| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 626, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01597444089456869, | |
| "grad_norm": 83.71863294460064, | |
| "learning_rate": 2e-05, | |
| "loss": 3.3466, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03194888178913738, | |
| "grad_norm": 8.309768229761245, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5926, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04792332268370607, | |
| "grad_norm": 5.5501283977443014, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9672, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06389776357827476, | |
| "grad_norm": 7.968011212878682, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8781, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07987220447284345, | |
| "grad_norm": 5.773155689191023, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7334, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09584664536741214, | |
| "grad_norm": 12.419688553572414, | |
| "learning_rate": 2e-05, | |
| "loss": 0.587, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11182108626198083, | |
| "grad_norm": 5.2316684575507475, | |
| "learning_rate": 2e-05, | |
| "loss": 0.415, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12779552715654952, | |
| "grad_norm": 5.527971827359797, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3943, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14376996805111822, | |
| "grad_norm": 3.888479455006538, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3001, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1597444089456869, | |
| "grad_norm": 3.565885610442197, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3127, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1757188498402556, | |
| "grad_norm": 3.8283750979248086, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2748, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 3.959745580093837, | |
| "learning_rate": 2e-05, | |
| "loss": 0.294, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.20766773162939298, | |
| "grad_norm": 3.427782759656506, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2843, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22364217252396165, | |
| "grad_norm": 3.54917757478626, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2709, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23961661341853036, | |
| "grad_norm": 3.998859283435439, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2334, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.25559105431309903, | |
| "grad_norm": 2.9777041483361493, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2673, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2715654952076677, | |
| "grad_norm": 3.1903834947337333, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2615, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.28753993610223644, | |
| "grad_norm": 3.4532885387117593, | |
| "learning_rate": 2e-05, | |
| "loss": 0.237, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3035143769968051, | |
| "grad_norm": 3.848662534075994, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2765, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3194888178913738, | |
| "grad_norm": 2.592443413879922, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2807, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3354632587859425, | |
| "grad_norm": 3.213168469956371, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2456, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3514376996805112, | |
| "grad_norm": 3.074169636941117, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2503, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.36741214057507987, | |
| "grad_norm": 2.417392170818873, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2495, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 3.114022999875041, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2283, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3993610223642173, | |
| "grad_norm": 3.5346691978020033, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2487, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.41533546325878595, | |
| "grad_norm": 2.8874485043167324, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2314, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.43130990415335463, | |
| "grad_norm": 2.6236532586559655, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2366, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4472843450479233, | |
| "grad_norm": 4.123647817876358, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2323, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.46325878594249204, | |
| "grad_norm": 3.017220393678832, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2676, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4792332268370607, | |
| "grad_norm": 3.5417915594547313, | |
| "learning_rate": 2e-05, | |
| "loss": 0.209, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4952076677316294, | |
| "grad_norm": 2.7257738584133846, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2327, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5111821086261981, | |
| "grad_norm": 3.744319886800019, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2752, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5271565495207667, | |
| "grad_norm": 3.708460813306105, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2473, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5431309904153354, | |
| "grad_norm": 2.629137546032223, | |
| "learning_rate": 2e-05, | |
| "loss": 0.207, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5591054313099042, | |
| "grad_norm": 2.5819587327031304, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2123, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 2.2655265122275843, | |
| "learning_rate": 2e-05, | |
| "loss": 0.191, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5910543130990416, | |
| "grad_norm": 2.6804526214625497, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2336, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6070287539936102, | |
| "grad_norm": 3.2628819241122047, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2293, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6230031948881789, | |
| "grad_norm": 2.3619065168507727, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2346, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6389776357827476, | |
| "grad_norm": 2.6055177798386575, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2297, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6549520766773163, | |
| "grad_norm": 2.279009387467159, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2059, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.670926517571885, | |
| "grad_norm": 2.2868092833043203, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2041, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6869009584664537, | |
| "grad_norm": 2.2145027994683835, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1944, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7028753993610224, | |
| "grad_norm": 2.7534202573670945, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2254, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7188498402555911, | |
| "grad_norm": 2.8212765347556243, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2314, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7348242811501597, | |
| "grad_norm": 3.041623121354941, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1936, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7507987220447284, | |
| "grad_norm": 2.5503401281825444, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2034, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 1.4590655767845273, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1889, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7827476038338658, | |
| "grad_norm": 2.0684938511926743, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2074, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.7987220447284346, | |
| "grad_norm": 2.2882466114625215, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1831, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8146964856230032, | |
| "grad_norm": 1.9988547836999553, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1839, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8306709265175719, | |
| "grad_norm": 3.0907590290493734, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1845, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8466453674121406, | |
| "grad_norm": 2.1647755741271673, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2048, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8626198083067093, | |
| "grad_norm": 2.6824247155851855, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1982, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8785942492012779, | |
| "grad_norm": 3.5383743017319724, | |
| "learning_rate": 2e-05, | |
| "loss": 0.181, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.8945686900958466, | |
| "grad_norm": 2.8311929588357314, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1947, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9105431309904153, | |
| "grad_norm": 1.9233735249140267, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1908, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9265175718849841, | |
| "grad_norm": 2.5582302790137716, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1843, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9424920127795527, | |
| "grad_norm": 2.815385399078734, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2073, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 2.0523335977866792, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2093, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9744408945686901, | |
| "grad_norm": 2.4835341876106254, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1911, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.9904153354632588, | |
| "grad_norm": 1.89871874035203, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1944, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0063897763578276, | |
| "grad_norm": 1.9465093494946675, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1646, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.0223642172523961, | |
| "grad_norm": 2.1130020505391043, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1165, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.038338658146965, | |
| "grad_norm": 2.4497863262645465, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1286, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0543130990415335, | |
| "grad_norm": 1.9251221087363877, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1316, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0702875399361023, | |
| "grad_norm": 3.163156158004402, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1391, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.0862619808306708, | |
| "grad_norm": 1.5544243471103054, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1378, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1022364217252396, | |
| "grad_norm": 1.8300180739910277, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1426, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.1182108626198084, | |
| "grad_norm": 4.267862119310627, | |
| "learning_rate": 2e-05, | |
| "loss": 0.123, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.134185303514377, | |
| "grad_norm": 2.2304810600659177, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1651, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 2.394550559267119, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1358, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.1661341853035143, | |
| "grad_norm": 1.7829450423345041, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1273, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.182108626198083, | |
| "grad_norm": 1.9929178446730769, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1246, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.1980830670926517, | |
| "grad_norm": 1.4920997161721226, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1054, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.2140575079872205, | |
| "grad_norm": 1.7830235203378444, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1528, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.230031948881789, | |
| "grad_norm": 2.2933280737888757, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1447, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.2460063897763578, | |
| "grad_norm": 2.030948890904391, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1188, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.2619808306709266, | |
| "grad_norm": 1.3988993353239416, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1255, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.2779552715654952, | |
| "grad_norm": 2.063734911750646, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1329, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.293929712460064, | |
| "grad_norm": 1.5683211620886752, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1503, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.3099041533546325, | |
| "grad_norm": 1.1469844339610817, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1243, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3258785942492013, | |
| "grad_norm": 1.914347504796738, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1412, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 1.408459123943724, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1192, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.3578274760383386, | |
| "grad_norm": 2.5298092191552675, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1254, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.3738019169329074, | |
| "grad_norm": 2.863611554543664, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1383, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.389776357827476, | |
| "grad_norm": 1.3937502030266669, | |
| "learning_rate": 2e-05, | |
| "loss": 0.123, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.4057507987220448, | |
| "grad_norm": 1.9311268379527509, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1464, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4217252396166133, | |
| "grad_norm": 1.5812762702306977, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1282, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.4376996805111821, | |
| "grad_norm": 1.4187576510125102, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1442, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.4536741214057507, | |
| "grad_norm": 2.3028021469112385, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1393, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.4696485623003195, | |
| "grad_norm": 1.6201932621396478, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1424, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.4856230031948883, | |
| "grad_norm": 6.491080482525731, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1645, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.5015974440894568, | |
| "grad_norm": 2.58391255169585, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1334, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5175718849840254, | |
| "grad_norm": 1.3586070147565672, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1315, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 2.733123473010616, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1314, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.549520766773163, | |
| "grad_norm": 1.3304270077738738, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1275, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.5654952076677318, | |
| "grad_norm": 1.9676141375354392, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1341, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.5814696485623003, | |
| "grad_norm": 1.2246193545539361, | |
| "learning_rate": 2e-05, | |
| "loss": 0.118, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.5974440894568689, | |
| "grad_norm": 1.8280491286099871, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1327, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6134185303514377, | |
| "grad_norm": 1.6446300714378677, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1324, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.6293929712460065, | |
| "grad_norm": 1.6264106258984186, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1181, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.645367412140575, | |
| "grad_norm": 1.2565344330053252, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1242, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.6613418530351438, | |
| "grad_norm": 3.531193389155649, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1449, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.6773162939297124, | |
| "grad_norm": 1.6720594005561888, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1241, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.6932907348242812, | |
| "grad_norm": 1.5686698135340418, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1313, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.70926517571885, | |
| "grad_norm": 1.349019597797907, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1204, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 1.6886302104516848, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1282, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.741214057507987, | |
| "grad_norm": 1.182710098581308, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1103, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.7571884984025559, | |
| "grad_norm": 1.561143117049643, | |
| "learning_rate": 2e-05, | |
| "loss": 0.136, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.7731629392971247, | |
| "grad_norm": 1.7553371062029604, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1533, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.7891373801916934, | |
| "grad_norm": 2.0761702224590497, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1423, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.805111821086262, | |
| "grad_norm": 1.3807418410219519, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1594, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.8210862619808306, | |
| "grad_norm": 1.373853735669036, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1283, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8370607028753994, | |
| "grad_norm": 1.7182062272516478, | |
| "learning_rate": 2e-05, | |
| "loss": 0.112, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.8530351437699681, | |
| "grad_norm": 1.664613239770966, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1145, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.8690095846645367, | |
| "grad_norm": 1.5615436452993707, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1239, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.8849840255591053, | |
| "grad_norm": 2.5876340444653083, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1536, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.900958466453674, | |
| "grad_norm": 1.8052711043442962, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1325, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 1.081108063876394, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1145, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9329073482428116, | |
| "grad_norm": 2.0564096508677006, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1465, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.9488817891373802, | |
| "grad_norm": 1.3664466495445207, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1168, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.9648562300319488, | |
| "grad_norm": 1.8759113558973934, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1324, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.9808306709265175, | |
| "grad_norm": 2.2588179463096294, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1334, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.9968051118210863, | |
| "grad_norm": 1.4113722123835157, | |
| "learning_rate": 2e-05, | |
| "loss": 0.128, | |
| "step": 625 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 626, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 313, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5387801591808.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |