Invalid JSON:Unexpected token 'I', ..."ad_norm": Infinity,
"... is not valid JSON
| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.016, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 8e-05, | |
| "grad_norm": 6.928949356079102, | |
| "learning_rate": 0.0, | |
| "loss": 12.014674186706543, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00016, | |
| "grad_norm": 8.920316696166992, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 15.713064193725586, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00024, | |
| "grad_norm": 8.561675071716309, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 13.12899398803711, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00032, | |
| "grad_norm": 8.729342460632324, | |
| "learning_rate": 0.0001, | |
| "loss": 16.589990615844727, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0004, | |
| "grad_norm": 7.973804950714111, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 13.032384872436523, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00048, | |
| "grad_norm": 6.11670446395874, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 8.09074592590332, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00056, | |
| "grad_norm": 12.821895599365234, | |
| "learning_rate": 0.0002, | |
| "loss": 13.565499305725098, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.00064, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019998688836656323, | |
| "loss": 9.580286026000977, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00072, | |
| "grad_norm": 10.717191696166992, | |
| "learning_rate": 0.00019998688836656323, | |
| "loss": 11.118619918823242, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0008, | |
| "grad_norm": 15.189062118530273, | |
| "learning_rate": 0.00019994755690455152, | |
| "loss": 10.562832832336426, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00088, | |
| "grad_norm": 11.298067092895508, | |
| "learning_rate": 0.0001998820159279591, | |
| "loss": 8.768858909606934, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.00096, | |
| "grad_norm": 10.482418060302734, | |
| "learning_rate": 0.00019979028262377118, | |
| "loss": 6.118743419647217, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.00104, | |
| "grad_norm": 13.8955078125, | |
| "learning_rate": 0.00019967238104745696, | |
| "loss": 6.71580696105957, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.00112, | |
| "grad_norm": 8.358887672424316, | |
| "learning_rate": 0.0001995283421166614, | |
| "loss": 6.379769325256348, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0012, | |
| "grad_norm": 11.862845420837402, | |
| "learning_rate": 0.00019935820360309777, | |
| "loss": 6.037031173706055, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.00128, | |
| "grad_norm": 8.839066505432129, | |
| "learning_rate": 0.00019916201012264254, | |
| "loss": 5.507349014282227, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.00136, | |
| "grad_norm": 8.00528621673584, | |
| "learning_rate": 0.00019893981312363562, | |
| "loss": 5.974639892578125, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.00144, | |
| "grad_norm": 5.869845867156982, | |
| "learning_rate": 0.00019869167087338907, | |
| "loss": 5.14662504196167, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.00152, | |
| "grad_norm": 7.9579997062683105, | |
| "learning_rate": 0.00019841764844290744, | |
| "loss": 5.114725112915039, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0016, | |
| "grad_norm": 6.74857234954834, | |
| "learning_rate": 0.0001981178176898239, | |
| "loss": 4.387443542480469, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.00168, | |
| "grad_norm": 4.906152725219727, | |
| "learning_rate": 0.00019779225723955707, | |
| "loss": 3.8158411979675293, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.00176, | |
| "grad_norm": 4.70120906829834, | |
| "learning_rate": 0.00019744105246469263, | |
| "loss": 3.4386539459228516, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.00184, | |
| "grad_norm": 3.952915906906128, | |
| "learning_rate": 0.00019706429546259593, | |
| "loss": 3.0041065216064453, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.00192, | |
| "grad_norm": 15.562743186950684, | |
| "learning_rate": 0.00019666208503126112, | |
| "loss": 4.854756832122803, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.002, | |
| "grad_norm": 4.657861709594727, | |
| "learning_rate": 0.00019623452664340306, | |
| "loss": 3.153581142425537, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.00208, | |
| "grad_norm": 7.487373352050781, | |
| "learning_rate": 0.00019578173241879872, | |
| "loss": 3.5299084186553955, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.00216, | |
| "grad_norm": 6.073024749755859, | |
| "learning_rate": 0.0001953038210948861, | |
| "loss": 2.894967555999756, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.00224, | |
| "grad_norm": 8.635567665100098, | |
| "learning_rate": 0.00019480091799562704, | |
| "loss": 3.047236680984497, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.00232, | |
| "grad_norm": 6.574955463409424, | |
| "learning_rate": 0.00019427315499864344, | |
| "loss": 3.163996696472168, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0024, | |
| "grad_norm": 8.933237075805664, | |
| "learning_rate": 0.00019372067050063438, | |
| "loss": 3.112985372543335, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.00248, | |
| "grad_norm": 6.229989528656006, | |
| "learning_rate": 0.00019314360938108425, | |
| "loss": 3.183608055114746, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.00256, | |
| "grad_norm": 4.343419551849365, | |
| "learning_rate": 0.00019254212296427044, | |
| "loss": 1.5173087120056152, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.00264, | |
| "grad_norm": 4.703884601593018, | |
| "learning_rate": 0.00019191636897958122, | |
| "loss": 2.5797841548919678, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.00272, | |
| "grad_norm": 4.645837783813477, | |
| "learning_rate": 0.00019126651152015403, | |
| "loss": 3.0819389820098877, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0028, | |
| "grad_norm": 4.055941581726074, | |
| "learning_rate": 0.0001905927209998447, | |
| "loss": 2.445788860321045, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.00288, | |
| "grad_norm": 3.485275983810425, | |
| "learning_rate": 0.00018989517410853955, | |
| "loss": 2.1609947681427, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.00296, | |
| "grad_norm": 2.7206485271453857, | |
| "learning_rate": 0.00018917405376582145, | |
| "loss": 1.184091567993164, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.00304, | |
| "grad_norm": 4.22371244430542, | |
| "learning_rate": 0.00018842954907300236, | |
| "loss": 2.033668279647827, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.00312, | |
| "grad_norm": 6.188478469848633, | |
| "learning_rate": 0.0001876618552635348, | |
| "loss": 3.973464012145996, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0032, | |
| "grad_norm": 5.991054058074951, | |
| "learning_rate": 0.00018687117365181512, | |
| "loss": 3.453714370727539, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.00328, | |
| "grad_norm": 3.2771706581115723, | |
| "learning_rate": 0.00018605771158039253, | |
| "loss": 2.016055107116699, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.00336, | |
| "grad_norm": 2.86710262298584, | |
| "learning_rate": 0.00018522168236559695, | |
| "loss": 1.0825953483581543, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.00344, | |
| "grad_norm": 3.313237190246582, | |
| "learning_rate": 0.00018436330524160047, | |
| "loss": 2.536651134490967, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.00352, | |
| "grad_norm": 4.080451488494873, | |
| "learning_rate": 0.00018348280530292713, | |
| "loss": 3.4808599948883057, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0036, | |
| "grad_norm": 3.049909830093384, | |
| "learning_rate": 0.00018258041344542566, | |
| "loss": 2.493288516998291, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.00368, | |
| "grad_norm": 6.047203540802002, | |
| "learning_rate": 0.0001816563663057211, | |
| "loss": 3.421189308166504, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.00376, | |
| "grad_norm": 6.8579206466674805, | |
| "learning_rate": 0.00018071090619916093, | |
| "loss": 4.0107502937316895, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.00384, | |
| "grad_norm": 8.001409530639648, | |
| "learning_rate": 0.00017974428105627208, | |
| "loss": 1.8383188247680664, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.00392, | |
| "grad_norm": 5.185579776763916, | |
| "learning_rate": 0.00017875674435774547, | |
| "loss": 2.074155807495117, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.004, | |
| "grad_norm": 3.5374512672424316, | |
| "learning_rate": 0.00017774855506796496, | |
| "loss": 1.4657212495803833, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.00408, | |
| "grad_norm": 3.688544750213623, | |
| "learning_rate": 0.00017671997756709863, | |
| "loss": 1.6368293762207031, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.00416, | |
| "grad_norm": 5.878078937530518, | |
| "learning_rate": 0.00017567128158176953, | |
| "loss": 2.581313133239746, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.00424, | |
| "grad_norm": 4.422114372253418, | |
| "learning_rate": 0.0001746027421143246, | |
| "loss": 2.660783290863037, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.00432, | |
| "grad_norm": 5.147724151611328, | |
| "learning_rate": 0.00017351463937072004, | |
| "loss": 2.2990241050720215, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0044, | |
| "grad_norm": 6.18672513961792, | |
| "learning_rate": 0.00017240725868704218, | |
| "loss": 1.4074926376342773, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.00448, | |
| "grad_norm": 5.2075042724609375, | |
| "learning_rate": 0.00017128089045468294, | |
| "loss": 1.5134844779968262, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.00456, | |
| "grad_norm": 3.1878983974456787, | |
| "learning_rate": 0.00017013583004418993, | |
| "loss": 2.699907064437866, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.00464, | |
| "grad_norm": 3.742098093032837, | |
| "learning_rate": 0.00016897237772781044, | |
| "loss": 1.6386728286743164, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.00472, | |
| "grad_norm": 2.6552226543426514, | |
| "learning_rate": 0.00016779083860075033, | |
| "loss": 1.1386427879333496, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0048, | |
| "grad_norm": 2.4843499660491943, | |
| "learning_rate": 0.00016659152250116812, | |
| "loss": 0.8249317407608032, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.00488, | |
| "grad_norm": 4.540428161621094, | |
| "learning_rate": 0.00016537474392892528, | |
| "loss": 3.6220569610595703, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.00496, | |
| "grad_norm": 4.440019130706787, | |
| "learning_rate": 0.000164140821963114, | |
| "loss": 2.0882022380828857, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.00504, | |
| "grad_norm": 5.714387893676758, | |
| "learning_rate": 0.00016289008017838445, | |
| "loss": 2.471445083618164, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.00512, | |
| "grad_norm": 3.9814000129699707, | |
| "learning_rate": 0.00016162284656009274, | |
| "loss": 1.4108766317367554, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.0052, | |
| "grad_norm": 2.511936664581299, | |
| "learning_rate": 0.00016033945341829248, | |
| "loss": 1.2236822843551636, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.00528, | |
| "grad_norm": 3.3108081817626953, | |
| "learning_rate": 0.00015904023730059228, | |
| "loss": 1.9047954082489014, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.00536, | |
| "grad_norm": 3.4142255783081055, | |
| "learning_rate": 0.00015772553890390197, | |
| "loss": 1.8660364151000977, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.00544, | |
| "grad_norm": 5.404369831085205, | |
| "learning_rate": 0.00015639570298509064, | |
| "loss": 2.1426186561584473, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.00552, | |
| "grad_norm": 4.093230724334717, | |
| "learning_rate": 0.00015505107827058036, | |
| "loss": 1.168108344078064, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0056, | |
| "grad_norm": 2.660120725631714, | |
| "learning_rate": 0.0001536920173648984, | |
| "loss": 2.291879892349243, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.00568, | |
| "grad_norm": 4.722398281097412, | |
| "learning_rate": 0.000152318876658213, | |
| "loss": 2.181743860244751, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.00576, | |
| "grad_norm": 2.6253931522369385, | |
| "learning_rate": 0.00015093201623287631, | |
| "loss": 1.9663147926330566, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.00584, | |
| "grad_norm": 3.9038803577423096, | |
| "learning_rate": 0.00014953179976899878, | |
| "loss": 2.0541481971740723, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.00592, | |
| "grad_norm": 6.097304821014404, | |
| "learning_rate": 0.00014811859444908052, | |
| "loss": 1.653533935546875, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.006, | |
| "grad_norm": 3.6705079078674316, | |
| "learning_rate": 0.00014669277086172406, | |
| "loss": 1.9317748546600342, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.00608, | |
| "grad_norm": 4.832096576690674, | |
| "learning_rate": 0.00014525470290445392, | |
| "loss": 1.969446063041687, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.00616, | |
| "grad_norm": 3.0651421546936035, | |
| "learning_rate": 0.00014380476768566824, | |
| "loss": 1.0908013582229614, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.00624, | |
| "grad_norm": 2.731213092803955, | |
| "learning_rate": 0.00014234334542574906, | |
| "loss": 1.504532814025879, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.00632, | |
| "grad_norm": 2.8738863468170166, | |
| "learning_rate": 0.00014087081935735564, | |
| "loss": 1.8904380798339844, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 3.4613351821899414, | |
| "learning_rate": 0.00013938757562492873, | |
| "loss": 2.320128917694092, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.00648, | |
| "grad_norm": 2.597461462020874, | |
| "learning_rate": 0.00013789400318343068, | |
| "loss": 0.9850823879241943, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.00656, | |
| "grad_norm": 4.481573104858398, | |
| "learning_rate": 0.00013639049369634876, | |
| "loss": 1.219327688217163, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.00664, | |
| "grad_norm": 5.184344291687012, | |
| "learning_rate": 0.00013487744143298822, | |
| "loss": 2.01560640335083, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.00672, | |
| "grad_norm": 3.655848979949951, | |
| "learning_rate": 0.00013335524316508208, | |
| "loss": 1.3672316074371338, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.0068, | |
| "grad_norm": 2.1263749599456787, | |
| "learning_rate": 0.0001318242980627444, | |
| "loss": 0.8222290277481079, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.00688, | |
| "grad_norm": 2.7781271934509277, | |
| "learning_rate": 0.00013028500758979506, | |
| "loss": 1.3532121181488037, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.00696, | |
| "grad_norm": 3.8022267818450928, | |
| "learning_rate": 0.00012873777539848283, | |
| "loss": 1.483093023300171, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.00704, | |
| "grad_norm": 4.301313400268555, | |
| "learning_rate": 0.0001271830072236343, | |
| "loss": 2.149625778198242, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.00712, | |
| "grad_norm": 1.896973729133606, | |
| "learning_rate": 0.00012562111077625722, | |
| "loss": 1.3007004261016846, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.0072, | |
| "grad_norm": 3.8339455127716064, | |
| "learning_rate": 0.00012405249563662537, | |
| "loss": 2.0180318355560303, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.00728, | |
| "grad_norm": 2.7834441661834717, | |
| "learning_rate": 0.00012247757314687297, | |
| "loss": 0.9584566354751587, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.00736, | |
| "grad_norm": 4.279788017272949, | |
| "learning_rate": 0.00012089675630312754, | |
| "loss": 2.384021759033203, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.00744, | |
| "grad_norm": 2.1873183250427246, | |
| "learning_rate": 0.00011931045964720881, | |
| "loss": 1.5363960266113281, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.00752, | |
| "grad_norm": 2.2680325508117676, | |
| "learning_rate": 0.0001177190991579223, | |
| "loss": 1.9146153926849365, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0076, | |
| "grad_norm": 6.569868564605713, | |
| "learning_rate": 0.00011612309214197599, | |
| "loss": 3.304594039916992, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.00768, | |
| "grad_norm": 2.8254189491271973, | |
| "learning_rate": 0.00011452285712454904, | |
| "loss": 1.9783649444580078, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.00776, | |
| "grad_norm": 2.8782193660736084, | |
| "learning_rate": 0.00011291881373954065, | |
| "loss": 1.1415950059890747, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.00784, | |
| "grad_norm": 4.276658058166504, | |
| "learning_rate": 0.00011131138261952845, | |
| "loss": 1.7967627048492432, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.00792, | |
| "grad_norm": 1.8674637079238892, | |
| "learning_rate": 0.00010970098528546481, | |
| "loss": 1.3417999744415283, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.008, | |
| "grad_norm": 2.835909366607666, | |
| "learning_rate": 0.00010808804403614043, | |
| "loss": 2.7765414714813232, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.00808, | |
| "grad_norm": 3.2777109146118164, | |
| "learning_rate": 0.00010647298183744359, | |
| "loss": 2.188383102416992, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.00816, | |
| "grad_norm": 2.4123282432556152, | |
| "learning_rate": 0.00010485622221144484, | |
| "loss": 2.901590347290039, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.00824, | |
| "grad_norm": 1.782806634902954, | |
| "learning_rate": 0.00010323818912533561, | |
| "loss": 0.8531882762908936, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.00832, | |
| "grad_norm": 2.6029393672943115, | |
| "learning_rate": 0.00010161930688025017, | |
| "loss": 2.415867805480957, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0084, | |
| "grad_norm": 2.7058908939361572, | |
| "learning_rate": 0.0001, | |
| "loss": 1.4201483726501465, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.00848, | |
| "grad_norm": 2.6447594165802, | |
| "learning_rate": 9.838069311974986e-05, | |
| "loss": 2.2144458293914795, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.00856, | |
| "grad_norm": 3.407778263092041, | |
| "learning_rate": 9.676181087466444e-05, | |
| "loss": 2.8273167610168457, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.00864, | |
| "grad_norm": 3.881974458694458, | |
| "learning_rate": 9.514377778855521e-05, | |
| "loss": 2.314471960067749, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.00872, | |
| "grad_norm": 3.9941506385803223, | |
| "learning_rate": 9.352701816255643e-05, | |
| "loss": 1.9356070756912231, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.0088, | |
| "grad_norm": 3.67986798286438, | |
| "learning_rate": 9.19119559638596e-05, | |
| "loss": 0.8660937547683716, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.00888, | |
| "grad_norm": 3.451327085494995, | |
| "learning_rate": 9.02990147145352e-05, | |
| "loss": 1.8435039520263672, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.00896, | |
| "grad_norm": 4.358551979064941, | |
| "learning_rate": 8.868861738047158e-05, | |
| "loss": 1.2189772129058838, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.00904, | |
| "grad_norm": 2.414902687072754, | |
| "learning_rate": 8.70811862604594e-05, | |
| "loss": 0.8712605237960815, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.00912, | |
| "grad_norm": 2.589571475982666, | |
| "learning_rate": 8.5477142875451e-05, | |
| "loss": 1.763231635093689, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.0092, | |
| "grad_norm": 3.0219295024871826, | |
| "learning_rate": 8.387690785802402e-05, | |
| "loss": 1.11235773563385, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.00928, | |
| "grad_norm": 1.9637902975082397, | |
| "learning_rate": 8.228090084207774e-05, | |
| "loss": 1.076027512550354, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.00936, | |
| "grad_norm": 2.4067959785461426, | |
| "learning_rate": 8.068954035279121e-05, | |
| "loss": 0.7609463930130005, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.00944, | |
| "grad_norm": 2.6905956268310547, | |
| "learning_rate": 7.91032436968725e-05, | |
| "loss": 1.5781116485595703, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.00952, | |
| "grad_norm": 3.6325156688690186, | |
| "learning_rate": 7.75224268531271e-05, | |
| "loss": 1.7823458909988403, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 3.7946057319641113, | |
| "learning_rate": 7.594750436337467e-05, | |
| "loss": 1.5035946369171143, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.00968, | |
| "grad_norm": 2.942676544189453, | |
| "learning_rate": 7.437888922374276e-05, | |
| "loss": 1.8382234573364258, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.00976, | |
| "grad_norm": 2.7739830017089844, | |
| "learning_rate": 7.281699277636572e-05, | |
| "loss": 1.751274585723877, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.00984, | |
| "grad_norm": 2.4406661987304688, | |
| "learning_rate": 7.126222460151719e-05, | |
| "loss": 0.7677597999572754, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.00992, | |
| "grad_norm": 2.1030468940734863, | |
| "learning_rate": 6.971499241020495e-05, | |
| "loss": 2.2280046939849854, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.3427834510803223, | |
| "learning_rate": 6.817570193725564e-05, | |
| "loss": 1.3863165378570557, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.01008, | |
| "grad_norm": 1.9932631254196167, | |
| "learning_rate": 6.664475683491796e-05, | |
| "loss": 0.7340274453163147, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.01016, | |
| "grad_norm": 3.0195388793945312, | |
| "learning_rate": 6.512255856701177e-05, | |
| "loss": 1.4951543807983398, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.01024, | |
| "grad_norm": 2.5189034938812256, | |
| "learning_rate": 6.360950630365126e-05, | |
| "loss": 0.5847827196121216, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.01032, | |
| "grad_norm": 3.0103163719177246, | |
| "learning_rate": 6.210599681656933e-05, | |
| "loss": 1.7431640625, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.0104, | |
| "grad_norm": 2.1604888439178467, | |
| "learning_rate": 6.061242437507131e-05, | |
| "loss": 0.8966880440711975, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.01048, | |
| "grad_norm": 2.067002296447754, | |
| "learning_rate": 5.9129180642644414e-05, | |
| "loss": 1.793744683265686, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.01056, | |
| "grad_norm": 2.590845823287964, | |
| "learning_rate": 5.765665457425102e-05, | |
| "loss": 1.7302422523498535, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.01064, | |
| "grad_norm": 3.48775053024292, | |
| "learning_rate": 5.6195232314331766e-05, | |
| "loss": 1.8507360219955444, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.01072, | |
| "grad_norm": 3.2789645195007324, | |
| "learning_rate": 5.474529709554612e-05, | |
| "loss": 2.3930463790893555, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0108, | |
| "grad_norm": 4.097634792327881, | |
| "learning_rate": 5.3307229138275936e-05, | |
| "loss": 1.4812769889831543, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.01088, | |
| "grad_norm": 3.504568576812744, | |
| "learning_rate": 5.1881405550919493e-05, | |
| "loss": 1.2586313486099243, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.01096, | |
| "grad_norm": 2.2558963298797607, | |
| "learning_rate": 5.0468200231001286e-05, | |
| "loss": 1.2390514612197876, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.01104, | |
| "grad_norm": 3.5328781604766846, | |
| "learning_rate": 4.9067983767123736e-05, | |
| "loss": 1.352766752243042, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.01112, | |
| "grad_norm": 3.3730530738830566, | |
| "learning_rate": 4.768112334178699e-05, | |
| "loss": 2.361686944961548, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.0112, | |
| "grad_norm": 3.0481679439544678, | |
| "learning_rate": 4.630798263510162e-05, | |
| "loss": 1.2501332759857178, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.01128, | |
| "grad_norm": 2.652317523956299, | |
| "learning_rate": 4.494892172941965e-05, | |
| "loss": 1.5895557403564453, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.01136, | |
| "grad_norm": 3.2051403522491455, | |
| "learning_rate": 4.360429701490934e-05, | |
| "loss": 1.1575316190719604, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.01144, | |
| "grad_norm": 3.064737319946289, | |
| "learning_rate": 4.227446109609809e-05, | |
| "loss": 1.1490418910980225, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.01152, | |
| "grad_norm": 3.624948024749756, | |
| "learning_rate": 4.0959762699407766e-05, | |
| "loss": 1.3190464973449707, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.0116, | |
| "grad_norm": 2.7830724716186523, | |
| "learning_rate": 3.966054658170754e-05, | |
| "loss": 2.2048940658569336, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.01168, | |
| "grad_norm": 3.638855218887329, | |
| "learning_rate": 3.8377153439907266e-05, | |
| "loss": 1.3956620693206787, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.01176, | |
| "grad_norm": 2.614692211151123, | |
| "learning_rate": 3.710991982161555e-05, | |
| "loss": 1.9561386108398438, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.01184, | |
| "grad_norm": 2.6306076049804688, | |
| "learning_rate": 3.585917803688603e-05, | |
| "loss": 2.6344668865203857, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.01192, | |
| "grad_norm": 2.817415237426758, | |
| "learning_rate": 3.4625256071074773e-05, | |
| "loss": 1.4757401943206787, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.012, | |
| "grad_norm": 3.840266704559326, | |
| "learning_rate": 3.340847749883191e-05, | |
| "loss": 1.3285486698150635, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.01208, | |
| "grad_norm": 3.7817294597625732, | |
| "learning_rate": 3.2209161399249674e-05, | |
| "loss": 1.2630081176757812, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.01216, | |
| "grad_norm": 3.292306661605835, | |
| "learning_rate": 3.102762227218957e-05, | |
| "loss": 1.7503516674041748, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.01224, | |
| "grad_norm": 2.8818812370300293, | |
| "learning_rate": 2.9864169955810084e-05, | |
| "loss": 1.488126277923584, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.01232, | |
| "grad_norm": 2.7769317626953125, | |
| "learning_rate": 2.8719109545317103e-05, | |
| "loss": 1.9848933219909668, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.0124, | |
| "grad_norm": 2.7237372398376465, | |
| "learning_rate": 2.759274131295787e-05, | |
| "loss": 0.954386293888092, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.01248, | |
| "grad_norm": 2.7596638202667236, | |
| "learning_rate": 2.6485360629279987e-05, | |
| "loss": 0.9503962993621826, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.01256, | |
| "grad_norm": 3.106092691421509, | |
| "learning_rate": 2.5397257885675397e-05, | |
| "loss": 2.1692826747894287, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.01264, | |
| "grad_norm": 3.0916266441345215, | |
| "learning_rate": 2.432871841823047e-05, | |
| "loss": 1.3709118366241455, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.01272, | |
| "grad_norm": 3.4759669303894043, | |
| "learning_rate": 2.3280022432901383e-05, | |
| "loss": 1.4861762523651123, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 2.9472529888153076, | |
| "learning_rate": 2.2251444932035094e-05, | |
| "loss": 1.0365121364593506, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.01288, | |
| "grad_norm": 2.2864620685577393, | |
| "learning_rate": 2.1243255642254578e-05, | |
| "loss": 0.6069309711456299, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.01296, | |
| "grad_norm": 1.494697093963623, | |
| "learning_rate": 2.025571894372794e-05, | |
| "loss": 0.5316790342330933, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.01304, | |
| "grad_norm": 2.833409309387207, | |
| "learning_rate": 1.9289093800839066e-05, | |
| "loss": 1.3331211805343628, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.01312, | |
| "grad_norm": 1.7902228832244873, | |
| "learning_rate": 1.8343633694278895e-05, | |
| "loss": 1.6830205917358398, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.0132, | |
| "grad_norm": 2.8428897857666016, | |
| "learning_rate": 1.741958655457436e-05, | |
| "loss": 1.3496726751327515, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.01328, | |
| "grad_norm": 1.9608421325683594, | |
| "learning_rate": 1.65171946970729e-05, | |
| "loss": 0.5171700716018677, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.01336, | |
| "grad_norm": 3.1063010692596436, | |
| "learning_rate": 1.563669475839956e-05, | |
| "loss": 1.6354092359542847, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.01344, | |
| "grad_norm": 2.7731094360351562, | |
| "learning_rate": 1.4778317634403083e-05, | |
| "loss": 1.6890660524368286, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.01352, | |
| "grad_norm": 2.616074562072754, | |
| "learning_rate": 1.3942288419607475e-05, | |
| "loss": 1.0017638206481934, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.0136, | |
| "grad_norm": 6.030595779418945, | |
| "learning_rate": 1.3128826348184887e-05, | |
| "loss": 2.2325081825256348, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.01368, | |
| "grad_norm": 2.899432420730591, | |
| "learning_rate": 1.233814473646524e-05, | |
| "loss": 1.6379659175872803, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.01376, | |
| "grad_norm": 2.3841640949249268, | |
| "learning_rate": 1.1570450926997655e-05, | |
| "loss": 0.8668946027755737, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.01384, | |
| "grad_norm": 1.8768247365951538, | |
| "learning_rate": 1.0825946234178574e-05, | |
| "loss": 0.6351206302642822, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.01392, | |
| "grad_norm": 4.373176574707031, | |
| "learning_rate": 1.010482589146048e-05, | |
| "loss": 1.6108734607696533, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.014, | |
| "grad_norm": 2.857194185256958, | |
| "learning_rate": 9.407279000155312e-06, | |
| "loss": 2.0244319438934326, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.01408, | |
| "grad_norm": 1.7663400173187256, | |
| "learning_rate": 8.733488479845997e-06, | |
| "loss": 0.5519629716873169, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.01416, | |
| "grad_norm": 2.0455868244171143, | |
| "learning_rate": 8.083631020418791e-06, | |
| "loss": 1.4491122961044312, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.01424, | |
| "grad_norm": 2.3110480308532715, | |
| "learning_rate": 7.457877035729588e-06, | |
| "loss": 0.7039970755577087, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.01432, | |
| "grad_norm": 3.0206592082977295, | |
| "learning_rate": 6.856390618915775e-06, | |
| "loss": 1.0675463676452637, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.0144, | |
| "grad_norm": 2.02497935295105, | |
| "learning_rate": 6.2793294993656494e-06, | |
| "loss": 1.504166841506958, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.01448, | |
| "grad_norm": 2.715156316757202, | |
| "learning_rate": 5.726845001356573e-06, | |
| "loss": 1.9781649112701416, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.01456, | |
| "grad_norm": 2.5113284587860107, | |
| "learning_rate": 5.199082004372957e-06, | |
| "loss": 1.5596675872802734, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.01464, | |
| "grad_norm": 3.491459608078003, | |
| "learning_rate": 4.6961789051139124e-06, | |
| "loss": 1.955146312713623, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.01472, | |
| "grad_norm": 4.690987586975098, | |
| "learning_rate": 4.2182675812012965e-06, | |
| "loss": 1.7040395736694336, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.0148, | |
| "grad_norm": 2.3959999084472656, | |
| "learning_rate": 3.7654733565969826e-06, | |
| "loss": 2.5348751544952393, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.01488, | |
| "grad_norm": 1.3934000730514526, | |
| "learning_rate": 3.3379149687388867e-06, | |
| "loss": 0.42651718854904175, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.01496, | |
| "grad_norm": 2.282785415649414, | |
| "learning_rate": 2.9357045374040825e-06, | |
| "loss": 1.5776516199111938, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.01504, | |
| "grad_norm": 1.904970645904541, | |
| "learning_rate": 2.5589475353073988e-06, | |
| "loss": 1.0290027856826782, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.01512, | |
| "grad_norm": 2.607666254043579, | |
| "learning_rate": 2.2077427604429433e-06, | |
| "loss": 1.63942551612854, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.0152, | |
| "grad_norm": 1.9027960300445557, | |
| "learning_rate": 1.882182310176095e-06, | |
| "loss": 0.957375168800354, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.01528, | |
| "grad_norm": 2.230802297592163, | |
| "learning_rate": 1.5823515570925763e-06, | |
| "loss": 0.6559663414955139, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.01536, | |
| "grad_norm": 2.59029221534729, | |
| "learning_rate": 1.30832912661093e-06, | |
| "loss": 1.4496815204620361, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.01544, | |
| "grad_norm": 2.604372978210449, | |
| "learning_rate": 1.0601868763643996e-06, | |
| "loss": 2.121373414993286, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.01552, | |
| "grad_norm": 2.929568290710449, | |
| "learning_rate": 8.379898773574924e-07, | |
| "loss": 1.1203787326812744, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.0156, | |
| "grad_norm": 2.514376401901245, | |
| "learning_rate": 6.41796396902239e-07, | |
| "loss": 0.9185475707054138, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.01568, | |
| "grad_norm": 2.5948593616485596, | |
| "learning_rate": 4.7165788333860536e-07, | |
| "loss": 0.7908981442451477, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.01576, | |
| "grad_norm": 2.6386489868164062, | |
| "learning_rate": 3.2761895254306287e-07, | |
| "loss": 0.893803060054779, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.01584, | |
| "grad_norm": 1.6470527648925781, | |
| "learning_rate": 2.0971737622883515e-07, | |
| "loss": 0.48487338423728943, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.01592, | |
| "grad_norm": 2.062701940536499, | |
| "learning_rate": 1.179840720409331e-07, | |
| "loss": 0.6738969683647156, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 2.6584014892578125, | |
| "learning_rate": 5.2443095448506674e-08, | |
| "loss": 1.9085880517959595, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1180489435152384.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |