| { | |
| "best_metric": 0.15615242719650269, | |
| "best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles10_seed1_q3_dropout_v2/checkpoint-469", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 2345, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010660980810234541, | |
| "grad_norm": 4.6948442459106445, | |
| "learning_rate": 4.989339019189766e-05, | |
| "loss": 0.2577, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021321961620469083, | |
| "grad_norm": 4.7025957107543945, | |
| "learning_rate": 4.978678038379531e-05, | |
| "loss": 0.1748, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031982942430703626, | |
| "grad_norm": 4.207946300506592, | |
| "learning_rate": 4.9680170575692967e-05, | |
| "loss": 0.2076, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042643923240938165, | |
| "grad_norm": 3.57958984375, | |
| "learning_rate": 4.957356076759062e-05, | |
| "loss": 0.164, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053304904051172705, | |
| "grad_norm": 2.662198305130005, | |
| "learning_rate": 4.9466950959488276e-05, | |
| "loss": 0.2137, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06396588486140725, | |
| "grad_norm": 5.8198089599609375, | |
| "learning_rate": 4.936034115138593e-05, | |
| "loss": 0.2325, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07462686567164178, | |
| "grad_norm": 8.999914169311523, | |
| "learning_rate": 4.9253731343283586e-05, | |
| "loss": 0.2484, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08528784648187633, | |
| "grad_norm": 5.918586254119873, | |
| "learning_rate": 4.914712153518124e-05, | |
| "loss": 0.3532, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09594882729211088, | |
| "grad_norm": 2.1160314083099365, | |
| "learning_rate": 4.904051172707889e-05, | |
| "loss": 0.1792, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10660980810234541, | |
| "grad_norm": 3.5747337341308594, | |
| "learning_rate": 4.893390191897655e-05, | |
| "loss": 0.1999, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11727078891257996, | |
| "grad_norm": 3.06170392036438, | |
| "learning_rate": 4.88272921108742e-05, | |
| "loss": 0.1272, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1279317697228145, | |
| "grad_norm": 3.509375810623169, | |
| "learning_rate": 4.872068230277186e-05, | |
| "loss": 0.1969, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13859275053304904, | |
| "grad_norm": 2.0117759704589844, | |
| "learning_rate": 4.861407249466951e-05, | |
| "loss": 0.1686, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14925373134328357, | |
| "grad_norm": 10.117158889770508, | |
| "learning_rate": 4.850746268656717e-05, | |
| "loss": 0.1797, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15991471215351813, | |
| "grad_norm": 4.7419753074646, | |
| "learning_rate": 4.840085287846482e-05, | |
| "loss": 0.199, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17057569296375266, | |
| "grad_norm": 2.7047812938690186, | |
| "learning_rate": 4.829424307036248e-05, | |
| "loss": 0.1092, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1812366737739872, | |
| "grad_norm": 7.600207328796387, | |
| "learning_rate": 4.8187633262260126e-05, | |
| "loss": 0.2466, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19189765458422176, | |
| "grad_norm": 5.312893867492676, | |
| "learning_rate": 4.808102345415779e-05, | |
| "loss": 0.1432, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2025586353944563, | |
| "grad_norm": 3.3284530639648438, | |
| "learning_rate": 4.7974413646055436e-05, | |
| "loss": 0.2515, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21321961620469082, | |
| "grad_norm": 3.6826746463775635, | |
| "learning_rate": 4.78678038379531e-05, | |
| "loss": 0.2147, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22388059701492538, | |
| "grad_norm": 4.016763210296631, | |
| "learning_rate": 4.7761194029850745e-05, | |
| "loss": 0.181, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2345415778251599, | |
| "grad_norm": 3.3017539978027344, | |
| "learning_rate": 4.765458422174841e-05, | |
| "loss": 0.1933, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24520255863539445, | |
| "grad_norm": 2.381120443344116, | |
| "learning_rate": 4.7547974413646055e-05, | |
| "loss": 0.235, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.255863539445629, | |
| "grad_norm": 5.675516605377197, | |
| "learning_rate": 4.7441364605543716e-05, | |
| "loss": 0.1975, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26652452025586354, | |
| "grad_norm": 3.24746036529541, | |
| "learning_rate": 4.7334754797441364e-05, | |
| "loss": 0.1543, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2771855010660981, | |
| "grad_norm": 2.446378469467163, | |
| "learning_rate": 4.7228144989339026e-05, | |
| "loss": 0.1585, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2878464818763326, | |
| "grad_norm": 3.961570978164673, | |
| "learning_rate": 4.7121535181236674e-05, | |
| "loss": 0.1647, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 5.520444393157959, | |
| "learning_rate": 4.7014925373134335e-05, | |
| "loss": 0.1789, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3091684434968017, | |
| "grad_norm": 3.7484569549560547, | |
| "learning_rate": 4.690831556503198e-05, | |
| "loss": 0.1416, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31982942430703626, | |
| "grad_norm": 4.208812713623047, | |
| "learning_rate": 4.6801705756929645e-05, | |
| "loss": 0.1616, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3304904051172708, | |
| "grad_norm": 4.549439430236816, | |
| "learning_rate": 4.669509594882729e-05, | |
| "loss": 0.3085, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3411513859275053, | |
| "grad_norm": 3.725412368774414, | |
| "learning_rate": 4.658848614072495e-05, | |
| "loss": 0.2696, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.35181236673773986, | |
| "grad_norm": 5.272983551025391, | |
| "learning_rate": 4.64818763326226e-05, | |
| "loss": 0.2292, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3624733475479744, | |
| "grad_norm": 1.6885418891906738, | |
| "learning_rate": 4.637526652452026e-05, | |
| "loss": 0.2302, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.373134328358209, | |
| "grad_norm": 3.5256693363189697, | |
| "learning_rate": 4.626865671641791e-05, | |
| "loss": 0.165, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3837953091684435, | |
| "grad_norm": 4.1506667137146, | |
| "learning_rate": 4.6162046908315566e-05, | |
| "loss": 0.1995, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39445628997867804, | |
| "grad_norm": 1.618788242340088, | |
| "learning_rate": 4.605543710021322e-05, | |
| "loss": 0.1993, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4051172707889126, | |
| "grad_norm": 2.8863253593444824, | |
| "learning_rate": 4.5948827292110876e-05, | |
| "loss": 0.1852, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4157782515991471, | |
| "grad_norm": 3.384495973587036, | |
| "learning_rate": 4.584221748400853e-05, | |
| "loss": 0.1871, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 1.4212799072265625, | |
| "learning_rate": 4.5735607675906185e-05, | |
| "loss": 0.0884, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43710021321961623, | |
| "grad_norm": 1.3817472457885742, | |
| "learning_rate": 4.562899786780384e-05, | |
| "loss": 0.1222, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44776119402985076, | |
| "grad_norm": 6.522399425506592, | |
| "learning_rate": 4.5522388059701495e-05, | |
| "loss": 0.1232, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4584221748400853, | |
| "grad_norm": 9.160868644714355, | |
| "learning_rate": 4.541577825159915e-05, | |
| "loss": 0.2099, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4690831556503198, | |
| "grad_norm": 4.965953826904297, | |
| "learning_rate": 4.5309168443496804e-05, | |
| "loss": 0.2236, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.47974413646055436, | |
| "grad_norm": 3.8943636417388916, | |
| "learning_rate": 4.520255863539446e-05, | |
| "loss": 0.2133, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4904051172707889, | |
| "grad_norm": 2.267188549041748, | |
| "learning_rate": 4.5095948827292114e-05, | |
| "loss": 0.1897, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5010660980810234, | |
| "grad_norm": 2.608792781829834, | |
| "learning_rate": 4.498933901918977e-05, | |
| "loss": 0.2031, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.511727078891258, | |
| "grad_norm": 4.1853227615356445, | |
| "learning_rate": 4.488272921108742e-05, | |
| "loss": 0.1873, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5223880597014925, | |
| "grad_norm": 3.608839273452759, | |
| "learning_rate": 4.477611940298508e-05, | |
| "loss": 0.2248, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5330490405117271, | |
| "grad_norm": 2.9814040660858154, | |
| "learning_rate": 4.466950959488273e-05, | |
| "loss": 0.2086, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5437100213219617, | |
| "grad_norm": 2.4261651039123535, | |
| "learning_rate": 4.456289978678039e-05, | |
| "loss": 0.1263, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5543710021321961, | |
| "grad_norm": 3.3815131187438965, | |
| "learning_rate": 4.445628997867804e-05, | |
| "loss": 0.1827, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5650319829424307, | |
| "grad_norm": 2.225668430328369, | |
| "learning_rate": 4.43496801705757e-05, | |
| "loss": 0.1938, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5756929637526652, | |
| "grad_norm": 4.9899001121521, | |
| "learning_rate": 4.424307036247335e-05, | |
| "loss": 0.1288, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5863539445628998, | |
| "grad_norm": 2.687403917312622, | |
| "learning_rate": 4.4136460554371006e-05, | |
| "loss": 0.1566, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 3.935621738433838, | |
| "learning_rate": 4.402985074626866e-05, | |
| "loss": 0.17, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6076759061833689, | |
| "grad_norm": 4.6521220207214355, | |
| "learning_rate": 4.3923240938166316e-05, | |
| "loss": 0.1722, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6183368869936035, | |
| "grad_norm": 7.111056804656982, | |
| "learning_rate": 4.381663113006397e-05, | |
| "loss": 0.2513, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6289978678038379, | |
| "grad_norm": 4.924593925476074, | |
| "learning_rate": 4.3710021321961625e-05, | |
| "loss": 0.216, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6396588486140725, | |
| "grad_norm": 2.6263926029205322, | |
| "learning_rate": 4.360341151385928e-05, | |
| "loss": 0.1049, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.650319829424307, | |
| "grad_norm": 3.3580148220062256, | |
| "learning_rate": 4.3496801705756935e-05, | |
| "loss": 0.1506, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6609808102345416, | |
| "grad_norm": 1.0859957933425903, | |
| "learning_rate": 4.339019189765459e-05, | |
| "loss": 0.1685, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6716417910447762, | |
| "grad_norm": 2.492729902267456, | |
| "learning_rate": 4.328358208955224e-05, | |
| "loss": 0.1925, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6823027718550106, | |
| "grad_norm": 3.3099021911621094, | |
| "learning_rate": 4.31769722814499e-05, | |
| "loss": 0.1356, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6929637526652452, | |
| "grad_norm": 3.3860507011413574, | |
| "learning_rate": 4.307036247334755e-05, | |
| "loss": 0.1809, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7036247334754797, | |
| "grad_norm": 1.5362133979797363, | |
| "learning_rate": 4.29637526652452e-05, | |
| "loss": 0.1506, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 2.5398333072662354, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.1997, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7249466950959488, | |
| "grad_norm": 1.6685487031936646, | |
| "learning_rate": 4.275053304904051e-05, | |
| "loss": 0.2233, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7356076759061834, | |
| "grad_norm": 3.7195467948913574, | |
| "learning_rate": 4.2643923240938166e-05, | |
| "loss": 0.1445, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "grad_norm": 3.792914867401123, | |
| "learning_rate": 4.253731343283582e-05, | |
| "loss": 0.1184, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7569296375266524, | |
| "grad_norm": 5.353825569152832, | |
| "learning_rate": 4.2430703624733475e-05, | |
| "loss": 0.1075, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.767590618336887, | |
| "grad_norm": 3.4059877395629883, | |
| "learning_rate": 4.232409381663113e-05, | |
| "loss": 0.1319, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7782515991471215, | |
| "grad_norm": 4.120956897735596, | |
| "learning_rate": 4.2217484008528785e-05, | |
| "loss": 0.1762, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7889125799573561, | |
| "grad_norm": 4.725405216217041, | |
| "learning_rate": 4.211087420042644e-05, | |
| "loss": 0.1381, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7995735607675906, | |
| "grad_norm": 7.871543884277344, | |
| "learning_rate": 4.2004264392324094e-05, | |
| "loss": 0.2403, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8102345415778252, | |
| "grad_norm": 2.8368208408355713, | |
| "learning_rate": 4.189765458422175e-05, | |
| "loss": 0.2203, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8208955223880597, | |
| "grad_norm": 4.245978832244873, | |
| "learning_rate": 4.1791044776119404e-05, | |
| "loss": 0.2889, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8315565031982942, | |
| "grad_norm": 3.8861076831817627, | |
| "learning_rate": 4.168443496801706e-05, | |
| "loss": 0.2999, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8422174840085288, | |
| "grad_norm": 2.694441318511963, | |
| "learning_rate": 4.157782515991471e-05, | |
| "loss": 0.1575, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 2.9438793659210205, | |
| "learning_rate": 4.147121535181237e-05, | |
| "loss": 0.1195, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8635394456289979, | |
| "grad_norm": 2.5647032260894775, | |
| "learning_rate": 4.136460554371002e-05, | |
| "loss": 0.1078, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8742004264392325, | |
| "grad_norm": 2.117713451385498, | |
| "learning_rate": 4.125799573560768e-05, | |
| "loss": 0.1186, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8848614072494669, | |
| "grad_norm": 5.229939937591553, | |
| "learning_rate": 4.115138592750533e-05, | |
| "loss": 0.1825, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8955223880597015, | |
| "grad_norm": 4.133886814117432, | |
| "learning_rate": 4.104477611940299e-05, | |
| "loss": 0.1792, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.906183368869936, | |
| "grad_norm": 3.489717721939087, | |
| "learning_rate": 4.093816631130064e-05, | |
| "loss": 0.1701, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9168443496801706, | |
| "grad_norm": 5.602441787719727, | |
| "learning_rate": 4.0831556503198296e-05, | |
| "loss": 0.3422, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9275053304904051, | |
| "grad_norm": 5.082653045654297, | |
| "learning_rate": 4.072494669509595e-05, | |
| "loss": 0.1916, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9381663113006397, | |
| "grad_norm": 4.722070693969727, | |
| "learning_rate": 4.0618336886993606e-05, | |
| "loss": 0.1329, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9488272921108742, | |
| "grad_norm": 5.7544403076171875, | |
| "learning_rate": 4.051172707889126e-05, | |
| "loss": 0.1514, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9594882729211087, | |
| "grad_norm": 1.7018496990203857, | |
| "learning_rate": 4.0405117270788915e-05, | |
| "loss": 0.1318, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9701492537313433, | |
| "grad_norm": 2.369692325592041, | |
| "learning_rate": 4.029850746268657e-05, | |
| "loss": 0.1063, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9808102345415778, | |
| "grad_norm": 3.717768430709839, | |
| "learning_rate": 4.0191897654584225e-05, | |
| "loss": 0.1615, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9914712153518124, | |
| "grad_norm": 2.74760365486145, | |
| "learning_rate": 4.008528784648188e-05, | |
| "loss": 0.0722, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9413333333333334, | |
| "eval_loss": 0.15615242719650269, | |
| "eval_runtime": 20.1495, | |
| "eval_samples_per_second": 186.108, | |
| "eval_steps_per_second": 5.856, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.0021321961620469, | |
| "grad_norm": 3.6047048568725586, | |
| "learning_rate": 3.997867803837953e-05, | |
| "loss": 0.1423, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0127931769722816, | |
| "grad_norm": 0.9384483098983765, | |
| "learning_rate": 3.987206823027719e-05, | |
| "loss": 0.0635, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.023454157782516, | |
| "grad_norm": 0.3238335847854614, | |
| "learning_rate": 3.976545842217484e-05, | |
| "loss": 0.0486, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0341151385927505, | |
| "grad_norm": 1.3718501329421997, | |
| "learning_rate": 3.96588486140725e-05, | |
| "loss": 0.0535, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.044776119402985, | |
| "grad_norm": 7.383277893066406, | |
| "learning_rate": 3.9552238805970146e-05, | |
| "loss": 0.0473, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0554371002132197, | |
| "grad_norm": 0.861047625541687, | |
| "learning_rate": 3.944562899786781e-05, | |
| "loss": 0.0218, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.0660980810234542, | |
| "grad_norm": 1.9219954013824463, | |
| "learning_rate": 3.9339019189765456e-05, | |
| "loss": 0.0186, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0767590618336886, | |
| "grad_norm": 0.13769158720970154, | |
| "learning_rate": 3.923240938166312e-05, | |
| "loss": 0.0097, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0874200426439233, | |
| "grad_norm": 8.93731689453125, | |
| "learning_rate": 3.9125799573560765e-05, | |
| "loss": 0.0629, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0980810234541578, | |
| "grad_norm": 0.49480921030044556, | |
| "learning_rate": 3.901918976545843e-05, | |
| "loss": 0.1475, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.1087420042643923, | |
| "grad_norm": 11.072365760803223, | |
| "learning_rate": 3.8912579957356075e-05, | |
| "loss": 0.0834, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1194029850746268, | |
| "grad_norm": 11.185342788696289, | |
| "learning_rate": 3.8805970149253736e-05, | |
| "loss": 0.0339, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.1300639658848615, | |
| "grad_norm": 8.913257598876953, | |
| "learning_rate": 3.8699360341151384e-05, | |
| "loss": 0.0646, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.140724946695096, | |
| "grad_norm": 1.4803158044815063, | |
| "learning_rate": 3.8592750533049046e-05, | |
| "loss": 0.1098, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1513859275053304, | |
| "grad_norm": 1.3932523727416992, | |
| "learning_rate": 3.8486140724946694e-05, | |
| "loss": 0.0909, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.1620469083155651, | |
| "grad_norm": 0.3472418189048767, | |
| "learning_rate": 3.8379530916844355e-05, | |
| "loss": 0.0284, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.1727078891257996, | |
| "grad_norm": 7.002597332000732, | |
| "learning_rate": 3.8272921108742e-05, | |
| "loss": 0.0309, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.183368869936034, | |
| "grad_norm": 0.1289745569229126, | |
| "learning_rate": 3.8166311300639665e-05, | |
| "loss": 0.0724, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1940298507462686, | |
| "grad_norm": 12.98277473449707, | |
| "learning_rate": 3.805970149253731e-05, | |
| "loss": 0.0944, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.2046908315565032, | |
| "grad_norm": 0.5267494320869446, | |
| "learning_rate": 3.7953091684434974e-05, | |
| "loss": 0.0986, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.2153518123667377, | |
| "grad_norm": 0.43557122349739075, | |
| "learning_rate": 3.784648187633262e-05, | |
| "loss": 0.0395, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.2260127931769722, | |
| "grad_norm": 2.4160540103912354, | |
| "learning_rate": 3.7739872068230284e-05, | |
| "loss": 0.0465, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.236673773987207, | |
| "grad_norm": 0.7843585014343262, | |
| "learning_rate": 3.763326226012793e-05, | |
| "loss": 0.0332, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2473347547974414, | |
| "grad_norm": 1.8062599897384644, | |
| "learning_rate": 3.752665245202559e-05, | |
| "loss": 0.0527, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.2579957356076759, | |
| "grad_norm": 0.13011713325977325, | |
| "learning_rate": 3.742004264392324e-05, | |
| "loss": 0.0353, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2686567164179103, | |
| "grad_norm": 5.145373821258545, | |
| "learning_rate": 3.73134328358209e-05, | |
| "loss": 0.0402, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.279317697228145, | |
| "grad_norm": 10.741586685180664, | |
| "learning_rate": 3.720682302771855e-05, | |
| "loss": 0.0333, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2899786780383795, | |
| "grad_norm": 0.9409611225128174, | |
| "learning_rate": 3.710021321961621e-05, | |
| "loss": 0.0514, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.3006396588486142, | |
| "grad_norm": 4.306687355041504, | |
| "learning_rate": 3.699360341151386e-05, | |
| "loss": 0.1023, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3113006396588487, | |
| "grad_norm": 5.564180850982666, | |
| "learning_rate": 3.6886993603411515e-05, | |
| "loss": 0.0423, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.3219616204690832, | |
| "grad_norm": 0.5213475227355957, | |
| "learning_rate": 3.678038379530917e-05, | |
| "loss": 0.1393, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.3326226012793176, | |
| "grad_norm": 12.431807518005371, | |
| "learning_rate": 3.6673773987206824e-05, | |
| "loss": 0.0907, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.3432835820895521, | |
| "grad_norm": 8.96229362487793, | |
| "learning_rate": 3.656716417910448e-05, | |
| "loss": 0.1039, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.3539445628997868, | |
| "grad_norm": 0.44193053245544434, | |
| "learning_rate": 3.6460554371002134e-05, | |
| "loss": 0.1137, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.3646055437100213, | |
| "grad_norm": 5.031518459320068, | |
| "learning_rate": 3.635394456289979e-05, | |
| "loss": 0.0297, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.375266524520256, | |
| "grad_norm": 9.284253120422363, | |
| "learning_rate": 3.624733475479744e-05, | |
| "loss": 0.0922, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.3859275053304905, | |
| "grad_norm": 4.879660129547119, | |
| "learning_rate": 3.61407249466951e-05, | |
| "loss": 0.0956, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.396588486140725, | |
| "grad_norm": 4.541709899902344, | |
| "learning_rate": 3.603411513859275e-05, | |
| "loss": 0.1053, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.4072494669509594, | |
| "grad_norm": 7.22124719619751, | |
| "learning_rate": 3.592750533049041e-05, | |
| "loss": 0.0589, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.417910447761194, | |
| "grad_norm": 0.20107485353946686, | |
| "learning_rate": 3.582089552238806e-05, | |
| "loss": 0.0759, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 10.484548568725586, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.0934, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.439232409381663, | |
| "grad_norm": 0.565690815448761, | |
| "learning_rate": 3.560767590618337e-05, | |
| "loss": 0.0973, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.4498933901918978, | |
| "grad_norm": 1.9878708124160767, | |
| "learning_rate": 3.5501066098081026e-05, | |
| "loss": 0.0198, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4605543710021323, | |
| "grad_norm": 5.344141960144043, | |
| "learning_rate": 3.539445628997868e-05, | |
| "loss": 0.0794, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.4712153518123667, | |
| "grad_norm": 24.157583236694336, | |
| "learning_rate": 3.5287846481876336e-05, | |
| "loss": 0.0479, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4818763326226012, | |
| "grad_norm": 6.547579288482666, | |
| "learning_rate": 3.518123667377399e-05, | |
| "loss": 0.0337, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "grad_norm": 1.685778021812439, | |
| "learning_rate": 3.5074626865671645e-05, | |
| "loss": 0.1428, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5031982942430704, | |
| "grad_norm": 11.218913078308105, | |
| "learning_rate": 3.496801705756929e-05, | |
| "loss": 0.0738, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.5138592750533049, | |
| "grad_norm": 6.199133396148682, | |
| "learning_rate": 3.4861407249466955e-05, | |
| "loss": 0.018, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.5245202558635396, | |
| "grad_norm": 4.0590949058532715, | |
| "learning_rate": 3.47547974413646e-05, | |
| "loss": 0.0363, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.535181236673774, | |
| "grad_norm": 0.11506767570972443, | |
| "learning_rate": 3.4648187633262264e-05, | |
| "loss": 0.0181, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.5458422174840085, | |
| "grad_norm": 2.7985172271728516, | |
| "learning_rate": 3.454157782515991e-05, | |
| "loss": 0.0154, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.556503198294243, | |
| "grad_norm": 4.3073835372924805, | |
| "learning_rate": 3.4434968017057574e-05, | |
| "loss": 0.0654, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5671641791044775, | |
| "grad_norm": 6.925653457641602, | |
| "learning_rate": 3.432835820895522e-05, | |
| "loss": 0.0606, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.5778251599147122, | |
| "grad_norm": 9.806808471679688, | |
| "learning_rate": 3.422174840085288e-05, | |
| "loss": 0.0963, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5884861407249466, | |
| "grad_norm": 1.7052072286605835, | |
| "learning_rate": 3.411513859275053e-05, | |
| "loss": 0.0184, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.5991471215351813, | |
| "grad_norm": 4.117446422576904, | |
| "learning_rate": 3.400852878464819e-05, | |
| "loss": 0.0717, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.6098081023454158, | |
| "grad_norm": 0.32086166739463806, | |
| "learning_rate": 3.390191897654584e-05, | |
| "loss": 0.0337, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.6204690831556503, | |
| "grad_norm": 13.948720932006836, | |
| "learning_rate": 3.37953091684435e-05, | |
| "loss": 0.1307, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.6311300639658848, | |
| "grad_norm": 9.503339767456055, | |
| "learning_rate": 3.368869936034115e-05, | |
| "loss": 0.0535, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.6417910447761193, | |
| "grad_norm": 3.2496016025543213, | |
| "learning_rate": 3.358208955223881e-05, | |
| "loss": 0.065, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.652452025586354, | |
| "grad_norm": 11.788161277770996, | |
| "learning_rate": 3.347547974413646e-05, | |
| "loss": 0.0444, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.6631130063965884, | |
| "grad_norm": 4.969666481018066, | |
| "learning_rate": 3.336886993603412e-05, | |
| "loss": 0.0108, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.6737739872068231, | |
| "grad_norm": 9.110553741455078, | |
| "learning_rate": 3.326226012793177e-05, | |
| "loss": 0.0946, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.6844349680170576, | |
| "grad_norm": 10.327310562133789, | |
| "learning_rate": 3.3155650319829424e-05, | |
| "loss": 0.1041, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.695095948827292, | |
| "grad_norm": 5.026795387268066, | |
| "learning_rate": 3.304904051172708e-05, | |
| "loss": 0.0967, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.7057569296375266, | |
| "grad_norm": 4.460590839385986, | |
| "learning_rate": 3.294243070362473e-05, | |
| "loss": 0.0294, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.716417910447761, | |
| "grad_norm": 2.505089044570923, | |
| "learning_rate": 3.283582089552239e-05, | |
| "loss": 0.1051, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.7270788912579957, | |
| "grad_norm": 0.11050331592559814, | |
| "learning_rate": 3.272921108742004e-05, | |
| "loss": 0.0313, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.7377398720682304, | |
| "grad_norm": 8.98729133605957, | |
| "learning_rate": 3.26226012793177e-05, | |
| "loss": 0.0845, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.748400852878465, | |
| "grad_norm": 7.458622455596924, | |
| "learning_rate": 3.251599147121535e-05, | |
| "loss": 0.0899, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7590618336886994, | |
| "grad_norm": 8.965328216552734, | |
| "learning_rate": 3.240938166311301e-05, | |
| "loss": 0.0565, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.7697228144989339, | |
| "grad_norm": 0.050269417464733124, | |
| "learning_rate": 3.230277185501066e-05, | |
| "loss": 0.0072, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.7803837953091683, | |
| "grad_norm": 4.382719993591309, | |
| "learning_rate": 3.2196162046908317e-05, | |
| "loss": 0.0413, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.7910447761194028, | |
| "grad_norm": 2.915208578109741, | |
| "learning_rate": 3.208955223880597e-05, | |
| "loss": 0.054, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.8017057569296375, | |
| "grad_norm": 5.750125885009766, | |
| "learning_rate": 3.1982942430703626e-05, | |
| "loss": 0.0584, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.8123667377398722, | |
| "grad_norm": 0.10096129775047302, | |
| "learning_rate": 3.187633262260128e-05, | |
| "loss": 0.121, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.8230277185501067, | |
| "grad_norm": 1.7602276802062988, | |
| "learning_rate": 3.1769722814498935e-05, | |
| "loss": 0.036, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.8336886993603412, | |
| "grad_norm": 10.816998481750488, | |
| "learning_rate": 3.166311300639659e-05, | |
| "loss": 0.0549, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.8443496801705757, | |
| "grad_norm": 0.7319279909133911, | |
| "learning_rate": 3.1556503198294245e-05, | |
| "loss": 0.0199, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.8550106609808101, | |
| "grad_norm": 0.08583816140890121, | |
| "learning_rate": 3.14498933901919e-05, | |
| "loss": 0.0681, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.8656716417910446, | |
| "grad_norm": 0.3255792260169983, | |
| "learning_rate": 3.1343283582089554e-05, | |
| "loss": 0.0263, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.8763326226012793, | |
| "grad_norm": 6.624326229095459, | |
| "learning_rate": 3.123667377398721e-05, | |
| "loss": 0.1362, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.886993603411514, | |
| "grad_norm": 3.0343503952026367, | |
| "learning_rate": 3.1130063965884864e-05, | |
| "loss": 0.0771, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.8976545842217485, | |
| "grad_norm": 0.38771405816078186, | |
| "learning_rate": 3.102345415778252e-05, | |
| "loss": 0.0586, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.908315565031983, | |
| "grad_norm": 0.7634451985359192, | |
| "learning_rate": 3.0916844349680173e-05, | |
| "loss": 0.1265, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.9189765458422174, | |
| "grad_norm": 5.297740459442139, | |
| "learning_rate": 3.081023454157783e-05, | |
| "loss": 0.1252, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.929637526652452, | |
| "grad_norm": 6.677194595336914, | |
| "learning_rate": 3.070362473347548e-05, | |
| "loss": 0.0275, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.9402985074626866, | |
| "grad_norm": 0.14601445198059082, | |
| "learning_rate": 3.059701492537314e-05, | |
| "loss": 0.0076, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.950959488272921, | |
| "grad_norm": 0.09449265152215958, | |
| "learning_rate": 3.0490405117270792e-05, | |
| "loss": 0.1322, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.9616204690831558, | |
| "grad_norm": 4.836129188537598, | |
| "learning_rate": 3.0383795309168444e-05, | |
| "loss": 0.0938, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.9722814498933903, | |
| "grad_norm": 3.539830207824707, | |
| "learning_rate": 3.0277185501066102e-05, | |
| "loss": 0.008, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.9829424307036247, | |
| "grad_norm": 1.1404074430465698, | |
| "learning_rate": 3.0170575692963753e-05, | |
| "loss": 0.0499, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.9936034115138592, | |
| "grad_norm": 0.09818711131811142, | |
| "learning_rate": 3.006396588486141e-05, | |
| "loss": 0.0062, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9370666666666667, | |
| "eval_loss": 0.24299149215221405, | |
| "eval_runtime": 19.9475, | |
| "eval_samples_per_second": 187.993, | |
| "eval_steps_per_second": 5.916, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 2.0042643923240937, | |
| "grad_norm": 2.3476803302764893, | |
| "learning_rate": 2.9957356076759063e-05, | |
| "loss": 0.0083, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.014925373134328, | |
| "grad_norm": 6.180583953857422, | |
| "learning_rate": 2.9850746268656714e-05, | |
| "loss": 0.024, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.025586353944563, | |
| "grad_norm": 0.03756539151072502, | |
| "learning_rate": 2.9744136460554372e-05, | |
| "loss": 0.0227, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.0362473347547976, | |
| "grad_norm": 0.04238955304026604, | |
| "learning_rate": 2.9637526652452023e-05, | |
| "loss": 0.01, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.046908315565032, | |
| "grad_norm": 0.07046917825937271, | |
| "learning_rate": 2.953091684434968e-05, | |
| "loss": 0.004, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.0575692963752665, | |
| "grad_norm": 0.0585479661822319, | |
| "learning_rate": 2.9424307036247333e-05, | |
| "loss": 0.0414, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.068230277185501, | |
| "grad_norm": 10.458964347839355, | |
| "learning_rate": 2.931769722814499e-05, | |
| "loss": 0.0226, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.0788912579957355, | |
| "grad_norm": 13.315876960754395, | |
| "learning_rate": 2.9211087420042642e-05, | |
| "loss": 0.0136, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.08955223880597, | |
| "grad_norm": 0.08008021116256714, | |
| "learning_rate": 2.91044776119403e-05, | |
| "loss": 0.0079, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.100213219616205, | |
| "grad_norm": 0.030625145882368088, | |
| "learning_rate": 2.8997867803837952e-05, | |
| "loss": 0.0648, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.1108742004264394, | |
| "grad_norm": 0.027961866930127144, | |
| "learning_rate": 2.889125799573561e-05, | |
| "loss": 0.0054, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.121535181236674, | |
| "grad_norm": 1.2943542003631592, | |
| "learning_rate": 2.878464818763326e-05, | |
| "loss": 0.0069, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.1321961620469083, | |
| "grad_norm": 0.10296867787837982, | |
| "learning_rate": 2.867803837953092e-05, | |
| "loss": 0.0051, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.026940099895000458, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.0045, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.1535181236673773, | |
| "grad_norm": 0.02851429209113121, | |
| "learning_rate": 2.846481876332623e-05, | |
| "loss": 0.0187, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.1641791044776117, | |
| "grad_norm": 0.2322418987751007, | |
| "learning_rate": 2.835820895522388e-05, | |
| "loss": 0.0069, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.1748400852878467, | |
| "grad_norm": 0.03120555356144905, | |
| "learning_rate": 2.825159914712154e-05, | |
| "loss": 0.0142, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.185501066098081, | |
| "grad_norm": 0.027482597157359123, | |
| "learning_rate": 2.814498933901919e-05, | |
| "loss": 0.0029, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.1961620469083156, | |
| "grad_norm": 0.024367960169911385, | |
| "learning_rate": 2.8038379530916848e-05, | |
| "loss": 0.0193, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.20682302771855, | |
| "grad_norm": 0.027932733297348022, | |
| "learning_rate": 2.79317697228145e-05, | |
| "loss": 0.0436, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.2174840085287846, | |
| "grad_norm": 0.05150827392935753, | |
| "learning_rate": 2.7825159914712157e-05, | |
| "loss": 0.0062, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.228144989339019, | |
| "grad_norm": 0.0276811420917511, | |
| "learning_rate": 2.771855010660981e-05, | |
| "loss": 0.0058, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.2388059701492535, | |
| "grad_norm": 0.03752530366182327, | |
| "learning_rate": 2.7611940298507467e-05, | |
| "loss": 0.0676, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.2494669509594885, | |
| "grad_norm": 0.16837607324123383, | |
| "learning_rate": 2.7505330490405118e-05, | |
| "loss": 0.0426, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.260127931769723, | |
| "grad_norm": 0.024154026061296463, | |
| "learning_rate": 2.7398720682302776e-05, | |
| "loss": 0.0176, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.2707889125799574, | |
| "grad_norm": 2.2850098609924316, | |
| "learning_rate": 2.7292110874200428e-05, | |
| "loss": 0.0594, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.281449893390192, | |
| "grad_norm": 0.0278911255300045, | |
| "learning_rate": 2.7185501066098086e-05, | |
| "loss": 0.0143, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.2921108742004264, | |
| "grad_norm": 11.531503677368164, | |
| "learning_rate": 2.7078891257995737e-05, | |
| "loss": 0.014, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.302771855010661, | |
| "grad_norm": 0.02558918669819832, | |
| "learning_rate": 2.6972281449893395e-05, | |
| "loss": 0.0036, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.3134328358208958, | |
| "grad_norm": 1.1388365030288696, | |
| "learning_rate": 2.6865671641791047e-05, | |
| "loss": 0.0377, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.3240938166311302, | |
| "grad_norm": 1.1085879802703857, | |
| "learning_rate": 2.6759061833688705e-05, | |
| "loss": 0.0062, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.3347547974413647, | |
| "grad_norm": 13.190254211425781, | |
| "learning_rate": 2.6652452025586356e-05, | |
| "loss": 0.0083, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.345415778251599, | |
| "grad_norm": 0.05368981882929802, | |
| "learning_rate": 2.6545842217484007e-05, | |
| "loss": 0.0068, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.3560767590618337, | |
| "grad_norm": 0.6162731051445007, | |
| "learning_rate": 2.6439232409381666e-05, | |
| "loss": 0.0498, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.366737739872068, | |
| "grad_norm": 0.02255820296704769, | |
| "learning_rate": 2.6332622601279317e-05, | |
| "loss": 0.0832, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.3773987206823026, | |
| "grad_norm": 0.021348414942622185, | |
| "learning_rate": 2.6226012793176975e-05, | |
| "loss": 0.029, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.388059701492537, | |
| "grad_norm": 0.4747634530067444, | |
| "learning_rate": 2.6119402985074626e-05, | |
| "loss": 0.0105, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.398720682302772, | |
| "grad_norm": 15.23071575164795, | |
| "learning_rate": 2.6012793176972285e-05, | |
| "loss": 0.0206, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.4093816631130065, | |
| "grad_norm": 0.027936460450291634, | |
| "learning_rate": 2.5906183368869936e-05, | |
| "loss": 0.0044, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.420042643923241, | |
| "grad_norm": 0.02519192174077034, | |
| "learning_rate": 2.5799573560767594e-05, | |
| "loss": 0.0203, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.4307036247334755, | |
| "grad_norm": 3.956721067428589, | |
| "learning_rate": 2.5692963752665245e-05, | |
| "loss": 0.0488, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.44136460554371, | |
| "grad_norm": 0.06295081228017807, | |
| "learning_rate": 2.5586353944562904e-05, | |
| "loss": 0.0346, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.4520255863539444, | |
| "grad_norm": 8.38576602935791, | |
| "learning_rate": 2.5479744136460555e-05, | |
| "loss": 0.047, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.4626865671641793, | |
| "grad_norm": 0.025208963081240654, | |
| "learning_rate": 2.537313432835821e-05, | |
| "loss": 0.0317, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.473347547974414, | |
| "grad_norm": 0.049614135175943375, | |
| "learning_rate": 2.5266524520255864e-05, | |
| "loss": 0.0387, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.4840085287846483, | |
| "grad_norm": 3.707105875015259, | |
| "learning_rate": 2.515991471215352e-05, | |
| "loss": 0.0503, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.4946695095948828, | |
| "grad_norm": 3.57007098197937, | |
| "learning_rate": 2.5053304904051174e-05, | |
| "loss": 0.0239, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.5053304904051172, | |
| "grad_norm": 0.021594833582639694, | |
| "learning_rate": 2.494669509594883e-05, | |
| "loss": 0.0424, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.5159914712153517, | |
| "grad_norm": 0.3944630026817322, | |
| "learning_rate": 2.4840085287846483e-05, | |
| "loss": 0.0042, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.526652452025586, | |
| "grad_norm": 0.14251954853534698, | |
| "learning_rate": 2.4733475479744138e-05, | |
| "loss": 0.0196, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.5373134328358207, | |
| "grad_norm": 0.022276824340224266, | |
| "learning_rate": 2.4626865671641793e-05, | |
| "loss": 0.0037, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.5479744136460556, | |
| "grad_norm": 15.048105239868164, | |
| "learning_rate": 2.4520255863539444e-05, | |
| "loss": 0.0177, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.55863539445629, | |
| "grad_norm": 0.031552329659461975, | |
| "learning_rate": 2.44136460554371e-05, | |
| "loss": 0.0039, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.5692963752665245, | |
| "grad_norm": 0.15429170429706573, | |
| "learning_rate": 2.4307036247334754e-05, | |
| "loss": 0.0063, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.579957356076759, | |
| "grad_norm": 0.030070830136537552, | |
| "learning_rate": 2.420042643923241e-05, | |
| "loss": 0.0407, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.5906183368869935, | |
| "grad_norm": 0.019874196499586105, | |
| "learning_rate": 2.4093816631130063e-05, | |
| "loss": 0.0152, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.6012793176972284, | |
| "grad_norm": 0.03877472132444382, | |
| "learning_rate": 2.3987206823027718e-05, | |
| "loss": 0.0466, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.611940298507463, | |
| "grad_norm": 12.384525299072266, | |
| "learning_rate": 2.3880597014925373e-05, | |
| "loss": 0.0507, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.6226012793176974, | |
| "grad_norm": 0.02228281833231449, | |
| "learning_rate": 2.3773987206823027e-05, | |
| "loss": 0.0028, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.633262260127932, | |
| "grad_norm": 0.02572835609316826, | |
| "learning_rate": 2.3667377398720682e-05, | |
| "loss": 0.0058, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.6439232409381663, | |
| "grad_norm": 0.12954434752464294, | |
| "learning_rate": 2.3560767590618337e-05, | |
| "loss": 0.0064, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.654584221748401, | |
| "grad_norm": 17.412492752075195, | |
| "learning_rate": 2.345415778251599e-05, | |
| "loss": 0.0247, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.6652452025586353, | |
| "grad_norm": 9.312185287475586, | |
| "learning_rate": 2.3347547974413646e-05, | |
| "loss": 0.0114, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.6759061833688698, | |
| "grad_norm": 0.2050735056400299, | |
| "learning_rate": 2.32409381663113e-05, | |
| "loss": 0.0027, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.6865671641791042, | |
| "grad_norm": 0.036513011902570724, | |
| "learning_rate": 2.3134328358208956e-05, | |
| "loss": 0.0115, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.697228144989339, | |
| "grad_norm": 0.11639884114265442, | |
| "learning_rate": 2.302771855010661e-05, | |
| "loss": 0.0388, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.7078891257995736, | |
| "grad_norm": 0.021872026845812798, | |
| "learning_rate": 2.2921108742004265e-05, | |
| "loss": 0.0025, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.718550106609808, | |
| "grad_norm": 4.843020439147949, | |
| "learning_rate": 2.281449893390192e-05, | |
| "loss": 0.0065, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.7292110874200426, | |
| "grad_norm": 0.022604284808039665, | |
| "learning_rate": 2.2707889125799575e-05, | |
| "loss": 0.042, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.739872068230277, | |
| "grad_norm": 0.02514186128973961, | |
| "learning_rate": 2.260127931769723e-05, | |
| "loss": 0.036, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.750533049040512, | |
| "grad_norm": 0.0391307957470417, | |
| "learning_rate": 2.2494669509594884e-05, | |
| "loss": 0.064, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.7611940298507465, | |
| "grad_norm": 12.592732429504395, | |
| "learning_rate": 2.238805970149254e-05, | |
| "loss": 0.0237, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.771855010660981, | |
| "grad_norm": 6.797610759735107, | |
| "learning_rate": 2.2281449893390194e-05, | |
| "loss": 0.0281, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.7825159914712154, | |
| "grad_norm": 0.6649022102355957, | |
| "learning_rate": 2.217484008528785e-05, | |
| "loss": 0.068, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.79317697228145, | |
| "grad_norm": 2.1860225200653076, | |
| "learning_rate": 2.2068230277185503e-05, | |
| "loss": 0.091, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.8038379530916844, | |
| "grad_norm": 0.17125117778778076, | |
| "learning_rate": 2.1961620469083158e-05, | |
| "loss": 0.0174, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.814498933901919, | |
| "grad_norm": 0.02357735112309456, | |
| "learning_rate": 2.1855010660980813e-05, | |
| "loss": 0.0026, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.8251599147121533, | |
| "grad_norm": 0.029319651424884796, | |
| "learning_rate": 2.1748400852878467e-05, | |
| "loss": 0.0608, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.835820895522388, | |
| "grad_norm": 7.288904666900635, | |
| "learning_rate": 2.164179104477612e-05, | |
| "loss": 0.0408, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.8464818763326227, | |
| "grad_norm": 0.0742633119225502, | |
| "learning_rate": 2.1535181236673773e-05, | |
| "loss": 0.0225, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.07969071716070175, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.0534, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.8678038379530917, | |
| "grad_norm": 0.03505810350179672, | |
| "learning_rate": 2.1321961620469083e-05, | |
| "loss": 0.0159, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.878464818763326, | |
| "grad_norm": 5.3611555099487305, | |
| "learning_rate": 2.1215351812366738e-05, | |
| "loss": 0.0403, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.8891257995735606, | |
| "grad_norm": 0.019943783059716225, | |
| "learning_rate": 2.1108742004264392e-05, | |
| "loss": 0.0534, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.8997867803837956, | |
| "grad_norm": 3.984166383743286, | |
| "learning_rate": 2.1002132196162047e-05, | |
| "loss": 0.0052, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.91044776119403, | |
| "grad_norm": 0.040945641696453094, | |
| "learning_rate": 2.0895522388059702e-05, | |
| "loss": 0.0027, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.9211087420042645, | |
| "grad_norm": 12.568500518798828, | |
| "learning_rate": 2.0788912579957357e-05, | |
| "loss": 0.0557, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.931769722814499, | |
| "grad_norm": 0.02530559152364731, | |
| "learning_rate": 2.068230277185501e-05, | |
| "loss": 0.0025, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.9424307036247335, | |
| "grad_norm": 0.0479360967874527, | |
| "learning_rate": 2.0575692963752666e-05, | |
| "loss": 0.0476, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.953091684434968, | |
| "grad_norm": 0.056971777230501175, | |
| "learning_rate": 2.046908315565032e-05, | |
| "loss": 0.03, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.9637526652452024, | |
| "grad_norm": 0.02350008301436901, | |
| "learning_rate": 2.0362473347547976e-05, | |
| "loss": 0.044, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.974413646055437, | |
| "grad_norm": 12.957428932189941, | |
| "learning_rate": 2.025586353944563e-05, | |
| "loss": 0.0218, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "grad_norm": 0.04227874428033829, | |
| "learning_rate": 2.0149253731343285e-05, | |
| "loss": 0.0325, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.9957356076759063, | |
| "grad_norm": 1.0541740655899048, | |
| "learning_rate": 2.004264392324094e-05, | |
| "loss": 0.0043, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.948, | |
| "eval_loss": 0.21167731285095215, | |
| "eval_runtime": 19.9973, | |
| "eval_samples_per_second": 187.525, | |
| "eval_steps_per_second": 5.901, | |
| "step": 1407 | |
| }, | |
| { | |
| "epoch": 3.0063965884861408, | |
| "grad_norm": 20.089067459106445, | |
| "learning_rate": 1.9936034115138594e-05, | |
| "loss": 0.0241, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.0170575692963753, | |
| "grad_norm": 0.01854545995593071, | |
| "learning_rate": 1.982942430703625e-05, | |
| "loss": 0.0022, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 3.0277185501066097, | |
| "grad_norm": 0.02019428461790085, | |
| "learning_rate": 1.9722814498933904e-05, | |
| "loss": 0.0024, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.038379530916844, | |
| "grad_norm": 0.02650708518922329, | |
| "learning_rate": 1.961620469083156e-05, | |
| "loss": 0.0037, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 3.0490405117270787, | |
| "grad_norm": 0.025915976613759995, | |
| "learning_rate": 1.9509594882729213e-05, | |
| "loss": 0.0027, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.0597014925373136, | |
| "grad_norm": 6.768576622009277, | |
| "learning_rate": 1.9402985074626868e-05, | |
| "loss": 0.0373, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 3.070362473347548, | |
| "grad_norm": 0.023793905973434448, | |
| "learning_rate": 1.9296375266524523e-05, | |
| "loss": 0.0182, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.0810234541577826, | |
| "grad_norm": 0.3890201151371002, | |
| "learning_rate": 1.9189765458422178e-05, | |
| "loss": 0.0026, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 3.091684434968017, | |
| "grad_norm": 0.02452036924660206, | |
| "learning_rate": 1.9083155650319832e-05, | |
| "loss": 0.0023, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.1023454157782515, | |
| "grad_norm": 0.02351544424891472, | |
| "learning_rate": 1.8976545842217487e-05, | |
| "loss": 0.0022, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 3.113006396588486, | |
| "grad_norm": 0.015399349853396416, | |
| "learning_rate": 1.8869936034115142e-05, | |
| "loss": 0.0054, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.1236673773987205, | |
| "grad_norm": 0.025128506124019623, | |
| "learning_rate": 1.8763326226012797e-05, | |
| "loss": 0.0021, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 3.1343283582089554, | |
| "grad_norm": 0.018281979486346245, | |
| "learning_rate": 1.865671641791045e-05, | |
| "loss": 0.0163, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.14498933901919, | |
| "grad_norm": 0.3955625295639038, | |
| "learning_rate": 1.8550106609808106e-05, | |
| "loss": 0.0327, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 3.1556503198294243, | |
| "grad_norm": 0.016859382390975952, | |
| "learning_rate": 1.8443496801705757e-05, | |
| "loss": 0.002, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.166311300639659, | |
| "grad_norm": 0.056564632803201675, | |
| "learning_rate": 1.8336886993603412e-05, | |
| "loss": 0.0299, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 3.1769722814498933, | |
| "grad_norm": 0.016592353582382202, | |
| "learning_rate": 1.8230277185501067e-05, | |
| "loss": 0.0019, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.1876332622601278, | |
| "grad_norm": 0.016828283667564392, | |
| "learning_rate": 1.812366737739872e-05, | |
| "loss": 0.0019, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 3.1982942430703627, | |
| "grad_norm": 0.02142864279448986, | |
| "learning_rate": 1.8017057569296376e-05, | |
| "loss": 0.002, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.208955223880597, | |
| "grad_norm": 0.6647986173629761, | |
| "learning_rate": 1.791044776119403e-05, | |
| "loss": 0.0027, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 3.2196162046908317, | |
| "grad_norm": 0.023662006482481956, | |
| "learning_rate": 1.7803837953091686e-05, | |
| "loss": 0.0023, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.230277185501066, | |
| "grad_norm": 0.01707913540303707, | |
| "learning_rate": 1.769722814498934e-05, | |
| "loss": 0.0019, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 3.2409381663113006, | |
| "grad_norm": 0.021021848544478416, | |
| "learning_rate": 1.7590618336886995e-05, | |
| "loss": 0.0019, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.251599147121535, | |
| "grad_norm": 0.015328795649111271, | |
| "learning_rate": 1.7484008528784647e-05, | |
| "loss": 0.0021, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 3.2622601279317696, | |
| "grad_norm": 0.3916546106338501, | |
| "learning_rate": 1.73773987206823e-05, | |
| "loss": 0.0397, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.272921108742004, | |
| "grad_norm": 0.022156652063131332, | |
| "learning_rate": 1.7270788912579956e-05, | |
| "loss": 0.0199, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 3.283582089552239, | |
| "grad_norm": 0.01748698204755783, | |
| "learning_rate": 1.716417910447761e-05, | |
| "loss": 0.0019, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.2942430703624734, | |
| "grad_norm": 0.020094819366931915, | |
| "learning_rate": 1.7057569296375266e-05, | |
| "loss": 0.0019, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 3.304904051172708, | |
| "grad_norm": 0.024203898385167122, | |
| "learning_rate": 1.695095948827292e-05, | |
| "loss": 0.007, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.3155650319829424, | |
| "grad_norm": 0.015299428254365921, | |
| "learning_rate": 1.6844349680170575e-05, | |
| "loss": 0.002, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 3.326226012793177, | |
| "grad_norm": 0.012815508991479874, | |
| "learning_rate": 1.673773987206823e-05, | |
| "loss": 0.0019, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.3368869936034113, | |
| "grad_norm": 0.01478462666273117, | |
| "learning_rate": 1.6631130063965885e-05, | |
| "loss": 0.0017, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 3.3475479744136463, | |
| "grad_norm": 0.02664569951593876, | |
| "learning_rate": 1.652452025586354e-05, | |
| "loss": 0.0389, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.3582089552238807, | |
| "grad_norm": 0.023107649758458138, | |
| "learning_rate": 1.6417910447761194e-05, | |
| "loss": 0.0025, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 3.368869936034115, | |
| "grad_norm": 0.10797708481550217, | |
| "learning_rate": 1.631130063965885e-05, | |
| "loss": 0.002, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.3795309168443497, | |
| "grad_norm": 0.013530361466109753, | |
| "learning_rate": 1.6204690831556504e-05, | |
| "loss": 0.0017, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 3.390191897654584, | |
| "grad_norm": 0.015169564634561539, | |
| "learning_rate": 1.6098081023454158e-05, | |
| "loss": 0.0019, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 3.4008528784648187, | |
| "grad_norm": 0.04296619072556496, | |
| "learning_rate": 1.5991471215351813e-05, | |
| "loss": 0.0018, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 3.411513859275053, | |
| "grad_norm": 0.01815551333129406, | |
| "learning_rate": 1.5884861407249468e-05, | |
| "loss": 0.0044, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.4221748400852876, | |
| "grad_norm": 0.03518333286046982, | |
| "learning_rate": 1.5778251599147122e-05, | |
| "loss": 0.0047, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 3.4328358208955225, | |
| "grad_norm": 0.015669243410229683, | |
| "learning_rate": 1.5671641791044777e-05, | |
| "loss": 0.0022, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 3.443496801705757, | |
| "grad_norm": 0.020631814375519753, | |
| "learning_rate": 1.5565031982942432e-05, | |
| "loss": 0.0017, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 3.4541577825159915, | |
| "grad_norm": 0.016420889645814896, | |
| "learning_rate": 1.5458422174840087e-05, | |
| "loss": 0.0017, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.464818763326226, | |
| "grad_norm": 0.015450740233063698, | |
| "learning_rate": 1.535181236673774e-05, | |
| "loss": 0.0017, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 3.4754797441364604, | |
| "grad_norm": 0.19196806848049164, | |
| "learning_rate": 1.5245202558635396e-05, | |
| "loss": 0.0023, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 3.486140724946695, | |
| "grad_norm": 0.01444052904844284, | |
| "learning_rate": 1.5138592750533051e-05, | |
| "loss": 0.0017, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 3.49680170575693, | |
| "grad_norm": 0.4485403597354889, | |
| "learning_rate": 1.5031982942430706e-05, | |
| "loss": 0.0567, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.5074626865671643, | |
| "grad_norm": 0.014176204800605774, | |
| "learning_rate": 1.4925373134328357e-05, | |
| "loss": 0.0017, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 3.518123667377399, | |
| "grad_norm": 0.014325299300253391, | |
| "learning_rate": 1.4818763326226012e-05, | |
| "loss": 0.0016, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.5287846481876333, | |
| "grad_norm": 0.014549434185028076, | |
| "learning_rate": 1.4712153518123666e-05, | |
| "loss": 0.0016, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 3.5394456289978677, | |
| "grad_norm": 0.01434430293738842, | |
| "learning_rate": 1.4605543710021321e-05, | |
| "loss": 0.0016, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 3.550106609808102, | |
| "grad_norm": 0.012784713879227638, | |
| "learning_rate": 1.4498933901918976e-05, | |
| "loss": 0.0043, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 3.5607675906183367, | |
| "grad_norm": 0.01848587580025196, | |
| "learning_rate": 1.439232409381663e-05, | |
| "loss": 0.0016, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.02839992754161358, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.0017, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 3.582089552238806, | |
| "grad_norm": 0.013904962688684464, | |
| "learning_rate": 1.417910447761194e-05, | |
| "loss": 0.0016, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 3.5927505330490406, | |
| "grad_norm": 0.017972519621253014, | |
| "learning_rate": 1.4072494669509595e-05, | |
| "loss": 0.0016, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 3.603411513859275, | |
| "grad_norm": 4.149586200714111, | |
| "learning_rate": 1.396588486140725e-05, | |
| "loss": 0.0406, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 3.6140724946695095, | |
| "grad_norm": 0.012793551199138165, | |
| "learning_rate": 1.3859275053304904e-05, | |
| "loss": 0.0021, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 3.624733475479744, | |
| "grad_norm": 0.015538396313786507, | |
| "learning_rate": 1.3752665245202559e-05, | |
| "loss": 0.0016, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.635394456289979, | |
| "grad_norm": 0.014050007797777653, | |
| "learning_rate": 1.3646055437100214e-05, | |
| "loss": 0.0429, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 3.6460554371002134, | |
| "grad_norm": 0.012542256154119968, | |
| "learning_rate": 1.3539445628997869e-05, | |
| "loss": 0.0016, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 3.656716417910448, | |
| "grad_norm": 0.012764188461005688, | |
| "learning_rate": 1.3432835820895523e-05, | |
| "loss": 0.0018, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 3.6673773987206824, | |
| "grad_norm": 0.013499289751052856, | |
| "learning_rate": 1.3326226012793178e-05, | |
| "loss": 0.0026, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.678038379530917, | |
| "grad_norm": 0.014842645265161991, | |
| "learning_rate": 1.3219616204690833e-05, | |
| "loss": 0.043, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 3.6886993603411513, | |
| "grad_norm": 0.02188076637685299, | |
| "learning_rate": 1.3113006396588488e-05, | |
| "loss": 0.0047, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.699360341151386, | |
| "grad_norm": 0.01476881094276905, | |
| "learning_rate": 1.3006396588486142e-05, | |
| "loss": 0.0019, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 3.7100213219616203, | |
| "grad_norm": 0.03763126581907272, | |
| "learning_rate": 1.2899786780383797e-05, | |
| "loss": 0.0016, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.7206823027718547, | |
| "grad_norm": 0.018351443111896515, | |
| "learning_rate": 1.2793176972281452e-05, | |
| "loss": 0.0016, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 3.7313432835820897, | |
| "grad_norm": 0.014618510380387306, | |
| "learning_rate": 1.2686567164179105e-05, | |
| "loss": 0.0017, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.742004264392324, | |
| "grad_norm": 0.012340176850557327, | |
| "learning_rate": 1.257995735607676e-05, | |
| "loss": 0.0017, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 3.7526652452025586, | |
| "grad_norm": 0.015355751849710941, | |
| "learning_rate": 1.2473347547974414e-05, | |
| "loss": 0.0015, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.763326226012793, | |
| "grad_norm": 0.01422917190939188, | |
| "learning_rate": 1.2366737739872069e-05, | |
| "loss": 0.0016, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 3.7739872068230276, | |
| "grad_norm": 18.438125610351562, | |
| "learning_rate": 1.2260127931769722e-05, | |
| "loss": 0.0297, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.7846481876332625, | |
| "grad_norm": 0.016914648935198784, | |
| "learning_rate": 1.2153518123667377e-05, | |
| "loss": 0.0015, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 3.795309168443497, | |
| "grad_norm": 0.013218486681580544, | |
| "learning_rate": 1.2046908315565032e-05, | |
| "loss": 0.0016, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.8059701492537314, | |
| "grad_norm": 0.018511895090341568, | |
| "learning_rate": 1.1940298507462686e-05, | |
| "loss": 0.0021, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 3.816631130063966, | |
| "grad_norm": 0.013273508287966251, | |
| "learning_rate": 1.1833688699360341e-05, | |
| "loss": 0.0015, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.8272921108742004, | |
| "grad_norm": 0.014320973306894302, | |
| "learning_rate": 1.1727078891257996e-05, | |
| "loss": 0.0019, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 3.837953091684435, | |
| "grad_norm": 0.0160856731235981, | |
| "learning_rate": 1.162046908315565e-05, | |
| "loss": 0.0017, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.8486140724946694, | |
| "grad_norm": 0.045672569423913956, | |
| "learning_rate": 1.1513859275053305e-05, | |
| "loss": 0.0017, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 3.859275053304904, | |
| "grad_norm": 0.011972334235906601, | |
| "learning_rate": 1.140724946695096e-05, | |
| "loss": 0.0015, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.8699360341151388, | |
| "grad_norm": 0.014581023715436459, | |
| "learning_rate": 1.1300639658848615e-05, | |
| "loss": 0.0177, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 3.8805970149253732, | |
| "grad_norm": 0.013506033457815647, | |
| "learning_rate": 1.119402985074627e-05, | |
| "loss": 0.0013, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.8912579957356077, | |
| "grad_norm": 0.3891560733318329, | |
| "learning_rate": 1.1087420042643924e-05, | |
| "loss": 0.0281, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 3.901918976545842, | |
| "grad_norm": 0.07957696169614792, | |
| "learning_rate": 1.0980810234541579e-05, | |
| "loss": 0.0401, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.9125799573560767, | |
| "grad_norm": 0.02362939342856407, | |
| "learning_rate": 1.0874200426439234e-05, | |
| "loss": 0.0015, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 3.923240938166311, | |
| "grad_norm": 0.02383952960371971, | |
| "learning_rate": 1.0767590618336887e-05, | |
| "loss": 0.0015, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.933901918976546, | |
| "grad_norm": 0.012055100873112679, | |
| "learning_rate": 1.0660980810234541e-05, | |
| "loss": 0.0014, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 3.9445628997867805, | |
| "grad_norm": 0.011509795673191547, | |
| "learning_rate": 1.0554371002132196e-05, | |
| "loss": 0.0013, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.955223880597015, | |
| "grad_norm": 0.012747013941407204, | |
| "learning_rate": 1.0447761194029851e-05, | |
| "loss": 0.0015, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 3.9658848614072495, | |
| "grad_norm": 0.013910112902522087, | |
| "learning_rate": 1.0341151385927506e-05, | |
| "loss": 0.0014, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.976545842217484, | |
| "grad_norm": 0.01466603484004736, | |
| "learning_rate": 1.023454157782516e-05, | |
| "loss": 0.0441, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 3.9872068230277184, | |
| "grad_norm": 0.015464485622942448, | |
| "learning_rate": 1.0127931769722815e-05, | |
| "loss": 0.0015, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.997867803837953, | |
| "grad_norm": 0.013403241522610188, | |
| "learning_rate": 1.002132196162047e-05, | |
| "loss": 0.0025, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9501333333333334, | |
| "eval_loss": 0.24279925227165222, | |
| "eval_runtime": 20.1516, | |
| "eval_samples_per_second": 186.09, | |
| "eval_steps_per_second": 5.856, | |
| "step": 1876 | |
| }, | |
| { | |
| "epoch": 4.008528784648187, | |
| "grad_norm": 0.013200121000409126, | |
| "learning_rate": 9.914712153518125e-06, | |
| "loss": 0.0023, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.019189765458422, | |
| "grad_norm": 0.0431358739733696, | |
| "learning_rate": 9.80810234541578e-06, | |
| "loss": 0.0015, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 4.029850746268656, | |
| "grad_norm": 0.013123239390552044, | |
| "learning_rate": 9.701492537313434e-06, | |
| "loss": 0.0013, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.040511727078891, | |
| "grad_norm": 0.012217454612255096, | |
| "learning_rate": 9.594882729211089e-06, | |
| "loss": 0.0013, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 4.051172707889126, | |
| "grad_norm": 0.01208677887916565, | |
| "learning_rate": 9.488272921108744e-06, | |
| "loss": 0.0015, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.061833688699361, | |
| "grad_norm": 0.012476786971092224, | |
| "learning_rate": 9.381663113006398e-06, | |
| "loss": 0.0014, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 4.072494669509595, | |
| "grad_norm": 0.012148105539381504, | |
| "learning_rate": 9.275053304904053e-06, | |
| "loss": 0.0014, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.08315565031983, | |
| "grad_norm": 0.011143679730594158, | |
| "learning_rate": 9.168443496801706e-06, | |
| "loss": 0.0013, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 4.093816631130064, | |
| "grad_norm": 0.013590450398623943, | |
| "learning_rate": 9.06183368869936e-06, | |
| "loss": 0.0015, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.104477611940299, | |
| "grad_norm": 0.01825527660548687, | |
| "learning_rate": 8.955223880597016e-06, | |
| "loss": 0.0016, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 4.115138592750533, | |
| "grad_norm": 0.010571052320301533, | |
| "learning_rate": 8.84861407249467e-06, | |
| "loss": 0.0013, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.1257995735607675, | |
| "grad_norm": 0.01558819692581892, | |
| "learning_rate": 8.742004264392323e-06, | |
| "loss": 0.0013, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 4.136460554371002, | |
| "grad_norm": 0.02532988414168358, | |
| "learning_rate": 8.635394456289978e-06, | |
| "loss": 0.0015, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.1471215351812365, | |
| "grad_norm": 0.009951004758477211, | |
| "learning_rate": 8.528784648187633e-06, | |
| "loss": 0.0014, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 4.157782515991471, | |
| "grad_norm": 0.011644040234386921, | |
| "learning_rate": 8.422174840085288e-06, | |
| "loss": 0.0014, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.1684434968017055, | |
| "grad_norm": 0.019300062209367752, | |
| "learning_rate": 8.315565031982942e-06, | |
| "loss": 0.0014, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 4.17910447761194, | |
| "grad_norm": 0.016809195280075073, | |
| "learning_rate": 8.208955223880597e-06, | |
| "loss": 0.0015, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.189765458422174, | |
| "grad_norm": 0.013822578825056553, | |
| "learning_rate": 8.102345415778252e-06, | |
| "loss": 0.0013, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 4.20042643923241, | |
| "grad_norm": 0.012465443462133408, | |
| "learning_rate": 7.995735607675907e-06, | |
| "loss": 0.0128, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 4.211087420042644, | |
| "grad_norm": 0.011877032928168774, | |
| "learning_rate": 7.889125799573561e-06, | |
| "loss": 0.0014, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 4.221748400852879, | |
| "grad_norm": 0.05958353728055954, | |
| "learning_rate": 7.782515991471216e-06, | |
| "loss": 0.0014, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.232409381663113, | |
| "grad_norm": 0.017867228016257286, | |
| "learning_rate": 7.67590618336887e-06, | |
| "loss": 0.0013, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 4.243070362473348, | |
| "grad_norm": 0.011448140256106853, | |
| "learning_rate": 7.5692963752665255e-06, | |
| "loss": 0.0013, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 4.253731343283582, | |
| "grad_norm": 0.02126247249543667, | |
| "learning_rate": 7.4626865671641785e-06, | |
| "loss": 0.0016, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 4.264392324093817, | |
| "grad_norm": 0.01147544290870428, | |
| "learning_rate": 7.356076759061833e-06, | |
| "loss": 0.0014, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.275053304904051, | |
| "grad_norm": 0.01126231998205185, | |
| "learning_rate": 7.249466950959488e-06, | |
| "loss": 0.0437, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.014390517957508564, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.0013, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 4.29637526652452, | |
| "grad_norm": 0.01252635195851326, | |
| "learning_rate": 7.0362473347547975e-06, | |
| "loss": 0.0456, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 4.3070362473347545, | |
| "grad_norm": 0.011789357289671898, | |
| "learning_rate": 6.929637526652452e-06, | |
| "loss": 0.0013, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 4.317697228144989, | |
| "grad_norm": 0.012756749987602234, | |
| "learning_rate": 6.823027718550107e-06, | |
| "loss": 0.0014, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 4.3283582089552235, | |
| "grad_norm": 0.023602314293384552, | |
| "learning_rate": 6.716417910447762e-06, | |
| "loss": 0.0014, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 4.339019189765459, | |
| "grad_norm": 0.011962332762777805, | |
| "learning_rate": 6.609808102345416e-06, | |
| "loss": 0.0361, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 4.349680170575693, | |
| "grad_norm": 0.010998536832630634, | |
| "learning_rate": 6.503198294243071e-06, | |
| "loss": 0.0013, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 4.360341151385928, | |
| "grad_norm": 0.012006332166492939, | |
| "learning_rate": 6.396588486140726e-06, | |
| "loss": 0.0014, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 4.371002132196162, | |
| "grad_norm": 0.14881190657615662, | |
| "learning_rate": 6.28997867803838e-06, | |
| "loss": 0.0014, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.381663113006397, | |
| "grad_norm": 0.01262710988521576, | |
| "learning_rate": 6.1833688699360345e-06, | |
| "loss": 0.0013, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 4.392324093816631, | |
| "grad_norm": 0.011607183143496513, | |
| "learning_rate": 6.076759061833688e-06, | |
| "loss": 0.0014, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 4.402985074626866, | |
| "grad_norm": 0.01482926495373249, | |
| "learning_rate": 5.970149253731343e-06, | |
| "loss": 0.0018, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 4.4136460554371, | |
| "grad_norm": 0.009752222336828709, | |
| "learning_rate": 5.863539445628998e-06, | |
| "loss": 0.0013, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 4.424307036247335, | |
| "grad_norm": 0.02456405945122242, | |
| "learning_rate": 5.756929637526653e-06, | |
| "loss": 0.0013, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 4.434968017057569, | |
| "grad_norm": 0.01045382209122181, | |
| "learning_rate": 5.650319829424307e-06, | |
| "loss": 0.0014, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 4.445628997867804, | |
| "grad_norm": 0.012066872790455818, | |
| "learning_rate": 5.543710021321962e-06, | |
| "loss": 0.0014, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 4.456289978678038, | |
| "grad_norm": 0.012539554387331009, | |
| "learning_rate": 5.437100213219617e-06, | |
| "loss": 0.0013, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 4.466950959488273, | |
| "grad_norm": 0.011873334646224976, | |
| "learning_rate": 5.330490405117271e-06, | |
| "loss": 0.0014, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 4.477611940298507, | |
| "grad_norm": 0.013598410412669182, | |
| "learning_rate": 5.2238805970149255e-06, | |
| "loss": 0.0013, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.4882729211087415, | |
| "grad_norm": 0.01245486456900835, | |
| "learning_rate": 5.11727078891258e-06, | |
| "loss": 0.0013, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 4.498933901918977, | |
| "grad_norm": 0.012206004001200199, | |
| "learning_rate": 5.010660980810235e-06, | |
| "loss": 0.0014, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 4.509594882729211, | |
| "grad_norm": 0.011170262470841408, | |
| "learning_rate": 4.90405117270789e-06, | |
| "loss": 0.0013, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 4.520255863539446, | |
| "grad_norm": 0.011509446427226067, | |
| "learning_rate": 4.797441364605544e-06, | |
| "loss": 0.0014, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 4.53091684434968, | |
| "grad_norm": 0.013342396356165409, | |
| "learning_rate": 4.690831556503199e-06, | |
| "loss": 0.0013, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 4.541577825159915, | |
| "grad_norm": 0.011811060830950737, | |
| "learning_rate": 4.584221748400853e-06, | |
| "loss": 0.0013, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 4.552238805970149, | |
| "grad_norm": 0.012742310762405396, | |
| "learning_rate": 4.477611940298508e-06, | |
| "loss": 0.0013, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 4.562899786780384, | |
| "grad_norm": 0.012248659506440163, | |
| "learning_rate": 4.371002132196162e-06, | |
| "loss": 0.0014, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 4.573560767590618, | |
| "grad_norm": 0.04519481211900711, | |
| "learning_rate": 4.264392324093816e-06, | |
| "loss": 0.0018, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 4.584221748400853, | |
| "grad_norm": 0.028292739763855934, | |
| "learning_rate": 4.157782515991471e-06, | |
| "loss": 0.0014, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 4.594882729211087, | |
| "grad_norm": 0.03468099236488342, | |
| "learning_rate": 4.051172707889126e-06, | |
| "loss": 0.0014, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 4.605543710021322, | |
| "grad_norm": 0.037209540605545044, | |
| "learning_rate": 3.944562899786781e-06, | |
| "loss": 0.0013, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 4.616204690831556, | |
| "grad_norm": 0.011522234417498112, | |
| "learning_rate": 3.837953091684435e-06, | |
| "loss": 0.0013, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 4.6268656716417915, | |
| "grad_norm": 0.011133146472275257, | |
| "learning_rate": 3.7313432835820893e-06, | |
| "loss": 0.0013, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 4.637526652452026, | |
| "grad_norm": 0.011684561148285866, | |
| "learning_rate": 3.624733475479744e-06, | |
| "loss": 0.0012, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 4.6481876332622605, | |
| "grad_norm": 0.011033386923372746, | |
| "learning_rate": 3.5181236673773987e-06, | |
| "loss": 0.0013, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 4.658848614072495, | |
| "grad_norm": 0.013793164864182472, | |
| "learning_rate": 3.4115138592750535e-06, | |
| "loss": 0.0013, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 4.669509594882729, | |
| "grad_norm": 0.012587225064635277, | |
| "learning_rate": 3.304904051172708e-06, | |
| "loss": 0.0012, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 4.680170575692964, | |
| "grad_norm": 0.03341620787978172, | |
| "learning_rate": 3.198294243070363e-06, | |
| "loss": 0.0014, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 4.690831556503198, | |
| "grad_norm": 0.01311581116169691, | |
| "learning_rate": 3.0916844349680173e-06, | |
| "loss": 0.0013, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.701492537313433, | |
| "grad_norm": 0.01578107662498951, | |
| "learning_rate": 2.9850746268656716e-06, | |
| "loss": 0.0013, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 4.712153518123667, | |
| "grad_norm": 0.012789585627615452, | |
| "learning_rate": 2.8784648187633263e-06, | |
| "loss": 0.0012, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 4.722814498933902, | |
| "grad_norm": 0.010789446532726288, | |
| "learning_rate": 2.771855010660981e-06, | |
| "loss": 0.0012, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 4.733475479744136, | |
| "grad_norm": 0.013990014791488647, | |
| "learning_rate": 2.6652452025586354e-06, | |
| "loss": 0.0013, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 4.744136460554371, | |
| "grad_norm": 0.013047412037849426, | |
| "learning_rate": 2.55863539445629e-06, | |
| "loss": 0.0014, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 4.754797441364605, | |
| "grad_norm": 0.011678214184939861, | |
| "learning_rate": 2.452025586353945e-06, | |
| "loss": 0.0013, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 4.76545842217484, | |
| "grad_norm": 0.01692238263785839, | |
| "learning_rate": 2.3454157782515996e-06, | |
| "loss": 0.0014, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 4.776119402985074, | |
| "grad_norm": 0.011887561529874802, | |
| "learning_rate": 2.238805970149254e-06, | |
| "loss": 0.0013, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 4.786780383795309, | |
| "grad_norm": 0.01334965880960226, | |
| "learning_rate": 2.132196162046908e-06, | |
| "loss": 0.0012, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 4.797441364605544, | |
| "grad_norm": 0.01241579744964838, | |
| "learning_rate": 2.025586353944563e-06, | |
| "loss": 0.0013, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 4.8081023454157785, | |
| "grad_norm": 0.013554994948208332, | |
| "learning_rate": 1.9189765458422177e-06, | |
| "loss": 0.0014, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 4.818763326226013, | |
| "grad_norm": 0.015750335529446602, | |
| "learning_rate": 1.812366737739872e-06, | |
| "loss": 0.0013, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 4.8294243070362475, | |
| "grad_norm": 0.01171982940286398, | |
| "learning_rate": 1.7057569296375267e-06, | |
| "loss": 0.0012, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 4.840085287846482, | |
| "grad_norm": 0.0744161307811737, | |
| "learning_rate": 1.5991471215351815e-06, | |
| "loss": 0.002, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 4.850746268656716, | |
| "grad_norm": 0.014584868215024471, | |
| "learning_rate": 1.4925373134328358e-06, | |
| "loss": 0.0012, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 4.861407249466951, | |
| "grad_norm": 0.013462378643453121, | |
| "learning_rate": 1.3859275053304905e-06, | |
| "loss": 0.0013, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 4.872068230277185, | |
| "grad_norm": 0.00995379127562046, | |
| "learning_rate": 1.279317697228145e-06, | |
| "loss": 0.0012, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 4.88272921108742, | |
| "grad_norm": 0.017240332439541817, | |
| "learning_rate": 1.1727078891257998e-06, | |
| "loss": 0.0014, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 4.893390191897654, | |
| "grad_norm": 0.0115845687687397, | |
| "learning_rate": 1.066098081023454e-06, | |
| "loss": 0.0013, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 4.904051172707889, | |
| "grad_norm": 0.022987473756074905, | |
| "learning_rate": 9.594882729211088e-07, | |
| "loss": 0.0014, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.914712153518123, | |
| "grad_norm": 0.011481226421892643, | |
| "learning_rate": 8.528784648187634e-07, | |
| "loss": 0.0017, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 4.925373134328359, | |
| "grad_norm": 0.010982918553054333, | |
| "learning_rate": 7.462686567164179e-07, | |
| "loss": 0.0012, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 4.936034115138593, | |
| "grad_norm": 0.010946305468678474, | |
| "learning_rate": 6.396588486140725e-07, | |
| "loss": 0.0012, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 4.946695095948828, | |
| "grad_norm": 0.01180151104927063, | |
| "learning_rate": 5.33049040511727e-07, | |
| "loss": 0.0013, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 4.957356076759062, | |
| "grad_norm": 0.05776381492614746, | |
| "learning_rate": 4.264392324093817e-07, | |
| "loss": 0.0013, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 4.968017057569297, | |
| "grad_norm": 0.01068054884672165, | |
| "learning_rate": 3.1982942430703626e-07, | |
| "loss": 0.0012, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 4.978678038379531, | |
| "grad_norm": 0.018672799691557884, | |
| "learning_rate": 2.1321961620469084e-07, | |
| "loss": 0.0013, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 4.9893390191897655, | |
| "grad_norm": 0.01208443846553564, | |
| "learning_rate": 1.0660980810234542e-07, | |
| "loss": 0.0012, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.02948063053190708, | |
| "learning_rate": 0.0, | |
| "loss": 0.0014, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9474666666666667, | |
| "eval_loss": 0.2692987024784088, | |
| "eval_runtime": 19.9174, | |
| "eval_samples_per_second": 188.277, | |
| "eval_steps_per_second": 5.924, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 2345, | |
| "total_flos": 5.8118992210944e+18, | |
| "train_loss": 0.05667482007509356, | |
| "train_runtime": 875.2593, | |
| "train_samples_per_second": 85.689, | |
| "train_steps_per_second": 2.679 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 2345, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.8118992210944e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |