| { | |
| "best_metric": 0.1182381734251976, | |
| "best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles6_seed1_q2_complexity\\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": 0.0714205801486969, | |
| "learning_rate": 4.989339019189766e-05, | |
| "loss": 0.0307, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021321961620469083, | |
| "grad_norm": 0.0004005177761428058, | |
| "learning_rate": 4.978678038379531e-05, | |
| "loss": 0.0371, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031982942430703626, | |
| "grad_norm": 8.92792797088623, | |
| "learning_rate": 4.9680170575692967e-05, | |
| "loss": 0.0957, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042643923240938165, | |
| "grad_norm": 15.518477439880371, | |
| "learning_rate": 4.957356076759062e-05, | |
| "loss": 0.1876, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053304904051172705, | |
| "grad_norm": 10.623318672180176, | |
| "learning_rate": 4.9466950959488276e-05, | |
| "loss": 0.1092, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06396588486140725, | |
| "grad_norm": 0.25674086809158325, | |
| "learning_rate": 4.936034115138593e-05, | |
| "loss": 0.0652, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07462686567164178, | |
| "grad_norm": 0.06139841303229332, | |
| "learning_rate": 4.9253731343283586e-05, | |
| "loss": 0.0885, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08528784648187633, | |
| "grad_norm": 0.05754043534398079, | |
| "learning_rate": 4.914712153518124e-05, | |
| "loss": 0.1568, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09594882729211088, | |
| "grad_norm": 3.1627166271209717, | |
| "learning_rate": 4.904051172707889e-05, | |
| "loss": 0.2158, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10660980810234541, | |
| "grad_norm": 0.12622235715389252, | |
| "learning_rate": 4.893390191897655e-05, | |
| "loss": 0.0433, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11727078891257996, | |
| "grad_norm": 6.332047462463379, | |
| "learning_rate": 4.88272921108742e-05, | |
| "loss": 0.07, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1279317697228145, | |
| "grad_norm": 6.619658946990967, | |
| "learning_rate": 4.872068230277186e-05, | |
| "loss": 0.1364, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13859275053304904, | |
| "grad_norm": 6.177881717681885, | |
| "learning_rate": 4.861407249466951e-05, | |
| "loss": 0.0273, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14925373134328357, | |
| "grad_norm": 4.713071346282959, | |
| "learning_rate": 4.850746268656717e-05, | |
| "loss": 0.0673, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15991471215351813, | |
| "grad_norm": 3.1095566749572754, | |
| "learning_rate": 4.840085287846482e-05, | |
| "loss": 0.0421, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17057569296375266, | |
| "grad_norm": 0.04303058609366417, | |
| "learning_rate": 4.829424307036248e-05, | |
| "loss": 0.0369, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1812366737739872, | |
| "grad_norm": 3.368828535079956, | |
| "learning_rate": 4.8187633262260126e-05, | |
| "loss": 0.0778, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19189765458422176, | |
| "grad_norm": 5.052602767944336, | |
| "learning_rate": 4.808102345415779e-05, | |
| "loss": 0.0972, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2025586353944563, | |
| "grad_norm": 4.186357021331787, | |
| "learning_rate": 4.7974413646055436e-05, | |
| "loss": 0.1447, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21321961620469082, | |
| "grad_norm": 0.6351255774497986, | |
| "learning_rate": 4.78678038379531e-05, | |
| "loss": 0.0799, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22388059701492538, | |
| "grad_norm": 4.8054423332214355, | |
| "learning_rate": 4.7761194029850745e-05, | |
| "loss": 0.0232, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2345415778251599, | |
| "grad_norm": 5.993706703186035, | |
| "learning_rate": 4.765458422174841e-05, | |
| "loss": 0.118, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24520255863539445, | |
| "grad_norm": 5.776984691619873, | |
| "learning_rate": 4.7547974413646055e-05, | |
| "loss": 0.0636, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.255863539445629, | |
| "grad_norm": 0.051732320338487625, | |
| "learning_rate": 4.7441364605543716e-05, | |
| "loss": 0.0371, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26652452025586354, | |
| "grad_norm": 0.27035900950431824, | |
| "learning_rate": 4.7334754797441364e-05, | |
| "loss": 0.0433, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2771855010660981, | |
| "grad_norm": 2.741495370864868, | |
| "learning_rate": 4.7228144989339026e-05, | |
| "loss": 0.1704, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2878464818763326, | |
| "grad_norm": 0.19535724818706512, | |
| "learning_rate": 4.7121535181236674e-05, | |
| "loss": 0.0772, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 3.0284807682037354, | |
| "learning_rate": 4.7014925373134335e-05, | |
| "loss": 0.0605, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3091684434968017, | |
| "grad_norm": 0.9240772128105164, | |
| "learning_rate": 4.690831556503198e-05, | |
| "loss": 0.0668, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31982942430703626, | |
| "grad_norm": 1.153721809387207, | |
| "learning_rate": 4.6801705756929645e-05, | |
| "loss": 0.0963, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3304904051172708, | |
| "grad_norm": 3.4412803649902344, | |
| "learning_rate": 4.669509594882729e-05, | |
| "loss": 0.0652, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3411513859275053, | |
| "grad_norm": 5.013424873352051, | |
| "learning_rate": 4.658848614072495e-05, | |
| "loss": 0.0264, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.35181236673773986, | |
| "grad_norm": 8.79882526397705, | |
| "learning_rate": 4.64818763326226e-05, | |
| "loss": 0.1177, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3624733475479744, | |
| "grad_norm": 3.925863265991211, | |
| "learning_rate": 4.637526652452026e-05, | |
| "loss": 0.0777, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.373134328358209, | |
| "grad_norm": 0.6153923869132996, | |
| "learning_rate": 4.626865671641791e-05, | |
| "loss": 0.0302, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3837953091684435, | |
| "grad_norm": 2.0949082374572754, | |
| "learning_rate": 4.6162046908315566e-05, | |
| "loss": 0.0595, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39445628997867804, | |
| "grad_norm": 4.5985260009765625, | |
| "learning_rate": 4.605543710021322e-05, | |
| "loss": 0.1175, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4051172707889126, | |
| "grad_norm": 0.8480165004730225, | |
| "learning_rate": 4.5948827292110876e-05, | |
| "loss": 0.0152, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4157782515991471, | |
| "grad_norm": 0.18910257518291473, | |
| "learning_rate": 4.584221748400853e-05, | |
| "loss": 0.0675, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 6.259706497192383, | |
| "learning_rate": 4.5735607675906185e-05, | |
| "loss": 0.1624, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43710021321961623, | |
| "grad_norm": 6.591126441955566, | |
| "learning_rate": 4.562899786780384e-05, | |
| "loss": 0.0605, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44776119402985076, | |
| "grad_norm": 3.2753217220306396, | |
| "learning_rate": 4.5522388059701495e-05, | |
| "loss": 0.104, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4584221748400853, | |
| "grad_norm": 6.949958324432373, | |
| "learning_rate": 4.541577825159915e-05, | |
| "loss": 0.0705, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4690831556503198, | |
| "grad_norm": 4.2345452308654785, | |
| "learning_rate": 4.5309168443496804e-05, | |
| "loss": 0.1018, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.47974413646055436, | |
| "grad_norm": 2.725113868713379, | |
| "learning_rate": 4.520255863539446e-05, | |
| "loss": 0.0366, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4904051172707889, | |
| "grad_norm": 3.393920421600342, | |
| "learning_rate": 4.5095948827292114e-05, | |
| "loss": 0.0637, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5010660980810234, | |
| "grad_norm": 5.699526786804199, | |
| "learning_rate": 4.498933901918977e-05, | |
| "loss": 0.0462, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.511727078891258, | |
| "grad_norm": 8.070642471313477, | |
| "learning_rate": 4.488272921108742e-05, | |
| "loss": 0.1336, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5223880597014925, | |
| "grad_norm": 2.0524590015411377, | |
| "learning_rate": 4.477611940298508e-05, | |
| "loss": 0.0778, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5330490405117271, | |
| "grad_norm": 8.874027252197266, | |
| "learning_rate": 4.466950959488273e-05, | |
| "loss": 0.094, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5437100213219617, | |
| "grad_norm": 4.671719551086426, | |
| "learning_rate": 4.456289978678039e-05, | |
| "loss": 0.0529, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5543710021321961, | |
| "grad_norm": 5.865716457366943, | |
| "learning_rate": 4.445628997867804e-05, | |
| "loss": 0.0458, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5650319829424307, | |
| "grad_norm": 2.4507577419281006, | |
| "learning_rate": 4.43496801705757e-05, | |
| "loss": 0.0289, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5756929637526652, | |
| "grad_norm": 2.382378339767456, | |
| "learning_rate": 4.424307036247335e-05, | |
| "loss": 0.0912, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5863539445628998, | |
| "grad_norm": 1.1386229991912842, | |
| "learning_rate": 4.4136460554371006e-05, | |
| "loss": 0.0387, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 1.5228124856948853, | |
| "learning_rate": 4.402985074626866e-05, | |
| "loss": 0.0443, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6076759061833689, | |
| "grad_norm": 4.459930896759033, | |
| "learning_rate": 4.3923240938166316e-05, | |
| "loss": 0.0544, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6183368869936035, | |
| "grad_norm": 0.01605360582470894, | |
| "learning_rate": 4.381663113006397e-05, | |
| "loss": 0.0293, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6289978678038379, | |
| "grad_norm": 3.7858338356018066, | |
| "learning_rate": 4.3710021321961625e-05, | |
| "loss": 0.1235, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6396588486140725, | |
| "grad_norm": 0.05360785126686096, | |
| "learning_rate": 4.360341151385928e-05, | |
| "loss": 0.006, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.650319829424307, | |
| "grad_norm": 2.8710269927978516, | |
| "learning_rate": 4.3496801705756935e-05, | |
| "loss": 0.0725, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6609808102345416, | |
| "grad_norm": 0.22796404361724854, | |
| "learning_rate": 4.339019189765459e-05, | |
| "loss": 0.0642, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6716417910447762, | |
| "grad_norm": 0.41445183753967285, | |
| "learning_rate": 4.328358208955224e-05, | |
| "loss": 0.0735, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6823027718550106, | |
| "grad_norm": 0.5575801730155945, | |
| "learning_rate": 4.31769722814499e-05, | |
| "loss": 0.1582, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6929637526652452, | |
| "grad_norm": 8.66786003112793, | |
| "learning_rate": 4.307036247334755e-05, | |
| "loss": 0.0393, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7036247334754797, | |
| "grad_norm": 8.188162803649902, | |
| "learning_rate": 4.29637526652452e-05, | |
| "loss": 0.0328, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 4.7807087898254395, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.0575, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7249466950959488, | |
| "grad_norm": 4.774600505828857, | |
| "learning_rate": 4.275053304904051e-05, | |
| "loss": 0.0693, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7356076759061834, | |
| "grad_norm": 0.24198925495147705, | |
| "learning_rate": 4.2643923240938166e-05, | |
| "loss": 0.0154, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "grad_norm": 6.401940822601318, | |
| "learning_rate": 4.253731343283582e-05, | |
| "loss": 0.0224, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7569296375266524, | |
| "grad_norm": 6.514515399932861, | |
| "learning_rate": 4.2430703624733475e-05, | |
| "loss": 0.043, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.767590618336887, | |
| "grad_norm": 3.973292350769043, | |
| "learning_rate": 4.232409381663113e-05, | |
| "loss": 0.0663, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7782515991471215, | |
| "grad_norm": 0.13549672067165375, | |
| "learning_rate": 4.2217484008528785e-05, | |
| "loss": 0.022, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7889125799573561, | |
| "grad_norm": 5.7366251945495605, | |
| "learning_rate": 4.211087420042644e-05, | |
| "loss": 0.0191, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7995735607675906, | |
| "grad_norm": 3.272925853729248, | |
| "learning_rate": 4.2004264392324094e-05, | |
| "loss": 0.0671, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8102345415778252, | |
| "grad_norm": 0.1337883323431015, | |
| "learning_rate": 4.189765458422175e-05, | |
| "loss": 0.0698, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8208955223880597, | |
| "grad_norm": 0.7873145937919617, | |
| "learning_rate": 4.1791044776119404e-05, | |
| "loss": 0.0538, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8315565031982942, | |
| "grad_norm": 4.945694923400879, | |
| "learning_rate": 4.168443496801706e-05, | |
| "loss": 0.1103, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8422174840085288, | |
| "grad_norm": 0.7360134124755859, | |
| "learning_rate": 4.157782515991471e-05, | |
| "loss": 0.0084, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 0.8266688585281372, | |
| "learning_rate": 4.147121535181237e-05, | |
| "loss": 0.0176, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8635394456289979, | |
| "grad_norm": 0.030890928581357002, | |
| "learning_rate": 4.136460554371002e-05, | |
| "loss": 0.0457, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8742004264392325, | |
| "grad_norm": 0.16308942437171936, | |
| "learning_rate": 4.125799573560768e-05, | |
| "loss": 0.0311, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8848614072494669, | |
| "grad_norm": 3.7586934566497803, | |
| "learning_rate": 4.115138592750533e-05, | |
| "loss": 0.0698, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8955223880597015, | |
| "grad_norm": 0.07764250785112381, | |
| "learning_rate": 4.104477611940299e-05, | |
| "loss": 0.0188, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.906183368869936, | |
| "grad_norm": 4.466598987579346, | |
| "learning_rate": 4.093816631130064e-05, | |
| "loss": 0.0593, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9168443496801706, | |
| "grad_norm": 0.17283818125724792, | |
| "learning_rate": 4.0831556503198296e-05, | |
| "loss": 0.0429, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9275053304904051, | |
| "grad_norm": 0.053455233573913574, | |
| "learning_rate": 4.072494669509595e-05, | |
| "loss": 0.0186, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9381663113006397, | |
| "grad_norm": 6.692505836486816, | |
| "learning_rate": 4.0618336886993606e-05, | |
| "loss": 0.0618, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9488272921108742, | |
| "grad_norm": 0.19959288835525513, | |
| "learning_rate": 4.051172707889126e-05, | |
| "loss": 0.0669, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9594882729211087, | |
| "grad_norm": 5.956345558166504, | |
| "learning_rate": 4.0405117270788915e-05, | |
| "loss": 0.0324, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9701492537313433, | |
| "grad_norm": 0.2815738320350647, | |
| "learning_rate": 4.029850746268657e-05, | |
| "loss": 0.0383, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9808102345415778, | |
| "grad_norm": 7.5461812019348145, | |
| "learning_rate": 4.0191897654584225e-05, | |
| "loss": 0.0607, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9914712153518124, | |
| "grad_norm": 8.231534957885742, | |
| "learning_rate": 4.008528784648188e-05, | |
| "loss": 0.0858, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9650666666666666, | |
| "eval_loss": 0.1182381734251976, | |
| "eval_runtime": 17.7307, | |
| "eval_samples_per_second": 211.497, | |
| "eval_steps_per_second": 6.655, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.0021321961620469, | |
| "grad_norm": 1.626712679862976, | |
| "learning_rate": 3.997867803837953e-05, | |
| "loss": 0.0185, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0127931769722816, | |
| "grad_norm": 0.044638343155384064, | |
| "learning_rate": 3.987206823027719e-05, | |
| "loss": 0.0412, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.023454157782516, | |
| "grad_norm": 0.020261414349079132, | |
| "learning_rate": 3.976545842217484e-05, | |
| "loss": 0.0137, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0341151385927505, | |
| "grad_norm": 0.05359117314219475, | |
| "learning_rate": 3.96588486140725e-05, | |
| "loss": 0.0062, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.044776119402985, | |
| "grad_norm": 0.8646987080574036, | |
| "learning_rate": 3.9552238805970146e-05, | |
| "loss": 0.0059, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0554371002132197, | |
| "grad_norm": 0.03817519545555115, | |
| "learning_rate": 3.944562899786781e-05, | |
| "loss": 0.0031, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.0660980810234542, | |
| "grad_norm": 6.9466681480407715, | |
| "learning_rate": 3.9339019189765456e-05, | |
| "loss": 0.0097, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0767590618336886, | |
| "grad_norm": 0.05334785580635071, | |
| "learning_rate": 3.923240938166312e-05, | |
| "loss": 0.0014, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0874200426439233, | |
| "grad_norm": 0.2097056657075882, | |
| "learning_rate": 3.9125799573560765e-05, | |
| "loss": 0.0011, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0980810234541578, | |
| "grad_norm": 0.09996387362480164, | |
| "learning_rate": 3.901918976545843e-05, | |
| "loss": 0.0018, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.1087420042643923, | |
| "grad_norm": 6.710844039916992, | |
| "learning_rate": 3.8912579957356075e-05, | |
| "loss": 0.011, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1194029850746268, | |
| "grad_norm": 0.36941811442375183, | |
| "learning_rate": 3.8805970149253736e-05, | |
| "loss": 0.0019, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.1300639658848615, | |
| "grad_norm": 0.21537742018699646, | |
| "learning_rate": 3.8699360341151384e-05, | |
| "loss": 0.0054, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.140724946695096, | |
| "grad_norm": 0.029234429821372032, | |
| "learning_rate": 3.8592750533049046e-05, | |
| "loss": 0.0052, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1513859275053304, | |
| "grad_norm": 0.01973758451640606, | |
| "learning_rate": 3.8486140724946694e-05, | |
| "loss": 0.0258, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.1620469083155651, | |
| "grad_norm": 1.7956969738006592, | |
| "learning_rate": 3.8379530916844355e-05, | |
| "loss": 0.0038, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.1727078891257996, | |
| "grad_norm": 0.039484452456235886, | |
| "learning_rate": 3.8272921108742e-05, | |
| "loss": 0.0012, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.183368869936034, | |
| "grad_norm": 5.022584915161133, | |
| "learning_rate": 3.8166311300639665e-05, | |
| "loss": 0.0028, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1940298507462686, | |
| "grad_norm": 0.1447264403104782, | |
| "learning_rate": 3.805970149253731e-05, | |
| "loss": 0.0371, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.2046908315565032, | |
| "grad_norm": 6.141026973724365, | |
| "learning_rate": 3.7953091684434974e-05, | |
| "loss": 0.006, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.2153518123667377, | |
| "grad_norm": 0.026512540876865387, | |
| "learning_rate": 3.784648187633262e-05, | |
| "loss": 0.0004, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.2260127931769722, | |
| "grad_norm": 0.0021177104208618402, | |
| "learning_rate": 3.7739872068230284e-05, | |
| "loss": 0.0141, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.236673773987207, | |
| "grad_norm": 0.12933123111724854, | |
| "learning_rate": 3.763326226012793e-05, | |
| "loss": 0.0005, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2473347547974414, | |
| "grad_norm": 0.05159159377217293, | |
| "learning_rate": 3.752665245202559e-05, | |
| "loss": 0.003, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.2579957356076759, | |
| "grad_norm": 0.015870722010731697, | |
| "learning_rate": 3.742004264392324e-05, | |
| "loss": 0.0015, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2686567164179103, | |
| "grad_norm": 0.007407957687973976, | |
| "learning_rate": 3.73134328358209e-05, | |
| "loss": 0.0181, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.279317697228145, | |
| "grad_norm": 0.04819609597325325, | |
| "learning_rate": 3.720682302771855e-05, | |
| "loss": 0.0011, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2899786780383795, | |
| "grad_norm": 0.1405772864818573, | |
| "learning_rate": 3.710021321961621e-05, | |
| "loss": 0.0116, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.3006396588486142, | |
| "grad_norm": 0.0038416755851358175, | |
| "learning_rate": 3.699360341151386e-05, | |
| "loss": 0.0003, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3113006396588487, | |
| "grad_norm": 0.013527607545256615, | |
| "learning_rate": 3.6886993603411515e-05, | |
| "loss": 0.0419, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.3219616204690832, | |
| "grad_norm": 0.04875783994793892, | |
| "learning_rate": 3.678038379530917e-05, | |
| "loss": 0.001, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.3326226012793176, | |
| "grad_norm": 0.014098434709012508, | |
| "learning_rate": 3.6673773987206824e-05, | |
| "loss": 0.0016, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.3432835820895521, | |
| "grad_norm": 13.899675369262695, | |
| "learning_rate": 3.656716417910448e-05, | |
| "loss": 0.0367, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.3539445628997868, | |
| "grad_norm": 0.0027879683766514063, | |
| "learning_rate": 3.6460554371002134e-05, | |
| "loss": 0.0017, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.3646055437100213, | |
| "grad_norm": 0.3567677140235901, | |
| "learning_rate": 3.635394456289979e-05, | |
| "loss": 0.0008, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.375266524520256, | |
| "grad_norm": 0.188824862241745, | |
| "learning_rate": 3.624733475479744e-05, | |
| "loss": 0.0103, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.3859275053304905, | |
| "grad_norm": 0.09603585302829742, | |
| "learning_rate": 3.61407249466951e-05, | |
| "loss": 0.0258, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.396588486140725, | |
| "grad_norm": 0.001065499964170158, | |
| "learning_rate": 3.603411513859275e-05, | |
| "loss": 0.0005, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.4072494669509594, | |
| "grad_norm": 13.907578468322754, | |
| "learning_rate": 3.592750533049041e-05, | |
| "loss": 0.0128, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.417910447761194, | |
| "grad_norm": 0.010527418926358223, | |
| "learning_rate": 3.582089552238806e-05, | |
| "loss": 0.0418, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.020130101591348648, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.0003, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.439232409381663, | |
| "grad_norm": 0.002067278139293194, | |
| "learning_rate": 3.560767590618337e-05, | |
| "loss": 0.0004, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.4498933901918978, | |
| "grad_norm": 0.34996917843818665, | |
| "learning_rate": 3.5501066098081026e-05, | |
| "loss": 0.0005, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4605543710021323, | |
| "grad_norm": 8.291366577148438, | |
| "learning_rate": 3.539445628997868e-05, | |
| "loss": 0.008, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.4712153518123667, | |
| "grad_norm": 20.71257781982422, | |
| "learning_rate": 3.5287846481876336e-05, | |
| "loss": 0.026, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4818763326226012, | |
| "grad_norm": 0.00548855634406209, | |
| "learning_rate": 3.518123667377399e-05, | |
| "loss": 0.0001, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "grad_norm": 0.012211128138005733, | |
| "learning_rate": 3.5074626865671645e-05, | |
| "loss": 0.0002, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5031982942430704, | |
| "grad_norm": 0.01268418226391077, | |
| "learning_rate": 3.496801705756929e-05, | |
| "loss": 0.0091, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.5138592750533049, | |
| "grad_norm": 0.01728120632469654, | |
| "learning_rate": 3.4861407249466955e-05, | |
| "loss": 0.039, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.5245202558635396, | |
| "grad_norm": 0.010496352799236774, | |
| "learning_rate": 3.47547974413646e-05, | |
| "loss": 0.0013, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.535181236673774, | |
| "grad_norm": 0.001391629921272397, | |
| "learning_rate": 3.4648187633262264e-05, | |
| "loss": 0.0005, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.5458422174840085, | |
| "grad_norm": 0.057142097502946854, | |
| "learning_rate": 3.454157782515991e-05, | |
| "loss": 0.0005, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.556503198294243, | |
| "grad_norm": 0.0026241871528327465, | |
| "learning_rate": 3.4434968017057574e-05, | |
| "loss": 0.0003, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5671641791044775, | |
| "grad_norm": 0.01787286438047886, | |
| "learning_rate": 3.432835820895522e-05, | |
| "loss": 0.0373, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.5778251599147122, | |
| "grad_norm": 0.000875612604431808, | |
| "learning_rate": 3.422174840085288e-05, | |
| "loss": 0.0001, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5884861407249466, | |
| "grad_norm": 0.014006942510604858, | |
| "learning_rate": 3.411513859275053e-05, | |
| "loss": 0.0003, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.5991471215351813, | |
| "grad_norm": 1.531774878501892, | |
| "learning_rate": 3.400852878464819e-05, | |
| "loss": 0.001, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.6098081023454158, | |
| "grad_norm": 0.0015476562548428774, | |
| "learning_rate": 3.390191897654584e-05, | |
| "loss": 0.0007, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.6204690831556503, | |
| "grad_norm": 0.012756136246025562, | |
| "learning_rate": 3.37953091684435e-05, | |
| "loss": 0.0239, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.6311300639658848, | |
| "grad_norm": 0.0007197922677733004, | |
| "learning_rate": 3.368869936034115e-05, | |
| "loss": 0.0071, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.6417910447761193, | |
| "grad_norm": 0.0055331019684672356, | |
| "learning_rate": 3.358208955223881e-05, | |
| "loss": 0.0104, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.652452025586354, | |
| "grad_norm": 0.02872060053050518, | |
| "learning_rate": 3.347547974413646e-05, | |
| "loss": 0.0002, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.6631130063965884, | |
| "grad_norm": 0.0017149861669167876, | |
| "learning_rate": 3.336886993603412e-05, | |
| "loss": 0.0098, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.6737739872068231, | |
| "grad_norm": 0.37227755784988403, | |
| "learning_rate": 3.326226012793177e-05, | |
| "loss": 0.0007, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.6844349680170576, | |
| "grad_norm": 0.01734251342713833, | |
| "learning_rate": 3.3155650319829424e-05, | |
| "loss": 0.0006, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.695095948827292, | |
| "grad_norm": 0.04362839087843895, | |
| "learning_rate": 3.304904051172708e-05, | |
| "loss": 0.003, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.7057569296375266, | |
| "grad_norm": 0.023149151355028152, | |
| "learning_rate": 3.294243070362473e-05, | |
| "loss": 0.0006, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.716417910447761, | |
| "grad_norm": 0.0008729319088160992, | |
| "learning_rate": 3.283582089552239e-05, | |
| "loss": 0.0003, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.7270788912579957, | |
| "grad_norm": 9.531977653503418, | |
| "learning_rate": 3.272921108742004e-05, | |
| "loss": 0.0113, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.7377398720682304, | |
| "grad_norm": 2.315152645111084, | |
| "learning_rate": 3.26226012793177e-05, | |
| "loss": 0.0654, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.748400852878465, | |
| "grad_norm": 0.02363394945859909, | |
| "learning_rate": 3.251599147121535e-05, | |
| "loss": 0.0003, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7590618336886994, | |
| "grad_norm": 0.44926193356513977, | |
| "learning_rate": 3.240938166311301e-05, | |
| "loss": 0.0003, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.7697228144989339, | |
| "grad_norm": 0.05977322906255722, | |
| "learning_rate": 3.230277185501066e-05, | |
| "loss": 0.014, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.7803837953091683, | |
| "grad_norm": 0.7710612416267395, | |
| "learning_rate": 3.2196162046908317e-05, | |
| "loss": 0.0112, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.7910447761194028, | |
| "grad_norm": 0.000869459065143019, | |
| "learning_rate": 3.208955223880597e-05, | |
| "loss": 0.0012, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.8017057569296375, | |
| "grad_norm": 0.4760114550590515, | |
| "learning_rate": 3.1982942430703626e-05, | |
| "loss": 0.0004, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.8123667377398722, | |
| "grad_norm": 0.0052476911805570126, | |
| "learning_rate": 3.187633262260128e-05, | |
| "loss": 0.0002, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.8230277185501067, | |
| "grad_norm": 8.89257526397705, | |
| "learning_rate": 3.1769722814498935e-05, | |
| "loss": 0.0269, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.8336886993603412, | |
| "grad_norm": 0.0020398315973579884, | |
| "learning_rate": 3.166311300639659e-05, | |
| "loss": 0.008, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.8443496801705757, | |
| "grad_norm": 0.3315824866294861, | |
| "learning_rate": 3.1556503198294245e-05, | |
| "loss": 0.0324, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.8550106609808101, | |
| "grad_norm": 0.019060449674725533, | |
| "learning_rate": 3.14498933901919e-05, | |
| "loss": 0.0183, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.8656716417910446, | |
| "grad_norm": 0.1318848729133606, | |
| "learning_rate": 3.1343283582089554e-05, | |
| "loss": 0.0004, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.8763326226012793, | |
| "grad_norm": 0.01659296080470085, | |
| "learning_rate": 3.123667377398721e-05, | |
| "loss": 0.0011, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.886993603411514, | |
| "grad_norm": 1.1072592735290527, | |
| "learning_rate": 3.1130063965884864e-05, | |
| "loss": 0.0247, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.8976545842217485, | |
| "grad_norm": 0.002012739423662424, | |
| "learning_rate": 3.102345415778252e-05, | |
| "loss": 0.0003, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.908315565031983, | |
| "grad_norm": 0.7516017556190491, | |
| "learning_rate": 3.0916844349680173e-05, | |
| "loss": 0.0009, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.9189765458422174, | |
| "grad_norm": 0.0022087525576353073, | |
| "learning_rate": 3.081023454157783e-05, | |
| "loss": 0.0874, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.929637526652452, | |
| "grad_norm": 0.0018624893855303526, | |
| "learning_rate": 3.070362473347548e-05, | |
| "loss": 0.0515, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.9402985074626866, | |
| "grad_norm": 0.06611283123493195, | |
| "learning_rate": 3.059701492537314e-05, | |
| "loss": 0.0011, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.950959488272921, | |
| "grad_norm": 0.18050549924373627, | |
| "learning_rate": 3.0490405117270792e-05, | |
| "loss": 0.0408, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.9616204690831558, | |
| "grad_norm": 0.013619977049529552, | |
| "learning_rate": 3.0383795309168444e-05, | |
| "loss": 0.0003, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.9722814498933903, | |
| "grad_norm": 0.012083163484930992, | |
| "learning_rate": 3.0277185501066102e-05, | |
| "loss": 0.0007, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.9829424307036247, | |
| "grad_norm": 0.007042413577437401, | |
| "learning_rate": 3.0170575692963753e-05, | |
| "loss": 0.0235, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.9936034115138592, | |
| "grad_norm": 0.02058466337621212, | |
| "learning_rate": 3.006396588486141e-05, | |
| "loss": 0.0003, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9698666666666667, | |
| "eval_loss": 0.13603569567203522, | |
| "eval_runtime": 17.5583, | |
| "eval_samples_per_second": 213.574, | |
| "eval_steps_per_second": 6.72, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 2.0042643923240937, | |
| "grad_norm": 0.0009510617819614708, | |
| "learning_rate": 2.9957356076759063e-05, | |
| "loss": 0.0002, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.014925373134328, | |
| "grad_norm": 0.0006816560053266585, | |
| "learning_rate": 2.9850746268656714e-05, | |
| "loss": 0.0001, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.025586353944563, | |
| "grad_norm": 0.015239764004945755, | |
| "learning_rate": 2.9744136460554372e-05, | |
| "loss": 0.0004, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.0362473347547976, | |
| "grad_norm": 0.033088065683841705, | |
| "learning_rate": 2.9637526652452023e-05, | |
| "loss": 0.0002, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.046908315565032, | |
| "grad_norm": 0.0019035555887967348, | |
| "learning_rate": 2.953091684434968e-05, | |
| "loss": 0.0001, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.0575692963752665, | |
| "grad_norm": 0.019931893795728683, | |
| "learning_rate": 2.9424307036247333e-05, | |
| "loss": 0.0193, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.068230277185501, | |
| "grad_norm": 0.013123378157615662, | |
| "learning_rate": 2.931769722814499e-05, | |
| "loss": 0.0035, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.0788912579957355, | |
| "grad_norm": 0.005247526802122593, | |
| "learning_rate": 2.9211087420042642e-05, | |
| "loss": 0.0002, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.08955223880597, | |
| "grad_norm": 0.00879177451133728, | |
| "learning_rate": 2.91044776119403e-05, | |
| "loss": 0.0003, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.100213219616205, | |
| "grad_norm": 0.0038365786895155907, | |
| "learning_rate": 2.8997867803837952e-05, | |
| "loss": 0.0037, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.1108742004264394, | |
| "grad_norm": 0.001064786803908646, | |
| "learning_rate": 2.889125799573561e-05, | |
| "loss": 0.0001, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.121535181236674, | |
| "grad_norm": 0.004799545276910067, | |
| "learning_rate": 2.878464818763326e-05, | |
| "loss": 0.0306, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.1321961620469083, | |
| "grad_norm": 9.035762786865234, | |
| "learning_rate": 2.867803837953092e-05, | |
| "loss": 0.0079, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.0010305867763236165, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.0002, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.1535181236673773, | |
| "grad_norm": 0.0045770746655762196, | |
| "learning_rate": 2.846481876332623e-05, | |
| "loss": 0.0006, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.1641791044776117, | |
| "grad_norm": 0.05552865192294121, | |
| "learning_rate": 2.835820895522388e-05, | |
| "loss": 0.0003, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.1748400852878467, | |
| "grad_norm": 0.01266010943800211, | |
| "learning_rate": 2.825159914712154e-05, | |
| "loss": 0.0004, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.185501066098081, | |
| "grad_norm": 0.14882685244083405, | |
| "learning_rate": 2.814498933901919e-05, | |
| "loss": 0.0004, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.1961620469083156, | |
| "grad_norm": 0.09102439135313034, | |
| "learning_rate": 2.8038379530916848e-05, | |
| "loss": 0.0008, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.20682302771855, | |
| "grad_norm": 0.000594578159507364, | |
| "learning_rate": 2.79317697228145e-05, | |
| "loss": 0.0002, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.2174840085287846, | |
| "grad_norm": 0.0005955426022410393, | |
| "learning_rate": 2.7825159914712157e-05, | |
| "loss": 0.0001, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.228144989339019, | |
| "grad_norm": 0.01295297872275114, | |
| "learning_rate": 2.771855010660981e-05, | |
| "loss": 0.0002, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.2388059701492535, | |
| "grad_norm": 0.0601358599960804, | |
| "learning_rate": 2.7611940298507467e-05, | |
| "loss": 0.0004, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.2494669509594885, | |
| "grad_norm": 0.0019663721323013306, | |
| "learning_rate": 2.7505330490405118e-05, | |
| "loss": 0.0001, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.260127931769723, | |
| "grad_norm": 0.0007762210443615913, | |
| "learning_rate": 2.7398720682302776e-05, | |
| "loss": 0.0001, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.2707889125799574, | |
| "grad_norm": 8.230362892150879, | |
| "learning_rate": 2.7292110874200428e-05, | |
| "loss": 0.0114, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.281449893390192, | |
| "grad_norm": 0.0019742657896131277, | |
| "learning_rate": 2.7185501066098086e-05, | |
| "loss": 0.0002, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.2921108742004264, | |
| "grad_norm": 0.0013049200642853975, | |
| "learning_rate": 2.7078891257995737e-05, | |
| "loss": 0.0002, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.302771855010661, | |
| "grad_norm": 0.0005671033868566155, | |
| "learning_rate": 2.6972281449893395e-05, | |
| "loss": 0.0001, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.3134328358208958, | |
| "grad_norm": 0.0004891786957159638, | |
| "learning_rate": 2.6865671641791047e-05, | |
| "loss": 0.0002, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.3240938166311302, | |
| "grad_norm": 0.0003661747614387423, | |
| "learning_rate": 2.6759061833688705e-05, | |
| "loss": 0.0088, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.3347547974413647, | |
| "grad_norm": 0.019423887133598328, | |
| "learning_rate": 2.6652452025586356e-05, | |
| "loss": 0.0001, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.345415778251599, | |
| "grad_norm": 0.0008206297643482685, | |
| "learning_rate": 2.6545842217484007e-05, | |
| "loss": 0.0003, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.3560767590618337, | |
| "grad_norm": 0.006124213803559542, | |
| "learning_rate": 2.6439232409381666e-05, | |
| "loss": 0.0001, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.366737739872068, | |
| "grad_norm": 12.176572799682617, | |
| "learning_rate": 2.6332622601279317e-05, | |
| "loss": 0.0096, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.3773987206823026, | |
| "grad_norm": 0.002056980039924383, | |
| "learning_rate": 2.6226012793176975e-05, | |
| "loss": 0.0001, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.388059701492537, | |
| "grad_norm": 0.019798509776592255, | |
| "learning_rate": 2.6119402985074626e-05, | |
| "loss": 0.0001, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.398720682302772, | |
| "grad_norm": 0.005863270256668329, | |
| "learning_rate": 2.6012793176972285e-05, | |
| "loss": 0.0001, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.4093816631130065, | |
| "grad_norm": 0.2485971599817276, | |
| "learning_rate": 2.5906183368869936e-05, | |
| "loss": 0.0004, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.420042643923241, | |
| "grad_norm": 2.3797671794891357, | |
| "learning_rate": 2.5799573560767594e-05, | |
| "loss": 0.001, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.4307036247334755, | |
| "grad_norm": 0.0006144079379737377, | |
| "learning_rate": 2.5692963752665245e-05, | |
| "loss": 0.0001, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.44136460554371, | |
| "grad_norm": 0.0013561258092522621, | |
| "learning_rate": 2.5586353944562904e-05, | |
| "loss": 0.0055, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.4520255863539444, | |
| "grad_norm": 0.007870888337492943, | |
| "learning_rate": 2.5479744136460555e-05, | |
| "loss": 0.0002, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.4626865671641793, | |
| "grad_norm": 0.023799536749720573, | |
| "learning_rate": 2.537313432835821e-05, | |
| "loss": 0.0003, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.473347547974414, | |
| "grad_norm": 0.19955310225486755, | |
| "learning_rate": 2.5266524520255864e-05, | |
| "loss": 0.0004, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.4840085287846483, | |
| "grad_norm": 0.03216109424829483, | |
| "learning_rate": 2.515991471215352e-05, | |
| "loss": 0.0001, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.4946695095948828, | |
| "grad_norm": 0.0009422688744962215, | |
| "learning_rate": 2.5053304904051174e-05, | |
| "loss": 0.017, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.5053304904051172, | |
| "grad_norm": 0.0025266490411013365, | |
| "learning_rate": 2.494669509594883e-05, | |
| "loss": 0.0001, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.5159914712153517, | |
| "grad_norm": 0.03144719451665878, | |
| "learning_rate": 2.4840085287846483e-05, | |
| "loss": 0.0002, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.526652452025586, | |
| "grad_norm": 0.004222916904836893, | |
| "learning_rate": 2.4733475479744138e-05, | |
| "loss": 0.0022, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.5373134328358207, | |
| "grad_norm": 0.23874346911907196, | |
| "learning_rate": 2.4626865671641793e-05, | |
| "loss": 0.0003, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.5479744136460556, | |
| "grad_norm": 0.008006093092262745, | |
| "learning_rate": 2.4520255863539444e-05, | |
| "loss": 0.0001, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.55863539445629, | |
| "grad_norm": 0.0004744882171507925, | |
| "learning_rate": 2.44136460554371e-05, | |
| "loss": 0.0023, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.5692963752665245, | |
| "grad_norm": 0.043702419847249985, | |
| "learning_rate": 2.4307036247334754e-05, | |
| "loss": 0.0001, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.579957356076759, | |
| "grad_norm": 0.01587110571563244, | |
| "learning_rate": 2.420042643923241e-05, | |
| "loss": 0.0002, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.5906183368869935, | |
| "grad_norm": 0.0004033471632283181, | |
| "learning_rate": 2.4093816631130063e-05, | |
| "loss": 0.0002, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.6012793176972284, | |
| "grad_norm": 0.003368718083947897, | |
| "learning_rate": 2.3987206823027718e-05, | |
| "loss": 0.0001, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.611940298507463, | |
| "grad_norm": 0.006782101932913065, | |
| "learning_rate": 2.3880597014925373e-05, | |
| "loss": 0.0001, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.6226012793176974, | |
| "grad_norm": 0.0014646780909970403, | |
| "learning_rate": 2.3773987206823027e-05, | |
| "loss": 0.0003, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.633262260127932, | |
| "grad_norm": 0.0032065703999251127, | |
| "learning_rate": 2.3667377398720682e-05, | |
| "loss": 0.0, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.6439232409381663, | |
| "grad_norm": 0.0025246513541787863, | |
| "learning_rate": 2.3560767590618337e-05, | |
| "loss": 0.0003, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.654584221748401, | |
| "grad_norm": 0.006225927267223597, | |
| "learning_rate": 2.345415778251599e-05, | |
| "loss": 0.0003, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.6652452025586353, | |
| "grad_norm": 0.003415194572880864, | |
| "learning_rate": 2.3347547974413646e-05, | |
| "loss": 0.0001, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.6759061833688698, | |
| "grad_norm": 0.0004717021365649998, | |
| "learning_rate": 2.32409381663113e-05, | |
| "loss": 0.0001, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.6865671641791042, | |
| "grad_norm": 0.0017549870535731316, | |
| "learning_rate": 2.3134328358208956e-05, | |
| "loss": 0.0, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.697228144989339, | |
| "grad_norm": 0.03759163245558739, | |
| "learning_rate": 2.302771855010661e-05, | |
| "loss": 0.0002, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.7078891257995736, | |
| "grad_norm": 0.0003970432444475591, | |
| "learning_rate": 2.2921108742004265e-05, | |
| "loss": 0.0001, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.718550106609808, | |
| "grad_norm": 0.0004572541802190244, | |
| "learning_rate": 2.281449893390192e-05, | |
| "loss": 0.0072, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.7292110874200426, | |
| "grad_norm": 0.00036305314279161394, | |
| "learning_rate": 2.2707889125799575e-05, | |
| "loss": 0.0001, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.739872068230277, | |
| "grad_norm": 0.0012093933764845133, | |
| "learning_rate": 2.260127931769723e-05, | |
| "loss": 0.0001, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.750533049040512, | |
| "grad_norm": 0.001646027434617281, | |
| "learning_rate": 2.2494669509594884e-05, | |
| "loss": 0.0001, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.7611940298507465, | |
| "grad_norm": 0.0057097249664366245, | |
| "learning_rate": 2.238805970149254e-05, | |
| "loss": 0.0, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.771855010660981, | |
| "grad_norm": 0.003947580698877573, | |
| "learning_rate": 2.2281449893390194e-05, | |
| "loss": 0.0001, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.7825159914712154, | |
| "grad_norm": 0.0011177313281223178, | |
| "learning_rate": 2.217484008528785e-05, | |
| "loss": 0.0001, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.79317697228145, | |
| "grad_norm": 0.00901961512863636, | |
| "learning_rate": 2.2068230277185503e-05, | |
| "loss": 0.0001, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.8038379530916844, | |
| "grad_norm": 0.5708050727844238, | |
| "learning_rate": 2.1961620469083158e-05, | |
| "loss": 0.0003, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.814498933901919, | |
| "grad_norm": 0.006386140827089548, | |
| "learning_rate": 2.1855010660980813e-05, | |
| "loss": 0.0001, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.8251599147121533, | |
| "grad_norm": 0.005727394018322229, | |
| "learning_rate": 2.1748400852878467e-05, | |
| "loss": 0.0002, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.835820895522388, | |
| "grad_norm": 0.0005577289848588407, | |
| "learning_rate": 2.164179104477612e-05, | |
| "loss": 0.0, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.8464818763326227, | |
| "grad_norm": 0.0015753644984215498, | |
| "learning_rate": 2.1535181236673773e-05, | |
| "loss": 0.0023, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.0005562901496887207, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.0001, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.8678038379530917, | |
| "grad_norm": 0.0017142018768936396, | |
| "learning_rate": 2.1321961620469083e-05, | |
| "loss": 0.0001, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.878464818763326, | |
| "grad_norm": 0.0013413095148280263, | |
| "learning_rate": 2.1215351812366738e-05, | |
| "loss": 0.0003, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.8891257995735606, | |
| "grad_norm": 0.09430250525474548, | |
| "learning_rate": 2.1108742004264392e-05, | |
| "loss": 0.0001, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.8997867803837956, | |
| "grad_norm": 0.02386680245399475, | |
| "learning_rate": 2.1002132196162047e-05, | |
| "loss": 0.0084, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.91044776119403, | |
| "grad_norm": 0.007424687035381794, | |
| "learning_rate": 2.0895522388059702e-05, | |
| "loss": 0.0001, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.9211087420042645, | |
| "grad_norm": 0.000830336706712842, | |
| "learning_rate": 2.0788912579957357e-05, | |
| "loss": 0.0001, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.931769722814499, | |
| "grad_norm": 0.0005704381037503481, | |
| "learning_rate": 2.068230277185501e-05, | |
| "loss": 0.0001, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.9424307036247335, | |
| "grad_norm": 0.00079008360626176, | |
| "learning_rate": 2.0575692963752666e-05, | |
| "loss": 0.0001, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.953091684434968, | |
| "grad_norm": 0.0024701335933059454, | |
| "learning_rate": 2.046908315565032e-05, | |
| "loss": 0.0001, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.9637526652452024, | |
| "grad_norm": 0.0005491616320796311, | |
| "learning_rate": 2.0362473347547976e-05, | |
| "loss": 0.0001, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.974413646055437, | |
| "grad_norm": 0.0011273273266851902, | |
| "learning_rate": 2.025586353944563e-05, | |
| "loss": 0.0001, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "grad_norm": 0.0006988812820054591, | |
| "learning_rate": 2.0149253731343285e-05, | |
| "loss": 0.0169, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.9957356076759063, | |
| "grad_norm": 0.00045983062591403723, | |
| "learning_rate": 2.004264392324094e-05, | |
| "loss": 0.0011, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9714666666666667, | |
| "eval_loss": 0.14409813284873962, | |
| "eval_runtime": 17.5343, | |
| "eval_samples_per_second": 213.867, | |
| "eval_steps_per_second": 6.73, | |
| "step": 1407 | |
| }, | |
| { | |
| "epoch": 3.0063965884861408, | |
| "grad_norm": 0.0011053835041821003, | |
| "learning_rate": 1.9936034115138594e-05, | |
| "loss": 0.0001, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.0170575692963753, | |
| "grad_norm": 0.0007745756884105504, | |
| "learning_rate": 1.982942430703625e-05, | |
| "loss": 0.0001, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 3.0277185501066097, | |
| "grad_norm": 0.002371458802372217, | |
| "learning_rate": 1.9722814498933904e-05, | |
| "loss": 0.0001, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.038379530916844, | |
| "grad_norm": 0.003524255705997348, | |
| "learning_rate": 1.961620469083156e-05, | |
| "loss": 0.0001, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 3.0490405117270787, | |
| "grad_norm": 0.0003550679248291999, | |
| "learning_rate": 1.9509594882729213e-05, | |
| "loss": 0.0, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.0597014925373136, | |
| "grad_norm": 0.003954087849706411, | |
| "learning_rate": 1.9402985074626868e-05, | |
| "loss": 0.0001, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 3.070362473347548, | |
| "grad_norm": 0.0004206807352602482, | |
| "learning_rate": 1.9296375266524523e-05, | |
| "loss": 0.0, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.0810234541577826, | |
| "grad_norm": 0.009241187013685703, | |
| "learning_rate": 1.9189765458422178e-05, | |
| "loss": 0.0001, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 3.091684434968017, | |
| "grad_norm": 0.0006860322318971157, | |
| "learning_rate": 1.9083155650319832e-05, | |
| "loss": 0.0001, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.1023454157782515, | |
| "grad_norm": 0.0003270851157139987, | |
| "learning_rate": 1.8976545842217487e-05, | |
| "loss": 0.0, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 3.113006396588486, | |
| "grad_norm": 0.0017470840830355883, | |
| "learning_rate": 1.8869936034115142e-05, | |
| "loss": 0.0, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.1236673773987205, | |
| "grad_norm": 0.0013792329700663686, | |
| "learning_rate": 1.8763326226012797e-05, | |
| "loss": 0.0001, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 3.1343283582089554, | |
| "grad_norm": 0.0008241386385634542, | |
| "learning_rate": 1.865671641791045e-05, | |
| "loss": 0.0, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.14498933901919, | |
| "grad_norm": 0.000447383732534945, | |
| "learning_rate": 1.8550106609808106e-05, | |
| "loss": 0.0, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 3.1556503198294243, | |
| "grad_norm": 0.0029486112762242556, | |
| "learning_rate": 1.8443496801705757e-05, | |
| "loss": 0.0024, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.166311300639659, | |
| "grad_norm": 0.000643405772279948, | |
| "learning_rate": 1.8336886993603412e-05, | |
| "loss": 0.0, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 3.1769722814498933, | |
| "grad_norm": 0.0009485495975241065, | |
| "learning_rate": 1.8230277185501067e-05, | |
| "loss": 0.0001, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.1876332622601278, | |
| "grad_norm": 0.0021624306682497263, | |
| "learning_rate": 1.812366737739872e-05, | |
| "loss": 0.0001, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 3.1982942430703627, | |
| "grad_norm": 0.0012437114492058754, | |
| "learning_rate": 1.8017057569296376e-05, | |
| "loss": 0.0002, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.208955223880597, | |
| "grad_norm": 0.0004210584447719157, | |
| "learning_rate": 1.791044776119403e-05, | |
| "loss": 0.0, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 3.2196162046908317, | |
| "grad_norm": 0.006191976834088564, | |
| "learning_rate": 1.7803837953091686e-05, | |
| "loss": 0.0, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.230277185501066, | |
| "grad_norm": 0.0034072811249643564, | |
| "learning_rate": 1.769722814498934e-05, | |
| "loss": 0.0001, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 3.2409381663113006, | |
| "grad_norm": 0.0012220110511407256, | |
| "learning_rate": 1.7590618336886995e-05, | |
| "loss": 0.0001, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.251599147121535, | |
| "grad_norm": 0.0008943532011471689, | |
| "learning_rate": 1.7484008528784647e-05, | |
| "loss": 0.0, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 3.2622601279317696, | |
| "grad_norm": 0.017330335453152657, | |
| "learning_rate": 1.73773987206823e-05, | |
| "loss": 0.0001, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.272921108742004, | |
| "grad_norm": 0.00029871691367588937, | |
| "learning_rate": 1.7270788912579956e-05, | |
| "loss": 0.0001, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 3.283582089552239, | |
| "grad_norm": 0.0006164326332509518, | |
| "learning_rate": 1.716417910447761e-05, | |
| "loss": 0.0, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.2942430703624734, | |
| "grad_norm": 0.0003607381077017635, | |
| "learning_rate": 1.7057569296375266e-05, | |
| "loss": 0.0002, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 3.304904051172708, | |
| "grad_norm": 0.00034220601082779467, | |
| "learning_rate": 1.695095948827292e-05, | |
| "loss": 0.0, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.3155650319829424, | |
| "grad_norm": 0.0003526586224325001, | |
| "learning_rate": 1.6844349680170575e-05, | |
| "loss": 0.0, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 3.326226012793177, | |
| "grad_norm": 0.0012586143566295505, | |
| "learning_rate": 1.673773987206823e-05, | |
| "loss": 0.0001, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.3368869936034113, | |
| "grad_norm": 0.005179602187126875, | |
| "learning_rate": 1.6631130063965885e-05, | |
| "loss": 0.0001, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 3.3475479744136463, | |
| "grad_norm": 0.0007326197810471058, | |
| "learning_rate": 1.652452025586354e-05, | |
| "loss": 0.0, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.3582089552238807, | |
| "grad_norm": 0.011357685551047325, | |
| "learning_rate": 1.6417910447761194e-05, | |
| "loss": 0.0001, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 3.368869936034115, | |
| "grad_norm": 0.0070250509306788445, | |
| "learning_rate": 1.631130063965885e-05, | |
| "loss": 0.0, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.3795309168443497, | |
| "grad_norm": 0.006196409463882446, | |
| "learning_rate": 1.6204690831556504e-05, | |
| "loss": 0.0, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 3.390191897654584, | |
| "grad_norm": 0.0013622590340673923, | |
| "learning_rate": 1.6098081023454158e-05, | |
| "loss": 0.0, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 3.4008528784648187, | |
| "grad_norm": 0.021805042400956154, | |
| "learning_rate": 1.5991471215351813e-05, | |
| "loss": 0.0001, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 3.411513859275053, | |
| "grad_norm": 0.0031194714829325676, | |
| "learning_rate": 1.5884861407249468e-05, | |
| "loss": 0.0002, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.4221748400852876, | |
| "grad_norm": 0.0006978256278671324, | |
| "learning_rate": 1.5778251599147122e-05, | |
| "loss": 0.0, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 3.4328358208955225, | |
| "grad_norm": 0.0003929181839339435, | |
| "learning_rate": 1.5671641791044777e-05, | |
| "loss": 0.0, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 3.443496801705757, | |
| "grad_norm": 0.0003352670173626393, | |
| "learning_rate": 1.5565031982942432e-05, | |
| "loss": 0.0, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 3.4541577825159915, | |
| "grad_norm": 0.001049940474331379, | |
| "learning_rate": 1.5458422174840087e-05, | |
| "loss": 0.0001, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.464818763326226, | |
| "grad_norm": 0.003111952682957053, | |
| "learning_rate": 1.535181236673774e-05, | |
| "loss": 0.0001, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 3.4754797441364604, | |
| "grad_norm": 0.002664284547790885, | |
| "learning_rate": 1.5245202558635396e-05, | |
| "loss": 0.0, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 3.486140724946695, | |
| "grad_norm": 0.0004285582690499723, | |
| "learning_rate": 1.5138592750533051e-05, | |
| "loss": 0.0, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 3.49680170575693, | |
| "grad_norm": 0.0004624539869837463, | |
| "learning_rate": 1.5031982942430706e-05, | |
| "loss": 0.0, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.5074626865671643, | |
| "grad_norm": 0.0014984884764999151, | |
| "learning_rate": 1.4925373134328357e-05, | |
| "loss": 0.0, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 3.518123667377399, | |
| "grad_norm": 0.0017825720133259892, | |
| "learning_rate": 1.4818763326226012e-05, | |
| "loss": 0.0, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.5287846481876333, | |
| "grad_norm": 0.00029846158577129245, | |
| "learning_rate": 1.4712153518123666e-05, | |
| "loss": 0.0, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 3.5394456289978677, | |
| "grad_norm": 0.0003302747500129044, | |
| "learning_rate": 1.4605543710021321e-05, | |
| "loss": 0.0, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 3.550106609808102, | |
| "grad_norm": 0.0009528864175081253, | |
| "learning_rate": 1.4498933901918976e-05, | |
| "loss": 0.0, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 3.5607675906183367, | |
| "grad_norm": 0.0008414736366830766, | |
| "learning_rate": 1.439232409381663e-05, | |
| "loss": 0.0001, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.0007819927996024489, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.0001, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 3.582089552238806, | |
| "grad_norm": 0.0005235728458501399, | |
| "learning_rate": 1.417910447761194e-05, | |
| "loss": 0.0, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 3.5927505330490406, | |
| "grad_norm": 0.0009267384302802384, | |
| "learning_rate": 1.4072494669509595e-05, | |
| "loss": 0.0, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 3.603411513859275, | |
| "grad_norm": 0.0036918336991220713, | |
| "learning_rate": 1.396588486140725e-05, | |
| "loss": 0.0001, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 3.6140724946695095, | |
| "grad_norm": 0.0010328283533453941, | |
| "learning_rate": 1.3859275053304904e-05, | |
| "loss": 0.0, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 3.624733475479744, | |
| "grad_norm": 0.001155112055130303, | |
| "learning_rate": 1.3752665245202559e-05, | |
| "loss": 0.0, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.635394456289979, | |
| "grad_norm": 0.0015656034229323268, | |
| "learning_rate": 1.3646055437100214e-05, | |
| "loss": 0.0, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 3.6460554371002134, | |
| "grad_norm": 0.0005122123984619975, | |
| "learning_rate": 1.3539445628997869e-05, | |
| "loss": 0.0, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 3.656716417910448, | |
| "grad_norm": 0.0012595773441717029, | |
| "learning_rate": 1.3432835820895523e-05, | |
| "loss": 0.0001, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 3.6673773987206824, | |
| "grad_norm": 0.006748442072421312, | |
| "learning_rate": 1.3326226012793178e-05, | |
| "loss": 0.0001, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.678038379530917, | |
| "grad_norm": 0.002126007340848446, | |
| "learning_rate": 1.3219616204690833e-05, | |
| "loss": 0.0, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 3.6886993603411513, | |
| "grad_norm": 0.04627062752842903, | |
| "learning_rate": 1.3113006396588488e-05, | |
| "loss": 0.0001, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.699360341151386, | |
| "grad_norm": 0.0007787595968693495, | |
| "learning_rate": 1.3006396588486142e-05, | |
| "loss": 0.0074, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 3.7100213219616203, | |
| "grad_norm": 0.004638800397515297, | |
| "learning_rate": 1.2899786780383797e-05, | |
| "loss": 0.0, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.7206823027718547, | |
| "grad_norm": 0.0003473752294667065, | |
| "learning_rate": 1.2793176972281452e-05, | |
| "loss": 0.0, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 3.7313432835820897, | |
| "grad_norm": 0.0002881583641283214, | |
| "learning_rate": 1.2686567164179105e-05, | |
| "loss": 0.0, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.742004264392324, | |
| "grad_norm": 0.0016327690100297332, | |
| "learning_rate": 1.257995735607676e-05, | |
| "loss": 0.0001, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 3.7526652452025586, | |
| "grad_norm": 0.0009497848222963512, | |
| "learning_rate": 1.2473347547974414e-05, | |
| "loss": 0.0, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.763326226012793, | |
| "grad_norm": 0.0006584698567166924, | |
| "learning_rate": 1.2366737739872069e-05, | |
| "loss": 0.0, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 3.7739872068230276, | |
| "grad_norm": 0.0017539839027449489, | |
| "learning_rate": 1.2260127931769722e-05, | |
| "loss": 0.0, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.7846481876332625, | |
| "grad_norm": 0.03169434890151024, | |
| "learning_rate": 1.2153518123667377e-05, | |
| "loss": 0.0001, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 3.795309168443497, | |
| "grad_norm": 0.0013038687175139785, | |
| "learning_rate": 1.2046908315565032e-05, | |
| "loss": 0.0, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.8059701492537314, | |
| "grad_norm": 0.00035355836735107005, | |
| "learning_rate": 1.1940298507462686e-05, | |
| "loss": 0.0, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 3.816631130063966, | |
| "grad_norm": 0.00602176645770669, | |
| "learning_rate": 1.1833688699360341e-05, | |
| "loss": 0.0, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.8272921108742004, | |
| "grad_norm": 0.001417440245859325, | |
| "learning_rate": 1.1727078891257996e-05, | |
| "loss": 0.0, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 3.837953091684435, | |
| "grad_norm": 0.0007981807575561106, | |
| "learning_rate": 1.162046908315565e-05, | |
| "loss": 0.0, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.8486140724946694, | |
| "grad_norm": 0.00032953996560536325, | |
| "learning_rate": 1.1513859275053305e-05, | |
| "loss": 0.0025, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 3.859275053304904, | |
| "grad_norm": 0.00046980270417407155, | |
| "learning_rate": 1.140724946695096e-05, | |
| "loss": 0.0, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.8699360341151388, | |
| "grad_norm": 0.0029039843939244747, | |
| "learning_rate": 1.1300639658848615e-05, | |
| "loss": 0.0, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 3.8805970149253732, | |
| "grad_norm": 0.0006313789053820074, | |
| "learning_rate": 1.119402985074627e-05, | |
| "loss": 0.0, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.8912579957356077, | |
| "grad_norm": 0.028328822925686836, | |
| "learning_rate": 1.1087420042643924e-05, | |
| "loss": 0.0001, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 3.901918976545842, | |
| "grad_norm": 0.0003210967988707125, | |
| "learning_rate": 1.0980810234541579e-05, | |
| "loss": 0.0, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.9125799573560767, | |
| "grad_norm": 0.0005060680559836328, | |
| "learning_rate": 1.0874200426439234e-05, | |
| "loss": 0.0, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 3.923240938166311, | |
| "grad_norm": 0.00031848024809733033, | |
| "learning_rate": 1.0767590618336887e-05, | |
| "loss": 0.0, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.933901918976546, | |
| "grad_norm": 0.0008760714554227889, | |
| "learning_rate": 1.0660980810234541e-05, | |
| "loss": 0.0, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 3.9445628997867805, | |
| "grad_norm": 0.0002840524248313159, | |
| "learning_rate": 1.0554371002132196e-05, | |
| "loss": 0.0001, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.955223880597015, | |
| "grad_norm": 0.0009671676089055836, | |
| "learning_rate": 1.0447761194029851e-05, | |
| "loss": 0.0, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 3.9658848614072495, | |
| "grad_norm": 0.0003127472009509802, | |
| "learning_rate": 1.0341151385927506e-05, | |
| "loss": 0.0, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.976545842217484, | |
| "grad_norm": 0.0003625962126534432, | |
| "learning_rate": 1.023454157782516e-05, | |
| "loss": 0.0, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 3.9872068230277184, | |
| "grad_norm": 0.00043455432751215994, | |
| "learning_rate": 1.0127931769722815e-05, | |
| "loss": 0.0, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.997867803837953, | |
| "grad_norm": 0.00046884582843631506, | |
| "learning_rate": 1.002132196162047e-05, | |
| "loss": 0.0, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9725333333333334, | |
| "eval_loss": 0.1464909166097641, | |
| "eval_runtime": 17.7302, | |
| "eval_samples_per_second": 211.504, | |
| "eval_steps_per_second": 6.655, | |
| "step": 1876 | |
| }, | |
| { | |
| "epoch": 4.008528784648187, | |
| "grad_norm": 0.006194824818521738, | |
| "learning_rate": 9.914712153518125e-06, | |
| "loss": 0.0, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.019189765458422, | |
| "grad_norm": 0.0008517757523804903, | |
| "learning_rate": 9.80810234541578e-06, | |
| "loss": 0.0, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 4.029850746268656, | |
| "grad_norm": 0.00042782819946296513, | |
| "learning_rate": 9.701492537313434e-06, | |
| "loss": 0.0, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.040511727078891, | |
| "grad_norm": 0.0006629274575971067, | |
| "learning_rate": 9.594882729211089e-06, | |
| "loss": 0.0, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 4.051172707889126, | |
| "grad_norm": 0.006160558667033911, | |
| "learning_rate": 9.488272921108744e-06, | |
| "loss": 0.0001, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.061833688699361, | |
| "grad_norm": 0.0018581621116027236, | |
| "learning_rate": 9.381663113006398e-06, | |
| "loss": 0.0, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 4.072494669509595, | |
| "grad_norm": 0.00031959457555785775, | |
| "learning_rate": 9.275053304904053e-06, | |
| "loss": 0.0, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.08315565031983, | |
| "grad_norm": 0.0004140738456044346, | |
| "learning_rate": 9.168443496801706e-06, | |
| "loss": 0.0, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 4.093816631130064, | |
| "grad_norm": 0.0011738823959603906, | |
| "learning_rate": 9.06183368869936e-06, | |
| "loss": 0.0, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.104477611940299, | |
| "grad_norm": 0.00555069325491786, | |
| "learning_rate": 8.955223880597016e-06, | |
| "loss": 0.0, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 4.115138592750533, | |
| "grad_norm": 0.000303637731121853, | |
| "learning_rate": 8.84861407249467e-06, | |
| "loss": 0.0, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.1257995735607675, | |
| "grad_norm": 0.0006620746571570635, | |
| "learning_rate": 8.742004264392323e-06, | |
| "loss": 0.0, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 4.136460554371002, | |
| "grad_norm": 0.00030262264772318304, | |
| "learning_rate": 8.635394456289978e-06, | |
| "loss": 0.0, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.1471215351812365, | |
| "grad_norm": 0.0018032094230875373, | |
| "learning_rate": 8.528784648187633e-06, | |
| "loss": 0.0001, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 4.157782515991471, | |
| "grad_norm": 0.0006144769722595811, | |
| "learning_rate": 8.422174840085288e-06, | |
| "loss": 0.0, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.1684434968017055, | |
| "grad_norm": 0.00028534428565762937, | |
| "learning_rate": 8.315565031982942e-06, | |
| "loss": 0.0001, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 4.17910447761194, | |
| "grad_norm": 0.0004092085873708129, | |
| "learning_rate": 8.208955223880597e-06, | |
| "loss": 0.0, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.189765458422174, | |
| "grad_norm": 0.001554907881654799, | |
| "learning_rate": 8.102345415778252e-06, | |
| "loss": 0.0, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 4.20042643923241, | |
| "grad_norm": 0.0002754669403657317, | |
| "learning_rate": 7.995735607675907e-06, | |
| "loss": 0.0, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 4.211087420042644, | |
| "grad_norm": 0.00036185182398185134, | |
| "learning_rate": 7.889125799573561e-06, | |
| "loss": 0.0, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 4.221748400852879, | |
| "grad_norm": 0.00029089066083543, | |
| "learning_rate": 7.782515991471216e-06, | |
| "loss": 0.0, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.232409381663113, | |
| "grad_norm": 0.0007173977210186422, | |
| "learning_rate": 7.67590618336887e-06, | |
| "loss": 0.0, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 4.243070362473348, | |
| "grad_norm": 0.0002844176779035479, | |
| "learning_rate": 7.5692963752665255e-06, | |
| "loss": 0.0, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 4.253731343283582, | |
| "grad_norm": 0.0004694333765655756, | |
| "learning_rate": 7.4626865671641785e-06, | |
| "loss": 0.0, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 4.264392324093817, | |
| "grad_norm": 0.0003098017768934369, | |
| "learning_rate": 7.356076759061833e-06, | |
| "loss": 0.0, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.275053304904051, | |
| "grad_norm": 0.0008835799526423216, | |
| "learning_rate": 7.249466950959488e-06, | |
| "loss": 0.0, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.0005121708963997662, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.0, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 4.29637526652452, | |
| "grad_norm": 0.000392212881706655, | |
| "learning_rate": 7.0362473347547975e-06, | |
| "loss": 0.0, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 4.3070362473347545, | |
| "grad_norm": 0.00047518289647996426, | |
| "learning_rate": 6.929637526652452e-06, | |
| "loss": 0.0, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 4.317697228144989, | |
| "grad_norm": 0.06014125421643257, | |
| "learning_rate": 6.823027718550107e-06, | |
| "loss": 0.0023, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 4.3283582089552235, | |
| "grad_norm": 0.0012079986045137048, | |
| "learning_rate": 6.716417910447762e-06, | |
| "loss": 0.0, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 4.339019189765459, | |
| "grad_norm": 0.0009479818982072175, | |
| "learning_rate": 6.609808102345416e-06, | |
| "loss": 0.0, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 4.349680170575693, | |
| "grad_norm": 0.0004274131206329912, | |
| "learning_rate": 6.503198294243071e-06, | |
| "loss": 0.0, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 4.360341151385928, | |
| "grad_norm": 0.007911008782684803, | |
| "learning_rate": 6.396588486140726e-06, | |
| "loss": 0.0, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 4.371002132196162, | |
| "grad_norm": 0.0010779683943837881, | |
| "learning_rate": 6.28997867803838e-06, | |
| "loss": 0.0, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.381663113006397, | |
| "grad_norm": 0.0009035211987793446, | |
| "learning_rate": 6.1833688699360345e-06, | |
| "loss": 0.0, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 4.392324093816631, | |
| "grad_norm": 0.0003453451790846884, | |
| "learning_rate": 6.076759061833688e-06, | |
| "loss": 0.0, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 4.402985074626866, | |
| "grad_norm": 0.0006489038933068514, | |
| "learning_rate": 5.970149253731343e-06, | |
| "loss": 0.0, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 4.4136460554371, | |
| "grad_norm": 0.0002653495466802269, | |
| "learning_rate": 5.863539445628998e-06, | |
| "loss": 0.0, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 4.424307036247335, | |
| "grad_norm": 0.0007485944661311805, | |
| "learning_rate": 5.756929637526653e-06, | |
| "loss": 0.0, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 4.434968017057569, | |
| "grad_norm": 0.0003160152118653059, | |
| "learning_rate": 5.650319829424307e-06, | |
| "loss": 0.0, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 4.445628997867804, | |
| "grad_norm": 0.002550690434873104, | |
| "learning_rate": 5.543710021321962e-06, | |
| "loss": 0.0, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 4.456289978678038, | |
| "grad_norm": 0.0005198578000999987, | |
| "learning_rate": 5.437100213219617e-06, | |
| "loss": 0.0, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 4.466950959488273, | |
| "grad_norm": 0.0006188583793118596, | |
| "learning_rate": 5.330490405117271e-06, | |
| "loss": 0.0, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 4.477611940298507, | |
| "grad_norm": 0.00027445191517472267, | |
| "learning_rate": 5.2238805970149255e-06, | |
| "loss": 0.0, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.4882729211087415, | |
| "grad_norm": 0.00042016259976662695, | |
| "learning_rate": 5.11727078891258e-06, | |
| "loss": 0.0, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 4.498933901918977, | |
| "grad_norm": 0.0013817092403769493, | |
| "learning_rate": 5.010660980810235e-06, | |
| "loss": 0.0, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 4.509594882729211, | |
| "grad_norm": 0.0003576301387511194, | |
| "learning_rate": 4.90405117270789e-06, | |
| "loss": 0.0, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 4.520255863539446, | |
| "grad_norm": 0.0010557634523138404, | |
| "learning_rate": 4.797441364605544e-06, | |
| "loss": 0.0, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 4.53091684434968, | |
| "grad_norm": 0.0027745207771658897, | |
| "learning_rate": 4.690831556503199e-06, | |
| "loss": 0.0, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 4.541577825159915, | |
| "grad_norm": 0.00030267093097791076, | |
| "learning_rate": 4.584221748400853e-06, | |
| "loss": 0.0, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 4.552238805970149, | |
| "grad_norm": 0.0003039832809008658, | |
| "learning_rate": 4.477611940298508e-06, | |
| "loss": 0.0, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 4.562899786780384, | |
| "grad_norm": 0.00027207224047742784, | |
| "learning_rate": 4.371002132196162e-06, | |
| "loss": 0.0, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 4.573560767590618, | |
| "grad_norm": 0.0009374875808134675, | |
| "learning_rate": 4.264392324093816e-06, | |
| "loss": 0.0, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 4.584221748400853, | |
| "grad_norm": 0.0004837230662815273, | |
| "learning_rate": 4.157782515991471e-06, | |
| "loss": 0.0, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 4.594882729211087, | |
| "grad_norm": 0.00026056167553178966, | |
| "learning_rate": 4.051172707889126e-06, | |
| "loss": 0.0, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 4.605543710021322, | |
| "grad_norm": 0.0036224022042006254, | |
| "learning_rate": 3.944562899786781e-06, | |
| "loss": 0.0, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 4.616204690831556, | |
| "grad_norm": 0.0003278214717283845, | |
| "learning_rate": 3.837953091684435e-06, | |
| "loss": 0.0, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 4.6268656716417915, | |
| "grad_norm": 0.0005697670858353376, | |
| "learning_rate": 3.7313432835820893e-06, | |
| "loss": 0.0, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 4.637526652452026, | |
| "grad_norm": 0.0027591376565396786, | |
| "learning_rate": 3.624733475479744e-06, | |
| "loss": 0.0, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 4.6481876332622605, | |
| "grad_norm": 0.00035470424336381257, | |
| "learning_rate": 3.5181236673773987e-06, | |
| "loss": 0.0, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 4.658848614072495, | |
| "grad_norm": 0.00028932015993632376, | |
| "learning_rate": 3.4115138592750535e-06, | |
| "loss": 0.0, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 4.669509594882729, | |
| "grad_norm": 0.00032757414737716317, | |
| "learning_rate": 3.304904051172708e-06, | |
| "loss": 0.0, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 4.680170575692964, | |
| "grad_norm": 0.00032305470085702837, | |
| "learning_rate": 3.198294243070363e-06, | |
| "loss": 0.0, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 4.690831556503198, | |
| "grad_norm": 0.0037091304548084736, | |
| "learning_rate": 3.0916844349680173e-06, | |
| "loss": 0.0001, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.701492537313433, | |
| "grad_norm": 0.006712085101753473, | |
| "learning_rate": 2.9850746268656716e-06, | |
| "loss": 0.0001, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 4.712153518123667, | |
| "grad_norm": 0.0004929785500280559, | |
| "learning_rate": 2.8784648187633263e-06, | |
| "loss": 0.0, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 4.722814498933902, | |
| "grad_norm": 0.1973581612110138, | |
| "learning_rate": 2.771855010660981e-06, | |
| "loss": 0.0067, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 4.733475479744136, | |
| "grad_norm": 0.000468151003587991, | |
| "learning_rate": 2.6652452025586354e-06, | |
| "loss": 0.0, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 4.744136460554371, | |
| "grad_norm": 0.00032925998675636947, | |
| "learning_rate": 2.55863539445629e-06, | |
| "loss": 0.0, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 4.754797441364605, | |
| "grad_norm": 0.0012265060795471072, | |
| "learning_rate": 2.452025586353945e-06, | |
| "loss": 0.0, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 4.76545842217484, | |
| "grad_norm": 0.0003857238916680217, | |
| "learning_rate": 2.3454157782515996e-06, | |
| "loss": 0.0, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 4.776119402985074, | |
| "grad_norm": 0.0008554061059840024, | |
| "learning_rate": 2.238805970149254e-06, | |
| "loss": 0.0, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 4.786780383795309, | |
| "grad_norm": 0.0002912227646447718, | |
| "learning_rate": 2.132196162046908e-06, | |
| "loss": 0.0, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 4.797441364605544, | |
| "grad_norm": 0.0003559317556209862, | |
| "learning_rate": 2.025586353944563e-06, | |
| "loss": 0.0, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 4.8081023454157785, | |
| "grad_norm": 0.002918532583862543, | |
| "learning_rate": 1.9189765458422177e-06, | |
| "loss": 0.0, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 4.818763326226013, | |
| "grad_norm": 0.0013300913851708174, | |
| "learning_rate": 1.812366737739872e-06, | |
| "loss": 0.0, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 4.8294243070362475, | |
| "grad_norm": 0.0003671310842037201, | |
| "learning_rate": 1.7057569296375267e-06, | |
| "loss": 0.0, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 4.840085287846482, | |
| "grad_norm": 0.0005498518003150821, | |
| "learning_rate": 1.5991471215351815e-06, | |
| "loss": 0.0, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 4.850746268656716, | |
| "grad_norm": 0.00034973243600688875, | |
| "learning_rate": 1.4925373134328358e-06, | |
| "loss": 0.0, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 4.861407249466951, | |
| "grad_norm": 0.0003210071299690753, | |
| "learning_rate": 1.3859275053304905e-06, | |
| "loss": 0.0, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 4.872068230277185, | |
| "grad_norm": 0.0002745548845268786, | |
| "learning_rate": 1.279317697228145e-06, | |
| "loss": 0.0, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 4.88272921108742, | |
| "grad_norm": 0.0006946906796656549, | |
| "learning_rate": 1.1727078891257998e-06, | |
| "loss": 0.0, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 4.893390191897654, | |
| "grad_norm": 0.001909334328956902, | |
| "learning_rate": 1.066098081023454e-06, | |
| "loss": 0.0023, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 4.904051172707889, | |
| "grad_norm": 0.00044869142584502697, | |
| "learning_rate": 9.594882729211088e-07, | |
| "loss": 0.0001, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.914712153518123, | |
| "grad_norm": 0.0003012365195900202, | |
| "learning_rate": 8.528784648187634e-07, | |
| "loss": 0.0, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 4.925373134328359, | |
| "grad_norm": 0.0003137913590762764, | |
| "learning_rate": 7.462686567164179e-07, | |
| "loss": 0.0, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 4.936034115138593, | |
| "grad_norm": 0.0013513383455574512, | |
| "learning_rate": 6.396588486140725e-07, | |
| "loss": 0.0, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 4.946695095948828, | |
| "grad_norm": 0.0002704511571209878, | |
| "learning_rate": 5.33049040511727e-07, | |
| "loss": 0.0, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 4.957356076759062, | |
| "grad_norm": 0.0002641172322910279, | |
| "learning_rate": 4.264392324093817e-07, | |
| "loss": 0.0, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 4.968017057569297, | |
| "grad_norm": 0.0018425466259941459, | |
| "learning_rate": 3.1982942430703626e-07, | |
| "loss": 0.0, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 4.978678038379531, | |
| "grad_norm": 0.001115600811317563, | |
| "learning_rate": 2.1321961620469084e-07, | |
| "loss": 0.0, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 4.9893390191897655, | |
| "grad_norm": 0.0002641855098772794, | |
| "learning_rate": 1.0660980810234542e-07, | |
| "loss": 0.0, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.0004063702072016895, | |
| "learning_rate": 0.0, | |
| "loss": 0.0, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9728, | |
| "eval_loss": 0.1473887413740158, | |
| "eval_runtime": 17.8169, | |
| "eval_samples_per_second": 210.474, | |
| "eval_steps_per_second": 6.623, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 2345, | |
| "total_flos": 5.8118992210944e+18, | |
| "train_loss": 0.016123040216125505, | |
| "train_runtime": 807.1262, | |
| "train_samples_per_second": 92.922, | |
| "train_steps_per_second": 2.905 | |
| } | |
| ], | |
| "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 | |
| } | |