diff --git "a/checkpoint-2500/trainer_state.json" "b/checkpoint-2500/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-2500/trainer_state.json" @@ -0,0 +1,17574 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.7290333736814802, + "eval_steps": 500, + "global_step": 2500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0006916825177243646, + "grad_norm": 0.17447508871555328, + "learning_rate": 0.0, + "loss": 2.4721, + "step": 1 + }, + { + "epoch": 0.0013833650354487291, + "grad_norm": 0.24757890403270721, + "learning_rate": 1.3698630136986302e-06, + "loss": 2.7788, + "step": 2 + }, + { + "epoch": 0.0020750475531730937, + "grad_norm": 0.20169955492019653, + "learning_rate": 2.7397260273972604e-06, + "loss": 2.817, + "step": 3 + }, + { + "epoch": 0.0027667300708974583, + "grad_norm": 0.2597520053386688, + "learning_rate": 4.10958904109589e-06, + "loss": 2.8883, + "step": 4 + }, + { + "epoch": 0.0034584125886218224, + "grad_norm": 0.2476697564125061, + "learning_rate": 5.479452054794521e-06, + "loss": 2.5194, + "step": 5 + }, + { + "epoch": 0.004150095106346187, + "grad_norm": 0.19448921084403992, + "learning_rate": 6.849315068493151e-06, + "loss": 2.0166, + "step": 6 + }, + { + "epoch": 0.0048417776240705515, + "grad_norm": 0.30129846930503845, + "learning_rate": 8.21917808219178e-06, + "loss": 2.0576, + "step": 7 + }, + { + "epoch": 0.0055334601417949165, + "grad_norm": 0.29668128490448, + "learning_rate": 9.589041095890411e-06, + "loss": 2.0307, + "step": 8 + }, + { + "epoch": 0.006225142659519281, + "grad_norm": 0.20030352473258972, + "learning_rate": 1.0958904109589042e-05, + "loss": 2.6198, + "step": 9 + }, + { + "epoch": 0.006916825177243645, + "grad_norm": 0.2806167006492615, + "learning_rate": 1.2328767123287671e-05, + "loss": 2.7184, + "step": 10 + }, + { + "epoch": 0.00760850769496801, + "grad_norm": 0.3155820965766907, + "learning_rate": 1.3698630136986302e-05, + "loss": 2.5599, + "step": 11 + }, + { + "epoch": 0.008300190212692375, + "grad_norm": 0.37342140078544617, + "learning_rate": 1.5068493150684931e-05, + "loss": 1.6116, + "step": 12 + }, + { + "epoch": 0.008991872730416739, + "grad_norm": 0.3110700845718384, + "learning_rate": 1.643835616438356e-05, + "loss": 2.03, + "step": 13 + }, + { + "epoch": 0.009683555248141103, + "grad_norm": 0.22983220219612122, + "learning_rate": 1.780821917808219e-05, + "loss": 3.0022, + "step": 14 + }, + { + "epoch": 0.010375237765865467, + "grad_norm": 0.20027688145637512, + "learning_rate": 1.9178082191780822e-05, + "loss": 2.3026, + "step": 15 + }, + { + "epoch": 0.011066920283589833, + "grad_norm": 0.2387463003396988, + "learning_rate": 2.0547945205479453e-05, + "loss": 2.9633, + "step": 16 + }, + { + "epoch": 0.011758602801314197, + "grad_norm": 0.21235717833042145, + "learning_rate": 2.1917808219178083e-05, + "loss": 2.8526, + "step": 17 + }, + { + "epoch": 0.012450285319038561, + "grad_norm": 0.29500314593315125, + "learning_rate": 2.328767123287671e-05, + "loss": 2.5277, + "step": 18 + }, + { + "epoch": 0.013141967836762926, + "grad_norm": 0.20916298031806946, + "learning_rate": 2.4657534246575342e-05, + "loss": 2.3433, + "step": 19 + }, + { + "epoch": 0.01383365035448729, + "grad_norm": 0.29092636704444885, + "learning_rate": 2.6027397260273973e-05, + "loss": 2.6725, + "step": 20 + }, + { + "epoch": 0.014525332872211656, + "grad_norm": 0.2366064488887787, + "learning_rate": 2.7397260273972603e-05, + "loss": 2.7656, + "step": 21 + }, + { + "epoch": 0.01521701538993602, + "grad_norm": 0.23627543449401855, + "learning_rate": 2.8767123287671234e-05, + "loss": 2.5008, + "step": 22 + }, + { + "epoch": 0.015908697907660384, + "grad_norm": 0.30645817518234253, + "learning_rate": 3.0136986301369862e-05, + "loss": 1.8266, + "step": 23 + }, + { + "epoch": 0.01660038042538475, + "grad_norm": 0.5485689043998718, + "learning_rate": 3.1506849315068496e-05, + "loss": 2.7961, + "step": 24 + }, + { + "epoch": 0.017292062943109112, + "grad_norm": 0.2701663076877594, + "learning_rate": 3.287671232876712e-05, + "loss": 2.7765, + "step": 25 + }, + { + "epoch": 0.017983745460833478, + "grad_norm": 0.28117069602012634, + "learning_rate": 3.424657534246575e-05, + "loss": 2.7523, + "step": 26 + }, + { + "epoch": 0.01867542797855784, + "grad_norm": 0.38187164068222046, + "learning_rate": 3.561643835616438e-05, + "loss": 2.2493, + "step": 27 + }, + { + "epoch": 0.019367110496282206, + "grad_norm": 0.23057831823825836, + "learning_rate": 3.698630136986301e-05, + "loss": 2.8409, + "step": 28 + }, + { + "epoch": 0.020058793014006572, + "grad_norm": 0.314023494720459, + "learning_rate": 3.8356164383561644e-05, + "loss": 2.7994, + "step": 29 + }, + { + "epoch": 0.020750475531730934, + "grad_norm": 0.5843497514724731, + "learning_rate": 3.9726027397260274e-05, + "loss": 2.8692, + "step": 30 + }, + { + "epoch": 0.0214421580494553, + "grad_norm": 0.3582310974597931, + "learning_rate": 4.1095890410958905e-05, + "loss": 2.5466, + "step": 31 + }, + { + "epoch": 0.022133840567179666, + "grad_norm": 0.2534823715686798, + "learning_rate": 4.2465753424657536e-05, + "loss": 2.5193, + "step": 32 + }, + { + "epoch": 0.02282552308490403, + "grad_norm": 0.34849318861961365, + "learning_rate": 4.383561643835617e-05, + "loss": 1.7904, + "step": 33 + }, + { + "epoch": 0.023517205602628394, + "grad_norm": 0.35676291584968567, + "learning_rate": 4.520547945205479e-05, + "loss": 1.6514, + "step": 34 + }, + { + "epoch": 0.024208888120352757, + "grad_norm": 0.24594038724899292, + "learning_rate": 4.657534246575342e-05, + "loss": 2.3192, + "step": 35 + }, + { + "epoch": 0.024900570638077123, + "grad_norm": 0.3562544882297516, + "learning_rate": 4.794520547945205e-05, + "loss": 2.8913, + "step": 36 + }, + { + "epoch": 0.02559225315580149, + "grad_norm": 0.4177575409412384, + "learning_rate": 4.9315068493150684e-05, + "loss": 2.4487, + "step": 37 + }, + { + "epoch": 0.02628393567352585, + "grad_norm": 0.3126346170902252, + "learning_rate": 5.068493150684932e-05, + "loss": 1.4528, + "step": 38 + }, + { + "epoch": 0.026975618191250217, + "grad_norm": 0.30482178926467896, + "learning_rate": 5.2054794520547945e-05, + "loss": 2.6085, + "step": 39 + }, + { + "epoch": 0.02766730070897458, + "grad_norm": 0.31493324041366577, + "learning_rate": 5.342465753424658e-05, + "loss": 1.8078, + "step": 40 + }, + { + "epoch": 0.028358983226698945, + "grad_norm": 0.4693777859210968, + "learning_rate": 5.479452054794521e-05, + "loss": 1.363, + "step": 41 + }, + { + "epoch": 0.02905066574442331, + "grad_norm": 0.6538149118423462, + "learning_rate": 5.616438356164384e-05, + "loss": 1.2077, + "step": 42 + }, + { + "epoch": 0.029742348262147673, + "grad_norm": 0.32594457268714905, + "learning_rate": 5.753424657534247e-05, + "loss": 2.2301, + "step": 43 + }, + { + "epoch": 0.03043403077987204, + "grad_norm": 0.35883522033691406, + "learning_rate": 5.89041095890411e-05, + "loss": 2.5199, + "step": 44 + }, + { + "epoch": 0.0311257132975964, + "grad_norm": 0.3142642080783844, + "learning_rate": 6.0273972602739724e-05, + "loss": 2.0425, + "step": 45 + }, + { + "epoch": 0.03181739581532077, + "grad_norm": 0.31059974431991577, + "learning_rate": 6.164383561643835e-05, + "loss": 2.6123, + "step": 46 + }, + { + "epoch": 0.03250907833304513, + "grad_norm": 0.24734480679035187, + "learning_rate": 6.301369863013699e-05, + "loss": 2.5415, + "step": 47 + }, + { + "epoch": 0.0332007608507695, + "grad_norm": 0.3828027546405792, + "learning_rate": 6.438356164383562e-05, + "loss": 2.3858, + "step": 48 + }, + { + "epoch": 0.03389244336849386, + "grad_norm": 0.32175979018211365, + "learning_rate": 6.575342465753424e-05, + "loss": 2.135, + "step": 49 + }, + { + "epoch": 0.034584125886218224, + "grad_norm": 0.5965486764907837, + "learning_rate": 6.712328767123288e-05, + "loss": 2.2111, + "step": 50 + }, + { + "epoch": 0.03527580840394259, + "grad_norm": 0.5561261177062988, + "learning_rate": 6.84931506849315e-05, + "loss": 1.725, + "step": 51 + }, + { + "epoch": 0.035967490921666956, + "grad_norm": 0.46931296586990356, + "learning_rate": 6.986301369863014e-05, + "loss": 2.5344, + "step": 52 + }, + { + "epoch": 0.03665917343939132, + "grad_norm": 0.4662770926952362, + "learning_rate": 7.123287671232876e-05, + "loss": 2.4031, + "step": 53 + }, + { + "epoch": 0.03735085595711568, + "grad_norm": 0.3315083980560303, + "learning_rate": 7.26027397260274e-05, + "loss": 2.0916, + "step": 54 + }, + { + "epoch": 0.03804253847484005, + "grad_norm": 0.36718690395355225, + "learning_rate": 7.397260273972603e-05, + "loss": 2.6542, + "step": 55 + }, + { + "epoch": 0.03873422099256441, + "grad_norm": 0.43732333183288574, + "learning_rate": 7.534246575342466e-05, + "loss": 2.3609, + "step": 56 + }, + { + "epoch": 0.039425903510288775, + "grad_norm": 0.5333709716796875, + "learning_rate": 7.671232876712329e-05, + "loss": 2.0736, + "step": 57 + }, + { + "epoch": 0.040117586028013144, + "grad_norm": 0.609667181968689, + "learning_rate": 7.808219178082192e-05, + "loss": 1.9863, + "step": 58 + }, + { + "epoch": 0.040809268545737507, + "grad_norm": 0.38461530208587646, + "learning_rate": 7.945205479452055e-05, + "loss": 2.0244, + "step": 59 + }, + { + "epoch": 0.04150095106346187, + "grad_norm": 0.37384161353111267, + "learning_rate": 8.082191780821919e-05, + "loss": 2.5084, + "step": 60 + }, + { + "epoch": 0.04219263358118624, + "grad_norm": 0.575130820274353, + "learning_rate": 8.219178082191781e-05, + "loss": 1.5943, + "step": 61 + }, + { + "epoch": 0.0428843160989106, + "grad_norm": 0.47650450468063354, + "learning_rate": 8.356164383561645e-05, + "loss": 2.4752, + "step": 62 + }, + { + "epoch": 0.04357599861663496, + "grad_norm": 0.36863836646080017, + "learning_rate": 8.493150684931507e-05, + "loss": 2.5023, + "step": 63 + }, + { + "epoch": 0.04426768113435933, + "grad_norm": 0.3590114116668701, + "learning_rate": 8.630136986301371e-05, + "loss": 2.1982, + "step": 64 + }, + { + "epoch": 0.044959363652083695, + "grad_norm": 0.4362975060939789, + "learning_rate": 8.767123287671233e-05, + "loss": 2.0463, + "step": 65 + }, + { + "epoch": 0.04565104616980806, + "grad_norm": 0.39568379521369934, + "learning_rate": 8.904109589041096e-05, + "loss": 2.5595, + "step": 66 + }, + { + "epoch": 0.04634272868753242, + "grad_norm": 0.41195258498191833, + "learning_rate": 9.041095890410958e-05, + "loss": 2.348, + "step": 67 + }, + { + "epoch": 0.04703441120525679, + "grad_norm": 0.5394885540008545, + "learning_rate": 9.178082191780822e-05, + "loss": 1.2385, + "step": 68 + }, + { + "epoch": 0.04772609372298115, + "grad_norm": 0.41482311487197876, + "learning_rate": 9.315068493150684e-05, + "loss": 2.7036, + "step": 69 + }, + { + "epoch": 0.048417776240705514, + "grad_norm": 0.4505786597728729, + "learning_rate": 9.452054794520548e-05, + "loss": 1.7636, + "step": 70 + }, + { + "epoch": 0.04910945875842988, + "grad_norm": 0.26118066906929016, + "learning_rate": 9.58904109589041e-05, + "loss": 1.7056, + "step": 71 + }, + { + "epoch": 0.049801141276154245, + "grad_norm": 0.7240022420883179, + "learning_rate": 9.726027397260274e-05, + "loss": 2.3171, + "step": 72 + }, + { + "epoch": 0.05049282379387861, + "grad_norm": 0.5030315518379211, + "learning_rate": 9.863013698630137e-05, + "loss": 2.1896, + "step": 73 + }, + { + "epoch": 0.05118450631160298, + "grad_norm": 0.7940630316734314, + "learning_rate": 0.0001, + "loss": 1.4996, + "step": 74 + }, + { + "epoch": 0.05187618882932734, + "grad_norm": 0.45342186093330383, + "learning_rate": 9.992716678805537e-05, + "loss": 2.1487, + "step": 75 + }, + { + "epoch": 0.0525678713470517, + "grad_norm": 0.4451451301574707, + "learning_rate": 9.985433357611071e-05, + "loss": 2.1561, + "step": 76 + }, + { + "epoch": 0.053259553864776064, + "grad_norm": 0.39434245228767395, + "learning_rate": 9.978150036416607e-05, + "loss": 1.8114, + "step": 77 + }, + { + "epoch": 0.053951236382500434, + "grad_norm": 0.5808930397033691, + "learning_rate": 9.970866715222141e-05, + "loss": 2.0528, + "step": 78 + }, + { + "epoch": 0.054642918900224796, + "grad_norm": 0.4741731882095337, + "learning_rate": 9.963583394027677e-05, + "loss": 1.988, + "step": 79 + }, + { + "epoch": 0.05533460141794916, + "grad_norm": 0.38925132155418396, + "learning_rate": 9.956300072833212e-05, + "loss": 1.1629, + "step": 80 + }, + { + "epoch": 0.05602628393567353, + "grad_norm": 0.5677796602249146, + "learning_rate": 9.949016751638748e-05, + "loss": 1.7325, + "step": 81 + }, + { + "epoch": 0.05671796645339789, + "grad_norm": 0.5454609990119934, + "learning_rate": 9.941733430444284e-05, + "loss": 2.0386, + "step": 82 + }, + { + "epoch": 0.05740964897112225, + "grad_norm": 0.42700880765914917, + "learning_rate": 9.934450109249819e-05, + "loss": 2.1646, + "step": 83 + }, + { + "epoch": 0.05810133148884662, + "grad_norm": 0.4880882203578949, + "learning_rate": 9.927166788055353e-05, + "loss": 1.8193, + "step": 84 + }, + { + "epoch": 0.058793014006570984, + "grad_norm": 0.4928286373615265, + "learning_rate": 9.919883466860888e-05, + "loss": 1.6243, + "step": 85 + }, + { + "epoch": 0.05948469652429535, + "grad_norm": 0.5179658532142639, + "learning_rate": 9.912600145666424e-05, + "loss": 1.6846, + "step": 86 + }, + { + "epoch": 0.060176379042019716, + "grad_norm": 0.6409149765968323, + "learning_rate": 9.905316824471959e-05, + "loss": 0.8419, + "step": 87 + }, + { + "epoch": 0.06086806155974408, + "grad_norm": 0.7547211050987244, + "learning_rate": 9.898033503277495e-05, + "loss": 2.399, + "step": 88 + }, + { + "epoch": 0.06155974407746844, + "grad_norm": 0.412672758102417, + "learning_rate": 9.890750182083031e-05, + "loss": 2.2212, + "step": 89 + }, + { + "epoch": 0.0622514265951928, + "grad_norm": 0.5434851050376892, + "learning_rate": 9.883466860888566e-05, + "loss": 2.3083, + "step": 90 + }, + { + "epoch": 0.06294310911291717, + "grad_norm": 0.4428962767124176, + "learning_rate": 9.876183539694101e-05, + "loss": 2.3586, + "step": 91 + }, + { + "epoch": 0.06363479163064154, + "grad_norm": 0.643883466720581, + "learning_rate": 9.868900218499635e-05, + "loss": 2.4555, + "step": 92 + }, + { + "epoch": 0.0643264741483659, + "grad_norm": 1.8360555171966553, + "learning_rate": 9.861616897305172e-05, + "loss": 2.3463, + "step": 93 + }, + { + "epoch": 0.06501815666609026, + "grad_norm": 0.5649489164352417, + "learning_rate": 9.854333576110706e-05, + "loss": 1.9054, + "step": 94 + }, + { + "epoch": 0.06570983918381462, + "grad_norm": 0.5043310523033142, + "learning_rate": 9.847050254916242e-05, + "loss": 2.0604, + "step": 95 + }, + { + "epoch": 0.066401521701539, + "grad_norm": 0.49205994606018066, + "learning_rate": 9.839766933721779e-05, + "loss": 2.5182, + "step": 96 + }, + { + "epoch": 0.06709320421926336, + "grad_norm": 0.9012492299079895, + "learning_rate": 9.832483612527313e-05, + "loss": 1.9895, + "step": 97 + }, + { + "epoch": 0.06778488673698772, + "grad_norm": 0.576653778553009, + "learning_rate": 9.825200291332848e-05, + "loss": 2.5201, + "step": 98 + }, + { + "epoch": 0.06847656925471209, + "grad_norm": 0.4745285212993622, + "learning_rate": 9.817916970138383e-05, + "loss": 1.9348, + "step": 99 + }, + { + "epoch": 0.06916825177243645, + "grad_norm": 0.5546420216560364, + "learning_rate": 9.810633648943919e-05, + "loss": 1.6732, + "step": 100 + }, + { + "epoch": 0.06985993429016081, + "grad_norm": 0.8806080222129822, + "learning_rate": 9.803350327749454e-05, + "loss": 1.713, + "step": 101 + }, + { + "epoch": 0.07055161680788519, + "grad_norm": 0.6029040217399597, + "learning_rate": 9.79606700655499e-05, + "loss": 1.5178, + "step": 102 + }, + { + "epoch": 0.07124329932560955, + "grad_norm": 0.56731778383255, + "learning_rate": 9.788783685360526e-05, + "loss": 1.6848, + "step": 103 + }, + { + "epoch": 0.07193498184333391, + "grad_norm": 0.5833808183670044, + "learning_rate": 9.781500364166059e-05, + "loss": 1.703, + "step": 104 + }, + { + "epoch": 0.07262666436105827, + "grad_norm": 0.724471390247345, + "learning_rate": 9.774217042971595e-05, + "loss": 2.227, + "step": 105 + }, + { + "epoch": 0.07331834687878264, + "grad_norm": 0.722097635269165, + "learning_rate": 9.76693372177713e-05, + "loss": 1.2173, + "step": 106 + }, + { + "epoch": 0.074010029396507, + "grad_norm": 0.7538579702377319, + "learning_rate": 9.759650400582666e-05, + "loss": 1.8546, + "step": 107 + }, + { + "epoch": 0.07470171191423136, + "grad_norm": 0.7746326327323914, + "learning_rate": 9.752367079388201e-05, + "loss": 2.4067, + "step": 108 + }, + { + "epoch": 0.07539339443195574, + "grad_norm": 0.5197188258171082, + "learning_rate": 9.745083758193737e-05, + "loss": 1.9193, + "step": 109 + }, + { + "epoch": 0.0760850769496801, + "grad_norm": 0.4536001682281494, + "learning_rate": 9.737800436999273e-05, + "loss": 2.2905, + "step": 110 + }, + { + "epoch": 0.07677675946740446, + "grad_norm": 0.6520904302597046, + "learning_rate": 9.730517115804807e-05, + "loss": 2.012, + "step": 111 + }, + { + "epoch": 0.07746844198512882, + "grad_norm": 0.5670090913772583, + "learning_rate": 9.723233794610343e-05, + "loss": 2.0289, + "step": 112 + }, + { + "epoch": 0.07816012450285319, + "grad_norm": 0.8609398603439331, + "learning_rate": 9.715950473415877e-05, + "loss": 1.332, + "step": 113 + }, + { + "epoch": 0.07885180702057755, + "grad_norm": 0.5880535244941711, + "learning_rate": 9.708667152221414e-05, + "loss": 1.7267, + "step": 114 + }, + { + "epoch": 0.07954348953830193, + "grad_norm": 0.6090431213378906, + "learning_rate": 9.701383831026948e-05, + "loss": 1.3968, + "step": 115 + }, + { + "epoch": 0.08023517205602629, + "grad_norm": 0.5701532363891602, + "learning_rate": 9.694100509832484e-05, + "loss": 2.1829, + "step": 116 + }, + { + "epoch": 0.08092685457375065, + "grad_norm": 0.6294519305229187, + "learning_rate": 9.68681718863802e-05, + "loss": 1.7385, + "step": 117 + }, + { + "epoch": 0.08161853709147501, + "grad_norm": 0.5141008496284485, + "learning_rate": 9.679533867443554e-05, + "loss": 1.8417, + "step": 118 + }, + { + "epoch": 0.08231021960919938, + "grad_norm": 1.0640853643417358, + "learning_rate": 9.67225054624909e-05, + "loss": 2.1031, + "step": 119 + }, + { + "epoch": 0.08300190212692374, + "grad_norm": 0.6235936284065247, + "learning_rate": 9.664967225054625e-05, + "loss": 1.9911, + "step": 120 + }, + { + "epoch": 0.0836935846446481, + "grad_norm": 0.8041340112686157, + "learning_rate": 9.657683903860161e-05, + "loss": 1.9172, + "step": 121 + }, + { + "epoch": 0.08438526716237248, + "grad_norm": 0.6447221040725708, + "learning_rate": 9.650400582665696e-05, + "loss": 2.0157, + "step": 122 + }, + { + "epoch": 0.08507694968009684, + "grad_norm": 0.7299740314483643, + "learning_rate": 9.643117261471232e-05, + "loss": 1.9157, + "step": 123 + }, + { + "epoch": 0.0857686321978212, + "grad_norm": 0.8950564861297607, + "learning_rate": 9.635833940276767e-05, + "loss": 1.4313, + "step": 124 + }, + { + "epoch": 0.08646031471554556, + "grad_norm": 0.5062771439552307, + "learning_rate": 9.628550619082301e-05, + "loss": 1.7742, + "step": 125 + }, + { + "epoch": 0.08715199723326993, + "grad_norm": 0.7482561469078064, + "learning_rate": 9.621267297887837e-05, + "loss": 1.6122, + "step": 126 + }, + { + "epoch": 0.08784367975099429, + "grad_norm": 0.4244556725025177, + "learning_rate": 9.613983976693372e-05, + "loss": 1.3825, + "step": 127 + }, + { + "epoch": 0.08853536226871866, + "grad_norm": 0.5349308252334595, + "learning_rate": 9.606700655498908e-05, + "loss": 1.9771, + "step": 128 + }, + { + "epoch": 0.08922704478644303, + "grad_norm": 0.5429610013961792, + "learning_rate": 9.599417334304443e-05, + "loss": 2.5704, + "step": 129 + }, + { + "epoch": 0.08991872730416739, + "grad_norm": 0.4875735938549042, + "learning_rate": 9.592134013109979e-05, + "loss": 2.0568, + "step": 130 + }, + { + "epoch": 0.09061040982189175, + "grad_norm": 0.5956497192382812, + "learning_rate": 9.584850691915514e-05, + "loss": 2.3604, + "step": 131 + }, + { + "epoch": 0.09130209233961611, + "grad_norm": 1.0286632776260376, + "learning_rate": 9.577567370721049e-05, + "loss": 1.278, + "step": 132 + }, + { + "epoch": 0.09199377485734048, + "grad_norm": 0.5342454314231873, + "learning_rate": 9.570284049526585e-05, + "loss": 1.7806, + "step": 133 + }, + { + "epoch": 0.09268545737506484, + "grad_norm": 0.5517177581787109, + "learning_rate": 9.56300072833212e-05, + "loss": 2.0241, + "step": 134 + }, + { + "epoch": 0.09337713989278922, + "grad_norm": 0.6514694094657898, + "learning_rate": 9.555717407137656e-05, + "loss": 2.002, + "step": 135 + }, + { + "epoch": 0.09406882241051358, + "grad_norm": 0.6356580257415771, + "learning_rate": 9.54843408594319e-05, + "loss": 2.2756, + "step": 136 + }, + { + "epoch": 0.09476050492823794, + "grad_norm": 0.6672503352165222, + "learning_rate": 9.541150764748726e-05, + "loss": 1.4608, + "step": 137 + }, + { + "epoch": 0.0954521874459623, + "grad_norm": 0.8465030193328857, + "learning_rate": 9.533867443554261e-05, + "loss": 2.0107, + "step": 138 + }, + { + "epoch": 0.09614386996368667, + "grad_norm": 0.6064625978469849, + "learning_rate": 9.526584122359796e-05, + "loss": 2.1404, + "step": 139 + }, + { + "epoch": 0.09683555248141103, + "grad_norm": 0.6728728413581848, + "learning_rate": 9.519300801165332e-05, + "loss": 1.7548, + "step": 140 + }, + { + "epoch": 0.0975272349991354, + "grad_norm": 0.7992367744445801, + "learning_rate": 9.512017479970867e-05, + "loss": 1.5071, + "step": 141 + }, + { + "epoch": 0.09821891751685977, + "grad_norm": 0.466144859790802, + "learning_rate": 9.504734158776403e-05, + "loss": 1.5908, + "step": 142 + }, + { + "epoch": 0.09891060003458413, + "grad_norm": 0.694807767868042, + "learning_rate": 9.497450837581938e-05, + "loss": 1.6261, + "step": 143 + }, + { + "epoch": 0.09960228255230849, + "grad_norm": 0.5561632513999939, + "learning_rate": 9.490167516387472e-05, + "loss": 1.2933, + "step": 144 + }, + { + "epoch": 0.10029396507003285, + "grad_norm": 0.6609554886817932, + "learning_rate": 9.482884195193008e-05, + "loss": 1.7889, + "step": 145 + }, + { + "epoch": 0.10098564758775722, + "grad_norm": 0.9272559285163879, + "learning_rate": 9.475600873998543e-05, + "loss": 1.2951, + "step": 146 + }, + { + "epoch": 0.10167733010548158, + "grad_norm": 1.5346007347106934, + "learning_rate": 9.46831755280408e-05, + "loss": 1.5395, + "step": 147 + }, + { + "epoch": 0.10236901262320595, + "grad_norm": 0.8031420111656189, + "learning_rate": 9.461034231609614e-05, + "loss": 2.2035, + "step": 148 + }, + { + "epoch": 0.10306069514093032, + "grad_norm": 0.5305848121643066, + "learning_rate": 9.45375091041515e-05, + "loss": 1.784, + "step": 149 + }, + { + "epoch": 0.10375237765865468, + "grad_norm": 0.6742457151412964, + "learning_rate": 9.446467589220685e-05, + "loss": 2.0042, + "step": 150 + }, + { + "epoch": 0.10444406017637904, + "grad_norm": 0.7282371520996094, + "learning_rate": 9.43918426802622e-05, + "loss": 2.2441, + "step": 151 + }, + { + "epoch": 0.1051357426941034, + "grad_norm": 0.5512478351593018, + "learning_rate": 9.431900946831756e-05, + "loss": 2.0307, + "step": 152 + }, + { + "epoch": 0.10582742521182777, + "grad_norm": 0.7224892377853394, + "learning_rate": 9.42461762563729e-05, + "loss": 2.4692, + "step": 153 + }, + { + "epoch": 0.10651910772955213, + "grad_norm": 0.7150198817253113, + "learning_rate": 9.417334304442827e-05, + "loss": 2.0194, + "step": 154 + }, + { + "epoch": 0.1072107902472765, + "grad_norm": 0.5024279356002808, + "learning_rate": 9.410050983248361e-05, + "loss": 2.0461, + "step": 155 + }, + { + "epoch": 0.10790247276500087, + "grad_norm": 1.0661766529083252, + "learning_rate": 9.402767662053898e-05, + "loss": 1.3458, + "step": 156 + }, + { + "epoch": 0.10859415528272523, + "grad_norm": 0.577153205871582, + "learning_rate": 9.395484340859432e-05, + "loss": 1.517, + "step": 157 + }, + { + "epoch": 0.10928583780044959, + "grad_norm": 0.6876221895217896, + "learning_rate": 9.388201019664967e-05, + "loss": 1.9622, + "step": 158 + }, + { + "epoch": 0.10997752031817395, + "grad_norm": 0.5576820373535156, + "learning_rate": 9.380917698470503e-05, + "loss": 1.6063, + "step": 159 + }, + { + "epoch": 0.11066920283589832, + "grad_norm": 0.8604760766029358, + "learning_rate": 9.373634377276038e-05, + "loss": 1.8693, + "step": 160 + }, + { + "epoch": 0.1113608853536227, + "grad_norm": 0.6998944282531738, + "learning_rate": 9.366351056081574e-05, + "loss": 1.8942, + "step": 161 + }, + { + "epoch": 0.11205256787134706, + "grad_norm": 0.7106878757476807, + "learning_rate": 9.359067734887109e-05, + "loss": 1.9999, + "step": 162 + }, + { + "epoch": 0.11274425038907142, + "grad_norm": 1.167450189590454, + "learning_rate": 9.351784413692645e-05, + "loss": 1.26, + "step": 163 + }, + { + "epoch": 0.11343593290679578, + "grad_norm": 0.6208682060241699, + "learning_rate": 9.34450109249818e-05, + "loss": 1.7881, + "step": 164 + }, + { + "epoch": 0.11412761542452014, + "grad_norm": 0.7046381235122681, + "learning_rate": 9.337217771303714e-05, + "loss": 1.7403, + "step": 165 + }, + { + "epoch": 0.1148192979422445, + "grad_norm": 0.6957246661186218, + "learning_rate": 9.32993445010925e-05, + "loss": 2.1148, + "step": 166 + }, + { + "epoch": 0.11551098045996887, + "grad_norm": 0.6169411540031433, + "learning_rate": 9.322651128914785e-05, + "loss": 1.9182, + "step": 167 + }, + { + "epoch": 0.11620266297769324, + "grad_norm": 0.8539403080940247, + "learning_rate": 9.315367807720321e-05, + "loss": 1.7015, + "step": 168 + }, + { + "epoch": 0.1168943454954176, + "grad_norm": 1.037312626838684, + "learning_rate": 9.308084486525856e-05, + "loss": 1.3205, + "step": 169 + }, + { + "epoch": 0.11758602801314197, + "grad_norm": 0.5971755981445312, + "learning_rate": 9.300801165331392e-05, + "loss": 1.1845, + "step": 170 + }, + { + "epoch": 0.11827771053086633, + "grad_norm": 0.5809158086776733, + "learning_rate": 9.293517844136927e-05, + "loss": 1.5852, + "step": 171 + }, + { + "epoch": 0.1189693930485907, + "grad_norm": 0.6503626704216003, + "learning_rate": 9.286234522942462e-05, + "loss": 1.9045, + "step": 172 + }, + { + "epoch": 0.11966107556631506, + "grad_norm": 0.5208730101585388, + "learning_rate": 9.278951201747998e-05, + "loss": 1.7599, + "step": 173 + }, + { + "epoch": 0.12035275808403943, + "grad_norm": 0.76451176404953, + "learning_rate": 9.271667880553533e-05, + "loss": 2.2624, + "step": 174 + }, + { + "epoch": 0.1210444406017638, + "grad_norm": 0.5927959084510803, + "learning_rate": 9.264384559359069e-05, + "loss": 1.5911, + "step": 175 + }, + { + "epoch": 0.12173612311948816, + "grad_norm": 0.6512097716331482, + "learning_rate": 9.257101238164603e-05, + "loss": 1.3409, + "step": 176 + }, + { + "epoch": 0.12242780563721252, + "grad_norm": 0.4837232530117035, + "learning_rate": 9.24981791697014e-05, + "loss": 1.2417, + "step": 177 + }, + { + "epoch": 0.12311948815493688, + "grad_norm": 0.6251150369644165, + "learning_rate": 9.242534595775674e-05, + "loss": 2.1377, + "step": 178 + }, + { + "epoch": 0.12381117067266124, + "grad_norm": 0.5964468717575073, + "learning_rate": 9.235251274581209e-05, + "loss": 1.616, + "step": 179 + }, + { + "epoch": 0.1245028531903856, + "grad_norm": 0.8185293674468994, + "learning_rate": 9.227967953386745e-05, + "loss": 2.372, + "step": 180 + }, + { + "epoch": 0.12519453570810998, + "grad_norm": 0.7391765117645264, + "learning_rate": 9.22068463219228e-05, + "loss": 2.0473, + "step": 181 + }, + { + "epoch": 0.12588621822583435, + "grad_norm": 0.6867517828941345, + "learning_rate": 9.213401310997816e-05, + "loss": 2.2592, + "step": 182 + }, + { + "epoch": 0.1265779007435587, + "grad_norm": 0.6889393329620361, + "learning_rate": 9.206117989803351e-05, + "loss": 1.8863, + "step": 183 + }, + { + "epoch": 0.12726958326128307, + "grad_norm": 0.6404229402542114, + "learning_rate": 9.198834668608885e-05, + "loss": 1.5859, + "step": 184 + }, + { + "epoch": 0.12796126577900743, + "grad_norm": 2.46905255317688, + "learning_rate": 9.191551347414422e-05, + "loss": 2.4798, + "step": 185 + }, + { + "epoch": 0.1286529482967318, + "grad_norm": 0.7490037083625793, + "learning_rate": 9.184268026219956e-05, + "loss": 2.1623, + "step": 186 + }, + { + "epoch": 0.12934463081445616, + "grad_norm": 0.5947558283805847, + "learning_rate": 9.176984705025492e-05, + "loss": 2.2375, + "step": 187 + }, + { + "epoch": 0.13003631333218052, + "grad_norm": 0.4829116463661194, + "learning_rate": 9.169701383831027e-05, + "loss": 1.66, + "step": 188 + }, + { + "epoch": 0.13072799584990488, + "grad_norm": 0.5807337164878845, + "learning_rate": 9.162418062636563e-05, + "loss": 1.3098, + "step": 189 + }, + { + "epoch": 0.13141967836762924, + "grad_norm": 0.8253926038742065, + "learning_rate": 9.155134741442098e-05, + "loss": 1.135, + "step": 190 + }, + { + "epoch": 0.13211136088535363, + "grad_norm": 0.7517785429954529, + "learning_rate": 9.147851420247633e-05, + "loss": 1.6342, + "step": 191 + }, + { + "epoch": 0.132803043403078, + "grad_norm": 0.791852593421936, + "learning_rate": 9.140568099053169e-05, + "loss": 1.3013, + "step": 192 + }, + { + "epoch": 0.13349472592080236, + "grad_norm": 0.6530910730361938, + "learning_rate": 9.133284777858704e-05, + "loss": 2.4537, + "step": 193 + }, + { + "epoch": 0.13418640843852672, + "grad_norm": 0.8071674704551697, + "learning_rate": 9.12600145666424e-05, + "loss": 1.1286, + "step": 194 + }, + { + "epoch": 0.13487809095625108, + "grad_norm": 0.5800924301147461, + "learning_rate": 9.118718135469774e-05, + "loss": 1.913, + "step": 195 + }, + { + "epoch": 0.13556977347397545, + "grad_norm": 0.6227284073829651, + "learning_rate": 9.11143481427531e-05, + "loss": 2.1569, + "step": 196 + }, + { + "epoch": 0.1362614559916998, + "grad_norm": 0.7267847657203674, + "learning_rate": 9.104151493080845e-05, + "loss": 2.4129, + "step": 197 + }, + { + "epoch": 0.13695313850942417, + "grad_norm": 0.6935005784034729, + "learning_rate": 9.09686817188638e-05, + "loss": 1.2993, + "step": 198 + }, + { + "epoch": 0.13764482102714853, + "grad_norm": 0.5625126957893372, + "learning_rate": 9.089584850691916e-05, + "loss": 1.6137, + "step": 199 + }, + { + "epoch": 0.1383365035448729, + "grad_norm": 0.6450534462928772, + "learning_rate": 9.082301529497451e-05, + "loss": 2.0705, + "step": 200 + }, + { + "epoch": 0.13902818606259726, + "grad_norm": 0.7700170874595642, + "learning_rate": 9.075018208302987e-05, + "loss": 1.2858, + "step": 201 + }, + { + "epoch": 0.13971986858032162, + "grad_norm": 0.4889370799064636, + "learning_rate": 9.067734887108522e-05, + "loss": 1.1423, + "step": 202 + }, + { + "epoch": 0.14041155109804598, + "grad_norm": 0.5464507341384888, + "learning_rate": 9.060451565914058e-05, + "loss": 1.0928, + "step": 203 + }, + { + "epoch": 0.14110323361577037, + "grad_norm": 0.7002052664756775, + "learning_rate": 9.053168244719593e-05, + "loss": 2.154, + "step": 204 + }, + { + "epoch": 0.14179491613349474, + "grad_norm": 0.73245769739151, + "learning_rate": 9.045884923525127e-05, + "loss": 2.3437, + "step": 205 + }, + { + "epoch": 0.1424865986512191, + "grad_norm": 0.6388504505157471, + "learning_rate": 9.038601602330664e-05, + "loss": 1.6209, + "step": 206 + }, + { + "epoch": 0.14317828116894346, + "grad_norm": 0.5278307199478149, + "learning_rate": 9.031318281136198e-05, + "loss": 1.1904, + "step": 207 + }, + { + "epoch": 0.14386996368666782, + "grad_norm": 0.6711840033531189, + "learning_rate": 9.024034959941734e-05, + "loss": 1.7129, + "step": 208 + }, + { + "epoch": 0.14456164620439219, + "grad_norm": 0.6358078122138977, + "learning_rate": 9.016751638747269e-05, + "loss": 1.6655, + "step": 209 + }, + { + "epoch": 0.14525332872211655, + "grad_norm": 0.6745946407318115, + "learning_rate": 9.009468317552805e-05, + "loss": 1.8217, + "step": 210 + }, + { + "epoch": 0.1459450112398409, + "grad_norm": 2.227377414703369, + "learning_rate": 9.00218499635834e-05, + "loss": 3.1095, + "step": 211 + }, + { + "epoch": 0.14663669375756527, + "grad_norm": 0.5836493968963623, + "learning_rate": 8.994901675163875e-05, + "loss": 1.9259, + "step": 212 + }, + { + "epoch": 0.14732837627528964, + "grad_norm": 0.6573939323425293, + "learning_rate": 8.987618353969411e-05, + "loss": 1.9532, + "step": 213 + }, + { + "epoch": 0.148020058793014, + "grad_norm": 0.8865386843681335, + "learning_rate": 8.980335032774946e-05, + "loss": 1.7339, + "step": 214 + }, + { + "epoch": 0.14871174131073836, + "grad_norm": 0.7364823818206787, + "learning_rate": 8.973051711580482e-05, + "loss": 1.2418, + "step": 215 + }, + { + "epoch": 0.14940342382846272, + "grad_norm": 0.7573681473731995, + "learning_rate": 8.965768390386016e-05, + "loss": 1.5159, + "step": 216 + }, + { + "epoch": 0.1500951063461871, + "grad_norm": 0.6876881718635559, + "learning_rate": 8.958485069191553e-05, + "loss": 2.2091, + "step": 217 + }, + { + "epoch": 0.15078678886391148, + "grad_norm": 0.6091864705085754, + "learning_rate": 8.951201747997087e-05, + "loss": 1.2208, + "step": 218 + }, + { + "epoch": 0.15147847138163584, + "grad_norm": 0.7913519740104675, + "learning_rate": 8.943918426802622e-05, + "loss": 2.3047, + "step": 219 + }, + { + "epoch": 0.1521701538993602, + "grad_norm": 1.3490211963653564, + "learning_rate": 8.936635105608158e-05, + "loss": 1.7839, + "step": 220 + }, + { + "epoch": 0.15286183641708456, + "grad_norm": 0.8339107036590576, + "learning_rate": 8.929351784413693e-05, + "loss": 2.1117, + "step": 221 + }, + { + "epoch": 0.15355351893480892, + "grad_norm": 1.5490026473999023, + "learning_rate": 8.922068463219229e-05, + "loss": 1.8909, + "step": 222 + }, + { + "epoch": 0.1542452014525333, + "grad_norm": 0.7403117418289185, + "learning_rate": 8.914785142024764e-05, + "loss": 1.838, + "step": 223 + }, + { + "epoch": 0.15493688397025765, + "grad_norm": 0.9296958446502686, + "learning_rate": 8.907501820830299e-05, + "loss": 1.8148, + "step": 224 + }, + { + "epoch": 0.155628566487982, + "grad_norm": 0.6639947891235352, + "learning_rate": 8.900218499635835e-05, + "loss": 1.8576, + "step": 225 + }, + { + "epoch": 0.15632024900570637, + "grad_norm": 0.7341317534446716, + "learning_rate": 8.89293517844137e-05, + "loss": 1.9323, + "step": 226 + }, + { + "epoch": 0.15701193152343074, + "grad_norm": 0.5162379145622253, + "learning_rate": 8.885651857246906e-05, + "loss": 2.1562, + "step": 227 + }, + { + "epoch": 0.1577036140411551, + "grad_norm": 0.6806768774986267, + "learning_rate": 8.87836853605244e-05, + "loss": 1.8393, + "step": 228 + }, + { + "epoch": 0.15839529655887946, + "grad_norm": 0.5701708197593689, + "learning_rate": 8.871085214857976e-05, + "loss": 1.4365, + "step": 229 + }, + { + "epoch": 0.15908697907660385, + "grad_norm": 0.7863550186157227, + "learning_rate": 8.863801893663511e-05, + "loss": 2.071, + "step": 230 + }, + { + "epoch": 0.15977866159432821, + "grad_norm": 0.6164423823356628, + "learning_rate": 8.856518572469046e-05, + "loss": 2.0573, + "step": 231 + }, + { + "epoch": 0.16047034411205258, + "grad_norm": 0.8759453892707825, + "learning_rate": 8.849235251274582e-05, + "loss": 1.8708, + "step": 232 + }, + { + "epoch": 0.16116202662977694, + "grad_norm": 1.3226007223129272, + "learning_rate": 8.841951930080117e-05, + "loss": 1.4042, + "step": 233 + }, + { + "epoch": 0.1618537091475013, + "grad_norm": 0.5054848790168762, + "learning_rate": 8.834668608885653e-05, + "loss": 1.2018, + "step": 234 + }, + { + "epoch": 0.16254539166522566, + "grad_norm": 0.6680052876472473, + "learning_rate": 8.827385287691188e-05, + "loss": 2.1112, + "step": 235 + }, + { + "epoch": 0.16323707418295003, + "grad_norm": 0.838471531867981, + "learning_rate": 8.820101966496724e-05, + "loss": 2.3499, + "step": 236 + }, + { + "epoch": 0.1639287567006744, + "grad_norm": 0.6450280547142029, + "learning_rate": 8.812818645302258e-05, + "loss": 1.5009, + "step": 237 + }, + { + "epoch": 0.16462043921839875, + "grad_norm": 0.5540733933448792, + "learning_rate": 8.805535324107793e-05, + "loss": 1.7475, + "step": 238 + }, + { + "epoch": 0.1653121217361231, + "grad_norm": 0.6765146255493164, + "learning_rate": 8.798252002913329e-05, + "loss": 2.0798, + "step": 239 + }, + { + "epoch": 0.16600380425384748, + "grad_norm": 0.6925728917121887, + "learning_rate": 8.790968681718864e-05, + "loss": 1.9599, + "step": 240 + }, + { + "epoch": 0.16669548677157184, + "grad_norm": 0.787634015083313, + "learning_rate": 8.7836853605244e-05, + "loss": 1.5637, + "step": 241 + }, + { + "epoch": 0.1673871692892962, + "grad_norm": 0.7191415429115295, + "learning_rate": 8.776402039329935e-05, + "loss": 2.1707, + "step": 242 + }, + { + "epoch": 0.1680788518070206, + "grad_norm": 0.7958317995071411, + "learning_rate": 8.769118718135471e-05, + "loss": 2.147, + "step": 243 + }, + { + "epoch": 0.16877053432474495, + "grad_norm": 0.6804454922676086, + "learning_rate": 8.761835396941004e-05, + "loss": 1.9832, + "step": 244 + }, + { + "epoch": 0.16946221684246932, + "grad_norm": 0.8284922242164612, + "learning_rate": 8.75455207574654e-05, + "loss": 1.2555, + "step": 245 + }, + { + "epoch": 0.17015389936019368, + "grad_norm": 0.7180774211883545, + "learning_rate": 8.747268754552075e-05, + "loss": 1.8228, + "step": 246 + }, + { + "epoch": 0.17084558187791804, + "grad_norm": 0.8632107377052307, + "learning_rate": 8.739985433357611e-05, + "loss": 1.1625, + "step": 247 + }, + { + "epoch": 0.1715372643956424, + "grad_norm": 0.5004351139068604, + "learning_rate": 8.732702112163147e-05, + "loss": 1.6948, + "step": 248 + }, + { + "epoch": 0.17222894691336676, + "grad_norm": 0.6219859719276428, + "learning_rate": 8.725418790968682e-05, + "loss": 1.6748, + "step": 249 + }, + { + "epoch": 0.17292062943109113, + "grad_norm": 0.8690136075019836, + "learning_rate": 8.718135469774218e-05, + "loss": 1.7746, + "step": 250 + }, + { + "epoch": 0.1736123119488155, + "grad_norm": 0.6689561605453491, + "learning_rate": 8.710852148579752e-05, + "loss": 1.1883, + "step": 251 + }, + { + "epoch": 0.17430399446653985, + "grad_norm": 1.0592402219772339, + "learning_rate": 8.703568827385288e-05, + "loss": 1.6322, + "step": 252 + }, + { + "epoch": 0.17499567698426421, + "grad_norm": 0.721960723400116, + "learning_rate": 8.696285506190823e-05, + "loss": 1.8509, + "step": 253 + }, + { + "epoch": 0.17568735950198858, + "grad_norm": 0.7862136960029602, + "learning_rate": 8.689002184996359e-05, + "loss": 1.5465, + "step": 254 + }, + { + "epoch": 0.17637904201971294, + "grad_norm": 0.62255859375, + "learning_rate": 8.681718863801895e-05, + "loss": 1.7961, + "step": 255 + }, + { + "epoch": 0.17707072453743733, + "grad_norm": 0.8009974956512451, + "learning_rate": 8.67443554260743e-05, + "loss": 1.7475, + "step": 256 + }, + { + "epoch": 0.1777624070551617, + "grad_norm": 0.6300957202911377, + "learning_rate": 8.667152221412966e-05, + "loss": 1.8628, + "step": 257 + }, + { + "epoch": 0.17845408957288605, + "grad_norm": 0.5678866505622864, + "learning_rate": 8.659868900218499e-05, + "loss": 1.2443, + "step": 258 + }, + { + "epoch": 0.17914577209061042, + "grad_norm": 0.6319582462310791, + "learning_rate": 8.652585579024035e-05, + "loss": 1.3356, + "step": 259 + }, + { + "epoch": 0.17983745460833478, + "grad_norm": 0.5843179821968079, + "learning_rate": 8.64530225782957e-05, + "loss": 1.9961, + "step": 260 + }, + { + "epoch": 0.18052913712605914, + "grad_norm": 0.6211555600166321, + "learning_rate": 8.638018936635106e-05, + "loss": 1.1243, + "step": 261 + }, + { + "epoch": 0.1812208196437835, + "grad_norm": 0.7266420722007751, + "learning_rate": 8.630735615440642e-05, + "loss": 2.2338, + "step": 262 + }, + { + "epoch": 0.18191250216150787, + "grad_norm": 0.8187587857246399, + "learning_rate": 8.623452294246177e-05, + "loss": 2.0237, + "step": 263 + }, + { + "epoch": 0.18260418467923223, + "grad_norm": 0.9444418549537659, + "learning_rate": 8.616168973051712e-05, + "loss": 1.6278, + "step": 264 + }, + { + "epoch": 0.1832958671969566, + "grad_norm": 0.5889193415641785, + "learning_rate": 8.608885651857246e-05, + "loss": 2.1453, + "step": 265 + }, + { + "epoch": 0.18398754971468095, + "grad_norm": 0.6819151639938354, + "learning_rate": 8.601602330662782e-05, + "loss": 1.2219, + "step": 266 + }, + { + "epoch": 0.18467923223240532, + "grad_norm": 0.8448613286018372, + "learning_rate": 8.594319009468317e-05, + "loss": 2.1193, + "step": 267 + }, + { + "epoch": 0.18537091475012968, + "grad_norm": 0.7247136235237122, + "learning_rate": 8.587035688273853e-05, + "loss": 1.6155, + "step": 268 + }, + { + "epoch": 0.18606259726785407, + "grad_norm": 0.7371733784675598, + "learning_rate": 8.57975236707939e-05, + "loss": 1.5466, + "step": 269 + }, + { + "epoch": 0.18675427978557843, + "grad_norm": 0.6676841974258423, + "learning_rate": 8.572469045884924e-05, + "loss": 2.1909, + "step": 270 + }, + { + "epoch": 0.1874459623033028, + "grad_norm": 0.7091754078865051, + "learning_rate": 8.565185724690459e-05, + "loss": 2.3795, + "step": 271 + }, + { + "epoch": 0.18813764482102716, + "grad_norm": 0.7999275922775269, + "learning_rate": 8.557902403495994e-05, + "loss": 2.0659, + "step": 272 + }, + { + "epoch": 0.18882932733875152, + "grad_norm": 0.5176097750663757, + "learning_rate": 8.55061908230153e-05, + "loss": 1.4036, + "step": 273 + }, + { + "epoch": 0.18952100985647588, + "grad_norm": 0.6579016447067261, + "learning_rate": 8.543335761107065e-05, + "loss": 1.9072, + "step": 274 + }, + { + "epoch": 0.19021269237420024, + "grad_norm": 1.0604159832000732, + "learning_rate": 8.536052439912601e-05, + "loss": 2.1682, + "step": 275 + }, + { + "epoch": 0.1909043748919246, + "grad_norm": 0.7157823443412781, + "learning_rate": 8.528769118718137e-05, + "loss": 1.9188, + "step": 276 + }, + { + "epoch": 0.19159605740964897, + "grad_norm": 0.691941499710083, + "learning_rate": 8.521485797523672e-05, + "loss": 1.8626, + "step": 277 + }, + { + "epoch": 0.19228773992737333, + "grad_norm": 0.5831330418586731, + "learning_rate": 8.514202476329206e-05, + "loss": 1.9975, + "step": 278 + }, + { + "epoch": 0.1929794224450977, + "grad_norm": 0.6630810499191284, + "learning_rate": 8.506919155134741e-05, + "loss": 2.0809, + "step": 279 + }, + { + "epoch": 0.19367110496282205, + "grad_norm": 0.601836621761322, + "learning_rate": 8.499635833940277e-05, + "loss": 1.2491, + "step": 280 + }, + { + "epoch": 0.19436278748054642, + "grad_norm": 0.7532071471214294, + "learning_rate": 8.492352512745812e-05, + "loss": 2.0672, + "step": 281 + }, + { + "epoch": 0.1950544699982708, + "grad_norm": 0.70450758934021, + "learning_rate": 8.485069191551348e-05, + "loss": 2.1466, + "step": 282 + }, + { + "epoch": 0.19574615251599517, + "grad_norm": 0.6915740966796875, + "learning_rate": 8.477785870356884e-05, + "loss": 1.4086, + "step": 283 + }, + { + "epoch": 0.19643783503371953, + "grad_norm": 0.949400007724762, + "learning_rate": 8.470502549162418e-05, + "loss": 1.6344, + "step": 284 + }, + { + "epoch": 0.1971295175514439, + "grad_norm": 0.7051184177398682, + "learning_rate": 8.463219227967954e-05, + "loss": 1.4375, + "step": 285 + }, + { + "epoch": 0.19782120006916826, + "grad_norm": 0.5827087163925171, + "learning_rate": 8.455935906773488e-05, + "loss": 1.4801, + "step": 286 + }, + { + "epoch": 0.19851288258689262, + "grad_norm": 0.6058622002601624, + "learning_rate": 8.448652585579024e-05, + "loss": 1.4378, + "step": 287 + }, + { + "epoch": 0.19920456510461698, + "grad_norm": 0.7779908180236816, + "learning_rate": 8.441369264384559e-05, + "loss": 1.6377, + "step": 288 + }, + { + "epoch": 0.19989624762234134, + "grad_norm": 0.6972940564155579, + "learning_rate": 8.434085943190095e-05, + "loss": 1.9437, + "step": 289 + }, + { + "epoch": 0.2005879301400657, + "grad_norm": 0.7289730906486511, + "learning_rate": 8.426802621995631e-05, + "loss": 1.5402, + "step": 290 + }, + { + "epoch": 0.20127961265779007, + "grad_norm": 0.6566206216812134, + "learning_rate": 8.419519300801165e-05, + "loss": 1.4304, + "step": 291 + }, + { + "epoch": 0.20197129517551443, + "grad_norm": 0.97121262550354, + "learning_rate": 8.412235979606701e-05, + "loss": 1.7801, + "step": 292 + }, + { + "epoch": 0.2026629776932388, + "grad_norm": 0.7539506554603577, + "learning_rate": 8.404952658412236e-05, + "loss": 1.5326, + "step": 293 + }, + { + "epoch": 0.20335466021096316, + "grad_norm": 0.8037034869194031, + "learning_rate": 8.397669337217772e-05, + "loss": 1.5559, + "step": 294 + }, + { + "epoch": 0.20404634272868752, + "grad_norm": 1.2110254764556885, + "learning_rate": 8.390386016023307e-05, + "loss": 1.7529, + "step": 295 + }, + { + "epoch": 0.2047380252464119, + "grad_norm": 0.9396728873252869, + "learning_rate": 8.383102694828843e-05, + "loss": 2.0101, + "step": 296 + }, + { + "epoch": 0.20542970776413627, + "grad_norm": 0.7678546905517578, + "learning_rate": 8.375819373634379e-05, + "loss": 2.0231, + "step": 297 + }, + { + "epoch": 0.20612139028186063, + "grad_norm": 1.1070085763931274, + "learning_rate": 8.368536052439912e-05, + "loss": 1.7999, + "step": 298 + }, + { + "epoch": 0.206813072799585, + "grad_norm": 0.7589883804321289, + "learning_rate": 8.361252731245448e-05, + "loss": 1.6723, + "step": 299 + }, + { + "epoch": 0.20750475531730936, + "grad_norm": 0.6468162536621094, + "learning_rate": 8.353969410050983e-05, + "loss": 1.3266, + "step": 300 + }, + { + "epoch": 0.20819643783503372, + "grad_norm": 0.7720943689346313, + "learning_rate": 8.346686088856519e-05, + "loss": 1.5897, + "step": 301 + }, + { + "epoch": 0.20888812035275808, + "grad_norm": 0.7349941730499268, + "learning_rate": 8.339402767662054e-05, + "loss": 1.9242, + "step": 302 + }, + { + "epoch": 0.20957980287048245, + "grad_norm": 0.7902266979217529, + "learning_rate": 8.33211944646759e-05, + "loss": 2.2367, + "step": 303 + }, + { + "epoch": 0.2102714853882068, + "grad_norm": 0.822661280632019, + "learning_rate": 8.324836125273125e-05, + "loss": 2.02, + "step": 304 + }, + { + "epoch": 0.21096316790593117, + "grad_norm": 0.6001129746437073, + "learning_rate": 8.31755280407866e-05, + "loss": 1.4506, + "step": 305 + }, + { + "epoch": 0.21165485042365553, + "grad_norm": 0.7674906253814697, + "learning_rate": 8.310269482884196e-05, + "loss": 1.9657, + "step": 306 + }, + { + "epoch": 0.2123465329413799, + "grad_norm": 0.5808192491531372, + "learning_rate": 8.30298616168973e-05, + "loss": 1.8088, + "step": 307 + }, + { + "epoch": 0.21303821545910426, + "grad_norm": 0.7100119590759277, + "learning_rate": 8.295702840495266e-05, + "loss": 1.9353, + "step": 308 + }, + { + "epoch": 0.21372989797682865, + "grad_norm": 0.7359102964401245, + "learning_rate": 8.288419519300801e-05, + "loss": 1.7844, + "step": 309 + }, + { + "epoch": 0.214421580494553, + "grad_norm": 0.7489526867866516, + "learning_rate": 8.281136198106337e-05, + "loss": 1.6238, + "step": 310 + }, + { + "epoch": 0.21511326301227737, + "grad_norm": 0.8063543438911438, + "learning_rate": 8.273852876911872e-05, + "loss": 1.6634, + "step": 311 + }, + { + "epoch": 0.21580494553000173, + "grad_norm": 0.7559211850166321, + "learning_rate": 8.266569555717407e-05, + "loss": 1.9663, + "step": 312 + }, + { + "epoch": 0.2164966280477261, + "grad_norm": 0.7710449695587158, + "learning_rate": 8.259286234522943e-05, + "loss": 2.0552, + "step": 313 + }, + { + "epoch": 0.21718831056545046, + "grad_norm": 1.0111525058746338, + "learning_rate": 8.252002913328478e-05, + "loss": 1.668, + "step": 314 + }, + { + "epoch": 0.21787999308317482, + "grad_norm": 0.7205057144165039, + "learning_rate": 8.244719592134014e-05, + "loss": 2.1546, + "step": 315 + }, + { + "epoch": 0.21857167560089918, + "grad_norm": 0.9559133052825928, + "learning_rate": 8.237436270939549e-05, + "loss": 1.6799, + "step": 316 + }, + { + "epoch": 0.21926335811862355, + "grad_norm": 0.7418698668479919, + "learning_rate": 8.230152949745085e-05, + "loss": 1.2123, + "step": 317 + }, + { + "epoch": 0.2199550406363479, + "grad_norm": 0.7784197330474854, + "learning_rate": 8.22286962855062e-05, + "loss": 1.1066, + "step": 318 + }, + { + "epoch": 0.22064672315407227, + "grad_norm": 0.754654586315155, + "learning_rate": 8.215586307356154e-05, + "loss": 2.0185, + "step": 319 + }, + { + "epoch": 0.22133840567179663, + "grad_norm": 0.6510013341903687, + "learning_rate": 8.20830298616169e-05, + "loss": 1.8316, + "step": 320 + }, + { + "epoch": 0.222030088189521, + "grad_norm": 1.541455626487732, + "learning_rate": 8.201019664967225e-05, + "loss": 1.3254, + "step": 321 + }, + { + "epoch": 0.2227217707072454, + "grad_norm": 0.8958970308303833, + "learning_rate": 8.193736343772761e-05, + "loss": 1.9616, + "step": 322 + }, + { + "epoch": 0.22341345322496975, + "grad_norm": 0.6273507475852966, + "learning_rate": 8.186453022578296e-05, + "loss": 1.13, + "step": 323 + }, + { + "epoch": 0.2241051357426941, + "grad_norm": 0.7866650223731995, + "learning_rate": 8.17916970138383e-05, + "loss": 1.784, + "step": 324 + }, + { + "epoch": 0.22479681826041847, + "grad_norm": 0.9650758504867554, + "learning_rate": 8.171886380189367e-05, + "loss": 0.9438, + "step": 325 + }, + { + "epoch": 0.22548850077814284, + "grad_norm": 0.7491368055343628, + "learning_rate": 8.164603058994901e-05, + "loss": 1.1472, + "step": 326 + }, + { + "epoch": 0.2261801832958672, + "grad_norm": 0.7516188025474548, + "learning_rate": 8.157319737800438e-05, + "loss": 1.841, + "step": 327 + }, + { + "epoch": 0.22687186581359156, + "grad_norm": 0.8555276989936829, + "learning_rate": 8.150036416605972e-05, + "loss": 2.0819, + "step": 328 + }, + { + "epoch": 0.22756354833131592, + "grad_norm": 0.8025880455970764, + "learning_rate": 8.142753095411508e-05, + "loss": 1.871, + "step": 329 + }, + { + "epoch": 0.22825523084904029, + "grad_norm": 0.8006131052970886, + "learning_rate": 8.135469774217043e-05, + "loss": 1.8293, + "step": 330 + }, + { + "epoch": 0.22894691336676465, + "grad_norm": 0.6898099184036255, + "learning_rate": 8.128186453022578e-05, + "loss": 1.9701, + "step": 331 + }, + { + "epoch": 0.229638595884489, + "grad_norm": 0.7827674150466919, + "learning_rate": 8.120903131828114e-05, + "loss": 1.9634, + "step": 332 + }, + { + "epoch": 0.23033027840221337, + "grad_norm": 0.7763857245445251, + "learning_rate": 8.113619810633649e-05, + "loss": 1.4908, + "step": 333 + }, + { + "epoch": 0.23102196091993774, + "grad_norm": 0.7040197253227234, + "learning_rate": 8.106336489439185e-05, + "loss": 1.9854, + "step": 334 + }, + { + "epoch": 0.23171364343766213, + "grad_norm": 1.0337374210357666, + "learning_rate": 8.09905316824472e-05, + "loss": 2.1542, + "step": 335 + }, + { + "epoch": 0.2324053259553865, + "grad_norm": 0.7228108048439026, + "learning_rate": 8.091769847050256e-05, + "loss": 1.9267, + "step": 336 + }, + { + "epoch": 0.23309700847311085, + "grad_norm": 0.6841875910758972, + "learning_rate": 8.08448652585579e-05, + "loss": 1.769, + "step": 337 + }, + { + "epoch": 0.2337886909908352, + "grad_norm": 0.6749361753463745, + "learning_rate": 8.077203204661325e-05, + "loss": 1.7758, + "step": 338 + }, + { + "epoch": 0.23448037350855958, + "grad_norm": 0.6411470174789429, + "learning_rate": 8.069919883466861e-05, + "loss": 1.2938, + "step": 339 + }, + { + "epoch": 0.23517205602628394, + "grad_norm": 0.6457031965255737, + "learning_rate": 8.062636562272396e-05, + "loss": 1.3416, + "step": 340 + }, + { + "epoch": 0.2358637385440083, + "grad_norm": 0.7273378372192383, + "learning_rate": 8.055353241077932e-05, + "loss": 1.5468, + "step": 341 + }, + { + "epoch": 0.23655542106173266, + "grad_norm": 0.7910048365592957, + "learning_rate": 8.048069919883467e-05, + "loss": 1.8577, + "step": 342 + }, + { + "epoch": 0.23724710357945702, + "grad_norm": 0.7172124981880188, + "learning_rate": 8.040786598689003e-05, + "loss": 2.2877, + "step": 343 + }, + { + "epoch": 0.2379387860971814, + "grad_norm": 0.6511913537979126, + "learning_rate": 8.033503277494538e-05, + "loss": 1.0914, + "step": 344 + }, + { + "epoch": 0.23863046861490575, + "grad_norm": 0.6148791909217834, + "learning_rate": 8.026219956300073e-05, + "loss": 1.3388, + "step": 345 + }, + { + "epoch": 0.2393221511326301, + "grad_norm": 0.7384858131408691, + "learning_rate": 8.018936635105609e-05, + "loss": 1.8186, + "step": 346 + }, + { + "epoch": 0.24001383365035447, + "grad_norm": 0.793538510799408, + "learning_rate": 8.011653313911143e-05, + "loss": 1.5962, + "step": 347 + }, + { + "epoch": 0.24070551616807886, + "grad_norm": 0.6316525936126709, + "learning_rate": 8.00436999271668e-05, + "loss": 1.4712, + "step": 348 + }, + { + "epoch": 0.24139719868580323, + "grad_norm": 0.7671139240264893, + "learning_rate": 7.997086671522214e-05, + "loss": 2.2086, + "step": 349 + }, + { + "epoch": 0.2420888812035276, + "grad_norm": 0.7423043251037598, + "learning_rate": 7.98980335032775e-05, + "loss": 2.1409, + "step": 350 + }, + { + "epoch": 0.24278056372125195, + "grad_norm": 0.834575891494751, + "learning_rate": 7.982520029133285e-05, + "loss": 1.5166, + "step": 351 + }, + { + "epoch": 0.24347224623897631, + "grad_norm": 1.0157177448272705, + "learning_rate": 7.97523670793882e-05, + "loss": 1.7054, + "step": 352 + }, + { + "epoch": 0.24416392875670068, + "grad_norm": 0.8765892386436462, + "learning_rate": 7.967953386744356e-05, + "loss": 2.4397, + "step": 353 + }, + { + "epoch": 0.24485561127442504, + "grad_norm": 0.6456695795059204, + "learning_rate": 7.960670065549891e-05, + "loss": 1.2428, + "step": 354 + }, + { + "epoch": 0.2455472937921494, + "grad_norm": 0.813552737236023, + "learning_rate": 7.953386744355427e-05, + "loss": 1.7926, + "step": 355 + }, + { + "epoch": 0.24623897630987376, + "grad_norm": 0.6702203750610352, + "learning_rate": 7.946103423160962e-05, + "loss": 1.6654, + "step": 356 + }, + { + "epoch": 0.24693065882759813, + "grad_norm": 0.9113941788673401, + "learning_rate": 7.938820101966498e-05, + "loss": 2.1639, + "step": 357 + }, + { + "epoch": 0.2476223413453225, + "grad_norm": 0.6006965041160583, + "learning_rate": 7.931536780772032e-05, + "loss": 1.815, + "step": 358 + }, + { + "epoch": 0.24831402386304685, + "grad_norm": 0.9116283059120178, + "learning_rate": 7.924253459577567e-05, + "loss": 1.5244, + "step": 359 + }, + { + "epoch": 0.2490057063807712, + "grad_norm": 0.8563610315322876, + "learning_rate": 7.916970138383103e-05, + "loss": 2.1287, + "step": 360 + }, + { + "epoch": 0.2496973888984956, + "grad_norm": 0.7254488468170166, + "learning_rate": 7.909686817188638e-05, + "loss": 1.1997, + "step": 361 + }, + { + "epoch": 0.25038907141621997, + "grad_norm": 0.8156582117080688, + "learning_rate": 7.902403495994174e-05, + "loss": 2.1275, + "step": 362 + }, + { + "epoch": 0.2510807539339443, + "grad_norm": 0.7504958510398865, + "learning_rate": 7.895120174799709e-05, + "loss": 2.1381, + "step": 363 + }, + { + "epoch": 0.2517724364516687, + "grad_norm": 0.6650155782699585, + "learning_rate": 7.887836853605244e-05, + "loss": 0.9823, + "step": 364 + }, + { + "epoch": 0.252464118969393, + "grad_norm": 0.6706410646438599, + "learning_rate": 7.88055353241078e-05, + "loss": 1.1854, + "step": 365 + }, + { + "epoch": 0.2531558014871174, + "grad_norm": 1.0841103792190552, + "learning_rate": 7.873270211216315e-05, + "loss": 1.4933, + "step": 366 + }, + { + "epoch": 0.25384748400484175, + "grad_norm": 0.7275784611701965, + "learning_rate": 7.86598689002185e-05, + "loss": 2.162, + "step": 367 + }, + { + "epoch": 0.25453916652256614, + "grad_norm": 0.994987964630127, + "learning_rate": 7.858703568827385e-05, + "loss": 1.6992, + "step": 368 + }, + { + "epoch": 0.25523084904029053, + "grad_norm": 1.160749912261963, + "learning_rate": 7.851420247632922e-05, + "loss": 1.2521, + "step": 369 + }, + { + "epoch": 0.25592253155801487, + "grad_norm": 0.8106790781021118, + "learning_rate": 7.844136926438456e-05, + "loss": 1.5949, + "step": 370 + }, + { + "epoch": 0.25661421407573926, + "grad_norm": 0.5805374383926392, + "learning_rate": 7.836853605243991e-05, + "loss": 1.713, + "step": 371 + }, + { + "epoch": 0.2573058965934636, + "grad_norm": 0.6805906295776367, + "learning_rate": 7.829570284049527e-05, + "loss": 1.7713, + "step": 372 + }, + { + "epoch": 0.257997579111188, + "grad_norm": 0.6964860558509827, + "learning_rate": 7.822286962855062e-05, + "loss": 1.8013, + "step": 373 + }, + { + "epoch": 0.2586892616289123, + "grad_norm": 1.2079120874404907, + "learning_rate": 7.815003641660598e-05, + "loss": 2.0171, + "step": 374 + }, + { + "epoch": 0.2593809441466367, + "grad_norm": 0.689630925655365, + "learning_rate": 7.807720320466133e-05, + "loss": 1.8816, + "step": 375 + }, + { + "epoch": 0.26007262666436104, + "grad_norm": 0.9079475402832031, + "learning_rate": 7.800436999271669e-05, + "loss": 1.5992, + "step": 376 + }, + { + "epoch": 0.26076430918208543, + "grad_norm": 0.4846475422382355, + "learning_rate": 7.793153678077204e-05, + "loss": 0.9038, + "step": 377 + }, + { + "epoch": 0.26145599169980976, + "grad_norm": 0.6311324238777161, + "learning_rate": 7.785870356882738e-05, + "loss": 1.757, + "step": 378 + }, + { + "epoch": 0.26214767421753415, + "grad_norm": 0.6497640609741211, + "learning_rate": 7.778587035688274e-05, + "loss": 1.799, + "step": 379 + }, + { + "epoch": 0.2628393567352585, + "grad_norm": 0.6967574954032898, + "learning_rate": 7.771303714493809e-05, + "loss": 2.0541, + "step": 380 + }, + { + "epoch": 0.2635310392529829, + "grad_norm": 0.8538269400596619, + "learning_rate": 7.764020393299345e-05, + "loss": 1.4781, + "step": 381 + }, + { + "epoch": 0.26422272177070727, + "grad_norm": 0.747353196144104, + "learning_rate": 7.75673707210488e-05, + "loss": 1.063, + "step": 382 + }, + { + "epoch": 0.2649144042884316, + "grad_norm": 0.7110087871551514, + "learning_rate": 7.749453750910416e-05, + "loss": 1.6301, + "step": 383 + }, + { + "epoch": 0.265606086806156, + "grad_norm": 0.8608129620552063, + "learning_rate": 7.742170429715951e-05, + "loss": 1.7459, + "step": 384 + }, + { + "epoch": 0.26629776932388033, + "grad_norm": 0.903325617313385, + "learning_rate": 7.734887108521486e-05, + "loss": 1.0401, + "step": 385 + }, + { + "epoch": 0.2669894518416047, + "grad_norm": 0.6887355446815491, + "learning_rate": 7.727603787327022e-05, + "loss": 1.6558, + "step": 386 + }, + { + "epoch": 0.26768113435932905, + "grad_norm": 0.8748055100440979, + "learning_rate": 7.720320466132557e-05, + "loss": 2.2414, + "step": 387 + }, + { + "epoch": 0.26837281687705344, + "grad_norm": 0.7315294146537781, + "learning_rate": 7.713037144938093e-05, + "loss": 1.7651, + "step": 388 + }, + { + "epoch": 0.2690644993947778, + "grad_norm": 0.7685129642486572, + "learning_rate": 7.705753823743627e-05, + "loss": 1.64, + "step": 389 + }, + { + "epoch": 0.26975618191250217, + "grad_norm": 0.7387799620628357, + "learning_rate": 7.698470502549163e-05, + "loss": 1.655, + "step": 390 + }, + { + "epoch": 0.2704478644302265, + "grad_norm": 0.6962683200836182, + "learning_rate": 7.691187181354698e-05, + "loss": 2.0759, + "step": 391 + }, + { + "epoch": 0.2711395469479509, + "grad_norm": 0.6928436160087585, + "learning_rate": 7.683903860160233e-05, + "loss": 1.902, + "step": 392 + }, + { + "epoch": 0.27183122946567523, + "grad_norm": 0.6286187767982483, + "learning_rate": 7.676620538965769e-05, + "loss": 1.1205, + "step": 393 + }, + { + "epoch": 0.2725229119833996, + "grad_norm": 0.8087384104728699, + "learning_rate": 7.669337217771304e-05, + "loss": 1.5928, + "step": 394 + }, + { + "epoch": 0.273214594501124, + "grad_norm": 0.6950211524963379, + "learning_rate": 7.66205389657684e-05, + "loss": 1.9022, + "step": 395 + }, + { + "epoch": 0.27390627701884834, + "grad_norm": 0.716234028339386, + "learning_rate": 7.654770575382375e-05, + "loss": 1.9103, + "step": 396 + }, + { + "epoch": 0.27459795953657273, + "grad_norm": 1.0911515951156616, + "learning_rate": 7.647487254187911e-05, + "loss": 2.1727, + "step": 397 + }, + { + "epoch": 0.27528964205429707, + "grad_norm": 1.0012110471725464, + "learning_rate": 7.640203932993446e-05, + "loss": 2.0671, + "step": 398 + }, + { + "epoch": 0.27598132457202146, + "grad_norm": 0.7271398305892944, + "learning_rate": 7.63292061179898e-05, + "loss": 1.8487, + "step": 399 + }, + { + "epoch": 0.2766730070897458, + "grad_norm": 0.9112443923950195, + "learning_rate": 7.625637290604516e-05, + "loss": 1.6583, + "step": 400 + }, + { + "epoch": 0.2773646896074702, + "grad_norm": 0.8109308481216431, + "learning_rate": 7.618353969410051e-05, + "loss": 1.6202, + "step": 401 + }, + { + "epoch": 0.2780563721251945, + "grad_norm": 0.7934839129447937, + "learning_rate": 7.611070648215587e-05, + "loss": 2.2605, + "step": 402 + }, + { + "epoch": 0.2787480546429189, + "grad_norm": 0.7166281342506409, + "learning_rate": 7.603787327021122e-05, + "loss": 1.5321, + "step": 403 + }, + { + "epoch": 0.27943973716064324, + "grad_norm": 0.6386390924453735, + "learning_rate": 7.596504005826657e-05, + "loss": 0.9431, + "step": 404 + }, + { + "epoch": 0.28013141967836763, + "grad_norm": 1.2669974565505981, + "learning_rate": 7.589220684632193e-05, + "loss": 0.7634, + "step": 405 + }, + { + "epoch": 0.28082310219609197, + "grad_norm": 0.7032380700111389, + "learning_rate": 7.581937363437728e-05, + "loss": 1.9926, + "step": 406 + }, + { + "epoch": 0.28151478471381636, + "grad_norm": 0.648324191570282, + "learning_rate": 7.574654042243264e-05, + "loss": 1.2027, + "step": 407 + }, + { + "epoch": 0.28220646723154075, + "grad_norm": 0.7746976613998413, + "learning_rate": 7.567370721048798e-05, + "loss": 1.3615, + "step": 408 + }, + { + "epoch": 0.2828981497492651, + "grad_norm": 0.8174494504928589, + "learning_rate": 7.560087399854335e-05, + "loss": 1.9945, + "step": 409 + }, + { + "epoch": 0.28358983226698947, + "grad_norm": 0.9701315760612488, + "learning_rate": 7.55280407865987e-05, + "loss": 1.4894, + "step": 410 + }, + { + "epoch": 0.2842815147847138, + "grad_norm": 0.9906868934631348, + "learning_rate": 7.545520757465404e-05, + "loss": 1.7937, + "step": 411 + }, + { + "epoch": 0.2849731973024382, + "grad_norm": 0.6380108594894409, + "learning_rate": 7.53823743627094e-05, + "loss": 1.6794, + "step": 412 + }, + { + "epoch": 0.28566487982016253, + "grad_norm": 0.6367055177688599, + "learning_rate": 7.530954115076475e-05, + "loss": 1.4019, + "step": 413 + }, + { + "epoch": 0.2863565623378869, + "grad_norm": 0.9779865741729736, + "learning_rate": 7.523670793882011e-05, + "loss": 1.8048, + "step": 414 + }, + { + "epoch": 0.28704824485561126, + "grad_norm": 0.6246545910835266, + "learning_rate": 7.516387472687546e-05, + "loss": 1.7138, + "step": 415 + }, + { + "epoch": 0.28773992737333565, + "grad_norm": 0.7418220639228821, + "learning_rate": 7.509104151493082e-05, + "loss": 1.9917, + "step": 416 + }, + { + "epoch": 0.28843160989106, + "grad_norm": 0.7883859276771545, + "learning_rate": 7.501820830298617e-05, + "loss": 1.2134, + "step": 417 + }, + { + "epoch": 0.28912329240878437, + "grad_norm": 0.8116041421890259, + "learning_rate": 7.494537509104151e-05, + "loss": 1.1634, + "step": 418 + }, + { + "epoch": 0.2898149749265087, + "grad_norm": 0.7708969712257385, + "learning_rate": 7.487254187909688e-05, + "loss": 1.8895, + "step": 419 + }, + { + "epoch": 0.2905066574442331, + "grad_norm": 1.3700909614562988, + "learning_rate": 7.479970866715222e-05, + "loss": 1.1294, + "step": 420 + }, + { + "epoch": 0.2911983399619575, + "grad_norm": 0.8107597827911377, + "learning_rate": 7.472687545520758e-05, + "loss": 1.0097, + "step": 421 + }, + { + "epoch": 0.2918900224796818, + "grad_norm": 0.8277159929275513, + "learning_rate": 7.465404224326293e-05, + "loss": 1.0633, + "step": 422 + }, + { + "epoch": 0.2925817049974062, + "grad_norm": 0.75505131483078, + "learning_rate": 7.458120903131829e-05, + "loss": 1.4971, + "step": 423 + }, + { + "epoch": 0.29327338751513055, + "grad_norm": 0.7761143445968628, + "learning_rate": 7.450837581937363e-05, + "loss": 1.569, + "step": 424 + }, + { + "epoch": 0.29396507003285494, + "grad_norm": 0.8429370522499084, + "learning_rate": 7.443554260742899e-05, + "loss": 2.147, + "step": 425 + }, + { + "epoch": 0.29465675255057927, + "grad_norm": 0.7506592273712158, + "learning_rate": 7.436270939548435e-05, + "loss": 2.1514, + "step": 426 + }, + { + "epoch": 0.29534843506830366, + "grad_norm": 0.6654348969459534, + "learning_rate": 7.42898761835397e-05, + "loss": 1.7697, + "step": 427 + }, + { + "epoch": 0.296040117586028, + "grad_norm": 1.179040551185608, + "learning_rate": 7.421704297159506e-05, + "loss": 1.6906, + "step": 428 + }, + { + "epoch": 0.2967318001037524, + "grad_norm": 0.6145846247673035, + "learning_rate": 7.41442097596504e-05, + "loss": 1.8308, + "step": 429 + }, + { + "epoch": 0.2974234826214767, + "grad_norm": 0.7995002269744873, + "learning_rate": 7.407137654770577e-05, + "loss": 1.7942, + "step": 430 + }, + { + "epoch": 0.2981151651392011, + "grad_norm": 0.707095742225647, + "learning_rate": 7.39985433357611e-05, + "loss": 1.6325, + "step": 431 + }, + { + "epoch": 0.29880684765692545, + "grad_norm": 0.5955487489700317, + "learning_rate": 7.392571012381646e-05, + "loss": 1.3466, + "step": 432 + }, + { + "epoch": 0.29949853017464984, + "grad_norm": 0.9258183240890503, + "learning_rate": 7.385287691187182e-05, + "loss": 1.8544, + "step": 433 + }, + { + "epoch": 0.3001902126923742, + "grad_norm": 0.8435434103012085, + "learning_rate": 7.378004369992717e-05, + "loss": 1.8884, + "step": 434 + }, + { + "epoch": 0.30088189521009856, + "grad_norm": 0.7008845210075378, + "learning_rate": 7.370721048798253e-05, + "loss": 1.8409, + "step": 435 + }, + { + "epoch": 0.30157357772782295, + "grad_norm": 0.7846685647964478, + "learning_rate": 7.363437727603788e-05, + "loss": 1.8821, + "step": 436 + }, + { + "epoch": 0.3022652602455473, + "grad_norm": 0.7767228484153748, + "learning_rate": 7.356154406409324e-05, + "loss": 2.0034, + "step": 437 + }, + { + "epoch": 0.3029569427632717, + "grad_norm": 0.787028431892395, + "learning_rate": 7.348871085214857e-05, + "loss": 1.8664, + "step": 438 + }, + { + "epoch": 0.303648625280996, + "grad_norm": 0.6583303809165955, + "learning_rate": 7.341587764020393e-05, + "loss": 1.1991, + "step": 439 + }, + { + "epoch": 0.3043403077987204, + "grad_norm": 0.8681322932243347, + "learning_rate": 7.33430444282593e-05, + "loss": 1.7124, + "step": 440 + }, + { + "epoch": 0.30503199031644473, + "grad_norm": 0.9781901240348816, + "learning_rate": 7.327021121631464e-05, + "loss": 2.0128, + "step": 441 + }, + { + "epoch": 0.3057236728341691, + "grad_norm": 0.6712149381637573, + "learning_rate": 7.319737800437e-05, + "loss": 1.4835, + "step": 442 + }, + { + "epoch": 0.30641535535189346, + "grad_norm": 0.7322747111320496, + "learning_rate": 7.312454479242535e-05, + "loss": 1.1381, + "step": 443 + }, + { + "epoch": 0.30710703786961785, + "grad_norm": 0.8576450347900391, + "learning_rate": 7.30517115804807e-05, + "loss": 1.8142, + "step": 444 + }, + { + "epoch": 0.3077987203873422, + "grad_norm": 0.6087522506713867, + "learning_rate": 7.297887836853605e-05, + "loss": 1.2711, + "step": 445 + }, + { + "epoch": 0.3084904029050666, + "grad_norm": 0.8685612082481384, + "learning_rate": 7.290604515659141e-05, + "loss": 1.2365, + "step": 446 + }, + { + "epoch": 0.30918208542279096, + "grad_norm": 0.7040896415710449, + "learning_rate": 7.283321194464677e-05, + "loss": 1.6385, + "step": 447 + }, + { + "epoch": 0.3098737679405153, + "grad_norm": 0.943147599697113, + "learning_rate": 7.276037873270212e-05, + "loss": 1.4699, + "step": 448 + }, + { + "epoch": 0.3105654504582397, + "grad_norm": 0.6641246676445007, + "learning_rate": 7.268754552075748e-05, + "loss": 1.6489, + "step": 449 + }, + { + "epoch": 0.311257132975964, + "grad_norm": 0.7420269250869751, + "learning_rate": 7.261471230881282e-05, + "loss": 2.1525, + "step": 450 + }, + { + "epoch": 0.3119488154936884, + "grad_norm": 0.7198522090911865, + "learning_rate": 7.254187909686817e-05, + "loss": 2.01, + "step": 451 + }, + { + "epoch": 0.31264049801141275, + "grad_norm": 0.8209108114242554, + "learning_rate": 7.246904588492352e-05, + "loss": 1.5205, + "step": 452 + }, + { + "epoch": 0.31333218052913714, + "grad_norm": 1.0207217931747437, + "learning_rate": 7.239621267297888e-05, + "loss": 1.589, + "step": 453 + }, + { + "epoch": 0.3140238630468615, + "grad_norm": 0.7939510345458984, + "learning_rate": 7.232337946103424e-05, + "loss": 1.9908, + "step": 454 + }, + { + "epoch": 0.31471554556458586, + "grad_norm": 0.6668927073478699, + "learning_rate": 7.225054624908959e-05, + "loss": 1.7476, + "step": 455 + }, + { + "epoch": 0.3154072280823102, + "grad_norm": 0.8068589568138123, + "learning_rate": 7.217771303714495e-05, + "loss": 1.7593, + "step": 456 + }, + { + "epoch": 0.3160989106000346, + "grad_norm": 0.6828214526176453, + "learning_rate": 7.21048798252003e-05, + "loss": 1.5636, + "step": 457 + }, + { + "epoch": 0.3167905931177589, + "grad_norm": 0.7853332757949829, + "learning_rate": 7.203204661325565e-05, + "loss": 1.1448, + "step": 458 + }, + { + "epoch": 0.3174822756354833, + "grad_norm": 1.0938345193862915, + "learning_rate": 7.195921340131099e-05, + "loss": 1.7145, + "step": 459 + }, + { + "epoch": 0.3181739581532077, + "grad_norm": 0.7735335826873779, + "learning_rate": 7.188638018936635e-05, + "loss": 1.2118, + "step": 460 + }, + { + "epoch": 0.31886564067093204, + "grad_norm": 0.6759296655654907, + "learning_rate": 7.181354697742171e-05, + "loss": 1.1214, + "step": 461 + }, + { + "epoch": 0.31955732318865643, + "grad_norm": 0.8163008689880371, + "learning_rate": 7.174071376547706e-05, + "loss": 1.9071, + "step": 462 + }, + { + "epoch": 0.32024900570638076, + "grad_norm": 0.8205277919769287, + "learning_rate": 7.166788055353242e-05, + "loss": 1.0775, + "step": 463 + }, + { + "epoch": 0.32094068822410515, + "grad_norm": 0.712774395942688, + "learning_rate": 7.159504734158776e-05, + "loss": 1.8912, + "step": 464 + }, + { + "epoch": 0.3216323707418295, + "grad_norm": 0.7423759698867798, + "learning_rate": 7.152221412964312e-05, + "loss": 2.0655, + "step": 465 + }, + { + "epoch": 0.3223240532595539, + "grad_norm": 0.6948613524436951, + "learning_rate": 7.144938091769847e-05, + "loss": 2.1957, + "step": 466 + }, + { + "epoch": 0.3230157357772782, + "grad_norm": 0.6164250373840332, + "learning_rate": 7.137654770575383e-05, + "loss": 0.9997, + "step": 467 + }, + { + "epoch": 0.3237074182950026, + "grad_norm": 0.7659111022949219, + "learning_rate": 7.130371449380919e-05, + "loss": 1.4006, + "step": 468 + }, + { + "epoch": 0.32439910081272694, + "grad_norm": 0.830442488193512, + "learning_rate": 7.123088128186454e-05, + "loss": 1.6562, + "step": 469 + }, + { + "epoch": 0.3250907833304513, + "grad_norm": 0.649169921875, + "learning_rate": 7.11580480699199e-05, + "loss": 1.8275, + "step": 470 + }, + { + "epoch": 0.32578246584817566, + "grad_norm": 0.7854671478271484, + "learning_rate": 7.108521485797523e-05, + "loss": 0.3147, + "step": 471 + }, + { + "epoch": 0.32647414836590005, + "grad_norm": 1.0381858348846436, + "learning_rate": 7.101238164603059e-05, + "loss": 2.0842, + "step": 472 + }, + { + "epoch": 0.32716583088362444, + "grad_norm": 0.7317079305648804, + "learning_rate": 7.093954843408594e-05, + "loss": 2.1326, + "step": 473 + }, + { + "epoch": 0.3278575134013488, + "grad_norm": 0.73940110206604, + "learning_rate": 7.08667152221413e-05, + "loss": 1.9647, + "step": 474 + }, + { + "epoch": 0.32854919591907317, + "grad_norm": 0.8855507969856262, + "learning_rate": 7.079388201019666e-05, + "loss": 1.9308, + "step": 475 + }, + { + "epoch": 0.3292408784367975, + "grad_norm": 0.7795459628105164, + "learning_rate": 7.072104879825201e-05, + "loss": 1.9297, + "step": 476 + }, + { + "epoch": 0.3299325609545219, + "grad_norm": 0.765084981918335, + "learning_rate": 7.064821558630737e-05, + "loss": 2.3063, + "step": 477 + }, + { + "epoch": 0.3306242434722462, + "grad_norm": 0.7164610624313354, + "learning_rate": 7.05753823743627e-05, + "loss": 1.8509, + "step": 478 + }, + { + "epoch": 0.3313159259899706, + "grad_norm": 0.6976935863494873, + "learning_rate": 7.050254916241806e-05, + "loss": 1.8351, + "step": 479 + }, + { + "epoch": 0.33200760850769495, + "grad_norm": 1.1228684186935425, + "learning_rate": 7.042971595047341e-05, + "loss": 1.4274, + "step": 480 + }, + { + "epoch": 0.33269929102541934, + "grad_norm": 0.7229692935943604, + "learning_rate": 7.035688273852877e-05, + "loss": 1.5104, + "step": 481 + }, + { + "epoch": 0.3333909735431437, + "grad_norm": 0.7485585808753967, + "learning_rate": 7.028404952658413e-05, + "loss": 1.1417, + "step": 482 + }, + { + "epoch": 0.33408265606086807, + "grad_norm": 1.7345014810562134, + "learning_rate": 7.021121631463948e-05, + "loss": 1.6614, + "step": 483 + }, + { + "epoch": 0.3347743385785924, + "grad_norm": 0.7559076547622681, + "learning_rate": 7.013838310269483e-05, + "loss": 1.8508, + "step": 484 + }, + { + "epoch": 0.3354660210963168, + "grad_norm": 1.2972337007522583, + "learning_rate": 7.006554989075018e-05, + "loss": 1.9322, + "step": 485 + }, + { + "epoch": 0.3361577036140412, + "grad_norm": 0.9719124436378479, + "learning_rate": 6.999271667880554e-05, + "loss": 1.6138, + "step": 486 + }, + { + "epoch": 0.3368493861317655, + "grad_norm": 0.7772285342216492, + "learning_rate": 6.991988346686089e-05, + "loss": 2.1904, + "step": 487 + }, + { + "epoch": 0.3375410686494899, + "grad_norm": 0.8187295198440552, + "learning_rate": 6.984705025491625e-05, + "loss": 1.4893, + "step": 488 + }, + { + "epoch": 0.33823275116721424, + "grad_norm": 0.8635819554328918, + "learning_rate": 6.977421704297161e-05, + "loss": 2.0885, + "step": 489 + }, + { + "epoch": 0.33892443368493863, + "grad_norm": 1.2180774211883545, + "learning_rate": 6.970138383102696e-05, + "loss": 1.1084, + "step": 490 + }, + { + "epoch": 0.33961611620266297, + "grad_norm": 0.8329484462738037, + "learning_rate": 6.96285506190823e-05, + "loss": 1.3746, + "step": 491 + }, + { + "epoch": 0.34030779872038736, + "grad_norm": 0.7301535606384277, + "learning_rate": 6.955571740713765e-05, + "loss": 1.2109, + "step": 492 + }, + { + "epoch": 0.3409994812381117, + "grad_norm": 0.740479052066803, + "learning_rate": 6.948288419519301e-05, + "loss": 1.9807, + "step": 493 + }, + { + "epoch": 0.3416911637558361, + "grad_norm": 0.9610648155212402, + "learning_rate": 6.941005098324836e-05, + "loss": 1.5924, + "step": 494 + }, + { + "epoch": 0.3423828462735604, + "grad_norm": 0.7890329360961914, + "learning_rate": 6.933721777130372e-05, + "loss": 1.7208, + "step": 495 + }, + { + "epoch": 0.3430745287912848, + "grad_norm": 0.7466354370117188, + "learning_rate": 6.926438455935908e-05, + "loss": 1.8483, + "step": 496 + }, + { + "epoch": 0.34376621130900914, + "grad_norm": 0.8114244341850281, + "learning_rate": 6.919155134741443e-05, + "loss": 1.9642, + "step": 497 + }, + { + "epoch": 0.34445789382673353, + "grad_norm": 0.8441615104675293, + "learning_rate": 6.911871813546978e-05, + "loss": 1.6589, + "step": 498 + }, + { + "epoch": 0.3451495763444579, + "grad_norm": 0.771629273891449, + "learning_rate": 6.904588492352512e-05, + "loss": 2.1515, + "step": 499 + }, + { + "epoch": 0.34584125886218225, + "grad_norm": 0.7773415446281433, + "learning_rate": 6.897305171158048e-05, + "loss": 1.1782, + "step": 500 + }, + { + "epoch": 0.34584125886218225, + "eval_loss": 1.5188031196594238, + "eval_runtime": 636.0478, + "eval_samples_per_second": 2.02, + "eval_steps_per_second": 1.011, + "step": 500 + }, + { + "epoch": 0.34653294137990664, + "grad_norm": 0.8539692759513855, + "learning_rate": 6.890021849963583e-05, + "loss": 2.2602, + "step": 501 + }, + { + "epoch": 0.347224623897631, + "grad_norm": 0.7637029886245728, + "learning_rate": 6.882738528769119e-05, + "loss": 1.8441, + "step": 502 + }, + { + "epoch": 0.34791630641535537, + "grad_norm": 0.7741591334342957, + "learning_rate": 6.875455207574655e-05, + "loss": 1.876, + "step": 503 + }, + { + "epoch": 0.3486079889330797, + "grad_norm": 0.7942149639129639, + "learning_rate": 6.868171886380189e-05, + "loss": 1.2218, + "step": 504 + }, + { + "epoch": 0.3492996714508041, + "grad_norm": 0.7626279592514038, + "learning_rate": 6.860888565185725e-05, + "loss": 1.563, + "step": 505 + }, + { + "epoch": 0.34999135396852843, + "grad_norm": 0.6218715310096741, + "learning_rate": 6.85360524399126e-05, + "loss": 1.7303, + "step": 506 + }, + { + "epoch": 0.3506830364862528, + "grad_norm": 0.8599730134010315, + "learning_rate": 6.846321922796796e-05, + "loss": 1.3347, + "step": 507 + }, + { + "epoch": 0.35137471900397715, + "grad_norm": 0.5834329128265381, + "learning_rate": 6.83903860160233e-05, + "loss": 1.1654, + "step": 508 + }, + { + "epoch": 0.35206640152170154, + "grad_norm": 0.8399999737739563, + "learning_rate": 6.831755280407867e-05, + "loss": 1.4099, + "step": 509 + }, + { + "epoch": 0.3527580840394259, + "grad_norm": 0.7133538126945496, + "learning_rate": 6.824471959213403e-05, + "loss": 0.9115, + "step": 510 + }, + { + "epoch": 0.35344976655715027, + "grad_norm": 0.7058927416801453, + "learning_rate": 6.817188638018936e-05, + "loss": 1.4206, + "step": 511 + }, + { + "epoch": 0.35414144907487466, + "grad_norm": 1.1560832262039185, + "learning_rate": 6.809905316824472e-05, + "loss": 1.5635, + "step": 512 + }, + { + "epoch": 0.354833131592599, + "grad_norm": 0.7324162721633911, + "learning_rate": 6.802621995630007e-05, + "loss": 2.045, + "step": 513 + }, + { + "epoch": 0.3555248141103234, + "grad_norm": 0.6711490750312805, + "learning_rate": 6.795338674435543e-05, + "loss": 2.0891, + "step": 514 + }, + { + "epoch": 0.3562164966280477, + "grad_norm": 0.795870304107666, + "learning_rate": 6.788055353241078e-05, + "loss": 2.1285, + "step": 515 + }, + { + "epoch": 0.3569081791457721, + "grad_norm": 1.913321852684021, + "learning_rate": 6.780772032046614e-05, + "loss": 2.1861, + "step": 516 + }, + { + "epoch": 0.35759986166349644, + "grad_norm": 0.7828819751739502, + "learning_rate": 6.77348871085215e-05, + "loss": 1.4283, + "step": 517 + }, + { + "epoch": 0.35829154418122083, + "grad_norm": 0.7590184211730957, + "learning_rate": 6.766205389657683e-05, + "loss": 1.6581, + "step": 518 + }, + { + "epoch": 0.35898322669894517, + "grad_norm": 0.8273248672485352, + "learning_rate": 6.75892206846322e-05, + "loss": 2.0783, + "step": 519 + }, + { + "epoch": 0.35967490921666956, + "grad_norm": 0.791447639465332, + "learning_rate": 6.751638747268754e-05, + "loss": 1.2507, + "step": 520 + }, + { + "epoch": 0.3603665917343939, + "grad_norm": 0.7864679098129272, + "learning_rate": 6.74435542607429e-05, + "loss": 1.4686, + "step": 521 + }, + { + "epoch": 0.3610582742521183, + "grad_norm": 0.7522057294845581, + "learning_rate": 6.737072104879825e-05, + "loss": 1.4506, + "step": 522 + }, + { + "epoch": 0.3617499567698426, + "grad_norm": 0.6190770864486694, + "learning_rate": 6.729788783685361e-05, + "loss": 2.0413, + "step": 523 + }, + { + "epoch": 0.362441639287567, + "grad_norm": 0.9738603830337524, + "learning_rate": 6.722505462490896e-05, + "loss": 1.315, + "step": 524 + }, + { + "epoch": 0.3631333218052914, + "grad_norm": 0.7005712389945984, + "learning_rate": 6.715222141296431e-05, + "loss": 1.528, + "step": 525 + }, + { + "epoch": 0.36382500432301573, + "grad_norm": 1.2126790285110474, + "learning_rate": 6.707938820101967e-05, + "loss": 2.0808, + "step": 526 + }, + { + "epoch": 0.3645166868407401, + "grad_norm": 0.9661760330200195, + "learning_rate": 6.700655498907502e-05, + "loss": 2.0187, + "step": 527 + }, + { + "epoch": 0.36520836935846446, + "grad_norm": 0.7546457648277283, + "learning_rate": 6.693372177713038e-05, + "loss": 1.4964, + "step": 528 + }, + { + "epoch": 0.36590005187618885, + "grad_norm": 0.8270094394683838, + "learning_rate": 6.686088856518573e-05, + "loss": 1.7801, + "step": 529 + }, + { + "epoch": 0.3665917343939132, + "grad_norm": 0.7802672386169434, + "learning_rate": 6.678805535324109e-05, + "loss": 1.4457, + "step": 530 + }, + { + "epoch": 0.3672834169116376, + "grad_norm": 0.7970958948135376, + "learning_rate": 6.671522214129643e-05, + "loss": 1.5796, + "step": 531 + }, + { + "epoch": 0.3679750994293619, + "grad_norm": 0.5904183983802795, + "learning_rate": 6.664238892935178e-05, + "loss": 1.013, + "step": 532 + }, + { + "epoch": 0.3686667819470863, + "grad_norm": 0.8535469174385071, + "learning_rate": 6.656955571740714e-05, + "loss": 1.3146, + "step": 533 + }, + { + "epoch": 0.36935846446481063, + "grad_norm": 0.722558319568634, + "learning_rate": 6.649672250546249e-05, + "loss": 1.2695, + "step": 534 + }, + { + "epoch": 0.370050146982535, + "grad_norm": 0.7470422983169556, + "learning_rate": 6.642388929351785e-05, + "loss": 2.1384, + "step": 535 + }, + { + "epoch": 0.37074182950025936, + "grad_norm": 1.035486102104187, + "learning_rate": 6.63510560815732e-05, + "loss": 1.6666, + "step": 536 + }, + { + "epoch": 0.37143351201798375, + "grad_norm": 0.8261759281158447, + "learning_rate": 6.627822286962856e-05, + "loss": 1.9752, + "step": 537 + }, + { + "epoch": 0.37212519453570814, + "grad_norm": 0.8254296183586121, + "learning_rate": 6.620538965768391e-05, + "loss": 1.9392, + "step": 538 + }, + { + "epoch": 0.37281687705343247, + "grad_norm": 0.9176802635192871, + "learning_rate": 6.613255644573925e-05, + "loss": 2.1566, + "step": 539 + }, + { + "epoch": 0.37350855957115686, + "grad_norm": 0.6909226179122925, + "learning_rate": 6.605972323379462e-05, + "loss": 1.4706, + "step": 540 + }, + { + "epoch": 0.3742002420888812, + "grad_norm": 1.0335818529129028, + "learning_rate": 6.598689002184996e-05, + "loss": 2.4501, + "step": 541 + }, + { + "epoch": 0.3748919246066056, + "grad_norm": 0.9764876365661621, + "learning_rate": 6.591405680990532e-05, + "loss": 2.2261, + "step": 542 + }, + { + "epoch": 0.3755836071243299, + "grad_norm": 0.8037734627723694, + "learning_rate": 6.584122359796067e-05, + "loss": 1.6487, + "step": 543 + }, + { + "epoch": 0.3762752896420543, + "grad_norm": 0.9658376574516296, + "learning_rate": 6.576839038601602e-05, + "loss": 2.0596, + "step": 544 + }, + { + "epoch": 0.37696697215977865, + "grad_norm": 0.7336639165878296, + "learning_rate": 6.569555717407138e-05, + "loss": 1.9689, + "step": 545 + }, + { + "epoch": 0.37765865467750304, + "grad_norm": 0.6894165873527527, + "learning_rate": 6.562272396212673e-05, + "loss": 1.8499, + "step": 546 + }, + { + "epoch": 0.37835033719522737, + "grad_norm": 0.7265276908874512, + "learning_rate": 6.554989075018209e-05, + "loss": 2.111, + "step": 547 + }, + { + "epoch": 0.37904201971295176, + "grad_norm": 0.8299263715744019, + "learning_rate": 6.547705753823744e-05, + "loss": 2.0565, + "step": 548 + }, + { + "epoch": 0.3797337022306761, + "grad_norm": 0.8597300052642822, + "learning_rate": 6.54042243262928e-05, + "loss": 1.7233, + "step": 549 + }, + { + "epoch": 0.3804253847484005, + "grad_norm": 0.7230535745620728, + "learning_rate": 6.533139111434814e-05, + "loss": 1.2764, + "step": 550 + }, + { + "epoch": 0.3811170672661249, + "grad_norm": 0.8922619819641113, + "learning_rate": 6.525855790240349e-05, + "loss": 2.1088, + "step": 551 + }, + { + "epoch": 0.3818087497838492, + "grad_norm": 0.7452985048294067, + "learning_rate": 6.518572469045885e-05, + "loss": 1.5725, + "step": 552 + }, + { + "epoch": 0.3825004323015736, + "grad_norm": 0.8344093561172485, + "learning_rate": 6.51128914785142e-05, + "loss": 1.8517, + "step": 553 + }, + { + "epoch": 0.38319211481929794, + "grad_norm": 0.6585731506347656, + "learning_rate": 6.504005826656956e-05, + "loss": 1.4584, + "step": 554 + }, + { + "epoch": 0.3838837973370223, + "grad_norm": 0.8440024256706238, + "learning_rate": 6.496722505462491e-05, + "loss": 1.547, + "step": 555 + }, + { + "epoch": 0.38457547985474666, + "grad_norm": 0.8570727705955505, + "learning_rate": 6.489439184268027e-05, + "loss": 1.7777, + "step": 556 + }, + { + "epoch": 0.38526716237247105, + "grad_norm": 0.8458102345466614, + "learning_rate": 6.482155863073562e-05, + "loss": 2.1761, + "step": 557 + }, + { + "epoch": 0.3859588448901954, + "grad_norm": 0.9602516293525696, + "learning_rate": 6.474872541879097e-05, + "loss": 1.224, + "step": 558 + }, + { + "epoch": 0.3866505274079198, + "grad_norm": 0.7018793821334839, + "learning_rate": 6.467589220684633e-05, + "loss": 1.8217, + "step": 559 + }, + { + "epoch": 0.3873422099256441, + "grad_norm": 0.796327531337738, + "learning_rate": 6.460305899490167e-05, + "loss": 1.9399, + "step": 560 + }, + { + "epoch": 0.3880338924433685, + "grad_norm": 0.6899397373199463, + "learning_rate": 6.453022578295704e-05, + "loss": 0.8488, + "step": 561 + }, + { + "epoch": 0.38872557496109283, + "grad_norm": 0.6222304701805115, + "learning_rate": 6.445739257101238e-05, + "loss": 0.7811, + "step": 562 + }, + { + "epoch": 0.3894172574788172, + "grad_norm": 0.7914639115333557, + "learning_rate": 6.438455935906774e-05, + "loss": 2.0951, + "step": 563 + }, + { + "epoch": 0.3901089399965416, + "grad_norm": 0.8506953120231628, + "learning_rate": 6.431172614712309e-05, + "loss": 1.3678, + "step": 564 + }, + { + "epoch": 0.39080062251426595, + "grad_norm": 0.6404629349708557, + "learning_rate": 6.423889293517844e-05, + "loss": 1.0092, + "step": 565 + }, + { + "epoch": 0.39149230503199034, + "grad_norm": 0.7604504227638245, + "learning_rate": 6.41660597232338e-05, + "loss": 1.931, + "step": 566 + }, + { + "epoch": 0.3921839875497147, + "grad_norm": 0.7102979421615601, + "learning_rate": 6.409322651128915e-05, + "loss": 1.3211, + "step": 567 + }, + { + "epoch": 0.39287567006743906, + "grad_norm": 0.7402350902557373, + "learning_rate": 6.402039329934451e-05, + "loss": 1.5225, + "step": 568 + }, + { + "epoch": 0.3935673525851634, + "grad_norm": 0.9871742129325867, + "learning_rate": 6.394756008739986e-05, + "loss": 0.7057, + "step": 569 + }, + { + "epoch": 0.3942590351028878, + "grad_norm": 0.7585322856903076, + "learning_rate": 6.387472687545522e-05, + "loss": 1.5484, + "step": 570 + }, + { + "epoch": 0.3949507176206121, + "grad_norm": 0.746465265750885, + "learning_rate": 6.380189366351056e-05, + "loss": 1.4404, + "step": 571 + }, + { + "epoch": 0.3956424001383365, + "grad_norm": 0.7660443186759949, + "learning_rate": 6.372906045156591e-05, + "loss": 1.8797, + "step": 572 + }, + { + "epoch": 0.39633408265606085, + "grad_norm": 0.7403152585029602, + "learning_rate": 6.365622723962127e-05, + "loss": 1.8291, + "step": 573 + }, + { + "epoch": 0.39702576517378524, + "grad_norm": 1.007710576057434, + "learning_rate": 6.358339402767662e-05, + "loss": 1.3267, + "step": 574 + }, + { + "epoch": 0.3977174476915096, + "grad_norm": 0.9685802459716797, + "learning_rate": 6.351056081573198e-05, + "loss": 2.0822, + "step": 575 + }, + { + "epoch": 0.39840913020923396, + "grad_norm": 0.8278066515922546, + "learning_rate": 6.343772760378733e-05, + "loss": 1.3005, + "step": 576 + }, + { + "epoch": 0.39910081272695835, + "grad_norm": 0.7012304067611694, + "learning_rate": 6.336489439184269e-05, + "loss": 0.9913, + "step": 577 + }, + { + "epoch": 0.3997924952446827, + "grad_norm": 1.449615716934204, + "learning_rate": 6.329206117989804e-05, + "loss": 1.3947, + "step": 578 + }, + { + "epoch": 0.4004841777624071, + "grad_norm": 1.1024595499038696, + "learning_rate": 6.321922796795339e-05, + "loss": 1.683, + "step": 579 + }, + { + "epoch": 0.4011758602801314, + "grad_norm": 1.3790334463119507, + "learning_rate": 6.314639475600875e-05, + "loss": 1.3824, + "step": 580 + }, + { + "epoch": 0.4018675427978558, + "grad_norm": 0.754559338092804, + "learning_rate": 6.30735615440641e-05, + "loss": 1.9189, + "step": 581 + }, + { + "epoch": 0.40255922531558014, + "grad_norm": 0.9193564653396606, + "learning_rate": 6.300072833211945e-05, + "loss": 1.7524, + "step": 582 + }, + { + "epoch": 0.40325090783330453, + "grad_norm": 0.6305272579193115, + "learning_rate": 6.29278951201748e-05, + "loss": 1.0415, + "step": 583 + }, + { + "epoch": 0.40394259035102886, + "grad_norm": 0.7553091645240784, + "learning_rate": 6.285506190823015e-05, + "loss": 1.4771, + "step": 584 + }, + { + "epoch": 0.40463427286875325, + "grad_norm": 0.9118616580963135, + "learning_rate": 6.278222869628551e-05, + "loss": 0.9804, + "step": 585 + }, + { + "epoch": 0.4053259553864776, + "grad_norm": 0.9057390093803406, + "learning_rate": 6.270939548434086e-05, + "loss": 1.8537, + "step": 586 + }, + { + "epoch": 0.406017637904202, + "grad_norm": 0.808927595615387, + "learning_rate": 6.263656227239622e-05, + "loss": 1.3161, + "step": 587 + }, + { + "epoch": 0.4067093204219263, + "grad_norm": 0.7657842040061951, + "learning_rate": 6.256372906045157e-05, + "loss": 1.3956, + "step": 588 + }, + { + "epoch": 0.4074010029396507, + "grad_norm": 0.7779245376586914, + "learning_rate": 6.249089584850693e-05, + "loss": 1.5218, + "step": 589 + }, + { + "epoch": 0.40809268545737504, + "grad_norm": 0.813758373260498, + "learning_rate": 6.241806263656228e-05, + "loss": 1.6086, + "step": 590 + }, + { + "epoch": 0.4087843679750994, + "grad_norm": 0.9651972055435181, + "learning_rate": 6.234522942461762e-05, + "loss": 2.0356, + "step": 591 + }, + { + "epoch": 0.4094760504928238, + "grad_norm": 0.9631683826446533, + "learning_rate": 6.227239621267298e-05, + "loss": 2.1457, + "step": 592 + }, + { + "epoch": 0.41016773301054815, + "grad_norm": 0.6648775339126587, + "learning_rate": 6.219956300072833e-05, + "loss": 1.3306, + "step": 593 + }, + { + "epoch": 0.41085941552827254, + "grad_norm": 1.3052250146865845, + "learning_rate": 6.212672978878369e-05, + "loss": 1.8813, + "step": 594 + }, + { + "epoch": 0.4115510980459969, + "grad_norm": 0.8607529401779175, + "learning_rate": 6.205389657683904e-05, + "loss": 1.9435, + "step": 595 + }, + { + "epoch": 0.41224278056372127, + "grad_norm": 0.7908481359481812, + "learning_rate": 6.19810633648944e-05, + "loss": 1.7544, + "step": 596 + }, + { + "epoch": 0.4129344630814456, + "grad_norm": 0.8440048694610596, + "learning_rate": 6.190823015294975e-05, + "loss": 1.8535, + "step": 597 + }, + { + "epoch": 0.41362614559917, + "grad_norm": 1.0176901817321777, + "learning_rate": 6.18353969410051e-05, + "loss": 1.5932, + "step": 598 + }, + { + "epoch": 0.4143178281168943, + "grad_norm": 0.8359641432762146, + "learning_rate": 6.176256372906046e-05, + "loss": 1.6704, + "step": 599 + }, + { + "epoch": 0.4150095106346187, + "grad_norm": 0.8534523248672485, + "learning_rate": 6.16897305171158e-05, + "loss": 1.7888, + "step": 600 + }, + { + "epoch": 0.41570119315234305, + "grad_norm": 0.8510312438011169, + "learning_rate": 6.161689730517117e-05, + "loss": 1.492, + "step": 601 + }, + { + "epoch": 0.41639287567006744, + "grad_norm": 0.9183124303817749, + "learning_rate": 6.154406409322651e-05, + "loss": 1.1547, + "step": 602 + }, + { + "epoch": 0.4170845581877918, + "grad_norm": 0.7638151049613953, + "learning_rate": 6.147123088128187e-05, + "loss": 1.198, + "step": 603 + }, + { + "epoch": 0.41777624070551617, + "grad_norm": 0.835200846195221, + "learning_rate": 6.139839766933721e-05, + "loss": 1.8262, + "step": 604 + }, + { + "epoch": 0.41846792322324056, + "grad_norm": 0.961685836315155, + "learning_rate": 6.132556445739257e-05, + "loss": 2.0815, + "step": 605 + }, + { + "epoch": 0.4191596057409649, + "grad_norm": 1.1949551105499268, + "learning_rate": 6.125273124544793e-05, + "loss": 2.0426, + "step": 606 + }, + { + "epoch": 0.4198512882586893, + "grad_norm": 0.6756623983383179, + "learning_rate": 6.117989803350328e-05, + "loss": 1.1775, + "step": 607 + }, + { + "epoch": 0.4205429707764136, + "grad_norm": 0.8606211543083191, + "learning_rate": 6.110706482155864e-05, + "loss": 1.1293, + "step": 608 + }, + { + "epoch": 0.421234653294138, + "grad_norm": 0.823352575302124, + "learning_rate": 6.103423160961399e-05, + "loss": 1.7001, + "step": 609 + }, + { + "epoch": 0.42192633581186234, + "grad_norm": 1.0490659475326538, + "learning_rate": 6.096139839766935e-05, + "loss": 1.6179, + "step": 610 + }, + { + "epoch": 0.42261801832958673, + "grad_norm": 0.972795307636261, + "learning_rate": 6.088856518572469e-05, + "loss": 2.1116, + "step": 611 + }, + { + "epoch": 0.42330970084731107, + "grad_norm": 0.795642077922821, + "learning_rate": 6.081573197378004e-05, + "loss": 1.7389, + "step": 612 + }, + { + "epoch": 0.42400138336503546, + "grad_norm": 0.796417772769928, + "learning_rate": 6.07428987618354e-05, + "loss": 1.2999, + "step": 613 + }, + { + "epoch": 0.4246930658827598, + "grad_norm": 0.8704690933227539, + "learning_rate": 6.067006554989075e-05, + "loss": 1.9508, + "step": 614 + }, + { + "epoch": 0.4253847484004842, + "grad_norm": 0.899750292301178, + "learning_rate": 6.0597232337946106e-05, + "loss": 1.9528, + "step": 615 + }, + { + "epoch": 0.4260764309182085, + "grad_norm": 0.7315880656242371, + "learning_rate": 6.052439912600146e-05, + "loss": 1.6963, + "step": 616 + }, + { + "epoch": 0.4267681134359329, + "grad_norm": 0.5647531151771545, + "learning_rate": 6.045156591405682e-05, + "loss": 1.1703, + "step": 617 + }, + { + "epoch": 0.4274597959536573, + "grad_norm": 0.9686012268066406, + "learning_rate": 6.037873270211216e-05, + "loss": 1.9199, + "step": 618 + }, + { + "epoch": 0.42815147847138163, + "grad_norm": 0.692129909992218, + "learning_rate": 6.0305899490167516e-05, + "loss": 0.9099, + "step": 619 + }, + { + "epoch": 0.428843160989106, + "grad_norm": 0.823519229888916, + "learning_rate": 6.023306627822287e-05, + "loss": 1.8828, + "step": 620 + }, + { + "epoch": 0.42953484350683035, + "grad_norm": 0.8990175127983093, + "learning_rate": 6.0160233066278225e-05, + "loss": 1.4318, + "step": 621 + }, + { + "epoch": 0.43022652602455475, + "grad_norm": 0.8949781656265259, + "learning_rate": 6.008739985433358e-05, + "loss": 2.0491, + "step": 622 + }, + { + "epoch": 0.4309182085422791, + "grad_norm": 1.0537045001983643, + "learning_rate": 6.001456664238893e-05, + "loss": 1.8443, + "step": 623 + }, + { + "epoch": 0.43160989106000347, + "grad_norm": 0.647614598274231, + "learning_rate": 5.994173343044428e-05, + "loss": 1.7638, + "step": 624 + }, + { + "epoch": 0.4323015735777278, + "grad_norm": 0.9621549844741821, + "learning_rate": 5.9868900218499635e-05, + "loss": 1.6114, + "step": 625 + }, + { + "epoch": 0.4329932560954522, + "grad_norm": 1.032599925994873, + "learning_rate": 5.979606700655499e-05, + "loss": 1.9123, + "step": 626 + }, + { + "epoch": 0.43368493861317653, + "grad_norm": 0.7238906621932983, + "learning_rate": 5.9723233794610344e-05, + "loss": 1.4896, + "step": 627 + }, + { + "epoch": 0.4343766211309009, + "grad_norm": 0.7447595596313477, + "learning_rate": 5.96504005826657e-05, + "loss": 1.5476, + "step": 628 + }, + { + "epoch": 0.43506830364862525, + "grad_norm": 0.6987977027893066, + "learning_rate": 5.957756737072105e-05, + "loss": 1.4778, + "step": 629 + }, + { + "epoch": 0.43575998616634964, + "grad_norm": 0.8449330925941467, + "learning_rate": 5.950473415877641e-05, + "loss": 1.8849, + "step": 630 + }, + { + "epoch": 0.43645166868407403, + "grad_norm": 1.0556013584136963, + "learning_rate": 5.9431900946831754e-05, + "loss": 2.2383, + "step": 631 + }, + { + "epoch": 0.43714335120179837, + "grad_norm": 0.6370812654495239, + "learning_rate": 5.935906773488711e-05, + "loss": 1.5145, + "step": 632 + }, + { + "epoch": 0.43783503371952276, + "grad_norm": 0.7922269701957703, + "learning_rate": 5.928623452294246e-05, + "loss": 1.5945, + "step": 633 + }, + { + "epoch": 0.4385267162372471, + "grad_norm": 0.9030670523643494, + "learning_rate": 5.921340131099782e-05, + "loss": 1.9727, + "step": 634 + }, + { + "epoch": 0.4392183987549715, + "grad_norm": 0.5330315828323364, + "learning_rate": 5.914056809905317e-05, + "loss": 1.1115, + "step": 635 + }, + { + "epoch": 0.4399100812726958, + "grad_norm": 0.7586291432380676, + "learning_rate": 5.9067734887108526e-05, + "loss": 1.3721, + "step": 636 + }, + { + "epoch": 0.4406017637904202, + "grad_norm": 0.8726906776428223, + "learning_rate": 5.899490167516388e-05, + "loss": 1.6429, + "step": 637 + }, + { + "epoch": 0.44129344630814454, + "grad_norm": 0.8924012184143066, + "learning_rate": 5.892206846321923e-05, + "loss": 1.7596, + "step": 638 + }, + { + "epoch": 0.44198512882586893, + "grad_norm": 0.9185534715652466, + "learning_rate": 5.884923525127458e-05, + "loss": 1.8224, + "step": 639 + }, + { + "epoch": 0.44267681134359327, + "grad_norm": 0.8137707114219666, + "learning_rate": 5.8776402039329936e-05, + "loss": 1.7631, + "step": 640 + }, + { + "epoch": 0.44336849386131766, + "grad_norm": 0.7134371399879456, + "learning_rate": 5.870356882738529e-05, + "loss": 1.8246, + "step": 641 + }, + { + "epoch": 0.444060176379042, + "grad_norm": 0.7280315160751343, + "learning_rate": 5.8630735615440645e-05, + "loss": 1.5007, + "step": 642 + }, + { + "epoch": 0.4447518588967664, + "grad_norm": 0.6889392733573914, + "learning_rate": 5.8557902403496e-05, + "loss": 1.2957, + "step": 643 + }, + { + "epoch": 0.4454435414144908, + "grad_norm": 0.6772429943084717, + "learning_rate": 5.8485069191551346e-05, + "loss": 1.2119, + "step": 644 + }, + { + "epoch": 0.4461352239322151, + "grad_norm": 0.7719807028770447, + "learning_rate": 5.84122359796067e-05, + "loss": 1.0095, + "step": 645 + }, + { + "epoch": 0.4468269064499395, + "grad_norm": 0.9226011037826538, + "learning_rate": 5.8339402767662055e-05, + "loss": 1.7821, + "step": 646 + }, + { + "epoch": 0.44751858896766383, + "grad_norm": 0.6220186948776245, + "learning_rate": 5.826656955571741e-05, + "loss": 1.1275, + "step": 647 + }, + { + "epoch": 0.4482102714853882, + "grad_norm": 0.6376112699508667, + "learning_rate": 5.8193736343772763e-05, + "loss": 1.4188, + "step": 648 + }, + { + "epoch": 0.44890195400311256, + "grad_norm": 0.7491716146469116, + "learning_rate": 5.812090313182812e-05, + "loss": 1.5822, + "step": 649 + }, + { + "epoch": 0.44959363652083695, + "grad_norm": 0.6484271287918091, + "learning_rate": 5.804806991988347e-05, + "loss": 1.6037, + "step": 650 + }, + { + "epoch": 0.4502853190385613, + "grad_norm": 1.5858505964279175, + "learning_rate": 5.797523670793882e-05, + "loss": 1.4725, + "step": 651 + }, + { + "epoch": 0.4509770015562857, + "grad_norm": 1.0042520761489868, + "learning_rate": 5.7902403495994174e-05, + "loss": 1.6445, + "step": 652 + }, + { + "epoch": 0.45166868407401, + "grad_norm": 1.0305542945861816, + "learning_rate": 5.782957028404953e-05, + "loss": 1.4438, + "step": 653 + }, + { + "epoch": 0.4523603665917344, + "grad_norm": 0.7213695049285889, + "learning_rate": 5.775673707210488e-05, + "loss": 1.6495, + "step": 654 + }, + { + "epoch": 0.45305204910945873, + "grad_norm": 0.8074948191642761, + "learning_rate": 5.768390386016024e-05, + "loss": 1.1349, + "step": 655 + }, + { + "epoch": 0.4537437316271831, + "grad_norm": 0.9579366445541382, + "learning_rate": 5.761107064821559e-05, + "loss": 1.421, + "step": 656 + }, + { + "epoch": 0.4544354141449075, + "grad_norm": 0.7831091284751892, + "learning_rate": 5.7538237436270945e-05, + "loss": 1.8771, + "step": 657 + }, + { + "epoch": 0.45512709666263185, + "grad_norm": 0.7758937478065491, + "learning_rate": 5.746540422432629e-05, + "loss": 1.7355, + "step": 658 + }, + { + "epoch": 0.45581877918035624, + "grad_norm": 0.8883340358734131, + "learning_rate": 5.739257101238165e-05, + "loss": 1.7985, + "step": 659 + }, + { + "epoch": 0.45651046169808057, + "grad_norm": 0.7148769497871399, + "learning_rate": 5.7319737800437e-05, + "loss": 2.0838, + "step": 660 + }, + { + "epoch": 0.45720214421580496, + "grad_norm": 0.7475634217262268, + "learning_rate": 5.7246904588492356e-05, + "loss": 1.1599, + "step": 661 + }, + { + "epoch": 0.4578938267335293, + "grad_norm": 0.9443761706352234, + "learning_rate": 5.717407137654771e-05, + "loss": 1.9477, + "step": 662 + }, + { + "epoch": 0.4585855092512537, + "grad_norm": 0.8133770823478699, + "learning_rate": 5.7101238164603064e-05, + "loss": 1.8349, + "step": 663 + }, + { + "epoch": 0.459277191768978, + "grad_norm": 0.7030238509178162, + "learning_rate": 5.702840495265841e-05, + "loss": 0.7075, + "step": 664 + }, + { + "epoch": 0.4599688742867024, + "grad_norm": 0.6615720987319946, + "learning_rate": 5.6955571740713766e-05, + "loss": 1.4996, + "step": 665 + }, + { + "epoch": 0.46066055680442675, + "grad_norm": 0.9358487725257874, + "learning_rate": 5.688273852876912e-05, + "loss": 1.4268, + "step": 666 + }, + { + "epoch": 0.46135223932215114, + "grad_norm": 0.6841732859611511, + "learning_rate": 5.6809905316824475e-05, + "loss": 1.7628, + "step": 667 + }, + { + "epoch": 0.46204392183987547, + "grad_norm": 0.7229800820350647, + "learning_rate": 5.673707210487983e-05, + "loss": 1.3031, + "step": 668 + }, + { + "epoch": 0.46273560435759986, + "grad_norm": 0.655465304851532, + "learning_rate": 5.666423889293518e-05, + "loss": 1.9593, + "step": 669 + }, + { + "epoch": 0.46342728687532425, + "grad_norm": 0.6893927454948425, + "learning_rate": 5.659140568099054e-05, + "loss": 0.9872, + "step": 670 + }, + { + "epoch": 0.4641189693930486, + "grad_norm": 1.0941834449768066, + "learning_rate": 5.6518572469045885e-05, + "loss": 1.4924, + "step": 671 + }, + { + "epoch": 0.464810651910773, + "grad_norm": 0.893545925617218, + "learning_rate": 5.644573925710124e-05, + "loss": 1.531, + "step": 672 + }, + { + "epoch": 0.4655023344284973, + "grad_norm": 0.7984673976898193, + "learning_rate": 5.6372906045156594e-05, + "loss": 1.79, + "step": 673 + }, + { + "epoch": 0.4661940169462217, + "grad_norm": 0.5964581370353699, + "learning_rate": 5.630007283321195e-05, + "loss": 1.1119, + "step": 674 + }, + { + "epoch": 0.46688569946394604, + "grad_norm": 0.7665301561355591, + "learning_rate": 5.62272396212673e-05, + "loss": 1.3566, + "step": 675 + }, + { + "epoch": 0.4675773819816704, + "grad_norm": 0.8093408346176147, + "learning_rate": 5.6154406409322656e-05, + "loss": 1.5565, + "step": 676 + }, + { + "epoch": 0.46826906449939476, + "grad_norm": 0.8169633150100708, + "learning_rate": 5.608157319737801e-05, + "loss": 2.2779, + "step": 677 + }, + { + "epoch": 0.46896074701711915, + "grad_norm": 0.8401588797569275, + "learning_rate": 5.600873998543336e-05, + "loss": 1.9076, + "step": 678 + }, + { + "epoch": 0.4696524295348435, + "grad_norm": 1.4105268716812134, + "learning_rate": 5.593590677348871e-05, + "loss": 1.4305, + "step": 679 + }, + { + "epoch": 0.4703441120525679, + "grad_norm": 1.0228116512298584, + "learning_rate": 5.586307356154407e-05, + "loss": 2.0295, + "step": 680 + }, + { + "epoch": 0.4710357945702922, + "grad_norm": 0.8564938306808472, + "learning_rate": 5.579024034959942e-05, + "loss": 1.6077, + "step": 681 + }, + { + "epoch": 0.4717274770880166, + "grad_norm": 0.8669772148132324, + "learning_rate": 5.5717407137654775e-05, + "loss": 1.584, + "step": 682 + }, + { + "epoch": 0.472419159605741, + "grad_norm": 0.5886315703392029, + "learning_rate": 5.564457392571013e-05, + "loss": 1.4101, + "step": 683 + }, + { + "epoch": 0.4731108421234653, + "grad_norm": 0.8294202089309692, + "learning_rate": 5.557174071376548e-05, + "loss": 1.8993, + "step": 684 + }, + { + "epoch": 0.4738025246411897, + "grad_norm": 1.0754139423370361, + "learning_rate": 5.549890750182083e-05, + "loss": 1.654, + "step": 685 + }, + { + "epoch": 0.47449420715891405, + "grad_norm": 0.7597669363021851, + "learning_rate": 5.5426074289876186e-05, + "loss": 2.0938, + "step": 686 + }, + { + "epoch": 0.47518588967663844, + "grad_norm": 0.7472283244132996, + "learning_rate": 5.535324107793154e-05, + "loss": 1.6874, + "step": 687 + }, + { + "epoch": 0.4758775721943628, + "grad_norm": 1.1373130083084106, + "learning_rate": 5.5280407865986894e-05, + "loss": 1.5681, + "step": 688 + }, + { + "epoch": 0.47656925471208716, + "grad_norm": 0.6962531805038452, + "learning_rate": 5.520757465404225e-05, + "loss": 1.2423, + "step": 689 + }, + { + "epoch": 0.4772609372298115, + "grad_norm": 0.8981777429580688, + "learning_rate": 5.51347414420976e-05, + "loss": 1.4854, + "step": 690 + }, + { + "epoch": 0.4779526197475359, + "grad_norm": 0.7226567268371582, + "learning_rate": 5.506190823015295e-05, + "loss": 2.2022, + "step": 691 + }, + { + "epoch": 0.4786443022652602, + "grad_norm": 0.9245135188102722, + "learning_rate": 5.4989075018208305e-05, + "loss": 1.3642, + "step": 692 + }, + { + "epoch": 0.4793359847829846, + "grad_norm": 1.0049363374710083, + "learning_rate": 5.491624180626366e-05, + "loss": 1.9998, + "step": 693 + }, + { + "epoch": 0.48002766730070895, + "grad_norm": 0.6878874897956848, + "learning_rate": 5.484340859431901e-05, + "loss": 1.9002, + "step": 694 + }, + { + "epoch": 0.48071934981843334, + "grad_norm": 1.1163276433944702, + "learning_rate": 5.477057538237437e-05, + "loss": 1.1385, + "step": 695 + }, + { + "epoch": 0.48141103233615773, + "grad_norm": 0.7806943655014038, + "learning_rate": 5.469774217042972e-05, + "loss": 1.0122, + "step": 696 + }, + { + "epoch": 0.48210271485388206, + "grad_norm": 0.9799820184707642, + "learning_rate": 5.4624908958485076e-05, + "loss": 2.0761, + "step": 697 + }, + { + "epoch": 0.48279439737160645, + "grad_norm": 0.5801637768745422, + "learning_rate": 5.4552075746540424e-05, + "loss": 0.8659, + "step": 698 + }, + { + "epoch": 0.4834860798893308, + "grad_norm": 0.7793147563934326, + "learning_rate": 5.447924253459578e-05, + "loss": 1.4955, + "step": 699 + }, + { + "epoch": 0.4841777624070552, + "grad_norm": 0.7253298759460449, + "learning_rate": 5.440640932265113e-05, + "loss": 0.8938, + "step": 700 + }, + { + "epoch": 0.4848694449247795, + "grad_norm": 0.8305007219314575, + "learning_rate": 5.4333576110706487e-05, + "loss": 1.3896, + "step": 701 + }, + { + "epoch": 0.4855611274425039, + "grad_norm": 0.8099757432937622, + "learning_rate": 5.426074289876184e-05, + "loss": 1.7134, + "step": 702 + }, + { + "epoch": 0.48625280996022824, + "grad_norm": 0.7706416845321655, + "learning_rate": 5.4187909686817195e-05, + "loss": 0.8483, + "step": 703 + }, + { + "epoch": 0.48694449247795263, + "grad_norm": 2.7785534858703613, + "learning_rate": 5.4115076474872536e-05, + "loss": 1.5148, + "step": 704 + }, + { + "epoch": 0.48763617499567696, + "grad_norm": 0.7698436975479126, + "learning_rate": 5.40422432629279e-05, + "loss": 2.031, + "step": 705 + }, + { + "epoch": 0.48832785751340135, + "grad_norm": 0.8159623742103577, + "learning_rate": 5.396941005098325e-05, + "loss": 2.0041, + "step": 706 + }, + { + "epoch": 0.4890195400311257, + "grad_norm": 0.6865976452827454, + "learning_rate": 5.3896576839038606e-05, + "loss": 1.2877, + "step": 707 + }, + { + "epoch": 0.4897112225488501, + "grad_norm": 0.9447488784790039, + "learning_rate": 5.382374362709396e-05, + "loss": 2.1087, + "step": 708 + }, + { + "epoch": 0.49040290506657447, + "grad_norm": 0.9315760731697083, + "learning_rate": 5.3750910415149314e-05, + "loss": 1.5459, + "step": 709 + }, + { + "epoch": 0.4910945875842988, + "grad_norm": 0.7628729939460754, + "learning_rate": 5.367807720320467e-05, + "loss": 1.6846, + "step": 710 + }, + { + "epoch": 0.4917862701020232, + "grad_norm": 0.7759138345718384, + "learning_rate": 5.360524399126001e-05, + "loss": 1.6179, + "step": 711 + }, + { + "epoch": 0.4924779526197475, + "grad_norm": 0.7866867184638977, + "learning_rate": 5.353241077931537e-05, + "loss": 2.2264, + "step": 712 + }, + { + "epoch": 0.4931696351374719, + "grad_norm": 0.7560875415802002, + "learning_rate": 5.3459577567370724e-05, + "loss": 1.5638, + "step": 713 + }, + { + "epoch": 0.49386131765519625, + "grad_norm": 0.6578519940376282, + "learning_rate": 5.338674435542608e-05, + "loss": 1.0194, + "step": 714 + }, + { + "epoch": 0.49455300017292064, + "grad_norm": 0.8581376671791077, + "learning_rate": 5.331391114348143e-05, + "loss": 1.7412, + "step": 715 + }, + { + "epoch": 0.495244682690645, + "grad_norm": 0.8838049173355103, + "learning_rate": 5.324107793153679e-05, + "loss": 2.0492, + "step": 716 + }, + { + "epoch": 0.49593636520836937, + "grad_norm": 0.7301944494247437, + "learning_rate": 5.316824471959214e-05, + "loss": 1.8718, + "step": 717 + }, + { + "epoch": 0.4966280477260937, + "grad_norm": 0.8501876592636108, + "learning_rate": 5.309541150764748e-05, + "loss": 1.9824, + "step": 718 + }, + { + "epoch": 0.4973197302438181, + "grad_norm": 0.8642913699150085, + "learning_rate": 5.3022578295702843e-05, + "loss": 2.0727, + "step": 719 + }, + { + "epoch": 0.4980114127615424, + "grad_norm": 0.7409586310386658, + "learning_rate": 5.29497450837582e-05, + "loss": 2.0216, + "step": 720 + }, + { + "epoch": 0.4987030952792668, + "grad_norm": 1.2024013996124268, + "learning_rate": 5.287691187181355e-05, + "loss": 1.534, + "step": 721 + }, + { + "epoch": 0.4993947777969912, + "grad_norm": 0.7344438433647156, + "learning_rate": 5.2804078659868906e-05, + "loss": 1.7692, + "step": 722 + }, + { + "epoch": 0.5000864603147156, + "grad_norm": 0.8432995080947876, + "learning_rate": 5.273124544792426e-05, + "loss": 1.1382, + "step": 723 + }, + { + "epoch": 0.5007781428324399, + "grad_norm": 0.8637827038764954, + "learning_rate": 5.26584122359796e-05, + "loss": 2.0577, + "step": 724 + }, + { + "epoch": 0.5014698253501643, + "grad_norm": 0.7211031913757324, + "learning_rate": 5.2585579024034956e-05, + "loss": 1.0286, + "step": 725 + }, + { + "epoch": 0.5021615078678886, + "grad_norm": 0.8980405330657959, + "learning_rate": 5.251274581209032e-05, + "loss": 1.6481, + "step": 726 + }, + { + "epoch": 0.502853190385613, + "grad_norm": 1.0838844776153564, + "learning_rate": 5.243991260014567e-05, + "loss": 1.8133, + "step": 727 + }, + { + "epoch": 0.5035448729033374, + "grad_norm": 0.7398082613945007, + "learning_rate": 5.2367079388201025e-05, + "loss": 1.9674, + "step": 728 + }, + { + "epoch": 0.5042365554210617, + "grad_norm": 0.9297565817832947, + "learning_rate": 5.229424617625638e-05, + "loss": 1.5748, + "step": 729 + }, + { + "epoch": 0.504928237938786, + "grad_norm": 1.0690850019454956, + "learning_rate": 5.2221412964311734e-05, + "loss": 1.5343, + "step": 730 + }, + { + "epoch": 0.5056199204565105, + "grad_norm": 0.8005755543708801, + "learning_rate": 5.2148579752367075e-05, + "loss": 1.7446, + "step": 731 + }, + { + "epoch": 0.5063116029742348, + "grad_norm": 0.9279308915138245, + "learning_rate": 5.207574654042243e-05, + "loss": 2.0124, + "step": 732 + }, + { + "epoch": 0.5070032854919592, + "grad_norm": 0.8205484747886658, + "learning_rate": 5.200291332847779e-05, + "loss": 1.9278, + "step": 733 + }, + { + "epoch": 0.5076949680096835, + "grad_norm": 1.0283253192901611, + "learning_rate": 5.1930080116533144e-05, + "loss": 1.6584, + "step": 734 + }, + { + "epoch": 0.508386650527408, + "grad_norm": 0.9164149165153503, + "learning_rate": 5.18572469045885e-05, + "loss": 1.5658, + "step": 735 + }, + { + "epoch": 0.5090783330451323, + "grad_norm": 0.8904675841331482, + "learning_rate": 5.178441369264385e-05, + "loss": 2.054, + "step": 736 + }, + { + "epoch": 0.5097700155628566, + "grad_norm": 0.731334924697876, + "learning_rate": 5.171158048069921e-05, + "loss": 1.8062, + "step": 737 + }, + { + "epoch": 0.5104616980805811, + "grad_norm": 1.436216115951538, + "learning_rate": 5.163874726875455e-05, + "loss": 2.4866, + "step": 738 + }, + { + "epoch": 0.5111533805983054, + "grad_norm": 0.7042586803436279, + "learning_rate": 5.15659140568099e-05, + "loss": 1.6228, + "step": 739 + }, + { + "epoch": 0.5118450631160297, + "grad_norm": 0.63124018907547, + "learning_rate": 5.149308084486526e-05, + "loss": 0.9537, + "step": 740 + }, + { + "epoch": 0.5125367456337541, + "grad_norm": 0.8653015494346619, + "learning_rate": 5.142024763292062e-05, + "loss": 2.0306, + "step": 741 + }, + { + "epoch": 0.5132284281514785, + "grad_norm": 0.8902332782745361, + "learning_rate": 5.134741442097597e-05, + "loss": 2.0709, + "step": 742 + }, + { + "epoch": 0.5139201106692028, + "grad_norm": 0.8607166409492493, + "learning_rate": 5.1274581209031326e-05, + "loss": 1.635, + "step": 743 + }, + { + "epoch": 0.5146117931869272, + "grad_norm": 0.8155879974365234, + "learning_rate": 5.120174799708667e-05, + "loss": 1.0902, + "step": 744 + }, + { + "epoch": 0.5153034757046515, + "grad_norm": 0.8006990551948547, + "learning_rate": 5.112891478514202e-05, + "loss": 1.2486, + "step": 745 + }, + { + "epoch": 0.515995158222376, + "grad_norm": 0.7772015929222107, + "learning_rate": 5.1056081573197375e-05, + "loss": 2.1048, + "step": 746 + }, + { + "epoch": 0.5166868407401003, + "grad_norm": 0.7376934289932251, + "learning_rate": 5.0983248361252736e-05, + "loss": 1.9748, + "step": 747 + }, + { + "epoch": 0.5173785232578246, + "grad_norm": 0.9872801899909973, + "learning_rate": 5.091041514930809e-05, + "loss": 1.3626, + "step": 748 + }, + { + "epoch": 0.5180702057755491, + "grad_norm": 0.6437541246414185, + "learning_rate": 5.0837581937363445e-05, + "loss": 1.8523, + "step": 749 + }, + { + "epoch": 0.5187618882932734, + "grad_norm": 0.8643835186958313, + "learning_rate": 5.07647487254188e-05, + "loss": 1.5464, + "step": 750 + }, + { + "epoch": 0.5194535708109977, + "grad_norm": 0.9346618056297302, + "learning_rate": 5.069191551347414e-05, + "loss": 1.7027, + "step": 751 + }, + { + "epoch": 0.5201452533287221, + "grad_norm": 0.8096154928207397, + "learning_rate": 5.0619082301529494e-05, + "loss": 1.4638, + "step": 752 + }, + { + "epoch": 0.5208369358464465, + "grad_norm": 0.8895930051803589, + "learning_rate": 5.054624908958485e-05, + "loss": 1.3399, + "step": 753 + }, + { + "epoch": 0.5215286183641709, + "grad_norm": 0.8496235013008118, + "learning_rate": 5.047341587764021e-05, + "loss": 0.9693, + "step": 754 + }, + { + "epoch": 0.5222203008818952, + "grad_norm": 0.7915313243865967, + "learning_rate": 5.0400582665695564e-05, + "loss": 1.4108, + "step": 755 + }, + { + "epoch": 0.5229119833996195, + "grad_norm": 0.8515884876251221, + "learning_rate": 5.032774945375092e-05, + "loss": 1.7595, + "step": 756 + }, + { + "epoch": 0.523603665917344, + "grad_norm": 0.7750046849250793, + "learning_rate": 5.025491624180627e-05, + "loss": 1.73, + "step": 757 + }, + { + "epoch": 0.5242953484350683, + "grad_norm": 0.8660963177680969, + "learning_rate": 5.018208302986161e-05, + "loss": 1.8401, + "step": 758 + }, + { + "epoch": 0.5249870309527926, + "grad_norm": 1.214325189590454, + "learning_rate": 5.010924981791697e-05, + "loss": 1.5471, + "step": 759 + }, + { + "epoch": 0.525678713470517, + "grad_norm": 1.0521632432937622, + "learning_rate": 5.003641660597232e-05, + "loss": 2.0549, + "step": 760 + }, + { + "epoch": 0.5263703959882414, + "grad_norm": 0.8383963108062744, + "learning_rate": 4.996358339402768e-05, + "loss": 1.9966, + "step": 761 + }, + { + "epoch": 0.5270620785059658, + "grad_norm": 0.9236989617347717, + "learning_rate": 4.989075018208304e-05, + "loss": 1.3734, + "step": 762 + }, + { + "epoch": 0.5277537610236901, + "grad_norm": 0.7525227665901184, + "learning_rate": 4.9817916970138385e-05, + "loss": 1.2582, + "step": 763 + }, + { + "epoch": 0.5284454435414145, + "grad_norm": 0.8061596751213074, + "learning_rate": 4.974508375819374e-05, + "loss": 2.0085, + "step": 764 + }, + { + "epoch": 0.5291371260591389, + "grad_norm": 1.0033149719238281, + "learning_rate": 4.967225054624909e-05, + "loss": 1.5858, + "step": 765 + }, + { + "epoch": 0.5298288085768632, + "grad_norm": 1.3156208992004395, + "learning_rate": 4.959941733430444e-05, + "loss": 1.8546, + "step": 766 + }, + { + "epoch": 0.5305204910945875, + "grad_norm": 0.7604823112487793, + "learning_rate": 4.9526584122359795e-05, + "loss": 1.7595, + "step": 767 + }, + { + "epoch": 0.531212173612312, + "grad_norm": 0.9594501852989197, + "learning_rate": 4.9453750910415156e-05, + "loss": 1.507, + "step": 768 + }, + { + "epoch": 0.5319038561300363, + "grad_norm": 1.176438808441162, + "learning_rate": 4.9380917698470504e-05, + "loss": 1.6748, + "step": 769 + }, + { + "epoch": 0.5325955386477607, + "grad_norm": 0.6893585324287415, + "learning_rate": 4.930808448652586e-05, + "loss": 1.326, + "step": 770 + }, + { + "epoch": 0.533287221165485, + "grad_norm": 0.8642687797546387, + "learning_rate": 4.923525127458121e-05, + "loss": 1.1461, + "step": 771 + }, + { + "epoch": 0.5339789036832094, + "grad_norm": 0.6313591003417969, + "learning_rate": 4.9162418062636567e-05, + "loss": 1.5494, + "step": 772 + }, + { + "epoch": 0.5346705862009338, + "grad_norm": 0.7172280550003052, + "learning_rate": 4.9089584850691914e-05, + "loss": 1.4831, + "step": 773 + }, + { + "epoch": 0.5353622687186581, + "grad_norm": 0.8146588206291199, + "learning_rate": 4.901675163874727e-05, + "loss": 1.913, + "step": 774 + }, + { + "epoch": 0.5360539512363826, + "grad_norm": 0.9037981033325195, + "learning_rate": 4.894391842680263e-05, + "loss": 1.7836, + "step": 775 + }, + { + "epoch": 0.5367456337541069, + "grad_norm": 0.7782710194587708, + "learning_rate": 4.887108521485798e-05, + "loss": 1.4689, + "step": 776 + }, + { + "epoch": 0.5374373162718312, + "grad_norm": 1.111527681350708, + "learning_rate": 4.879825200291333e-05, + "loss": 1.6938, + "step": 777 + }, + { + "epoch": 0.5381289987895556, + "grad_norm": 1.2994364500045776, + "learning_rate": 4.8725418790968686e-05, + "loss": 1.529, + "step": 778 + }, + { + "epoch": 0.53882068130728, + "grad_norm": 0.7461956143379211, + "learning_rate": 4.865258557902403e-05, + "loss": 1.1475, + "step": 779 + }, + { + "epoch": 0.5395123638250043, + "grad_norm": 0.6900501847267151, + "learning_rate": 4.857975236707939e-05, + "loss": 1.8377, + "step": 780 + }, + { + "epoch": 0.5402040463427287, + "grad_norm": 0.7007532715797424, + "learning_rate": 4.850691915513474e-05, + "loss": 1.4117, + "step": 781 + }, + { + "epoch": 0.540895728860453, + "grad_norm": 0.8662421703338623, + "learning_rate": 4.84340859431901e-05, + "loss": 1.7831, + "step": 782 + }, + { + "epoch": 0.5415874113781775, + "grad_norm": 1.2154408693313599, + "learning_rate": 4.836125273124545e-05, + "loss": 1.9868, + "step": 783 + }, + { + "epoch": 0.5422790938959018, + "grad_norm": 0.822338342666626, + "learning_rate": 4.8288419519300804e-05, + "loss": 1.8789, + "step": 784 + }, + { + "epoch": 0.5429707764136261, + "grad_norm": 0.9804028868675232, + "learning_rate": 4.821558630735616e-05, + "loss": 2.2036, + "step": 785 + }, + { + "epoch": 0.5436624589313505, + "grad_norm": 0.9605854749679565, + "learning_rate": 4.8142753095411506e-05, + "loss": 2.0455, + "step": 786 + }, + { + "epoch": 0.5443541414490749, + "grad_norm": 0.7948291301727295, + "learning_rate": 4.806991988346686e-05, + "loss": 1.529, + "step": 787 + }, + { + "epoch": 0.5450458239667992, + "grad_norm": 0.7471349239349365, + "learning_rate": 4.7997086671522215e-05, + "loss": 1.5366, + "step": 788 + }, + { + "epoch": 0.5457375064845236, + "grad_norm": 1.0526931285858154, + "learning_rate": 4.792425345957757e-05, + "loss": 1.2396, + "step": 789 + }, + { + "epoch": 0.546429189002248, + "grad_norm": 1.0676295757293701, + "learning_rate": 4.7851420247632923e-05, + "loss": 1.6639, + "step": 790 + }, + { + "epoch": 0.5471208715199724, + "grad_norm": 0.9911693334579468, + "learning_rate": 4.777858703568828e-05, + "loss": 1.8713, + "step": 791 + }, + { + "epoch": 0.5478125540376967, + "grad_norm": 0.5897760391235352, + "learning_rate": 4.770575382374363e-05, + "loss": 1.4593, + "step": 792 + }, + { + "epoch": 0.548504236555421, + "grad_norm": 0.836234986782074, + "learning_rate": 4.763292061179898e-05, + "loss": 1.3373, + "step": 793 + }, + { + "epoch": 0.5491959190731455, + "grad_norm": 1.0234719514846802, + "learning_rate": 4.7560087399854334e-05, + "loss": 1.3758, + "step": 794 + }, + { + "epoch": 0.5498876015908698, + "grad_norm": 0.6555935144424438, + "learning_rate": 4.748725418790969e-05, + "loss": 1.0075, + "step": 795 + }, + { + "epoch": 0.5505792841085941, + "grad_norm": 0.7756921052932739, + "learning_rate": 4.741442097596504e-05, + "loss": 1.5594, + "step": 796 + }, + { + "epoch": 0.5512709666263185, + "grad_norm": 0.8114323616027832, + "learning_rate": 4.73415877640204e-05, + "loss": 1.9788, + "step": 797 + }, + { + "epoch": 0.5519626491440429, + "grad_norm": 1.0409117937088013, + "learning_rate": 4.726875455207575e-05, + "loss": 1.4971, + "step": 798 + }, + { + "epoch": 0.5526543316617673, + "grad_norm": 0.963651716709137, + "learning_rate": 4.71959213401311e-05, + "loss": 1.8012, + "step": 799 + }, + { + "epoch": 0.5533460141794916, + "grad_norm": 0.7546253204345703, + "learning_rate": 4.712308812818645e-05, + "loss": 2.0704, + "step": 800 + }, + { + "epoch": 0.554037696697216, + "grad_norm": 0.7837393283843994, + "learning_rate": 4.705025491624181e-05, + "loss": 1.9567, + "step": 801 + }, + { + "epoch": 0.5547293792149404, + "grad_norm": 0.9750471711158752, + "learning_rate": 4.697742170429716e-05, + "loss": 1.8202, + "step": 802 + }, + { + "epoch": 0.5554210617326647, + "grad_norm": 0.656570315361023, + "learning_rate": 4.6904588492352516e-05, + "loss": 1.9521, + "step": 803 + }, + { + "epoch": 0.556112744250389, + "grad_norm": 0.7636544108390808, + "learning_rate": 4.683175528040787e-05, + "loss": 1.6885, + "step": 804 + }, + { + "epoch": 0.5568044267681135, + "grad_norm": 0.6734775900840759, + "learning_rate": 4.6758922068463224e-05, + "loss": 1.4624, + "step": 805 + }, + { + "epoch": 0.5574961092858378, + "grad_norm": 0.7191459536552429, + "learning_rate": 4.668608885651857e-05, + "loss": 1.9275, + "step": 806 + }, + { + "epoch": 0.5581877918035621, + "grad_norm": 0.7927097082138062, + "learning_rate": 4.6613255644573926e-05, + "loss": 1.3038, + "step": 807 + }, + { + "epoch": 0.5588794743212865, + "grad_norm": 0.7056376934051514, + "learning_rate": 4.654042243262928e-05, + "loss": 1.4776, + "step": 808 + }, + { + "epoch": 0.5595711568390109, + "grad_norm": 0.6944011449813843, + "learning_rate": 4.6467589220684635e-05, + "loss": 1.6238, + "step": 809 + }, + { + "epoch": 0.5602628393567353, + "grad_norm": 0.6770850419998169, + "learning_rate": 4.639475600873999e-05, + "loss": 2.0007, + "step": 810 + }, + { + "epoch": 0.5609545218744596, + "grad_norm": 0.6012453436851501, + "learning_rate": 4.632192279679534e-05, + "loss": 1.3431, + "step": 811 + }, + { + "epoch": 0.5616462043921839, + "grad_norm": 0.8556742072105408, + "learning_rate": 4.62490895848507e-05, + "loss": 1.9744, + "step": 812 + }, + { + "epoch": 0.5623378869099084, + "grad_norm": 0.9271799921989441, + "learning_rate": 4.6176256372906045e-05, + "loss": 2.0651, + "step": 813 + }, + { + "epoch": 0.5630295694276327, + "grad_norm": 0.7125703692436218, + "learning_rate": 4.61034231609614e-05, + "loss": 1.519, + "step": 814 + }, + { + "epoch": 0.563721251945357, + "grad_norm": 0.721099853515625, + "learning_rate": 4.6030589949016754e-05, + "loss": 1.4397, + "step": 815 + }, + { + "epoch": 0.5644129344630815, + "grad_norm": 0.8622443675994873, + "learning_rate": 4.595775673707211e-05, + "loss": 1.5684, + "step": 816 + }, + { + "epoch": 0.5651046169808058, + "grad_norm": 0.8319584727287292, + "learning_rate": 4.588492352512746e-05, + "loss": 1.8106, + "step": 817 + }, + { + "epoch": 0.5657962994985302, + "grad_norm": 0.7659180760383606, + "learning_rate": 4.5812090313182816e-05, + "loss": 1.976, + "step": 818 + }, + { + "epoch": 0.5664879820162545, + "grad_norm": 0.658638060092926, + "learning_rate": 4.5739257101238164e-05, + "loss": 1.6156, + "step": 819 + }, + { + "epoch": 0.5671796645339789, + "grad_norm": 0.7831644415855408, + "learning_rate": 4.566642388929352e-05, + "loss": 1.7169, + "step": 820 + }, + { + "epoch": 0.5678713470517033, + "grad_norm": 0.8587795495986938, + "learning_rate": 4.559359067734887e-05, + "loss": 1.8784, + "step": 821 + }, + { + "epoch": 0.5685630295694276, + "grad_norm": 0.6490662693977356, + "learning_rate": 4.552075746540423e-05, + "loss": 1.2044, + "step": 822 + }, + { + "epoch": 0.569254712087152, + "grad_norm": 0.8098453879356384, + "learning_rate": 4.544792425345958e-05, + "loss": 2.0041, + "step": 823 + }, + { + "epoch": 0.5699463946048764, + "grad_norm": 0.9287192225456238, + "learning_rate": 4.5375091041514935e-05, + "loss": 2.0794, + "step": 824 + }, + { + "epoch": 0.5706380771226007, + "grad_norm": 0.6558881998062134, + "learning_rate": 4.530225782957029e-05, + "loss": 1.1153, + "step": 825 + }, + { + "epoch": 0.5713297596403251, + "grad_norm": 0.9448340535163879, + "learning_rate": 4.522942461762564e-05, + "loss": 1.8924, + "step": 826 + }, + { + "epoch": 0.5720214421580495, + "grad_norm": 0.9030715823173523, + "learning_rate": 4.515659140568099e-05, + "loss": 1.487, + "step": 827 + }, + { + "epoch": 0.5727131246757738, + "grad_norm": 0.8691999316215515, + "learning_rate": 4.5083758193736346e-05, + "loss": 1.709, + "step": 828 + }, + { + "epoch": 0.5734048071934982, + "grad_norm": 0.802944540977478, + "learning_rate": 4.50109249817917e-05, + "loss": 1.5163, + "step": 829 + }, + { + "epoch": 0.5740964897112225, + "grad_norm": 0.7917433381080627, + "learning_rate": 4.4938091769847054e-05, + "loss": 1.5238, + "step": 830 + }, + { + "epoch": 0.574788172228947, + "grad_norm": 0.8852903842926025, + "learning_rate": 4.486525855790241e-05, + "loss": 1.4459, + "step": 831 + }, + { + "epoch": 0.5754798547466713, + "grad_norm": 0.8966250419616699, + "learning_rate": 4.479242534595776e-05, + "loss": 1.4654, + "step": 832 + }, + { + "epoch": 0.5761715372643956, + "grad_norm": 1.3262107372283936, + "learning_rate": 4.471959213401311e-05, + "loss": 1.5453, + "step": 833 + }, + { + "epoch": 0.57686321978212, + "grad_norm": 0.7559188604354858, + "learning_rate": 4.4646758922068465e-05, + "loss": 1.5227, + "step": 834 + }, + { + "epoch": 0.5775549022998444, + "grad_norm": 0.9680230617523193, + "learning_rate": 4.457392571012382e-05, + "loss": 1.8038, + "step": 835 + }, + { + "epoch": 0.5782465848175687, + "grad_norm": 0.8509007096290588, + "learning_rate": 4.450109249817917e-05, + "loss": 1.8162, + "step": 836 + }, + { + "epoch": 0.5789382673352931, + "grad_norm": 1.4922651052474976, + "learning_rate": 4.442825928623453e-05, + "loss": 1.5922, + "step": 837 + }, + { + "epoch": 0.5796299498530174, + "grad_norm": 1.0057311058044434, + "learning_rate": 4.435542607428988e-05, + "loss": 1.6555, + "step": 838 + }, + { + "epoch": 0.5803216323707419, + "grad_norm": 0.7542306184768677, + "learning_rate": 4.428259286234523e-05, + "loss": 1.7829, + "step": 839 + }, + { + "epoch": 0.5810133148884662, + "grad_norm": 0.7741349935531616, + "learning_rate": 4.4209759650400584e-05, + "loss": 1.6429, + "step": 840 + }, + { + "epoch": 0.5817049974061905, + "grad_norm": 0.6510467529296875, + "learning_rate": 4.413692643845594e-05, + "loss": 1.6901, + "step": 841 + }, + { + "epoch": 0.582396679923915, + "grad_norm": 0.5973620414733887, + "learning_rate": 4.406409322651129e-05, + "loss": 1.8884, + "step": 842 + }, + { + "epoch": 0.5830883624416393, + "grad_norm": 0.7993059158325195, + "learning_rate": 4.3991260014566647e-05, + "loss": 1.9353, + "step": 843 + }, + { + "epoch": 0.5837800449593636, + "grad_norm": 0.9664002656936646, + "learning_rate": 4.3918426802622e-05, + "loss": 1.5514, + "step": 844 + }, + { + "epoch": 0.584471727477088, + "grad_norm": 1.7549793720245361, + "learning_rate": 4.3845593590677355e-05, + "loss": 1.3181, + "step": 845 + }, + { + "epoch": 0.5851634099948124, + "grad_norm": 0.681312084197998, + "learning_rate": 4.37727603787327e-05, + "loss": 1.3649, + "step": 846 + }, + { + "epoch": 0.5858550925125368, + "grad_norm": 1.23701012134552, + "learning_rate": 4.369992716678806e-05, + "loss": 1.7591, + "step": 847 + }, + { + "epoch": 0.5865467750302611, + "grad_norm": 0.6490361094474792, + "learning_rate": 4.362709395484341e-05, + "loss": 0.9716, + "step": 848 + }, + { + "epoch": 0.5872384575479854, + "grad_norm": 1.2864362001419067, + "learning_rate": 4.355426074289876e-05, + "loss": 1.2925, + "step": 849 + }, + { + "epoch": 0.5879301400657099, + "grad_norm": 0.8159788250923157, + "learning_rate": 4.348142753095411e-05, + "loss": 1.6182, + "step": 850 + }, + { + "epoch": 0.5886218225834342, + "grad_norm": 0.9866076111793518, + "learning_rate": 4.3408594319009474e-05, + "loss": 1.0921, + "step": 851 + }, + { + "epoch": 0.5893135051011585, + "grad_norm": 0.7778481841087341, + "learning_rate": 4.333576110706483e-05, + "loss": 1.8508, + "step": 852 + }, + { + "epoch": 0.590005187618883, + "grad_norm": 0.7202186584472656, + "learning_rate": 4.3262927895120176e-05, + "loss": 2.0207, + "step": 853 + }, + { + "epoch": 0.5906968701366073, + "grad_norm": 0.764106035232544, + "learning_rate": 4.319009468317553e-05, + "loss": 1.6755, + "step": 854 + }, + { + "epoch": 0.5913885526543317, + "grad_norm": 0.993752121925354, + "learning_rate": 4.3117261471230884e-05, + "loss": 2.061, + "step": 855 + }, + { + "epoch": 0.592080235172056, + "grad_norm": 0.9889275431632996, + "learning_rate": 4.304442825928623e-05, + "loss": 2.0183, + "step": 856 + }, + { + "epoch": 0.5927719176897804, + "grad_norm": 0.7371869683265686, + "learning_rate": 4.2971595047341586e-05, + "loss": 1.8013, + "step": 857 + }, + { + "epoch": 0.5934636002075048, + "grad_norm": 0.8173306584358215, + "learning_rate": 4.289876183539695e-05, + "loss": 1.8409, + "step": 858 + }, + { + "epoch": 0.5941552827252291, + "grad_norm": 0.8699966669082642, + "learning_rate": 4.2825928623452295e-05, + "loss": 1.1185, + "step": 859 + }, + { + "epoch": 0.5948469652429534, + "grad_norm": 0.8679646253585815, + "learning_rate": 4.275309541150765e-05, + "loss": 1.7849, + "step": 860 + }, + { + "epoch": 0.5955386477606779, + "grad_norm": 0.7400544881820679, + "learning_rate": 4.2680262199563003e-05, + "loss": 1.4845, + "step": 861 + }, + { + "epoch": 0.5962303302784022, + "grad_norm": 0.9016357064247131, + "learning_rate": 4.260742898761836e-05, + "loss": 1.6556, + "step": 862 + }, + { + "epoch": 0.5969220127961266, + "grad_norm": 1.0750559568405151, + "learning_rate": 4.2534595775673705e-05, + "loss": 1.2046, + "step": 863 + }, + { + "epoch": 0.5976136953138509, + "grad_norm": 0.9797933101654053, + "learning_rate": 4.246176256372906e-05, + "loss": 1.9871, + "step": 864 + }, + { + "epoch": 0.5983053778315753, + "grad_norm": 0.9287171959877014, + "learning_rate": 4.238892935178442e-05, + "loss": 2.0108, + "step": 865 + }, + { + "epoch": 0.5989970603492997, + "grad_norm": 1.2748676538467407, + "learning_rate": 4.231609613983977e-05, + "loss": 1.8482, + "step": 866 + }, + { + "epoch": 0.599688742867024, + "grad_norm": 0.7953082919120789, + "learning_rate": 4.224326292789512e-05, + "loss": 1.9459, + "step": 867 + }, + { + "epoch": 0.6003804253847485, + "grad_norm": 0.6961773037910461, + "learning_rate": 4.217042971595048e-05, + "loss": 1.4124, + "step": 868 + }, + { + "epoch": 0.6010721079024728, + "grad_norm": 0.7349817156791687, + "learning_rate": 4.2097596504005824e-05, + "loss": 1.3384, + "step": 869 + }, + { + "epoch": 0.6017637904201971, + "grad_norm": 0.8988999724388123, + "learning_rate": 4.202476329206118e-05, + "loss": 2.1146, + "step": 870 + }, + { + "epoch": 0.6024554729379215, + "grad_norm": 0.7482579946517944, + "learning_rate": 4.195193008011653e-05, + "loss": 1.9195, + "step": 871 + }, + { + "epoch": 0.6031471554556459, + "grad_norm": 0.584476888179779, + "learning_rate": 4.1879096868171894e-05, + "loss": 1.1254, + "step": 872 + }, + { + "epoch": 0.6038388379733702, + "grad_norm": 0.9063954949378967, + "learning_rate": 4.180626365622724e-05, + "loss": 1.655, + "step": 873 + }, + { + "epoch": 0.6045305204910946, + "grad_norm": 1.784583568572998, + "learning_rate": 4.1733430444282596e-05, + "loss": 1.6756, + "step": 874 + }, + { + "epoch": 0.6052222030088189, + "grad_norm": 1.1327074766159058, + "learning_rate": 4.166059723233795e-05, + "loss": 1.6496, + "step": 875 + }, + { + "epoch": 0.6059138855265433, + "grad_norm": 0.6708528399467468, + "learning_rate": 4.15877640203933e-05, + "loss": 1.4513, + "step": 876 + }, + { + "epoch": 0.6066055680442677, + "grad_norm": 0.8013389110565186, + "learning_rate": 4.151493080844865e-05, + "loss": 1.2203, + "step": 877 + }, + { + "epoch": 0.607297250561992, + "grad_norm": 0.9567274451255798, + "learning_rate": 4.1442097596504006e-05, + "loss": 1.6625, + "step": 878 + }, + { + "epoch": 0.6079889330797165, + "grad_norm": 0.7005029916763306, + "learning_rate": 4.136926438455936e-05, + "loss": 1.7659, + "step": 879 + }, + { + "epoch": 0.6086806155974408, + "grad_norm": 0.7626811861991882, + "learning_rate": 4.1296431172614715e-05, + "loss": 1.9302, + "step": 880 + }, + { + "epoch": 0.6093722981151651, + "grad_norm": 0.8964532017707825, + "learning_rate": 4.122359796067007e-05, + "loss": 1.5873, + "step": 881 + }, + { + "epoch": 0.6100639806328895, + "grad_norm": 1.0764849185943604, + "learning_rate": 4.115076474872542e-05, + "loss": 1.7932, + "step": 882 + }, + { + "epoch": 0.6107556631506139, + "grad_norm": 0.6796227097511292, + "learning_rate": 4.107793153678077e-05, + "loss": 1.4911, + "step": 883 + }, + { + "epoch": 0.6114473456683382, + "grad_norm": 0.8213468790054321, + "learning_rate": 4.1005098324836125e-05, + "loss": 1.101, + "step": 884 + }, + { + "epoch": 0.6121390281860626, + "grad_norm": 0.7105699777603149, + "learning_rate": 4.093226511289148e-05, + "loss": 1.583, + "step": 885 + }, + { + "epoch": 0.6128307107037869, + "grad_norm": 0.8570541143417358, + "learning_rate": 4.0859431900946834e-05, + "loss": 1.4963, + "step": 886 + }, + { + "epoch": 0.6135223932215114, + "grad_norm": 0.9847023487091064, + "learning_rate": 4.078659868900219e-05, + "loss": 1.7706, + "step": 887 + }, + { + "epoch": 0.6142140757392357, + "grad_norm": 0.7415423393249512, + "learning_rate": 4.071376547705754e-05, + "loss": 1.7496, + "step": 888 + }, + { + "epoch": 0.61490575825696, + "grad_norm": 0.8785969614982605, + "learning_rate": 4.064093226511289e-05, + "loss": 1.3073, + "step": 889 + }, + { + "epoch": 0.6155974407746844, + "grad_norm": 0.8583815097808838, + "learning_rate": 4.0568099053168244e-05, + "loss": 1.5086, + "step": 890 + }, + { + "epoch": 0.6162891232924088, + "grad_norm": 4.068897724151611, + "learning_rate": 4.04952658412236e-05, + "loss": 1.9179, + "step": 891 + }, + { + "epoch": 0.6169808058101331, + "grad_norm": 0.781101644039154, + "learning_rate": 4.042243262927895e-05, + "loss": 1.7548, + "step": 892 + }, + { + "epoch": 0.6176724883278575, + "grad_norm": 0.7089797854423523, + "learning_rate": 4.034959941733431e-05, + "loss": 0.8754, + "step": 893 + }, + { + "epoch": 0.6183641708455819, + "grad_norm": 0.8193789124488831, + "learning_rate": 4.027676620538966e-05, + "loss": 1.8492, + "step": 894 + }, + { + "epoch": 0.6190558533633063, + "grad_norm": 1.1896588802337646, + "learning_rate": 4.0203932993445015e-05, + "loss": 1.3831, + "step": 895 + }, + { + "epoch": 0.6197475358810306, + "grad_norm": 0.7735778093338013, + "learning_rate": 4.013109978150036e-05, + "loss": 1.3693, + "step": 896 + }, + { + "epoch": 0.6204392183987549, + "grad_norm": 0.8395145535469055, + "learning_rate": 4.005826656955572e-05, + "loss": 2.0921, + "step": 897 + }, + { + "epoch": 0.6211309009164794, + "grad_norm": 0.8883162140846252, + "learning_rate": 3.998543335761107e-05, + "loss": 1.1215, + "step": 898 + }, + { + "epoch": 0.6218225834342037, + "grad_norm": 0.7758648991584778, + "learning_rate": 3.9912600145666426e-05, + "loss": 2.1886, + "step": 899 + }, + { + "epoch": 0.622514265951928, + "grad_norm": 1.0782335996627808, + "learning_rate": 3.983976693372178e-05, + "loss": 1.6127, + "step": 900 + }, + { + "epoch": 0.6232059484696524, + "grad_norm": 0.7295915484428406, + "learning_rate": 3.9766933721777134e-05, + "loss": 1.4735, + "step": 901 + }, + { + "epoch": 0.6238976309873768, + "grad_norm": 0.8186253309249878, + "learning_rate": 3.969410050983249e-05, + "loss": 1.9706, + "step": 902 + }, + { + "epoch": 0.6245893135051012, + "grad_norm": 0.9425630569458008, + "learning_rate": 3.9621267297887836e-05, + "loss": 1.7289, + "step": 903 + }, + { + "epoch": 0.6252809960228255, + "grad_norm": 1.7592432498931885, + "learning_rate": 3.954843408594319e-05, + "loss": 1.7906, + "step": 904 + }, + { + "epoch": 0.6259726785405499, + "grad_norm": 0.8689557909965515, + "learning_rate": 3.9475600873998545e-05, + "loss": 1.7125, + "step": 905 + }, + { + "epoch": 0.6266643610582743, + "grad_norm": 0.7336760759353638, + "learning_rate": 3.94027676620539e-05, + "loss": 1.2892, + "step": 906 + }, + { + "epoch": 0.6273560435759986, + "grad_norm": 1.0234981775283813, + "learning_rate": 3.932993445010925e-05, + "loss": 1.4871, + "step": 907 + }, + { + "epoch": 0.628047726093723, + "grad_norm": 0.62737637758255, + "learning_rate": 3.925710123816461e-05, + "loss": 1.4537, + "step": 908 + }, + { + "epoch": 0.6287394086114474, + "grad_norm": 0.8133891224861145, + "learning_rate": 3.9184268026219955e-05, + "loss": 1.9147, + "step": 909 + }, + { + "epoch": 0.6294310911291717, + "grad_norm": 0.8863133192062378, + "learning_rate": 3.911143481427531e-05, + "loss": 1.4371, + "step": 910 + }, + { + "epoch": 0.6301227736468961, + "grad_norm": 0.8457992076873779, + "learning_rate": 3.9038601602330664e-05, + "loss": 1.4767, + "step": 911 + }, + { + "epoch": 0.6308144561646204, + "grad_norm": 0.9177572727203369, + "learning_rate": 3.896576839038602e-05, + "loss": 0.9703, + "step": 912 + }, + { + "epoch": 0.6315061386823448, + "grad_norm": 0.798324465751648, + "learning_rate": 3.889293517844137e-05, + "loss": 1.21, + "step": 913 + }, + { + "epoch": 0.6321978212000692, + "grad_norm": 0.8874804973602295, + "learning_rate": 3.8820101966496726e-05, + "loss": 1.5766, + "step": 914 + }, + { + "epoch": 0.6328895037177935, + "grad_norm": 0.7579106092453003, + "learning_rate": 3.874726875455208e-05, + "loss": 2.0235, + "step": 915 + }, + { + "epoch": 0.6335811862355178, + "grad_norm": 0.8905123472213745, + "learning_rate": 3.867443554260743e-05, + "loss": 1.4235, + "step": 916 + }, + { + "epoch": 0.6342728687532423, + "grad_norm": 0.7457150816917419, + "learning_rate": 3.860160233066278e-05, + "loss": 1.005, + "step": 917 + }, + { + "epoch": 0.6349645512709666, + "grad_norm": 0.7743381261825562, + "learning_rate": 3.852876911871814e-05, + "loss": 1.5984, + "step": 918 + }, + { + "epoch": 0.635656233788691, + "grad_norm": 0.9183463454246521, + "learning_rate": 3.845593590677349e-05, + "loss": 1.9001, + "step": 919 + }, + { + "epoch": 0.6363479163064154, + "grad_norm": 0.9490715861320496, + "learning_rate": 3.8383102694828845e-05, + "loss": 1.6814, + "step": 920 + }, + { + "epoch": 0.6370395988241397, + "grad_norm": 0.7596708536148071, + "learning_rate": 3.83102694828842e-05, + "loss": 1.3015, + "step": 921 + }, + { + "epoch": 0.6377312813418641, + "grad_norm": 0.8084985017776489, + "learning_rate": 3.8237436270939554e-05, + "loss": 1.4776, + "step": 922 + }, + { + "epoch": 0.6384229638595884, + "grad_norm": 1.5200563669204712, + "learning_rate": 3.81646030589949e-05, + "loss": 1.5271, + "step": 923 + }, + { + "epoch": 0.6391146463773129, + "grad_norm": 0.9168257117271423, + "learning_rate": 3.8091769847050256e-05, + "loss": 0.865, + "step": 924 + }, + { + "epoch": 0.6398063288950372, + "grad_norm": 0.8897766470909119, + "learning_rate": 3.801893663510561e-05, + "loss": 1.7975, + "step": 925 + }, + { + "epoch": 0.6404980114127615, + "grad_norm": 0.8529428839683533, + "learning_rate": 3.7946103423160964e-05, + "loss": 1.5159, + "step": 926 + }, + { + "epoch": 0.6411896939304859, + "grad_norm": 0.9676948189735413, + "learning_rate": 3.787327021121632e-05, + "loss": 0.9847, + "step": 927 + }, + { + "epoch": 0.6418813764482103, + "grad_norm": 0.8767004013061523, + "learning_rate": 3.780043699927167e-05, + "loss": 2.0668, + "step": 928 + }, + { + "epoch": 0.6425730589659346, + "grad_norm": 0.7345457673072815, + "learning_rate": 3.772760378732702e-05, + "loss": 1.8291, + "step": 929 + }, + { + "epoch": 0.643264741483659, + "grad_norm": 1.1258862018585205, + "learning_rate": 3.7654770575382375e-05, + "loss": 1.9181, + "step": 930 + }, + { + "epoch": 0.6439564240013833, + "grad_norm": 1.059696078300476, + "learning_rate": 3.758193736343773e-05, + "loss": 1.1648, + "step": 931 + }, + { + "epoch": 0.6446481065191078, + "grad_norm": 0.875996470451355, + "learning_rate": 3.750910415149308e-05, + "loss": 1.8461, + "step": 932 + }, + { + "epoch": 0.6453397890368321, + "grad_norm": 0.7454437017440796, + "learning_rate": 3.743627093954844e-05, + "loss": 1.8346, + "step": 933 + }, + { + "epoch": 0.6460314715545564, + "grad_norm": 1.0619977712631226, + "learning_rate": 3.736343772760379e-05, + "loss": 1.1181, + "step": 934 + }, + { + "epoch": 0.6467231540722809, + "grad_norm": 0.7725276350975037, + "learning_rate": 3.7290604515659146e-05, + "loss": 1.3964, + "step": 935 + }, + { + "epoch": 0.6474148365900052, + "grad_norm": 0.9066617488861084, + "learning_rate": 3.7217771303714494e-05, + "loss": 1.8135, + "step": 936 + }, + { + "epoch": 0.6481065191077295, + "grad_norm": 0.7097616195678711, + "learning_rate": 3.714493809176985e-05, + "loss": 2.001, + "step": 937 + }, + { + "epoch": 0.6487982016254539, + "grad_norm": 0.9724728465080261, + "learning_rate": 3.70721048798252e-05, + "loss": 2.0217, + "step": 938 + }, + { + "epoch": 0.6494898841431783, + "grad_norm": 0.7304125428199768, + "learning_rate": 3.699927166788055e-05, + "loss": 1.5426, + "step": 939 + }, + { + "epoch": 0.6501815666609027, + "grad_norm": 0.886914074420929, + "learning_rate": 3.692643845593591e-05, + "loss": 1.6769, + "step": 940 + }, + { + "epoch": 0.650873249178627, + "grad_norm": 1.057144284248352, + "learning_rate": 3.6853605243991265e-05, + "loss": 1.4546, + "step": 941 + }, + { + "epoch": 0.6515649316963513, + "grad_norm": 0.9073127508163452, + "learning_rate": 3.678077203204662e-05, + "loss": 1.6194, + "step": 942 + }, + { + "epoch": 0.6522566142140758, + "grad_norm": 0.9237063527107239, + "learning_rate": 3.670793882010197e-05, + "loss": 1.4516, + "step": 943 + }, + { + "epoch": 0.6529482967318001, + "grad_norm": 0.7701244354248047, + "learning_rate": 3.663510560815732e-05, + "loss": 1.7599, + "step": 944 + }, + { + "epoch": 0.6536399792495244, + "grad_norm": 0.7972707152366638, + "learning_rate": 3.6562272396212676e-05, + "loss": 1.9227, + "step": 945 + }, + { + "epoch": 0.6543316617672489, + "grad_norm": 0.6878631711006165, + "learning_rate": 3.648943918426802e-05, + "loss": 1.2225, + "step": 946 + }, + { + "epoch": 0.6550233442849732, + "grad_norm": 0.6954039335250854, + "learning_rate": 3.6416605972323384e-05, + "loss": 1.3706, + "step": 947 + }, + { + "epoch": 0.6557150268026976, + "grad_norm": 0.8597190380096436, + "learning_rate": 3.634377276037874e-05, + "loss": 2.0226, + "step": 948 + }, + { + "epoch": 0.6564067093204219, + "grad_norm": 0.8058347702026367, + "learning_rate": 3.6270939548434086e-05, + "loss": 1.8904, + "step": 949 + }, + { + "epoch": 0.6570983918381463, + "grad_norm": 0.7838011384010315, + "learning_rate": 3.619810633648944e-05, + "loss": 1.1612, + "step": 950 + }, + { + "epoch": 0.6577900743558707, + "grad_norm": 0.9218417406082153, + "learning_rate": 3.6125273124544795e-05, + "loss": 1.9737, + "step": 951 + }, + { + "epoch": 0.658481756873595, + "grad_norm": 0.7816157937049866, + "learning_rate": 3.605243991260015e-05, + "loss": 0.86, + "step": 952 + }, + { + "epoch": 0.6591734393913193, + "grad_norm": 0.8489840626716614, + "learning_rate": 3.5979606700655496e-05, + "loss": 2.0043, + "step": 953 + }, + { + "epoch": 0.6598651219090438, + "grad_norm": 1.0207152366638184, + "learning_rate": 3.590677348871086e-05, + "loss": 1.4315, + "step": 954 + }, + { + "epoch": 0.6605568044267681, + "grad_norm": 0.8000298738479614, + "learning_rate": 3.583394027676621e-05, + "loss": 1.9495, + "step": 955 + }, + { + "epoch": 0.6612484869444925, + "grad_norm": 0.8676748871803284, + "learning_rate": 3.576110706482156e-05, + "loss": 1.7678, + "step": 956 + }, + { + "epoch": 0.6619401694622168, + "grad_norm": 0.9304187297821045, + "learning_rate": 3.5688273852876913e-05, + "loss": 2.0562, + "step": 957 + }, + { + "epoch": 0.6626318519799412, + "grad_norm": 1.0136486291885376, + "learning_rate": 3.561544064093227e-05, + "loss": 1.7028, + "step": 958 + }, + { + "epoch": 0.6633235344976656, + "grad_norm": 0.735945463180542, + "learning_rate": 3.5542607428987615e-05, + "loss": 1.3553, + "step": 959 + }, + { + "epoch": 0.6640152170153899, + "grad_norm": 0.7199145555496216, + "learning_rate": 3.546977421704297e-05, + "loss": 1.3213, + "step": 960 + }, + { + "epoch": 0.6647068995331143, + "grad_norm": 0.8235880732536316, + "learning_rate": 3.539694100509833e-05, + "loss": 2.0212, + "step": 961 + }, + { + "epoch": 0.6653985820508387, + "grad_norm": 1.330508828163147, + "learning_rate": 3.5324107793153685e-05, + "loss": 1.8431, + "step": 962 + }, + { + "epoch": 0.666090264568563, + "grad_norm": 0.7981826663017273, + "learning_rate": 3.525127458120903e-05, + "loss": 1.1926, + "step": 963 + }, + { + "epoch": 0.6667819470862874, + "grad_norm": 0.9556028842926025, + "learning_rate": 3.517844136926439e-05, + "loss": 1.1749, + "step": 964 + }, + { + "epoch": 0.6674736296040118, + "grad_norm": 0.9828068614006042, + "learning_rate": 3.510560815731974e-05, + "loss": 1.4142, + "step": 965 + }, + { + "epoch": 0.6681653121217361, + "grad_norm": 1.730177879333496, + "learning_rate": 3.503277494537509e-05, + "loss": 1.8473, + "step": 966 + }, + { + "epoch": 0.6688569946394605, + "grad_norm": 0.7303544878959656, + "learning_rate": 3.495994173343044e-05, + "loss": 1.3033, + "step": 967 + }, + { + "epoch": 0.6695486771571848, + "grad_norm": 0.7854034900665283, + "learning_rate": 3.4887108521485804e-05, + "loss": 1.2283, + "step": 968 + }, + { + "epoch": 0.6702403596749092, + "grad_norm": 0.8185420036315918, + "learning_rate": 3.481427530954115e-05, + "loss": 2.0494, + "step": 969 + }, + { + "epoch": 0.6709320421926336, + "grad_norm": 0.7962689995765686, + "learning_rate": 3.4741442097596506e-05, + "loss": 1.9523, + "step": 970 + }, + { + "epoch": 0.6716237247103579, + "grad_norm": 0.9332857131958008, + "learning_rate": 3.466860888565186e-05, + "loss": 1.6991, + "step": 971 + }, + { + "epoch": 0.6723154072280824, + "grad_norm": 0.7964072823524475, + "learning_rate": 3.4595775673707214e-05, + "loss": 1.6715, + "step": 972 + }, + { + "epoch": 0.6730070897458067, + "grad_norm": 0.8391403555870056, + "learning_rate": 3.452294246176256e-05, + "loss": 1.8878, + "step": 973 + }, + { + "epoch": 0.673698772263531, + "grad_norm": 0.8566287159919739, + "learning_rate": 3.4450109249817916e-05, + "loss": 1.7805, + "step": 974 + }, + { + "epoch": 0.6743904547812554, + "grad_norm": 0.9097205400466919, + "learning_rate": 3.437727603787328e-05, + "loss": 1.4648, + "step": 975 + }, + { + "epoch": 0.6750821372989798, + "grad_norm": 0.9566125273704529, + "learning_rate": 3.4304442825928625e-05, + "loss": 1.4794, + "step": 976 + }, + { + "epoch": 0.6757738198167041, + "grad_norm": 0.89398193359375, + "learning_rate": 3.423160961398398e-05, + "loss": 2.0932, + "step": 977 + }, + { + "epoch": 0.6764655023344285, + "grad_norm": 0.8663723468780518, + "learning_rate": 3.415877640203933e-05, + "loss": 1.8625, + "step": 978 + }, + { + "epoch": 0.6771571848521528, + "grad_norm": 0.751266598701477, + "learning_rate": 3.408594319009468e-05, + "loss": 1.0625, + "step": 979 + }, + { + "epoch": 0.6778488673698773, + "grad_norm": 0.8049336671829224, + "learning_rate": 3.4013109978150035e-05, + "loss": 1.3388, + "step": 980 + }, + { + "epoch": 0.6785405498876016, + "grad_norm": 0.8025321364402771, + "learning_rate": 3.394027676620539e-05, + "loss": 1.4493, + "step": 981 + }, + { + "epoch": 0.6792322324053259, + "grad_norm": 0.8283219933509827, + "learning_rate": 3.386744355426075e-05, + "loss": 1.1765, + "step": 982 + }, + { + "epoch": 0.6799239149230503, + "grad_norm": 1.0506592988967896, + "learning_rate": 3.37946103423161e-05, + "loss": 1.2619, + "step": 983 + }, + { + "epoch": 0.6806155974407747, + "grad_norm": 0.5937513709068298, + "learning_rate": 3.372177713037145e-05, + "loss": 0.9384, + "step": 984 + }, + { + "epoch": 0.681307279958499, + "grad_norm": 0.6834938526153564, + "learning_rate": 3.3648943918426806e-05, + "loss": 1.0192, + "step": 985 + }, + { + "epoch": 0.6819989624762234, + "grad_norm": 1.1329364776611328, + "learning_rate": 3.3576110706482154e-05, + "loss": 2.2567, + "step": 986 + }, + { + "epoch": 0.6826906449939478, + "grad_norm": 0.7666782736778259, + "learning_rate": 3.350327749453751e-05, + "loss": 1.8783, + "step": 987 + }, + { + "epoch": 0.6833823275116722, + "grad_norm": 1.2282352447509766, + "learning_rate": 3.343044428259286e-05, + "loss": 1.5103, + "step": 988 + }, + { + "epoch": 0.6840740100293965, + "grad_norm": 0.7626004219055176, + "learning_rate": 3.335761107064822e-05, + "loss": 1.3147, + "step": 989 + }, + { + "epoch": 0.6847656925471208, + "grad_norm": 1.0253794193267822, + "learning_rate": 3.328477785870357e-05, + "loss": 1.7115, + "step": 990 + }, + { + "epoch": 0.6854573750648453, + "grad_norm": 0.890225887298584, + "learning_rate": 3.3211944646758925e-05, + "loss": 1.6278, + "step": 991 + }, + { + "epoch": 0.6861490575825696, + "grad_norm": 1.0109024047851562, + "learning_rate": 3.313911143481428e-05, + "loss": 1.5538, + "step": 992 + }, + { + "epoch": 0.686840740100294, + "grad_norm": 0.8567763566970825, + "learning_rate": 3.306627822286963e-05, + "loss": 1.7241, + "step": 993 + }, + { + "epoch": 0.6875324226180183, + "grad_norm": 0.7532871961593628, + "learning_rate": 3.299344501092498e-05, + "loss": 1.3609, + "step": 994 + }, + { + "epoch": 0.6882241051357427, + "grad_norm": 0.9524408578872681, + "learning_rate": 3.2920611798980336e-05, + "loss": 1.9429, + "step": 995 + }, + { + "epoch": 0.6889157876534671, + "grad_norm": 0.6177567839622498, + "learning_rate": 3.284777858703569e-05, + "loss": 0.9766, + "step": 996 + }, + { + "epoch": 0.6896074701711914, + "grad_norm": 1.1738656759262085, + "learning_rate": 3.2774945375091044e-05, + "loss": 1.6276, + "step": 997 + }, + { + "epoch": 0.6902991526889158, + "grad_norm": 1.018707036972046, + "learning_rate": 3.27021121631464e-05, + "loss": 0.8201, + "step": 998 + }, + { + "epoch": 0.6909908352066402, + "grad_norm": 0.8498096466064453, + "learning_rate": 3.2629278951201746e-05, + "loss": 0.9932, + "step": 999 + }, + { + "epoch": 0.6916825177243645, + "grad_norm": 0.8145414590835571, + "learning_rate": 3.25564457392571e-05, + "loss": 1.9895, + "step": 1000 + }, + { + "epoch": 0.6916825177243645, + "eval_loss": 1.4734511375427246, + "eval_runtime": 636.6322, + "eval_samples_per_second": 2.018, + "eval_steps_per_second": 1.01, + "step": 1000 + }, + { + "epoch": 0.6923742002420888, + "grad_norm": 0.9223456382751465, + "learning_rate": 3.2483612527312455e-05, + "loss": 1.7351, + "step": 1001 + }, + { + "epoch": 0.6930658827598133, + "grad_norm": 0.6885508894920349, + "learning_rate": 3.241077931536781e-05, + "loss": 0.8462, + "step": 1002 + }, + { + "epoch": 0.6937575652775376, + "grad_norm": 0.9678956866264343, + "learning_rate": 3.233794610342316e-05, + "loss": 1.7896, + "step": 1003 + }, + { + "epoch": 0.694449247795262, + "grad_norm": 0.9409008622169495, + "learning_rate": 3.226511289147852e-05, + "loss": 1.1164, + "step": 1004 + }, + { + "epoch": 0.6951409303129863, + "grad_norm": 0.959503710269928, + "learning_rate": 3.219227967953387e-05, + "loss": 1.8561, + "step": 1005 + }, + { + "epoch": 0.6958326128307107, + "grad_norm": 0.7940098643302917, + "learning_rate": 3.211944646758922e-05, + "loss": 1.2933, + "step": 1006 + }, + { + "epoch": 0.6965242953484351, + "grad_norm": 1.091283917427063, + "learning_rate": 3.2046613255644574e-05, + "loss": 1.8583, + "step": 1007 + }, + { + "epoch": 0.6972159778661594, + "grad_norm": 0.8690599799156189, + "learning_rate": 3.197378004369993e-05, + "loss": 1.2886, + "step": 1008 + }, + { + "epoch": 0.6979076603838837, + "grad_norm": 0.9679297804832458, + "learning_rate": 3.190094683175528e-05, + "loss": 1.8224, + "step": 1009 + }, + { + "epoch": 0.6985993429016082, + "grad_norm": 0.8979253172874451, + "learning_rate": 3.1828113619810637e-05, + "loss": 1.0664, + "step": 1010 + }, + { + "epoch": 0.6992910254193325, + "grad_norm": 0.849543035030365, + "learning_rate": 3.175528040786599e-05, + "loss": 1.5253, + "step": 1011 + }, + { + "epoch": 0.6999827079370569, + "grad_norm": 0.8288280963897705, + "learning_rate": 3.1682447195921345e-05, + "loss": 1.0364, + "step": 1012 + }, + { + "epoch": 0.7006743904547813, + "grad_norm": 0.9843714833259583, + "learning_rate": 3.160961398397669e-05, + "loss": 1.3689, + "step": 1013 + }, + { + "epoch": 0.7013660729725056, + "grad_norm": 0.8555485010147095, + "learning_rate": 3.153678077203205e-05, + "loss": 1.8278, + "step": 1014 + }, + { + "epoch": 0.70205775549023, + "grad_norm": 1.4214497804641724, + "learning_rate": 3.14639475600874e-05, + "loss": 1.4655, + "step": 1015 + }, + { + "epoch": 0.7027494380079543, + "grad_norm": 1.129608392715454, + "learning_rate": 3.1391114348142756e-05, + "loss": 1.6663, + "step": 1016 + }, + { + "epoch": 0.7034411205256788, + "grad_norm": 0.7891383171081543, + "learning_rate": 3.131828113619811e-05, + "loss": 1.7844, + "step": 1017 + }, + { + "epoch": 0.7041328030434031, + "grad_norm": 0.6413160562515259, + "learning_rate": 3.1245447924253464e-05, + "loss": 1.4443, + "step": 1018 + }, + { + "epoch": 0.7048244855611274, + "grad_norm": 0.836772084236145, + "learning_rate": 3.117261471230881e-05, + "loss": 1.8385, + "step": 1019 + }, + { + "epoch": 0.7055161680788518, + "grad_norm": 0.7464689016342163, + "learning_rate": 3.1099781500364166e-05, + "loss": 1.697, + "step": 1020 + }, + { + "epoch": 0.7062078505965762, + "grad_norm": 0.8900326490402222, + "learning_rate": 3.102694828841952e-05, + "loss": 2.2027, + "step": 1021 + }, + { + "epoch": 0.7068995331143005, + "grad_norm": 0.9313298463821411, + "learning_rate": 3.0954115076474874e-05, + "loss": 1.9712, + "step": 1022 + }, + { + "epoch": 0.7075912156320249, + "grad_norm": 0.8505009412765503, + "learning_rate": 3.088128186453023e-05, + "loss": 1.7874, + "step": 1023 + }, + { + "epoch": 0.7082828981497493, + "grad_norm": 1.010100245475769, + "learning_rate": 3.080844865258558e-05, + "loss": 1.7268, + "step": 1024 + }, + { + "epoch": 0.7089745806674737, + "grad_norm": 0.9934255480766296, + "learning_rate": 3.073561544064094e-05, + "loss": 0.9234, + "step": 1025 + }, + { + "epoch": 0.709666263185198, + "grad_norm": 1.2760083675384521, + "learning_rate": 3.0662782228696285e-05, + "loss": 1.5318, + "step": 1026 + }, + { + "epoch": 0.7103579457029223, + "grad_norm": 1.0074844360351562, + "learning_rate": 3.058994901675164e-05, + "loss": 1.9285, + "step": 1027 + }, + { + "epoch": 0.7110496282206468, + "grad_norm": 0.8921088576316833, + "learning_rate": 3.0517115804806993e-05, + "loss": 1.701, + "step": 1028 + }, + { + "epoch": 0.7117413107383711, + "grad_norm": 0.7134079337120056, + "learning_rate": 3.0444282592862344e-05, + "loss": 1.3919, + "step": 1029 + }, + { + "epoch": 0.7124329932560954, + "grad_norm": 0.7399762868881226, + "learning_rate": 3.03714493809177e-05, + "loss": 1.6299, + "step": 1030 + }, + { + "epoch": 0.7131246757738198, + "grad_norm": 0.7438504099845886, + "learning_rate": 3.0298616168973053e-05, + "loss": 1.4501, + "step": 1031 + }, + { + "epoch": 0.7138163582915442, + "grad_norm": 0.9777660369873047, + "learning_rate": 3.022578295702841e-05, + "loss": 1.8551, + "step": 1032 + }, + { + "epoch": 0.7145080408092686, + "grad_norm": 0.7377002239227295, + "learning_rate": 3.0152949745083758e-05, + "loss": 1.7002, + "step": 1033 + }, + { + "epoch": 0.7151997233269929, + "grad_norm": 0.7134854197502136, + "learning_rate": 3.0080116533139112e-05, + "loss": 2.0436, + "step": 1034 + }, + { + "epoch": 0.7158914058447172, + "grad_norm": 1.086472511291504, + "learning_rate": 3.0007283321194467e-05, + "loss": 1.6442, + "step": 1035 + }, + { + "epoch": 0.7165830883624417, + "grad_norm": 0.5974130630493164, + "learning_rate": 2.9934450109249818e-05, + "loss": 0.6907, + "step": 1036 + }, + { + "epoch": 0.717274770880166, + "grad_norm": 1.6432491540908813, + "learning_rate": 2.9861616897305172e-05, + "loss": 1.7509, + "step": 1037 + }, + { + "epoch": 0.7179664533978903, + "grad_norm": 0.7549503445625305, + "learning_rate": 2.9788783685360526e-05, + "loss": 1.3029, + "step": 1038 + }, + { + "epoch": 0.7186581359156148, + "grad_norm": 0.7781493663787842, + "learning_rate": 2.9715950473415877e-05, + "loss": 1.8372, + "step": 1039 + }, + { + "epoch": 0.7193498184333391, + "grad_norm": 0.8380869626998901, + "learning_rate": 2.964311726147123e-05, + "loss": 0.9865, + "step": 1040 + }, + { + "epoch": 0.7200415009510635, + "grad_norm": 0.7673766613006592, + "learning_rate": 2.9570284049526586e-05, + "loss": 1.5038, + "step": 1041 + }, + { + "epoch": 0.7207331834687878, + "grad_norm": 0.8372862935066223, + "learning_rate": 2.949745083758194e-05, + "loss": 1.3747, + "step": 1042 + }, + { + "epoch": 0.7214248659865122, + "grad_norm": 0.8145121932029724, + "learning_rate": 2.942461762563729e-05, + "loss": 1.7284, + "step": 1043 + }, + { + "epoch": 0.7221165485042366, + "grad_norm": 0.7420305013656616, + "learning_rate": 2.9351784413692645e-05, + "loss": 1.4102, + "step": 1044 + }, + { + "epoch": 0.7228082310219609, + "grad_norm": 0.6267819404602051, + "learning_rate": 2.9278951201748e-05, + "loss": 1.4778, + "step": 1045 + }, + { + "epoch": 0.7234999135396852, + "grad_norm": 0.9034648537635803, + "learning_rate": 2.920611798980335e-05, + "loss": 1.5424, + "step": 1046 + }, + { + "epoch": 0.7241915960574097, + "grad_norm": 0.8761317133903503, + "learning_rate": 2.9133284777858705e-05, + "loss": 2.1288, + "step": 1047 + }, + { + "epoch": 0.724883278575134, + "grad_norm": 0.8351477384567261, + "learning_rate": 2.906045156591406e-05, + "loss": 1.6834, + "step": 1048 + }, + { + "epoch": 0.7255749610928584, + "grad_norm": 0.6748347878456116, + "learning_rate": 2.898761835396941e-05, + "loss": 1.7173, + "step": 1049 + }, + { + "epoch": 0.7262666436105828, + "grad_norm": 0.794681191444397, + "learning_rate": 2.8914785142024764e-05, + "loss": 1.9143, + "step": 1050 + }, + { + "epoch": 0.7269583261283071, + "grad_norm": 0.7815545797348022, + "learning_rate": 2.884195193008012e-05, + "loss": 1.7722, + "step": 1051 + }, + { + "epoch": 0.7276500086460315, + "grad_norm": 1.7133221626281738, + "learning_rate": 2.8769118718135473e-05, + "loss": 1.354, + "step": 1052 + }, + { + "epoch": 0.7283416911637558, + "grad_norm": 0.9203530550003052, + "learning_rate": 2.8696285506190824e-05, + "loss": 1.6095, + "step": 1053 + }, + { + "epoch": 0.7290333736814802, + "grad_norm": 0.4724554121494293, + "learning_rate": 2.8623452294246178e-05, + "loss": 0.5242, + "step": 1054 + }, + { + "epoch": 0.7297250561992046, + "grad_norm": 1.373349666595459, + "learning_rate": 2.8550619082301532e-05, + "loss": 0.8815, + "step": 1055 + }, + { + "epoch": 0.7304167387169289, + "grad_norm": 0.7870145440101624, + "learning_rate": 2.8477785870356883e-05, + "loss": 1.8316, + "step": 1056 + }, + { + "epoch": 0.7311084212346532, + "grad_norm": 1.0465471744537354, + "learning_rate": 2.8404952658412237e-05, + "loss": 1.8831, + "step": 1057 + }, + { + "epoch": 0.7318001037523777, + "grad_norm": 0.7755276560783386, + "learning_rate": 2.833211944646759e-05, + "loss": 1.0587, + "step": 1058 + }, + { + "epoch": 0.732491786270102, + "grad_norm": 1.0767560005187988, + "learning_rate": 2.8259286234522943e-05, + "loss": 2.2136, + "step": 1059 + }, + { + "epoch": 0.7331834687878264, + "grad_norm": 1.0757145881652832, + "learning_rate": 2.8186453022578297e-05, + "loss": 2.0378, + "step": 1060 + }, + { + "epoch": 0.7338751513055507, + "grad_norm": 0.6956034302711487, + "learning_rate": 2.811361981063365e-05, + "loss": 0.9547, + "step": 1061 + }, + { + "epoch": 0.7345668338232751, + "grad_norm": 0.9913384914398193, + "learning_rate": 2.8040786598689005e-05, + "loss": 1.7596, + "step": 1062 + }, + { + "epoch": 0.7352585163409995, + "grad_norm": 0.8847383856773376, + "learning_rate": 2.7967953386744356e-05, + "loss": 1.7152, + "step": 1063 + }, + { + "epoch": 0.7359501988587238, + "grad_norm": 0.9734081029891968, + "learning_rate": 2.789512017479971e-05, + "loss": 1.358, + "step": 1064 + }, + { + "epoch": 0.7366418813764483, + "grad_norm": 0.8713496923446655, + "learning_rate": 2.7822286962855065e-05, + "loss": 1.766, + "step": 1065 + }, + { + "epoch": 0.7373335638941726, + "grad_norm": 0.8987798094749451, + "learning_rate": 2.7749453750910416e-05, + "loss": 1.4112, + "step": 1066 + }, + { + "epoch": 0.7380252464118969, + "grad_norm": 0.7988868951797485, + "learning_rate": 2.767662053896577e-05, + "loss": 1.8304, + "step": 1067 + }, + { + "epoch": 0.7387169289296213, + "grad_norm": 0.6583023071289062, + "learning_rate": 2.7603787327021124e-05, + "loss": 1.4216, + "step": 1068 + }, + { + "epoch": 0.7394086114473457, + "grad_norm": 0.8360283970832825, + "learning_rate": 2.7530954115076475e-05, + "loss": 1.1174, + "step": 1069 + }, + { + "epoch": 0.74010029396507, + "grad_norm": 0.7958647608757019, + "learning_rate": 2.745812090313183e-05, + "loss": 1.8732, + "step": 1070 + }, + { + "epoch": 0.7407919764827944, + "grad_norm": 0.8186526894569397, + "learning_rate": 2.7385287691187184e-05, + "loss": 1.6917, + "step": 1071 + }, + { + "epoch": 0.7414836590005187, + "grad_norm": 0.7415515184402466, + "learning_rate": 2.7312454479242538e-05, + "loss": 1.8473, + "step": 1072 + }, + { + "epoch": 0.7421753415182432, + "grad_norm": 0.826209306716919, + "learning_rate": 2.723962126729789e-05, + "loss": 1.989, + "step": 1073 + }, + { + "epoch": 0.7428670240359675, + "grad_norm": 1.1724082231521606, + "learning_rate": 2.7166788055353243e-05, + "loss": 2.0115, + "step": 1074 + }, + { + "epoch": 0.7435587065536918, + "grad_norm": 0.8580861687660217, + "learning_rate": 2.7093954843408598e-05, + "loss": 1.9135, + "step": 1075 + }, + { + "epoch": 0.7442503890714163, + "grad_norm": 1.016176462173462, + "learning_rate": 2.702112163146395e-05, + "loss": 1.7682, + "step": 1076 + }, + { + "epoch": 0.7449420715891406, + "grad_norm": 0.8663372993469238, + "learning_rate": 2.6948288419519303e-05, + "loss": 2.1598, + "step": 1077 + }, + { + "epoch": 0.7456337541068649, + "grad_norm": 0.7362541556358337, + "learning_rate": 2.6875455207574657e-05, + "loss": 1.8869, + "step": 1078 + }, + { + "epoch": 0.7463254366245893, + "grad_norm": 1.2881722450256348, + "learning_rate": 2.6802621995630005e-05, + "loss": 1.7314, + "step": 1079 + }, + { + "epoch": 0.7470171191423137, + "grad_norm": 0.7441843748092651, + "learning_rate": 2.6729788783685362e-05, + "loss": 1.3527, + "step": 1080 + }, + { + "epoch": 0.7477088016600381, + "grad_norm": 0.8295880556106567, + "learning_rate": 2.6656955571740717e-05, + "loss": 1.1347, + "step": 1081 + }, + { + "epoch": 0.7484004841777624, + "grad_norm": 0.9294624924659729, + "learning_rate": 2.658412235979607e-05, + "loss": 2.1327, + "step": 1082 + }, + { + "epoch": 0.7490921666954867, + "grad_norm": 0.8970943689346313, + "learning_rate": 2.6511289147851422e-05, + "loss": 1.5343, + "step": 1083 + }, + { + "epoch": 0.7497838492132112, + "grad_norm": 0.754835844039917, + "learning_rate": 2.6438455935906776e-05, + "loss": 0.9905, + "step": 1084 + }, + { + "epoch": 0.7504755317309355, + "grad_norm": 0.6362099051475525, + "learning_rate": 2.636562272396213e-05, + "loss": 1.1349, + "step": 1085 + }, + { + "epoch": 0.7511672142486598, + "grad_norm": 0.7959074378013611, + "learning_rate": 2.6292789512017478e-05, + "loss": 1.9783, + "step": 1086 + }, + { + "epoch": 0.7518588967663842, + "grad_norm": 1.3375893831253052, + "learning_rate": 2.6219956300072835e-05, + "loss": 0.9572, + "step": 1087 + }, + { + "epoch": 0.7525505792841086, + "grad_norm": 0.9002304673194885, + "learning_rate": 2.614712308812819e-05, + "loss": 1.3006, + "step": 1088 + }, + { + "epoch": 0.753242261801833, + "grad_norm": 0.5876688361167908, + "learning_rate": 2.6074289876183537e-05, + "loss": 1.0762, + "step": 1089 + }, + { + "epoch": 0.7539339443195573, + "grad_norm": 0.7274353504180908, + "learning_rate": 2.6001456664238895e-05, + "loss": 1.1997, + "step": 1090 + }, + { + "epoch": 0.7546256268372817, + "grad_norm": 0.8360443115234375, + "learning_rate": 2.592862345229425e-05, + "loss": 1.5214, + "step": 1091 + }, + { + "epoch": 0.7553173093550061, + "grad_norm": 0.9110802412033081, + "learning_rate": 2.5855790240349604e-05, + "loss": 1.6864, + "step": 1092 + }, + { + "epoch": 0.7560089918727304, + "grad_norm": 0.7017329931259155, + "learning_rate": 2.578295702840495e-05, + "loss": 1.0758, + "step": 1093 + }, + { + "epoch": 0.7567006743904547, + "grad_norm": 0.9383176565170288, + "learning_rate": 2.571012381646031e-05, + "loss": 1.9367, + "step": 1094 + }, + { + "epoch": 0.7573923569081792, + "grad_norm": 0.5317110419273376, + "learning_rate": 2.5637290604515663e-05, + "loss": 0.5657, + "step": 1095 + }, + { + "epoch": 0.7580840394259035, + "grad_norm": 1.021541953086853, + "learning_rate": 2.556445739257101e-05, + "loss": 1.7779, + "step": 1096 + }, + { + "epoch": 0.7587757219436279, + "grad_norm": 0.6099836826324463, + "learning_rate": 2.5491624180626368e-05, + "loss": 0.625, + "step": 1097 + }, + { + "epoch": 0.7594674044613522, + "grad_norm": 0.6868588328361511, + "learning_rate": 2.5418790968681723e-05, + "loss": 0.8857, + "step": 1098 + }, + { + "epoch": 0.7601590869790766, + "grad_norm": 0.722844123840332, + "learning_rate": 2.534595775673707e-05, + "loss": 0.9726, + "step": 1099 + }, + { + "epoch": 0.760850769496801, + "grad_norm": 0.6676250100135803, + "learning_rate": 2.5273124544792424e-05, + "loss": 1.3346, + "step": 1100 + }, + { + "epoch": 0.7615424520145253, + "grad_norm": 1.0060703754425049, + "learning_rate": 2.5200291332847782e-05, + "loss": 1.4589, + "step": 1101 + }, + { + "epoch": 0.7622341345322498, + "grad_norm": 0.9362277388572693, + "learning_rate": 2.5127458120903136e-05, + "loss": 1.8174, + "step": 1102 + }, + { + "epoch": 0.7629258170499741, + "grad_norm": 0.7632747888565063, + "learning_rate": 2.5054624908958484e-05, + "loss": 1.2281, + "step": 1103 + }, + { + "epoch": 0.7636174995676984, + "grad_norm": 1.2207857370376587, + "learning_rate": 2.498179169701384e-05, + "loss": 2.126, + "step": 1104 + }, + { + "epoch": 0.7643091820854228, + "grad_norm": 1.2145875692367554, + "learning_rate": 2.4908958485069192e-05, + "loss": 1.8128, + "step": 1105 + }, + { + "epoch": 0.7650008646031472, + "grad_norm": 0.7407699823379517, + "learning_rate": 2.4836125273124547e-05, + "loss": 1.4689, + "step": 1106 + }, + { + "epoch": 0.7656925471208715, + "grad_norm": 0.7252597212791443, + "learning_rate": 2.4763292061179898e-05, + "loss": 1.5147, + "step": 1107 + }, + { + "epoch": 0.7663842296385959, + "grad_norm": 1.0968459844589233, + "learning_rate": 2.4690458849235252e-05, + "loss": 0.9713, + "step": 1108 + }, + { + "epoch": 0.7670759121563202, + "grad_norm": 0.9213295578956604, + "learning_rate": 2.4617625637290606e-05, + "loss": 1.7205, + "step": 1109 + }, + { + "epoch": 0.7677675946740447, + "grad_norm": 0.8879563212394714, + "learning_rate": 2.4544792425345957e-05, + "loss": 1.8252, + "step": 1110 + }, + { + "epoch": 0.768459277191769, + "grad_norm": 0.8051623702049255, + "learning_rate": 2.4471959213401315e-05, + "loss": 1.0521, + "step": 1111 + }, + { + "epoch": 0.7691509597094933, + "grad_norm": 1.2763499021530151, + "learning_rate": 2.4399126001456666e-05, + "loss": 1.6888, + "step": 1112 + }, + { + "epoch": 0.7698426422272177, + "grad_norm": 0.7829091548919678, + "learning_rate": 2.4326292789512017e-05, + "loss": 1.8972, + "step": 1113 + }, + { + "epoch": 0.7705343247449421, + "grad_norm": 0.9532336592674255, + "learning_rate": 2.425345957756737e-05, + "loss": 1.6777, + "step": 1114 + }, + { + "epoch": 0.7712260072626664, + "grad_norm": 0.7734281420707703, + "learning_rate": 2.4180626365622725e-05, + "loss": 1.9136, + "step": 1115 + }, + { + "epoch": 0.7719176897803908, + "grad_norm": 0.7758744359016418, + "learning_rate": 2.410779315367808e-05, + "loss": 1.606, + "step": 1116 + }, + { + "epoch": 0.7726093722981152, + "grad_norm": 0.7138837575912476, + "learning_rate": 2.403495994173343e-05, + "loss": 1.3624, + "step": 1117 + }, + { + "epoch": 0.7733010548158396, + "grad_norm": 1.141640543937683, + "learning_rate": 2.3962126729788785e-05, + "loss": 1.4454, + "step": 1118 + }, + { + "epoch": 0.7739927373335639, + "grad_norm": 0.9187803268432617, + "learning_rate": 2.388929351784414e-05, + "loss": 1.875, + "step": 1119 + }, + { + "epoch": 0.7746844198512882, + "grad_norm": 1.2046056985855103, + "learning_rate": 2.381646030589949e-05, + "loss": 1.575, + "step": 1120 + }, + { + "epoch": 0.7753761023690127, + "grad_norm": 0.8147513270378113, + "learning_rate": 2.3743627093954844e-05, + "loss": 1.417, + "step": 1121 + }, + { + "epoch": 0.776067784886737, + "grad_norm": 0.8273805379867554, + "learning_rate": 2.36707938820102e-05, + "loss": 1.4772, + "step": 1122 + }, + { + "epoch": 0.7767594674044613, + "grad_norm": 0.8634644746780396, + "learning_rate": 2.359796067006555e-05, + "loss": 1.7216, + "step": 1123 + }, + { + "epoch": 0.7774511499221857, + "grad_norm": 1.0923925638198853, + "learning_rate": 2.3525127458120904e-05, + "loss": 1.7762, + "step": 1124 + }, + { + "epoch": 0.7781428324399101, + "grad_norm": 0.788434624671936, + "learning_rate": 2.3452294246176258e-05, + "loss": 1.9594, + "step": 1125 + }, + { + "epoch": 0.7788345149576344, + "grad_norm": 1.1660581827163696, + "learning_rate": 2.3379461034231612e-05, + "loss": 1.4304, + "step": 1126 + }, + { + "epoch": 0.7795261974753588, + "grad_norm": 0.856432318687439, + "learning_rate": 2.3306627822286963e-05, + "loss": 1.4878, + "step": 1127 + }, + { + "epoch": 0.7802178799930832, + "grad_norm": 0.7626217007637024, + "learning_rate": 2.3233794610342317e-05, + "loss": 1.4872, + "step": 1128 + }, + { + "epoch": 0.7809095625108076, + "grad_norm": 0.7981844544410706, + "learning_rate": 2.316096139839767e-05, + "loss": 1.7608, + "step": 1129 + }, + { + "epoch": 0.7816012450285319, + "grad_norm": 0.9393052458763123, + "learning_rate": 2.3088128186453022e-05, + "loss": 1.784, + "step": 1130 + }, + { + "epoch": 0.7822929275462562, + "grad_norm": 0.8461928963661194, + "learning_rate": 2.3015294974508377e-05, + "loss": 1.2737, + "step": 1131 + }, + { + "epoch": 0.7829846100639807, + "grad_norm": 0.8578364849090576, + "learning_rate": 2.294246176256373e-05, + "loss": 1.8538, + "step": 1132 + }, + { + "epoch": 0.783676292581705, + "grad_norm": 0.9262241125106812, + "learning_rate": 2.2869628550619082e-05, + "loss": 2.2347, + "step": 1133 + }, + { + "epoch": 0.7843679750994293, + "grad_norm": 0.7736664414405823, + "learning_rate": 2.2796795338674436e-05, + "loss": 1.3659, + "step": 1134 + }, + { + "epoch": 0.7850596576171537, + "grad_norm": 0.798596978187561, + "learning_rate": 2.272396212672979e-05, + "loss": 1.7982, + "step": 1135 + }, + { + "epoch": 0.7857513401348781, + "grad_norm": 0.8106185793876648, + "learning_rate": 2.2651128914785145e-05, + "loss": 1.7942, + "step": 1136 + }, + { + "epoch": 0.7864430226526025, + "grad_norm": 0.9546266198158264, + "learning_rate": 2.2578295702840496e-05, + "loss": 1.974, + "step": 1137 + }, + { + "epoch": 0.7871347051703268, + "grad_norm": 0.7522141337394714, + "learning_rate": 2.250546249089585e-05, + "loss": 1.7433, + "step": 1138 + }, + { + "epoch": 0.7878263876880511, + "grad_norm": 0.6415102481842041, + "learning_rate": 2.2432629278951204e-05, + "loss": 1.0512, + "step": 1139 + }, + { + "epoch": 0.7885180702057756, + "grad_norm": 0.937563955783844, + "learning_rate": 2.2359796067006555e-05, + "loss": 1.7341, + "step": 1140 + }, + { + "epoch": 0.7892097527234999, + "grad_norm": 0.8371694087982178, + "learning_rate": 2.228696285506191e-05, + "loss": 1.8038, + "step": 1141 + }, + { + "epoch": 0.7899014352412242, + "grad_norm": 0.666323721408844, + "learning_rate": 2.2214129643117264e-05, + "loss": 1.281, + "step": 1142 + }, + { + "epoch": 0.7905931177589487, + "grad_norm": 0.7282040119171143, + "learning_rate": 2.2141296431172615e-05, + "loss": 1.6639, + "step": 1143 + }, + { + "epoch": 0.791284800276673, + "grad_norm": 1.0112649202346802, + "learning_rate": 2.206846321922797e-05, + "loss": 1.8349, + "step": 1144 + }, + { + "epoch": 0.7919764827943974, + "grad_norm": 0.986565887928009, + "learning_rate": 2.1995630007283323e-05, + "loss": 0.9161, + "step": 1145 + }, + { + "epoch": 0.7926681653121217, + "grad_norm": 0.8928730487823486, + "learning_rate": 2.1922796795338678e-05, + "loss": 1.7917, + "step": 1146 + }, + { + "epoch": 0.7933598478298461, + "grad_norm": 1.5609923601150513, + "learning_rate": 2.184996358339403e-05, + "loss": 1.5677, + "step": 1147 + }, + { + "epoch": 0.7940515303475705, + "grad_norm": 0.9750341773033142, + "learning_rate": 2.177713037144938e-05, + "loss": 1.4404, + "step": 1148 + }, + { + "epoch": 0.7947432128652948, + "grad_norm": 0.47437143325805664, + "learning_rate": 2.1704297159504737e-05, + "loss": 0.442, + "step": 1149 + }, + { + "epoch": 0.7954348953830191, + "grad_norm": 0.8163256645202637, + "learning_rate": 2.1631463947560088e-05, + "loss": 1.9685, + "step": 1150 + }, + { + "epoch": 0.7961265779007436, + "grad_norm": 1.0846987962722778, + "learning_rate": 2.1558630735615442e-05, + "loss": 1.6867, + "step": 1151 + }, + { + "epoch": 0.7968182604184679, + "grad_norm": 0.9128690361976624, + "learning_rate": 2.1485797523670793e-05, + "loss": 1.6015, + "step": 1152 + }, + { + "epoch": 0.7975099429361923, + "grad_norm": 1.0973516702651978, + "learning_rate": 2.1412964311726147e-05, + "loss": 1.4411, + "step": 1153 + }, + { + "epoch": 0.7982016254539167, + "grad_norm": 0.9682608842849731, + "learning_rate": 2.1340131099781502e-05, + "loss": 2.1468, + "step": 1154 + }, + { + "epoch": 0.798893307971641, + "grad_norm": 1.1797525882720947, + "learning_rate": 2.1267297887836853e-05, + "loss": 1.5891, + "step": 1155 + }, + { + "epoch": 0.7995849904893654, + "grad_norm": 0.7655579447746277, + "learning_rate": 2.119446467589221e-05, + "loss": 1.6008, + "step": 1156 + }, + { + "epoch": 0.8002766730070897, + "grad_norm": 0.8085634112358093, + "learning_rate": 2.112163146394756e-05, + "loss": 1.8382, + "step": 1157 + }, + { + "epoch": 0.8009683555248142, + "grad_norm": 0.8497790694236755, + "learning_rate": 2.1048798252002912e-05, + "loss": 1.8467, + "step": 1158 + }, + { + "epoch": 0.8016600380425385, + "grad_norm": 1.1677498817443848, + "learning_rate": 2.0975965040058266e-05, + "loss": 1.7032, + "step": 1159 + }, + { + "epoch": 0.8023517205602628, + "grad_norm": 1.2180325984954834, + "learning_rate": 2.090313182811362e-05, + "loss": 1.1789, + "step": 1160 + }, + { + "epoch": 0.8030434030779872, + "grad_norm": 0.8025382161140442, + "learning_rate": 2.0830298616168975e-05, + "loss": 1.0759, + "step": 1161 + }, + { + "epoch": 0.8037350855957116, + "grad_norm": 0.8110843300819397, + "learning_rate": 2.0757465404224326e-05, + "loss": 1.4129, + "step": 1162 + }, + { + "epoch": 0.8044267681134359, + "grad_norm": 0.8053744435310364, + "learning_rate": 2.068463219227968e-05, + "loss": 1.7426, + "step": 1163 + }, + { + "epoch": 0.8051184506311603, + "grad_norm": 0.951020359992981, + "learning_rate": 2.0611798980335034e-05, + "loss": 1.8264, + "step": 1164 + }, + { + "epoch": 0.8058101331488846, + "grad_norm": 0.8345422148704529, + "learning_rate": 2.0538965768390385e-05, + "loss": 1.8637, + "step": 1165 + }, + { + "epoch": 0.8065018156666091, + "grad_norm": 0.9381097555160522, + "learning_rate": 2.046613255644574e-05, + "loss": 1.9496, + "step": 1166 + }, + { + "epoch": 0.8071934981843334, + "grad_norm": 0.6921238899230957, + "learning_rate": 2.0393299344501094e-05, + "loss": 1.4406, + "step": 1167 + }, + { + "epoch": 0.8078851807020577, + "grad_norm": 0.7710753083229065, + "learning_rate": 2.0320466132556445e-05, + "loss": 1.5617, + "step": 1168 + }, + { + "epoch": 0.8085768632197822, + "grad_norm": 0.7170714139938354, + "learning_rate": 2.02476329206118e-05, + "loss": 1.9137, + "step": 1169 + }, + { + "epoch": 0.8092685457375065, + "grad_norm": 0.873906672000885, + "learning_rate": 2.0174799708667153e-05, + "loss": 2.014, + "step": 1170 + }, + { + "epoch": 0.8099602282552308, + "grad_norm": 0.8897404074668884, + "learning_rate": 2.0101966496722508e-05, + "loss": 1.3305, + "step": 1171 + }, + { + "epoch": 0.8106519107729552, + "grad_norm": 0.7641202807426453, + "learning_rate": 2.002913328477786e-05, + "loss": 1.6831, + "step": 1172 + }, + { + "epoch": 0.8113435932906796, + "grad_norm": 0.9235252737998962, + "learning_rate": 1.9956300072833213e-05, + "loss": 1.8274, + "step": 1173 + }, + { + "epoch": 0.812035275808404, + "grad_norm": 1.036197543144226, + "learning_rate": 1.9883466860888567e-05, + "loss": 1.68, + "step": 1174 + }, + { + "epoch": 0.8127269583261283, + "grad_norm": 0.9558598399162292, + "learning_rate": 1.9810633648943918e-05, + "loss": 1.6004, + "step": 1175 + }, + { + "epoch": 0.8134186408438526, + "grad_norm": 0.72776198387146, + "learning_rate": 1.9737800436999272e-05, + "loss": 1.3257, + "step": 1176 + }, + { + "epoch": 0.8141103233615771, + "grad_norm": 0.5898051261901855, + "learning_rate": 1.9664967225054627e-05, + "loss": 1.1236, + "step": 1177 + }, + { + "epoch": 0.8148020058793014, + "grad_norm": 0.8061662316322327, + "learning_rate": 1.9592134013109978e-05, + "loss": 1.4763, + "step": 1178 + }, + { + "epoch": 0.8154936883970257, + "grad_norm": 0.7124134302139282, + "learning_rate": 1.9519300801165332e-05, + "loss": 1.9014, + "step": 1179 + }, + { + "epoch": 0.8161853709147501, + "grad_norm": 0.767233669757843, + "learning_rate": 1.9446467589220686e-05, + "loss": 2.225, + "step": 1180 + }, + { + "epoch": 0.8168770534324745, + "grad_norm": 0.9079988598823547, + "learning_rate": 1.937363437727604e-05, + "loss": 1.5216, + "step": 1181 + }, + { + "epoch": 0.8175687359501989, + "grad_norm": 1.01505446434021, + "learning_rate": 1.930080116533139e-05, + "loss": 1.2661, + "step": 1182 + }, + { + "epoch": 0.8182604184679232, + "grad_norm": 0.7502846717834473, + "learning_rate": 1.9227967953386746e-05, + "loss": 1.6943, + "step": 1183 + }, + { + "epoch": 0.8189521009856476, + "grad_norm": 0.7551614046096802, + "learning_rate": 1.91551347414421e-05, + "loss": 1.273, + "step": 1184 + }, + { + "epoch": 0.819643783503372, + "grad_norm": 0.7072368860244751, + "learning_rate": 1.908230152949745e-05, + "loss": 1.4101, + "step": 1185 + }, + { + "epoch": 0.8203354660210963, + "grad_norm": 0.709496796131134, + "learning_rate": 1.9009468317552805e-05, + "loss": 1.0502, + "step": 1186 + }, + { + "epoch": 0.8210271485388206, + "grad_norm": 0.8042691349983215, + "learning_rate": 1.893663510560816e-05, + "loss": 1.5302, + "step": 1187 + }, + { + "epoch": 0.8217188310565451, + "grad_norm": 0.8842705488204956, + "learning_rate": 1.886380189366351e-05, + "loss": 1.8909, + "step": 1188 + }, + { + "epoch": 0.8224105135742694, + "grad_norm": 0.8364051580429077, + "learning_rate": 1.8790968681718865e-05, + "loss": 2.0175, + "step": 1189 + }, + { + "epoch": 0.8231021960919938, + "grad_norm": 0.7898297905921936, + "learning_rate": 1.871813546977422e-05, + "loss": 1.775, + "step": 1190 + }, + { + "epoch": 0.8237938786097181, + "grad_norm": 0.9381147623062134, + "learning_rate": 1.8645302257829573e-05, + "loss": 1.256, + "step": 1191 + }, + { + "epoch": 0.8244855611274425, + "grad_norm": 0.965323805809021, + "learning_rate": 1.8572469045884924e-05, + "loss": 1.3012, + "step": 1192 + }, + { + "epoch": 0.8251772436451669, + "grad_norm": 0.9088322520256042, + "learning_rate": 1.8499635833940275e-05, + "loss": 1.6961, + "step": 1193 + }, + { + "epoch": 0.8258689261628912, + "grad_norm": 0.6530706882476807, + "learning_rate": 1.8426802621995633e-05, + "loss": 1.202, + "step": 1194 + }, + { + "epoch": 0.8265606086806156, + "grad_norm": 1.110731601715088, + "learning_rate": 1.8353969410050983e-05, + "loss": 1.627, + "step": 1195 + }, + { + "epoch": 0.82725229119834, + "grad_norm": 0.9196153879165649, + "learning_rate": 1.8281136198106338e-05, + "loss": 1.6238, + "step": 1196 + }, + { + "epoch": 0.8279439737160643, + "grad_norm": 0.8053334951400757, + "learning_rate": 1.8208302986161692e-05, + "loss": 1.498, + "step": 1197 + }, + { + "epoch": 0.8286356562337887, + "grad_norm": 0.7577121257781982, + "learning_rate": 1.8135469774217043e-05, + "loss": 1.7706, + "step": 1198 + }, + { + "epoch": 0.8293273387515131, + "grad_norm": 0.9525589942932129, + "learning_rate": 1.8062636562272397e-05, + "loss": 0.9962, + "step": 1199 + }, + { + "epoch": 0.8300190212692374, + "grad_norm": 0.889854371547699, + "learning_rate": 1.7989803350327748e-05, + "loss": 1.8276, + "step": 1200 + }, + { + "epoch": 0.8307107037869618, + "grad_norm": 1.3769193887710571, + "learning_rate": 1.7916970138383106e-05, + "loss": 1.3189, + "step": 1201 + }, + { + "epoch": 0.8314023863046861, + "grad_norm": 0.955964207649231, + "learning_rate": 1.7844136926438457e-05, + "loss": 1.9168, + "step": 1202 + }, + { + "epoch": 0.8320940688224105, + "grad_norm": 0.6694774031639099, + "learning_rate": 1.7771303714493808e-05, + "loss": 1.367, + "step": 1203 + }, + { + "epoch": 0.8327857513401349, + "grad_norm": 0.9121772050857544, + "learning_rate": 1.7698470502549165e-05, + "loss": 1.5888, + "step": 1204 + }, + { + "epoch": 0.8334774338578592, + "grad_norm": 0.8669543862342834, + "learning_rate": 1.7625637290604516e-05, + "loss": 1.4607, + "step": 1205 + }, + { + "epoch": 0.8341691163755836, + "grad_norm": 0.9893675446510315, + "learning_rate": 1.755280407865987e-05, + "loss": 1.4227, + "step": 1206 + }, + { + "epoch": 0.834860798893308, + "grad_norm": 0.846275269985199, + "learning_rate": 1.747997086671522e-05, + "loss": 1.7157, + "step": 1207 + }, + { + "epoch": 0.8355524814110323, + "grad_norm": 0.7746529579162598, + "learning_rate": 1.7407137654770576e-05, + "loss": 1.5266, + "step": 1208 + }, + { + "epoch": 0.8362441639287567, + "grad_norm": 0.7803775072097778, + "learning_rate": 1.733430444282593e-05, + "loss": 1.8217, + "step": 1209 + }, + { + "epoch": 0.8369358464464811, + "grad_norm": 0.875978410243988, + "learning_rate": 1.726147123088128e-05, + "loss": 1.8471, + "step": 1210 + }, + { + "epoch": 0.8376275289642054, + "grad_norm": 1.1349574327468872, + "learning_rate": 1.718863801893664e-05, + "loss": 1.9108, + "step": 1211 + }, + { + "epoch": 0.8383192114819298, + "grad_norm": 0.9728168249130249, + "learning_rate": 1.711580480699199e-05, + "loss": 1.3895, + "step": 1212 + }, + { + "epoch": 0.8390108939996541, + "grad_norm": 0.8211981058120728, + "learning_rate": 1.704297159504734e-05, + "loss": 1.8612, + "step": 1213 + }, + { + "epoch": 0.8397025765173786, + "grad_norm": 0.7289793491363525, + "learning_rate": 1.6970138383102695e-05, + "loss": 1.3072, + "step": 1214 + }, + { + "epoch": 0.8403942590351029, + "grad_norm": 0.9708792567253113, + "learning_rate": 1.689730517115805e-05, + "loss": 1.9577, + "step": 1215 + }, + { + "epoch": 0.8410859415528272, + "grad_norm": 0.8946198225021362, + "learning_rate": 1.6824471959213403e-05, + "loss": 1.3665, + "step": 1216 + }, + { + "epoch": 0.8417776240705516, + "grad_norm": 1.0570414066314697, + "learning_rate": 1.6751638747268754e-05, + "loss": 1.0713, + "step": 1217 + }, + { + "epoch": 0.842469306588276, + "grad_norm": 1.1281256675720215, + "learning_rate": 1.667880553532411e-05, + "loss": 1.6448, + "step": 1218 + }, + { + "epoch": 0.8431609891060003, + "grad_norm": 1.0511953830718994, + "learning_rate": 1.6605972323379463e-05, + "loss": 1.4774, + "step": 1219 + }, + { + "epoch": 0.8438526716237247, + "grad_norm": 0.9247922301292419, + "learning_rate": 1.6533139111434814e-05, + "loss": 1.6477, + "step": 1220 + }, + { + "epoch": 0.8445443541414491, + "grad_norm": 0.7894918322563171, + "learning_rate": 1.6460305899490168e-05, + "loss": 1.583, + "step": 1221 + }, + { + "epoch": 0.8452360366591735, + "grad_norm": 0.916254997253418, + "learning_rate": 1.6387472687545522e-05, + "loss": 1.6072, + "step": 1222 + }, + { + "epoch": 0.8459277191768978, + "grad_norm": 2.19047474861145, + "learning_rate": 1.6314639475600873e-05, + "loss": 1.6593, + "step": 1223 + }, + { + "epoch": 0.8466194016946221, + "grad_norm": 1.136637568473816, + "learning_rate": 1.6241806263656227e-05, + "loss": 1.0327, + "step": 1224 + }, + { + "epoch": 0.8473110842123466, + "grad_norm": 0.7224776744842529, + "learning_rate": 1.616897305171158e-05, + "loss": 1.5504, + "step": 1225 + }, + { + "epoch": 0.8480027667300709, + "grad_norm": 1.0250519514083862, + "learning_rate": 1.6096139839766936e-05, + "loss": 1.5905, + "step": 1226 + }, + { + "epoch": 0.8486944492477952, + "grad_norm": 0.668803334236145, + "learning_rate": 1.6023306627822287e-05, + "loss": 1.4418, + "step": 1227 + }, + { + "epoch": 0.8493861317655196, + "grad_norm": 0.8995612263679504, + "learning_rate": 1.595047341587764e-05, + "loss": 1.7372, + "step": 1228 + }, + { + "epoch": 0.850077814283244, + "grad_norm": 0.706891655921936, + "learning_rate": 1.5877640203932995e-05, + "loss": 1.6232, + "step": 1229 + }, + { + "epoch": 0.8507694968009684, + "grad_norm": 0.965289831161499, + "learning_rate": 1.5804806991988346e-05, + "loss": 1.8065, + "step": 1230 + }, + { + "epoch": 0.8514611793186927, + "grad_norm": 0.8709790706634521, + "learning_rate": 1.57319737800437e-05, + "loss": 0.8363, + "step": 1231 + }, + { + "epoch": 0.852152861836417, + "grad_norm": 0.851180911064148, + "learning_rate": 1.5659140568099055e-05, + "loss": 1.9668, + "step": 1232 + }, + { + "epoch": 0.8528445443541415, + "grad_norm": 0.8270988464355469, + "learning_rate": 1.5586307356154406e-05, + "loss": 1.6041, + "step": 1233 + }, + { + "epoch": 0.8535362268718658, + "grad_norm": 0.8693057894706726, + "learning_rate": 1.551347414420976e-05, + "loss": 1.478, + "step": 1234 + }, + { + "epoch": 0.8542279093895901, + "grad_norm": 0.8406581878662109, + "learning_rate": 1.5440640932265114e-05, + "loss": 1.7629, + "step": 1235 + }, + { + "epoch": 0.8549195919073146, + "grad_norm": 0.7432942986488342, + "learning_rate": 1.536780772032047e-05, + "loss": 1.2746, + "step": 1236 + }, + { + "epoch": 0.8556112744250389, + "grad_norm": 1.00314199924469, + "learning_rate": 1.529497450837582e-05, + "loss": 1.8321, + "step": 1237 + }, + { + "epoch": 0.8563029569427633, + "grad_norm": 0.9806699156761169, + "learning_rate": 1.5222141296431172e-05, + "loss": 2.2027, + "step": 1238 + }, + { + "epoch": 0.8569946394604876, + "grad_norm": 0.9438709616661072, + "learning_rate": 1.5149308084486526e-05, + "loss": 1.8418, + "step": 1239 + }, + { + "epoch": 0.857686321978212, + "grad_norm": 0.8315860629081726, + "learning_rate": 1.5076474872541879e-05, + "loss": 1.8644, + "step": 1240 + }, + { + "epoch": 0.8583780044959364, + "grad_norm": 0.7097284197807312, + "learning_rate": 1.5003641660597233e-05, + "loss": 1.8832, + "step": 1241 + }, + { + "epoch": 0.8590696870136607, + "grad_norm": 0.7777730822563171, + "learning_rate": 1.4930808448652586e-05, + "loss": 2.185, + "step": 1242 + }, + { + "epoch": 0.859761369531385, + "grad_norm": 0.801007866859436, + "learning_rate": 1.4857975236707939e-05, + "loss": 1.4709, + "step": 1243 + }, + { + "epoch": 0.8604530520491095, + "grad_norm": 0.8698675036430359, + "learning_rate": 1.4785142024763293e-05, + "loss": 1.4306, + "step": 1244 + }, + { + "epoch": 0.8611447345668338, + "grad_norm": 0.5594910979270935, + "learning_rate": 1.4712308812818645e-05, + "loss": 1.0724, + "step": 1245 + }, + { + "epoch": 0.8618364170845582, + "grad_norm": 1.0255452394485474, + "learning_rate": 1.4639475600874e-05, + "loss": 2.0901, + "step": 1246 + }, + { + "epoch": 0.8625280996022826, + "grad_norm": 0.7796461582183838, + "learning_rate": 1.4566642388929352e-05, + "loss": 1.618, + "step": 1247 + }, + { + "epoch": 0.8632197821200069, + "grad_norm": 0.8455306887626648, + "learning_rate": 1.4493809176984705e-05, + "loss": 1.5911, + "step": 1248 + }, + { + "epoch": 0.8639114646377313, + "grad_norm": 0.8485004901885986, + "learning_rate": 1.442097596504006e-05, + "loss": 1.6083, + "step": 1249 + }, + { + "epoch": 0.8646031471554556, + "grad_norm": 0.8033997416496277, + "learning_rate": 1.4348142753095412e-05, + "loss": 1.831, + "step": 1250 + }, + { + "epoch": 0.86529482967318, + "grad_norm": 0.7714558839797974, + "learning_rate": 1.4275309541150766e-05, + "loss": 1.4212, + "step": 1251 + }, + { + "epoch": 0.8659865121909044, + "grad_norm": 0.8830355405807495, + "learning_rate": 1.4202476329206119e-05, + "loss": 1.9342, + "step": 1252 + }, + { + "epoch": 0.8666781947086287, + "grad_norm": 0.9375443458557129, + "learning_rate": 1.4129643117261471e-05, + "loss": 1.5649, + "step": 1253 + }, + { + "epoch": 0.8673698772263531, + "grad_norm": 0.6834514141082764, + "learning_rate": 1.4056809905316826e-05, + "loss": 1.4737, + "step": 1254 + }, + { + "epoch": 0.8680615597440775, + "grad_norm": 0.7813445925712585, + "learning_rate": 1.3983976693372178e-05, + "loss": 1.419, + "step": 1255 + }, + { + "epoch": 0.8687532422618018, + "grad_norm": 0.9051533341407776, + "learning_rate": 1.3911143481427532e-05, + "loss": 1.7336, + "step": 1256 + }, + { + "epoch": 0.8694449247795262, + "grad_norm": 0.7639695405960083, + "learning_rate": 1.3838310269482885e-05, + "loss": 1.568, + "step": 1257 + }, + { + "epoch": 0.8701366072972505, + "grad_norm": 0.7403379678726196, + "learning_rate": 1.3765477057538238e-05, + "loss": 1.5993, + "step": 1258 + }, + { + "epoch": 0.870828289814975, + "grad_norm": 0.9866239428520203, + "learning_rate": 1.3692643845593592e-05, + "loss": 1.3154, + "step": 1259 + }, + { + "epoch": 0.8715199723326993, + "grad_norm": 0.690859317779541, + "learning_rate": 1.3619810633648945e-05, + "loss": 1.2129, + "step": 1260 + }, + { + "epoch": 0.8722116548504236, + "grad_norm": 1.2022666931152344, + "learning_rate": 1.3546977421704299e-05, + "loss": 1.9456, + "step": 1261 + }, + { + "epoch": 0.8729033373681481, + "grad_norm": 0.7174875140190125, + "learning_rate": 1.3474144209759651e-05, + "loss": 1.5667, + "step": 1262 + }, + { + "epoch": 0.8735950198858724, + "grad_norm": 0.7700484991073608, + "learning_rate": 1.3401310997815002e-05, + "loss": 0.9548, + "step": 1263 + }, + { + "epoch": 0.8742867024035967, + "grad_norm": 1.03740394115448, + "learning_rate": 1.3328477785870358e-05, + "loss": 1.8943, + "step": 1264 + }, + { + "epoch": 0.8749783849213211, + "grad_norm": 1.2701075077056885, + "learning_rate": 1.3255644573925711e-05, + "loss": 1.8975, + "step": 1265 + }, + { + "epoch": 0.8756700674390455, + "grad_norm": 0.8156731128692627, + "learning_rate": 1.3182811361981065e-05, + "loss": 1.8794, + "step": 1266 + }, + { + "epoch": 0.8763617499567699, + "grad_norm": 0.6429268717765808, + "learning_rate": 1.3109978150036418e-05, + "loss": 1.0624, + "step": 1267 + }, + { + "epoch": 0.8770534324744942, + "grad_norm": 1.0503954887390137, + "learning_rate": 1.3037144938091769e-05, + "loss": 1.9622, + "step": 1268 + }, + { + "epoch": 0.8777451149922185, + "grad_norm": 1.1268541812896729, + "learning_rate": 1.2964311726147125e-05, + "loss": 1.941, + "step": 1269 + }, + { + "epoch": 0.878436797509943, + "grad_norm": 0.9450888633728027, + "learning_rate": 1.2891478514202476e-05, + "loss": 1.9332, + "step": 1270 + }, + { + "epoch": 0.8791284800276673, + "grad_norm": 0.8041183352470398, + "learning_rate": 1.2818645302257832e-05, + "loss": 0.9917, + "step": 1271 + }, + { + "epoch": 0.8798201625453916, + "grad_norm": 0.8963249325752258, + "learning_rate": 1.2745812090313184e-05, + "loss": 1.5991, + "step": 1272 + }, + { + "epoch": 0.8805118450631161, + "grad_norm": 0.8786289691925049, + "learning_rate": 1.2672978878368535e-05, + "loss": 1.4649, + "step": 1273 + }, + { + "epoch": 0.8812035275808404, + "grad_norm": 0.7974990606307983, + "learning_rate": 1.2600145666423891e-05, + "loss": 1.4704, + "step": 1274 + }, + { + "epoch": 0.8818952100985648, + "grad_norm": 0.8088487982749939, + "learning_rate": 1.2527312454479242e-05, + "loss": 1.3816, + "step": 1275 + }, + { + "epoch": 0.8825868926162891, + "grad_norm": 0.8868317604064941, + "learning_rate": 1.2454479242534596e-05, + "loss": 1.5239, + "step": 1276 + }, + { + "epoch": 0.8832785751340135, + "grad_norm": 0.8497899174690247, + "learning_rate": 1.2381646030589949e-05, + "loss": 1.4348, + "step": 1277 + }, + { + "epoch": 0.8839702576517379, + "grad_norm": 0.7080361843109131, + "learning_rate": 1.2308812818645303e-05, + "loss": 1.4534, + "step": 1278 + }, + { + "epoch": 0.8846619401694622, + "grad_norm": 0.8808167576789856, + "learning_rate": 1.2235979606700657e-05, + "loss": 1.6104, + "step": 1279 + }, + { + "epoch": 0.8853536226871865, + "grad_norm": 0.9049322009086609, + "learning_rate": 1.2163146394756008e-05, + "loss": 1.419, + "step": 1280 + }, + { + "epoch": 0.886045305204911, + "grad_norm": 0.8116534948348999, + "learning_rate": 1.2090313182811363e-05, + "loss": 1.021, + "step": 1281 + }, + { + "epoch": 0.8867369877226353, + "grad_norm": 0.9728810787200928, + "learning_rate": 1.2017479970866715e-05, + "loss": 1.6904, + "step": 1282 + }, + { + "epoch": 0.8874286702403597, + "grad_norm": 0.8273593783378601, + "learning_rate": 1.194464675892207e-05, + "loss": 1.1989, + "step": 1283 + }, + { + "epoch": 0.888120352758084, + "grad_norm": 0.8844332098960876, + "learning_rate": 1.1871813546977422e-05, + "loss": 1.6771, + "step": 1284 + }, + { + "epoch": 0.8888120352758084, + "grad_norm": 1.0074676275253296, + "learning_rate": 1.1798980335032775e-05, + "loss": 1.6171, + "step": 1285 + }, + { + "epoch": 0.8895037177935328, + "grad_norm": 0.8480923175811768, + "learning_rate": 1.1726147123088129e-05, + "loss": 1.9028, + "step": 1286 + }, + { + "epoch": 0.8901954003112571, + "grad_norm": 0.8386601209640503, + "learning_rate": 1.1653313911143482e-05, + "loss": 1.883, + "step": 1287 + }, + { + "epoch": 0.8908870828289815, + "grad_norm": 0.8075034022331238, + "learning_rate": 1.1580480699198836e-05, + "loss": 1.7361, + "step": 1288 + }, + { + "epoch": 0.8915787653467059, + "grad_norm": 0.9218458533287048, + "learning_rate": 1.1507647487254188e-05, + "loss": 1.9189, + "step": 1289 + }, + { + "epoch": 0.8922704478644302, + "grad_norm": 1.0705146789550781, + "learning_rate": 1.1434814275309541e-05, + "loss": 2.0755, + "step": 1290 + }, + { + "epoch": 0.8929621303821546, + "grad_norm": 1.3390390872955322, + "learning_rate": 1.1361981063364895e-05, + "loss": 1.5268, + "step": 1291 + }, + { + "epoch": 0.893653812899879, + "grad_norm": 0.770948052406311, + "learning_rate": 1.1289147851420248e-05, + "loss": 1.2664, + "step": 1292 + }, + { + "epoch": 0.8943454954176033, + "grad_norm": 1.244935393333435, + "learning_rate": 1.1216314639475602e-05, + "loss": 1.7128, + "step": 1293 + }, + { + "epoch": 0.8950371779353277, + "grad_norm": 0.7564862370491028, + "learning_rate": 1.1143481427530955e-05, + "loss": 1.7408, + "step": 1294 + }, + { + "epoch": 0.895728860453052, + "grad_norm": 0.8390173316001892, + "learning_rate": 1.1070648215586307e-05, + "loss": 1.2566, + "step": 1295 + }, + { + "epoch": 0.8964205429707764, + "grad_norm": 0.7898053526878357, + "learning_rate": 1.0997815003641662e-05, + "loss": 1.666, + "step": 1296 + }, + { + "epoch": 0.8971122254885008, + "grad_norm": 0.5722017884254456, + "learning_rate": 1.0924981791697014e-05, + "loss": 1.1444, + "step": 1297 + }, + { + "epoch": 0.8978039080062251, + "grad_norm": 0.9312048554420471, + "learning_rate": 1.0852148579752369e-05, + "loss": 0.811, + "step": 1298 + }, + { + "epoch": 0.8984955905239496, + "grad_norm": 1.001647710800171, + "learning_rate": 1.0779315367807721e-05, + "loss": 1.3638, + "step": 1299 + }, + { + "epoch": 0.8991872730416739, + "grad_norm": 0.9583906531333923, + "learning_rate": 1.0706482155863074e-05, + "loss": 1.6276, + "step": 1300 + }, + { + "epoch": 0.8998789555593982, + "grad_norm": 0.9671969413757324, + "learning_rate": 1.0633648943918426e-05, + "loss": 2.0144, + "step": 1301 + }, + { + "epoch": 0.9005706380771226, + "grad_norm": 0.9044082164764404, + "learning_rate": 1.056081573197378e-05, + "loss": 1.7702, + "step": 1302 + }, + { + "epoch": 0.901262320594847, + "grad_norm": 0.6864272356033325, + "learning_rate": 1.0487982520029133e-05, + "loss": 1.3465, + "step": 1303 + }, + { + "epoch": 0.9019540031125713, + "grad_norm": 0.7952218055725098, + "learning_rate": 1.0415149308084487e-05, + "loss": 1.8879, + "step": 1304 + }, + { + "epoch": 0.9026456856302957, + "grad_norm": 1.7466024160385132, + "learning_rate": 1.034231609613984e-05, + "loss": 1.5671, + "step": 1305 + }, + { + "epoch": 0.90333736814802, + "grad_norm": 1.093386173248291, + "learning_rate": 1.0269482884195193e-05, + "loss": 1.771, + "step": 1306 + }, + { + "epoch": 0.9040290506657445, + "grad_norm": 1.2970846891403198, + "learning_rate": 1.0196649672250547e-05, + "loss": 1.9003, + "step": 1307 + }, + { + "epoch": 0.9047207331834688, + "grad_norm": 0.8338832259178162, + "learning_rate": 1.01238164603059e-05, + "loss": 1.8831, + "step": 1308 + }, + { + "epoch": 0.9054124157011931, + "grad_norm": 1.1791330575942993, + "learning_rate": 1.0050983248361254e-05, + "loss": 2.0394, + "step": 1309 + }, + { + "epoch": 0.9061040982189175, + "grad_norm": 1.3057067394256592, + "learning_rate": 9.978150036416606e-06, + "loss": 1.752, + "step": 1310 + }, + { + "epoch": 0.9067957807366419, + "grad_norm": 0.78617924451828, + "learning_rate": 9.905316824471959e-06, + "loss": 1.5791, + "step": 1311 + }, + { + "epoch": 0.9074874632543662, + "grad_norm": 0.866249144077301, + "learning_rate": 9.832483612527313e-06, + "loss": 2.0739, + "step": 1312 + }, + { + "epoch": 0.9081791457720906, + "grad_norm": 0.936693549156189, + "learning_rate": 9.759650400582666e-06, + "loss": 1.6054, + "step": 1313 + }, + { + "epoch": 0.908870828289815, + "grad_norm": 0.7746766209602356, + "learning_rate": 9.68681718863802e-06, + "loss": 1.2682, + "step": 1314 + }, + { + "epoch": 0.9095625108075394, + "grad_norm": 0.8422493934631348, + "learning_rate": 9.613983976693373e-06, + "loss": 1.6796, + "step": 1315 + }, + { + "epoch": 0.9102541933252637, + "grad_norm": 0.9587863087654114, + "learning_rate": 9.541150764748725e-06, + "loss": 1.9267, + "step": 1316 + }, + { + "epoch": 0.910945875842988, + "grad_norm": 1.1121776103973389, + "learning_rate": 9.46831755280408e-06, + "loss": 0.9557, + "step": 1317 + }, + { + "epoch": 0.9116375583607125, + "grad_norm": 1.2082459926605225, + "learning_rate": 9.395484340859432e-06, + "loss": 1.0349, + "step": 1318 + }, + { + "epoch": 0.9123292408784368, + "grad_norm": 0.9303831458091736, + "learning_rate": 9.322651128914787e-06, + "loss": 1.0255, + "step": 1319 + }, + { + "epoch": 0.9130209233961611, + "grad_norm": 0.856451153755188, + "learning_rate": 9.249817916970137e-06, + "loss": 1.1031, + "step": 1320 + }, + { + "epoch": 0.9137126059138855, + "grad_norm": 0.9590927958488464, + "learning_rate": 9.176984705025492e-06, + "loss": 1.8872, + "step": 1321 + }, + { + "epoch": 0.9144042884316099, + "grad_norm": 0.8161438703536987, + "learning_rate": 9.104151493080846e-06, + "loss": 1.9037, + "step": 1322 + }, + { + "epoch": 0.9150959709493343, + "grad_norm": 1.196974754333496, + "learning_rate": 9.031318281136199e-06, + "loss": 1.2044, + "step": 1323 + }, + { + "epoch": 0.9157876534670586, + "grad_norm": 0.7898424863815308, + "learning_rate": 8.958485069191553e-06, + "loss": 1.725, + "step": 1324 + }, + { + "epoch": 0.916479335984783, + "grad_norm": 0.9460667967796326, + "learning_rate": 8.885651857246904e-06, + "loss": 2.0682, + "step": 1325 + }, + { + "epoch": 0.9171710185025074, + "grad_norm": 0.6884534955024719, + "learning_rate": 8.812818645302258e-06, + "loss": 1.3313, + "step": 1326 + }, + { + "epoch": 0.9178627010202317, + "grad_norm": 0.8076276183128357, + "learning_rate": 8.73998543335761e-06, + "loss": 1.1363, + "step": 1327 + }, + { + "epoch": 0.918554383537956, + "grad_norm": 0.8317755460739136, + "learning_rate": 8.667152221412965e-06, + "loss": 1.3865, + "step": 1328 + }, + { + "epoch": 0.9192460660556805, + "grad_norm": 0.7593029141426086, + "learning_rate": 8.59431900946832e-06, + "loss": 2.0115, + "step": 1329 + }, + { + "epoch": 0.9199377485734048, + "grad_norm": 1.1669124364852905, + "learning_rate": 8.52148579752367e-06, + "loss": 2.6219, + "step": 1330 + }, + { + "epoch": 0.9206294310911292, + "grad_norm": 0.7803002595901489, + "learning_rate": 8.448652585579024e-06, + "loss": 2.0944, + "step": 1331 + }, + { + "epoch": 0.9213211136088535, + "grad_norm": 1.0606745481491089, + "learning_rate": 8.375819373634377e-06, + "loss": 1.5527, + "step": 1332 + }, + { + "epoch": 0.9220127961265779, + "grad_norm": 0.9968279600143433, + "learning_rate": 8.302986161689731e-06, + "loss": 1.3545, + "step": 1333 + }, + { + "epoch": 0.9227044786443023, + "grad_norm": 0.9095673561096191, + "learning_rate": 8.230152949745084e-06, + "loss": 1.0772, + "step": 1334 + }, + { + "epoch": 0.9233961611620266, + "grad_norm": 0.7902427911758423, + "learning_rate": 8.157319737800437e-06, + "loss": 1.7249, + "step": 1335 + }, + { + "epoch": 0.9240878436797509, + "grad_norm": 0.9093700647354126, + "learning_rate": 8.08448652585579e-06, + "loss": 1.6361, + "step": 1336 + }, + { + "epoch": 0.9247795261974754, + "grad_norm": 0.7257758378982544, + "learning_rate": 8.011653313911143e-06, + "loss": 1.0136, + "step": 1337 + }, + { + "epoch": 0.9254712087151997, + "grad_norm": 1.0346702337265015, + "learning_rate": 7.938820101966498e-06, + "loss": 1.8582, + "step": 1338 + }, + { + "epoch": 0.9261628912329241, + "grad_norm": 0.661464512348175, + "learning_rate": 7.86598689002185e-06, + "loss": 1.1277, + "step": 1339 + }, + { + "epoch": 0.9268545737506485, + "grad_norm": 0.6260141730308533, + "learning_rate": 7.793153678077203e-06, + "loss": 1.2754, + "step": 1340 + }, + { + "epoch": 0.9275462562683728, + "grad_norm": 0.666012167930603, + "learning_rate": 7.720320466132557e-06, + "loss": 0.9829, + "step": 1341 + }, + { + "epoch": 0.9282379387860972, + "grad_norm": 0.9367914795875549, + "learning_rate": 7.64748725418791e-06, + "loss": 1.727, + "step": 1342 + }, + { + "epoch": 0.9289296213038215, + "grad_norm": 1.1356720924377441, + "learning_rate": 7.574654042243263e-06, + "loss": 1.7825, + "step": 1343 + }, + { + "epoch": 0.929621303821546, + "grad_norm": 0.9468403458595276, + "learning_rate": 7.501820830298617e-06, + "loss": 1.8361, + "step": 1344 + }, + { + "epoch": 0.9303129863392703, + "grad_norm": 1.14466392993927, + "learning_rate": 7.428987618353969e-06, + "loss": 1.5967, + "step": 1345 + }, + { + "epoch": 0.9310046688569946, + "grad_norm": 0.7425711750984192, + "learning_rate": 7.356154406409323e-06, + "loss": 1.1371, + "step": 1346 + }, + { + "epoch": 0.931696351374719, + "grad_norm": 0.971847414970398, + "learning_rate": 7.283321194464676e-06, + "loss": 2.007, + "step": 1347 + }, + { + "epoch": 0.9323880338924434, + "grad_norm": 0.7165243029594421, + "learning_rate": 7.21048798252003e-06, + "loss": 1.7195, + "step": 1348 + }, + { + "epoch": 0.9330797164101677, + "grad_norm": 0.9244675040245056, + "learning_rate": 7.137654770575383e-06, + "loss": 1.3981, + "step": 1349 + }, + { + "epoch": 0.9337713989278921, + "grad_norm": 0.7251322865486145, + "learning_rate": 7.064821558630736e-06, + "loss": 1.6066, + "step": 1350 + }, + { + "epoch": 0.9344630814456165, + "grad_norm": 1.3074641227722168, + "learning_rate": 6.991988346686089e-06, + "loss": 1.1456, + "step": 1351 + }, + { + "epoch": 0.9351547639633409, + "grad_norm": 0.9026491641998291, + "learning_rate": 6.9191551347414425e-06, + "loss": 1.4573, + "step": 1352 + }, + { + "epoch": 0.9358464464810652, + "grad_norm": 0.7197074294090271, + "learning_rate": 6.846321922796796e-06, + "loss": 1.6819, + "step": 1353 + }, + { + "epoch": 0.9365381289987895, + "grad_norm": 0.8270043134689331, + "learning_rate": 6.773488710852149e-06, + "loss": 1.2796, + "step": 1354 + }, + { + "epoch": 0.937229811516514, + "grad_norm": 0.8307384848594666, + "learning_rate": 6.700655498907501e-06, + "loss": 1.4652, + "step": 1355 + }, + { + "epoch": 0.9379214940342383, + "grad_norm": 0.5234942436218262, + "learning_rate": 6.6278222869628554e-06, + "loss": 0.3653, + "step": 1356 + }, + { + "epoch": 0.9386131765519626, + "grad_norm": 0.713591992855072, + "learning_rate": 6.554989075018209e-06, + "loss": 1.3439, + "step": 1357 + }, + { + "epoch": 0.939304859069687, + "grad_norm": 0.902199387550354, + "learning_rate": 6.482155863073562e-06, + "loss": 1.7389, + "step": 1358 + }, + { + "epoch": 0.9399965415874114, + "grad_norm": 0.8864997625350952, + "learning_rate": 6.409322651128916e-06, + "loss": 1.6, + "step": 1359 + }, + { + "epoch": 0.9406882241051358, + "grad_norm": 0.9748254418373108, + "learning_rate": 6.3364894391842675e-06, + "loss": 1.6671, + "step": 1360 + }, + { + "epoch": 0.9413799066228601, + "grad_norm": 0.8912494778633118, + "learning_rate": 6.263656227239621e-06, + "loss": 1.6931, + "step": 1361 + }, + { + "epoch": 0.9420715891405844, + "grad_norm": 0.9208559393882751, + "learning_rate": 6.190823015294974e-06, + "loss": 1.7124, + "step": 1362 + }, + { + "epoch": 0.9427632716583089, + "grad_norm": 1.513280987739563, + "learning_rate": 6.117989803350329e-06, + "loss": 1.5457, + "step": 1363 + }, + { + "epoch": 0.9434549541760332, + "grad_norm": 0.8193607926368713, + "learning_rate": 6.045156591405681e-06, + "loss": 1.6974, + "step": 1364 + }, + { + "epoch": 0.9441466366937575, + "grad_norm": 0.9019060730934143, + "learning_rate": 5.972323379461035e-06, + "loss": 1.3058, + "step": 1365 + }, + { + "epoch": 0.944838319211482, + "grad_norm": 0.9550915360450745, + "learning_rate": 5.899490167516387e-06, + "loss": 1.8829, + "step": 1366 + }, + { + "epoch": 0.9455300017292063, + "grad_norm": 1.0974806547164917, + "learning_rate": 5.826656955571741e-06, + "loss": 1.5233, + "step": 1367 + }, + { + "epoch": 0.9462216842469306, + "grad_norm": 1.074586272239685, + "learning_rate": 5.753823743627094e-06, + "loss": 1.4755, + "step": 1368 + }, + { + "epoch": 0.946913366764655, + "grad_norm": 0.9504185318946838, + "learning_rate": 5.680990531682448e-06, + "loss": 1.9579, + "step": 1369 + }, + { + "epoch": 0.9476050492823794, + "grad_norm": 0.8342894911766052, + "learning_rate": 5.608157319737801e-06, + "loss": 1.2366, + "step": 1370 + }, + { + "epoch": 0.9482967318001038, + "grad_norm": 0.8063657879829407, + "learning_rate": 5.535324107793154e-06, + "loss": 1.9314, + "step": 1371 + }, + { + "epoch": 0.9489884143178281, + "grad_norm": 0.9030585885047913, + "learning_rate": 5.462490895848507e-06, + "loss": 1.4967, + "step": 1372 + }, + { + "epoch": 0.9496800968355524, + "grad_norm": 0.7804280519485474, + "learning_rate": 5.3896576839038606e-06, + "loss": 0.9585, + "step": 1373 + }, + { + "epoch": 0.9503717793532769, + "grad_norm": 0.9998471140861511, + "learning_rate": 5.316824471959213e-06, + "loss": 1.4133, + "step": 1374 + }, + { + "epoch": 0.9510634618710012, + "grad_norm": 0.8258745074272156, + "learning_rate": 5.243991260014567e-06, + "loss": 1.3937, + "step": 1375 + }, + { + "epoch": 0.9517551443887255, + "grad_norm": 0.9427676200866699, + "learning_rate": 5.17115804806992e-06, + "loss": 2.1021, + "step": 1376 + }, + { + "epoch": 0.95244682690645, + "grad_norm": 0.8124197125434875, + "learning_rate": 5.0983248361252735e-06, + "loss": 1.7599, + "step": 1377 + }, + { + "epoch": 0.9531385094241743, + "grad_norm": 0.8654883503913879, + "learning_rate": 5.025491624180627e-06, + "loss": 1.7338, + "step": 1378 + }, + { + "epoch": 0.9538301919418987, + "grad_norm": 0.7667081356048584, + "learning_rate": 4.9526584122359795e-06, + "loss": 2.0584, + "step": 1379 + }, + { + "epoch": 0.954521874459623, + "grad_norm": 0.9698280096054077, + "learning_rate": 4.879825200291333e-06, + "loss": 1.39, + "step": 1380 + }, + { + "epoch": 0.9552135569773474, + "grad_norm": 0.8826647996902466, + "learning_rate": 4.806991988346686e-06, + "loss": 1.6265, + "step": 1381 + }, + { + "epoch": 0.9559052394950718, + "grad_norm": 0.9939996600151062, + "learning_rate": 4.73415877640204e-06, + "loss": 1.4502, + "step": 1382 + }, + { + "epoch": 0.9565969220127961, + "grad_norm": 0.6261147856712341, + "learning_rate": 4.661325564457393e-06, + "loss": 0.4957, + "step": 1383 + }, + { + "epoch": 0.9572886045305204, + "grad_norm": 0.8702592253684998, + "learning_rate": 4.588492352512746e-06, + "loss": 1.7885, + "step": 1384 + }, + { + "epoch": 0.9579802870482449, + "grad_norm": 0.9081373810768127, + "learning_rate": 4.515659140568099e-06, + "loss": 1.8912, + "step": 1385 + }, + { + "epoch": 0.9586719695659692, + "grad_norm": 0.6326776146888733, + "learning_rate": 4.442825928623452e-06, + "loss": 1.2118, + "step": 1386 + }, + { + "epoch": 0.9593636520836936, + "grad_norm": 0.8432955145835876, + "learning_rate": 4.369992716678805e-06, + "loss": 1.6837, + "step": 1387 + }, + { + "epoch": 0.9600553346014179, + "grad_norm": 0.9558162093162537, + "learning_rate": 4.29715950473416e-06, + "loss": 1.8122, + "step": 1388 + }, + { + "epoch": 0.9607470171191423, + "grad_norm": 0.7647939324378967, + "learning_rate": 4.224326292789512e-06, + "loss": 1.994, + "step": 1389 + }, + { + "epoch": 0.9614386996368667, + "grad_norm": 1.095019817352295, + "learning_rate": 4.151493080844866e-06, + "loss": 2.1668, + "step": 1390 + }, + { + "epoch": 0.962130382154591, + "grad_norm": 1.0153563022613525, + "learning_rate": 4.078659868900218e-06, + "loss": 2.0185, + "step": 1391 + }, + { + "epoch": 0.9628220646723155, + "grad_norm": 0.712698221206665, + "learning_rate": 4.005826656955572e-06, + "loss": 1.4266, + "step": 1392 + }, + { + "epoch": 0.9635137471900398, + "grad_norm": 0.7006120085716248, + "learning_rate": 3.932993445010925e-06, + "loss": 1.1372, + "step": 1393 + }, + { + "epoch": 0.9642054297077641, + "grad_norm": 1.0687493085861206, + "learning_rate": 3.860160233066279e-06, + "loss": 2.0922, + "step": 1394 + }, + { + "epoch": 0.9648971122254885, + "grad_norm": 0.9177038669586182, + "learning_rate": 3.7873270211216316e-06, + "loss": 1.9107, + "step": 1395 + }, + { + "epoch": 0.9655887947432129, + "grad_norm": 0.9137991666793823, + "learning_rate": 3.7144938091769846e-06, + "loss": 1.0229, + "step": 1396 + }, + { + "epoch": 0.9662804772609372, + "grad_norm": 0.9297300577163696, + "learning_rate": 3.641660597232338e-06, + "loss": 2.1492, + "step": 1397 + }, + { + "epoch": 0.9669721597786616, + "grad_norm": 0.7305806279182434, + "learning_rate": 3.5688273852876915e-06, + "loss": 1.3146, + "step": 1398 + }, + { + "epoch": 0.9676638422963859, + "grad_norm": 0.7297166585922241, + "learning_rate": 3.4959941733430445e-06, + "loss": 1.1816, + "step": 1399 + }, + { + "epoch": 0.9683555248141104, + "grad_norm": 0.9093372225761414, + "learning_rate": 3.423160961398398e-06, + "loss": 2.0061, + "step": 1400 + }, + { + "epoch": 0.9690472073318347, + "grad_norm": 0.9646827578544617, + "learning_rate": 3.3503277494537506e-06, + "loss": 1.6416, + "step": 1401 + }, + { + "epoch": 0.969738889849559, + "grad_norm": 0.94194096326828, + "learning_rate": 3.2774945375091044e-06, + "loss": 1.2457, + "step": 1402 + }, + { + "epoch": 0.9704305723672834, + "grad_norm": 1.0533934831619263, + "learning_rate": 3.204661325564458e-06, + "loss": 1.7572, + "step": 1403 + }, + { + "epoch": 0.9711222548850078, + "grad_norm": 0.7030104398727417, + "learning_rate": 3.1318281136198105e-06, + "loss": 1.4678, + "step": 1404 + }, + { + "epoch": 0.9718139374027321, + "grad_norm": 0.7260177135467529, + "learning_rate": 3.0589949016751643e-06, + "loss": 1.5277, + "step": 1405 + }, + { + "epoch": 0.9725056199204565, + "grad_norm": 0.8522548675537109, + "learning_rate": 2.9861616897305174e-06, + "loss": 2.0796, + "step": 1406 + }, + { + "epoch": 0.9731973024381809, + "grad_norm": 0.7276301383972168, + "learning_rate": 2.9133284777858704e-06, + "loss": 0.8843, + "step": 1407 + }, + { + "epoch": 0.9738889849559053, + "grad_norm": 0.8044816851615906, + "learning_rate": 2.840495265841224e-06, + "loss": 1.8257, + "step": 1408 + }, + { + "epoch": 0.9745806674736296, + "grad_norm": 0.9322279095649719, + "learning_rate": 2.767662053896577e-06, + "loss": 1.7733, + "step": 1409 + }, + { + "epoch": 0.9752723499913539, + "grad_norm": 0.7970879673957825, + "learning_rate": 2.6948288419519303e-06, + "loss": 1.6663, + "step": 1410 + }, + { + "epoch": 0.9759640325090784, + "grad_norm": 0.8964112401008606, + "learning_rate": 2.6219956300072833e-06, + "loss": 1.3942, + "step": 1411 + }, + { + "epoch": 0.9766557150268027, + "grad_norm": 0.7014464735984802, + "learning_rate": 2.5491624180626367e-06, + "loss": 1.7089, + "step": 1412 + }, + { + "epoch": 0.977347397544527, + "grad_norm": 0.8625869154930115, + "learning_rate": 2.4763292061179898e-06, + "loss": 0.9857, + "step": 1413 + }, + { + "epoch": 0.9780390800622514, + "grad_norm": 0.7947872877120972, + "learning_rate": 2.403495994173343e-06, + "loss": 2.0671, + "step": 1414 + }, + { + "epoch": 0.9787307625799758, + "grad_norm": 0.8284644484519958, + "learning_rate": 2.3306627822286966e-06, + "loss": 1.6046, + "step": 1415 + }, + { + "epoch": 0.9794224450977002, + "grad_norm": 0.6849550008773804, + "learning_rate": 2.2578295702840497e-06, + "loss": 1.0991, + "step": 1416 + }, + { + "epoch": 0.9801141276154245, + "grad_norm": 1.1793265342712402, + "learning_rate": 2.1849963583394027e-06, + "loss": 1.6424, + "step": 1417 + }, + { + "epoch": 0.9808058101331489, + "grad_norm": 0.8303595781326294, + "learning_rate": 2.112163146394756e-06, + "loss": 1.8512, + "step": 1418 + }, + { + "epoch": 0.9814974926508733, + "grad_norm": 0.9628428816795349, + "learning_rate": 2.039329934450109e-06, + "loss": 1.7375, + "step": 1419 + }, + { + "epoch": 0.9821891751685976, + "grad_norm": 1.1076195240020752, + "learning_rate": 1.9664967225054626e-06, + "loss": 1.3424, + "step": 1420 + }, + { + "epoch": 0.9828808576863219, + "grad_norm": 1.2490711212158203, + "learning_rate": 1.8936635105608158e-06, + "loss": 1.6675, + "step": 1421 + }, + { + "epoch": 0.9835725402040464, + "grad_norm": 0.9343705773353577, + "learning_rate": 1.820830298616169e-06, + "loss": 1.5664, + "step": 1422 + }, + { + "epoch": 0.9842642227217707, + "grad_norm": 0.9502846002578735, + "learning_rate": 1.7479970866715223e-06, + "loss": 1.7591, + "step": 1423 + }, + { + "epoch": 0.984955905239495, + "grad_norm": 0.7205687165260315, + "learning_rate": 1.6751638747268753e-06, + "loss": 1.2958, + "step": 1424 + }, + { + "epoch": 0.9856475877572194, + "grad_norm": 0.9856358170509338, + "learning_rate": 1.602330662782229e-06, + "loss": 1.7782, + "step": 1425 + }, + { + "epoch": 0.9863392702749438, + "grad_norm": 0.7743490934371948, + "learning_rate": 1.5294974508375822e-06, + "loss": 1.7658, + "step": 1426 + }, + { + "epoch": 0.9870309527926682, + "grad_norm": 1.2497276067733765, + "learning_rate": 1.4566642388929352e-06, + "loss": 1.1109, + "step": 1427 + }, + { + "epoch": 0.9877226353103925, + "grad_norm": 0.6770827174186707, + "learning_rate": 1.3838310269482884e-06, + "loss": 1.3271, + "step": 1428 + }, + { + "epoch": 0.9884143178281168, + "grad_norm": 0.857930064201355, + "learning_rate": 1.3109978150036416e-06, + "loss": 1.5296, + "step": 1429 + }, + { + "epoch": 0.9891060003458413, + "grad_norm": 0.7909390330314636, + "learning_rate": 1.2381646030589949e-06, + "loss": 1.5466, + "step": 1430 + }, + { + "epoch": 0.9897976828635656, + "grad_norm": 0.7727596759796143, + "learning_rate": 1.1653313911143483e-06, + "loss": 1.3115, + "step": 1431 + }, + { + "epoch": 0.99048936538129, + "grad_norm": 0.8144175410270691, + "learning_rate": 1.0924981791697013e-06, + "loss": 1.7048, + "step": 1432 + }, + { + "epoch": 0.9911810478990144, + "grad_norm": 0.767493724822998, + "learning_rate": 1.0196649672250546e-06, + "loss": 1.162, + "step": 1433 + }, + { + "epoch": 0.9918727304167387, + "grad_norm": 1.894450306892395, + "learning_rate": 9.468317552804079e-07, + "loss": 1.9863, + "step": 1434 + }, + { + "epoch": 0.9925644129344631, + "grad_norm": 0.9905784130096436, + "learning_rate": 8.739985433357611e-07, + "loss": 2.1525, + "step": 1435 + }, + { + "epoch": 0.9932560954521874, + "grad_norm": 0.9424154758453369, + "learning_rate": 8.011653313911145e-07, + "loss": 1.9599, + "step": 1436 + }, + { + "epoch": 0.9939477779699118, + "grad_norm": 0.7488706111907959, + "learning_rate": 7.283321194464676e-07, + "loss": 1.7235, + "step": 1437 + }, + { + "epoch": 0.9946394604876362, + "grad_norm": 1.0373190641403198, + "learning_rate": 6.554989075018208e-07, + "loss": 1.7826, + "step": 1438 + }, + { + "epoch": 0.9953311430053605, + "grad_norm": 0.8698166012763977, + "learning_rate": 5.826656955571742e-07, + "loss": 2.011, + "step": 1439 + }, + { + "epoch": 0.9960228255230849, + "grad_norm": 0.7344587445259094, + "learning_rate": 5.098324836125273e-07, + "loss": 1.1814, + "step": 1440 + }, + { + "epoch": 0.9967145080408093, + "grad_norm": 1.0096945762634277, + "learning_rate": 4.3699927166788057e-07, + "loss": 1.504, + "step": 1441 + }, + { + "epoch": 0.9974061905585336, + "grad_norm": 0.8708990216255188, + "learning_rate": 3.641660597232338e-07, + "loss": 2.2336, + "step": 1442 + }, + { + "epoch": 0.998097873076258, + "grad_norm": 0.9140763282775879, + "learning_rate": 2.913328477785871e-07, + "loss": 1.9224, + "step": 1443 + }, + { + "epoch": 0.9987895555939824, + "grad_norm": 0.7608925104141235, + "learning_rate": 2.1849963583394028e-07, + "loss": 1.5217, + "step": 1444 + }, + { + "epoch": 0.9994812381117067, + "grad_norm": 0.6970881819725037, + "learning_rate": 1.4566642388929354e-07, + "loss": 1.9148, + "step": 1445 + }, + { + "epoch": 1.0, + "grad_norm": 0.7492889761924744, + "learning_rate": 7.283321194464677e-08, + "loss": 1.5782, + "step": 1446 + }, + { + "epoch": 1.0006916825177243, + "grad_norm": 0.863744854927063, + "learning_rate": 0.0, + "loss": 1.4985, + "step": 1447 + }, + { + "epoch": 1.0013833650354487, + "grad_norm": 0.9803378582000732, + "learning_rate": 2.1025043680838673e-05, + "loss": 1.8497, + "step": 1448 + }, + { + "epoch": 1.002075047553173, + "grad_norm": 1.0503430366516113, + "learning_rate": 2.0966802562609203e-05, + "loss": 0.8252, + "step": 1449 + }, + { + "epoch": 1.0027667300708976, + "grad_norm": 0.9179481863975525, + "learning_rate": 2.090856144437973e-05, + "loss": 0.871, + "step": 1450 + }, + { + "epoch": 1.003458412588622, + "grad_norm": 0.756814181804657, + "learning_rate": 2.0850320326150264e-05, + "loss": 1.8423, + "step": 1451 + }, + { + "epoch": 1.0041500951063462, + "grad_norm": 0.8344918489456177, + "learning_rate": 2.079207920792079e-05, + "loss": 1.0372, + "step": 1452 + }, + { + "epoch": 1.0048417776240706, + "grad_norm": 0.8150106072425842, + "learning_rate": 2.0733838089691325e-05, + "loss": 1.3791, + "step": 1453 + }, + { + "epoch": 1.005533460141795, + "grad_norm": 0.973647952079773, + "learning_rate": 2.067559697146185e-05, + "loss": 1.7459, + "step": 1454 + }, + { + "epoch": 1.0062251426595192, + "grad_norm": 1.036309838294983, + "learning_rate": 2.0617355853232385e-05, + "loss": 1.1696, + "step": 1455 + }, + { + "epoch": 1.0069168251772436, + "grad_norm": 0.7406970262527466, + "learning_rate": 2.0559114735002912e-05, + "loss": 1.8236, + "step": 1456 + }, + { + "epoch": 1.007608507694968, + "grad_norm": 0.7419610619544983, + "learning_rate": 2.0500873616773442e-05, + "loss": 1.7502, + "step": 1457 + }, + { + "epoch": 1.0083001902126925, + "grad_norm": 0.9418008327484131, + "learning_rate": 2.0442632498543973e-05, + "loss": 1.7512, + "step": 1458 + }, + { + "epoch": 1.0089918727304168, + "grad_norm": 1.4829647541046143, + "learning_rate": 2.0384391380314503e-05, + "loss": 1.6802, + "step": 1459 + }, + { + "epoch": 1.0096835552481411, + "grad_norm": 0.7008660435676575, + "learning_rate": 2.0326150262085033e-05, + "loss": 1.0512, + "step": 1460 + }, + { + "epoch": 1.0103752377658655, + "grad_norm": 1.1753841638565063, + "learning_rate": 2.0267909143855564e-05, + "loss": 1.1687, + "step": 1461 + }, + { + "epoch": 1.0110669202835898, + "grad_norm": 0.8660651445388794, + "learning_rate": 2.0209668025626094e-05, + "loss": 1.4298, + "step": 1462 + }, + { + "epoch": 1.0117586028013141, + "grad_norm": 0.876896321773529, + "learning_rate": 2.0151426907396624e-05, + "loss": 1.5521, + "step": 1463 + }, + { + "epoch": 1.0124502853190385, + "grad_norm": 0.870000958442688, + "learning_rate": 2.0093185789167155e-05, + "loss": 1.2283, + "step": 1464 + }, + { + "epoch": 1.013141967836763, + "grad_norm": 1.0520581007003784, + "learning_rate": 2.0034944670937685e-05, + "loss": 1.5771, + "step": 1465 + }, + { + "epoch": 1.0138336503544874, + "grad_norm": 0.9469524621963501, + "learning_rate": 1.9976703552708212e-05, + "loss": 1.9631, + "step": 1466 + }, + { + "epoch": 1.0145253328722117, + "grad_norm": 0.8725298643112183, + "learning_rate": 1.9918462434478742e-05, + "loss": 1.0068, + "step": 1467 + }, + { + "epoch": 1.015217015389936, + "grad_norm": 0.9582999348640442, + "learning_rate": 1.9860221316249272e-05, + "loss": 1.5826, + "step": 1468 + }, + { + "epoch": 1.0159086979076604, + "grad_norm": 0.7669278979301453, + "learning_rate": 1.9801980198019803e-05, + "loss": 1.4152, + "step": 1469 + }, + { + "epoch": 1.0166003804253847, + "grad_norm": 0.8162096738815308, + "learning_rate": 1.9743739079790333e-05, + "loss": 1.0294, + "step": 1470 + }, + { + "epoch": 1.017292062943109, + "grad_norm": 0.8054408431053162, + "learning_rate": 1.9685497961560863e-05, + "loss": 1.3641, + "step": 1471 + }, + { + "epoch": 1.0179837454608336, + "grad_norm": 0.6561583876609802, + "learning_rate": 1.9627256843331394e-05, + "loss": 1.0095, + "step": 1472 + }, + { + "epoch": 1.018675427978558, + "grad_norm": 1.0789417028427124, + "learning_rate": 1.956901572510192e-05, + "loss": 1.949, + "step": 1473 + }, + { + "epoch": 1.0193671104962823, + "grad_norm": 0.8590020537376404, + "learning_rate": 1.9510774606872454e-05, + "loss": 1.63, + "step": 1474 + }, + { + "epoch": 1.0200587930140066, + "grad_norm": 2.1856565475463867, + "learning_rate": 1.945253348864298e-05, + "loss": 1.6432, + "step": 1475 + }, + { + "epoch": 1.020750475531731, + "grad_norm": 0.8429316878318787, + "learning_rate": 1.9394292370413515e-05, + "loss": 2.0053, + "step": 1476 + }, + { + "epoch": 1.0214421580494553, + "grad_norm": 0.6636189222335815, + "learning_rate": 1.933605125218404e-05, + "loss": 1.3913, + "step": 1477 + }, + { + "epoch": 1.0221338405671796, + "grad_norm": 0.7353069186210632, + "learning_rate": 1.9277810133954575e-05, + "loss": 1.6338, + "step": 1478 + }, + { + "epoch": 1.022825523084904, + "grad_norm": 0.9759920239448547, + "learning_rate": 1.9219569015725102e-05, + "loss": 1.7212, + "step": 1479 + }, + { + "epoch": 1.0235172056026285, + "grad_norm": 0.6762087345123291, + "learning_rate": 1.9161327897495633e-05, + "loss": 1.7112, + "step": 1480 + }, + { + "epoch": 1.0242088881203528, + "grad_norm": 0.6899464726448059, + "learning_rate": 1.9103086779266163e-05, + "loss": 1.1365, + "step": 1481 + }, + { + "epoch": 1.0249005706380772, + "grad_norm": 0.634777843952179, + "learning_rate": 1.9044845661036693e-05, + "loss": 1.3237, + "step": 1482 + }, + { + "epoch": 1.0255922531558015, + "grad_norm": 0.8964893817901611, + "learning_rate": 1.8986604542807223e-05, + "loss": 1.5688, + "step": 1483 + }, + { + "epoch": 1.0262839356735258, + "grad_norm": 0.7786408066749573, + "learning_rate": 1.8928363424577754e-05, + "loss": 1.741, + "step": 1484 + }, + { + "epoch": 1.0269756181912502, + "grad_norm": 0.7621763944625854, + "learning_rate": 1.8870122306348284e-05, + "loss": 2.0862, + "step": 1485 + }, + { + "epoch": 1.0276673007089745, + "grad_norm": 1.3445504903793335, + "learning_rate": 1.8811881188118814e-05, + "loss": 1.4772, + "step": 1486 + }, + { + "epoch": 1.028358983226699, + "grad_norm": 0.8518914580345154, + "learning_rate": 1.875364006988934e-05, + "loss": 1.6635, + "step": 1487 + }, + { + "epoch": 1.0290506657444234, + "grad_norm": 0.7836101651191711, + "learning_rate": 1.869539895165987e-05, + "loss": 1.2657, + "step": 1488 + }, + { + "epoch": 1.0297423482621477, + "grad_norm": 0.8802850842475891, + "learning_rate": 1.8637157833430402e-05, + "loss": 1.9376, + "step": 1489 + }, + { + "epoch": 1.030434030779872, + "grad_norm": 0.6951125264167786, + "learning_rate": 1.8578916715200932e-05, + "loss": 0.8851, + "step": 1490 + }, + { + "epoch": 1.0311257132975964, + "grad_norm": 0.8488371968269348, + "learning_rate": 1.8520675596971462e-05, + "loss": 2.0302, + "step": 1491 + }, + { + "epoch": 1.0318173958153207, + "grad_norm": 0.8759697079658508, + "learning_rate": 1.8462434478741993e-05, + "loss": 2.0752, + "step": 1492 + }, + { + "epoch": 1.032509078333045, + "grad_norm": 0.8570151329040527, + "learning_rate": 1.8404193360512523e-05, + "loss": 1.8307, + "step": 1493 + }, + { + "epoch": 1.0332007608507694, + "grad_norm": 0.8472738265991211, + "learning_rate": 1.834595224228305e-05, + "loss": 1.2439, + "step": 1494 + }, + { + "epoch": 1.033892443368494, + "grad_norm": 0.7950757741928101, + "learning_rate": 1.8287711124053584e-05, + "loss": 1.8829, + "step": 1495 + }, + { + "epoch": 1.0345841258862183, + "grad_norm": 1.00944983959198, + "learning_rate": 1.822947000582411e-05, + "loss": 1.9019, + "step": 1496 + }, + { + "epoch": 1.0352758084039426, + "grad_norm": 0.9507315754890442, + "learning_rate": 1.8171228887594644e-05, + "loss": 1.5145, + "step": 1497 + }, + { + "epoch": 1.035967490921667, + "grad_norm": 0.9261516332626343, + "learning_rate": 1.811298776936517e-05, + "loss": 1.5145, + "step": 1498 + }, + { + "epoch": 1.0366591734393913, + "grad_norm": 0.7590171694755554, + "learning_rate": 1.8054746651135705e-05, + "loss": 0.8552, + "step": 1499 + }, + { + "epoch": 1.0373508559571156, + "grad_norm": 1.0986384153366089, + "learning_rate": 1.7996505532906232e-05, + "loss": 1.7503, + "step": 1500 + }, + { + "epoch": 1.0373508559571156, + "eval_loss": 1.4565975666046143, + "eval_runtime": 563.6011, + "eval_samples_per_second": 2.28, + "eval_steps_per_second": 1.141, + "step": 1500 + }, + { + "epoch": 1.03804253847484, + "grad_norm": 0.673653781414032, + "learning_rate": 1.7938264414676762e-05, + "loss": 1.1308, + "step": 1501 + }, + { + "epoch": 1.0387342209925645, + "grad_norm": 0.9013132452964783, + "learning_rate": 1.7880023296447292e-05, + "loss": 2.1572, + "step": 1502 + }, + { + "epoch": 1.0394259035102889, + "grad_norm": 0.6884844899177551, + "learning_rate": 1.7821782178217823e-05, + "loss": 1.4862, + "step": 1503 + }, + { + "epoch": 1.0401175860280132, + "grad_norm": 0.7120798230171204, + "learning_rate": 1.7763541059988353e-05, + "loss": 1.641, + "step": 1504 + }, + { + "epoch": 1.0408092685457375, + "grad_norm": 0.8523894548416138, + "learning_rate": 1.7705299941758883e-05, + "loss": 1.4092, + "step": 1505 + }, + { + "epoch": 1.0415009510634619, + "grad_norm": 0.8717172741889954, + "learning_rate": 1.7647058823529414e-05, + "loss": 1.1293, + "step": 1506 + }, + { + "epoch": 1.0421926335811862, + "grad_norm": 0.8149329423904419, + "learning_rate": 1.7588817705299944e-05, + "loss": 1.1035, + "step": 1507 + }, + { + "epoch": 1.0428843160989105, + "grad_norm": 0.9907032251358032, + "learning_rate": 1.753057658707047e-05, + "loss": 1.9464, + "step": 1508 + }, + { + "epoch": 1.0435759986166349, + "grad_norm": 1.0136642456054688, + "learning_rate": 1.7472335468841005e-05, + "loss": 1.1966, + "step": 1509 + }, + { + "epoch": 1.0442676811343594, + "grad_norm": 0.7494779825210571, + "learning_rate": 1.741409435061153e-05, + "loss": 1.8103, + "step": 1510 + }, + { + "epoch": 1.0449593636520838, + "grad_norm": 0.9566720128059387, + "learning_rate": 1.7355853232382062e-05, + "loss": 1.6682, + "step": 1511 + }, + { + "epoch": 1.045651046169808, + "grad_norm": 0.9102514386177063, + "learning_rate": 1.7297612114152592e-05, + "loss": 2.0524, + "step": 1512 + }, + { + "epoch": 1.0463427286875324, + "grad_norm": 0.7713015675544739, + "learning_rate": 1.7239370995923122e-05, + "loss": 1.8451, + "step": 1513 + }, + { + "epoch": 1.0470344112052568, + "grad_norm": 0.7905024886131287, + "learning_rate": 1.7181129877693653e-05, + "loss": 1.3569, + "step": 1514 + }, + { + "epoch": 1.047726093722981, + "grad_norm": 0.8789911866188049, + "learning_rate": 1.7122888759464183e-05, + "loss": 1.5025, + "step": 1515 + }, + { + "epoch": 1.0484177762407054, + "grad_norm": 1.6176868677139282, + "learning_rate": 1.7064647641234713e-05, + "loss": 1.3858, + "step": 1516 + }, + { + "epoch": 1.04910945875843, + "grad_norm": 0.8238159418106079, + "learning_rate": 1.700640652300524e-05, + "loss": 1.1574, + "step": 1517 + }, + { + "epoch": 1.0498011412761543, + "grad_norm": 1.1071362495422363, + "learning_rate": 1.6948165404775774e-05, + "loss": 2.107, + "step": 1518 + }, + { + "epoch": 1.0504928237938786, + "grad_norm": 0.6677478551864624, + "learning_rate": 1.68899242865463e-05, + "loss": 1.0157, + "step": 1519 + }, + { + "epoch": 1.051184506311603, + "grad_norm": 1.5846837759017944, + "learning_rate": 1.6831683168316834e-05, + "loss": 1.4526, + "step": 1520 + }, + { + "epoch": 1.0518761888293273, + "grad_norm": 0.8592979907989502, + "learning_rate": 1.677344205008736e-05, + "loss": 1.5912, + "step": 1521 + }, + { + "epoch": 1.0525678713470517, + "grad_norm": 1.162194848060608, + "learning_rate": 1.671520093185789e-05, + "loss": 1.7149, + "step": 1522 + }, + { + "epoch": 1.053259553864776, + "grad_norm": 0.8566716909408569, + "learning_rate": 1.6656959813628422e-05, + "loss": 1.693, + "step": 1523 + }, + { + "epoch": 1.0539512363825003, + "grad_norm": 0.846211850643158, + "learning_rate": 1.6598718695398952e-05, + "loss": 2.0591, + "step": 1524 + }, + { + "epoch": 1.0546429189002249, + "grad_norm": 1.1176985502243042, + "learning_rate": 1.6540477577169483e-05, + "loss": 1.5019, + "step": 1525 + }, + { + "epoch": 1.0553346014179492, + "grad_norm": 0.8028376698493958, + "learning_rate": 1.6482236458940013e-05, + "loss": 1.8108, + "step": 1526 + }, + { + "epoch": 1.0560262839356735, + "grad_norm": 0.9885690212249756, + "learning_rate": 1.6423995340710543e-05, + "loss": 2.0553, + "step": 1527 + }, + { + "epoch": 1.0567179664533979, + "grad_norm": 1.017757534980774, + "learning_rate": 1.6365754222481073e-05, + "loss": 1.2021, + "step": 1528 + }, + { + "epoch": 1.0574096489711222, + "grad_norm": 0.7753478288650513, + "learning_rate": 1.63075131042516e-05, + "loss": 1.1547, + "step": 1529 + }, + { + "epoch": 1.0581013314888466, + "grad_norm": 0.8402890563011169, + "learning_rate": 1.6249271986022134e-05, + "loss": 1.623, + "step": 1530 + }, + { + "epoch": 1.0587930140065709, + "grad_norm": 0.7326244115829468, + "learning_rate": 1.619103086779266e-05, + "loss": 1.4355, + "step": 1531 + }, + { + "epoch": 1.0594846965242954, + "grad_norm": 0.8238964676856995, + "learning_rate": 1.6132789749563195e-05, + "loss": 1.4867, + "step": 1532 + }, + { + "epoch": 1.0601763790420198, + "grad_norm": 0.8625259399414062, + "learning_rate": 1.607454863133372e-05, + "loss": 1.9233, + "step": 1533 + }, + { + "epoch": 1.0608680615597441, + "grad_norm": 1.1380106210708618, + "learning_rate": 1.6016307513104252e-05, + "loss": 1.4181, + "step": 1534 + }, + { + "epoch": 1.0615597440774684, + "grad_norm": 0.9015654921531677, + "learning_rate": 1.5958066394874782e-05, + "loss": 1.768, + "step": 1535 + }, + { + "epoch": 1.0622514265951928, + "grad_norm": 0.9468401670455933, + "learning_rate": 1.5899825276645312e-05, + "loss": 1.9714, + "step": 1536 + }, + { + "epoch": 1.0629431091129171, + "grad_norm": 0.7965930700302124, + "learning_rate": 1.5841584158415843e-05, + "loss": 1.7377, + "step": 1537 + }, + { + "epoch": 1.0636347916306415, + "grad_norm": 0.9804380536079407, + "learning_rate": 1.578334304018637e-05, + "loss": 2.0444, + "step": 1538 + }, + { + "epoch": 1.0643264741483658, + "grad_norm": 0.6031072735786438, + "learning_rate": 1.5725101921956903e-05, + "loss": 1.6853, + "step": 1539 + }, + { + "epoch": 1.0650181566660903, + "grad_norm": 0.9756177663803101, + "learning_rate": 1.566686080372743e-05, + "loss": 1.7688, + "step": 1540 + }, + { + "epoch": 1.0657098391838147, + "grad_norm": 0.7981095314025879, + "learning_rate": 1.5608619685497964e-05, + "loss": 1.6502, + "step": 1541 + }, + { + "epoch": 1.066401521701539, + "grad_norm": 0.988097608089447, + "learning_rate": 1.555037856726849e-05, + "loss": 1.0343, + "step": 1542 + }, + { + "epoch": 1.0670932042192633, + "grad_norm": 0.7080538868904114, + "learning_rate": 1.5492137449039025e-05, + "loss": 1.1162, + "step": 1543 + }, + { + "epoch": 1.0677848867369877, + "grad_norm": 0.8531929850578308, + "learning_rate": 1.543389633080955e-05, + "loss": 1.8998, + "step": 1544 + }, + { + "epoch": 1.068476569254712, + "grad_norm": 0.8813366293907166, + "learning_rate": 1.5375655212580082e-05, + "loss": 1.0619, + "step": 1545 + }, + { + "epoch": 1.0691682517724364, + "grad_norm": 0.862993061542511, + "learning_rate": 1.5317414094350612e-05, + "loss": 1.7948, + "step": 1546 + }, + { + "epoch": 1.069859934290161, + "grad_norm": 0.7223910689353943, + "learning_rate": 1.5259172976121142e-05, + "loss": 1.303, + "step": 1547 + }, + { + "epoch": 1.0705516168078852, + "grad_norm": 0.8272300958633423, + "learning_rate": 1.5200931857891673e-05, + "loss": 1.5962, + "step": 1548 + }, + { + "epoch": 1.0712432993256096, + "grad_norm": 0.9725604057312012, + "learning_rate": 1.5142690739662201e-05, + "loss": 1.6826, + "step": 1549 + }, + { + "epoch": 1.071934981843334, + "grad_norm": 0.8255423903465271, + "learning_rate": 1.5084449621432733e-05, + "loss": 2.0776, + "step": 1550 + }, + { + "epoch": 1.0726266643610582, + "grad_norm": 0.7376166582107544, + "learning_rate": 1.5026208503203262e-05, + "loss": 0.912, + "step": 1551 + }, + { + "epoch": 1.0733183468787826, + "grad_norm": 0.8306238055229187, + "learning_rate": 1.496796738497379e-05, + "loss": 1.3008, + "step": 1552 + }, + { + "epoch": 1.074010029396507, + "grad_norm": 0.7455337047576904, + "learning_rate": 1.4909726266744323e-05, + "loss": 1.8351, + "step": 1553 + }, + { + "epoch": 1.0747017119142313, + "grad_norm": 0.8191561698913574, + "learning_rate": 1.4851485148514851e-05, + "loss": 1.9757, + "step": 1554 + }, + { + "epoch": 1.0753933944319558, + "grad_norm": 0.7372782230377197, + "learning_rate": 1.4793244030285383e-05, + "loss": 1.9232, + "step": 1555 + }, + { + "epoch": 1.0760850769496801, + "grad_norm": 0.7686094045639038, + "learning_rate": 1.4735002912055912e-05, + "loss": 1.2343, + "step": 1556 + }, + { + "epoch": 1.0767767594674045, + "grad_norm": 0.751352071762085, + "learning_rate": 1.4676761793826444e-05, + "loss": 1.245, + "step": 1557 + }, + { + "epoch": 1.0774684419851288, + "grad_norm": 1.0581755638122559, + "learning_rate": 1.4618520675596972e-05, + "loss": 1.8101, + "step": 1558 + }, + { + "epoch": 1.0781601245028531, + "grad_norm": 0.8349416851997375, + "learning_rate": 1.4560279557367501e-05, + "loss": 1.773, + "step": 1559 + }, + { + "epoch": 1.0788518070205775, + "grad_norm": 1.1016970872879028, + "learning_rate": 1.4502038439138033e-05, + "loss": 1.4782, + "step": 1560 + }, + { + "epoch": 1.079543489538302, + "grad_norm": 0.7285605669021606, + "learning_rate": 1.4443797320908562e-05, + "loss": 1.6937, + "step": 1561 + }, + { + "epoch": 1.0802351720560264, + "grad_norm": 0.7619128823280334, + "learning_rate": 1.4385556202679094e-05, + "loss": 1.0654, + "step": 1562 + }, + { + "epoch": 1.0809268545737507, + "grad_norm": 0.9540428519248962, + "learning_rate": 1.4327315084449622e-05, + "loss": 1.3818, + "step": 1563 + }, + { + "epoch": 1.081618537091475, + "grad_norm": 0.8722973465919495, + "learning_rate": 1.4269073966220154e-05, + "loss": 1.8706, + "step": 1564 + }, + { + "epoch": 1.0823102196091994, + "grad_norm": 0.6846266388893127, + "learning_rate": 1.4210832847990683e-05, + "loss": 0.9266, + "step": 1565 + }, + { + "epoch": 1.0830019021269237, + "grad_norm": 1.6851447820663452, + "learning_rate": 1.4152591729761211e-05, + "loss": 1.5706, + "step": 1566 + }, + { + "epoch": 1.083693584644648, + "grad_norm": 0.7721676230430603, + "learning_rate": 1.4094350611531743e-05, + "loss": 1.8559, + "step": 1567 + }, + { + "epoch": 1.0843852671623724, + "grad_norm": 0.7722626328468323, + "learning_rate": 1.4036109493302272e-05, + "loss": 1.4473, + "step": 1568 + }, + { + "epoch": 1.085076949680097, + "grad_norm": 0.9184356927871704, + "learning_rate": 1.3977868375072802e-05, + "loss": 1.5392, + "step": 1569 + }, + { + "epoch": 1.0857686321978213, + "grad_norm": 0.9129490852355957, + "learning_rate": 1.3919627256843331e-05, + "loss": 1.7303, + "step": 1570 + }, + { + "epoch": 1.0864603147155456, + "grad_norm": 0.8343762159347534, + "learning_rate": 1.3861386138613863e-05, + "loss": 1.1604, + "step": 1571 + }, + { + "epoch": 1.08715199723327, + "grad_norm": 0.908096432685852, + "learning_rate": 1.3803145020384391e-05, + "loss": 1.8513, + "step": 1572 + }, + { + "epoch": 1.0878436797509943, + "grad_norm": 0.9411665201187134, + "learning_rate": 1.374490390215492e-05, + "loss": 1.4829, + "step": 1573 + }, + { + "epoch": 1.0885353622687186, + "grad_norm": 1.0297383069992065, + "learning_rate": 1.3686662783925452e-05, + "loss": 1.465, + "step": 1574 + }, + { + "epoch": 1.089227044786443, + "grad_norm": 0.7959553003311157, + "learning_rate": 1.362842166569598e-05, + "loss": 1.0267, + "step": 1575 + }, + { + "epoch": 1.0899187273041675, + "grad_norm": 0.8858935236930847, + "learning_rate": 1.3570180547466513e-05, + "loss": 1.5978, + "step": 1576 + }, + { + "epoch": 1.0906104098218918, + "grad_norm": 0.6778844594955444, + "learning_rate": 1.3511939429237041e-05, + "loss": 1.1873, + "step": 1577 + }, + { + "epoch": 1.0913020923396162, + "grad_norm": 0.7167158722877502, + "learning_rate": 1.3453698311007573e-05, + "loss": 1.7835, + "step": 1578 + }, + { + "epoch": 1.0919937748573405, + "grad_norm": 0.9060688614845276, + "learning_rate": 1.3395457192778102e-05, + "loss": 1.8769, + "step": 1579 + }, + { + "epoch": 1.0926854573750648, + "grad_norm": 0.6919119358062744, + "learning_rate": 1.333721607454863e-05, + "loss": 1.0191, + "step": 1580 + }, + { + "epoch": 1.0933771398927892, + "grad_norm": 0.8415911793708801, + "learning_rate": 1.3278974956319162e-05, + "loss": 1.3884, + "step": 1581 + }, + { + "epoch": 1.0940688224105135, + "grad_norm": 0.899627685546875, + "learning_rate": 1.3220733838089691e-05, + "loss": 1.8216, + "step": 1582 + }, + { + "epoch": 1.0947605049282378, + "grad_norm": 0.9176965951919556, + "learning_rate": 1.3162492719860223e-05, + "loss": 2.0036, + "step": 1583 + }, + { + "epoch": 1.0954521874459624, + "grad_norm": 0.8543194532394409, + "learning_rate": 1.3104251601630752e-05, + "loss": 1.794, + "step": 1584 + }, + { + "epoch": 1.0961438699636867, + "grad_norm": 0.9835048913955688, + "learning_rate": 1.3046010483401284e-05, + "loss": 1.123, + "step": 1585 + }, + { + "epoch": 1.096835552481411, + "grad_norm": 0.8706053495407104, + "learning_rate": 1.2987769365171812e-05, + "loss": 1.5749, + "step": 1586 + }, + { + "epoch": 1.0975272349991354, + "grad_norm": 1.199723243713379, + "learning_rate": 1.2929528246942341e-05, + "loss": 1.5023, + "step": 1587 + }, + { + "epoch": 1.0982189175168597, + "grad_norm": 0.809876024723053, + "learning_rate": 1.2871287128712873e-05, + "loss": 1.6761, + "step": 1588 + }, + { + "epoch": 1.098910600034584, + "grad_norm": 1.0492621660232544, + "learning_rate": 1.2813046010483402e-05, + "loss": 1.9406, + "step": 1589 + }, + { + "epoch": 1.0996022825523084, + "grad_norm": 0.8627383708953857, + "learning_rate": 1.2754804892253932e-05, + "loss": 0.9306, + "step": 1590 + }, + { + "epoch": 1.100293965070033, + "grad_norm": 1.2382993698120117, + "learning_rate": 1.2696563774024462e-05, + "loss": 1.875, + "step": 1591 + }, + { + "epoch": 1.1009856475877573, + "grad_norm": 0.5542892217636108, + "learning_rate": 1.2638322655794992e-05, + "loss": 0.9045, + "step": 1592 + }, + { + "epoch": 1.1016773301054816, + "grad_norm": 1.0427334308624268, + "learning_rate": 1.2580081537565521e-05, + "loss": 1.2303, + "step": 1593 + }, + { + "epoch": 1.102369012623206, + "grad_norm": 0.8489208817481995, + "learning_rate": 1.2521840419336051e-05, + "loss": 1.7907, + "step": 1594 + }, + { + "epoch": 1.1030606951409303, + "grad_norm": 1.0355641841888428, + "learning_rate": 1.2463599301106582e-05, + "loss": 0.8085, + "step": 1595 + }, + { + "epoch": 1.1037523776586546, + "grad_norm": 0.7226613759994507, + "learning_rate": 1.2405358182877112e-05, + "loss": 1.5537, + "step": 1596 + }, + { + "epoch": 1.104444060176379, + "grad_norm": 1.0447916984558105, + "learning_rate": 1.2347117064647642e-05, + "loss": 1.8019, + "step": 1597 + }, + { + "epoch": 1.1051357426941033, + "grad_norm": 0.8970907330513, + "learning_rate": 1.228887594641817e-05, + "loss": 1.9532, + "step": 1598 + }, + { + "epoch": 1.1058274252118279, + "grad_norm": 0.8990969061851501, + "learning_rate": 1.2230634828188701e-05, + "loss": 2.0669, + "step": 1599 + }, + { + "epoch": 1.1065191077295522, + "grad_norm": 0.7818430662155151, + "learning_rate": 1.2172393709959231e-05, + "loss": 1.7287, + "step": 1600 + }, + { + "epoch": 1.1072107902472765, + "grad_norm": 0.9390760064125061, + "learning_rate": 1.2114152591729762e-05, + "loss": 0.707, + "step": 1601 + }, + { + "epoch": 1.1079024727650009, + "grad_norm": 0.8881179094314575, + "learning_rate": 1.2055911473500292e-05, + "loss": 1.9263, + "step": 1602 + }, + { + "epoch": 1.1085941552827252, + "grad_norm": 0.7376513481140137, + "learning_rate": 1.1997670355270822e-05, + "loss": 1.6354, + "step": 1603 + }, + { + "epoch": 1.1092858378004495, + "grad_norm": 0.7851370573043823, + "learning_rate": 1.1939429237041353e-05, + "loss": 1.8352, + "step": 1604 + }, + { + "epoch": 1.1099775203181739, + "grad_norm": 1.1918026208877563, + "learning_rate": 1.1881188118811881e-05, + "loss": 1.5637, + "step": 1605 + }, + { + "epoch": 1.1106692028358984, + "grad_norm": 0.9654248952865601, + "learning_rate": 1.1822947000582412e-05, + "loss": 1.9586, + "step": 1606 + }, + { + "epoch": 1.1113608853536228, + "grad_norm": 1.2490615844726562, + "learning_rate": 1.1764705882352942e-05, + "loss": 1.1162, + "step": 1607 + }, + { + "epoch": 1.112052567871347, + "grad_norm": 0.702061116695404, + "learning_rate": 1.1706464764123472e-05, + "loss": 1.3764, + "step": 1608 + }, + { + "epoch": 1.1127442503890714, + "grad_norm": 0.8088762760162354, + "learning_rate": 1.1648223645894002e-05, + "loss": 1.8542, + "step": 1609 + }, + { + "epoch": 1.1134359329067958, + "grad_norm": 0.9740604758262634, + "learning_rate": 1.1589982527664533e-05, + "loss": 1.8837, + "step": 1610 + }, + { + "epoch": 1.11412761542452, + "grad_norm": 0.8064848184585571, + "learning_rate": 1.1531741409435063e-05, + "loss": 1.4857, + "step": 1611 + }, + { + "epoch": 1.1148192979422444, + "grad_norm": 0.7576624155044556, + "learning_rate": 1.1473500291205592e-05, + "loss": 1.8967, + "step": 1612 + }, + { + "epoch": 1.1155109804599688, + "grad_norm": 1.0941357612609863, + "learning_rate": 1.1415259172976122e-05, + "loss": 2.0201, + "step": 1613 + }, + { + "epoch": 1.1162026629776933, + "grad_norm": 0.5964834094047546, + "learning_rate": 1.1357018054746652e-05, + "loss": 0.9747, + "step": 1614 + }, + { + "epoch": 1.1168943454954177, + "grad_norm": 0.9126596450805664, + "learning_rate": 1.1298776936517181e-05, + "loss": 1.7304, + "step": 1615 + }, + { + "epoch": 1.117586028013142, + "grad_norm": 0.9090709686279297, + "learning_rate": 1.1240535818287711e-05, + "loss": 1.7617, + "step": 1616 + }, + { + "epoch": 1.1182777105308663, + "grad_norm": 1.1052805185317993, + "learning_rate": 1.1182294700058241e-05, + "loss": 1.5669, + "step": 1617 + }, + { + "epoch": 1.1189693930485907, + "grad_norm": 0.7590151429176331, + "learning_rate": 1.1124053581828772e-05, + "loss": 1.5984, + "step": 1618 + }, + { + "epoch": 1.119661075566315, + "grad_norm": 0.9704794883728027, + "learning_rate": 1.10658124635993e-05, + "loss": 1.8408, + "step": 1619 + }, + { + "epoch": 1.1203527580840393, + "grad_norm": 1.1911158561706543, + "learning_rate": 1.100757134536983e-05, + "loss": 0.9104, + "step": 1620 + }, + { + "epoch": 1.121044440601764, + "grad_norm": 0.9407263398170471, + "learning_rate": 1.0949330227140361e-05, + "loss": 1.5961, + "step": 1621 + }, + { + "epoch": 1.1217361231194882, + "grad_norm": 0.8614974021911621, + "learning_rate": 1.0891089108910891e-05, + "loss": 1.2822, + "step": 1622 + }, + { + "epoch": 1.1224278056372126, + "grad_norm": 0.7882705330848694, + "learning_rate": 1.0832847990681422e-05, + "loss": 1.7405, + "step": 1623 + }, + { + "epoch": 1.123119488154937, + "grad_norm": 1.5602566003799438, + "learning_rate": 1.0774606872451952e-05, + "loss": 1.4524, + "step": 1624 + }, + { + "epoch": 1.1238111706726612, + "grad_norm": 0.9316334128379822, + "learning_rate": 1.0716365754222482e-05, + "loss": 1.3574, + "step": 1625 + }, + { + "epoch": 1.1245028531903856, + "grad_norm": 1.1098911762237549, + "learning_rate": 1.065812463599301e-05, + "loss": 1.5294, + "step": 1626 + }, + { + "epoch": 1.12519453570811, + "grad_norm": 0.8112794160842896, + "learning_rate": 1.0599883517763541e-05, + "loss": 1.9297, + "step": 1627 + }, + { + "epoch": 1.1258862182258342, + "grad_norm": 0.8933188915252686, + "learning_rate": 1.0541642399534071e-05, + "loss": 1.68, + "step": 1628 + }, + { + "epoch": 1.1265779007435588, + "grad_norm": 0.9110146164894104, + "learning_rate": 1.0483401281304602e-05, + "loss": 2.2353, + "step": 1629 + }, + { + "epoch": 1.1272695832612831, + "grad_norm": 0.8418118953704834, + "learning_rate": 1.0425160163075132e-05, + "loss": 1.6014, + "step": 1630 + }, + { + "epoch": 1.1279612657790075, + "grad_norm": 0.9635080695152283, + "learning_rate": 1.0366919044845662e-05, + "loss": 1.8902, + "step": 1631 + }, + { + "epoch": 1.1286529482967318, + "grad_norm": 1.1626524925231934, + "learning_rate": 1.0308677926616193e-05, + "loss": 1.7483, + "step": 1632 + }, + { + "epoch": 1.1293446308144561, + "grad_norm": 0.8520684242248535, + "learning_rate": 1.0250436808386721e-05, + "loss": 1.7917, + "step": 1633 + }, + { + "epoch": 1.1300363133321805, + "grad_norm": 0.7874677181243896, + "learning_rate": 1.0192195690157252e-05, + "loss": 1.4263, + "step": 1634 + }, + { + "epoch": 1.1307279958499048, + "grad_norm": 0.8110877871513367, + "learning_rate": 1.0133954571927782e-05, + "loss": 1.7277, + "step": 1635 + }, + { + "epoch": 1.1314196783676294, + "grad_norm": 1.2249890565872192, + "learning_rate": 1.0075713453698312e-05, + "loss": 1.8784, + "step": 1636 + }, + { + "epoch": 1.1321113608853537, + "grad_norm": 1.0384804010391235, + "learning_rate": 1.0017472335468842e-05, + "loss": 2.0303, + "step": 1637 + }, + { + "epoch": 1.132803043403078, + "grad_norm": 0.9578065276145935, + "learning_rate": 9.959231217239371e-06, + "loss": 1.5147, + "step": 1638 + }, + { + "epoch": 1.1334947259208024, + "grad_norm": 0.9212397336959839, + "learning_rate": 9.900990099009901e-06, + "loss": 1.5085, + "step": 1639 + }, + { + "epoch": 1.1341864084385267, + "grad_norm": 1.0539535284042358, + "learning_rate": 9.842748980780432e-06, + "loss": 1.3367, + "step": 1640 + }, + { + "epoch": 1.134878090956251, + "grad_norm": 0.7783110737800598, + "learning_rate": 9.78450786255096e-06, + "loss": 1.1612, + "step": 1641 + }, + { + "epoch": 1.1355697734739754, + "grad_norm": 0.8694387674331665, + "learning_rate": 9.72626674432149e-06, + "loss": 1.6028, + "step": 1642 + }, + { + "epoch": 1.1362614559916997, + "grad_norm": 1.076819658279419, + "learning_rate": 9.66802562609202e-06, + "loss": 1.1279, + "step": 1643 + }, + { + "epoch": 1.1369531385094243, + "grad_norm": 1.5457978248596191, + "learning_rate": 9.609784507862551e-06, + "loss": 1.7353, + "step": 1644 + }, + { + "epoch": 1.1376448210271486, + "grad_norm": 0.8007596135139465, + "learning_rate": 9.551543389633081e-06, + "loss": 1.6712, + "step": 1645 + }, + { + "epoch": 1.138336503544873, + "grad_norm": 0.8414595127105713, + "learning_rate": 9.493302271403612e-06, + "loss": 1.735, + "step": 1646 + }, + { + "epoch": 1.1390281860625973, + "grad_norm": 0.817112386226654, + "learning_rate": 9.435061153174142e-06, + "loss": 1.9924, + "step": 1647 + }, + { + "epoch": 1.1397198685803216, + "grad_norm": 0.8054618239402771, + "learning_rate": 9.37682003494467e-06, + "loss": 1.8575, + "step": 1648 + }, + { + "epoch": 1.140411551098046, + "grad_norm": 0.8680825233459473, + "learning_rate": 9.318578916715201e-06, + "loss": 2.1419, + "step": 1649 + }, + { + "epoch": 1.1411032336157705, + "grad_norm": 0.8252211809158325, + "learning_rate": 9.260337798485731e-06, + "loss": 1.7489, + "step": 1650 + }, + { + "epoch": 1.1417949161334948, + "grad_norm": 0.6045031547546387, + "learning_rate": 9.202096680256262e-06, + "loss": 0.9681, + "step": 1651 + }, + { + "epoch": 1.1424865986512192, + "grad_norm": 1.1018991470336914, + "learning_rate": 9.143855562026792e-06, + "loss": 1.8272, + "step": 1652 + }, + { + "epoch": 1.1431782811689435, + "grad_norm": 0.9452030062675476, + "learning_rate": 9.085614443797322e-06, + "loss": 1.6236, + "step": 1653 + }, + { + "epoch": 1.1438699636866678, + "grad_norm": 1.0537296533584595, + "learning_rate": 9.027373325567852e-06, + "loss": 1.0604, + "step": 1654 + }, + { + "epoch": 1.1445616462043922, + "grad_norm": 0.7857002019882202, + "learning_rate": 8.969132207338381e-06, + "loss": 1.1005, + "step": 1655 + }, + { + "epoch": 1.1452533287221165, + "grad_norm": 0.8211665749549866, + "learning_rate": 8.910891089108911e-06, + "loss": 1.6964, + "step": 1656 + }, + { + "epoch": 1.1459450112398408, + "grad_norm": 1.3052046298980713, + "learning_rate": 8.852649970879442e-06, + "loss": 1.6691, + "step": 1657 + }, + { + "epoch": 1.1466366937575652, + "grad_norm": 0.8832458853721619, + "learning_rate": 8.794408852649972e-06, + "loss": 1.8245, + "step": 1658 + }, + { + "epoch": 1.1473283762752897, + "grad_norm": 0.6758939027786255, + "learning_rate": 8.736167734420502e-06, + "loss": 0.8366, + "step": 1659 + }, + { + "epoch": 1.148020058793014, + "grad_norm": 0.8165250420570374, + "learning_rate": 8.677926616191031e-06, + "loss": 1.2328, + "step": 1660 + }, + { + "epoch": 1.1487117413107384, + "grad_norm": 0.9432193636894226, + "learning_rate": 8.619685497961561e-06, + "loss": 2.1603, + "step": 1661 + }, + { + "epoch": 1.1494034238284627, + "grad_norm": 0.8237228393554688, + "learning_rate": 8.561444379732091e-06, + "loss": 1.6826, + "step": 1662 + }, + { + "epoch": 1.150095106346187, + "grad_norm": 0.7904053926467896, + "learning_rate": 8.50320326150262e-06, + "loss": 1.9237, + "step": 1663 + }, + { + "epoch": 1.1507867888639114, + "grad_norm": 0.7902259826660156, + "learning_rate": 8.44496214327315e-06, + "loss": 1.9479, + "step": 1664 + }, + { + "epoch": 1.151478471381636, + "grad_norm": 0.7349541783332825, + "learning_rate": 8.38672102504368e-06, + "loss": 1.0345, + "step": 1665 + }, + { + "epoch": 1.1521701538993603, + "grad_norm": 0.9517996907234192, + "learning_rate": 8.328479906814211e-06, + "loss": 1.871, + "step": 1666 + }, + { + "epoch": 1.1528618364170846, + "grad_norm": 0.7846324443817139, + "learning_rate": 8.270238788584741e-06, + "loss": 2.0113, + "step": 1667 + }, + { + "epoch": 1.153553518934809, + "grad_norm": 0.7118953466415405, + "learning_rate": 8.211997670355272e-06, + "loss": 1.0855, + "step": 1668 + }, + { + "epoch": 1.1542452014525333, + "grad_norm": 0.8612886667251587, + "learning_rate": 8.1537565521258e-06, + "loss": 1.3707, + "step": 1669 + }, + { + "epoch": 1.1549368839702576, + "grad_norm": 0.8379912972450256, + "learning_rate": 8.09551543389633e-06, + "loss": 1.8068, + "step": 1670 + }, + { + "epoch": 1.155628566487982, + "grad_norm": 0.7577520608901978, + "learning_rate": 8.03727431566686e-06, + "loss": 0.9441, + "step": 1671 + }, + { + "epoch": 1.1563202490057063, + "grad_norm": 0.8887487053871155, + "learning_rate": 7.979033197437391e-06, + "loss": 1.8997, + "step": 1672 + }, + { + "epoch": 1.1570119315234306, + "grad_norm": 0.9183244109153748, + "learning_rate": 7.920792079207921e-06, + "loss": 1.2895, + "step": 1673 + }, + { + "epoch": 1.1577036140411552, + "grad_norm": 0.7634314894676208, + "learning_rate": 7.862550960978452e-06, + "loss": 1.1826, + "step": 1674 + }, + { + "epoch": 1.1583952965588795, + "grad_norm": 1.25795316696167, + "learning_rate": 7.804309842748982e-06, + "loss": 1.8899, + "step": 1675 + }, + { + "epoch": 1.1590869790766039, + "grad_norm": 1.096569299697876, + "learning_rate": 7.746068724519512e-06, + "loss": 1.8991, + "step": 1676 + }, + { + "epoch": 1.1597786615943282, + "grad_norm": 0.5702176094055176, + "learning_rate": 7.687827606290041e-06, + "loss": 0.8812, + "step": 1677 + }, + { + "epoch": 1.1604703441120525, + "grad_norm": 0.8095030188560486, + "learning_rate": 7.629586488060571e-06, + "loss": 1.9199, + "step": 1678 + }, + { + "epoch": 1.1611620266297769, + "grad_norm": 1.7219574451446533, + "learning_rate": 7.571345369831101e-06, + "loss": 1.5043, + "step": 1679 + }, + { + "epoch": 1.1618537091475014, + "grad_norm": 0.9227262735366821, + "learning_rate": 7.513104251601631e-06, + "loss": 1.3494, + "step": 1680 + }, + { + "epoch": 1.1625453916652257, + "grad_norm": 0.7733169794082642, + "learning_rate": 7.454863133372161e-06, + "loss": 0.9736, + "step": 1681 + }, + { + "epoch": 1.16323707418295, + "grad_norm": 0.7111944556236267, + "learning_rate": 7.3966220151426916e-06, + "loss": 1.7746, + "step": 1682 + }, + { + "epoch": 1.1639287567006744, + "grad_norm": 0.9316900372505188, + "learning_rate": 7.338380896913222e-06, + "loss": 1.4266, + "step": 1683 + }, + { + "epoch": 1.1646204392183988, + "grad_norm": 0.7173492908477783, + "learning_rate": 7.2801397786837505e-06, + "loss": 1.4059, + "step": 1684 + }, + { + "epoch": 1.165312121736123, + "grad_norm": 0.812737762928009, + "learning_rate": 7.221898660454281e-06, + "loss": 1.5358, + "step": 1685 + }, + { + "epoch": 1.1660038042538474, + "grad_norm": 0.8015706539154053, + "learning_rate": 7.163657542224811e-06, + "loss": 1.6875, + "step": 1686 + }, + { + "epoch": 1.1666954867715718, + "grad_norm": 0.8840973377227783, + "learning_rate": 7.105416423995341e-06, + "loss": 1.0127, + "step": 1687 + }, + { + "epoch": 1.167387169289296, + "grad_norm": 1.1448097229003906, + "learning_rate": 7.047175305765872e-06, + "loss": 1.4883, + "step": 1688 + }, + { + "epoch": 1.1680788518070206, + "grad_norm": 0.8772367238998413, + "learning_rate": 6.988934187536401e-06, + "loss": 1.3213, + "step": 1689 + }, + { + "epoch": 1.168770534324745, + "grad_norm": 1.3608027696609497, + "learning_rate": 6.9306930693069314e-06, + "loss": 2.0195, + "step": 1690 + }, + { + "epoch": 1.1694622168424693, + "grad_norm": 1.2655911445617676, + "learning_rate": 6.87245195107746e-06, + "loss": 1.7195, + "step": 1691 + }, + { + "epoch": 1.1701538993601937, + "grad_norm": 0.5347901582717896, + "learning_rate": 6.81421083284799e-06, + "loss": 0.6424, + "step": 1692 + }, + { + "epoch": 1.170845581877918, + "grad_norm": 0.9062162041664124, + "learning_rate": 6.755969714618521e-06, + "loss": 1.6686, + "step": 1693 + }, + { + "epoch": 1.1715372643956423, + "grad_norm": 0.9029771089553833, + "learning_rate": 6.697728596389051e-06, + "loss": 1.4612, + "step": 1694 + }, + { + "epoch": 1.1722289469133669, + "grad_norm": 1.0943589210510254, + "learning_rate": 6.639487478159581e-06, + "loss": 1.5256, + "step": 1695 + }, + { + "epoch": 1.1729206294310912, + "grad_norm": 0.9593904614448547, + "learning_rate": 6.5812463599301115e-06, + "loss": 1.438, + "step": 1696 + }, + { + "epoch": 1.1736123119488155, + "grad_norm": 0.7761796712875366, + "learning_rate": 6.523005241700642e-06, + "loss": 1.3875, + "step": 1697 + }, + { + "epoch": 1.1743039944665399, + "grad_norm": 0.8188440799713135, + "learning_rate": 6.4647641234711704e-06, + "loss": 1.53, + "step": 1698 + }, + { + "epoch": 1.1749956769842642, + "grad_norm": 1.0318803787231445, + "learning_rate": 6.406523005241701e-06, + "loss": 1.6939, + "step": 1699 + }, + { + "epoch": 1.1756873595019885, + "grad_norm": 0.7635033130645752, + "learning_rate": 6.348281887012231e-06, + "loss": 1.9347, + "step": 1700 + }, + { + "epoch": 1.1763790420197129, + "grad_norm": 0.9503921866416931, + "learning_rate": 6.2900407687827605e-06, + "loss": 1.8576, + "step": 1701 + }, + { + "epoch": 1.1770707245374372, + "grad_norm": 0.7102924585342407, + "learning_rate": 6.231799650553291e-06, + "loss": 1.4608, + "step": 1702 + }, + { + "epoch": 1.1777624070551618, + "grad_norm": 0.8875777721405029, + "learning_rate": 6.173558532323821e-06, + "loss": 1.7386, + "step": 1703 + }, + { + "epoch": 1.178454089572886, + "grad_norm": 1.0264475345611572, + "learning_rate": 6.1153174140943506e-06, + "loss": 1.9825, + "step": 1704 + }, + { + "epoch": 1.1791457720906104, + "grad_norm": 0.7291073799133301, + "learning_rate": 6.057076295864881e-06, + "loss": 1.0899, + "step": 1705 + }, + { + "epoch": 1.1798374546083348, + "grad_norm": 0.8950873017311096, + "learning_rate": 5.998835177635411e-06, + "loss": 1.9644, + "step": 1706 + }, + { + "epoch": 1.1805291371260591, + "grad_norm": 0.6388865113258362, + "learning_rate": 5.940594059405941e-06, + "loss": 1.0341, + "step": 1707 + }, + { + "epoch": 1.1812208196437834, + "grad_norm": 0.8564591407775879, + "learning_rate": 5.882352941176471e-06, + "loss": 1.6944, + "step": 1708 + }, + { + "epoch": 1.1819125021615078, + "grad_norm": 1.463777780532837, + "learning_rate": 5.824111822947001e-06, + "loss": 2.1109, + "step": 1709 + }, + { + "epoch": 1.1826041846792323, + "grad_norm": 1.2916083335876465, + "learning_rate": 5.7658707047175315e-06, + "loss": 1.3081, + "step": 1710 + }, + { + "epoch": 1.1832958671969567, + "grad_norm": 0.883579432964325, + "learning_rate": 5.707629586488061e-06, + "loss": 1.6848, + "step": 1711 + }, + { + "epoch": 1.183987549714681, + "grad_norm": 1.019521951675415, + "learning_rate": 5.6493884682585904e-06, + "loss": 2.0684, + "step": 1712 + }, + { + "epoch": 1.1846792322324053, + "grad_norm": 0.9618213772773743, + "learning_rate": 5.591147350029121e-06, + "loss": 1.645, + "step": 1713 + }, + { + "epoch": 1.1853709147501297, + "grad_norm": 1.555221676826477, + "learning_rate": 5.53290623179965e-06, + "loss": 2.1313, + "step": 1714 + }, + { + "epoch": 1.186062597267854, + "grad_norm": 0.8156567215919495, + "learning_rate": 5.4746651135701805e-06, + "loss": 1.239, + "step": 1715 + }, + { + "epoch": 1.1867542797855783, + "grad_norm": 0.8247396349906921, + "learning_rate": 5.416423995340711e-06, + "loss": 1.9384, + "step": 1716 + }, + { + "epoch": 1.1874459623033027, + "grad_norm": 0.8407225608825684, + "learning_rate": 5.358182877111241e-06, + "loss": 1.474, + "step": 1717 + }, + { + "epoch": 1.1881376448210272, + "grad_norm": 0.677727997303009, + "learning_rate": 5.2999417588817705e-06, + "loss": 1.3451, + "step": 1718 + }, + { + "epoch": 1.1888293273387516, + "grad_norm": 0.993644654750824, + "learning_rate": 5.241700640652301e-06, + "loss": 1.0386, + "step": 1719 + }, + { + "epoch": 1.189521009856476, + "grad_norm": 0.8841768503189087, + "learning_rate": 5.183459522422831e-06, + "loss": 1.346, + "step": 1720 + }, + { + "epoch": 1.1902126923742002, + "grad_norm": 1.0056674480438232, + "learning_rate": 5.125218404193361e-06, + "loss": 2.1375, + "step": 1721 + }, + { + "epoch": 1.1909043748919246, + "grad_norm": 0.8851880431175232, + "learning_rate": 5.066977285963891e-06, + "loss": 1.8277, + "step": 1722 + }, + { + "epoch": 1.191596057409649, + "grad_norm": 0.7013739347457886, + "learning_rate": 5.008736167734421e-06, + "loss": 1.3859, + "step": 1723 + }, + { + "epoch": 1.1922877399273732, + "grad_norm": 0.997570276260376, + "learning_rate": 4.950495049504951e-06, + "loss": 1.593, + "step": 1724 + }, + { + "epoch": 1.1929794224450978, + "grad_norm": 0.7194474339485168, + "learning_rate": 4.89225393127548e-06, + "loss": 1.0988, + "step": 1725 + }, + { + "epoch": 1.1936711049628221, + "grad_norm": 1.3475464582443237, + "learning_rate": 4.83401281304601e-06, + "loss": 1.7412, + "step": 1726 + }, + { + "epoch": 1.1943627874805465, + "grad_norm": 0.9524636268615723, + "learning_rate": 4.775771694816541e-06, + "loss": 1.8068, + "step": 1727 + }, + { + "epoch": 1.1950544699982708, + "grad_norm": 1.0491656064987183, + "learning_rate": 4.717530576587071e-06, + "loss": 1.3985, + "step": 1728 + }, + { + "epoch": 1.1957461525159951, + "grad_norm": 0.8438884019851685, + "learning_rate": 4.6592894583576005e-06, + "loss": 1.6725, + "step": 1729 + }, + { + "epoch": 1.1964378350337195, + "grad_norm": 0.8291139006614685, + "learning_rate": 4.601048340128131e-06, + "loss": 1.3476, + "step": 1730 + }, + { + "epoch": 1.1971295175514438, + "grad_norm": 0.8813645839691162, + "learning_rate": 4.542807221898661e-06, + "loss": 2.0227, + "step": 1731 + }, + { + "epoch": 1.1978212000691681, + "grad_norm": 0.9611250758171082, + "learning_rate": 4.4845661036691905e-06, + "loss": 1.0414, + "step": 1732 + }, + { + "epoch": 1.1985128825868927, + "grad_norm": 0.8533077239990234, + "learning_rate": 4.426324985439721e-06, + "loss": 2.1895, + "step": 1733 + }, + { + "epoch": 1.199204565104617, + "grad_norm": 1.1292895078659058, + "learning_rate": 4.368083867210251e-06, + "loss": 1.3055, + "step": 1734 + }, + { + "epoch": 1.1998962476223414, + "grad_norm": 0.9011284112930298, + "learning_rate": 4.309842748980781e-06, + "loss": 1.8156, + "step": 1735 + }, + { + "epoch": 1.2005879301400657, + "grad_norm": 0.8826439380645752, + "learning_rate": 4.25160163075131e-06, + "loss": 1.3445, + "step": 1736 + }, + { + "epoch": 1.20127961265779, + "grad_norm": 0.9942179918289185, + "learning_rate": 4.19336051252184e-06, + "loss": 1.8752, + "step": 1737 + }, + { + "epoch": 1.2019712951755144, + "grad_norm": 0.7776837944984436, + "learning_rate": 4.135119394292371e-06, + "loss": 1.9512, + "step": 1738 + }, + { + "epoch": 1.2026629776932387, + "grad_norm": 0.8008607029914856, + "learning_rate": 4.0768782760629e-06, + "loss": 1.825, + "step": 1739 + }, + { + "epoch": 1.2033546602109633, + "grad_norm": 0.8627616167068481, + "learning_rate": 4.01863715783343e-06, + "loss": 1.5149, + "step": 1740 + }, + { + "epoch": 1.2040463427286876, + "grad_norm": 0.8164024949073792, + "learning_rate": 3.960396039603961e-06, + "loss": 1.0616, + "step": 1741 + }, + { + "epoch": 1.204738025246412, + "grad_norm": 0.8973370790481567, + "learning_rate": 3.902154921374491e-06, + "loss": 1.7584, + "step": 1742 + }, + { + "epoch": 1.2054297077641363, + "grad_norm": 0.940290629863739, + "learning_rate": 3.8439138031450205e-06, + "loss": 1.7521, + "step": 1743 + }, + { + "epoch": 1.2061213902818606, + "grad_norm": 0.7406190633773804, + "learning_rate": 3.7856726849155503e-06, + "loss": 1.1551, + "step": 1744 + }, + { + "epoch": 1.206813072799585, + "grad_norm": 0.792534589767456, + "learning_rate": 3.7274315666860806e-06, + "loss": 1.4167, + "step": 1745 + }, + { + "epoch": 1.2075047553173093, + "grad_norm": 0.6060773730278015, + "learning_rate": 3.669190448456611e-06, + "loss": 0.9178, + "step": 1746 + }, + { + "epoch": 1.2081964378350336, + "grad_norm": 1.111598253250122, + "learning_rate": 3.6109493302271404e-06, + "loss": 1.7898, + "step": 1747 + }, + { + "epoch": 1.2088881203527582, + "grad_norm": 0.7094188928604126, + "learning_rate": 3.5527082119976707e-06, + "loss": 1.5905, + "step": 1748 + }, + { + "epoch": 1.2095798028704825, + "grad_norm": 0.953357458114624, + "learning_rate": 3.4944670937682006e-06, + "loss": 1.6589, + "step": 1749 + }, + { + "epoch": 1.2102714853882068, + "grad_norm": 0.9556488990783691, + "learning_rate": 3.43622597553873e-06, + "loss": 1.0608, + "step": 1750 + }, + { + "epoch": 1.2109631679059312, + "grad_norm": 0.9832695126533508, + "learning_rate": 3.3779848573092603e-06, + "loss": 1.334, + "step": 1751 + }, + { + "epoch": 1.2116548504236555, + "grad_norm": 0.7971144914627075, + "learning_rate": 3.3197437390797906e-06, + "loss": 1.7318, + "step": 1752 + }, + { + "epoch": 1.2123465329413798, + "grad_norm": 1.296964168548584, + "learning_rate": 3.261502620850321e-06, + "loss": 1.6727, + "step": 1753 + }, + { + "epoch": 1.2130382154591042, + "grad_norm": 0.8589297533035278, + "learning_rate": 3.2032615026208504e-06, + "loss": 1.5484, + "step": 1754 + }, + { + "epoch": 1.2137298979768287, + "grad_norm": 0.8144680261611938, + "learning_rate": 3.1450203843913803e-06, + "loss": 1.0397, + "step": 1755 + }, + { + "epoch": 1.214421580494553, + "grad_norm": 0.9721055030822754, + "learning_rate": 3.0867792661619106e-06, + "loss": 1.1358, + "step": 1756 + }, + { + "epoch": 1.2151132630122774, + "grad_norm": 0.6962984800338745, + "learning_rate": 3.0285381479324404e-06, + "loss": 1.2004, + "step": 1757 + }, + { + "epoch": 1.2158049455300017, + "grad_norm": 0.8922805190086365, + "learning_rate": 2.9702970297029703e-06, + "loss": 1.7155, + "step": 1758 + }, + { + "epoch": 1.216496628047726, + "grad_norm": 0.9580974578857422, + "learning_rate": 2.9120559114735006e-06, + "loss": 1.9653, + "step": 1759 + }, + { + "epoch": 1.2171883105654504, + "grad_norm": 0.8019964694976807, + "learning_rate": 2.8538147932440305e-06, + "loss": 1.5887, + "step": 1760 + }, + { + "epoch": 1.2178799930831747, + "grad_norm": 0.8755665421485901, + "learning_rate": 2.7955736750145604e-06, + "loss": 1.8611, + "step": 1761 + }, + { + "epoch": 1.218571675600899, + "grad_norm": 0.7570388317108154, + "learning_rate": 2.7373325567850902e-06, + "loss": 1.3914, + "step": 1762 + }, + { + "epoch": 1.2192633581186236, + "grad_norm": 0.9185116291046143, + "learning_rate": 2.6790914385556205e-06, + "loss": 1.4699, + "step": 1763 + }, + { + "epoch": 1.219955040636348, + "grad_norm": 0.7744626402854919, + "learning_rate": 2.6208503203261504e-06, + "loss": 1.9008, + "step": 1764 + }, + { + "epoch": 1.2206467231540723, + "grad_norm": 0.834476113319397, + "learning_rate": 2.5626092020966803e-06, + "loss": 0.8271, + "step": 1765 + }, + { + "epoch": 1.2213384056717966, + "grad_norm": 1.4558451175689697, + "learning_rate": 2.5043680838672106e-06, + "loss": 2.2102, + "step": 1766 + }, + { + "epoch": 1.222030088189521, + "grad_norm": 0.9548596143722534, + "learning_rate": 2.44612696563774e-06, + "loss": 1.5597, + "step": 1767 + }, + { + "epoch": 1.2227217707072453, + "grad_norm": 0.7513933181762695, + "learning_rate": 2.3878858474082704e-06, + "loss": 1.5013, + "step": 1768 + }, + { + "epoch": 1.2234134532249699, + "grad_norm": 0.7574390172958374, + "learning_rate": 2.3296447291788002e-06, + "loss": 1.7306, + "step": 1769 + }, + { + "epoch": 1.2241051357426942, + "grad_norm": 0.963781476020813, + "learning_rate": 2.2714036109493305e-06, + "loss": 1.4079, + "step": 1770 + }, + { + "epoch": 1.2247968182604185, + "grad_norm": 0.8196379542350769, + "learning_rate": 2.2131624927198604e-06, + "loss": 1.2213, + "step": 1771 + }, + { + "epoch": 1.2254885007781429, + "grad_norm": 0.809852123260498, + "learning_rate": 2.1549213744903903e-06, + "loss": 1.4478, + "step": 1772 + }, + { + "epoch": 1.2261801832958672, + "grad_norm": 0.9147219061851501, + "learning_rate": 2.09668025626092e-06, + "loss": 0.9827, + "step": 1773 + }, + { + "epoch": 1.2268718658135915, + "grad_norm": 0.9042532444000244, + "learning_rate": 2.03843913803145e-06, + "loss": 1.9345, + "step": 1774 + }, + { + "epoch": 1.2275635483313159, + "grad_norm": 0.7653542160987854, + "learning_rate": 1.9801980198019803e-06, + "loss": 1.268, + "step": 1775 + }, + { + "epoch": 1.2282552308490402, + "grad_norm": 0.8734069466590881, + "learning_rate": 1.9219569015725102e-06, + "loss": 1.0422, + "step": 1776 + }, + { + "epoch": 1.2289469133667645, + "grad_norm": 0.8974725008010864, + "learning_rate": 1.8637157833430403e-06, + "loss": 1.8183, + "step": 1777 + }, + { + "epoch": 1.229638595884489, + "grad_norm": 0.9170926809310913, + "learning_rate": 1.8054746651135702e-06, + "loss": 1.6404, + "step": 1778 + }, + { + "epoch": 1.2303302784022134, + "grad_norm": 0.7619123458862305, + "learning_rate": 1.7472335468841003e-06, + "loss": 1.4713, + "step": 1779 + }, + { + "epoch": 1.2310219609199378, + "grad_norm": 0.8837324380874634, + "learning_rate": 1.6889924286546302e-06, + "loss": 1.4467, + "step": 1780 + }, + { + "epoch": 1.231713643437662, + "grad_norm": 0.9955554008483887, + "learning_rate": 1.6307513104251605e-06, + "loss": 1.7185, + "step": 1781 + }, + { + "epoch": 1.2324053259553864, + "grad_norm": 1.7202911376953125, + "learning_rate": 1.5725101921956901e-06, + "loss": 1.2339, + "step": 1782 + }, + { + "epoch": 1.2330970084731108, + "grad_norm": 1.0741084814071655, + "learning_rate": 1.5142690739662202e-06, + "loss": 1.6243, + "step": 1783 + }, + { + "epoch": 1.2337886909908353, + "grad_norm": 0.7952813506126404, + "learning_rate": 1.4560279557367503e-06, + "loss": 1.9482, + "step": 1784 + }, + { + "epoch": 1.2344803735085597, + "grad_norm": 0.8114420175552368, + "learning_rate": 1.3977868375072802e-06, + "loss": 1.8339, + "step": 1785 + }, + { + "epoch": 1.235172056026284, + "grad_norm": 0.8544707894325256, + "learning_rate": 1.3395457192778103e-06, + "loss": 2.0403, + "step": 1786 + }, + { + "epoch": 1.2358637385440083, + "grad_norm": 0.883529007434845, + "learning_rate": 1.2813046010483402e-06, + "loss": 1.4886, + "step": 1787 + }, + { + "epoch": 1.2365554210617327, + "grad_norm": 1.9958866834640503, + "learning_rate": 1.22306348281887e-06, + "loss": 0.8203, + "step": 1788 + }, + { + "epoch": 1.237247103579457, + "grad_norm": 1.0635721683502197, + "learning_rate": 1.1648223645894001e-06, + "loss": 1.5776, + "step": 1789 + }, + { + "epoch": 1.2379387860971813, + "grad_norm": 0.7094911932945251, + "learning_rate": 1.1065812463599302e-06, + "loss": 1.2963, + "step": 1790 + }, + { + "epoch": 1.2386304686149057, + "grad_norm": 0.8129479289054871, + "learning_rate": 1.04834012813046e-06, + "loss": 1.2952, + "step": 1791 + }, + { + "epoch": 1.23932215113263, + "grad_norm": 0.8642808198928833, + "learning_rate": 9.900990099009902e-07, + "loss": 1.6511, + "step": 1792 + }, + { + "epoch": 1.2400138336503546, + "grad_norm": 0.8090636134147644, + "learning_rate": 9.318578916715202e-07, + "loss": 1.5167, + "step": 1793 + }, + { + "epoch": 1.240705516168079, + "grad_norm": 0.8204816579818726, + "learning_rate": 8.736167734420501e-07, + "loss": 1.3854, + "step": 1794 + }, + { + "epoch": 1.2413971986858032, + "grad_norm": 0.828066885471344, + "learning_rate": 8.153756552125802e-07, + "loss": 1.7426, + "step": 1795 + }, + { + "epoch": 1.2420888812035276, + "grad_norm": 1.0011173486709595, + "learning_rate": 7.571345369831101e-07, + "loss": 1.7903, + "step": 1796 + }, + { + "epoch": 1.242780563721252, + "grad_norm": 1.3271926641464233, + "learning_rate": 6.988934187536401e-07, + "loss": 1.169, + "step": 1797 + }, + { + "epoch": 1.2434722462389762, + "grad_norm": 0.9071324467658997, + "learning_rate": 6.406523005241701e-07, + "loss": 1.7922, + "step": 1798 + }, + { + "epoch": 1.2441639287567008, + "grad_norm": 1.0562552213668823, + "learning_rate": 5.824111822947001e-07, + "loss": 1.5748, + "step": 1799 + }, + { + "epoch": 1.2448556112744251, + "grad_norm": 0.9828399419784546, + "learning_rate": 5.2417006406523e-07, + "loss": 1.8798, + "step": 1800 + }, + { + "epoch": 1.2455472937921495, + "grad_norm": 0.7894175052642822, + "learning_rate": 4.659289458357601e-07, + "loss": 1.2337, + "step": 1801 + }, + { + "epoch": 1.2462389763098738, + "grad_norm": 1.3144441843032837, + "learning_rate": 4.076878276062901e-07, + "loss": 1.4897, + "step": 1802 + }, + { + "epoch": 1.2469306588275981, + "grad_norm": 0.9511979222297668, + "learning_rate": 3.4944670937682005e-07, + "loss": 1.0088, + "step": 1803 + }, + { + "epoch": 1.2476223413453225, + "grad_norm": 1.0620826482772827, + "learning_rate": 2.9120559114735003e-07, + "loss": 2.12, + "step": 1804 + }, + { + "epoch": 1.2483140238630468, + "grad_norm": 0.9103230237960815, + "learning_rate": 2.3296447291788004e-07, + "loss": 2.079, + "step": 1805 + }, + { + "epoch": 1.2490057063807711, + "grad_norm": 0.8028068542480469, + "learning_rate": 1.7472335468841002e-07, + "loss": 1.1473, + "step": 1806 + }, + { + "epoch": 1.2496973888984957, + "grad_norm": 0.7902282476425171, + "learning_rate": 1.1648223645894002e-07, + "loss": 1.5735, + "step": 1807 + }, + { + "epoch": 1.25038907141622, + "grad_norm": 0.9187193512916565, + "learning_rate": 5.824111822947001e-08, + "loss": 1.9526, + "step": 1808 + }, + { + "epoch": 1.2510807539339444, + "grad_norm": 0.9034796357154846, + "learning_rate": 0.0, + "loss": 1.9859, + "step": 1809 + }, + { + "epoch": 1.2517724364516687, + "grad_norm": 0.8242852687835693, + "learning_rate": 3.942482708409174e-05, + "loss": 0.9241, + "step": 1810 + }, + { + "epoch": 1.252464118969393, + "grad_norm": 0.8046451807022095, + "learning_rate": 3.938842373498362e-05, + "loss": 1.0936, + "step": 1811 + }, + { + "epoch": 1.2531558014871174, + "grad_norm": 1.1966800689697266, + "learning_rate": 3.93520203858755e-05, + "loss": 1.4015, + "step": 1812 + }, + { + "epoch": 1.2538474840048417, + "grad_norm": 0.9291172623634338, + "learning_rate": 3.931561703676738e-05, + "loss": 2.0242, + "step": 1813 + }, + { + "epoch": 1.2545391665225663, + "grad_norm": 0.8558546900749207, + "learning_rate": 3.9279213687659264e-05, + "loss": 1.5019, + "step": 1814 + }, + { + "epoch": 1.2552308490402906, + "grad_norm": 1.8218317031860352, + "learning_rate": 3.924281033855115e-05, + "loss": 1.1266, + "step": 1815 + }, + { + "epoch": 1.255922531558015, + "grad_norm": 0.7981867790222168, + "learning_rate": 3.9206406989443026e-05, + "loss": 1.4552, + "step": 1816 + }, + { + "epoch": 1.2566142140757393, + "grad_norm": 0.6677556037902832, + "learning_rate": 3.917000364033491e-05, + "loss": 1.6076, + "step": 1817 + }, + { + "epoch": 1.2573058965934636, + "grad_norm": 0.788378894329071, + "learning_rate": 3.9133600291226794e-05, + "loss": 1.6708, + "step": 1818 + }, + { + "epoch": 1.257997579111188, + "grad_norm": 0.7965348958969116, + "learning_rate": 3.9097196942118675e-05, + "loss": 1.654, + "step": 1819 + }, + { + "epoch": 1.2586892616289123, + "grad_norm": 1.294364094734192, + "learning_rate": 3.9060793593010556e-05, + "loss": 1.74, + "step": 1820 + }, + { + "epoch": 1.2593809441466366, + "grad_norm": 0.7612819671630859, + "learning_rate": 3.9024390243902444e-05, + "loss": 1.7262, + "step": 1821 + }, + { + "epoch": 1.260072626664361, + "grad_norm": 0.97246915102005, + "learning_rate": 3.898798689479432e-05, + "loss": 1.3967, + "step": 1822 + }, + { + "epoch": 1.2607643091820855, + "grad_norm": 0.6338122487068176, + "learning_rate": 3.8951583545686206e-05, + "loss": 0.8912, + "step": 1823 + }, + { + "epoch": 1.2614559916998098, + "grad_norm": 0.8113964200019836, + "learning_rate": 3.891518019657809e-05, + "loss": 1.5932, + "step": 1824 + }, + { + "epoch": 1.2621476742175342, + "grad_norm": 0.757221519947052, + "learning_rate": 3.887877684746997e-05, + "loss": 1.6811, + "step": 1825 + }, + { + "epoch": 1.2628393567352585, + "grad_norm": 0.8660487532615662, + "learning_rate": 3.884237349836185e-05, + "loss": 2.0209, + "step": 1826 + }, + { + "epoch": 1.2635310392529828, + "grad_norm": 0.9298242926597595, + "learning_rate": 3.8805970149253736e-05, + "loss": 1.4099, + "step": 1827 + }, + { + "epoch": 1.2642227217707074, + "grad_norm": 0.7871847748756409, + "learning_rate": 3.876956680014561e-05, + "loss": 0.9616, + "step": 1828 + }, + { + "epoch": 1.2649144042884317, + "grad_norm": 0.8489501476287842, + "learning_rate": 3.87331634510375e-05, + "loss": 1.5154, + "step": 1829 + }, + { + "epoch": 1.265606086806156, + "grad_norm": 0.9907186031341553, + "learning_rate": 3.869676010192938e-05, + "loss": 1.6095, + "step": 1830 + }, + { + "epoch": 1.2662977693238804, + "grad_norm": 0.9827702641487122, + "learning_rate": 3.866035675282127e-05, + "loss": 0.9608, + "step": 1831 + }, + { + "epoch": 1.2669894518416047, + "grad_norm": 0.8035033941268921, + "learning_rate": 3.862395340371314e-05, + "loss": 1.5481, + "step": 1832 + }, + { + "epoch": 1.267681134359329, + "grad_norm": 1.0978895425796509, + "learning_rate": 3.858755005460503e-05, + "loss": 2.0443, + "step": 1833 + }, + { + "epoch": 1.2683728168770534, + "grad_norm": 0.8758592009544373, + "learning_rate": 3.855114670549691e-05, + "loss": 1.6489, + "step": 1834 + }, + { + "epoch": 1.2690644993947777, + "grad_norm": 0.9485347867012024, + "learning_rate": 3.851474335638879e-05, + "loss": 1.4746, + "step": 1835 + }, + { + "epoch": 1.269756181912502, + "grad_norm": 0.9407802224159241, + "learning_rate": 3.847834000728067e-05, + "loss": 1.5258, + "step": 1836 + }, + { + "epoch": 1.2704478644302264, + "grad_norm": 0.82390958070755, + "learning_rate": 3.844193665817255e-05, + "loss": 1.9283, + "step": 1837 + }, + { + "epoch": 1.271139546947951, + "grad_norm": 0.8061322569847107, + "learning_rate": 3.8405533309064434e-05, + "loss": 1.7803, + "step": 1838 + }, + { + "epoch": 1.2718312294656753, + "grad_norm": 0.8548954725265503, + "learning_rate": 3.836912995995632e-05, + "loss": 1.0745, + "step": 1839 + }, + { + "epoch": 1.2725229119833996, + "grad_norm": 0.8848943710327148, + "learning_rate": 3.83327266108482e-05, + "loss": 1.4076, + "step": 1840 + }, + { + "epoch": 1.273214594501124, + "grad_norm": 0.7797129154205322, + "learning_rate": 3.829632326174008e-05, + "loss": 1.7962, + "step": 1841 + }, + { + "epoch": 1.2739062770188483, + "grad_norm": 0.8673209547996521, + "learning_rate": 3.8259919912631964e-05, + "loss": 1.8038, + "step": 1842 + }, + { + "epoch": 1.2745979595365728, + "grad_norm": 1.3388570547103882, + "learning_rate": 3.8223516563523845e-05, + "loss": 1.9272, + "step": 1843 + }, + { + "epoch": 1.2752896420542972, + "grad_norm": 0.9810792207717896, + "learning_rate": 3.8187113214415726e-05, + "loss": 1.8324, + "step": 1844 + }, + { + "epoch": 1.2759813245720215, + "grad_norm": 0.8849295377731323, + "learning_rate": 3.815070986530761e-05, + "loss": 1.8651, + "step": 1845 + }, + { + "epoch": 1.2766730070897458, + "grad_norm": 1.7068865299224854, + "learning_rate": 3.8114306516199495e-05, + "loss": 1.5318, + "step": 1846 + }, + { + "epoch": 1.2773646896074702, + "grad_norm": 0.9444654583930969, + "learning_rate": 3.807790316709137e-05, + "loss": 1.5704, + "step": 1847 + }, + { + "epoch": 1.2780563721251945, + "grad_norm": 0.8557124137878418, + "learning_rate": 3.8041499817983257e-05, + "loss": 2.1305, + "step": 1848 + }, + { + "epoch": 1.2787480546429189, + "grad_norm": 0.8498656749725342, + "learning_rate": 3.800509646887514e-05, + "loss": 1.3427, + "step": 1849 + }, + { + "epoch": 1.2794397371606432, + "grad_norm": 0.7638754844665527, + "learning_rate": 3.796869311976702e-05, + "loss": 0.8919, + "step": 1850 + }, + { + "epoch": 1.2801314196783675, + "grad_norm": 0.8892665505409241, + "learning_rate": 3.79322897706589e-05, + "loss": 0.7198, + "step": 1851 + }, + { + "epoch": 1.2808231021960919, + "grad_norm": 0.8356907367706299, + "learning_rate": 3.789588642155079e-05, + "loss": 1.8621, + "step": 1852 + }, + { + "epoch": 1.2815147847138164, + "grad_norm": 0.7617167234420776, + "learning_rate": 3.785948307244266e-05, + "loss": 1.1182, + "step": 1853 + }, + { + "epoch": 1.2822064672315407, + "grad_norm": 0.6191325187683105, + "learning_rate": 3.782307972333455e-05, + "loss": 1.1352, + "step": 1854 + }, + { + "epoch": 1.282898149749265, + "grad_norm": 1.0343198776245117, + "learning_rate": 3.778667637422643e-05, + "loss": 1.9038, + "step": 1855 + }, + { + "epoch": 1.2835898322669894, + "grad_norm": 1.0529627799987793, + "learning_rate": 3.775027302511831e-05, + "loss": 1.3542, + "step": 1856 + }, + { + "epoch": 1.2842815147847138, + "grad_norm": 1.4223628044128418, + "learning_rate": 3.771386967601019e-05, + "loss": 1.6522, + "step": 1857 + }, + { + "epoch": 1.2849731973024383, + "grad_norm": 0.8065588474273682, + "learning_rate": 3.767746632690208e-05, + "loss": 1.6015, + "step": 1858 + }, + { + "epoch": 1.2856648798201626, + "grad_norm": 0.7395001649856567, + "learning_rate": 3.7641062977793954e-05, + "loss": 1.3136, + "step": 1859 + }, + { + "epoch": 1.286356562337887, + "grad_norm": 1.0379223823547363, + "learning_rate": 3.760465962868584e-05, + "loss": 1.6277, + "step": 1860 + }, + { + "epoch": 1.2870482448556113, + "grad_norm": 0.716241717338562, + "learning_rate": 3.756825627957772e-05, + "loss": 1.6013, + "step": 1861 + }, + { + "epoch": 1.2877399273733356, + "grad_norm": 0.8964675068855286, + "learning_rate": 3.753185293046961e-05, + "loss": 1.8281, + "step": 1862 + }, + { + "epoch": 1.28843160989106, + "grad_norm": 0.9805247783660889, + "learning_rate": 3.7495449581361484e-05, + "loss": 1.1535, + "step": 1863 + }, + { + "epoch": 1.2891232924087843, + "grad_norm": 0.9833081364631653, + "learning_rate": 3.745904623225337e-05, + "loss": 1.0096, + "step": 1864 + }, + { + "epoch": 1.2898149749265087, + "grad_norm": 0.9487015008926392, + "learning_rate": 3.742264288314525e-05, + "loss": 1.8207, + "step": 1865 + }, + { + "epoch": 1.290506657444233, + "grad_norm": 1.1372509002685547, + "learning_rate": 3.7386239534037134e-05, + "loss": 0.9241, + "step": 1866 + }, + { + "epoch": 1.2911983399619575, + "grad_norm": 1.0568435192108154, + "learning_rate": 3.7349836184929015e-05, + "loss": 1.059, + "step": 1867 + }, + { + "epoch": 1.2918900224796819, + "grad_norm": 0.74294513463974, + "learning_rate": 3.73134328358209e-05, + "loss": 0.9375, + "step": 1868 + }, + { + "epoch": 1.2925817049974062, + "grad_norm": 0.9022307991981506, + "learning_rate": 3.727702948671278e-05, + "loss": 1.3262, + "step": 1869 + }, + { + "epoch": 1.2932733875151305, + "grad_norm": 0.9156614542007446, + "learning_rate": 3.7240626137604665e-05, + "loss": 1.5394, + "step": 1870 + }, + { + "epoch": 1.2939650700328549, + "grad_norm": 0.9429082870483398, + "learning_rate": 3.7204222788496546e-05, + "loss": 1.9851, + "step": 1871 + }, + { + "epoch": 1.2946567525505792, + "grad_norm": 0.8855860829353333, + "learning_rate": 3.7167819439388426e-05, + "loss": 1.9623, + "step": 1872 + }, + { + "epoch": 1.2953484350683038, + "grad_norm": 0.7447137236595154, + "learning_rate": 3.713141609028031e-05, + "loss": 1.7209, + "step": 1873 + }, + { + "epoch": 1.296040117586028, + "grad_norm": 1.0196491479873657, + "learning_rate": 3.709501274117219e-05, + "loss": 1.4934, + "step": 1874 + }, + { + "epoch": 1.2967318001037524, + "grad_norm": 0.7354358434677124, + "learning_rate": 3.705860939206407e-05, + "loss": 1.7464, + "step": 1875 + }, + { + "epoch": 1.2974234826214768, + "grad_norm": 0.9382926821708679, + "learning_rate": 3.702220604295595e-05, + "loss": 1.618, + "step": 1876 + }, + { + "epoch": 1.298115165139201, + "grad_norm": 0.8664951920509338, + "learning_rate": 3.698580269384784e-05, + "loss": 1.4829, + "step": 1877 + }, + { + "epoch": 1.2988068476569254, + "grad_norm": 0.7127748131752014, + "learning_rate": 3.694939934473972e-05, + "loss": 1.2425, + "step": 1878 + }, + { + "epoch": 1.2994985301746498, + "grad_norm": 1.1230412721633911, + "learning_rate": 3.69129959956316e-05, + "loss": 1.7808, + "step": 1879 + }, + { + "epoch": 1.3001902126923741, + "grad_norm": 1.0134071111679077, + "learning_rate": 3.687659264652348e-05, + "loss": 1.7617, + "step": 1880 + }, + { + "epoch": 1.3008818952100984, + "grad_norm": 0.7539078593254089, + "learning_rate": 3.684018929741536e-05, + "loss": 1.6487, + "step": 1881 + }, + { + "epoch": 1.301573577727823, + "grad_norm": 1.0024137496948242, + "learning_rate": 3.680378594830724e-05, + "loss": 1.7021, + "step": 1882 + }, + { + "epoch": 1.3022652602455473, + "grad_norm": 0.9032376408576965, + "learning_rate": 3.676738259919913e-05, + "loss": 1.9292, + "step": 1883 + }, + { + "epoch": 1.3029569427632717, + "grad_norm": 0.8564134836196899, + "learning_rate": 3.6730979250091005e-05, + "loss": 1.6919, + "step": 1884 + }, + { + "epoch": 1.303648625280996, + "grad_norm": 0.7483653426170349, + "learning_rate": 3.669457590098289e-05, + "loss": 1.0905, + "step": 1885 + }, + { + "epoch": 1.3043403077987203, + "grad_norm": 1.19503653049469, + "learning_rate": 3.665817255187477e-05, + "loss": 1.5814, + "step": 1886 + }, + { + "epoch": 1.3050319903164447, + "grad_norm": 0.9574710130691528, + "learning_rate": 3.6621769202766654e-05, + "loss": 1.8714, + "step": 1887 + }, + { + "epoch": 1.3057236728341692, + "grad_norm": 0.7887849807739258, + "learning_rate": 3.6585365853658535e-05, + "loss": 1.4556, + "step": 1888 + }, + { + "epoch": 1.3064153553518936, + "grad_norm": 0.7369945049285889, + "learning_rate": 3.654896250455042e-05, + "loss": 0.9844, + "step": 1889 + }, + { + "epoch": 1.307107037869618, + "grad_norm": 1.0574779510498047, + "learning_rate": 3.6512559155442304e-05, + "loss": 1.7363, + "step": 1890 + }, + { + "epoch": 1.3077987203873422, + "grad_norm": 0.7746357917785645, + "learning_rate": 3.6476155806334185e-05, + "loss": 1.2119, + "step": 1891 + }, + { + "epoch": 1.3084904029050666, + "grad_norm": 1.3173068761825562, + "learning_rate": 3.6439752457226066e-05, + "loss": 1.2764, + "step": 1892 + }, + { + "epoch": 1.309182085422791, + "grad_norm": 0.8748396039009094, + "learning_rate": 3.6403349108117953e-05, + "loss": 1.5357, + "step": 1893 + }, + { + "epoch": 1.3098737679405152, + "grad_norm": 0.9714013338088989, + "learning_rate": 3.636694575900983e-05, + "loss": 1.3777, + "step": 1894 + }, + { + "epoch": 1.3105654504582396, + "grad_norm": 0.8243861198425293, + "learning_rate": 3.6330542409901715e-05, + "loss": 1.5617, + "step": 1895 + }, + { + "epoch": 1.311257132975964, + "grad_norm": 0.9486716985702515, + "learning_rate": 3.6294139060793596e-05, + "loss": 2.0681, + "step": 1896 + }, + { + "epoch": 1.3119488154936885, + "grad_norm": 0.8115946054458618, + "learning_rate": 3.625773571168548e-05, + "loss": 1.8801, + "step": 1897 + }, + { + "epoch": 1.3126404980114128, + "grad_norm": 0.8780468702316284, + "learning_rate": 3.622133236257736e-05, + "loss": 1.3265, + "step": 1898 + }, + { + "epoch": 1.3133321805291371, + "grad_norm": 0.9462249279022217, + "learning_rate": 3.6184929013469246e-05, + "loss": 1.4264, + "step": 1899 + }, + { + "epoch": 1.3140238630468615, + "grad_norm": 0.9610809087753296, + "learning_rate": 3.614852566436112e-05, + "loss": 1.7539, + "step": 1900 + }, + { + "epoch": 1.3147155455645858, + "grad_norm": 0.7879363894462585, + "learning_rate": 3.611212231525301e-05, + "loss": 1.6434, + "step": 1901 + }, + { + "epoch": 1.3154072280823101, + "grad_norm": 0.9264380931854248, + "learning_rate": 3.607571896614489e-05, + "loss": 1.6235, + "step": 1902 + }, + { + "epoch": 1.3160989106000347, + "grad_norm": 0.849011242389679, + "learning_rate": 3.603931561703677e-05, + "loss": 1.4447, + "step": 1903 + }, + { + "epoch": 1.316790593117759, + "grad_norm": 0.9442479014396667, + "learning_rate": 3.600291226792865e-05, + "loss": 1.0444, + "step": 1904 + }, + { + "epoch": 1.3174822756354834, + "grad_norm": 1.0553207397460938, + "learning_rate": 3.596650891882054e-05, + "loss": 1.587, + "step": 1905 + }, + { + "epoch": 1.3181739581532077, + "grad_norm": 0.9437171220779419, + "learning_rate": 3.593010556971241e-05, + "loss": 1.1938, + "step": 1906 + }, + { + "epoch": 1.318865640670932, + "grad_norm": 0.7971818447113037, + "learning_rate": 3.58937022206043e-05, + "loss": 1.0695, + "step": 1907 + }, + { + "epoch": 1.3195573231886564, + "grad_norm": 0.9962893128395081, + "learning_rate": 3.585729887149618e-05, + "loss": 1.8475, + "step": 1908 + }, + { + "epoch": 1.3202490057063807, + "grad_norm": 1.122597098350525, + "learning_rate": 3.582089552238806e-05, + "loss": 1.0026, + "step": 1909 + }, + { + "epoch": 1.320940688224105, + "grad_norm": 0.84935462474823, + "learning_rate": 3.578449217327994e-05, + "loss": 1.8014, + "step": 1910 + }, + { + "epoch": 1.3216323707418294, + "grad_norm": 0.8386709690093994, + "learning_rate": 3.5748088824171824e-05, + "loss": 1.9097, + "step": 1911 + }, + { + "epoch": 1.322324053259554, + "grad_norm": 0.853274941444397, + "learning_rate": 3.5711685475063705e-05, + "loss": 2.0259, + "step": 1912 + }, + { + "epoch": 1.3230157357772783, + "grad_norm": 0.6767750978469849, + "learning_rate": 3.5675282125955586e-05, + "loss": 0.9174, + "step": 1913 + }, + { + "epoch": 1.3237074182950026, + "grad_norm": 0.9412535429000854, + "learning_rate": 3.5638878776847474e-05, + "loss": 1.3135, + "step": 1914 + }, + { + "epoch": 1.324399100812727, + "grad_norm": 0.9713802337646484, + "learning_rate": 3.560247542773935e-05, + "loss": 1.4882, + "step": 1915 + }, + { + "epoch": 1.3250907833304513, + "grad_norm": 0.7832048535346985, + "learning_rate": 3.5566072078631236e-05, + "loss": 1.7031, + "step": 1916 + }, + { + "epoch": 1.3257824658481756, + "grad_norm": 0.5951961278915405, + "learning_rate": 3.5529668729523117e-05, + "loss": 0.2102, + "step": 1917 + }, + { + "epoch": 1.3264741483659002, + "grad_norm": 1.3314907550811768, + "learning_rate": 3.5493265380415e-05, + "loss": 2.0361, + "step": 1918 + }, + { + "epoch": 1.3271658308836245, + "grad_norm": 0.8842023015022278, + "learning_rate": 3.545686203130688e-05, + "loss": 2.0366, + "step": 1919 + }, + { + "epoch": 1.3278575134013488, + "grad_norm": 0.8390107750892639, + "learning_rate": 3.5420458682198766e-05, + "loss": 1.7833, + "step": 1920 + }, + { + "epoch": 1.3285491959190732, + "grad_norm": 1.0433529615402222, + "learning_rate": 3.538405533309065e-05, + "loss": 1.7994, + "step": 1921 + }, + { + "epoch": 1.3292408784367975, + "grad_norm": 0.9450101852416992, + "learning_rate": 3.534765198398253e-05, + "loss": 1.8912, + "step": 1922 + }, + { + "epoch": 1.3299325609545218, + "grad_norm": 0.9666965007781982, + "learning_rate": 3.531124863487441e-05, + "loss": 2.1481, + "step": 1923 + }, + { + "epoch": 1.3306242434722462, + "grad_norm": 0.845268964767456, + "learning_rate": 3.52748452857663e-05, + "loss": 1.7217, + "step": 1924 + }, + { + "epoch": 1.3313159259899705, + "grad_norm": 1.0283581018447876, + "learning_rate": 3.523844193665817e-05, + "loss": 1.7777, + "step": 1925 + }, + { + "epoch": 1.3320076085076948, + "grad_norm": 1.2970536947250366, + "learning_rate": 3.520203858755006e-05, + "loss": 1.2992, + "step": 1926 + }, + { + "epoch": 1.3326992910254194, + "grad_norm": 0.8032466769218445, + "learning_rate": 3.516563523844194e-05, + "loss": 1.4745, + "step": 1927 + }, + { + "epoch": 1.3333909735431437, + "grad_norm": 1.101788878440857, + "learning_rate": 3.512923188933382e-05, + "loss": 1.0427, + "step": 1928 + }, + { + "epoch": 1.334082656060868, + "grad_norm": 1.4565415382385254, + "learning_rate": 3.50928285402257e-05, + "loss": 1.5888, + "step": 1929 + }, + { + "epoch": 1.3347743385785924, + "grad_norm": 1.0797853469848633, + "learning_rate": 3.505642519111759e-05, + "loss": 1.8314, + "step": 1930 + }, + { + "epoch": 1.3354660210963167, + "grad_norm": 1.1963329315185547, + "learning_rate": 3.5020021842009463e-05, + "loss": 1.7397, + "step": 1931 + }, + { + "epoch": 1.3361577036140413, + "grad_norm": 1.0609445571899414, + "learning_rate": 3.498361849290135e-05, + "loss": 1.4553, + "step": 1932 + }, + { + "epoch": 1.3368493861317656, + "grad_norm": 0.9074013233184814, + "learning_rate": 3.494721514379323e-05, + "loss": 1.994, + "step": 1933 + }, + { + "epoch": 1.33754106864949, + "grad_norm": 0.9575091600418091, + "learning_rate": 3.491081179468511e-05, + "loss": 1.3343, + "step": 1934 + }, + { + "epoch": 1.3382327511672143, + "grad_norm": 0.9111261963844299, + "learning_rate": 3.4874408445576994e-05, + "loss": 2.0059, + "step": 1935 + }, + { + "epoch": 1.3389244336849386, + "grad_norm": 1.0967391729354858, + "learning_rate": 3.483800509646888e-05, + "loss": 0.8606, + "step": 1936 + }, + { + "epoch": 1.339616116202663, + "grad_norm": 0.9879076480865479, + "learning_rate": 3.4801601747360756e-05, + "loss": 1.3205, + "step": 1937 + }, + { + "epoch": 1.3403077987203873, + "grad_norm": 0.8380864262580872, + "learning_rate": 3.4765198398252644e-05, + "loss": 1.1052, + "step": 1938 + }, + { + "epoch": 1.3409994812381116, + "grad_norm": 0.7907826900482178, + "learning_rate": 3.4728795049144525e-05, + "loss": 1.8246, + "step": 1939 + }, + { + "epoch": 1.341691163755836, + "grad_norm": 1.152637004852295, + "learning_rate": 3.4692391700036406e-05, + "loss": 1.4916, + "step": 1940 + }, + { + "epoch": 1.3423828462735603, + "grad_norm": 0.8820608258247375, + "learning_rate": 3.4655988350928286e-05, + "loss": 1.5205, + "step": 1941 + }, + { + "epoch": 1.3430745287912849, + "grad_norm": 0.9089112281799316, + "learning_rate": 3.461958500182017e-05, + "loss": 1.7229, + "step": 1942 + }, + { + "epoch": 1.3437662113090092, + "grad_norm": 1.005041241645813, + "learning_rate": 3.458318165271205e-05, + "loss": 1.8108, + "step": 1943 + }, + { + "epoch": 1.3444578938267335, + "grad_norm": 0.9245075583457947, + "learning_rate": 3.4546778303603936e-05, + "loss": 1.5649, + "step": 1944 + }, + { + "epoch": 1.3451495763444579, + "grad_norm": 0.9393779635429382, + "learning_rate": 3.451037495449582e-05, + "loss": 2.0629, + "step": 1945 + }, + { + "epoch": 1.3458412588621822, + "grad_norm": 0.849787175655365, + "learning_rate": 3.44739716053877e-05, + "loss": 0.9961, + "step": 1946 + }, + { + "epoch": 1.3465329413799068, + "grad_norm": 1.092171549797058, + "learning_rate": 3.443756825627958e-05, + "loss": 2.1455, + "step": 1947 + }, + { + "epoch": 1.347224623897631, + "grad_norm": 1.2719589471817017, + "learning_rate": 3.440116490717146e-05, + "loss": 1.7918, + "step": 1948 + }, + { + "epoch": 1.3479163064153554, + "grad_norm": 0.9361221194267273, + "learning_rate": 3.436476155806334e-05, + "loss": 1.8331, + "step": 1949 + }, + { + "epoch": 1.3486079889330798, + "grad_norm": 0.8719432950019836, + "learning_rate": 3.432835820895522e-05, + "loss": 1.139, + "step": 1950 + }, + { + "epoch": 1.349299671450804, + "grad_norm": 0.9212896227836609, + "learning_rate": 3.429195485984711e-05, + "loss": 1.5082, + "step": 1951 + }, + { + "epoch": 1.3499913539685284, + "grad_norm": 0.7980116605758667, + "learning_rate": 3.425555151073899e-05, + "loss": 1.7042, + "step": 1952 + }, + { + "epoch": 1.3506830364862528, + "grad_norm": 0.9115065336227417, + "learning_rate": 3.421914816163087e-05, + "loss": 1.2251, + "step": 1953 + }, + { + "epoch": 1.351374719003977, + "grad_norm": 0.6713042259216309, + "learning_rate": 3.418274481252275e-05, + "loss": 1.122, + "step": 1954 + }, + { + "epoch": 1.3520664015217014, + "grad_norm": 0.8126210570335388, + "learning_rate": 3.414634146341464e-05, + "loss": 1.2804, + "step": 1955 + }, + { + "epoch": 1.3527580840394258, + "grad_norm": 0.7522339820861816, + "learning_rate": 3.4109938114306514e-05, + "loss": 0.7906, + "step": 1956 + }, + { + "epoch": 1.3534497665571503, + "grad_norm": 0.793449878692627, + "learning_rate": 3.40735347651984e-05, + "loss": 1.2429, + "step": 1957 + }, + { + "epoch": 1.3541414490748747, + "grad_norm": 1.1302376985549927, + "learning_rate": 3.403713141609028e-05, + "loss": 1.4192, + "step": 1958 + }, + { + "epoch": 1.354833131592599, + "grad_norm": 0.927330732345581, + "learning_rate": 3.4000728066982164e-05, + "loss": 1.9934, + "step": 1959 + }, + { + "epoch": 1.3555248141103233, + "grad_norm": 0.8644328713417053, + "learning_rate": 3.3964324717874045e-05, + "loss": 1.9719, + "step": 1960 + }, + { + "epoch": 1.3562164966280477, + "grad_norm": 0.9866162538528442, + "learning_rate": 3.392792136876593e-05, + "loss": 2.0058, + "step": 1961 + }, + { + "epoch": 1.3569081791457722, + "grad_norm": 1.3088966608047485, + "learning_rate": 3.389151801965781e-05, + "loss": 2.0066, + "step": 1962 + }, + { + "epoch": 1.3575998616634966, + "grad_norm": 0.7904307246208191, + "learning_rate": 3.3855114670549694e-05, + "loss": 1.239, + "step": 1963 + }, + { + "epoch": 1.358291544181221, + "grad_norm": 0.8573948740959167, + "learning_rate": 3.3818711321441575e-05, + "loss": 1.5207, + "step": 1964 + }, + { + "epoch": 1.3589832266989452, + "grad_norm": 0.9484941959381104, + "learning_rate": 3.3782307972333456e-05, + "loss": 1.9436, + "step": 1965 + }, + { + "epoch": 1.3596749092166696, + "grad_norm": 1.0695719718933105, + "learning_rate": 3.374590462322534e-05, + "loss": 1.2806, + "step": 1966 + }, + { + "epoch": 1.360366591734394, + "grad_norm": 0.8883374333381653, + "learning_rate": 3.3709501274117225e-05, + "loss": 1.4222, + "step": 1967 + }, + { + "epoch": 1.3610582742521182, + "grad_norm": 0.8962006568908691, + "learning_rate": 3.36730979250091e-05, + "loss": 1.3712, + "step": 1968 + }, + { + "epoch": 1.3617499567698426, + "grad_norm": 0.7431996464729309, + "learning_rate": 3.363669457590099e-05, + "loss": 1.8663, + "step": 1969 + }, + { + "epoch": 1.362441639287567, + "grad_norm": 1.1721320152282715, + "learning_rate": 3.360029122679287e-05, + "loss": 1.2521, + "step": 1970 + }, + { + "epoch": 1.3631333218052915, + "grad_norm": 0.8025596737861633, + "learning_rate": 3.356388787768475e-05, + "loss": 1.4624, + "step": 1971 + }, + { + "epoch": 1.3638250043230158, + "grad_norm": 1.3165680170059204, + "learning_rate": 3.352748452857663e-05, + "loss": 1.9478, + "step": 1972 + }, + { + "epoch": 1.3645166868407401, + "grad_norm": 0.9428910613059998, + "learning_rate": 3.349108117946852e-05, + "loss": 1.8116, + "step": 1973 + }, + { + "epoch": 1.3652083693584645, + "grad_norm": 0.8140659928321838, + "learning_rate": 3.345467783036039e-05, + "loss": 1.3329, + "step": 1974 + }, + { + "epoch": 1.3659000518761888, + "grad_norm": 0.8956738710403442, + "learning_rate": 3.341827448125228e-05, + "loss": 1.6468, + "step": 1975 + }, + { + "epoch": 1.3665917343939131, + "grad_norm": 0.8602350950241089, + "learning_rate": 3.338187113214416e-05, + "loss": 1.3494, + "step": 1976 + }, + { + "epoch": 1.3672834169116377, + "grad_norm": 0.8748160004615784, + "learning_rate": 3.334546778303604e-05, + "loss": 1.4456, + "step": 1977 + }, + { + "epoch": 1.367975099429362, + "grad_norm": 0.5959149599075317, + "learning_rate": 3.330906443392792e-05, + "loss": 0.8933, + "step": 1978 + }, + { + "epoch": 1.3686667819470864, + "grad_norm": 0.8277064561843872, + "learning_rate": 3.32726610848198e-05, + "loss": 1.1484, + "step": 1979 + }, + { + "epoch": 1.3693584644648107, + "grad_norm": 0.8472768068313599, + "learning_rate": 3.3236257735711684e-05, + "loss": 1.2087, + "step": 1980 + }, + { + "epoch": 1.370050146982535, + "grad_norm": 0.8168089389801025, + "learning_rate": 3.3199854386603565e-05, + "loss": 1.9683, + "step": 1981 + }, + { + "epoch": 1.3707418295002594, + "grad_norm": 1.3133728504180908, + "learning_rate": 3.316345103749545e-05, + "loss": 1.5552, + "step": 1982 + }, + { + "epoch": 1.3714335120179837, + "grad_norm": 0.897013247013092, + "learning_rate": 3.3127047688387334e-05, + "loss": 1.7741, + "step": 1983 + }, + { + "epoch": 1.372125194535708, + "grad_norm": 0.984239935874939, + "learning_rate": 3.3090644339279215e-05, + "loss": 1.8397, + "step": 1984 + }, + { + "epoch": 1.3728168770534324, + "grad_norm": 0.9961678385734558, + "learning_rate": 3.3054240990171096e-05, + "loss": 1.9499, + "step": 1985 + }, + { + "epoch": 1.373508559571157, + "grad_norm": 0.7989017963409424, + "learning_rate": 3.301783764106298e-05, + "loss": 1.4258, + "step": 1986 + }, + { + "epoch": 1.3742002420888813, + "grad_norm": 1.1241167783737183, + "learning_rate": 3.298143429195486e-05, + "loss": 2.2852, + "step": 1987 + }, + { + "epoch": 1.3748919246066056, + "grad_norm": 1.1435675621032715, + "learning_rate": 3.2945030942846745e-05, + "loss": 2.0182, + "step": 1988 + }, + { + "epoch": 1.37558360712433, + "grad_norm": 0.9687975645065308, + "learning_rate": 3.2908627593738626e-05, + "loss": 1.6482, + "step": 1989 + }, + { + "epoch": 1.3762752896420543, + "grad_norm": 1.0680760145187378, + "learning_rate": 3.287222424463051e-05, + "loss": 1.8309, + "step": 1990 + }, + { + "epoch": 1.3769669721597786, + "grad_norm": 0.8759111166000366, + "learning_rate": 3.283582089552239e-05, + "loss": 1.8496, + "step": 1991 + }, + { + "epoch": 1.3776586546775031, + "grad_norm": 0.8297052383422852, + "learning_rate": 3.2799417546414276e-05, + "loss": 1.7293, + "step": 1992 + }, + { + "epoch": 1.3783503371952275, + "grad_norm": 0.9033782482147217, + "learning_rate": 3.276301419730615e-05, + "loss": 2.0258, + "step": 1993 + }, + { + "epoch": 1.3790420197129518, + "grad_norm": 1.040075659751892, + "learning_rate": 3.272661084819804e-05, + "loss": 1.8762, + "step": 1994 + }, + { + "epoch": 1.3797337022306762, + "grad_norm": 0.9219354391098022, + "learning_rate": 3.269020749908992e-05, + "loss": 1.55, + "step": 1995 + }, + { + "epoch": 1.3804253847484005, + "grad_norm": 0.7530835866928101, + "learning_rate": 3.26538041499818e-05, + "loss": 1.1414, + "step": 1996 + }, + { + "epoch": 1.3811170672661248, + "grad_norm": 1.0707637071609497, + "learning_rate": 3.261740080087368e-05, + "loss": 1.9501, + "step": 1997 + }, + { + "epoch": 1.3818087497838492, + "grad_norm": 0.8039554953575134, + "learning_rate": 3.258099745176557e-05, + "loss": 1.5128, + "step": 1998 + }, + { + "epoch": 1.3825004323015735, + "grad_norm": 0.9372995495796204, + "learning_rate": 3.254459410265744e-05, + "loss": 1.7541, + "step": 1999 + }, + { + "epoch": 1.3831921148192978, + "grad_norm": 0.8266523480415344, + "learning_rate": 3.250819075354933e-05, + "loss": 1.4367, + "step": 2000 + }, + { + "epoch": 1.3831921148192978, + "eval_loss": 1.4428718090057373, + "eval_runtime": 572.781, + "eval_samples_per_second": 2.243, + "eval_steps_per_second": 1.123, + "step": 2000 + }, + { + "epoch": 1.3838837973370224, + "grad_norm": 0.984535813331604, + "learning_rate": 3.247178740444121e-05, + "loss": 1.4718, + "step": 2001 + }, + { + "epoch": 1.3845754798547467, + "grad_norm": 0.8770325779914856, + "learning_rate": 3.243538405533309e-05, + "loss": 1.6298, + "step": 2002 + }, + { + "epoch": 1.385267162372471, + "grad_norm": 0.8822327852249146, + "learning_rate": 3.239898070622497e-05, + "loss": 1.982, + "step": 2003 + }, + { + "epoch": 1.3859588448901954, + "grad_norm": 1.0327270030975342, + "learning_rate": 3.236257735711686e-05, + "loss": 1.1729, + "step": 2004 + }, + { + "epoch": 1.3866505274079197, + "grad_norm": 0.8722877502441406, + "learning_rate": 3.2326174008008735e-05, + "loss": 1.75, + "step": 2005 + }, + { + "epoch": 1.387342209925644, + "grad_norm": 0.9822987914085388, + "learning_rate": 3.228977065890062e-05, + "loss": 1.7807, + "step": 2006 + }, + { + "epoch": 1.3880338924433686, + "grad_norm": 0.8557208180427551, + "learning_rate": 3.2253367309792504e-05, + "loss": 0.8672, + "step": 2007 + }, + { + "epoch": 1.388725574961093, + "grad_norm": 0.8146331310272217, + "learning_rate": 3.2216963960684385e-05, + "loss": 0.7328, + "step": 2008 + }, + { + "epoch": 1.3894172574788173, + "grad_norm": 0.9947080016136169, + "learning_rate": 3.2180560611576266e-05, + "loss": 2.082, + "step": 2009 + }, + { + "epoch": 1.3901089399965416, + "grad_norm": 0.9812456965446472, + "learning_rate": 3.2144157262468146e-05, + "loss": 1.3078, + "step": 2010 + }, + { + "epoch": 1.390800622514266, + "grad_norm": 0.8116563558578491, + "learning_rate": 3.210775391336003e-05, + "loss": 0.9362, + "step": 2011 + }, + { + "epoch": 1.3914923050319903, + "grad_norm": 0.8665092587471008, + "learning_rate": 3.2071350564251915e-05, + "loss": 1.7722, + "step": 2012 + }, + { + "epoch": 1.3921839875497146, + "grad_norm": 0.9438626766204834, + "learning_rate": 3.2034947215143796e-05, + "loss": 1.3206, + "step": 2013 + }, + { + "epoch": 1.392875670067439, + "grad_norm": 0.8886246681213379, + "learning_rate": 3.199854386603568e-05, + "loss": 1.4046, + "step": 2014 + }, + { + "epoch": 1.3935673525851633, + "grad_norm": 1.079876184463501, + "learning_rate": 3.196214051692756e-05, + "loss": 0.6202, + "step": 2015 + }, + { + "epoch": 1.3942590351028878, + "grad_norm": 0.8678843975067139, + "learning_rate": 3.192573716781944e-05, + "loss": 1.4741, + "step": 2016 + }, + { + "epoch": 1.3949507176206122, + "grad_norm": 0.7784777283668518, + "learning_rate": 3.188933381871133e-05, + "loss": 1.2453, + "step": 2017 + }, + { + "epoch": 1.3956424001383365, + "grad_norm": 0.9093233346939087, + "learning_rate": 3.18529304696032e-05, + "loss": 1.7637, + "step": 2018 + }, + { + "epoch": 1.3963340826560608, + "grad_norm": 0.8888777494430542, + "learning_rate": 3.181652712049509e-05, + "loss": 1.7317, + "step": 2019 + }, + { + "epoch": 1.3970257651737852, + "grad_norm": 1.071500539779663, + "learning_rate": 3.178012377138697e-05, + "loss": 1.1548, + "step": 2020 + }, + { + "epoch": 1.3977174476915095, + "grad_norm": 1.8188599348068237, + "learning_rate": 3.174372042227885e-05, + "loss": 1.8737, + "step": 2021 + }, + { + "epoch": 1.398409130209234, + "grad_norm": 1.022908329963684, + "learning_rate": 3.170731707317073e-05, + "loss": 1.17, + "step": 2022 + }, + { + "epoch": 1.3991008127269584, + "grad_norm": 0.7291505336761475, + "learning_rate": 3.167091372406262e-05, + "loss": 0.9002, + "step": 2023 + }, + { + "epoch": 1.3997924952446827, + "grad_norm": 1.122520089149475, + "learning_rate": 3.163451037495449e-05, + "loss": 1.3312, + "step": 2024 + }, + { + "epoch": 1.400484177762407, + "grad_norm": 1.1598308086395264, + "learning_rate": 3.159810702584638e-05, + "loss": 1.5041, + "step": 2025 + }, + { + "epoch": 1.4011758602801314, + "grad_norm": 1.453706979751587, + "learning_rate": 3.156170367673826e-05, + "loss": 1.3363, + "step": 2026 + }, + { + "epoch": 1.4018675427978557, + "grad_norm": 0.9163658022880554, + "learning_rate": 3.152530032763014e-05, + "loss": 1.871, + "step": 2027 + }, + { + "epoch": 1.40255922531558, + "grad_norm": 1.0080070495605469, + "learning_rate": 3.1488896978522024e-05, + "loss": 1.6916, + "step": 2028 + }, + { + "epoch": 1.4032509078333044, + "grad_norm": 0.8330860733985901, + "learning_rate": 3.145249362941391e-05, + "loss": 1.0437, + "step": 2029 + }, + { + "epoch": 1.4039425903510288, + "grad_norm": 0.8617184162139893, + "learning_rate": 3.1416090280305786e-05, + "loss": 1.4751, + "step": 2030 + }, + { + "epoch": 1.4046342728687533, + "grad_norm": 0.8956760764122009, + "learning_rate": 3.1379686931197674e-05, + "loss": 0.8605, + "step": 2031 + }, + { + "epoch": 1.4053259553864776, + "grad_norm": 1.022942066192627, + "learning_rate": 3.1343283582089554e-05, + "loss": 1.9217, + "step": 2032 + }, + { + "epoch": 1.406017637904202, + "grad_norm": 0.972957968711853, + "learning_rate": 3.1306880232981435e-05, + "loss": 1.2501, + "step": 2033 + }, + { + "epoch": 1.4067093204219263, + "grad_norm": 0.8603571653366089, + "learning_rate": 3.1270476883873316e-05, + "loss": 1.3215, + "step": 2034 + }, + { + "epoch": 1.4074010029396506, + "grad_norm": 0.9371677041053772, + "learning_rate": 3.1234073534765204e-05, + "loss": 1.4054, + "step": 2035 + }, + { + "epoch": 1.408092685457375, + "grad_norm": 0.8988721966743469, + "learning_rate": 3.119767018565708e-05, + "loss": 1.4167, + "step": 2036 + }, + { + "epoch": 1.4087843679750995, + "grad_norm": 1.142337679862976, + "learning_rate": 3.1161266836548966e-05, + "loss": 2.0087, + "step": 2037 + }, + { + "epoch": 1.4094760504928239, + "grad_norm": 1.1030882596969604, + "learning_rate": 3.112486348744085e-05, + "loss": 1.9914, + "step": 2038 + }, + { + "epoch": 1.4101677330105482, + "grad_norm": 0.951513946056366, + "learning_rate": 3.108846013833273e-05, + "loss": 1.3162, + "step": 2039 + }, + { + "epoch": 1.4108594155282725, + "grad_norm": 1.359693169593811, + "learning_rate": 3.105205678922461e-05, + "loss": 1.6323, + "step": 2040 + }, + { + "epoch": 1.4115510980459969, + "grad_norm": 1.0670546293258667, + "learning_rate": 3.1015653440116497e-05, + "loss": 1.8598, + "step": 2041 + }, + { + "epoch": 1.4122427805637212, + "grad_norm": 0.8523070812225342, + "learning_rate": 3.097925009100837e-05, + "loss": 1.5866, + "step": 2042 + }, + { + "epoch": 1.4129344630814455, + "grad_norm": 0.9881057739257812, + "learning_rate": 3.094284674190026e-05, + "loss": 1.7549, + "step": 2043 + }, + { + "epoch": 1.4136261455991699, + "grad_norm": 1.243102788925171, + "learning_rate": 3.090644339279214e-05, + "loss": 1.5402, + "step": 2044 + }, + { + "epoch": 1.4143178281168942, + "grad_norm": 1.0429972410202026, + "learning_rate": 3.087004004368402e-05, + "loss": 1.5623, + "step": 2045 + }, + { + "epoch": 1.4150095106346188, + "grad_norm": 0.9752019047737122, + "learning_rate": 3.08336366945759e-05, + "loss": 1.676, + "step": 2046 + }, + { + "epoch": 1.415701193152343, + "grad_norm": 0.9358912110328674, + "learning_rate": 3.079723334546778e-05, + "loss": 1.4046, + "step": 2047 + }, + { + "epoch": 1.4163928756700674, + "grad_norm": 1.0242204666137695, + "learning_rate": 3.076082999635967e-05, + "loss": 1.0768, + "step": 2048 + }, + { + "epoch": 1.4170845581877918, + "grad_norm": 0.8519956469535828, + "learning_rate": 3.0724426647251544e-05, + "loss": 1.0261, + "step": 2049 + }, + { + "epoch": 1.417776240705516, + "grad_norm": 1.011920690536499, + "learning_rate": 3.068802329814343e-05, + "loss": 1.7517, + "step": 2050 + }, + { + "epoch": 1.4184679232232407, + "grad_norm": 1.0758455991744995, + "learning_rate": 3.065161994903531e-05, + "loss": 1.8425, + "step": 2051 + }, + { + "epoch": 1.419159605740965, + "grad_norm": 1.2541645765304565, + "learning_rate": 3.0615216599927194e-05, + "loss": 1.9531, + "step": 2052 + }, + { + "epoch": 1.4198512882586893, + "grad_norm": 0.8418980836868286, + "learning_rate": 3.0578813250819075e-05, + "loss": 1.1586, + "step": 2053 + }, + { + "epoch": 1.4205429707764137, + "grad_norm": 1.1952518224716187, + "learning_rate": 3.054240990171096e-05, + "loss": 1.1757, + "step": 2054 + }, + { + "epoch": 1.421234653294138, + "grad_norm": 0.8220242261886597, + "learning_rate": 3.050600655260284e-05, + "loss": 1.581, + "step": 2055 + }, + { + "epoch": 1.4219263358118623, + "grad_norm": 1.1030998229980469, + "learning_rate": 3.0469603203494724e-05, + "loss": 1.3963, + "step": 2056 + }, + { + "epoch": 1.4226180183295867, + "grad_norm": 1.0655429363250732, + "learning_rate": 3.043319985438661e-05, + "loss": 1.953, + "step": 2057 + }, + { + "epoch": 1.423309700847311, + "grad_norm": 0.948377788066864, + "learning_rate": 3.0396796505278486e-05, + "loss": 1.6345, + "step": 2058 + }, + { + "epoch": 1.4240013833650353, + "grad_norm": 0.9457731246948242, + "learning_rate": 3.036039315617037e-05, + "loss": 1.2298, + "step": 2059 + }, + { + "epoch": 1.4246930658827597, + "grad_norm": 0.9740118384361267, + "learning_rate": 3.032398980706225e-05, + "loss": 1.8855, + "step": 2060 + }, + { + "epoch": 1.4253847484004842, + "grad_norm": 1.0778017044067383, + "learning_rate": 3.0287586457954132e-05, + "loss": 1.8038, + "step": 2061 + }, + { + "epoch": 1.4260764309182086, + "grad_norm": 0.7892261743545532, + "learning_rate": 3.0251183108846017e-05, + "loss": 1.5528, + "step": 2062 + }, + { + "epoch": 1.426768113435933, + "grad_norm": 0.6538057923316956, + "learning_rate": 3.0214779759737898e-05, + "loss": 1.0946, + "step": 2063 + }, + { + "epoch": 1.4274597959536572, + "grad_norm": 0.9798827767372131, + "learning_rate": 3.017837641062978e-05, + "loss": 1.857, + "step": 2064 + }, + { + "epoch": 1.4281514784713816, + "grad_norm": 0.7944064140319824, + "learning_rate": 3.014197306152166e-05, + "loss": 0.8624, + "step": 2065 + }, + { + "epoch": 1.4288431609891061, + "grad_norm": 0.9933128952980042, + "learning_rate": 3.0105569712413544e-05, + "loss": 1.7867, + "step": 2066 + }, + { + "epoch": 1.4295348435068305, + "grad_norm": 1.0323796272277832, + "learning_rate": 3.0069166363305425e-05, + "loss": 1.4007, + "step": 2067 + }, + { + "epoch": 1.4302265260245548, + "grad_norm": 0.9992436170578003, + "learning_rate": 3.0032763014197306e-05, + "loss": 1.9203, + "step": 2068 + }, + { + "epoch": 1.4309182085422791, + "grad_norm": 1.204050898551941, + "learning_rate": 2.999635966508919e-05, + "loss": 1.7265, + "step": 2069 + }, + { + "epoch": 1.4316098910600035, + "grad_norm": 0.7516879439353943, + "learning_rate": 2.9959956315981068e-05, + "loss": 1.6953, + "step": 2070 + }, + { + "epoch": 1.4323015735777278, + "grad_norm": 1.1178338527679443, + "learning_rate": 2.9923552966872952e-05, + "loss": 1.512, + "step": 2071 + }, + { + "epoch": 1.4329932560954521, + "grad_norm": 1.11270010471344, + "learning_rate": 2.9887149617764836e-05, + "loss": 1.6694, + "step": 2072 + }, + { + "epoch": 1.4336849386131765, + "grad_norm": 0.9197645783424377, + "learning_rate": 2.9850746268656714e-05, + "loss": 1.4221, + "step": 2073 + }, + { + "epoch": 1.4343766211309008, + "grad_norm": 1.053852915763855, + "learning_rate": 2.98143429195486e-05, + "loss": 1.5718, + "step": 2074 + }, + { + "epoch": 1.4350683036486251, + "grad_norm": 0.8827266097068787, + "learning_rate": 2.9777939570440483e-05, + "loss": 1.4309, + "step": 2075 + }, + { + "epoch": 1.4357599861663497, + "grad_norm": 1.0066460371017456, + "learning_rate": 2.9741536221332367e-05, + "loss": 1.8799, + "step": 2076 + }, + { + "epoch": 1.436451668684074, + "grad_norm": 1.2522358894348145, + "learning_rate": 2.9705132872224245e-05, + "loss": 2.0703, + "step": 2077 + }, + { + "epoch": 1.4371433512017984, + "grad_norm": 0.8221448063850403, + "learning_rate": 2.966872952311613e-05, + "loss": 1.4469, + "step": 2078 + }, + { + "epoch": 1.4378350337195227, + "grad_norm": 0.9565039277076721, + "learning_rate": 2.9632326174008013e-05, + "loss": 1.4525, + "step": 2079 + }, + { + "epoch": 1.438526716237247, + "grad_norm": 1.0112320184707642, + "learning_rate": 2.959592282489989e-05, + "loss": 1.8909, + "step": 2080 + }, + { + "epoch": 1.4392183987549716, + "grad_norm": 0.6184213161468506, + "learning_rate": 2.9559519475791775e-05, + "loss": 1.0124, + "step": 2081 + }, + { + "epoch": 1.439910081272696, + "grad_norm": 0.8376033306121826, + "learning_rate": 2.952311612668366e-05, + "loss": 1.2771, + "step": 2082 + }, + { + "epoch": 1.4406017637904203, + "grad_norm": 0.953520655632019, + "learning_rate": 2.9486712777575537e-05, + "loss": 1.5206, + "step": 2083 + }, + { + "epoch": 1.4412934463081446, + "grad_norm": 0.9576830267906189, + "learning_rate": 2.945030942846742e-05, + "loss": 1.5874, + "step": 2084 + }, + { + "epoch": 1.441985128825869, + "grad_norm": 1.042127251625061, + "learning_rate": 2.9413906079359306e-05, + "loss": 1.697, + "step": 2085 + }, + { + "epoch": 1.4426768113435933, + "grad_norm": 0.9355104565620422, + "learning_rate": 2.9377502730251183e-05, + "loss": 1.6515, + "step": 2086 + }, + { + "epoch": 1.4433684938613176, + "grad_norm": 0.8054540753364563, + "learning_rate": 2.9341099381143068e-05, + "loss": 1.7362, + "step": 2087 + }, + { + "epoch": 1.444060176379042, + "grad_norm": 0.883931040763855, + "learning_rate": 2.9304696032034952e-05, + "loss": 1.3874, + "step": 2088 + }, + { + "epoch": 1.4447518588967663, + "grad_norm": 0.7654348015785217, + "learning_rate": 2.926829268292683e-05, + "loss": 1.2725, + "step": 2089 + }, + { + "epoch": 1.4454435414144908, + "grad_norm": 0.7754483819007874, + "learning_rate": 2.9231889333818714e-05, + "loss": 1.1652, + "step": 2090 + }, + { + "epoch": 1.4461352239322152, + "grad_norm": 1.050100326538086, + "learning_rate": 2.9195485984710598e-05, + "loss": 1.0042, + "step": 2091 + }, + { + "epoch": 1.4468269064499395, + "grad_norm": 1.0527033805847168, + "learning_rate": 2.9159082635602476e-05, + "loss": 1.629, + "step": 2092 + }, + { + "epoch": 1.4475185889676638, + "grad_norm": 0.8082405924797058, + "learning_rate": 2.912267928649436e-05, + "loss": 1.0546, + "step": 2093 + }, + { + "epoch": 1.4482102714853882, + "grad_norm": 0.8122817873954773, + "learning_rate": 2.908627593738624e-05, + "loss": 1.372, + "step": 2094 + }, + { + "epoch": 1.4489019540031125, + "grad_norm": 0.9192789196968079, + "learning_rate": 2.9049872588278122e-05, + "loss": 1.5171, + "step": 2095 + }, + { + "epoch": 1.449593636520837, + "grad_norm": 0.7754972577095032, + "learning_rate": 2.9013469239170006e-05, + "loss": 1.5841, + "step": 2096 + }, + { + "epoch": 1.4502853190385614, + "grad_norm": 1.4227865934371948, + "learning_rate": 2.8977065890061887e-05, + "loss": 1.2395, + "step": 2097 + }, + { + "epoch": 1.4509770015562857, + "grad_norm": 1.1727687120437622, + "learning_rate": 2.8940662540953768e-05, + "loss": 1.5705, + "step": 2098 + }, + { + "epoch": 1.45166868407401, + "grad_norm": 1.0853270292282104, + "learning_rate": 2.890425919184565e-05, + "loss": 1.3697, + "step": 2099 + }, + { + "epoch": 1.4523603665917344, + "grad_norm": 0.7672690749168396, + "learning_rate": 2.8867855842737534e-05, + "loss": 1.4192, + "step": 2100 + }, + { + "epoch": 1.4530520491094587, + "grad_norm": 0.8678781986236572, + "learning_rate": 2.8831452493629414e-05, + "loss": 1.0606, + "step": 2101 + }, + { + "epoch": 1.453743731627183, + "grad_norm": 1.2555452585220337, + "learning_rate": 2.8795049144521295e-05, + "loss": 1.4279, + "step": 2102 + }, + { + "epoch": 1.4544354141449074, + "grad_norm": 0.984279453754425, + "learning_rate": 2.875864579541318e-05, + "loss": 1.7996, + "step": 2103 + }, + { + "epoch": 1.4551270966626317, + "grad_norm": 0.9172074794769287, + "learning_rate": 2.8722242446305057e-05, + "loss": 1.689, + "step": 2104 + }, + { + "epoch": 1.4558187791803563, + "grad_norm": 1.1231496334075928, + "learning_rate": 2.868583909719694e-05, + "loss": 1.7618, + "step": 2105 + }, + { + "epoch": 1.4565104616980806, + "grad_norm": 0.7898075580596924, + "learning_rate": 2.8649435748088826e-05, + "loss": 1.9489, + "step": 2106 + }, + { + "epoch": 1.457202144215805, + "grad_norm": 0.9540377855300903, + "learning_rate": 2.861303239898071e-05, + "loss": 1.1625, + "step": 2107 + }, + { + "epoch": 1.4578938267335293, + "grad_norm": 1.1656274795532227, + "learning_rate": 2.8576629049872588e-05, + "loss": 1.9003, + "step": 2108 + }, + { + "epoch": 1.4585855092512536, + "grad_norm": 0.9742414951324463, + "learning_rate": 2.8540225700764472e-05, + "loss": 1.7539, + "step": 2109 + }, + { + "epoch": 1.459277191768978, + "grad_norm": 0.7401710152626038, + "learning_rate": 2.8503822351656357e-05, + "loss": 0.7433, + "step": 2110 + }, + { + "epoch": 1.4599688742867025, + "grad_norm": 0.7515429854393005, + "learning_rate": 2.8467419002548234e-05, + "loss": 1.4108, + "step": 2111 + }, + { + "epoch": 1.4606605568044269, + "grad_norm": 1.014967679977417, + "learning_rate": 2.843101565344012e-05, + "loss": 1.3691, + "step": 2112 + }, + { + "epoch": 1.4613522393221512, + "grad_norm": 0.7807340621948242, + "learning_rate": 2.8394612304332003e-05, + "loss": 1.6938, + "step": 2113 + }, + { + "epoch": 1.4620439218398755, + "grad_norm": 0.8181832432746887, + "learning_rate": 2.835820895522388e-05, + "loss": 1.2001, + "step": 2114 + }, + { + "epoch": 1.4627356043575999, + "grad_norm": 0.7824162244796753, + "learning_rate": 2.8321805606115765e-05, + "loss": 1.8433, + "step": 2115 + }, + { + "epoch": 1.4634272868753242, + "grad_norm": 0.7915189862251282, + "learning_rate": 2.828540225700765e-05, + "loss": 0.9591, + "step": 2116 + }, + { + "epoch": 1.4641189693930485, + "grad_norm": 1.091099739074707, + "learning_rate": 2.8248998907899527e-05, + "loss": 1.3694, + "step": 2117 + }, + { + "epoch": 1.4648106519107729, + "grad_norm": 1.0215061902999878, + "learning_rate": 2.821259555879141e-05, + "loss": 1.4408, + "step": 2118 + }, + { + "epoch": 1.4655023344284972, + "grad_norm": 0.9416251182556152, + "learning_rate": 2.8176192209683295e-05, + "loss": 1.7343, + "step": 2119 + }, + { + "epoch": 1.4661940169462218, + "grad_norm": 0.7310605049133301, + "learning_rate": 2.8139788860575173e-05, + "loss": 1.0775, + "step": 2120 + }, + { + "epoch": 1.466885699463946, + "grad_norm": 1.1415340900421143, + "learning_rate": 2.8103385511467057e-05, + "loss": 1.2196, + "step": 2121 + }, + { + "epoch": 1.4675773819816704, + "grad_norm": 0.894534707069397, + "learning_rate": 2.806698216235894e-05, + "loss": 1.4169, + "step": 2122 + }, + { + "epoch": 1.4682690644993948, + "grad_norm": 1.0294947624206543, + "learning_rate": 2.803057881325082e-05, + "loss": 2.1316, + "step": 2123 + }, + { + "epoch": 1.468960747017119, + "grad_norm": 1.0295460224151611, + "learning_rate": 2.7994175464142703e-05, + "loss": 1.8167, + "step": 2124 + }, + { + "epoch": 1.4696524295348434, + "grad_norm": 1.4340204000473022, + "learning_rate": 2.7957772115034588e-05, + "loss": 1.1795, + "step": 2125 + }, + { + "epoch": 1.470344112052568, + "grad_norm": 1.02447509765625, + "learning_rate": 2.7921368765926465e-05, + "loss": 2.0563, + "step": 2126 + }, + { + "epoch": 1.4710357945702923, + "grad_norm": 0.9650735259056091, + "learning_rate": 2.788496541681835e-05, + "loss": 1.4544, + "step": 2127 + }, + { + "epoch": 1.4717274770880167, + "grad_norm": 1.0105457305908203, + "learning_rate": 2.784856206771023e-05, + "loss": 1.4324, + "step": 2128 + }, + { + "epoch": 1.472419159605741, + "grad_norm": 0.6995699405670166, + "learning_rate": 2.781215871860211e-05, + "loss": 1.3468, + "step": 2129 + }, + { + "epoch": 1.4731108421234653, + "grad_norm": 0.9325497150421143, + "learning_rate": 2.7775755369493996e-05, + "loss": 1.8303, + "step": 2130 + }, + { + "epoch": 1.4738025246411897, + "grad_norm": 1.4274487495422363, + "learning_rate": 2.7739352020385877e-05, + "loss": 1.4923, + "step": 2131 + }, + { + "epoch": 1.474494207158914, + "grad_norm": 0.8967704772949219, + "learning_rate": 2.7702948671277758e-05, + "loss": 2.006, + "step": 2132 + }, + { + "epoch": 1.4751858896766383, + "grad_norm": 0.8790166974067688, + "learning_rate": 2.766654532216964e-05, + "loss": 1.6006, + "step": 2133 + }, + { + "epoch": 1.4758775721943627, + "grad_norm": 1.218833088874817, + "learning_rate": 2.7630141973061523e-05, + "loss": 1.3981, + "step": 2134 + }, + { + "epoch": 1.4765692547120872, + "grad_norm": 0.8063285946846008, + "learning_rate": 2.7593738623953404e-05, + "loss": 1.1325, + "step": 2135 + }, + { + "epoch": 1.4772609372298116, + "grad_norm": 0.9671421051025391, + "learning_rate": 2.7557335274845285e-05, + "loss": 1.3994, + "step": 2136 + }, + { + "epoch": 1.477952619747536, + "grad_norm": 0.8792144656181335, + "learning_rate": 2.752093192573717e-05, + "loss": 2.0558, + "step": 2137 + }, + { + "epoch": 1.4786443022652602, + "grad_norm": 1.1556845903396606, + "learning_rate": 2.7484528576629054e-05, + "loss": 1.3411, + "step": 2138 + }, + { + "epoch": 1.4793359847829846, + "grad_norm": 1.064486026763916, + "learning_rate": 2.744812522752093e-05, + "loss": 1.9644, + "step": 2139 + }, + { + "epoch": 1.480027667300709, + "grad_norm": 0.9538613557815552, + "learning_rate": 2.7411721878412815e-05, + "loss": 1.8002, + "step": 2140 + }, + { + "epoch": 1.4807193498184335, + "grad_norm": 1.4818669557571411, + "learning_rate": 2.73753185293047e-05, + "loss": 1.0879, + "step": 2141 + }, + { + "epoch": 1.4814110323361578, + "grad_norm": 1.0517104864120483, + "learning_rate": 2.7338915180196577e-05, + "loss": 1.0992, + "step": 2142 + }, + { + "epoch": 1.4821027148538821, + "grad_norm": 1.2164661884307861, + "learning_rate": 2.7302511831088462e-05, + "loss": 1.9742, + "step": 2143 + }, + { + "epoch": 1.4827943973716065, + "grad_norm": 0.7196162939071655, + "learning_rate": 2.7266108481980346e-05, + "loss": 0.829, + "step": 2144 + }, + { + "epoch": 1.4834860798893308, + "grad_norm": 0.95005202293396, + "learning_rate": 2.7229705132872224e-05, + "loss": 1.4282, + "step": 2145 + }, + { + "epoch": 1.4841777624070551, + "grad_norm": 0.8048756718635559, + "learning_rate": 2.7193301783764108e-05, + "loss": 0.8343, + "step": 2146 + }, + { + "epoch": 1.4848694449247795, + "grad_norm": 0.8858820796012878, + "learning_rate": 2.7156898434655992e-05, + "loss": 1.3228, + "step": 2147 + }, + { + "epoch": 1.4855611274425038, + "grad_norm": 0.9090560078620911, + "learning_rate": 2.712049508554787e-05, + "loss": 1.692, + "step": 2148 + }, + { + "epoch": 1.4862528099602281, + "grad_norm": 0.8048760890960693, + "learning_rate": 2.7084091736439754e-05, + "loss": 0.828, + "step": 2149 + }, + { + "epoch": 1.4869444924779527, + "grad_norm": 1.728994369506836, + "learning_rate": 2.704768838733164e-05, + "loss": 1.2424, + "step": 2150 + }, + { + "epoch": 1.487636174995677, + "grad_norm": 0.9087968468666077, + "learning_rate": 2.7011285038223516e-05, + "loss": 1.9992, + "step": 2151 + }, + { + "epoch": 1.4883278575134014, + "grad_norm": 0.9076192378997803, + "learning_rate": 2.69748816891154e-05, + "loss": 1.8559, + "step": 2152 + }, + { + "epoch": 1.4890195400311257, + "grad_norm": 0.7927855253219604, + "learning_rate": 2.6938478340007285e-05, + "loss": 1.2985, + "step": 2153 + }, + { + "epoch": 1.48971122254885, + "grad_norm": 1.0480645895004272, + "learning_rate": 2.6902074990899162e-05, + "loss": 1.9779, + "step": 2154 + }, + { + "epoch": 1.4904029050665746, + "grad_norm": 1.1005717515945435, + "learning_rate": 2.6865671641791047e-05, + "loss": 1.466, + "step": 2155 + }, + { + "epoch": 1.491094587584299, + "grad_norm": 0.8802543878555298, + "learning_rate": 2.682926829268293e-05, + "loss": 1.5946, + "step": 2156 + }, + { + "epoch": 1.4917862701020232, + "grad_norm": 0.8819108009338379, + "learning_rate": 2.679286494357481e-05, + "loss": 1.571, + "step": 2157 + }, + { + "epoch": 1.4924779526197476, + "grad_norm": 0.9346900582313538, + "learning_rate": 2.6756461594466693e-05, + "loss": 2.2181, + "step": 2158 + }, + { + "epoch": 1.493169635137472, + "grad_norm": 0.9447158575057983, + "learning_rate": 2.6720058245358577e-05, + "loss": 1.4636, + "step": 2159 + }, + { + "epoch": 1.4938613176551963, + "grad_norm": 0.7425116896629333, + "learning_rate": 2.6683654896250455e-05, + "loss": 0.9839, + "step": 2160 + }, + { + "epoch": 1.4945530001729206, + "grad_norm": 1.0036354064941406, + "learning_rate": 2.664725154714234e-05, + "loss": 1.6515, + "step": 2161 + }, + { + "epoch": 1.495244682690645, + "grad_norm": 1.0199100971221924, + "learning_rate": 2.661084819803422e-05, + "loss": 1.86, + "step": 2162 + }, + { + "epoch": 1.4959363652083693, + "grad_norm": 0.8444830179214478, + "learning_rate": 2.65744448489261e-05, + "loss": 1.7584, + "step": 2163 + }, + { + "epoch": 1.4966280477260936, + "grad_norm": 0.8935509920120239, + "learning_rate": 2.6538041499817985e-05, + "loss": 1.9101, + "step": 2164 + }, + { + "epoch": 1.4973197302438181, + "grad_norm": 1.1125240325927734, + "learning_rate": 2.6501638150709866e-05, + "loss": 2.0366, + "step": 2165 + }, + { + "epoch": 1.4980114127615425, + "grad_norm": 0.8527866005897522, + "learning_rate": 2.6465234801601747e-05, + "loss": 1.944, + "step": 2166 + }, + { + "epoch": 1.4987030952792668, + "grad_norm": 1.4828540086746216, + "learning_rate": 2.6428831452493628e-05, + "loss": 1.5715, + "step": 2167 + }, + { + "epoch": 1.4993947777969912, + "grad_norm": 0.866712212562561, + "learning_rate": 2.6392428103385513e-05, + "loss": 1.7244, + "step": 2168 + }, + { + "epoch": 1.5000864603147157, + "grad_norm": 1.0088051557540894, + "learning_rate": 2.6356024754277397e-05, + "loss": 1.0077, + "step": 2169 + }, + { + "epoch": 1.50077814283244, + "grad_norm": 0.9489467144012451, + "learning_rate": 2.6319621405169274e-05, + "loss": 1.8881, + "step": 2170 + }, + { + "epoch": 1.5014698253501644, + "grad_norm": 0.8433712720870972, + "learning_rate": 2.628321805606116e-05, + "loss": 0.955, + "step": 2171 + }, + { + "epoch": 1.5021615078678887, + "grad_norm": 1.0661237239837646, + "learning_rate": 2.6246814706953043e-05, + "loss": 1.6466, + "step": 2172 + }, + { + "epoch": 1.502853190385613, + "grad_norm": 1.4430923461914062, + "learning_rate": 2.621041135784492e-05, + "loss": 1.8077, + "step": 2173 + }, + { + "epoch": 1.5035448729033374, + "grad_norm": 0.9212530851364136, + "learning_rate": 2.6174008008736805e-05, + "loss": 1.8886, + "step": 2174 + }, + { + "epoch": 1.5042365554210617, + "grad_norm": 1.1725152730941772, + "learning_rate": 2.613760465962869e-05, + "loss": 1.578, + "step": 2175 + }, + { + "epoch": 1.504928237938786, + "grad_norm": 1.1502147912979126, + "learning_rate": 2.6101201310520567e-05, + "loss": 1.455, + "step": 2176 + }, + { + "epoch": 1.5056199204565104, + "grad_norm": 0.9008479118347168, + "learning_rate": 2.606479796141245e-05, + "loss": 1.6361, + "step": 2177 + }, + { + "epoch": 1.5063116029742347, + "grad_norm": 0.8758420944213867, + "learning_rate": 2.6028394612304336e-05, + "loss": 1.8129, + "step": 2178 + }, + { + "epoch": 1.507003285491959, + "grad_norm": 0.9053659439086914, + "learning_rate": 2.5991991263196213e-05, + "loss": 1.7446, + "step": 2179 + }, + { + "epoch": 1.5076949680096834, + "grad_norm": 1.2413023710250854, + "learning_rate": 2.5955587914088097e-05, + "loss": 1.5854, + "step": 2180 + }, + { + "epoch": 1.508386650527408, + "grad_norm": 1.2506285905838013, + "learning_rate": 2.5919184564979982e-05, + "loss": 1.5796, + "step": 2181 + }, + { + "epoch": 1.5090783330451323, + "grad_norm": 1.1174542903900146, + "learning_rate": 2.588278121587186e-05, + "loss": 1.8836, + "step": 2182 + }, + { + "epoch": 1.5097700155628566, + "grad_norm": 0.8305802345275879, + "learning_rate": 2.5846377866763744e-05, + "loss": 1.7852, + "step": 2183 + }, + { + "epoch": 1.5104616980805812, + "grad_norm": 1.6766009330749512, + "learning_rate": 2.5809974517655628e-05, + "loss": 2.2966, + "step": 2184 + }, + { + "epoch": 1.5111533805983055, + "grad_norm": 0.9089788198471069, + "learning_rate": 2.5773571168547506e-05, + "loss": 1.5303, + "step": 2185 + }, + { + "epoch": 1.5118450631160298, + "grad_norm": 0.7708346843719482, + "learning_rate": 2.573716781943939e-05, + "loss": 0.8766, + "step": 2186 + }, + { + "epoch": 1.5125367456337542, + "grad_norm": 1.0795036554336548, + "learning_rate": 2.5700764470331274e-05, + "loss": 2.0073, + "step": 2187 + }, + { + "epoch": 1.5132284281514785, + "grad_norm": 1.0379574298858643, + "learning_rate": 2.5664361121223152e-05, + "loss": 1.9833, + "step": 2188 + }, + { + "epoch": 1.5139201106692028, + "grad_norm": 1.0953075885772705, + "learning_rate": 2.5627957772115036e-05, + "loss": 1.5865, + "step": 2189 + }, + { + "epoch": 1.5146117931869272, + "grad_norm": 0.9112504124641418, + "learning_rate": 2.559155442300692e-05, + "loss": 0.9652, + "step": 2190 + }, + { + "epoch": 1.5153034757046515, + "grad_norm": 0.9524039030075073, + "learning_rate": 2.5555151073898798e-05, + "loss": 1.1672, + "step": 2191 + }, + { + "epoch": 1.5159951582223758, + "grad_norm": 0.9237524271011353, + "learning_rate": 2.5518747724790682e-05, + "loss": 2.0534, + "step": 2192 + }, + { + "epoch": 1.5166868407401002, + "grad_norm": 0.8292364478111267, + "learning_rate": 2.5482344375682567e-05, + "loss": 1.8897, + "step": 2193 + }, + { + "epoch": 1.5173785232578245, + "grad_norm": 1.0136380195617676, + "learning_rate": 2.5445941026574444e-05, + "loss": 1.3357, + "step": 2194 + }, + { + "epoch": 1.518070205775549, + "grad_norm": 0.7681485414505005, + "learning_rate": 2.540953767746633e-05, + "loss": 1.8097, + "step": 2195 + }, + { + "epoch": 1.5187618882932734, + "grad_norm": 1.0271018743515015, + "learning_rate": 2.537313432835821e-05, + "loss": 1.5457, + "step": 2196 + }, + { + "epoch": 1.5194535708109977, + "grad_norm": 1.2221665382385254, + "learning_rate": 2.533673097925009e-05, + "loss": 1.6692, + "step": 2197 + }, + { + "epoch": 1.520145253328722, + "grad_norm": 0.8557051420211792, + "learning_rate": 2.5300327630141975e-05, + "loss": 1.4016, + "step": 2198 + }, + { + "epoch": 1.5208369358464466, + "grad_norm": 0.9504237174987793, + "learning_rate": 2.5263924281033856e-05, + "loss": 1.2589, + "step": 2199 + }, + { + "epoch": 1.521528618364171, + "grad_norm": 1.0647205114364624, + "learning_rate": 2.522752093192574e-05, + "loss": 0.982, + "step": 2200 + }, + { + "epoch": 1.5222203008818953, + "grad_norm": 0.8775227665901184, + "learning_rate": 2.5191117582817618e-05, + "loss": 1.3885, + "step": 2201 + }, + { + "epoch": 1.5229119833996196, + "grad_norm": 1.0006057024002075, + "learning_rate": 2.5154714233709502e-05, + "loss": 1.6314, + "step": 2202 + }, + { + "epoch": 1.523603665917344, + "grad_norm": 0.941718578338623, + "learning_rate": 2.5118310884601386e-05, + "loss": 1.6693, + "step": 2203 + }, + { + "epoch": 1.5242953484350683, + "grad_norm": 1.0183758735656738, + "learning_rate": 2.5081907535493264e-05, + "loss": 1.7892, + "step": 2204 + }, + { + "epoch": 1.5249870309527926, + "grad_norm": 1.084665060043335, + "learning_rate": 2.504550418638515e-05, + "loss": 1.441, + "step": 2205 + }, + { + "epoch": 1.525678713470517, + "grad_norm": 1.113876223564148, + "learning_rate": 2.5009100837277033e-05, + "loss": 1.868, + "step": 2206 + }, + { + "epoch": 1.5263703959882413, + "grad_norm": 1.0181273221969604, + "learning_rate": 2.4972697488168914e-05, + "loss": 1.9074, + "step": 2207 + }, + { + "epoch": 1.5270620785059656, + "grad_norm": 1.0213264226913452, + "learning_rate": 2.4936294139060795e-05, + "loss": 1.3006, + "step": 2208 + }, + { + "epoch": 1.52775376102369, + "grad_norm": 0.8019144535064697, + "learning_rate": 2.4899890789952676e-05, + "loss": 1.183, + "step": 2209 + }, + { + "epoch": 1.5284454435414145, + "grad_norm": 0.9494912028312683, + "learning_rate": 2.486348744084456e-05, + "loss": 1.9953, + "step": 2210 + }, + { + "epoch": 1.5291371260591389, + "grad_norm": 1.0867944955825806, + "learning_rate": 2.482708409173644e-05, + "loss": 1.4977, + "step": 2211 + }, + { + "epoch": 1.5298288085768632, + "grad_norm": 1.195742130279541, + "learning_rate": 2.4790680742628322e-05, + "loss": 1.712, + "step": 2212 + }, + { + "epoch": 1.5305204910945875, + "grad_norm": 0.9193386435508728, + "learning_rate": 2.4754277393520206e-05, + "loss": 1.6745, + "step": 2213 + }, + { + "epoch": 1.531212173612312, + "grad_norm": 1.0849440097808838, + "learning_rate": 2.4717874044412087e-05, + "loss": 1.3801, + "step": 2214 + }, + { + "epoch": 1.5319038561300364, + "grad_norm": 1.2076828479766846, + "learning_rate": 2.4681470695303968e-05, + "loss": 1.5294, + "step": 2215 + }, + { + "epoch": 1.5325955386477608, + "grad_norm": 0.8005238175392151, + "learning_rate": 2.4645067346195852e-05, + "loss": 1.2335, + "step": 2216 + }, + { + "epoch": 1.533287221165485, + "grad_norm": 1.2941259145736694, + "learning_rate": 2.4608663997087733e-05, + "loss": 1.1751, + "step": 2217 + }, + { + "epoch": 1.5339789036832094, + "grad_norm": 0.7172979712486267, + "learning_rate": 2.4572260647979614e-05, + "loss": 1.4751, + "step": 2218 + }, + { + "epoch": 1.5346705862009338, + "grad_norm": 0.7608055472373962, + "learning_rate": 2.45358572988715e-05, + "loss": 1.3703, + "step": 2219 + }, + { + "epoch": 1.535362268718658, + "grad_norm": 0.921215832233429, + "learning_rate": 2.449945394976338e-05, + "loss": 1.8559, + "step": 2220 + }, + { + "epoch": 1.5360539512363824, + "grad_norm": 0.9999344944953918, + "learning_rate": 2.446305060065526e-05, + "loss": 1.6561, + "step": 2221 + }, + { + "epoch": 1.5367456337541068, + "grad_norm": 0.7347996830940247, + "learning_rate": 2.4426647251547145e-05, + "loss": 1.3104, + "step": 2222 + }, + { + "epoch": 1.5374373162718311, + "grad_norm": 1.355538010597229, + "learning_rate": 2.4390243902439026e-05, + "loss": 1.5452, + "step": 2223 + }, + { + "epoch": 1.5381289987895554, + "grad_norm": 1.40723717212677, + "learning_rate": 2.435384055333091e-05, + "loss": 1.5086, + "step": 2224 + }, + { + "epoch": 1.53882068130728, + "grad_norm": 0.8438430428504944, + "learning_rate": 2.431743720422279e-05, + "loss": 1.0649, + "step": 2225 + }, + { + "epoch": 1.5395123638250043, + "grad_norm": 0.8540118336677551, + "learning_rate": 2.4281033855114672e-05, + "loss": 1.6785, + "step": 2226 + }, + { + "epoch": 1.5402040463427287, + "grad_norm": 0.722993016242981, + "learning_rate": 2.4244630506006556e-05, + "loss": 1.2971, + "step": 2227 + }, + { + "epoch": 1.540895728860453, + "grad_norm": 0.9366788864135742, + "learning_rate": 2.4208227156898437e-05, + "loss": 1.6815, + "step": 2228 + }, + { + "epoch": 1.5415874113781776, + "grad_norm": 1.4829459190368652, + "learning_rate": 2.4171823807790318e-05, + "loss": 1.9242, + "step": 2229 + }, + { + "epoch": 1.542279093895902, + "grad_norm": 0.9441123604774475, + "learning_rate": 2.41354204586822e-05, + "loss": 1.7927, + "step": 2230 + }, + { + "epoch": 1.5429707764136262, + "grad_norm": 1.15497624874115, + "learning_rate": 2.4099017109574083e-05, + "loss": 2.0602, + "step": 2231 + }, + { + "epoch": 1.5436624589313506, + "grad_norm": 1.0702331066131592, + "learning_rate": 2.4062613760465964e-05, + "loss": 1.85, + "step": 2232 + }, + { + "epoch": 1.544354141449075, + "grad_norm": 0.9456023573875427, + "learning_rate": 2.4026210411357845e-05, + "loss": 1.54, + "step": 2233 + }, + { + "epoch": 1.5450458239667992, + "grad_norm": 1.041280746459961, + "learning_rate": 2.3989807062249726e-05, + "loss": 1.5169, + "step": 2234 + }, + { + "epoch": 1.5457375064845236, + "grad_norm": 1.4801952838897705, + "learning_rate": 2.3953403713141607e-05, + "loss": 1.1921, + "step": 2235 + }, + { + "epoch": 1.546429189002248, + "grad_norm": 1.2367304563522339, + "learning_rate": 2.391700036403349e-05, + "loss": 1.5813, + "step": 2236 + }, + { + "epoch": 1.5471208715199722, + "grad_norm": 1.2282288074493408, + "learning_rate": 2.3880597014925373e-05, + "loss": 1.7608, + "step": 2237 + }, + { + "epoch": 1.5478125540376966, + "grad_norm": 0.7776118516921997, + "learning_rate": 2.3844193665817257e-05, + "loss": 1.3951, + "step": 2238 + }, + { + "epoch": 1.548504236555421, + "grad_norm": 0.9500800371170044, + "learning_rate": 2.3807790316709138e-05, + "loss": 1.2471, + "step": 2239 + }, + { + "epoch": 1.5491959190731455, + "grad_norm": 2.875310182571411, + "learning_rate": 2.377138696760102e-05, + "loss": 1.2059, + "step": 2240 + }, + { + "epoch": 1.5498876015908698, + "grad_norm": 0.7105834484100342, + "learning_rate": 2.3734983618492903e-05, + "loss": 0.9645, + "step": 2241 + }, + { + "epoch": 1.5505792841085941, + "grad_norm": 0.9164943099021912, + "learning_rate": 2.3698580269384784e-05, + "loss": 1.4913, + "step": 2242 + }, + { + "epoch": 1.5512709666263185, + "grad_norm": 0.9220166206359863, + "learning_rate": 2.3662176920276665e-05, + "loss": 1.8357, + "step": 2243 + }, + { + "epoch": 1.551962649144043, + "grad_norm": 1.4042824506759644, + "learning_rate": 2.362577357116855e-05, + "loss": 1.3325, + "step": 2244 + }, + { + "epoch": 1.5526543316617674, + "grad_norm": 1.2085990905761719, + "learning_rate": 2.358937022206043e-05, + "loss": 1.7539, + "step": 2245 + }, + { + "epoch": 1.5533460141794917, + "grad_norm": 0.8625283241271973, + "learning_rate": 2.355296687295231e-05, + "loss": 1.941, + "step": 2246 + }, + { + "epoch": 1.554037696697216, + "grad_norm": 1.0419166088104248, + "learning_rate": 2.3516563523844196e-05, + "loss": 1.8716, + "step": 2247 + }, + { + "epoch": 1.5547293792149404, + "grad_norm": 1.1237373352050781, + "learning_rate": 2.3480160174736077e-05, + "loss": 1.8145, + "step": 2248 + }, + { + "epoch": 1.5554210617326647, + "grad_norm": 0.8355576992034912, + "learning_rate": 2.3443756825627957e-05, + "loss": 1.8728, + "step": 2249 + }, + { + "epoch": 1.556112744250389, + "grad_norm": 0.9098324179649353, + "learning_rate": 2.3407353476519842e-05, + "loss": 1.521, + "step": 2250 + }, + { + "epoch": 1.5568044267681134, + "grad_norm": 0.8380647897720337, + "learning_rate": 2.3370950127411723e-05, + "loss": 1.4171, + "step": 2251 + }, + { + "epoch": 1.5574961092858377, + "grad_norm": 0.9279719591140747, + "learning_rate": 2.3334546778303604e-05, + "loss": 1.8595, + "step": 2252 + }, + { + "epoch": 1.558187791803562, + "grad_norm": 1.0741068124771118, + "learning_rate": 2.3298143429195488e-05, + "loss": 1.2131, + "step": 2253 + }, + { + "epoch": 1.5588794743212864, + "grad_norm": 0.8023629188537598, + "learning_rate": 2.326174008008737e-05, + "loss": 1.4076, + "step": 2254 + }, + { + "epoch": 1.559571156839011, + "grad_norm": 0.8634189367294312, + "learning_rate": 2.3225336730979253e-05, + "loss": 1.5695, + "step": 2255 + }, + { + "epoch": 1.5602628393567353, + "grad_norm": 0.8278612494468689, + "learning_rate": 2.3188933381871134e-05, + "loss": 1.8979, + "step": 2256 + }, + { + "epoch": 1.5609545218744596, + "grad_norm": 0.7097908854484558, + "learning_rate": 2.3152530032763015e-05, + "loss": 1.2414, + "step": 2257 + }, + { + "epoch": 1.561646204392184, + "grad_norm": 0.9384331703186035, + "learning_rate": 2.31161266836549e-05, + "loss": 1.8887, + "step": 2258 + }, + { + "epoch": 1.5623378869099085, + "grad_norm": 1.1173044443130493, + "learning_rate": 2.307972333454678e-05, + "loss": 2.0172, + "step": 2259 + }, + { + "epoch": 1.5630295694276328, + "grad_norm": 0.9721001982688904, + "learning_rate": 2.304331998543866e-05, + "loss": 1.512, + "step": 2260 + }, + { + "epoch": 1.5637212519453572, + "grad_norm": 0.851543128490448, + "learning_rate": 2.3006916636330546e-05, + "loss": 1.3716, + "step": 2261 + }, + { + "epoch": 1.5644129344630815, + "grad_norm": 0.8371115922927856, + "learning_rate": 2.2970513287222427e-05, + "loss": 1.4036, + "step": 2262 + }, + { + "epoch": 1.5651046169808058, + "grad_norm": 0.9390556216239929, + "learning_rate": 2.2934109938114308e-05, + "loss": 1.7812, + "step": 2263 + }, + { + "epoch": 1.5657962994985302, + "grad_norm": 0.8929088711738586, + "learning_rate": 2.2897706589006192e-05, + "loss": 1.9498, + "step": 2264 + }, + { + "epoch": 1.5664879820162545, + "grad_norm": 0.8331410884857178, + "learning_rate": 2.2861303239898073e-05, + "loss": 1.5209, + "step": 2265 + }, + { + "epoch": 1.5671796645339788, + "grad_norm": 1.0900577306747437, + "learning_rate": 2.2824899890789954e-05, + "loss": 1.6262, + "step": 2266 + }, + { + "epoch": 1.5678713470517032, + "grad_norm": 1.0398414134979248, + "learning_rate": 2.2788496541681835e-05, + "loss": 1.8023, + "step": 2267 + }, + { + "epoch": 1.5685630295694275, + "grad_norm": 0.7122235894203186, + "learning_rate": 2.2752093192573716e-05, + "loss": 1.1728, + "step": 2268 + }, + { + "epoch": 1.5692547120871518, + "grad_norm": 1.1669389009475708, + "learning_rate": 2.27156898434656e-05, + "loss": 1.9754, + "step": 2269 + }, + { + "epoch": 1.5699463946048764, + "grad_norm": 1.1289633512496948, + "learning_rate": 2.267928649435748e-05, + "loss": 1.9609, + "step": 2270 + }, + { + "epoch": 1.5706380771226007, + "grad_norm": 0.7589194774627686, + "learning_rate": 2.2642883145249362e-05, + "loss": 1.0469, + "step": 2271 + }, + { + "epoch": 1.571329759640325, + "grad_norm": 1.0977452993392944, + "learning_rate": 2.2606479796141246e-05, + "loss": 1.8149, + "step": 2272 + }, + { + "epoch": 1.5720214421580496, + "grad_norm": 1.1783114671707153, + "learning_rate": 2.2570076447033127e-05, + "loss": 1.3949, + "step": 2273 + }, + { + "epoch": 1.572713124675774, + "grad_norm": 1.1496466398239136, + "learning_rate": 2.253367309792501e-05, + "loss": 1.7596, + "step": 2274 + }, + { + "epoch": 1.5734048071934983, + "grad_norm": 0.9879779815673828, + "learning_rate": 2.2497269748816893e-05, + "loss": 1.5737, + "step": 2275 + }, + { + "epoch": 1.5740964897112226, + "grad_norm": 1.0273772478103638, + "learning_rate": 2.2460866399708774e-05, + "loss": 1.4897, + "step": 2276 + }, + { + "epoch": 1.574788172228947, + "grad_norm": 0.8760284185409546, + "learning_rate": 2.2424463050600655e-05, + "loss": 1.405, + "step": 2277 + }, + { + "epoch": 1.5754798547466713, + "grad_norm": 1.0425629615783691, + "learning_rate": 2.238805970149254e-05, + "loss": 1.4208, + "step": 2278 + }, + { + "epoch": 1.5761715372643956, + "grad_norm": 1.2762773036956787, + "learning_rate": 2.235165635238442e-05, + "loss": 1.3426, + "step": 2279 + }, + { + "epoch": 1.57686321978212, + "grad_norm": 0.8973204493522644, + "learning_rate": 2.23152530032763e-05, + "loss": 1.4547, + "step": 2280 + }, + { + "epoch": 1.5775549022998443, + "grad_norm": 1.0758110284805298, + "learning_rate": 2.2278849654168185e-05, + "loss": 1.649, + "step": 2281 + }, + { + "epoch": 1.5782465848175686, + "grad_norm": 0.9393559694290161, + "learning_rate": 2.2242446305060066e-05, + "loss": 1.722, + "step": 2282 + }, + { + "epoch": 1.578938267335293, + "grad_norm": 1.8895992040634155, + "learning_rate": 2.2206042955951947e-05, + "loss": 1.667, + "step": 2283 + }, + { + "epoch": 1.5796299498530173, + "grad_norm": 1.1462339162826538, + "learning_rate": 2.216963960684383e-05, + "loss": 1.6178, + "step": 2284 + }, + { + "epoch": 1.5803216323707419, + "grad_norm": 0.8290567994117737, + "learning_rate": 2.2133236257735712e-05, + "loss": 1.6832, + "step": 2285 + }, + { + "epoch": 1.5810133148884662, + "grad_norm": 0.8535481095314026, + "learning_rate": 2.2096832908627597e-05, + "loss": 1.6348, + "step": 2286 + }, + { + "epoch": 1.5817049974061905, + "grad_norm": 0.7832224369049072, + "learning_rate": 2.2060429559519478e-05, + "loss": 1.6979, + "step": 2287 + }, + { + "epoch": 1.582396679923915, + "grad_norm": 0.7348320484161377, + "learning_rate": 2.202402621041136e-05, + "loss": 1.7401, + "step": 2288 + }, + { + "epoch": 1.5830883624416394, + "grad_norm": 0.9076175689697266, + "learning_rate": 2.1987622861303243e-05, + "loss": 1.8583, + "step": 2289 + }, + { + "epoch": 1.5837800449593638, + "grad_norm": 1.0359013080596924, + "learning_rate": 2.1951219512195124e-05, + "loss": 1.3975, + "step": 2290 + }, + { + "epoch": 1.584471727477088, + "grad_norm": 1.7890931367874146, + "learning_rate": 2.1914816163087005e-05, + "loss": 1.311, + "step": 2291 + }, + { + "epoch": 1.5851634099948124, + "grad_norm": 0.8306867480278015, + "learning_rate": 2.187841281397889e-05, + "loss": 1.2419, + "step": 2292 + }, + { + "epoch": 1.5858550925125368, + "grad_norm": 1.2653504610061646, + "learning_rate": 2.184200946487077e-05, + "loss": 1.6349, + "step": 2293 + }, + { + "epoch": 1.586546775030261, + "grad_norm": 0.7140494585037231, + "learning_rate": 2.180560611576265e-05, + "loss": 0.9266, + "step": 2294 + }, + { + "epoch": 1.5872384575479854, + "grad_norm": 1.0886075496673584, + "learning_rate": 2.1769202766654535e-05, + "loss": 1.101, + "step": 2295 + }, + { + "epoch": 1.5879301400657098, + "grad_norm": 0.8796497583389282, + "learning_rate": 2.1732799417546416e-05, + "loss": 1.4517, + "step": 2296 + }, + { + "epoch": 1.588621822583434, + "grad_norm": 1.195204734802246, + "learning_rate": 2.1696396068438297e-05, + "loss": 1.0265, + "step": 2297 + }, + { + "epoch": 1.5893135051011584, + "grad_norm": 0.9413547515869141, + "learning_rate": 2.165999271933018e-05, + "loss": 1.8275, + "step": 2298 + }, + { + "epoch": 1.590005187618883, + "grad_norm": 0.8125203847885132, + "learning_rate": 2.1623589370222063e-05, + "loss": 1.9603, + "step": 2299 + }, + { + "epoch": 1.5906968701366073, + "grad_norm": 0.8235089182853699, + "learning_rate": 2.1587186021113943e-05, + "loss": 1.6365, + "step": 2300 + }, + { + "epoch": 1.5913885526543317, + "grad_norm": 1.2348897457122803, + "learning_rate": 2.1550782672005824e-05, + "loss": 1.9568, + "step": 2301 + }, + { + "epoch": 1.592080235172056, + "grad_norm": 1.0779099464416504, + "learning_rate": 2.1514379322897705e-05, + "loss": 1.947, + "step": 2302 + }, + { + "epoch": 1.5927719176897805, + "grad_norm": 0.8659629821777344, + "learning_rate": 2.147797597378959e-05, + "loss": 1.6857, + "step": 2303 + }, + { + "epoch": 1.5934636002075049, + "grad_norm": 0.984544575214386, + "learning_rate": 2.144157262468147e-05, + "loss": 1.8058, + "step": 2304 + }, + { + "epoch": 1.5941552827252292, + "grad_norm": 1.080067753791809, + "learning_rate": 2.140516927557335e-05, + "loss": 1.0279, + "step": 2305 + }, + { + "epoch": 1.5948469652429536, + "grad_norm": 0.9962109923362732, + "learning_rate": 2.1368765926465236e-05, + "loss": 1.6811, + "step": 2306 + }, + { + "epoch": 1.5955386477606779, + "grad_norm": 0.8938770294189453, + "learning_rate": 2.1332362577357117e-05, + "loss": 1.5296, + "step": 2307 + }, + { + "epoch": 1.5962303302784022, + "grad_norm": 1.041804313659668, + "learning_rate": 2.1295959228248998e-05, + "loss": 1.6468, + "step": 2308 + }, + { + "epoch": 1.5969220127961266, + "grad_norm": 1.3163366317749023, + "learning_rate": 2.1259555879140882e-05, + "loss": 1.1924, + "step": 2309 + }, + { + "epoch": 1.597613695313851, + "grad_norm": 1.208646297454834, + "learning_rate": 2.1223152530032763e-05, + "loss": 1.9227, + "step": 2310 + }, + { + "epoch": 1.5983053778315752, + "grad_norm": 1.1080639362335205, + "learning_rate": 2.1186749180924644e-05, + "loss": 1.9495, + "step": 2311 + }, + { + "epoch": 1.5989970603492996, + "grad_norm": 1.3428796529769897, + "learning_rate": 2.115034583181653e-05, + "loss": 1.7533, + "step": 2312 + }, + { + "epoch": 1.599688742867024, + "grad_norm": 0.9917942881584167, + "learning_rate": 2.111394248270841e-05, + "loss": 1.9576, + "step": 2313 + }, + { + "epoch": 1.6003804253847485, + "grad_norm": 0.8607562184333801, + "learning_rate": 2.107753913360029e-05, + "loss": 1.3447, + "step": 2314 + }, + { + "epoch": 1.6010721079024728, + "grad_norm": 0.8866775631904602, + "learning_rate": 2.1041135784492175e-05, + "loss": 1.2276, + "step": 2315 + }, + { + "epoch": 1.6017637904201971, + "grad_norm": 1.1829619407653809, + "learning_rate": 2.1004732435384056e-05, + "loss": 2.0294, + "step": 2316 + }, + { + "epoch": 1.6024554729379215, + "grad_norm": 0.8793166875839233, + "learning_rate": 2.096832908627594e-05, + "loss": 1.8243, + "step": 2317 + }, + { + "epoch": 1.603147155455646, + "grad_norm": 0.6399521231651306, + "learning_rate": 2.093192573716782e-05, + "loss": 1.1092, + "step": 2318 + }, + { + "epoch": 1.6038388379733703, + "grad_norm": 1.0232120752334595, + "learning_rate": 2.0895522388059702e-05, + "loss": 1.6328, + "step": 2319 + }, + { + "epoch": 1.6045305204910947, + "grad_norm": 1.0168575048446655, + "learning_rate": 2.0859119038951586e-05, + "loss": 1.4407, + "step": 2320 + }, + { + "epoch": 1.605222203008819, + "grad_norm": 1.2204807996749878, + "learning_rate": 2.0822715689843467e-05, + "loss": 1.4877, + "step": 2321 + }, + { + "epoch": 1.6059138855265433, + "grad_norm": 0.8492107391357422, + "learning_rate": 2.0786312340735348e-05, + "loss": 1.4081, + "step": 2322 + }, + { + "epoch": 1.6066055680442677, + "grad_norm": 0.8629893064498901, + "learning_rate": 2.0749908991627232e-05, + "loss": 1.2145, + "step": 2323 + }, + { + "epoch": 1.607297250561992, + "grad_norm": 0.9798515439033508, + "learning_rate": 2.0713505642519113e-05, + "loss": 1.4678, + "step": 2324 + }, + { + "epoch": 1.6079889330797164, + "grad_norm": 0.8434598445892334, + "learning_rate": 2.0677102293410994e-05, + "loss": 1.6696, + "step": 2325 + }, + { + "epoch": 1.6086806155974407, + "grad_norm": 0.8641858100891113, + "learning_rate": 2.064069894430288e-05, + "loss": 1.8231, + "step": 2326 + }, + { + "epoch": 1.609372298115165, + "grad_norm": 1.0745980739593506, + "learning_rate": 2.060429559519476e-05, + "loss": 1.5422, + "step": 2327 + }, + { + "epoch": 1.6100639806328894, + "grad_norm": 1.2168126106262207, + "learning_rate": 2.056789224608664e-05, + "loss": 1.8395, + "step": 2328 + }, + { + "epoch": 1.610755663150614, + "grad_norm": 0.8866637349128723, + "learning_rate": 2.0531488896978525e-05, + "loss": 1.4186, + "step": 2329 + }, + { + "epoch": 1.6114473456683382, + "grad_norm": 0.8915860652923584, + "learning_rate": 2.0495085547870406e-05, + "loss": 1.0821, + "step": 2330 + }, + { + "epoch": 1.6121390281860626, + "grad_norm": 0.7546595931053162, + "learning_rate": 2.0458682198762287e-05, + "loss": 1.5044, + "step": 2331 + }, + { + "epoch": 1.612830710703787, + "grad_norm": 1.006158709526062, + "learning_rate": 2.042227884965417e-05, + "loss": 1.4342, + "step": 2332 + }, + { + "epoch": 1.6135223932215115, + "grad_norm": 1.2282028198242188, + "learning_rate": 2.0385875500546052e-05, + "loss": 1.6489, + "step": 2333 + }, + { + "epoch": 1.6142140757392358, + "grad_norm": 0.7992381453514099, + "learning_rate": 2.0349472151437933e-05, + "loss": 1.6385, + "step": 2334 + }, + { + "epoch": 1.6149057582569601, + "grad_norm": 0.9843496680259705, + "learning_rate": 2.0313068802329814e-05, + "loss": 1.2549, + "step": 2335 + }, + { + "epoch": 1.6155974407746845, + "grad_norm": 0.9226531982421875, + "learning_rate": 2.0276665453221695e-05, + "loss": 1.4241, + "step": 2336 + }, + { + "epoch": 1.6162891232924088, + "grad_norm": 1.138067364692688, + "learning_rate": 2.024026210411358e-05, + "loss": 1.7577, + "step": 2337 + }, + { + "epoch": 1.6169808058101331, + "grad_norm": 0.8949439525604248, + "learning_rate": 2.020385875500546e-05, + "loss": 1.6527, + "step": 2338 + }, + { + "epoch": 1.6176724883278575, + "grad_norm": 0.898450493812561, + "learning_rate": 2.016745540589734e-05, + "loss": 0.8489, + "step": 2339 + }, + { + "epoch": 1.6183641708455818, + "grad_norm": 0.9169157147407532, + "learning_rate": 2.0131052056789225e-05, + "loss": 1.7538, + "step": 2340 + }, + { + "epoch": 1.6190558533633062, + "grad_norm": 1.51553475856781, + "learning_rate": 2.0094648707681106e-05, + "loss": 1.4571, + "step": 2341 + }, + { + "epoch": 1.6197475358810305, + "grad_norm": 0.8794975280761719, + "learning_rate": 2.0058245358572987e-05, + "loss": 1.3612, + "step": 2342 + }, + { + "epoch": 1.6204392183987548, + "grad_norm": 0.9341661930084229, + "learning_rate": 2.0021842009464872e-05, + "loss": 2.1021, + "step": 2343 + }, + { + "epoch": 1.6211309009164794, + "grad_norm": 0.9817913174629211, + "learning_rate": 1.9985438660356753e-05, + "loss": 1.092, + "step": 2344 + }, + { + "epoch": 1.6218225834342037, + "grad_norm": 0.8946571350097656, + "learning_rate": 1.9949035311248634e-05, + "loss": 2.1064, + "step": 2345 + }, + { + "epoch": 1.622514265951928, + "grad_norm": 0.9536589980125427, + "learning_rate": 1.9912631962140518e-05, + "loss": 1.4338, + "step": 2346 + }, + { + "epoch": 1.6232059484696524, + "grad_norm": 0.9259676337242126, + "learning_rate": 1.98762286130324e-05, + "loss": 1.4849, + "step": 2347 + }, + { + "epoch": 1.623897630987377, + "grad_norm": 0.9228977560997009, + "learning_rate": 1.9839825263924283e-05, + "loss": 1.9277, + "step": 2348 + }, + { + "epoch": 1.6245893135051013, + "grad_norm": 1.0991334915161133, + "learning_rate": 1.9803421914816164e-05, + "loss": 1.678, + "step": 2349 + }, + { + "epoch": 1.6252809960228256, + "grad_norm": 1.80892014503479, + "learning_rate": 1.9767018565708045e-05, + "loss": 1.5363, + "step": 2350 + }, + { + "epoch": 1.62597267854055, + "grad_norm": 0.9532687067985535, + "learning_rate": 1.973061521659993e-05, + "loss": 1.6928, + "step": 2351 + }, + { + "epoch": 1.6266643610582743, + "grad_norm": 0.8402472138404846, + "learning_rate": 1.969421186749181e-05, + "loss": 1.2206, + "step": 2352 + }, + { + "epoch": 1.6273560435759986, + "grad_norm": 1.1433659791946411, + "learning_rate": 1.965780851838369e-05, + "loss": 1.3224, + "step": 2353 + }, + { + "epoch": 1.628047726093723, + "grad_norm": 0.7204102277755737, + "learning_rate": 1.9621405169275576e-05, + "loss": 1.41, + "step": 2354 + }, + { + "epoch": 1.6287394086114473, + "grad_norm": 0.9019567966461182, + "learning_rate": 1.9585001820167457e-05, + "loss": 1.8849, + "step": 2355 + }, + { + "epoch": 1.6294310911291716, + "grad_norm": 0.8422867655754089, + "learning_rate": 1.9548598471059338e-05, + "loss": 1.3503, + "step": 2356 + }, + { + "epoch": 1.630122773646896, + "grad_norm": 0.8603572845458984, + "learning_rate": 1.9512195121951222e-05, + "loss": 1.3527, + "step": 2357 + }, + { + "epoch": 1.6308144561646203, + "grad_norm": 0.9578169584274292, + "learning_rate": 1.9475791772843103e-05, + "loss": 0.8533, + "step": 2358 + }, + { + "epoch": 1.6315061386823448, + "grad_norm": 0.8656398057937622, + "learning_rate": 1.9439388423734984e-05, + "loss": 1.0955, + "step": 2359 + }, + { + "epoch": 1.6321978212000692, + "grad_norm": 1.182929277420044, + "learning_rate": 1.9402985074626868e-05, + "loss": 1.6282, + "step": 2360 + }, + { + "epoch": 1.6328895037177935, + "grad_norm": 0.8781986832618713, + "learning_rate": 1.936658172551875e-05, + "loss": 2.0018, + "step": 2361 + }, + { + "epoch": 1.6335811862355178, + "grad_norm": 0.9545982480049133, + "learning_rate": 1.9330178376410633e-05, + "loss": 1.2459, + "step": 2362 + }, + { + "epoch": 1.6342728687532424, + "grad_norm": 1.070780634880066, + "learning_rate": 1.9293775027302514e-05, + "loss": 1.0964, + "step": 2363 + }, + { + "epoch": 1.6349645512709667, + "grad_norm": 0.8855124115943909, + "learning_rate": 1.9257371678194395e-05, + "loss": 1.4871, + "step": 2364 + }, + { + "epoch": 1.635656233788691, + "grad_norm": 0.9848822951316833, + "learning_rate": 1.9220968329086276e-05, + "loss": 1.7118, + "step": 2365 + }, + { + "epoch": 1.6363479163064154, + "grad_norm": 1.001775860786438, + "learning_rate": 1.918456497997816e-05, + "loss": 1.5732, + "step": 2366 + }, + { + "epoch": 1.6370395988241397, + "grad_norm": 0.8310945630073547, + "learning_rate": 1.914816163087004e-05, + "loss": 1.2186, + "step": 2367 + }, + { + "epoch": 1.637731281341864, + "grad_norm": 1.0571595430374146, + "learning_rate": 1.9111758281761923e-05, + "loss": 1.3877, + "step": 2368 + }, + { + "epoch": 1.6384229638595884, + "grad_norm": 1.4988269805908203, + "learning_rate": 1.9075354932653803e-05, + "loss": 1.5295, + "step": 2369 + }, + { + "epoch": 1.6391146463773127, + "grad_norm": 1.0751508474349976, + "learning_rate": 1.9038951583545684e-05, + "loss": 0.7709, + "step": 2370 + }, + { + "epoch": 1.639806328895037, + "grad_norm": 1.3799232244491577, + "learning_rate": 1.900254823443757e-05, + "loss": 1.7996, + "step": 2371 + }, + { + "epoch": 1.6404980114127614, + "grad_norm": 1.0418415069580078, + "learning_rate": 1.896614488532945e-05, + "loss": 1.4951, + "step": 2372 + }, + { + "epoch": 1.6411896939304857, + "grad_norm": 1.0699201822280884, + "learning_rate": 1.892974153622133e-05, + "loss": 0.9344, + "step": 2373 + }, + { + "epoch": 1.6418813764482103, + "grad_norm": 0.9429408311843872, + "learning_rate": 1.8893338187113215e-05, + "loss": 1.9607, + "step": 2374 + }, + { + "epoch": 1.6425730589659346, + "grad_norm": 0.9036697745323181, + "learning_rate": 1.8856934838005096e-05, + "loss": 1.8237, + "step": 2375 + }, + { + "epoch": 1.643264741483659, + "grad_norm": 1.2165155410766602, + "learning_rate": 1.8820531488896977e-05, + "loss": 1.8432, + "step": 2376 + }, + { + "epoch": 1.6439564240013833, + "grad_norm": 0.8559542894363403, + "learning_rate": 1.878412813978886e-05, + "loss": 1.0861, + "step": 2377 + }, + { + "epoch": 1.6446481065191079, + "grad_norm": 0.9780797362327576, + "learning_rate": 1.8747724790680742e-05, + "loss": 1.7229, + "step": 2378 + }, + { + "epoch": 1.6453397890368322, + "grad_norm": 0.8456622958183289, + "learning_rate": 1.8711321441572627e-05, + "loss": 1.6706, + "step": 2379 + }, + { + "epoch": 1.6460314715545565, + "grad_norm": 1.3006799221038818, + "learning_rate": 1.8674918092464507e-05, + "loss": 1.1094, + "step": 2380 + }, + { + "epoch": 1.6467231540722809, + "grad_norm": 0.8623272180557251, + "learning_rate": 1.863851474335639e-05, + "loss": 1.3869, + "step": 2381 + }, + { + "epoch": 1.6474148365900052, + "grad_norm": 0.9243130683898926, + "learning_rate": 1.8602111394248273e-05, + "loss": 1.7079, + "step": 2382 + }, + { + "epoch": 1.6481065191077295, + "grad_norm": 0.8512841463088989, + "learning_rate": 1.8565708045140154e-05, + "loss": 1.9186, + "step": 2383 + }, + { + "epoch": 1.6487982016254539, + "grad_norm": 1.1149475574493408, + "learning_rate": 1.8529304696032035e-05, + "loss": 1.9283, + "step": 2384 + }, + { + "epoch": 1.6494898841431782, + "grad_norm": 0.8714730739593506, + "learning_rate": 1.849290134692392e-05, + "loss": 1.445, + "step": 2385 + }, + { + "epoch": 1.6501815666609025, + "grad_norm": 0.923741340637207, + "learning_rate": 1.84564979978158e-05, + "loss": 1.6285, + "step": 2386 + }, + { + "epoch": 1.6508732491786269, + "grad_norm": 1.2058165073394775, + "learning_rate": 1.842009464870768e-05, + "loss": 1.5193, + "step": 2387 + }, + { + "epoch": 1.6515649316963512, + "grad_norm": 1.027906894683838, + "learning_rate": 1.8383691299599565e-05, + "loss": 1.6151, + "step": 2388 + }, + { + "epoch": 1.6522566142140758, + "grad_norm": 0.8688392639160156, + "learning_rate": 1.8347287950491446e-05, + "loss": 1.3428, + "step": 2389 + }, + { + "epoch": 1.6529482967318, + "grad_norm": 0.9275920987129211, + "learning_rate": 1.8310884601383327e-05, + "loss": 1.7093, + "step": 2390 + }, + { + "epoch": 1.6536399792495244, + "grad_norm": 0.9836880564689636, + "learning_rate": 1.827448125227521e-05, + "loss": 1.8134, + "step": 2391 + }, + { + "epoch": 1.654331661767249, + "grad_norm": 0.8987468481063843, + "learning_rate": 1.8238077903167092e-05, + "loss": 1.2157, + "step": 2392 + }, + { + "epoch": 1.6550233442849733, + "grad_norm": 0.7679839134216309, + "learning_rate": 1.8201674554058977e-05, + "loss": 1.3538, + "step": 2393 + }, + { + "epoch": 1.6557150268026977, + "grad_norm": 1.0226809978485107, + "learning_rate": 1.8165271204950858e-05, + "loss": 1.9941, + "step": 2394 + }, + { + "epoch": 1.656406709320422, + "grad_norm": 0.9852961897850037, + "learning_rate": 1.812886785584274e-05, + "loss": 1.7164, + "step": 2395 + }, + { + "epoch": 1.6570983918381463, + "grad_norm": 1.0636413097381592, + "learning_rate": 1.8092464506734623e-05, + "loss": 1.1651, + "step": 2396 + }, + { + "epoch": 1.6577900743558707, + "grad_norm": 1.0011451244354248, + "learning_rate": 1.8056061157626504e-05, + "loss": 1.868, + "step": 2397 + }, + { + "epoch": 1.658481756873595, + "grad_norm": 0.939164936542511, + "learning_rate": 1.8019657808518385e-05, + "loss": 0.8075, + "step": 2398 + }, + { + "epoch": 1.6591734393913193, + "grad_norm": 0.941457211971283, + "learning_rate": 1.798325445941027e-05, + "loss": 1.9064, + "step": 2399 + }, + { + "epoch": 1.6598651219090437, + "grad_norm": 1.0357470512390137, + "learning_rate": 1.794685111030215e-05, + "loss": 1.3495, + "step": 2400 + }, + { + "epoch": 1.660556804426768, + "grad_norm": 0.9081909656524658, + "learning_rate": 1.791044776119403e-05, + "loss": 1.9126, + "step": 2401 + }, + { + "epoch": 1.6612484869444923, + "grad_norm": 0.9275761842727661, + "learning_rate": 1.7874044412085912e-05, + "loss": 1.6704, + "step": 2402 + }, + { + "epoch": 1.6619401694622167, + "grad_norm": 1.0169730186462402, + "learning_rate": 1.7837641062977793e-05, + "loss": 1.9949, + "step": 2403 + }, + { + "epoch": 1.6626318519799412, + "grad_norm": 1.1368589401245117, + "learning_rate": 1.7801237713869674e-05, + "loss": 1.6442, + "step": 2404 + }, + { + "epoch": 1.6633235344976656, + "grad_norm": 0.8694601058959961, + "learning_rate": 1.7764834364761558e-05, + "loss": 1.3597, + "step": 2405 + }, + { + "epoch": 1.66401521701539, + "grad_norm": 0.7513114213943481, + "learning_rate": 1.772843101565344e-05, + "loss": 1.3165, + "step": 2406 + }, + { + "epoch": 1.6647068995331145, + "grad_norm": 0.989966094493866, + "learning_rate": 1.7692027666545324e-05, + "loss": 2.0479, + "step": 2407 + }, + { + "epoch": 1.6653985820508388, + "grad_norm": 1.7258808612823486, + "learning_rate": 1.7655624317437205e-05, + "loss": 1.9259, + "step": 2408 + }, + { + "epoch": 1.6660902645685631, + "grad_norm": 0.946379542350769, + "learning_rate": 1.7619220968329085e-05, + "loss": 1.206, + "step": 2409 + }, + { + "epoch": 1.6667819470862875, + "grad_norm": 1.1271051168441772, + "learning_rate": 1.758281761922097e-05, + "loss": 1.0387, + "step": 2410 + }, + { + "epoch": 1.6674736296040118, + "grad_norm": 1.1737791299819946, + "learning_rate": 1.754641427011285e-05, + "loss": 1.2887, + "step": 2411 + }, + { + "epoch": 1.6681653121217361, + "grad_norm": 0.9748665690422058, + "learning_rate": 1.7510010921004732e-05, + "loss": 1.8055, + "step": 2412 + }, + { + "epoch": 1.6688569946394605, + "grad_norm": 0.8215218782424927, + "learning_rate": 1.7473607571896616e-05, + "loss": 1.2058, + "step": 2413 + }, + { + "epoch": 1.6695486771571848, + "grad_norm": 0.8643574714660645, + "learning_rate": 1.7437204222788497e-05, + "loss": 1.1155, + "step": 2414 + }, + { + "epoch": 1.6702403596749091, + "grad_norm": 0.9250410795211792, + "learning_rate": 1.7400800873680378e-05, + "loss": 1.9442, + "step": 2415 + }, + { + "epoch": 1.6709320421926335, + "grad_norm": 0.8802753686904907, + "learning_rate": 1.7364397524572262e-05, + "loss": 1.9279, + "step": 2416 + }, + { + "epoch": 1.6716237247103578, + "grad_norm": 1.0128893852233887, + "learning_rate": 1.7327994175464143e-05, + "loss": 1.5791, + "step": 2417 + }, + { + "epoch": 1.6723154072280824, + "grad_norm": 0.9950559139251709, + "learning_rate": 1.7291590826356024e-05, + "loss": 1.5734, + "step": 2418 + }, + { + "epoch": 1.6730070897458067, + "grad_norm": 0.934482991695404, + "learning_rate": 1.725518747724791e-05, + "loss": 1.7531, + "step": 2419 + }, + { + "epoch": 1.673698772263531, + "grad_norm": 1.2302453517913818, + "learning_rate": 1.721878412813979e-05, + "loss": 1.7979, + "step": 2420 + }, + { + "epoch": 1.6743904547812554, + "grad_norm": 1.6799739599227905, + "learning_rate": 1.718238077903167e-05, + "loss": 1.5029, + "step": 2421 + }, + { + "epoch": 1.67508213729898, + "grad_norm": 1.1403224468231201, + "learning_rate": 1.7145977429923555e-05, + "loss": 1.3767, + "step": 2422 + }, + { + "epoch": 1.6757738198167043, + "grad_norm": 1.005231499671936, + "learning_rate": 1.7109574080815436e-05, + "loss": 2.0528, + "step": 2423 + }, + { + "epoch": 1.6764655023344286, + "grad_norm": 1.0098246335983276, + "learning_rate": 1.707317073170732e-05, + "loss": 1.757, + "step": 2424 + }, + { + "epoch": 1.677157184852153, + "grad_norm": 0.870312511920929, + "learning_rate": 1.70367673825992e-05, + "loss": 1.0476, + "step": 2425 + }, + { + "epoch": 1.6778488673698773, + "grad_norm": 0.9728682637214661, + "learning_rate": 1.7000364033491082e-05, + "loss": 1.3328, + "step": 2426 + }, + { + "epoch": 1.6785405498876016, + "grad_norm": 1.094130277633667, + "learning_rate": 1.6963960684382966e-05, + "loss": 1.5296, + "step": 2427 + }, + { + "epoch": 1.679232232405326, + "grad_norm": 0.8974465131759644, + "learning_rate": 1.6927557335274847e-05, + "loss": 1.0728, + "step": 2428 + }, + { + "epoch": 1.6799239149230503, + "grad_norm": 0.9955480098724365, + "learning_rate": 1.6891153986166728e-05, + "loss": 1.2188, + "step": 2429 + }, + { + "epoch": 1.6806155974407746, + "grad_norm": 0.704064667224884, + "learning_rate": 1.6854750637058613e-05, + "loss": 0.9037, + "step": 2430 + }, + { + "epoch": 1.681307279958499, + "grad_norm": 0.8005505800247192, + "learning_rate": 1.6818347287950493e-05, + "loss": 1.0082, + "step": 2431 + }, + { + "epoch": 1.6819989624762233, + "grad_norm": 1.2470619678497314, + "learning_rate": 1.6781943938842374e-05, + "loss": 2.1728, + "step": 2432 + }, + { + "epoch": 1.6826906449939478, + "grad_norm": 0.8063885569572449, + "learning_rate": 1.674554058973426e-05, + "loss": 1.757, + "step": 2433 + }, + { + "epoch": 1.6833823275116722, + "grad_norm": 1.0915642976760864, + "learning_rate": 1.670913724062614e-05, + "loss": 1.4087, + "step": 2434 + }, + { + "epoch": 1.6840740100293965, + "grad_norm": 0.7927170395851135, + "learning_rate": 1.667273389151802e-05, + "loss": 1.1917, + "step": 2435 + }, + { + "epoch": 1.6847656925471208, + "grad_norm": 1.1966607570648193, + "learning_rate": 1.66363305424099e-05, + "loss": 1.7187, + "step": 2436 + }, + { + "epoch": 1.6854573750648454, + "grad_norm": 1.14612877368927, + "learning_rate": 1.6599927193301783e-05, + "loss": 1.7027, + "step": 2437 + }, + { + "epoch": 1.6861490575825697, + "grad_norm": 1.1655855178833008, + "learning_rate": 1.6563523844193667e-05, + "loss": 1.603, + "step": 2438 + }, + { + "epoch": 1.686840740100294, + "grad_norm": 0.9268144369125366, + "learning_rate": 1.6527120495085548e-05, + "loss": 1.7074, + "step": 2439 + }, + { + "epoch": 1.6875324226180184, + "grad_norm": 0.8364464640617371, + "learning_rate": 1.649071714597743e-05, + "loss": 1.3336, + "step": 2440 + }, + { + "epoch": 1.6882241051357427, + "grad_norm": 1.137802004814148, + "learning_rate": 1.6454313796869313e-05, + "loss": 1.913, + "step": 2441 + }, + { + "epoch": 1.688915787653467, + "grad_norm": 1.2386194467544556, + "learning_rate": 1.6417910447761194e-05, + "loss": 1.0058, + "step": 2442 + }, + { + "epoch": 1.6896074701711914, + "grad_norm": 1.1942088603973389, + "learning_rate": 1.6381507098653075e-05, + "loss": 1.5135, + "step": 2443 + }, + { + "epoch": 1.6902991526889157, + "grad_norm": 1.052964687347412, + "learning_rate": 1.634510374954496e-05, + "loss": 0.7595, + "step": 2444 + }, + { + "epoch": 1.69099083520664, + "grad_norm": 1.000148057937622, + "learning_rate": 1.630870040043684e-05, + "loss": 1.0357, + "step": 2445 + }, + { + "epoch": 1.6916825177243644, + "grad_norm": 0.8445965051651001, + "learning_rate": 1.627229705132872e-05, + "loss": 1.9409, + "step": 2446 + }, + { + "epoch": 1.6923742002420887, + "grad_norm": 1.0974277257919312, + "learning_rate": 1.6235893702220606e-05, + "loss": 1.6184, + "step": 2447 + }, + { + "epoch": 1.6930658827598133, + "grad_norm": 0.7980025410652161, + "learning_rate": 1.6199490353112487e-05, + "loss": 0.8208, + "step": 2448 + }, + { + "epoch": 1.6937575652775376, + "grad_norm": 1.0979562997817993, + "learning_rate": 1.6163087004004367e-05, + "loss": 1.745, + "step": 2449 + }, + { + "epoch": 1.694449247795262, + "grad_norm": 1.1840740442276, + "learning_rate": 1.6126683654896252e-05, + "loss": 1.0616, + "step": 2450 + }, + { + "epoch": 1.6951409303129863, + "grad_norm": 1.1019084453582764, + "learning_rate": 1.6090280305788133e-05, + "loss": 1.8189, + "step": 2451 + }, + { + "epoch": 1.6958326128307109, + "grad_norm": 1.218327522277832, + "learning_rate": 1.6053876956680014e-05, + "loss": 1.3346, + "step": 2452 + }, + { + "epoch": 1.6965242953484352, + "grad_norm": 1.1250942945480347, + "learning_rate": 1.6017473607571898e-05, + "loss": 1.7778, + "step": 2453 + }, + { + "epoch": 1.6972159778661595, + "grad_norm": 1.0627609491348267, + "learning_rate": 1.598107025846378e-05, + "loss": 1.1783, + "step": 2454 + }, + { + "epoch": 1.6979076603838839, + "grad_norm": 1.224642276763916, + "learning_rate": 1.5944666909355663e-05, + "loss": 1.7735, + "step": 2455 + }, + { + "epoch": 1.6985993429016082, + "grad_norm": 1.1235707998275757, + "learning_rate": 1.5908263560247544e-05, + "loss": 1.1083, + "step": 2456 + }, + { + "epoch": 1.6992910254193325, + "grad_norm": 0.9425327777862549, + "learning_rate": 1.5871860211139425e-05, + "loss": 1.4332, + "step": 2457 + }, + { + "epoch": 1.6999827079370569, + "grad_norm": 0.9698038101196289, + "learning_rate": 1.583545686203131e-05, + "loss": 0.978, + "step": 2458 + }, + { + "epoch": 1.7006743904547812, + "grad_norm": 1.0467294454574585, + "learning_rate": 1.579905351292319e-05, + "loss": 1.3348, + "step": 2459 + }, + { + "epoch": 1.7013660729725055, + "grad_norm": 0.9501770734786987, + "learning_rate": 1.576265016381507e-05, + "loss": 1.7046, + "step": 2460 + }, + { + "epoch": 1.7020577554902299, + "grad_norm": 1.0488495826721191, + "learning_rate": 1.5726246814706956e-05, + "loss": 1.1148, + "step": 2461 + }, + { + "epoch": 1.7027494380079542, + "grad_norm": 1.3002995252609253, + "learning_rate": 1.5689843465598837e-05, + "loss": 1.6453, + "step": 2462 + }, + { + "epoch": 1.7034411205256788, + "grad_norm": 0.9870662093162537, + "learning_rate": 1.5653440116490718e-05, + "loss": 1.6659, + "step": 2463 + }, + { + "epoch": 1.704132803043403, + "grad_norm": 0.6940680146217346, + "learning_rate": 1.5617036767382602e-05, + "loss": 1.3961, + "step": 2464 + }, + { + "epoch": 1.7048244855611274, + "grad_norm": 0.8727228045463562, + "learning_rate": 1.5580633418274483e-05, + "loss": 1.7241, + "step": 2465 + }, + { + "epoch": 1.7055161680788518, + "grad_norm": 0.8964528441429138, + "learning_rate": 1.5544230069166364e-05, + "loss": 1.6518, + "step": 2466 + }, + { + "epoch": 1.7062078505965763, + "grad_norm": 0.9496326446533203, + "learning_rate": 1.5507826720058248e-05, + "loss": 2.0925, + "step": 2467 + }, + { + "epoch": 1.7068995331143006, + "grad_norm": 1.083885669708252, + "learning_rate": 1.547142337095013e-05, + "loss": 1.8879, + "step": 2468 + }, + { + "epoch": 1.707591215632025, + "grad_norm": 0.993247389793396, + "learning_rate": 1.543502002184201e-05, + "loss": 1.6604, + "step": 2469 + }, + { + "epoch": 1.7082828981497493, + "grad_norm": 1.279847502708435, + "learning_rate": 1.539861667273389e-05, + "loss": 1.7554, + "step": 2470 + }, + { + "epoch": 1.7089745806674737, + "grad_norm": 0.9252359867095947, + "learning_rate": 1.5362213323625772e-05, + "loss": 0.847, + "step": 2471 + }, + { + "epoch": 1.709666263185198, + "grad_norm": 1.6128175258636475, + "learning_rate": 1.5325809974517656e-05, + "loss": 1.5952, + "step": 2472 + }, + { + "epoch": 1.7103579457029223, + "grad_norm": 1.2497482299804688, + "learning_rate": 1.5289406625409537e-05, + "loss": 1.8887, + "step": 2473 + }, + { + "epoch": 1.7110496282206467, + "grad_norm": 0.9129915237426758, + "learning_rate": 1.525300327630142e-05, + "loss": 1.6574, + "step": 2474 + }, + { + "epoch": 1.711741310738371, + "grad_norm": 1.0423600673675537, + "learning_rate": 1.5216599927193304e-05, + "loss": 1.4154, + "step": 2475 + }, + { + "epoch": 1.7124329932560953, + "grad_norm": 0.7967844605445862, + "learning_rate": 1.5180196578085185e-05, + "loss": 1.5242, + "step": 2476 + }, + { + "epoch": 1.7131246757738197, + "grad_norm": 0.9126850962638855, + "learning_rate": 1.5143793228977066e-05, + "loss": 1.4249, + "step": 2477 + }, + { + "epoch": 1.7138163582915442, + "grad_norm": 0.948121190071106, + "learning_rate": 1.5107389879868949e-05, + "loss": 1.7527, + "step": 2478 + }, + { + "epoch": 1.7145080408092686, + "grad_norm": 0.8831592798233032, + "learning_rate": 1.507098653076083e-05, + "loss": 1.6812, + "step": 2479 + }, + { + "epoch": 1.7151997233269929, + "grad_norm": 0.8274893164634705, + "learning_rate": 1.5034583181652712e-05, + "loss": 1.9424, + "step": 2480 + }, + { + "epoch": 1.7158914058447172, + "grad_norm": 1.2506767511367798, + "learning_rate": 1.4998179832544595e-05, + "loss": 1.5627, + "step": 2481 + }, + { + "epoch": 1.7165830883624418, + "grad_norm": 0.853093147277832, + "learning_rate": 1.4961776483436476e-05, + "loss": 0.7283, + "step": 2482 + }, + { + "epoch": 1.7172747708801661, + "grad_norm": 1.7281954288482666, + "learning_rate": 1.4925373134328357e-05, + "loss": 1.7007, + "step": 2483 + }, + { + "epoch": 1.7179664533978904, + "grad_norm": 0.9560598134994507, + "learning_rate": 1.4888969785220241e-05, + "loss": 1.3463, + "step": 2484 + }, + { + "epoch": 1.7186581359156148, + "grad_norm": 0.931280255317688, + "learning_rate": 1.4852566436112122e-05, + "loss": 1.8012, + "step": 2485 + }, + { + "epoch": 1.7193498184333391, + "grad_norm": 1.0191988945007324, + "learning_rate": 1.4816163087004007e-05, + "loss": 1.0877, + "step": 2486 + }, + { + "epoch": 1.7200415009510635, + "grad_norm": 0.8088228106498718, + "learning_rate": 1.4779759737895888e-05, + "loss": 1.4176, + "step": 2487 + }, + { + "epoch": 1.7207331834687878, + "grad_norm": 0.9902637004852295, + "learning_rate": 1.4743356388787769e-05, + "loss": 1.3056, + "step": 2488 + }, + { + "epoch": 1.7214248659865121, + "grad_norm": 1.0314936637878418, + "learning_rate": 1.4706953039679653e-05, + "loss": 1.6047, + "step": 2489 + }, + { + "epoch": 1.7221165485042365, + "grad_norm": 0.9210548400878906, + "learning_rate": 1.4670549690571534e-05, + "loss": 1.3797, + "step": 2490 + }, + { + "epoch": 1.7228082310219608, + "grad_norm": 0.6789326667785645, + "learning_rate": 1.4634146341463415e-05, + "loss": 1.4157, + "step": 2491 + }, + { + "epoch": 1.7234999135396851, + "grad_norm": 1.0114710330963135, + "learning_rate": 1.4597742992355299e-05, + "loss": 1.3783, + "step": 2492 + }, + { + "epoch": 1.7241915960574097, + "grad_norm": 1.0227901935577393, + "learning_rate": 1.456133964324718e-05, + "loss": 1.9873, + "step": 2493 + }, + { + "epoch": 1.724883278575134, + "grad_norm": 0.9514389634132385, + "learning_rate": 1.4524936294139061e-05, + "loss": 1.5695, + "step": 2494 + }, + { + "epoch": 1.7255749610928584, + "grad_norm": 0.8479339480400085, + "learning_rate": 1.4488532945030944e-05, + "loss": 1.6232, + "step": 2495 + }, + { + "epoch": 1.726266643610583, + "grad_norm": 0.9300490021705627, + "learning_rate": 1.4452129595922825e-05, + "loss": 1.7569, + "step": 2496 + }, + { + "epoch": 1.7269583261283072, + "grad_norm": 0.9302891492843628, + "learning_rate": 1.4415726246814707e-05, + "loss": 1.7697, + "step": 2497 + }, + { + "epoch": 1.7276500086460316, + "grad_norm": 1.7044777870178223, + "learning_rate": 1.437932289770659e-05, + "loss": 1.2168, + "step": 2498 + }, + { + "epoch": 1.728341691163756, + "grad_norm": 1.2160143852233887, + "learning_rate": 1.434291954859847e-05, + "loss": 1.7096, + "step": 2499 + }, + { + "epoch": 1.7290333736814802, + "grad_norm": 0.5079087018966675, + "learning_rate": 1.4306516199490355e-05, + "loss": 0.5207, + "step": 2500 + }, + { + "epoch": 1.7290333736814802, + "eval_loss": 1.4318758249282837, + "eval_runtime": 568.9702, + "eval_samples_per_second": 2.258, + "eval_steps_per_second": 1.13, + "step": 2500 + } + ], + "logging_steps": 1, + "max_steps": 2892, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 7.910455612243968e+16, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}