": 0.051915775011872965, | |
| "learning_rate": 1.4160874341577446e-05, | |
| "loss": 0.0521, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.5669291338582676, | |
| "grad_norm": 0.05481705129948632, | |
| "learning_rate": 1.368389199756075e-05, | |
| "loss": 0.0752, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.574803149606299, | |
| "grad_norm": 0.05217737291956296, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 0.0736, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.574803149606299, | |
| "eval_loss": 0.06070369854569435, | |
| "eval_runtime": 6.4146, | |
| "eval_samples_per_second": 0.935, | |
| "eval_steps_per_second": 0.312, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.5826771653543306, | |
| "grad_norm": 0.04678999221530293, | |
| "learning_rate": 1.275070099662815e-05, | |
| "loss": 0.0707, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.590551181102362, | |
| "grad_norm": 0.05124530088869734, | |
| "learning_rate": 1.2294669511155193e-05, | |
| "loss": 0.0559, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.5984251968503937, | |
| "grad_norm": 0.05350993198205092, | |
| "learning_rate": 1.1845796590009683e-05, | |
| "loss": 0.0768, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.6062992125984252, | |
| "grad_norm": 0.05328374514051326, | |
| "learning_rate": 1.1404167454183957e-05, | |
| "loss": 0.0638, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.6141732283464567, | |
| "grad_norm": 0.049069065326141866, | |
| "learning_rate": 1.0969865949397901e-05, | |
| "loss": 0.0701, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.6220472440944882, | |
| "grad_norm": 0.051454451624391905, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 0.0535, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.6299212598425197, | |
| "grad_norm": 0.04437769885060027, | |
| "learning_rate": 1.012357424421455e-05, | |
| "loss": 0.0521, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.6377952755905512, | |
| "grad_norm": 0.04825172010643271, | |
| "learning_rate": 9.711744716951093e-06, | |
| "loss": 0.0689, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.6456692913385826, | |
| "grad_norm": 0.06554675030762395, | |
| "learning_rate": 9.307564136490254e-06, | |
| "loss": 0.0612, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.6535433070866141, | |
| "grad_norm": 0.05866366054712351, | |
| "learning_rate": 8.911109238737747e-06, | |
| "loss": 0.0791, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.6614173228346458, | |
| "grad_norm": 0.04336852944598847, | |
| "learning_rate": 8.522455292835934e-06, | |
| "loss": 0.0539, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.6692913385826773, | |
| "grad_norm": 0.051947799814034856, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 0.0639, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.6771653543307088, | |
| "grad_norm": 0.04351599998197119, | |
| "learning_rate": 7.768843913876756e-06, | |
| "loss": 0.0482, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.6850393700787403, | |
| "grad_norm": 0.06480464685922364, | |
| "learning_rate": 7.404029558083653e-06, | |
| "loss": 0.0662, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.6929133858267718, | |
| "grad_norm": 0.04654351665030391, | |
| "learning_rate": 7.047302281505736e-06, | |
| "loss": 0.0566, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.7007874015748032, | |
| "grad_norm": 0.04652775620652711, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 0.0571, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.7086614173228347, | |
| "grad_norm": 0.047224618696043545, | |
| "learning_rate": 6.3583783243009285e-06, | |
| "loss": 0.0592, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.7165354330708662, | |
| "grad_norm": 0.053615270349897666, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 0.0646, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.7244094488188977, | |
| "grad_norm": 0.04665853226611898, | |
| "learning_rate": 5.702595201436101e-06, | |
| "loss": 0.0574, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.7322834645669292, | |
| "grad_norm": 0.05099339647037513, | |
| "learning_rate": 5.387288069080299e-06, | |
| "loss": 0.0614, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.7401574803149606, | |
| "grad_norm": 0.043289965587830936, | |
| "learning_rate": 5.080450905401057e-06, | |
| "loss": 0.0468, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.7480314960629921, | |
| "grad_norm": 0.0692890477383107, | |
| "learning_rate": 4.782141965121128e-06, | |
| "loss": 0.0493, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.7559055118110236, | |
| "grad_norm": 0.0403442528103141, | |
| "learning_rate": 4.492417883833155e-06, | |
| "loss": 0.0476, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.763779527559055, | |
| "grad_norm": 0.04191520401385937, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 0.052, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.7716535433070866, | |
| "grad_norm": 0.054916613323571846, | |
| "learning_rate": 3.9389426807471766e-06, | |
| "loss": 0.0653, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.779527559055118, | |
| "grad_norm": 0.04750102972177334, | |
| "learning_rate": 3.675296639259912e-06, | |
| "loss": 0.0578, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.7874015748031495, | |
| "grad_norm": 0.0588876838593015, | |
| "learning_rate": 3.420445597436056e-06, | |
| "loss": 0.0565, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.795275590551181, | |
| "grad_norm": 0.050319040408150564, | |
| "learning_rate": 3.1744379401472677e-06, | |
| "loss": 0.0546, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.8031496062992125, | |
| "grad_norm": 0.05754277450687693, | |
| "learning_rate": 2.9373203733000232e-06, | |
| "loss": 0.0686, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.811023622047244, | |
| "grad_norm": 0.05586902250961728, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 0.0673, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.8188976377952755, | |
| "grad_norm": 0.05258380689743355, | |
| "learning_rate": 2.4899338868464404e-06, | |
| "loss": 0.0656, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.826771653543307, | |
| "grad_norm": 0.0444146353656455, | |
| "learning_rate": 2.2797499060246253e-06, | |
| "loss": 0.0543, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.8346456692913384, | |
| "grad_norm": 0.050066903014096574, | |
| "learning_rate": 2.0786258770873647e-06, | |
| "loss": 0.0599, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.84251968503937, | |
| "grad_norm": 0.045770675687925415, | |
| "learning_rate": 1.8865999845374793e-06, | |
| "loss": 0.0599, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.8503937007874016, | |
| "grad_norm": 0.05464584148563528, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 0.0685, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.858267716535433, | |
| "grad_norm": 0.04968886797769799, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 0.068, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.8661417322834646, | |
| "grad_norm": 0.0552386892230046, | |
| "learning_rate": 1.3654670190718034e-06, | |
| "loss": 0.0668, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.874015748031496, | |
| "grad_norm": 0.042676978639831414, | |
| "learning_rate": 1.210180868628219e-06, | |
| "loss": 0.053, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.8818897637795275, | |
| "grad_norm": 0.04702355316793717, | |
| "learning_rate": 1.064157733632276e-06, | |
| "loss": 0.0513, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.889763779527559, | |
| "grad_norm": 0.07317497229013392, | |
| "learning_rate": 9.274253373791064e-07, | |
| "loss": 0.0609, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.8976377952755905, | |
| "grad_norm": 0.04768857996113459, | |
| "learning_rate": 8.000096392660029e-07, | |
| "loss": 0.0523, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.905511811023622, | |
| "grad_norm": 0.0600627581282451, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 0.0756, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.9133858267716537, | |
| "grad_norm": 0.048224067594923814, | |
| "learning_rate": 5.732233263245845e-07, | |
| "loss": 0.0568, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.9212598425196852, | |
| "grad_norm": 0.051208440325568635, | |
| "learning_rate": 4.738957681248379e-07, | |
| "loss": 0.0576, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.9291338582677167, | |
| "grad_norm": 0.05305714441239726, | |
| "learning_rate": 3.839710131477492e-07, | |
| "loss": 0.0578, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.9370078740157481, | |
| "grad_norm": 0.0539897244444409, | |
| "learning_rate": 3.034661341025258e-07, | |
| "loss": 0.0586, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.9448818897637796, | |
| "grad_norm": 0.04186116850231722, | |
| "learning_rate": 2.323964152831426e-07, | |
| "loss": 0.0511, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.952755905511811, | |
| "grad_norm": 0.049234512985121694, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 0.0597, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.9606299212598426, | |
| "grad_norm": 0.0551149693713877, | |
| "learning_rate": 1.1861463635077785e-07, | |
| "loss": 0.0684, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.968503937007874, | |
| "grad_norm": 0.05304684823721321, | |
| "learning_rate": 7.59241783341913e-08, | |
| "loss": 0.0682, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.9763779527559056, | |
| "grad_norm": 0.07253818377949055, | |
| "learning_rate": 4.2712080634949024e-08, | |
| "loss": 0.0637, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.984251968503937, | |
| "grad_norm": 0.053065918527288605, | |
| "learning_rate": 1.8984648752429225e-08, | |
| "loss": 0.0693, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.9921259842519685, | |
| "grad_norm": 0.05734495408696275, | |
| "learning_rate": 4.746387470044855e-09, | |
| "loss": 0.061, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.05703286573729645, | |
| "learning_rate": 0.0, | |
| "loss": 0.0489, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 254, | |
| "total_flos": 643226180517888.0, | |
| "train_loss": 0.07837846240131405, | |
| "train_runtime": 1925.3112, | |
| "train_samples_per_second": 0.525, | |
| "train_steps_per_second": 0.132 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 254, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 643226180517888.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |