| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.979633401221996, | |
| "eval_steps": 500, | |
| "global_step": 2450, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004073319755600814, | |
| "grad_norm": 12.219289779663086, | |
| "learning_rate": 8.16326530612245e-07, | |
| "loss": 2.0786, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.020366598778004074, | |
| "grad_norm": 12.530974388122559, | |
| "learning_rate": 4.081632653061224e-06, | |
| "loss": 2.0708, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04073319755600815, | |
| "grad_norm": 10.991973876953125, | |
| "learning_rate": 8.163265306122448e-06, | |
| "loss": 1.9589, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06109979633401222, | |
| "grad_norm": 11.441458702087402, | |
| "learning_rate": 1.2244897959183674e-05, | |
| "loss": 1.6878, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0814663951120163, | |
| "grad_norm": 4.176149845123291, | |
| "learning_rate": 1.6326530612244897e-05, | |
| "loss": 1.3224, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.10183299389002037, | |
| "grad_norm": 2.458606004714966, | |
| "learning_rate": 2.0408163265306123e-05, | |
| "loss": 1.1912, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.12219959266802444, | |
| "grad_norm": 1.30384361743927, | |
| "learning_rate": 2.448979591836735e-05, | |
| "loss": 1.1123, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1425661914460285, | |
| "grad_norm": 1.0811594724655151, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 1.0462, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1629327902240326, | |
| "grad_norm": 0.7156168222427368, | |
| "learning_rate": 3.265306122448979e-05, | |
| "loss": 0.992, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.18329938900203666, | |
| "grad_norm": 0.6059057116508484, | |
| "learning_rate": 3.673469387755102e-05, | |
| "loss": 0.9393, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.20366598778004075, | |
| "grad_norm": 0.5178468227386475, | |
| "learning_rate": 4.0816326530612245e-05, | |
| "loss": 0.8967, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2240325865580448, | |
| "grad_norm": 0.45643067359924316, | |
| "learning_rate": 4.4897959183673474e-05, | |
| "loss": 0.8665, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.24439918533604887, | |
| "grad_norm": 0.4627813994884491, | |
| "learning_rate": 4.89795918367347e-05, | |
| "loss": 0.8554, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.26476578411405294, | |
| "grad_norm": 0.7451216578483582, | |
| "learning_rate": 5.3061224489795926e-05, | |
| "loss": 0.8335, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.285132382892057, | |
| "grad_norm": 0.4662891924381256, | |
| "learning_rate": 5.714285714285714e-05, | |
| "loss": 0.827, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3054989816700611, | |
| "grad_norm": 0.42819878458976746, | |
| "learning_rate": 6.122448979591838e-05, | |
| "loss": 0.816, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3258655804480652, | |
| "grad_norm": 0.4619337022304535, | |
| "learning_rate": 6.530612244897959e-05, | |
| "loss": 0.7968, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.34623217922606925, | |
| "grad_norm": 0.4922891855239868, | |
| "learning_rate": 6.938775510204082e-05, | |
| "loss": 0.7954, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3665987780040733, | |
| "grad_norm": 0.5278509259223938, | |
| "learning_rate": 7.346938775510205e-05, | |
| "loss": 0.7953, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3869653767820774, | |
| "grad_norm": 0.912343442440033, | |
| "learning_rate": 7.755102040816327e-05, | |
| "loss": 0.7773, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.4073319755600815, | |
| "grad_norm": 0.5746473073959351, | |
| "learning_rate": 8.163265306122449e-05, | |
| "loss": 0.775, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.42769857433808556, | |
| "grad_norm": 0.4901811480522156, | |
| "learning_rate": 8.571428571428571e-05, | |
| "loss": 0.7708, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4480651731160896, | |
| "grad_norm": 0.596023678779602, | |
| "learning_rate": 8.979591836734695e-05, | |
| "loss": 0.761, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4684317718940937, | |
| "grad_norm": 0.9318987131118774, | |
| "learning_rate": 9.387755102040817e-05, | |
| "loss": 0.7572, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.48879837067209775, | |
| "grad_norm": 0.648429811000824, | |
| "learning_rate": 9.79591836734694e-05, | |
| "loss": 0.7652, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5091649694501018, | |
| "grad_norm": 0.4797033369541168, | |
| "learning_rate": 0.00010204081632653062, | |
| "loss": 0.7563, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5295315682281059, | |
| "grad_norm": 0.9584823250770569, | |
| "learning_rate": 0.00010612244897959185, | |
| "loss": 0.7532, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5498981670061099, | |
| "grad_norm": 0.6396449208259583, | |
| "learning_rate": 0.00011020408163265306, | |
| "loss": 0.7438, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.570264765784114, | |
| "grad_norm": 0.8127761483192444, | |
| "learning_rate": 0.00011428571428571428, | |
| "loss": 0.7367, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5906313645621182, | |
| "grad_norm": 0.6745628118515015, | |
| "learning_rate": 0.00011836734693877552, | |
| "loss": 0.7284, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.6109979633401222, | |
| "grad_norm": 0.5591239929199219, | |
| "learning_rate": 0.00012244897959183676, | |
| "loss": 0.7337, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6313645621181263, | |
| "grad_norm": 0.7267659902572632, | |
| "learning_rate": 0.00012653061224489798, | |
| "loss": 0.7264, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.6517311608961304, | |
| "grad_norm": 0.5548218488693237, | |
| "learning_rate": 0.00013061224489795917, | |
| "loss": 0.7301, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6720977596741344, | |
| "grad_norm": 0.9011363387107849, | |
| "learning_rate": 0.0001346938775510204, | |
| "loss": 0.72, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6924643584521385, | |
| "grad_norm": 0.5213197469711304, | |
| "learning_rate": 0.00013877551020408165, | |
| "loss": 0.7243, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7128309572301426, | |
| "grad_norm": 0.476764440536499, | |
| "learning_rate": 0.00014285714285714287, | |
| "loss": 0.7198, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7331975560081466, | |
| "grad_norm": 0.7126782536506653, | |
| "learning_rate": 0.0001469387755102041, | |
| "loss": 0.7177, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7535641547861507, | |
| "grad_norm": 0.4389437735080719, | |
| "learning_rate": 0.0001510204081632653, | |
| "loss": 0.7143, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7739307535641547, | |
| "grad_norm": 0.5171541571617126, | |
| "learning_rate": 0.00015510204081632654, | |
| "loss": 0.714, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7942973523421588, | |
| "grad_norm": 0.5747414827346802, | |
| "learning_rate": 0.00015918367346938776, | |
| "loss": 0.7192, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.814663951120163, | |
| "grad_norm": 0.703008770942688, | |
| "learning_rate": 0.00016326530612244898, | |
| "loss": 0.7094, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.835030549898167, | |
| "grad_norm": 0.41933196783065796, | |
| "learning_rate": 0.00016734693877551023, | |
| "loss": 0.7027, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8553971486761711, | |
| "grad_norm": 0.8496442437171936, | |
| "learning_rate": 0.00017142857142857143, | |
| "loss": 0.7086, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.8757637474541752, | |
| "grad_norm": 0.44953274726867676, | |
| "learning_rate": 0.00017551020408163265, | |
| "loss": 0.7104, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8961303462321792, | |
| "grad_norm": 0.49215880036354065, | |
| "learning_rate": 0.0001795918367346939, | |
| "loss": 0.6998, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.9164969450101833, | |
| "grad_norm": 0.4515603482723236, | |
| "learning_rate": 0.00018367346938775512, | |
| "loss": 0.6994, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9368635437881874, | |
| "grad_norm": 0.48363256454467773, | |
| "learning_rate": 0.00018775510204081634, | |
| "loss": 0.697, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9572301425661914, | |
| "grad_norm": 0.504764974117279, | |
| "learning_rate": 0.00019183673469387756, | |
| "loss": 0.6937, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9775967413441955, | |
| "grad_norm": 0.6177113056182861, | |
| "learning_rate": 0.0001959183673469388, | |
| "loss": 0.702, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9979633401221996, | |
| "grad_norm": 0.6133261919021606, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6982, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9979633401221996, | |
| "eval_loss": 1.8247634172439575, | |
| "eval_runtime": 0.3801, | |
| "eval_samples_per_second": 31.574, | |
| "eval_steps_per_second": 2.631, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.0183299389002036, | |
| "grad_norm": 0.6387724876403809, | |
| "learning_rate": 0.00019999746258949147, | |
| "loss": 0.6786, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0386965376782078, | |
| "grad_norm": 0.4179505705833435, | |
| "learning_rate": 0.00019998985048673486, | |
| "loss": 0.6806, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0590631364562118, | |
| "grad_norm": 0.7707039713859558, | |
| "learning_rate": 0.0001999771640780308, | |
| "loss": 0.6828, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.079429735234216, | |
| "grad_norm": 0.735935628414154, | |
| "learning_rate": 0.0001999594040071918, | |
| "loss": 0.6832, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0997963340122199, | |
| "grad_norm": 0.6434056758880615, | |
| "learning_rate": 0.00019993657117550973, | |
| "loss": 0.681, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.120162932790224, | |
| "grad_norm": 0.7113597393035889, | |
| "learning_rate": 0.00019990866674170983, | |
| "loss": 0.6735, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.140529531568228, | |
| "grad_norm": 0.7098330855369568, | |
| "learning_rate": 0.00019987569212189224, | |
| "loss": 0.6786, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1608961303462322, | |
| "grad_norm": 0.38251993060112, | |
| "learning_rate": 0.0001998376489894599, | |
| "loss": 0.6794, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1812627291242364, | |
| "grad_norm": 0.6434310674667358, | |
| "learning_rate": 0.00019979453927503364, | |
| "loss": 0.6846, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.2016293279022403, | |
| "grad_norm": 0.5492526292800903, | |
| "learning_rate": 0.00019974636516635434, | |
| "loss": 0.6697, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.2219959266802445, | |
| "grad_norm": 0.48028430342674255, | |
| "learning_rate": 0.00019969312910817183, | |
| "loss": 0.6778, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2423625254582484, | |
| "grad_norm": 0.35454341769218445, | |
| "learning_rate": 0.00019963483380212068, | |
| "loss": 0.6671, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2627291242362526, | |
| "grad_norm": 0.3555687665939331, | |
| "learning_rate": 0.00019957148220658345, | |
| "loss": 0.6646, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2830957230142566, | |
| "grad_norm": 0.5027297735214233, | |
| "learning_rate": 0.00019950307753654017, | |
| "loss": 0.6723, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.3034623217922607, | |
| "grad_norm": 0.34120357036590576, | |
| "learning_rate": 0.00019942962326340537, | |
| "loss": 0.6669, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.3238289205702647, | |
| "grad_norm": 0.3805827498435974, | |
| "learning_rate": 0.000199351123114852, | |
| "loss": 0.6682, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.3441955193482689, | |
| "grad_norm": 0.4035264551639557, | |
| "learning_rate": 0.00019926758107462206, | |
| "loss": 0.6665, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.364562118126273, | |
| "grad_norm": 0.43114498257637024, | |
| "learning_rate": 0.0001991790013823246, | |
| "loss": 0.6687, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.384928716904277, | |
| "grad_norm": 0.43574780225753784, | |
| "learning_rate": 0.00019908538853322048, | |
| "loss": 0.6648, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.405295315682281, | |
| "grad_norm": 0.36844050884246826, | |
| "learning_rate": 0.00019898674727799419, | |
| "loss": 0.6635, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.4256619144602851, | |
| "grad_norm": 0.360999196767807, | |
| "learning_rate": 0.00019888308262251285, | |
| "loss": 0.6715, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.4460285132382893, | |
| "grad_norm": 0.47190627455711365, | |
| "learning_rate": 0.0001987743998275723, | |
| "loss": 0.6595, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4663951120162932, | |
| "grad_norm": 0.4380181133747101, | |
| "learning_rate": 0.00019866070440862976, | |
| "loss": 0.6589, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4867617107942974, | |
| "grad_norm": 0.3873290419578552, | |
| "learning_rate": 0.00019854200213552424, | |
| "loss": 0.6649, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.5071283095723014, | |
| "grad_norm": 0.4168381094932556, | |
| "learning_rate": 0.00019841829903218376, | |
| "loss": 0.6673, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.5274949083503055, | |
| "grad_norm": 0.4359332025051117, | |
| "learning_rate": 0.00019828960137631928, | |
| "loss": 0.6666, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.5478615071283097, | |
| "grad_norm": 0.33967724442481995, | |
| "learning_rate": 0.00019815591569910654, | |
| "loss": 0.6663, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.5682281059063137, | |
| "grad_norm": 0.33921679854393005, | |
| "learning_rate": 0.00019801724878485438, | |
| "loss": 0.6557, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5885947046843176, | |
| "grad_norm": 0.3741433024406433, | |
| "learning_rate": 0.00019787360767066053, | |
| "loss": 0.6522, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.6089613034623218, | |
| "grad_norm": 0.3441908359527588, | |
| "learning_rate": 0.0001977249996460544, | |
| "loss": 0.653, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.629327902240326, | |
| "grad_norm": 0.36859628558158875, | |
| "learning_rate": 0.00019757143225262728, | |
| "loss": 0.6582, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.64969450101833, | |
| "grad_norm": 0.3605320453643799, | |
| "learning_rate": 0.00019741291328364955, | |
| "loss": 0.6546, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6700610997963339, | |
| "grad_norm": 0.6110367178916931, | |
| "learning_rate": 0.00019724945078367513, | |
| "loss": 0.6567, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.690427698574338, | |
| "grad_norm": 0.3475343585014343, | |
| "learning_rate": 0.00019708105304813332, | |
| "loss": 0.6658, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.7107942973523422, | |
| "grad_norm": 0.45948854088783264, | |
| "learning_rate": 0.0001969077286229078, | |
| "loss": 0.6584, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.7311608961303462, | |
| "grad_norm": 0.31532639265060425, | |
| "learning_rate": 0.00019672948630390294, | |
| "loss": 0.6579, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.7515274949083504, | |
| "grad_norm": 0.3328857421875, | |
| "learning_rate": 0.00019654633513659743, | |
| "loss": 0.6583, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7718940936863543, | |
| "grad_norm": 0.3420257866382599, | |
| "learning_rate": 0.00019635828441558515, | |
| "loss": 0.6493, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7922606924643585, | |
| "grad_norm": 0.31616052985191345, | |
| "learning_rate": 0.00019616534368410365, | |
| "loss": 0.6536, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.8126272912423627, | |
| "grad_norm": 0.30928483605384827, | |
| "learning_rate": 0.0001959675227335497, | |
| "loss": 0.6583, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.8329938900203666, | |
| "grad_norm": 0.3872964084148407, | |
| "learning_rate": 0.00019576483160298246, | |
| "loss": 0.6467, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.8533604887983706, | |
| "grad_norm": 0.3449136018753052, | |
| "learning_rate": 0.0001955572805786141, | |
| "loss": 0.6513, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8737270875763747, | |
| "grad_norm": 0.3003346621990204, | |
| "learning_rate": 0.00019534488019328752, | |
| "loss": 0.6542, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.894093686354379, | |
| "grad_norm": 0.3554345965385437, | |
| "learning_rate": 0.0001951276412259422, | |
| "loss": 0.6496, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.9144602851323829, | |
| "grad_norm": 0.3109475374221802, | |
| "learning_rate": 0.00019490557470106686, | |
| "loss": 0.6544, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.9348268839103868, | |
| "grad_norm": 0.35385647416114807, | |
| "learning_rate": 0.00019467869188814023, | |
| "loss": 0.6456, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.955193482688391, | |
| "grad_norm": 0.3411131203174591, | |
| "learning_rate": 0.00019444700430105893, | |
| "loss": 0.6478, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9755600814663952, | |
| "grad_norm": 0.3149469792842865, | |
| "learning_rate": 0.00019421052369755334, | |
| "loss": 0.6541, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9959266802443993, | |
| "grad_norm": 0.4090039134025574, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 0.6596, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.833816647529602, | |
| "eval_runtime": 0.3787, | |
| "eval_samples_per_second": 31.686, | |
| "eval_steps_per_second": 2.641, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 2.016293279022403, | |
| "grad_norm": 0.4153566062450409, | |
| "learning_rate": 0.0001937232316877668, | |
| "loss": 0.6285, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.0366598778004072, | |
| "grad_norm": 0.41273799538612366, | |
| "learning_rate": 0.00019347244501068312, | |
| "loss": 0.6309, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0570264765784114, | |
| "grad_norm": 0.40306323766708374, | |
| "learning_rate": 0.00019321691477431485, | |
| "loss": 0.6294, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.0773930753564156, | |
| "grad_norm": 0.37823185324668884, | |
| "learning_rate": 0.00019295665394636414, | |
| "loss": 0.6333, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.0977596741344193, | |
| "grad_norm": 0.31898233294487, | |
| "learning_rate": 0.0001926916757346022, | |
| "loss": 0.6217, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.1181262729124235, | |
| "grad_norm": 0.3543696105480194, | |
| "learning_rate": 0.00019242199358619893, | |
| "loss": 0.6268, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.1384928716904277, | |
| "grad_norm": 0.4415835738182068, | |
| "learning_rate": 0.00019214762118704076, | |
| "loss": 0.6252, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.158859470468432, | |
| "grad_norm": 0.40828588604927063, | |
| "learning_rate": 0.00019186857246103584, | |
| "loss": 0.624, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.179226069246436, | |
| "grad_norm": 0.37648454308509827, | |
| "learning_rate": 0.00019158486156940757, | |
| "loss": 0.6228, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.1995926680244398, | |
| "grad_norm": 0.3546181321144104, | |
| "learning_rate": 0.0001912965029099759, | |
| "loss": 0.62, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.219959266802444, | |
| "grad_norm": 0.3437698185443878, | |
| "learning_rate": 0.00019100351111642666, | |
| "loss": 0.6228, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.240325865580448, | |
| "grad_norm": 0.3919028639793396, | |
| "learning_rate": 0.000190705901057569, | |
| "loss": 0.623, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.2606924643584523, | |
| "grad_norm": 0.36673861742019653, | |
| "learning_rate": 0.00019040368783658073, | |
| "loss": 0.6222, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.281059063136456, | |
| "grad_norm": 0.3100906312465668, | |
| "learning_rate": 0.0001900968867902419, | |
| "loss": 0.6276, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.30142566191446, | |
| "grad_norm": 0.33423855900764465, | |
| "learning_rate": 0.00018978551348815652, | |
| "loss": 0.6304, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.3217922606924644, | |
| "grad_norm": 0.3217325210571289, | |
| "learning_rate": 0.00018946958373196231, | |
| "loss": 0.6255, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.3421588594704685, | |
| "grad_norm": 0.32845550775527954, | |
| "learning_rate": 0.00018914911355452895, | |
| "loss": 0.6276, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.3625254582484727, | |
| "grad_norm": 0.3502614498138428, | |
| "learning_rate": 0.0001888241192191444, | |
| "loss": 0.6345, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.3828920570264764, | |
| "grad_norm": 0.3093807101249695, | |
| "learning_rate": 0.00018849461721868945, | |
| "loss": 0.6231, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.4032586558044806, | |
| "grad_norm": 0.34325292706489563, | |
| "learning_rate": 0.0001881606242748009, | |
| "loss": 0.6356, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.423625254582485, | |
| "grad_norm": 0.3460986018180847, | |
| "learning_rate": 0.00018782215733702286, | |
| "loss": 0.6228, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.443991853360489, | |
| "grad_norm": 0.31046947836875916, | |
| "learning_rate": 0.00018747923358194662, | |
| "loss": 0.6264, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.4643584521384927, | |
| "grad_norm": 0.318236380815506, | |
| "learning_rate": 0.00018713187041233896, | |
| "loss": 0.6251, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.484725050916497, | |
| "grad_norm": 0.33597373962402344, | |
| "learning_rate": 0.000186780085456259, | |
| "loss": 0.6333, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.505091649694501, | |
| "grad_norm": 0.32059183716773987, | |
| "learning_rate": 0.0001864238965661637, | |
| "loss": 0.6322, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.525458248472505, | |
| "grad_norm": 0.31177854537963867, | |
| "learning_rate": 0.00018606332181800165, | |
| "loss": 0.6254, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.5458248472505094, | |
| "grad_norm": 0.35773152112960815, | |
| "learning_rate": 0.00018569837951029595, | |
| "loss": 0.6282, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.566191446028513, | |
| "grad_norm": 0.31747111678123474, | |
| "learning_rate": 0.00018532908816321558, | |
| "loss": 0.6229, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.5865580448065173, | |
| "grad_norm": 0.3057750463485718, | |
| "learning_rate": 0.0001849554665176354, | |
| "loss": 0.6257, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.6069246435845215, | |
| "grad_norm": 0.30599597096443176, | |
| "learning_rate": 0.0001845775335341852, | |
| "loss": 0.6199, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.627291242362525, | |
| "grad_norm": 0.3067176043987274, | |
| "learning_rate": 0.0001841953083922875, | |
| "loss": 0.6244, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.6476578411405294, | |
| "grad_norm": 0.32527825236320496, | |
| "learning_rate": 0.00018380881048918405, | |
| "loss": 0.6258, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.6680244399185336, | |
| "grad_norm": 0.3099977970123291, | |
| "learning_rate": 0.00018341805943895178, | |
| "loss": 0.6298, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.6883910386965377, | |
| "grad_norm": 0.3218511641025543, | |
| "learning_rate": 0.00018302307507150702, | |
| "loss": 0.624, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.708757637474542, | |
| "grad_norm": 0.3705076277256012, | |
| "learning_rate": 0.0001826238774315995, | |
| "loss": 0.6219, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.729124236252546, | |
| "grad_norm": 0.4094727337360382, | |
| "learning_rate": 0.00018222048677779494, | |
| "loss": 0.6244, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.74949083503055, | |
| "grad_norm": 0.327188640832901, | |
| "learning_rate": 0.00018181292358144703, | |
| "loss": 0.6195, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.769857433808554, | |
| "grad_norm": 0.3251926302909851, | |
| "learning_rate": 0.0001814012085256585, | |
| "loss": 0.6235, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.790224032586558, | |
| "grad_norm": 0.320218950510025, | |
| "learning_rate": 0.00018098536250423154, | |
| "loss": 0.6236, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.810590631364562, | |
| "grad_norm": 0.32766249775886536, | |
| "learning_rate": 0.00018056540662060745, | |
| "loss": 0.6188, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.830957230142566, | |
| "grad_norm": 0.35727354884147644, | |
| "learning_rate": 0.00018014136218679567, | |
| "loss": 0.619, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.8513238289205702, | |
| "grad_norm": 0.32284310460090637, | |
| "learning_rate": 0.00017971325072229226, | |
| "loss": 0.6224, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.8716904276985744, | |
| "grad_norm": 0.34942692518234253, | |
| "learning_rate": 0.00017928109395298777, | |
| "loss": 0.617, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.8920570264765786, | |
| "grad_norm": 0.3333948850631714, | |
| "learning_rate": 0.00017884491381006478, | |
| "loss": 0.6269, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.9124236252545828, | |
| "grad_norm": 0.3345637321472168, | |
| "learning_rate": 0.00017840473242888486, | |
| "loss": 0.6226, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.9327902240325865, | |
| "grad_norm": 0.3559291958808899, | |
| "learning_rate": 0.0001779605721478652, | |
| "loss": 0.6173, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.9531568228105907, | |
| "grad_norm": 0.3167003393173218, | |
| "learning_rate": 0.0001775124555073452, | |
| "loss": 0.6224, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.973523421588595, | |
| "grad_norm": 0.3369687497615814, | |
| "learning_rate": 0.00017706040524844221, | |
| "loss": 0.6155, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.9938900203665986, | |
| "grad_norm": 0.30602747201919556, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 0.6197, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.9979633401221997, | |
| "eval_loss": 1.8432329893112183, | |
| "eval_runtime": 0.4594, | |
| "eval_samples_per_second": 26.122, | |
| "eval_steps_per_second": 2.177, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 3.0142566191446027, | |
| "grad_norm": 0.3442608416080475, | |
| "learning_rate": 0.00017614459583691346, | |
| "loss": 0.6108, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.034623217922607, | |
| "grad_norm": 0.30495738983154297, | |
| "learning_rate": 0.0001756808831599762, | |
| "loss": 0.6031, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.054989816700611, | |
| "grad_norm": 0.34981706738471985, | |
| "learning_rate": 0.0001752133298136744, | |
| "loss": 0.5959, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0753564154786153, | |
| "grad_norm": 0.32501649856567383, | |
| "learning_rate": 0.00017474195952550355, | |
| "loss": 0.5964, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.095723014256619, | |
| "grad_norm": 0.3137637972831726, | |
| "learning_rate": 0.00017426679621666212, | |
| "loss": 0.5885, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.116089613034623, | |
| "grad_norm": 0.3315506875514984, | |
| "learning_rate": 0.00017378786400083757, | |
| "loss": 0.5914, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.1364562118126273, | |
| "grad_norm": 0.33116060495376587, | |
| "learning_rate": 0.00017330518718298264, | |
| "loss": 0.5997, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.1568228105906315, | |
| "grad_norm": 0.33854183554649353, | |
| "learning_rate": 0.0001728187902580819, | |
| "loss": 0.5971, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.1771894093686353, | |
| "grad_norm": 0.31813180446624756, | |
| "learning_rate": 0.00017232869790990879, | |
| "loss": 0.5994, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.1975560081466394, | |
| "grad_norm": 0.30412253737449646, | |
| "learning_rate": 0.00017183493500977278, | |
| "loss": 0.5979, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.2179226069246436, | |
| "grad_norm": 0.3499339520931244, | |
| "learning_rate": 0.0001713375266152572, | |
| "loss": 0.5981, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.2382892057026478, | |
| "grad_norm": 0.3247123956680298, | |
| "learning_rate": 0.00017083649796894795, | |
| "loss": 0.5955, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.258655804480652, | |
| "grad_norm": 0.3245786726474762, | |
| "learning_rate": 0.00017033187449715196, | |
| "loss": 0.6019, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.2790224032586557, | |
| "grad_norm": 0.3269343674182892, | |
| "learning_rate": 0.00016982368180860728, | |
| "loss": 0.5979, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.29938900203666, | |
| "grad_norm": 0.4949992299079895, | |
| "learning_rate": 0.00016931194569318327, | |
| "loss": 0.6037, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.319755600814664, | |
| "grad_norm": 0.3468710780143738, | |
| "learning_rate": 0.00016879669212057187, | |
| "loss": 0.5943, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.340122199592668, | |
| "grad_norm": 0.323321670293808, | |
| "learning_rate": 0.00016827794723896968, | |
| "loss": 0.6003, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.360488798370672, | |
| "grad_norm": 0.3203057646751404, | |
| "learning_rate": 0.00016775573737375096, | |
| "loss": 0.6023, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.380855397148676, | |
| "grad_norm": 0.32133370637893677, | |
| "learning_rate": 0.0001672300890261317, | |
| "loss": 0.6043, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.4012219959266803, | |
| "grad_norm": 0.32316601276397705, | |
| "learning_rate": 0.0001667010288718247, | |
| "loss": 0.6051, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.4215885947046845, | |
| "grad_norm": 0.34438490867614746, | |
| "learning_rate": 0.00016616858375968595, | |
| "loss": 0.6039, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.4419551934826886, | |
| "grad_norm": 0.3012474477291107, | |
| "learning_rate": 0.0001656327807103518, | |
| "loss": 0.599, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.4623217922606924, | |
| "grad_norm": 0.3228248059749603, | |
| "learning_rate": 0.0001650936469148681, | |
| "loss": 0.604, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.4826883910386965, | |
| "grad_norm": 0.3279968798160553, | |
| "learning_rate": 0.00016455120973331, | |
| "loss": 0.5953, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.5030549898167007, | |
| "grad_norm": 0.3526194393634796, | |
| "learning_rate": 0.0001640054966933935, | |
| "loss": 0.6029, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.5234215885947044, | |
| "grad_norm": 0.37160396575927734, | |
| "learning_rate": 0.00016345653548907873, | |
| "loss": 0.6005, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.5437881873727086, | |
| "grad_norm": 0.33020690083503723, | |
| "learning_rate": 0.00016290435397916424, | |
| "loss": 0.6049, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.564154786150713, | |
| "grad_norm": 0.33525606989860535, | |
| "learning_rate": 0.00016234898018587337, | |
| "loss": 0.5942, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.584521384928717, | |
| "grad_norm": 0.33215492963790894, | |
| "learning_rate": 0.00016179044229343206, | |
| "loss": 0.6, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.604887983706721, | |
| "grad_norm": 0.3076721429824829, | |
| "learning_rate": 0.00016122876864663868, | |
| "loss": 0.5984, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.6252545824847253, | |
| "grad_norm": 0.379398375749588, | |
| "learning_rate": 0.00016066398774942554, | |
| "loss": 0.6005, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.645621181262729, | |
| "grad_norm": 0.3342379629611969, | |
| "learning_rate": 0.00016009612826341227, | |
| "loss": 0.5987, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.6659877800407332, | |
| "grad_norm": 0.3370605707168579, | |
| "learning_rate": 0.00015952521900645144, | |
| "loss": 0.6013, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.6863543788187374, | |
| "grad_norm": 0.3159192204475403, | |
| "learning_rate": 0.000158951288951166, | |
| "loss": 0.5992, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.706720977596741, | |
| "grad_norm": 0.3914301097393036, | |
| "learning_rate": 0.000158374367223479, | |
| "loss": 0.5994, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.7270875763747453, | |
| "grad_norm": 0.33679434657096863, | |
| "learning_rate": 0.00015779448310113554, | |
| "loss": 0.5949, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.7474541751527495, | |
| "grad_norm": 0.3156144618988037, | |
| "learning_rate": 0.00015721166601221698, | |
| "loss": 0.595, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.7678207739307537, | |
| "grad_norm": 0.3130475878715515, | |
| "learning_rate": 0.0001566259455336474, | |
| "loss": 0.596, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.788187372708758, | |
| "grad_norm": 0.303627610206604, | |
| "learning_rate": 0.00015603735138969272, | |
| "loss": 0.6036, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.8085539714867616, | |
| "grad_norm": 0.32147976756095886, | |
| "learning_rate": 0.0001554459134504523, | |
| "loss": 0.5971, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.8289205702647657, | |
| "grad_norm": 0.33086270093917847, | |
| "learning_rate": 0.000154851661730343, | |
| "loss": 0.5979, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.84928716904277, | |
| "grad_norm": 0.3000107407569885, | |
| "learning_rate": 0.00015425462638657595, | |
| "loss": 0.5914, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.869653767820774, | |
| "grad_norm": 0.292559415102005, | |
| "learning_rate": 0.0001536548377176263, | |
| "loss": 0.5941, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.890020366598778, | |
| "grad_norm": 0.30751997232437134, | |
| "learning_rate": 0.00015305232616169548, | |
| "loss": 0.6094, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.910386965376782, | |
| "grad_norm": 0.3731157183647156, | |
| "learning_rate": 0.00015244712229516656, | |
| "loss": 0.5955, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.930753564154786, | |
| "grad_norm": 0.3387170732021332, | |
| "learning_rate": 0.00015183925683105254, | |
| "loss": 0.5972, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.9511201629327903, | |
| "grad_norm": 0.3645428717136383, | |
| "learning_rate": 0.0001512287606174377, | |
| "loss": 0.5998, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.9714867617107945, | |
| "grad_norm": 0.29603078961372375, | |
| "learning_rate": 0.0001506156646359123, | |
| "loss": 0.596, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.9918533604887982, | |
| "grad_norm": 0.30991217494010925, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.6011, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.870729923248291, | |
| "eval_runtime": 0.3769, | |
| "eval_samples_per_second": 31.837, | |
| "eval_steps_per_second": 2.653, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 4.012219959266803, | |
| "grad_norm": 0.33605799078941345, | |
| "learning_rate": 0.00014938179795357916, | |
| "loss": 0.5858, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 4.032586558044806, | |
| "grad_norm": 0.3564818203449249, | |
| "learning_rate": 0.00014876108986929717, | |
| "loss": 0.5735, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.05295315682281, | |
| "grad_norm": 0.31967151165008545, | |
| "learning_rate": 0.00014813790724697832, | |
| "loss": 0.579, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.0733197556008145, | |
| "grad_norm": 0.33131125569343567, | |
| "learning_rate": 0.0001475122817120253, | |
| "loss": 0.57, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.093686354378819, | |
| "grad_norm": 0.3302120864391327, | |
| "learning_rate": 0.00014688424501381424, | |
| "loss": 0.5658, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.114052953156823, | |
| "grad_norm": 0.356545627117157, | |
| "learning_rate": 0.00014625382902408356, | |
| "loss": 0.5722, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.134419551934827, | |
| "grad_norm": 0.30831289291381836, | |
| "learning_rate": 0.0001456210657353163, | |
| "loss": 0.5775, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.154786150712831, | |
| "grad_norm": 0.32479575276374817, | |
| "learning_rate": 0.00014498598725911691, | |
| "loss": 0.5762, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.175152749490835, | |
| "grad_norm": 0.3275298476219177, | |
| "learning_rate": 0.00014434862582458135, | |
| "loss": 0.5742, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.195519348268839, | |
| "grad_norm": 0.32563847303390503, | |
| "learning_rate": 0.00014370901377666167, | |
| "loss": 0.5696, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.215885947046843, | |
| "grad_norm": 0.3277311325073242, | |
| "learning_rate": 0.00014306718357452446, | |
| "loss": 0.5683, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.236252545824847, | |
| "grad_norm": 0.3310577869415283, | |
| "learning_rate": 0.00014242316778990372, | |
| "loss": 0.5724, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.256619144602851, | |
| "grad_norm": 0.3116869330406189, | |
| "learning_rate": 0.00014177699910544793, | |
| "loss": 0.5748, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 4.276985743380855, | |
| "grad_norm": 0.4322071969509125, | |
| "learning_rate": 0.00014112871031306119, | |
| "loss": 0.5848, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.2973523421588595, | |
| "grad_norm": 0.3503554165363312, | |
| "learning_rate": 0.00014047833431223938, | |
| "loss": 0.5732, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.317718940936864, | |
| "grad_norm": 0.31962850689888, | |
| "learning_rate": 0.00013982590410840056, | |
| "loss": 0.5774, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.338085539714868, | |
| "grad_norm": 0.32895031571388245, | |
| "learning_rate": 0.00013917145281120983, | |
| "loss": 0.5765, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.358452138492872, | |
| "grad_norm": 0.3837604522705078, | |
| "learning_rate": 0.00013851501363289906, | |
| "loss": 0.5817, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.378818737270876, | |
| "grad_norm": 0.35425829887390137, | |
| "learning_rate": 0.0001378566198865818, | |
| "loss": 0.58, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.3991853360488795, | |
| "grad_norm": 0.33126357197761536, | |
| "learning_rate": 0.00013719630498456212, | |
| "loss": 0.5827, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.419551934826884, | |
| "grad_norm": 0.3265593945980072, | |
| "learning_rate": 0.00013653410243663952, | |
| "loss": 0.5826, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.439918533604888, | |
| "grad_norm": 0.4005591869354248, | |
| "learning_rate": 0.00013587004584840804, | |
| "loss": 0.5795, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.460285132382892, | |
| "grad_norm": 0.36212101578712463, | |
| "learning_rate": 0.00013520416891955102, | |
| "loss": 0.5711, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.480651731160896, | |
| "grad_norm": 0.340218186378479, | |
| "learning_rate": 0.00013453650544213076, | |
| "loss": 0.5821, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.5010183299389, | |
| "grad_norm": 0.31296971440315247, | |
| "learning_rate": 0.00013386708929887377, | |
| "loss": 0.5798, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.521384928716905, | |
| "grad_norm": 0.31503763794898987, | |
| "learning_rate": 0.00013319595446145116, | |
| "loss": 0.5757, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.541751527494909, | |
| "grad_norm": 0.3179926574230194, | |
| "learning_rate": 0.00013252313498875472, | |
| "loss": 0.5822, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.562118126272912, | |
| "grad_norm": 0.3273943066596985, | |
| "learning_rate": 0.00013184866502516845, | |
| "loss": 0.5796, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.582484725050916, | |
| "grad_norm": 0.3253306448459625, | |
| "learning_rate": 0.00013117257879883583, | |
| "loss": 0.5789, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.60285132382892, | |
| "grad_norm": 0.34257039427757263, | |
| "learning_rate": 0.00013049491061992274, | |
| "loss": 0.5798, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.6232179226069245, | |
| "grad_norm": 0.32451459765434265, | |
| "learning_rate": 0.00012981569487887637, | |
| "loss": 0.5766, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.643584521384929, | |
| "grad_norm": 0.31603989005088806, | |
| "learning_rate": 0.0001291349660446799, | |
| "loss": 0.5765, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.663951120162933, | |
| "grad_norm": 0.3706742525100708, | |
| "learning_rate": 0.00012845275866310324, | |
| "loss": 0.5809, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.684317718940937, | |
| "grad_norm": 0.32328835129737854, | |
| "learning_rate": 0.00012776910735495003, | |
| "loss": 0.5784, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.704684317718941, | |
| "grad_norm": 0.3305988907814026, | |
| "learning_rate": 0.00012708404681430053, | |
| "loss": 0.5678, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.725050916496945, | |
| "grad_norm": 0.33260881900787354, | |
| "learning_rate": 0.00012639761180675098, | |
| "loss": 0.5895, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.745417515274949, | |
| "grad_norm": 0.34639254212379456, | |
| "learning_rate": 0.0001257098371676495, | |
| "loss": 0.5823, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.765784114052953, | |
| "grad_norm": 0.3303966224193573, | |
| "learning_rate": 0.0001250207578003279, | |
| "loss": 0.5716, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.786150712830957, | |
| "grad_norm": 0.3739718496799469, | |
| "learning_rate": 0.0001243304086743309, | |
| "loss": 0.5811, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.806517311608961, | |
| "grad_norm": 0.36287903785705566, | |
| "learning_rate": 0.0001236388248236409, | |
| "loss": 0.5791, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.826883910386965, | |
| "grad_norm": 0.3199651837348938, | |
| "learning_rate": 0.00012294604134490056, | |
| "loss": 0.5769, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.84725050916497, | |
| "grad_norm": 0.3307904899120331, | |
| "learning_rate": 0.00012225209339563145, | |
| "loss": 0.5781, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.867617107942974, | |
| "grad_norm": 0.3095554709434509, | |
| "learning_rate": 0.00012155701619244997, | |
| "loss": 0.5776, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.887983706720978, | |
| "grad_norm": 0.318012535572052, | |
| "learning_rate": 0.0001208608450092801, | |
| "loss": 0.5816, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.908350305498981, | |
| "grad_norm": 0.31566402316093445, | |
| "learning_rate": 0.00012016361517556334, | |
| "loss": 0.578, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.928716904276985, | |
| "grad_norm": 0.32222023606300354, | |
| "learning_rate": 0.00011946536207446586, | |
| "loss": 0.5776, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.94908350305499, | |
| "grad_norm": 0.31605854630470276, | |
| "learning_rate": 0.00011876612114108277, | |
| "loss": 0.5793, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.969450101832994, | |
| "grad_norm": 0.3147432804107666, | |
| "learning_rate": 0.0001180659278606399, | |
| "loss": 0.5753, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.989816700610998, | |
| "grad_norm": 0.3298138380050659, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.5805, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.997963340122199, | |
| "eval_loss": 1.9009366035461426, | |
| "eval_runtime": 0.3793, | |
| "eval_samples_per_second": 31.637, | |
| "eval_steps_per_second": 2.636, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 5.010183299389002, | |
| "grad_norm": 0.3077872395515442, | |
| "learning_rate": 0.00011666282643932458, | |
| "loss": 0.5659, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 5.030549898167006, | |
| "grad_norm": 0.3151206374168396, | |
| "learning_rate": 0.00011595998950333793, | |
| "loss": 0.5451, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.05091649694501, | |
| "grad_norm": 0.31535080075263977, | |
| "learning_rate": 0.00011525634262644964, | |
| "loss": 0.5478, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 5.071283095723015, | |
| "grad_norm": 0.3317250907421112, | |
| "learning_rate": 0.00011455192151747932, | |
| "loss": 0.5568, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 5.091649694501018, | |
| "grad_norm": 0.34287726879119873, | |
| "learning_rate": 0.0001138467619245374, | |
| "loss": 0.5529, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.112016293279022, | |
| "grad_norm": 0.34103691577911377, | |
| "learning_rate": 0.00011314089963321119, | |
| "loss": 0.5546, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 5.132382892057026, | |
| "grad_norm": 0.34115347266197205, | |
| "learning_rate": 0.00011243437046474853, | |
| "loss": 0.5571, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 5.15274949083503, | |
| "grad_norm": 0.350697785615921, | |
| "learning_rate": 0.0001117272102742402, | |
| "loss": 0.5534, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 5.173116089613035, | |
| "grad_norm": 0.34019774198532104, | |
| "learning_rate": 0.00011101945494880012, | |
| "loss": 0.5536, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 5.193482688391039, | |
| "grad_norm": 0.34322109818458557, | |
| "learning_rate": 0.00011031114040574437, | |
| "loss": 0.5577, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 5.213849287169043, | |
| "grad_norm": 0.33681756258010864, | |
| "learning_rate": 0.00010960230259076818, | |
| "loss": 0.5555, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 5.234215885947047, | |
| "grad_norm": 0.335644394159317, | |
| "learning_rate": 0.00010889297747612202, | |
| "loss": 0.5559, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 5.254582484725051, | |
| "grad_norm": 0.33518001437187195, | |
| "learning_rate": 0.00010818320105878584, | |
| "loss": 0.5547, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 5.274949083503055, | |
| "grad_norm": 0.32509303092956543, | |
| "learning_rate": 0.00010747300935864243, | |
| "loss": 0.5539, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 5.295315682281059, | |
| "grad_norm": 0.33121034502983093, | |
| "learning_rate": 0.0001067624384166495, | |
| "loss": 0.5609, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.315682281059063, | |
| "grad_norm": 0.34210404753685, | |
| "learning_rate": 0.00010605152429301055, | |
| "loss": 0.5629, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 5.336048879837067, | |
| "grad_norm": 0.3202488422393799, | |
| "learning_rate": 0.0001053403030653449, | |
| "loss": 0.5616, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 5.356415478615071, | |
| "grad_norm": 0.33203092217445374, | |
| "learning_rate": 0.00010462881082685691, | |
| "loss": 0.5625, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 5.3767820773930755, | |
| "grad_norm": 0.37128540873527527, | |
| "learning_rate": 0.00010391708368450427, | |
| "loss": 0.5531, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 5.39714867617108, | |
| "grad_norm": 0.32188987731933594, | |
| "learning_rate": 0.00010320515775716555, | |
| "loss": 0.5559, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 5.417515274949084, | |
| "grad_norm": 0.33244988322257996, | |
| "learning_rate": 0.0001024930691738073, | |
| "loss": 0.5579, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 5.437881873727088, | |
| "grad_norm": 0.3311491310596466, | |
| "learning_rate": 0.00010178085407165066, | |
| "loss": 0.5545, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 5.458248472505091, | |
| "grad_norm": 0.3205713629722595, | |
| "learning_rate": 0.00010106854859433734, | |
| "loss": 0.5596, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 5.478615071283095, | |
| "grad_norm": 0.3317239284515381, | |
| "learning_rate": 0.00010035618889009535, | |
| "loss": 0.5669, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 5.4989816700611, | |
| "grad_norm": 0.3288976848125458, | |
| "learning_rate": 9.96438111099047e-05, | |
| "loss": 0.5633, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 5.519348268839104, | |
| "grad_norm": 0.33682990074157715, | |
| "learning_rate": 9.893145140566269e-05, | |
| "loss": 0.5603, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 5.539714867617108, | |
| "grad_norm": 0.3263765871524811, | |
| "learning_rate": 9.821914592834935e-05, | |
| "loss": 0.5604, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 5.560081466395112, | |
| "grad_norm": 0.3341580033302307, | |
| "learning_rate": 9.750693082619273e-05, | |
| "loss": 0.5627, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 5.580448065173116, | |
| "grad_norm": 0.3362567126750946, | |
| "learning_rate": 9.679484224283449e-05, | |
| "loss": 0.5595, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 5.6008146639511205, | |
| "grad_norm": 0.33394739031791687, | |
| "learning_rate": 9.608291631549574e-05, | |
| "loss": 0.5548, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 5.621181262729124, | |
| "grad_norm": 0.3413245975971222, | |
| "learning_rate": 9.537118917314311e-05, | |
| "loss": 0.5509, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 5.641547861507128, | |
| "grad_norm": 0.32499754428863525, | |
| "learning_rate": 9.46596969346551e-05, | |
| "loss": 0.5582, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 5.661914460285132, | |
| "grad_norm": 0.3440452218055725, | |
| "learning_rate": 9.39484757069895e-05, | |
| "loss": 0.5638, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 5.682281059063136, | |
| "grad_norm": 0.3295697271823883, | |
| "learning_rate": 9.323756158335053e-05, | |
| "loss": 0.5592, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 5.7026476578411405, | |
| "grad_norm": 0.32334938645362854, | |
| "learning_rate": 9.252699064135758e-05, | |
| "loss": 0.5609, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 5.723014256619145, | |
| "grad_norm": 0.33664849400520325, | |
| "learning_rate": 9.181679894121421e-05, | |
| "loss": 0.5584, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 5.743380855397149, | |
| "grad_norm": 0.3422396779060364, | |
| "learning_rate": 9.1107022523878e-05, | |
| "loss": 0.5553, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 5.763747454175153, | |
| "grad_norm": 0.3184050917625427, | |
| "learning_rate": 9.039769740923183e-05, | |
| "loss": 0.5604, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 5.784114052953157, | |
| "grad_norm": 0.326419860124588, | |
| "learning_rate": 8.968885959425567e-05, | |
| "loss": 0.5645, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 5.804480651731161, | |
| "grad_norm": 0.33074718713760376, | |
| "learning_rate": 8.898054505119989e-05, | |
| "loss": 0.5654, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 5.824847250509165, | |
| "grad_norm": 0.3398876488208771, | |
| "learning_rate": 8.827278972575983e-05, | |
| "loss": 0.5635, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 5.845213849287169, | |
| "grad_norm": 0.33468544483184814, | |
| "learning_rate": 8.756562953525152e-05, | |
| "loss": 0.5557, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 5.865580448065173, | |
| "grad_norm": 0.3315589427947998, | |
| "learning_rate": 8.685910036678883e-05, | |
| "loss": 0.5626, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 5.885947046843177, | |
| "grad_norm": 0.3360842764377594, | |
| "learning_rate": 8.615323807546258e-05, | |
| "loss": 0.5588, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 5.906313645621181, | |
| "grad_norm": 0.3325459063053131, | |
| "learning_rate": 8.54480784825207e-05, | |
| "loss": 0.5569, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 5.9266802443991855, | |
| "grad_norm": 0.3266001343727112, | |
| "learning_rate": 8.474365737355038e-05, | |
| "loss": 0.5617, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 5.94704684317719, | |
| "grad_norm": 0.3289891481399536, | |
| "learning_rate": 8.404001049666211e-05, | |
| "loss": 0.5572, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 5.967413441955194, | |
| "grad_norm": 0.32171157002449036, | |
| "learning_rate": 8.333717356067543e-05, | |
| "loss": 0.5571, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 5.987780040733197, | |
| "grad_norm": 0.3251187205314636, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 0.5585, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.9298146963119507, | |
| "eval_runtime": 0.3781, | |
| "eval_samples_per_second": 31.738, | |
| "eval_steps_per_second": 2.645, | |
| "step": 1473 | |
| }, | |
| { | |
| "epoch": 6.008146639511201, | |
| "grad_norm": 0.32297611236572266, | |
| "learning_rate": 8.193407213936012e-05, | |
| "loss": 0.5502, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 6.0285132382892055, | |
| "grad_norm": 0.32624468207359314, | |
| "learning_rate": 8.123387885891725e-05, | |
| "loss": 0.5355, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 6.04887983706721, | |
| "grad_norm": 0.34197378158569336, | |
| "learning_rate": 8.053463792553416e-05, | |
| "loss": 0.5415, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 6.069246435845214, | |
| "grad_norm": 0.3497610092163086, | |
| "learning_rate": 7.98363848244367e-05, | |
| "loss": 0.5418, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 6.089613034623218, | |
| "grad_norm": 0.35511311888694763, | |
| "learning_rate": 7.913915499071993e-05, | |
| "loss": 0.5353, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 6.109979633401222, | |
| "grad_norm": 0.33935195207595825, | |
| "learning_rate": 7.844298380755003e-05, | |
| "loss": 0.5371, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.130346232179226, | |
| "grad_norm": 0.33899983763694763, | |
| "learning_rate": 7.774790660436858e-05, | |
| "loss": 0.5379, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 6.1507128309572305, | |
| "grad_norm": 0.35454773902893066, | |
| "learning_rate": 7.705395865509947e-05, | |
| "loss": 0.5353, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 6.171079429735234, | |
| "grad_norm": 0.3619644045829773, | |
| "learning_rate": 7.636117517635912e-05, | |
| "loss": 0.5402, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 6.191446028513238, | |
| "grad_norm": 0.34515708684921265, | |
| "learning_rate": 7.566959132566915e-05, | |
| "loss": 0.5377, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 6.211812627291242, | |
| "grad_norm": 0.347515344619751, | |
| "learning_rate": 7.497924219967209e-05, | |
| "loss": 0.535, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 6.232179226069246, | |
| "grad_norm": 0.34020859003067017, | |
| "learning_rate": 7.429016283235053e-05, | |
| "loss": 0.5406, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 6.2525458248472505, | |
| "grad_norm": 0.34483760595321655, | |
| "learning_rate": 7.360238819324903e-05, | |
| "loss": 0.5377, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 6.272912423625255, | |
| "grad_norm": 0.34881579875946045, | |
| "learning_rate": 7.291595318569951e-05, | |
| "loss": 0.5372, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 6.293279022403259, | |
| "grad_norm": 0.37821272015571594, | |
| "learning_rate": 7.223089264505e-05, | |
| "loss": 0.5383, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 6.313645621181263, | |
| "grad_norm": 0.35509517788887024, | |
| "learning_rate": 7.154724133689677e-05, | |
| "loss": 0.5464, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 6.334012219959266, | |
| "grad_norm": 0.34138378500938416, | |
| "learning_rate": 7.086503395532012e-05, | |
| "loss": 0.5366, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 6.3543788187372705, | |
| "grad_norm": 0.3617904484272003, | |
| "learning_rate": 7.018430512112366e-05, | |
| "loss": 0.541, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 6.374745417515275, | |
| "grad_norm": 0.3442498743534088, | |
| "learning_rate": 6.950508938007729e-05, | |
| "loss": 0.5388, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 6.395112016293279, | |
| "grad_norm": 0.3558500111103058, | |
| "learning_rate": 6.88274212011642e-05, | |
| "loss": 0.5363, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 6.415478615071283, | |
| "grad_norm": 0.35669848322868347, | |
| "learning_rate": 6.815133497483157e-05, | |
| "loss": 0.5462, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 6.435845213849287, | |
| "grad_norm": 0.3383665084838867, | |
| "learning_rate": 6.74768650112453e-05, | |
| "loss": 0.5389, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 6.456211812627291, | |
| "grad_norm": 0.3450273275375366, | |
| "learning_rate": 6.680404553854885e-05, | |
| "loss": 0.5378, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 6.4765784114052956, | |
| "grad_norm": 0.34580737352371216, | |
| "learning_rate": 6.613291070112624e-05, | |
| "loss": 0.5401, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.4969450101833, | |
| "grad_norm": 0.34363240003585815, | |
| "learning_rate": 6.546349455786926e-05, | |
| "loss": 0.5378, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 6.517311608961304, | |
| "grad_norm": 0.35278892517089844, | |
| "learning_rate": 6.479583108044899e-05, | |
| "loss": 0.5387, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.537678207739307, | |
| "grad_norm": 0.36023351550102234, | |
| "learning_rate": 6.412995415159197e-05, | |
| "loss": 0.5421, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 6.558044806517311, | |
| "grad_norm": 0.3527635633945465, | |
| "learning_rate": 6.34658975633605e-05, | |
| "loss": 0.5469, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 6.5784114052953155, | |
| "grad_norm": 0.3474122881889343, | |
| "learning_rate": 6.28036950154379e-05, | |
| "loss": 0.5468, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 6.59877800407332, | |
| "grad_norm": 0.336265504360199, | |
| "learning_rate": 6.214338011341824e-05, | |
| "loss": 0.5459, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 6.619144602851324, | |
| "grad_norm": 0.3502029776573181, | |
| "learning_rate": 6.148498636710092e-05, | |
| "loss": 0.5399, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 6.639511201629328, | |
| "grad_norm": 0.3454211950302124, | |
| "learning_rate": 6.082854718879021e-05, | |
| "loss": 0.5376, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 6.659877800407332, | |
| "grad_norm": 0.3441862165927887, | |
| "learning_rate": 6.017409589159946e-05, | |
| "loss": 0.543, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 6.680244399185336, | |
| "grad_norm": 0.33739784359931946, | |
| "learning_rate": 5.952166568776062e-05, | |
| "loss": 0.5441, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 6.70061099796334, | |
| "grad_norm": 0.3396100103855133, | |
| "learning_rate": 5.887128968693887e-05, | |
| "loss": 0.5471, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 6.720977596741344, | |
| "grad_norm": 0.35324475169181824, | |
| "learning_rate": 5.822300089455211e-05, | |
| "loss": 0.5408, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 6.741344195519348, | |
| "grad_norm": 0.3377906084060669, | |
| "learning_rate": 5.7576832210096245e-05, | |
| "loss": 0.5416, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 6.761710794297352, | |
| "grad_norm": 0.3413607180118561, | |
| "learning_rate": 5.6932816425475554e-05, | |
| "loss": 0.5425, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 6.782077393075356, | |
| "grad_norm": 0.3485510051250458, | |
| "learning_rate": 5.629098622333837e-05, | |
| "loss": 0.5459, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 6.802443991853361, | |
| "grad_norm": 0.3555828332901001, | |
| "learning_rate": 5.5651374175418656e-05, | |
| "loss": 0.5473, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 6.822810590631365, | |
| "grad_norm": 0.34203040599823, | |
| "learning_rate": 5.5014012740883115e-05, | |
| "loss": 0.539, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 6.843177189409369, | |
| "grad_norm": 0.34898054599761963, | |
| "learning_rate": 5.43789342646837e-05, | |
| "loss": 0.5388, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 6.863543788187373, | |
| "grad_norm": 0.3571433126926422, | |
| "learning_rate": 5.37461709759165e-05, | |
| "loss": 0.5486, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 6.883910386965377, | |
| "grad_norm": 0.34400326013565063, | |
| "learning_rate": 5.3115754986185774e-05, | |
| "loss": 0.545, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 6.904276985743381, | |
| "grad_norm": 0.3509522080421448, | |
| "learning_rate": 5.248771828797474e-05, | |
| "loss": 0.5438, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 6.924643584521385, | |
| "grad_norm": 0.33699506521224976, | |
| "learning_rate": 5.1862092753021754e-05, | |
| "loss": 0.5396, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 6.945010183299389, | |
| "grad_norm": 0.3432750105857849, | |
| "learning_rate": 5.123891013070288e-05, | |
| "loss": 0.5468, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 6.965376782077393, | |
| "grad_norm": 0.3491397500038147, | |
| "learning_rate": 5.061820204642085e-05, | |
| "loss": 0.5467, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 6.985743380855397, | |
| "grad_norm": 0.34977835416793823, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.5413, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 6.997963340122199, | |
| "eval_loss": 1.9540338516235352, | |
| "eval_runtime": 0.3818, | |
| "eval_samples_per_second": 31.43, | |
| "eval_steps_per_second": 2.619, | |
| "step": 1718 | |
| }, | |
| { | |
| "epoch": 7.006109979633401, | |
| "grad_norm": 0.35306668281555176, | |
| "learning_rate": 4.938433536408771e-05, | |
| "loss": 0.5349, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 7.026476578411406, | |
| "grad_norm": 0.3659164011478424, | |
| "learning_rate": 4.8771239382562287e-05, | |
| "loss": 0.5277, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 7.04684317718941, | |
| "grad_norm": 0.35312947630882263, | |
| "learning_rate": 4.8160743168947496e-05, | |
| "loss": 0.5283, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 7.067209775967413, | |
| "grad_norm": 0.37476396560668945, | |
| "learning_rate": 4.755287770483349e-05, | |
| "loss": 0.5254, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 7.087576374745417, | |
| "grad_norm": 0.37969648838043213, | |
| "learning_rate": 4.694767383830453e-05, | |
| "loss": 0.5267, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 7.107942973523421, | |
| "grad_norm": 0.3799395263195038, | |
| "learning_rate": 4.634516228237372e-05, | |
| "loss": 0.5191, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 7.128309572301426, | |
| "grad_norm": 0.3764006495475769, | |
| "learning_rate": 4.574537361342407e-05, | |
| "loss": 0.5243, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.14867617107943, | |
| "grad_norm": 0.34774523973464966, | |
| "learning_rate": 4.514833826965705e-05, | |
| "loss": 0.5261, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 7.169042769857434, | |
| "grad_norm": 0.3704371750354767, | |
| "learning_rate": 4.4554086549547715e-05, | |
| "loss": 0.5313, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 7.189409368635438, | |
| "grad_norm": 0.3685317039489746, | |
| "learning_rate": 4.3962648610307286e-05, | |
| "loss": 0.5301, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 7.209775967413442, | |
| "grad_norm": 0.35810336470603943, | |
| "learning_rate": 4.337405446635264e-05, | |
| "loss": 0.5253, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 7.2301425661914465, | |
| "grad_norm": 0.3616923987865448, | |
| "learning_rate": 4.278833398778306e-05, | |
| "loss": 0.5272, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 7.25050916496945, | |
| "grad_norm": 0.3660266399383545, | |
| "learning_rate": 4.2205516898864463e-05, | |
| "loss": 0.5304, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 7.270875763747454, | |
| "grad_norm": 0.36601680517196655, | |
| "learning_rate": 4.1625632776521037e-05, | |
| "loss": 0.5327, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 7.291242362525458, | |
| "grad_norm": 0.36126160621643066, | |
| "learning_rate": 4.1048711048834033e-05, | |
| "loss": 0.5256, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 7.311608961303462, | |
| "grad_norm": 0.3732444941997528, | |
| "learning_rate": 4.0474780993548566e-05, | |
| "loss": 0.5261, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 7.3319755600814664, | |
| "grad_norm": 0.36716046929359436, | |
| "learning_rate": 3.990387173658774e-05, | |
| "loss": 0.53, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.352342158859471, | |
| "grad_norm": 0.3701815903186798, | |
| "learning_rate": 3.933601225057446e-05, | |
| "loss": 0.532, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 7.372708757637475, | |
| "grad_norm": 0.3797333836555481, | |
| "learning_rate": 3.8771231353361326e-05, | |
| "loss": 0.524, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 7.393075356415479, | |
| "grad_norm": 0.364581435918808, | |
| "learning_rate": 3.820955770656798e-05, | |
| "loss": 0.524, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 7.413441955193482, | |
| "grad_norm": 0.36119377613067627, | |
| "learning_rate": 3.7651019814126654e-05, | |
| "loss": 0.5268, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 7.433808553971486, | |
| "grad_norm": 0.35705694556236267, | |
| "learning_rate": 3.7095646020835754e-05, | |
| "loss": 0.524, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 7.454175152749491, | |
| "grad_norm": 0.3591172993183136, | |
| "learning_rate": 3.654346451092129e-05, | |
| "loss": 0.5306, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 7.474541751527495, | |
| "grad_norm": 0.3496874272823334, | |
| "learning_rate": 3.5994503306606497e-05, | |
| "loss": 0.5292, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 7.494908350305499, | |
| "grad_norm": 0.3553576171398163, | |
| "learning_rate": 3.544879026669005e-05, | |
| "loss": 0.5273, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 7.515274949083503, | |
| "grad_norm": 0.3603392541408539, | |
| "learning_rate": 3.4906353085131914e-05, | |
| "loss": 0.5367, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 7.535641547861507, | |
| "grad_norm": 0.3681369721889496, | |
| "learning_rate": 3.436721928964819e-05, | |
| "loss": 0.5243, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 7.5560081466395115, | |
| "grad_norm": 0.36058658361434937, | |
| "learning_rate": 3.383141624031408e-05, | |
| "loss": 0.5177, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 7.576374745417516, | |
| "grad_norm": 0.3690408170223236, | |
| "learning_rate": 3.329897112817529e-05, | |
| "loss": 0.5168, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 7.59674134419552, | |
| "grad_norm": 0.3683675229549408, | |
| "learning_rate": 3.276991097386831e-05, | |
| "loss": 0.5259, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 7.617107942973523, | |
| "grad_norm": 0.3538174033164978, | |
| "learning_rate": 3.2244262626249075e-05, | |
| "loss": 0.5241, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 7.637474541751527, | |
| "grad_norm": 0.357802152633667, | |
| "learning_rate": 3.172205276103033e-05, | |
| "loss": 0.5203, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 7.6578411405295315, | |
| "grad_norm": 0.3668569028377533, | |
| "learning_rate": 3.120330787942815e-05, | |
| "loss": 0.5295, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 7.678207739307536, | |
| "grad_norm": 0.3501560389995575, | |
| "learning_rate": 3.068805430681675e-05, | |
| "loss": 0.534, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 7.69857433808554, | |
| "grad_norm": 0.35364505648612976, | |
| "learning_rate": 3.0176318191392726e-05, | |
| "loss": 0.5311, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 7.718940936863544, | |
| "grad_norm": 0.3579385578632355, | |
| "learning_rate": 2.966812550284803e-05, | |
| "loss": 0.5261, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 7.739307535641548, | |
| "grad_norm": 0.3566991090774536, | |
| "learning_rate": 2.916350203105207e-05, | |
| "loss": 0.5238, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 7.7596741344195515, | |
| "grad_norm": 0.3725852966308594, | |
| "learning_rate": 2.8662473384742773e-05, | |
| "loss": 0.5279, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 7.780040733197556, | |
| "grad_norm": 0.3531646132469177, | |
| "learning_rate": 2.8165064990227252e-05, | |
| "loss": 0.5266, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 7.80040733197556, | |
| "grad_norm": 0.36464017629623413, | |
| "learning_rate": 2.76713020900912e-05, | |
| "loss": 0.5345, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 7.820773930753564, | |
| "grad_norm": 0.36327338218688965, | |
| "learning_rate": 2.718120974191809e-05, | |
| "loss": 0.5211, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 7.841140529531568, | |
| "grad_norm": 0.3583681583404541, | |
| "learning_rate": 2.669481281701739e-05, | |
| "loss": 0.531, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 7.861507128309572, | |
| "grad_norm": 0.36573490500450134, | |
| "learning_rate": 2.6212135999162445e-05, | |
| "loss": 0.5236, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 7.8818737270875765, | |
| "grad_norm": 0.3562600910663605, | |
| "learning_rate": 2.573320378333789e-05, | |
| "loss": 0.5273, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 7.902240325865581, | |
| "grad_norm": 0.3540510833263397, | |
| "learning_rate": 2.525804047449648e-05, | |
| "loss": 0.53, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 7.922606924643585, | |
| "grad_norm": 0.35902467370033264, | |
| "learning_rate": 2.478667018632562e-05, | |
| "loss": 0.5273, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 7.942973523421589, | |
| "grad_norm": 0.37109845876693726, | |
| "learning_rate": 2.4319116840023813e-05, | |
| "loss": 0.5305, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 7.963340122199592, | |
| "grad_norm": 0.36434054374694824, | |
| "learning_rate": 2.3855404163086558e-05, | |
| "loss": 0.5287, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 7.9837067209775965, | |
| "grad_norm": 0.35995838046073914, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 0.5295, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.981424331665039, | |
| "eval_runtime": 0.3785, | |
| "eval_samples_per_second": 31.705, | |
| "eval_steps_per_second": 2.642, | |
| "step": 1964 | |
| }, | |
| { | |
| "epoch": 8.004073319755602, | |
| "grad_norm": 0.3557198941707611, | |
| "learning_rate": 2.2939594751557802e-05, | |
| "loss": 0.5225, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 8.024439918533606, | |
| "grad_norm": 0.362589955329895, | |
| "learning_rate": 2.248754449265483e-05, | |
| "loss": 0.5222, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 8.044806517311608, | |
| "grad_norm": 0.36456504464149475, | |
| "learning_rate": 2.2039427852134788e-05, | |
| "loss": 0.5174, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 8.065173116089612, | |
| "grad_norm": 0.35784122347831726, | |
| "learning_rate": 2.1595267571115163e-05, | |
| "loss": 0.5102, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 8.085539714867616, | |
| "grad_norm": 0.3696603775024414, | |
| "learning_rate": 2.1155086189935224e-05, | |
| "loss": 0.5194, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 8.10590631364562, | |
| "grad_norm": 0.3742619454860687, | |
| "learning_rate": 2.0718906047012242e-05, | |
| "loss": 0.5176, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 8.126272912423625, | |
| "grad_norm": 0.368327260017395, | |
| "learning_rate": 2.0286749277707782e-05, | |
| "loss": 0.5175, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 8.146639511201629, | |
| "grad_norm": 0.3673515319824219, | |
| "learning_rate": 1.985863781320435e-05, | |
| "loss": 0.5188, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.167006109979633, | |
| "grad_norm": 0.3596293330192566, | |
| "learning_rate": 1.9434593379392562e-05, | |
| "loss": 0.5179, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 8.187372708757637, | |
| "grad_norm": 0.36756983399391174, | |
| "learning_rate": 1.9014637495768483e-05, | |
| "loss": 0.5137, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 8.207739307535642, | |
| "grad_norm": 0.3634319603443146, | |
| "learning_rate": 1.8598791474341514e-05, | |
| "loss": 0.5173, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 8.228105906313646, | |
| "grad_norm": 0.37804871797561646, | |
| "learning_rate": 1.8187076418552974e-05, | |
| "loss": 0.5177, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 8.24847250509165, | |
| "grad_norm": 0.3717120885848999, | |
| "learning_rate": 1.777951322220508e-05, | |
| "loss": 0.5151, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 8.268839103869654, | |
| "grad_norm": 0.3710237443447113, | |
| "learning_rate": 1.7376122568400532e-05, | |
| "loss": 0.5186, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 8.289205702647658, | |
| "grad_norm": 0.369852215051651, | |
| "learning_rate": 1.697692492849299e-05, | |
| "loss": 0.5146, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 8.309572301425662, | |
| "grad_norm": 0.3662493824958801, | |
| "learning_rate": 1.658194056104825e-05, | |
| "loss": 0.5163, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 8.329938900203667, | |
| "grad_norm": 0.3654170334339142, | |
| "learning_rate": 1.619118951081594e-05, | |
| "loss": 0.5182, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 8.35030549898167, | |
| "grad_norm": 0.3634709119796753, | |
| "learning_rate": 1.580469160771253e-05, | |
| "loss": 0.5182, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 8.370672097759675, | |
| "grad_norm": 0.3728494942188263, | |
| "learning_rate": 1.54224664658148e-05, | |
| "loss": 0.5193, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 8.391038696537677, | |
| "grad_norm": 0.3711146116256714, | |
| "learning_rate": 1.504453348236461e-05, | |
| "loss": 0.5183, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 8.411405295315681, | |
| "grad_norm": 0.36539244651794434, | |
| "learning_rate": 1.467091183678444e-05, | |
| "loss": 0.5167, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 8.431771894093686, | |
| "grad_norm": 0.367108017206192, | |
| "learning_rate": 1.430162048970407e-05, | |
| "loss": 0.525, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 8.45213849287169, | |
| "grad_norm": 0.3688015341758728, | |
| "learning_rate": 1.3936678181998374e-05, | |
| "loss": 0.5202, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 8.472505091649694, | |
| "grad_norm": 0.36969494819641113, | |
| "learning_rate": 1.357610343383634e-05, | |
| "loss": 0.5214, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 8.492871690427698, | |
| "grad_norm": 0.36953508853912354, | |
| "learning_rate": 1.3219914543741008e-05, | |
| "loss": 0.5217, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 8.513238289205702, | |
| "grad_norm": 0.3687889575958252, | |
| "learning_rate": 1.286812958766106e-05, | |
| "loss": 0.5153, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 8.533604887983707, | |
| "grad_norm": 0.36163973808288574, | |
| "learning_rate": 1.2520766418053408e-05, | |
| "loss": 0.5167, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 8.55397148676171, | |
| "grad_norm": 0.37320175766944885, | |
| "learning_rate": 1.2177842662977135e-05, | |
| "loss": 0.5144, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.574338085539715, | |
| "grad_norm": 0.36868005990982056, | |
| "learning_rate": 1.1839375725199098e-05, | |
| "loss": 0.5227, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 8.594704684317719, | |
| "grad_norm": 0.36514589190483093, | |
| "learning_rate": 1.1505382781310559e-05, | |
| "loss": 0.5157, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 8.615071283095723, | |
| "grad_norm": 0.3700886070728302, | |
| "learning_rate": 1.1175880780855608e-05, | |
| "loss": 0.5187, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 8.635437881873727, | |
| "grad_norm": 0.3674730956554413, | |
| "learning_rate": 1.0850886445471054e-05, | |
| "loss": 0.5153, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 8.655804480651732, | |
| "grad_norm": 0.37101948261260986, | |
| "learning_rate": 1.0530416268037702e-05, | |
| "loss": 0.5149, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 8.676171079429736, | |
| "grad_norm": 0.36893007159233093, | |
| "learning_rate": 1.021448651184349e-05, | |
| "loss": 0.5183, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 8.69653767820774, | |
| "grad_norm": 0.3680102527141571, | |
| "learning_rate": 9.903113209758096e-06, | |
| "loss": 0.5163, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 8.716904276985744, | |
| "grad_norm": 0.36099961400032043, | |
| "learning_rate": 9.596312163419274e-06, | |
| "loss": 0.5188, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 8.737270875763748, | |
| "grad_norm": 0.36749574542045593, | |
| "learning_rate": 9.294098942430996e-06, | |
| "loss": 0.5181, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 8.757637474541752, | |
| "grad_norm": 0.36886829137802124, | |
| "learning_rate": 8.99648888357335e-06, | |
| "loss": 0.5171, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 8.778004073319755, | |
| "grad_norm": 0.3683699071407318, | |
| "learning_rate": 8.703497090024116e-06, | |
| "loss": 0.5118, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 8.798370672097759, | |
| "grad_norm": 0.3667095899581909, | |
| "learning_rate": 8.415138430592428e-06, | |
| "loss": 0.5157, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 8.818737270875763, | |
| "grad_norm": 0.36200326681137085, | |
| "learning_rate": 8.131427538964164e-06, | |
| "loss": 0.5167, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 8.839103869653767, | |
| "grad_norm": 0.3657187521457672, | |
| "learning_rate": 7.852378812959227e-06, | |
| "loss": 0.5155, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 8.859470468431772, | |
| "grad_norm": 0.3640839457511902, | |
| "learning_rate": 7.578006413801075e-06, | |
| "loss": 0.5179, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 8.879837067209776, | |
| "grad_norm": 0.3747439980506897, | |
| "learning_rate": 7.308324265397836e-06, | |
| "loss": 0.5154, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 8.90020366598778, | |
| "grad_norm": 0.366439551115036, | |
| "learning_rate": 7.0433460536358685e-06, | |
| "loss": 0.5147, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 8.920570264765784, | |
| "grad_norm": 0.37617239356040955, | |
| "learning_rate": 6.783085225685148e-06, | |
| "loss": 0.52, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 8.940936863543788, | |
| "grad_norm": 0.3725389540195465, | |
| "learning_rate": 6.527554989316897e-06, | |
| "loss": 0.517, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 8.961303462321792, | |
| "grad_norm": 0.3703247606754303, | |
| "learning_rate": 6.276768312233228e-06, | |
| "loss": 0.52, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 8.981670061099797, | |
| "grad_norm": 0.35923415422439575, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 0.5154, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 8.9979633401222, | |
| "eval_loss": 1.9978961944580078, | |
| "eval_runtime": 0.3817, | |
| "eval_samples_per_second": 31.441, | |
| "eval_steps_per_second": 2.62, | |
| "step": 2209 | |
| }, | |
| { | |
| "epoch": 9.0020366598778, | |
| "grad_norm": 0.35924673080444336, | |
| "learning_rate": 5.789476302446662e-06, | |
| "loss": 0.5142, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 9.022403258655805, | |
| "grad_norm": 0.3611084818840027, | |
| "learning_rate": 5.552995698941088e-06, | |
| "loss": 0.5042, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 9.04276985743381, | |
| "grad_norm": 0.3653542101383209, | |
| "learning_rate": 5.321308111859791e-06, | |
| "loss": 0.514, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 9.063136456211813, | |
| "grad_norm": 0.36559605598449707, | |
| "learning_rate": 5.094425298933136e-06, | |
| "loss": 0.508, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 9.083503054989817, | |
| "grad_norm": 0.3634701669216156, | |
| "learning_rate": 4.872358774057806e-06, | |
| "loss": 0.5073, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 9.103869653767822, | |
| "grad_norm": 0.36651164293289185, | |
| "learning_rate": 4.655119806712482e-06, | |
| "loss": 0.5087, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 9.124236252545824, | |
| "grad_norm": 0.36734244227409363, | |
| "learning_rate": 4.442719421385922e-06, | |
| "loss": 0.5099, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 9.144602851323828, | |
| "grad_norm": 0.3621695339679718, | |
| "learning_rate": 4.235168397017541e-06, | |
| "loss": 0.5093, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 9.164969450101832, | |
| "grad_norm": 0.36267393827438354, | |
| "learning_rate": 4.0324772664503296e-06, | |
| "loss": 0.5087, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.185336048879837, | |
| "grad_norm": 0.36996185779571533, | |
| "learning_rate": 3.8346563158963785e-06, | |
| "loss": 0.5129, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 9.20570264765784, | |
| "grad_norm": 0.3650117814540863, | |
| "learning_rate": 3.641715584414862e-06, | |
| "loss": 0.5145, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 9.226069246435845, | |
| "grad_norm": 0.37506943941116333, | |
| "learning_rate": 3.453664863402595e-06, | |
| "loss": 0.5174, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 9.246435845213849, | |
| "grad_norm": 0.37266090512275696, | |
| "learning_rate": 3.270513696097055e-06, | |
| "loss": 0.5149, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 9.266802443991853, | |
| "grad_norm": 0.371489554643631, | |
| "learning_rate": 3.092271377092215e-06, | |
| "loss": 0.5155, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 9.287169042769857, | |
| "grad_norm": 0.36911362409591675, | |
| "learning_rate": 2.9189469518666967e-06, | |
| "loss": 0.5125, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 9.307535641547862, | |
| "grad_norm": 0.37086358666419983, | |
| "learning_rate": 2.7505492163248934e-06, | |
| "loss": 0.5116, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 9.327902240325866, | |
| "grad_norm": 0.37481385469436646, | |
| "learning_rate": 2.587086716350473e-06, | |
| "loss": 0.5127, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 9.34826883910387, | |
| "grad_norm": 0.3686206638813019, | |
| "learning_rate": 2.4285677473727118e-06, | |
| "loss": 0.5168, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 9.368635437881874, | |
| "grad_norm": 0.37265482544898987, | |
| "learning_rate": 2.2750003539455998e-06, | |
| "loss": 0.5067, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 9.389002036659878, | |
| "grad_norm": 0.36506715416908264, | |
| "learning_rate": 2.1263923293394774e-06, | |
| "loss": 0.5135, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 9.409368635437882, | |
| "grad_norm": 0.36821919679641724, | |
| "learning_rate": 1.9827512151456173e-06, | |
| "loss": 0.514, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 9.429735234215887, | |
| "grad_norm": 0.3649906814098358, | |
| "learning_rate": 1.8440843008934561e-06, | |
| "loss": 0.5127, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 9.45010183299389, | |
| "grad_norm": 0.36437344551086426, | |
| "learning_rate": 1.7103986236807313e-06, | |
| "loss": 0.5087, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 9.470468431771893, | |
| "grad_norm": 0.36268892884254456, | |
| "learning_rate": 1.5817009678162685e-06, | |
| "loss": 0.5157, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 9.490835030549897, | |
| "grad_norm": 0.37203824520111084, | |
| "learning_rate": 1.4579978644757464e-06, | |
| "loss": 0.5097, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 9.511201629327902, | |
| "grad_norm": 0.3669569492340088, | |
| "learning_rate": 1.339295591370271e-06, | |
| "loss": 0.5134, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 9.531568228105906, | |
| "grad_norm": 0.37304890155792236, | |
| "learning_rate": 1.2256001724277321e-06, | |
| "loss": 0.5177, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 9.55193482688391, | |
| "grad_norm": 0.36801040172576904, | |
| "learning_rate": 1.1169173774871478e-06, | |
| "loss": 0.5135, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 9.572301425661914, | |
| "grad_norm": 0.3649773895740509, | |
| "learning_rate": 1.013252722005842e-06, | |
| "loss": 0.5096, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 9.592668024439918, | |
| "grad_norm": 0.3708580732345581, | |
| "learning_rate": 9.146114667795358e-07, | |
| "loss": 0.5084, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 9.613034623217922, | |
| "grad_norm": 0.3649328351020813, | |
| "learning_rate": 8.209986176753948e-07, | |
| "loss": 0.5143, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 9.633401221995927, | |
| "grad_norm": 0.3666220009326935, | |
| "learning_rate": 7.324189253779312e-07, | |
| "loss": 0.508, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 9.65376782077393, | |
| "grad_norm": 0.36750587821006775, | |
| "learning_rate": 6.488768851480087e-07, | |
| "loss": 0.51, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 9.674134419551935, | |
| "grad_norm": 0.36811190843582153, | |
| "learning_rate": 5.703767365946466e-07, | |
| "loss": 0.512, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 9.69450101832994, | |
| "grad_norm": 0.37304404377937317, | |
| "learning_rate": 4.969224634598591e-07, | |
| "loss": 0.5131, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 9.714867617107943, | |
| "grad_norm": 0.37467247247695923, | |
| "learning_rate": 4.2851779341654964e-07, | |
| "loss": 0.5167, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 9.735234215885948, | |
| "grad_norm": 0.37464529275894165, | |
| "learning_rate": 3.651661978793075e-07, | |
| "loss": 0.5116, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 9.755600814663952, | |
| "grad_norm": 0.36434635519981384, | |
| "learning_rate": 3.068708918281926e-07, | |
| "loss": 0.5088, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 9.775967413441956, | |
| "grad_norm": 0.3670872747898102, | |
| "learning_rate": 2.536348336456551e-07, | |
| "loss": 0.5172, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 9.79633401221996, | |
| "grad_norm": 0.36443930864334106, | |
| "learning_rate": 2.054607249663665e-07, | |
| "loss": 0.5135, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 9.816700610997962, | |
| "grad_norm": 0.3700142204761505, | |
| "learning_rate": 1.6235101054011824e-07, | |
| "loss": 0.5085, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 9.837067209775967, | |
| "grad_norm": 0.3709539473056793, | |
| "learning_rate": 1.2430787810776555e-07, | |
| "loss": 0.5164, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 9.85743380855397, | |
| "grad_norm": 0.36750784516334534, | |
| "learning_rate": 9.133325829017158e-08, | |
| "loss": 0.5163, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 9.877800407331975, | |
| "grad_norm": 0.3683048188686371, | |
| "learning_rate": 6.342882449029696e-08, | |
| "loss": 0.5113, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 9.89816700610998, | |
| "grad_norm": 0.37231889367103577, | |
| "learning_rate": 4.059599280819004e-08, | |
| "loss": 0.5136, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 9.918533604887983, | |
| "grad_norm": 0.36660727858543396, | |
| "learning_rate": 2.2835921969210917e-08, | |
| "loss": 0.5183, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 9.938900203665987, | |
| "grad_norm": 0.3696270287036896, | |
| "learning_rate": 1.0149513265145238e-08, | |
| "loss": 0.5079, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 9.959266802443992, | |
| "grad_norm": 0.36083751916885376, | |
| "learning_rate": 2.5374105085518295e-09, | |
| "loss": 0.5151, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 9.979633401221996, | |
| "grad_norm": 0.3677830100059509, | |
| "learning_rate": 0.0, | |
| "loss": 0.508, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 9.979633401221996, | |
| "eval_loss": 2.0012366771698, | |
| "eval_runtime": 0.3701, | |
| "eval_samples_per_second": 32.425, | |
| "eval_steps_per_second": 2.702, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 9.979633401221996, | |
| "step": 2450, | |
| "total_flos": 6.866432485157241e+18, | |
| "train_loss": 0.5986210168137842, | |
| "train_runtime": 9994.3917, | |
| "train_samples_per_second": 15.694, | |
| "train_steps_per_second": 0.245 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 2450, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.866432485157241e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |