| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.2835164835164834, |
| "eval_steps": 500, |
| "global_step": 195, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.006593406593406593, |
| "grad_norm": 5.459638595581055, |
| "learning_rate": 0.0, |
| "loss": 5.1572, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.013186813186813187, |
| "grad_norm": 4.408932209014893, |
| "learning_rate": 4e-05, |
| "loss": 4.941, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.01978021978021978, |
| "grad_norm": 5.487926959991455, |
| "learning_rate": 8e-05, |
| "loss": 5.5698, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.026373626373626374, |
| "grad_norm": 5.361940860748291, |
| "learning_rate": 0.00012, |
| "loss": 5.4254, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.03296703296703297, |
| "grad_norm": 5.856925010681152, |
| "learning_rate": 0.00016, |
| "loss": 4.9638, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.03956043956043956, |
| "grad_norm": 6.2031073570251465, |
| "learning_rate": 0.0002, |
| "loss": 4.1384, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.046153846153846156, |
| "grad_norm": 6.153454780578613, |
| "learning_rate": 0.00019955654101995565, |
| "loss": 2.5665, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.05274725274725275, |
| "grad_norm": 2.060917615890503, |
| "learning_rate": 0.00019911308203991133, |
| "loss": 1.335, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.05934065934065934, |
| "grad_norm": 1.5414468050003052, |
| "learning_rate": 0.00019866962305986697, |
| "loss": 1.2202, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.06593406593406594, |
| "grad_norm": 2.760967493057251, |
| "learning_rate": 0.00019822616407982261, |
| "loss": 1.3177, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.07252747252747253, |
| "grad_norm": 1.4582066535949707, |
| "learning_rate": 0.00019778270509977829, |
| "loss": 1.1973, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.07912087912087912, |
| "grad_norm": 0.800369918346405, |
| "learning_rate": 0.00019733924611973393, |
| "loss": 1.1367, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.08571428571428572, |
| "grad_norm": 0.8436974287033081, |
| "learning_rate": 0.0001968957871396896, |
| "loss": 1.2465, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.09230769230769231, |
| "grad_norm": 0.9210023283958435, |
| "learning_rate": 0.00019645232815964525, |
| "loss": 1.0176, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0989010989010989, |
| "grad_norm": 0.8066878318786621, |
| "learning_rate": 0.00019600886917960092, |
| "loss": 0.9148, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.1054945054945055, |
| "grad_norm": 0.8594912886619568, |
| "learning_rate": 0.00019556541019955653, |
| "loss": 1.0468, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.11208791208791209, |
| "grad_norm": 0.6237018704414368, |
| "learning_rate": 0.0001951219512195122, |
| "loss": 1.0676, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.11868131868131868, |
| "grad_norm": 0.6330975294113159, |
| "learning_rate": 0.00019467849223946785, |
| "loss": 0.9913, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.12527472527472527, |
| "grad_norm": 0.7443146705627441, |
| "learning_rate": 0.00019423503325942352, |
| "loss": 0.8187, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.13186813186813187, |
| "grad_norm": 0.7366983890533447, |
| "learning_rate": 0.00019379157427937917, |
| "loss": 1.1666, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.13846153846153847, |
| "grad_norm": 0.571020781993866, |
| "learning_rate": 0.00019334811529933484, |
| "loss": 0.9126, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.14505494505494507, |
| "grad_norm": 0.6713335514068604, |
| "learning_rate": 0.00019290465631929045, |
| "loss": 0.7393, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.15164835164835164, |
| "grad_norm": 0.5838486552238464, |
| "learning_rate": 0.00019246119733924613, |
| "loss": 0.9377, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.15824175824175823, |
| "grad_norm": 0.6766859889030457, |
| "learning_rate": 0.00019201773835920177, |
| "loss": 1.1663, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.16483516483516483, |
| "grad_norm": 0.6486108303070068, |
| "learning_rate": 0.00019157427937915744, |
| "loss": 0.8836, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.17142857142857143, |
| "grad_norm": 0.6390822529792786, |
| "learning_rate": 0.00019113082039911309, |
| "loss": 1.0328, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.17802197802197803, |
| "grad_norm": 0.6051312685012817, |
| "learning_rate": 0.00019068736141906876, |
| "loss": 1.1173, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.18461538461538463, |
| "grad_norm": 0.539639949798584, |
| "learning_rate": 0.0001902439024390244, |
| "loss": 0.8195, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.1912087912087912, |
| "grad_norm": 0.6244184970855713, |
| "learning_rate": 0.00018980044345898005, |
| "loss": 0.9956, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.1978021978021978, |
| "grad_norm": 0.5312393307685852, |
| "learning_rate": 0.00018935698447893572, |
| "loss": 0.9427, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.2043956043956044, |
| "grad_norm": 0.6213347911834717, |
| "learning_rate": 0.00018891352549889136, |
| "loss": 1.0347, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.210989010989011, |
| "grad_norm": 0.5339341759681702, |
| "learning_rate": 0.00018847006651884703, |
| "loss": 0.8984, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.2175824175824176, |
| "grad_norm": 0.5304893851280212, |
| "learning_rate": 0.00018802660753880268, |
| "loss": 0.8964, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.22417582417582418, |
| "grad_norm": 0.6589417457580566, |
| "learning_rate": 0.00018758314855875832, |
| "loss": 1.3518, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.23076923076923078, |
| "grad_norm": 0.6399314403533936, |
| "learning_rate": 0.00018713968957871397, |
| "loss": 1.1438, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.23736263736263735, |
| "grad_norm": 0.545724630355835, |
| "learning_rate": 0.00018669623059866964, |
| "loss": 0.9772, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.24395604395604395, |
| "grad_norm": 0.5593492984771729, |
| "learning_rate": 0.00018625277161862528, |
| "loss": 0.883, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.25054945054945055, |
| "grad_norm": 0.5673443078994751, |
| "learning_rate": 0.00018580931263858095, |
| "loss": 0.9052, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.2571428571428571, |
| "grad_norm": 0.6377619504928589, |
| "learning_rate": 0.0001853658536585366, |
| "loss": 0.9566, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.26373626373626374, |
| "grad_norm": 0.5556446313858032, |
| "learning_rate": 0.00018492239467849224, |
| "loss": 0.8046, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.2703296703296703, |
| "grad_norm": 0.5751636624336243, |
| "learning_rate": 0.0001844789356984479, |
| "loss": 0.9647, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.27692307692307694, |
| "grad_norm": 0.8823965191841125, |
| "learning_rate": 0.00018403547671840356, |
| "loss": 0.914, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.2835164835164835, |
| "grad_norm": 0.4730704128742218, |
| "learning_rate": 0.0001835920177383592, |
| "loss": 0.8514, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.29010989010989013, |
| "grad_norm": 0.527775764465332, |
| "learning_rate": 0.00018314855875831487, |
| "loss": 0.8588, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.2967032967032967, |
| "grad_norm": 1.2125602960586548, |
| "learning_rate": 0.00018270509977827052, |
| "loss": 0.9817, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.3032967032967033, |
| "grad_norm": 0.46695390343666077, |
| "learning_rate": 0.00018226164079822616, |
| "loss": 0.7875, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.3098901098901099, |
| "grad_norm": 0.45550069212913513, |
| "learning_rate": 0.00018181818181818183, |
| "loss": 0.7247, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.31648351648351647, |
| "grad_norm": 0.5238901972770691, |
| "learning_rate": 0.00018137472283813748, |
| "loss": 0.788, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.3230769230769231, |
| "grad_norm": 0.5668091773986816, |
| "learning_rate": 0.00018093126385809312, |
| "loss": 0.8173, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.32967032967032966, |
| "grad_norm": 0.5564643740653992, |
| "learning_rate": 0.0001804878048780488, |
| "loss": 0.9231, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.3362637362637363, |
| "grad_norm": 0.4617767035961151, |
| "learning_rate": 0.00018004434589800444, |
| "loss": 0.8297, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.34285714285714286, |
| "grad_norm": 0.47752270102500916, |
| "learning_rate": 0.00017960088691796008, |
| "loss": 0.7544, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.34945054945054943, |
| "grad_norm": 0.535720705986023, |
| "learning_rate": 0.00017915742793791575, |
| "loss": 0.9407, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.35604395604395606, |
| "grad_norm": 0.47505563497543335, |
| "learning_rate": 0.0001787139689578714, |
| "loss": 0.8374, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.3626373626373626, |
| "grad_norm": 0.6126910448074341, |
| "learning_rate": 0.00017827050997782707, |
| "loss": 1.0878, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.36923076923076925, |
| "grad_norm": 0.5779691934585571, |
| "learning_rate": 0.00017782705099778271, |
| "loss": 1.0181, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.3758241758241758, |
| "grad_norm": 0.5656658411026001, |
| "learning_rate": 0.00017738359201773839, |
| "loss": 0.8647, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.3824175824175824, |
| "grad_norm": 0.5828663110733032, |
| "learning_rate": 0.000176940133037694, |
| "loss": 0.7128, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.389010989010989, |
| "grad_norm": 0.5121160745620728, |
| "learning_rate": 0.00017649667405764967, |
| "loss": 0.9738, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.3956043956043956, |
| "grad_norm": 0.621181309223175, |
| "learning_rate": 0.00017605321507760532, |
| "loss": 0.808, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.4021978021978022, |
| "grad_norm": 0.5183335542678833, |
| "learning_rate": 0.000175609756097561, |
| "loss": 0.8703, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.4087912087912088, |
| "grad_norm": 0.6565868258476257, |
| "learning_rate": 0.00017516629711751663, |
| "loss": 0.9312, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.4153846153846154, |
| "grad_norm": 0.6722279191017151, |
| "learning_rate": 0.0001747228381374723, |
| "loss": 1.0824, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.421978021978022, |
| "grad_norm": 0.6071406006813049, |
| "learning_rate": 0.00017427937915742792, |
| "loss": 0.891, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.42857142857142855, |
| "grad_norm": 0.513313889503479, |
| "learning_rate": 0.0001738359201773836, |
| "loss": 0.742, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.4351648351648352, |
| "grad_norm": 0.5047631859779358, |
| "learning_rate": 0.00017339246119733924, |
| "loss": 0.9423, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.44175824175824174, |
| "grad_norm": 0.5646599531173706, |
| "learning_rate": 0.0001729490022172949, |
| "loss": 0.8566, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.44835164835164837, |
| "grad_norm": 0.5328014492988586, |
| "learning_rate": 0.00017250554323725056, |
| "loss": 0.8397, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.45494505494505494, |
| "grad_norm": 0.6697370409965515, |
| "learning_rate": 0.00017206208425720623, |
| "loss": 0.9044, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.46153846153846156, |
| "grad_norm": 0.7905986309051514, |
| "learning_rate": 0.00017161862527716187, |
| "loss": 0.9489, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.46813186813186813, |
| "grad_norm": 0.5284698605537415, |
| "learning_rate": 0.00017117516629711752, |
| "loss": 0.7049, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.4747252747252747, |
| "grad_norm": 0.5368447303771973, |
| "learning_rate": 0.0001707317073170732, |
| "loss": 0.8274, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.48131868131868133, |
| "grad_norm": 0.5713925361633301, |
| "learning_rate": 0.00017028824833702883, |
| "loss": 0.6653, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.4879120879120879, |
| "grad_norm": 0.5431435108184814, |
| "learning_rate": 0.0001698447893569845, |
| "loss": 0.9738, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.4945054945054945, |
| "grad_norm": 0.5508645176887512, |
| "learning_rate": 0.00016940133037694015, |
| "loss": 0.7946, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.5010989010989011, |
| "grad_norm": 0.5687035918235779, |
| "learning_rate": 0.00016895787139689582, |
| "loss": 1.1262, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.5076923076923077, |
| "grad_norm": 0.499746710062027, |
| "learning_rate": 0.00016851441241685144, |
| "loss": 1.0047, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.5142857142857142, |
| "grad_norm": 0.59474116563797, |
| "learning_rate": 0.0001680709534368071, |
| "loss": 1.1529, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.5208791208791209, |
| "grad_norm": 0.6204425096511841, |
| "learning_rate": 0.00016762749445676275, |
| "loss": 0.8883, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.5274725274725275, |
| "grad_norm": 0.5115718245506287, |
| "learning_rate": 0.00016718403547671842, |
| "loss": 0.8317, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.5340659340659341, |
| "grad_norm": 0.5380851030349731, |
| "learning_rate": 0.00016674057649667407, |
| "loss": 0.9325, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.5406593406593406, |
| "grad_norm": 0.5884971618652344, |
| "learning_rate": 0.00016629711751662974, |
| "loss": 1.0196, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.5472527472527473, |
| "grad_norm": 0.6122893691062927, |
| "learning_rate": 0.00016585365853658536, |
| "loss": 0.787, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.5538461538461539, |
| "grad_norm": 0.5914639830589294, |
| "learning_rate": 0.00016541019955654103, |
| "loss": 0.928, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.5604395604395604, |
| "grad_norm": 0.6377210021018982, |
| "learning_rate": 0.00016496674057649667, |
| "loss": 0.953, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.567032967032967, |
| "grad_norm": 0.5960861444473267, |
| "learning_rate": 0.00016452328159645234, |
| "loss": 0.8375, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.5736263736263736, |
| "grad_norm": 0.5654808282852173, |
| "learning_rate": 0.000164079822616408, |
| "loss": 0.7931, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.5802197802197803, |
| "grad_norm": 0.6073900461196899, |
| "learning_rate": 0.00016363636363636366, |
| "loss": 1.0419, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.5868131868131868, |
| "grad_norm": 0.5416488647460938, |
| "learning_rate": 0.0001631929046563193, |
| "loss": 0.8963, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.5934065934065934, |
| "grad_norm": 0.48589223623275757, |
| "learning_rate": 0.00016274944567627495, |
| "loss": 0.8119, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 0.5589356422424316, |
| "learning_rate": 0.0001623059866962306, |
| "loss": 0.9856, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.6065934065934065, |
| "grad_norm": 0.5352998375892639, |
| "learning_rate": 0.00016186252771618626, |
| "loss": 0.7669, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.6131868131868132, |
| "grad_norm": 0.6389907002449036, |
| "learning_rate": 0.0001614190687361419, |
| "loss": 0.7909, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.6197802197802198, |
| "grad_norm": 0.7931464314460754, |
| "learning_rate": 0.00016097560975609758, |
| "loss": 1.1146, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.6263736263736264, |
| "grad_norm": 0.6886119246482849, |
| "learning_rate": 0.00016053215077605322, |
| "loss": 0.8083, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.6329670329670329, |
| "grad_norm": 0.5980899333953857, |
| "learning_rate": 0.00016008869179600887, |
| "loss": 0.8838, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.6395604395604395, |
| "grad_norm": 0.6072768568992615, |
| "learning_rate": 0.00015964523281596454, |
| "loss": 0.9462, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.6461538461538462, |
| "grad_norm": 0.5868003964424133, |
| "learning_rate": 0.00015920177383592018, |
| "loss": 0.8632, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.6527472527472528, |
| "grad_norm": 0.7429962158203125, |
| "learning_rate": 0.00015875831485587586, |
| "loss": 1.1878, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.6593406593406593, |
| "grad_norm": 0.49301764369010925, |
| "learning_rate": 0.0001583148558758315, |
| "loss": 0.86, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.6659340659340659, |
| "grad_norm": 0.4976905584335327, |
| "learning_rate": 0.00015787139689578714, |
| "loss": 0.9493, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.6725274725274726, |
| "grad_norm": 0.5512967705726624, |
| "learning_rate": 0.0001574279379157428, |
| "loss": 0.9698, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.6791208791208792, |
| "grad_norm": 0.5553017258644104, |
| "learning_rate": 0.00015698447893569846, |
| "loss": 0.9741, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.6857142857142857, |
| "grad_norm": 0.5939576625823975, |
| "learning_rate": 0.0001565410199556541, |
| "loss": 0.8432, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.6923076923076923, |
| "grad_norm": 0.576961100101471, |
| "learning_rate": 0.00015609756097560978, |
| "loss": 0.7377, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.6989010989010989, |
| "grad_norm": 0.6039689779281616, |
| "learning_rate": 0.00015565410199556542, |
| "loss": 1.1159, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.7054945054945055, |
| "grad_norm": 0.650936484336853, |
| "learning_rate": 0.00015521064301552106, |
| "loss": 0.8777, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.7120879120879121, |
| "grad_norm": 0.5367209315299988, |
| "learning_rate": 0.0001547671840354767, |
| "loss": 0.9016, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.7186813186813187, |
| "grad_norm": 0.5427932143211365, |
| "learning_rate": 0.00015432372505543238, |
| "loss": 0.7885, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.7252747252747253, |
| "grad_norm": 0.5590482950210571, |
| "learning_rate": 0.00015388026607538802, |
| "loss": 0.831, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.7318681318681318, |
| "grad_norm": 0.6897863745689392, |
| "learning_rate": 0.0001534368070953437, |
| "loss": 1.0135, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.7384615384615385, |
| "grad_norm": 0.5914762020111084, |
| "learning_rate": 0.00015299334811529934, |
| "loss": 0.881, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.7450549450549451, |
| "grad_norm": 0.8431822061538696, |
| "learning_rate": 0.00015254988913525498, |
| "loss": 0.8999, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.7516483516483516, |
| "grad_norm": 0.5025018453598022, |
| "learning_rate": 0.00015210643015521066, |
| "loss": 0.771, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.7582417582417582, |
| "grad_norm": 0.6190459728240967, |
| "learning_rate": 0.0001516629711751663, |
| "loss": 0.8828, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.7648351648351648, |
| "grad_norm": 0.5529432892799377, |
| "learning_rate": 0.00015121951219512197, |
| "loss": 0.7698, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.7714285714285715, |
| "grad_norm": 0.5121442675590515, |
| "learning_rate": 0.00015077605321507762, |
| "loss": 0.8353, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.778021978021978, |
| "grad_norm": 0.5955220460891724, |
| "learning_rate": 0.0001503325942350333, |
| "loss": 0.6533, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.7846153846153846, |
| "grad_norm": 0.59442538022995, |
| "learning_rate": 0.0001498891352549889, |
| "loss": 0.7851, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.7912087912087912, |
| "grad_norm": 0.5546603798866272, |
| "learning_rate": 0.00014944567627494458, |
| "loss": 1.1005, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.7978021978021979, |
| "grad_norm": 0.5451670289039612, |
| "learning_rate": 0.00014900221729490022, |
| "loss": 0.7612, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.8043956043956044, |
| "grad_norm": 0.5197620391845703, |
| "learning_rate": 0.0001485587583148559, |
| "loss": 0.837, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.810989010989011, |
| "grad_norm": 0.5600094199180603, |
| "learning_rate": 0.00014811529933481154, |
| "loss": 0.8021, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.8175824175824176, |
| "grad_norm": 0.6268289685249329, |
| "learning_rate": 0.0001476718403547672, |
| "loss": 0.8966, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.8241758241758241, |
| "grad_norm": 0.5550496578216553, |
| "learning_rate": 0.00014722838137472282, |
| "loss": 0.8343, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.8307692307692308, |
| "grad_norm": 0.6128094792366028, |
| "learning_rate": 0.0001467849223946785, |
| "loss": 1.1045, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.8373626373626374, |
| "grad_norm": 0.5269573926925659, |
| "learning_rate": 0.00014634146341463414, |
| "loss": 0.7257, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.843956043956044, |
| "grad_norm": 0.6352261304855347, |
| "learning_rate": 0.0001458980044345898, |
| "loss": 0.7755, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.8505494505494505, |
| "grad_norm": 0.5489485263824463, |
| "learning_rate": 0.00014545454545454546, |
| "loss": 0.7397, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.8571428571428571, |
| "grad_norm": 0.5395745038986206, |
| "learning_rate": 0.00014501108647450113, |
| "loss": 0.7006, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.8637362637362638, |
| "grad_norm": 0.567113995552063, |
| "learning_rate": 0.00014456762749445675, |
| "loss": 0.7323, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.8703296703296703, |
| "grad_norm": 0.5484573841094971, |
| "learning_rate": 0.00014412416851441242, |
| "loss": 0.7767, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.8769230769230769, |
| "grad_norm": 0.5116643309593201, |
| "learning_rate": 0.00014368070953436806, |
| "loss": 0.8164, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.8835164835164835, |
| "grad_norm": 0.5965120196342468, |
| "learning_rate": 0.00014323725055432373, |
| "loss": 0.9261, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.8901098901098901, |
| "grad_norm": 0.619005560874939, |
| "learning_rate": 0.00014279379157427938, |
| "loss": 0.8345, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.8967032967032967, |
| "grad_norm": 0.6372644305229187, |
| "learning_rate": 0.00014235033259423505, |
| "loss": 0.8322, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.9032967032967033, |
| "grad_norm": 0.5704457759857178, |
| "learning_rate": 0.0001419068736141907, |
| "loss": 0.7773, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.9098901098901099, |
| "grad_norm": 0.6067429184913635, |
| "learning_rate": 0.00014146341463414634, |
| "loss": 0.9886, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.9164835164835164, |
| "grad_norm": 0.5782042145729065, |
| "learning_rate": 0.000141019955654102, |
| "loss": 0.8481, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.9230769230769231, |
| "grad_norm": 0.5294620990753174, |
| "learning_rate": 0.00014057649667405765, |
| "loss": 0.8596, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.9296703296703297, |
| "grad_norm": 0.5170754194259644, |
| "learning_rate": 0.00014013303769401332, |
| "loss": 0.8392, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.9362637362637363, |
| "grad_norm": 0.6627360582351685, |
| "learning_rate": 0.00013968957871396897, |
| "loss": 1.0295, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.9428571428571428, |
| "grad_norm": 0.5979954600334167, |
| "learning_rate": 0.00013924611973392464, |
| "loss": 0.8471, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.9494505494505494, |
| "grad_norm": 0.5813125371932983, |
| "learning_rate": 0.00013880266075388026, |
| "loss": 0.8533, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.9560439560439561, |
| "grad_norm": 0.6231347918510437, |
| "learning_rate": 0.00013835920177383593, |
| "loss": 0.8619, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.9626373626373627, |
| "grad_norm": 0.5202571749687195, |
| "learning_rate": 0.00013791574279379157, |
| "loss": 0.6541, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.9692307692307692, |
| "grad_norm": 0.5599802732467651, |
| "learning_rate": 0.00013747228381374724, |
| "loss": 0.7213, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.9758241758241758, |
| "grad_norm": 0.5238470435142517, |
| "learning_rate": 0.0001370288248337029, |
| "loss": 0.7997, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.9824175824175824, |
| "grad_norm": 0.6066797971725464, |
| "learning_rate": 0.00013658536585365856, |
| "loss": 0.7616, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.989010989010989, |
| "grad_norm": 0.5322993993759155, |
| "learning_rate": 0.00013614190687361418, |
| "loss": 0.9426, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.9956043956043956, |
| "grad_norm": 0.5406814813613892, |
| "learning_rate": 0.00013569844789356985, |
| "loss": 0.9232, |
| "step": 151 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.6926743984222412, |
| "learning_rate": 0.0001352549889135255, |
| "loss": 0.7743, |
| "step": 152 |
| }, |
| { |
| "epoch": 1.0065934065934066, |
| "grad_norm": 0.6528266668319702, |
| "learning_rate": 0.00013481152993348116, |
| "loss": 0.8807, |
| "step": 153 |
| }, |
| { |
| "epoch": 1.0131868131868131, |
| "grad_norm": 0.5777136087417603, |
| "learning_rate": 0.0001343680709534368, |
| "loss": 0.749, |
| "step": 154 |
| }, |
| { |
| "epoch": 1.0197802197802197, |
| "grad_norm": 0.5486526489257812, |
| "learning_rate": 0.00013392461197339248, |
| "loss": 0.8516, |
| "step": 155 |
| }, |
| { |
| "epoch": 1.0263736263736263, |
| "grad_norm": 0.5391802787780762, |
| "learning_rate": 0.00013348115299334812, |
| "loss": 0.7205, |
| "step": 156 |
| }, |
| { |
| "epoch": 1.032967032967033, |
| "grad_norm": 0.5297644734382629, |
| "learning_rate": 0.00013303769401330377, |
| "loss": 0.6941, |
| "step": 157 |
| }, |
| { |
| "epoch": 1.0395604395604396, |
| "grad_norm": 0.5092700123786926, |
| "learning_rate": 0.00013259423503325944, |
| "loss": 0.5409, |
| "step": 158 |
| }, |
| { |
| "epoch": 1.0461538461538462, |
| "grad_norm": 0.5707080364227295, |
| "learning_rate": 0.00013215077605321509, |
| "loss": 0.6905, |
| "step": 159 |
| }, |
| { |
| "epoch": 1.0527472527472528, |
| "grad_norm": 0.606627881526947, |
| "learning_rate": 0.00013170731707317076, |
| "loss": 0.7107, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.0593406593406594, |
| "grad_norm": 0.5382817387580872, |
| "learning_rate": 0.0001312638580931264, |
| "loss": 0.7526, |
| "step": 161 |
| }, |
| { |
| "epoch": 1.065934065934066, |
| "grad_norm": 0.558792769908905, |
| "learning_rate": 0.00013082039911308205, |
| "loss": 0.7253, |
| "step": 162 |
| }, |
| { |
| "epoch": 1.0725274725274725, |
| "grad_norm": 0.5459743142127991, |
| "learning_rate": 0.0001303769401330377, |
| "loss": 0.6518, |
| "step": 163 |
| }, |
| { |
| "epoch": 1.079120879120879, |
| "grad_norm": 0.5695344805717468, |
| "learning_rate": 0.00012993348115299336, |
| "loss": 0.6852, |
| "step": 164 |
| }, |
| { |
| "epoch": 1.0857142857142856, |
| "grad_norm": 0.6128743886947632, |
| "learning_rate": 0.000129490022172949, |
| "loss": 0.6762, |
| "step": 165 |
| }, |
| { |
| "epoch": 1.0923076923076924, |
| "grad_norm": 0.6627138257026672, |
| "learning_rate": 0.00012904656319290468, |
| "loss": 1.0677, |
| "step": 166 |
| }, |
| { |
| "epoch": 1.098901098901099, |
| "grad_norm": 0.7332176566123962, |
| "learning_rate": 0.00012860310421286032, |
| "loss": 0.8847, |
| "step": 167 |
| }, |
| { |
| "epoch": 1.1054945054945056, |
| "grad_norm": 0.5791655778884888, |
| "learning_rate": 0.00012815964523281597, |
| "loss": 0.746, |
| "step": 168 |
| }, |
| { |
| "epoch": 1.1120879120879121, |
| "grad_norm": 0.6921065449714661, |
| "learning_rate": 0.0001277161862527716, |
| "loss": 0.6869, |
| "step": 169 |
| }, |
| { |
| "epoch": 1.1186813186813187, |
| "grad_norm": 0.5876157879829407, |
| "learning_rate": 0.00012727272727272728, |
| "loss": 0.6389, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.1252747252747253, |
| "grad_norm": 0.6788483262062073, |
| "learning_rate": 0.00012682926829268293, |
| "loss": 0.6412, |
| "step": 171 |
| }, |
| { |
| "epoch": 1.1318681318681318, |
| "grad_norm": 0.5909616947174072, |
| "learning_rate": 0.0001263858093126386, |
| "loss": 0.6374, |
| "step": 172 |
| }, |
| { |
| "epoch": 1.1384615384615384, |
| "grad_norm": 0.6279222369194031, |
| "learning_rate": 0.00012594235033259424, |
| "loss": 0.7105, |
| "step": 173 |
| }, |
| { |
| "epoch": 1.145054945054945, |
| "grad_norm": 0.5922663807868958, |
| "learning_rate": 0.00012549889135254989, |
| "loss": 0.7109, |
| "step": 174 |
| }, |
| { |
| "epoch": 1.1516483516483516, |
| "grad_norm": 0.7478090524673462, |
| "learning_rate": 0.00012505543237250553, |
| "loss": 0.677, |
| "step": 175 |
| }, |
| { |
| "epoch": 1.1582417582417581, |
| "grad_norm": 0.5965471863746643, |
| "learning_rate": 0.0001246119733924612, |
| "loss": 0.7968, |
| "step": 176 |
| }, |
| { |
| "epoch": 1.164835164835165, |
| "grad_norm": 0.6833251714706421, |
| "learning_rate": 0.00012416851441241685, |
| "loss": 0.8221, |
| "step": 177 |
| }, |
| { |
| "epoch": 1.1714285714285715, |
| "grad_norm": 0.6009984016418457, |
| "learning_rate": 0.00012372505543237252, |
| "loss": 0.5564, |
| "step": 178 |
| }, |
| { |
| "epoch": 1.178021978021978, |
| "grad_norm": 0.7059131264686584, |
| "learning_rate": 0.00012328159645232816, |
| "loss": 0.9398, |
| "step": 179 |
| }, |
| { |
| "epoch": 1.1846153846153846, |
| "grad_norm": 0.6330722570419312, |
| "learning_rate": 0.0001228381374722838, |
| "loss": 0.6585, |
| "step": 180 |
| }, |
| { |
| "epoch": 1.1912087912087912, |
| "grad_norm": 0.754753053188324, |
| "learning_rate": 0.00012239467849223948, |
| "loss": 0.7682, |
| "step": 181 |
| }, |
| { |
| "epoch": 1.1978021978021978, |
| "grad_norm": 0.6728742122650146, |
| "learning_rate": 0.00012195121951219512, |
| "loss": 0.7416, |
| "step": 182 |
| }, |
| { |
| "epoch": 1.2043956043956043, |
| "grad_norm": 0.646062970161438, |
| "learning_rate": 0.00012150776053215078, |
| "loss": 0.7183, |
| "step": 183 |
| }, |
| { |
| "epoch": 1.210989010989011, |
| "grad_norm": 0.6689327955245972, |
| "learning_rate": 0.00012106430155210644, |
| "loss": 0.7409, |
| "step": 184 |
| }, |
| { |
| "epoch": 1.2175824175824177, |
| "grad_norm": 0.6284791231155396, |
| "learning_rate": 0.0001206208425720621, |
| "loss": 0.7137, |
| "step": 185 |
| }, |
| { |
| "epoch": 1.2241758241758243, |
| "grad_norm": 0.6613520383834839, |
| "learning_rate": 0.00012017738359201774, |
| "loss": 0.7389, |
| "step": 186 |
| }, |
| { |
| "epoch": 1.2307692307692308, |
| "grad_norm": 0.855766773223877, |
| "learning_rate": 0.0001197339246119734, |
| "loss": 0.8416, |
| "step": 187 |
| }, |
| { |
| "epoch": 1.2373626373626374, |
| "grad_norm": 0.7242646217346191, |
| "learning_rate": 0.00011929046563192906, |
| "loss": 0.6823, |
| "step": 188 |
| }, |
| { |
| "epoch": 1.243956043956044, |
| "grad_norm": 0.7441504597663879, |
| "learning_rate": 0.00011884700665188471, |
| "loss": 0.7099, |
| "step": 189 |
| }, |
| { |
| "epoch": 1.2505494505494505, |
| "grad_norm": 0.6454845070838928, |
| "learning_rate": 0.00011840354767184036, |
| "loss": 0.627, |
| "step": 190 |
| }, |
| { |
| "epoch": 1.2571428571428571, |
| "grad_norm": 0.6478668451309204, |
| "learning_rate": 0.00011796008869179602, |
| "loss": 0.5941, |
| "step": 191 |
| }, |
| { |
| "epoch": 1.2637362637362637, |
| "grad_norm": 0.7972174882888794, |
| "learning_rate": 0.00011751662971175166, |
| "loss": 0.7645, |
| "step": 192 |
| }, |
| { |
| "epoch": 1.2703296703296703, |
| "grad_norm": 0.8269811868667603, |
| "learning_rate": 0.00011707317073170732, |
| "loss": 0.7868, |
| "step": 193 |
| }, |
| { |
| "epoch": 1.2769230769230768, |
| "grad_norm": 0.7218356728553772, |
| "learning_rate": 0.00011662971175166298, |
| "loss": 0.6267, |
| "step": 194 |
| }, |
| { |
| "epoch": 1.2835164835164834, |
| "grad_norm": 0.585889458656311, |
| "learning_rate": 0.00011618625277161863, |
| "loss": 0.4764, |
| "step": 195 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 456, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 15, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2.0405267076071424e+16, |
| "train_batch_size": 22, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
