| { | |
| "best_metric": 0.059450652450323105, | |
| "best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles3_seed1_q1\\checkpoint-938", | |
| "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": 7.523022174835205, | |
| "learning_rate": 4.989339019189766e-05, | |
| "loss": 0.0624, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021321961620469083, | |
| "grad_norm": 0.07260505110025406, | |
| "learning_rate": 4.978678038379531e-05, | |
| "loss": 0.0251, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031982942430703626, | |
| "grad_norm": 12.401585578918457, | |
| "learning_rate": 4.9680170575692967e-05, | |
| "loss": 0.0088, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042643923240938165, | |
| "grad_norm": 13.994823455810547, | |
| "learning_rate": 4.957356076759062e-05, | |
| "loss": 0.0352, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053304904051172705, | |
| "grad_norm": 6.336690425872803, | |
| "learning_rate": 4.9466950959488276e-05, | |
| "loss": 0.0758, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06396588486140725, | |
| "grad_norm": 8.967635154724121, | |
| "learning_rate": 4.936034115138593e-05, | |
| "loss": 0.1398, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07462686567164178, | |
| "grad_norm": 1.0860027074813843, | |
| "learning_rate": 4.9253731343283586e-05, | |
| "loss": 0.1171, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08528784648187633, | |
| "grad_norm": 10.433364868164062, | |
| "learning_rate": 4.914712153518124e-05, | |
| "loss": 0.1258, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09594882729211088, | |
| "grad_norm": 12.0237398147583, | |
| "learning_rate": 4.904051172707889e-05, | |
| "loss": 0.1628, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10660980810234541, | |
| "grad_norm": 0.047775354236364365, | |
| "learning_rate": 4.893390191897655e-05, | |
| "loss": 0.0825, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11727078891257996, | |
| "grad_norm": 0.9382216334342957, | |
| "learning_rate": 4.88272921108742e-05, | |
| "loss": 0.0328, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1279317697228145, | |
| "grad_norm": 4.386844635009766, | |
| "learning_rate": 4.872068230277186e-05, | |
| "loss": 0.0104, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13859275053304904, | |
| "grad_norm": 0.0282029677182436, | |
| "learning_rate": 4.861407249466951e-05, | |
| "loss": 0.0909, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14925373134328357, | |
| "grad_norm": 0.10068627446889877, | |
| "learning_rate": 4.850746268656717e-05, | |
| "loss": 0.0533, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15991471215351813, | |
| "grad_norm": 0.6938403844833374, | |
| "learning_rate": 4.840085287846482e-05, | |
| "loss": 0.0468, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17057569296375266, | |
| "grad_norm": 2.18245530128479, | |
| "learning_rate": 4.829424307036248e-05, | |
| "loss": 0.0537, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1812366737739872, | |
| "grad_norm": 0.6952275037765503, | |
| "learning_rate": 4.8187633262260126e-05, | |
| "loss": 0.0502, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19189765458422176, | |
| "grad_norm": 0.10946524888277054, | |
| "learning_rate": 4.808102345415779e-05, | |
| "loss": 0.0098, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2025586353944563, | |
| "grad_norm": 0.023197470232844353, | |
| "learning_rate": 4.7974413646055436e-05, | |
| "loss": 0.0131, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21321961620469082, | |
| "grad_norm": 0.6925906538963318, | |
| "learning_rate": 4.78678038379531e-05, | |
| "loss": 0.0141, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22388059701492538, | |
| "grad_norm": 3.7687132358551025, | |
| "learning_rate": 4.7761194029850745e-05, | |
| "loss": 0.0293, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2345415778251599, | |
| "grad_norm": 0.34725797176361084, | |
| "learning_rate": 4.765458422174841e-05, | |
| "loss": 0.0026, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24520255863539445, | |
| "grad_norm": 8.361950874328613, | |
| "learning_rate": 4.7547974413646055e-05, | |
| "loss": 0.1031, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.255863539445629, | |
| "grad_norm": 3.192587375640869, | |
| "learning_rate": 4.7441364605543716e-05, | |
| "loss": 0.0303, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26652452025586354, | |
| "grad_norm": 0.13406366109848022, | |
| "learning_rate": 4.7334754797441364e-05, | |
| "loss": 0.0921, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2771855010660981, | |
| "grad_norm": 6.989262104034424, | |
| "learning_rate": 4.7228144989339026e-05, | |
| "loss": 0.0903, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2878464818763326, | |
| "grad_norm": 0.008396037854254246, | |
| "learning_rate": 4.7121535181236674e-05, | |
| "loss": 0.0009, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 13.521650314331055, | |
| "learning_rate": 4.7014925373134335e-05, | |
| "loss": 0.1023, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3091684434968017, | |
| "grad_norm": 0.11828146874904633, | |
| "learning_rate": 4.690831556503198e-05, | |
| "loss": 0.045, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31982942430703626, | |
| "grad_norm": 0.23041707277297974, | |
| "learning_rate": 4.6801705756929645e-05, | |
| "loss": 0.0303, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3304904051172708, | |
| "grad_norm": 1.239055871963501, | |
| "learning_rate": 4.669509594882729e-05, | |
| "loss": 0.0026, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3411513859275053, | |
| "grad_norm": 4.681065082550049, | |
| "learning_rate": 4.658848614072495e-05, | |
| "loss": 0.0364, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.35181236673773986, | |
| "grad_norm": 5.469776153564453, | |
| "learning_rate": 4.64818763326226e-05, | |
| "loss": 0.0121, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3624733475479744, | |
| "grad_norm": 2.5136218070983887, | |
| "learning_rate": 4.637526652452026e-05, | |
| "loss": 0.0625, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.373134328358209, | |
| "grad_norm": 5.015021324157715, | |
| "learning_rate": 4.626865671641791e-05, | |
| "loss": 0.0346, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3837953091684435, | |
| "grad_norm": 5.595672607421875, | |
| "learning_rate": 4.6162046908315566e-05, | |
| "loss": 0.1033, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39445628997867804, | |
| "grad_norm": 1.053650140762329, | |
| "learning_rate": 4.605543710021322e-05, | |
| "loss": 0.0572, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4051172707889126, | |
| "grad_norm": 0.6538744568824768, | |
| "learning_rate": 4.5948827292110876e-05, | |
| "loss": 0.018, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4157782515991471, | |
| "grad_norm": 1.1336654424667358, | |
| "learning_rate": 4.584221748400853e-05, | |
| "loss": 0.0264, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 0.10380210727453232, | |
| "learning_rate": 4.5735607675906185e-05, | |
| "loss": 0.049, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43710021321961623, | |
| "grad_norm": 0.2752641439437866, | |
| "learning_rate": 4.562899786780384e-05, | |
| "loss": 0.0372, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44776119402985076, | |
| "grad_norm": 0.018331853672862053, | |
| "learning_rate": 4.5522388059701495e-05, | |
| "loss": 0.0784, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4584221748400853, | |
| "grad_norm": 4.259432315826416, | |
| "learning_rate": 4.541577825159915e-05, | |
| "loss": 0.1846, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4690831556503198, | |
| "grad_norm": 2.1121721267700195, | |
| "learning_rate": 4.5309168443496804e-05, | |
| "loss": 0.089, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.47974413646055436, | |
| "grad_norm": 0.08846390247344971, | |
| "learning_rate": 4.520255863539446e-05, | |
| "loss": 0.0203, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4904051172707889, | |
| "grad_norm": 3.529513359069824, | |
| "learning_rate": 4.5095948827292114e-05, | |
| "loss": 0.0619, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5010660980810234, | |
| "grad_norm": 2.0754308700561523, | |
| "learning_rate": 4.498933901918977e-05, | |
| "loss": 0.0643, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.511727078891258, | |
| "grad_norm": 0.41601184010505676, | |
| "learning_rate": 4.488272921108742e-05, | |
| "loss": 0.0792, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5223880597014925, | |
| "grad_norm": 0.21950437128543854, | |
| "learning_rate": 4.477611940298508e-05, | |
| "loss": 0.0295, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5330490405117271, | |
| "grad_norm": 0.38914090394973755, | |
| "learning_rate": 4.466950959488273e-05, | |
| "loss": 0.0392, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5437100213219617, | |
| "grad_norm": 1.4780489206314087, | |
| "learning_rate": 4.456289978678039e-05, | |
| "loss": 0.0068, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5543710021321961, | |
| "grad_norm": 1.410050630569458, | |
| "learning_rate": 4.445628997867804e-05, | |
| "loss": 0.0052, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5650319829424307, | |
| "grad_norm": 1.8445395231246948, | |
| "learning_rate": 4.43496801705757e-05, | |
| "loss": 0.0472, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5756929637526652, | |
| "grad_norm": 1.1135804653167725, | |
| "learning_rate": 4.424307036247335e-05, | |
| "loss": 0.0457, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5863539445628998, | |
| "grad_norm": 0.1754743605852127, | |
| "learning_rate": 4.4136460554371006e-05, | |
| "loss": 0.0249, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 5.77360200881958, | |
| "learning_rate": 4.402985074626866e-05, | |
| "loss": 0.0621, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6076759061833689, | |
| "grad_norm": 0.1517520397901535, | |
| "learning_rate": 4.3923240938166316e-05, | |
| "loss": 0.0742, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6183368869936035, | |
| "grad_norm": 6.860696315765381, | |
| "learning_rate": 4.381663113006397e-05, | |
| "loss": 0.0445, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6289978678038379, | |
| "grad_norm": 9.31261920928955, | |
| "learning_rate": 4.3710021321961625e-05, | |
| "loss": 0.0586, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6396588486140725, | |
| "grad_norm": 0.022177331149578094, | |
| "learning_rate": 4.360341151385928e-05, | |
| "loss": 0.1295, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.650319829424307, | |
| "grad_norm": 2.3088161945343018, | |
| "learning_rate": 4.3496801705756935e-05, | |
| "loss": 0.0358, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6609808102345416, | |
| "grad_norm": 0.9260404109954834, | |
| "learning_rate": 4.339019189765459e-05, | |
| "loss": 0.0124, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6716417910447762, | |
| "grad_norm": 0.6644256114959717, | |
| "learning_rate": 4.328358208955224e-05, | |
| "loss": 0.0386, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6823027718550106, | |
| "grad_norm": 0.13432615995407104, | |
| "learning_rate": 4.31769722814499e-05, | |
| "loss": 0.0253, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6929637526652452, | |
| "grad_norm": 0.011860758997499943, | |
| "learning_rate": 4.307036247334755e-05, | |
| "loss": 0.0418, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7036247334754797, | |
| "grad_norm": 0.03158905729651451, | |
| "learning_rate": 4.29637526652452e-05, | |
| "loss": 0.0185, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 6.030971527099609, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.0248, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7249466950959488, | |
| "grad_norm": 0.25387847423553467, | |
| "learning_rate": 4.275053304904051e-05, | |
| "loss": 0.0046, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7356076759061834, | |
| "grad_norm": 6.103703498840332, | |
| "learning_rate": 4.2643923240938166e-05, | |
| "loss": 0.0573, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "grad_norm": 4.099830150604248, | |
| "learning_rate": 4.253731343283582e-05, | |
| "loss": 0.0132, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7569296375266524, | |
| "grad_norm": 8.336865425109863, | |
| "learning_rate": 4.2430703624733475e-05, | |
| "loss": 0.0711, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.767590618336887, | |
| "grad_norm": 0.04767255485057831, | |
| "learning_rate": 4.232409381663113e-05, | |
| "loss": 0.0543, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7782515991471215, | |
| "grad_norm": 2.82381534576416, | |
| "learning_rate": 4.2217484008528785e-05, | |
| "loss": 0.0295, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7889125799573561, | |
| "grad_norm": 3.60748553276062, | |
| "learning_rate": 4.211087420042644e-05, | |
| "loss": 0.028, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7995735607675906, | |
| "grad_norm": 0.21805140376091003, | |
| "learning_rate": 4.2004264392324094e-05, | |
| "loss": 0.0682, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8102345415778252, | |
| "grad_norm": 2.768376588821411, | |
| "learning_rate": 4.189765458422175e-05, | |
| "loss": 0.1194, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8208955223880597, | |
| "grad_norm": 0.9290714263916016, | |
| "learning_rate": 4.1791044776119404e-05, | |
| "loss": 0.0444, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8315565031982942, | |
| "grad_norm": 3.394944667816162, | |
| "learning_rate": 4.168443496801706e-05, | |
| "loss": 0.0494, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8422174840085288, | |
| "grad_norm": 0.16371995210647583, | |
| "learning_rate": 4.157782515991471e-05, | |
| "loss": 0.0337, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 7.573339462280273, | |
| "learning_rate": 4.147121535181237e-05, | |
| "loss": 0.0512, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8635394456289979, | |
| "grad_norm": 0.3734154999256134, | |
| "learning_rate": 4.136460554371002e-05, | |
| "loss": 0.1004, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8742004264392325, | |
| "grad_norm": 0.3404262661933899, | |
| "learning_rate": 4.125799573560768e-05, | |
| "loss": 0.0386, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8848614072494669, | |
| "grad_norm": 3.594224214553833, | |
| "learning_rate": 4.115138592750533e-05, | |
| "loss": 0.0347, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8955223880597015, | |
| "grad_norm": 0.39725932478904724, | |
| "learning_rate": 4.104477611940299e-05, | |
| "loss": 0.0246, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.906183368869936, | |
| "grad_norm": 0.10839869827032089, | |
| "learning_rate": 4.093816631130064e-05, | |
| "loss": 0.056, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9168443496801706, | |
| "grad_norm": 0.05952761694788933, | |
| "learning_rate": 4.0831556503198296e-05, | |
| "loss": 0.058, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9275053304904051, | |
| "grad_norm": 1.0791202783584595, | |
| "learning_rate": 4.072494669509595e-05, | |
| "loss": 0.006, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9381663113006397, | |
| "grad_norm": 0.011694577522575855, | |
| "learning_rate": 4.0618336886993606e-05, | |
| "loss": 0.0078, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9488272921108742, | |
| "grad_norm": 0.2500549852848053, | |
| "learning_rate": 4.051172707889126e-05, | |
| "loss": 0.0076, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9594882729211087, | |
| "grad_norm": 4.215610504150391, | |
| "learning_rate": 4.0405117270788915e-05, | |
| "loss": 0.0377, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9701492537313433, | |
| "grad_norm": 2.921285390853882, | |
| "learning_rate": 4.029850746268657e-05, | |
| "loss": 0.0395, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9808102345415778, | |
| "grad_norm": 6.361176013946533, | |
| "learning_rate": 4.0191897654584225e-05, | |
| "loss": 0.0361, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9914712153518124, | |
| "grad_norm": 4.146510601043701, | |
| "learning_rate": 4.008528784648188e-05, | |
| "loss": 0.0888, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9682666666666667, | |
| "eval_loss": 0.12556278705596924, | |
| "eval_runtime": 18.1564, | |
| "eval_samples_per_second": 206.538, | |
| "eval_steps_per_second": 6.499, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.0021321961620469, | |
| "grad_norm": 0.01982228457927704, | |
| "learning_rate": 3.997867803837953e-05, | |
| "loss": 0.0171, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0127931769722816, | |
| "grad_norm": 0.10318928211927414, | |
| "learning_rate": 3.987206823027719e-05, | |
| "loss": 0.086, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.023454157782516, | |
| "grad_norm": 0.0037166441325098276, | |
| "learning_rate": 3.976545842217484e-05, | |
| "loss": 0.0152, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0341151385927505, | |
| "grad_norm": 0.2977741062641144, | |
| "learning_rate": 3.96588486140725e-05, | |
| "loss": 0.0264, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.044776119402985, | |
| "grad_norm": 0.01882185786962509, | |
| "learning_rate": 3.9552238805970146e-05, | |
| "loss": 0.022, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0554371002132197, | |
| "grad_norm": 0.3763905167579651, | |
| "learning_rate": 3.944562899786781e-05, | |
| "loss": 0.025, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.0660980810234542, | |
| "grad_norm": 0.05068141222000122, | |
| "learning_rate": 3.9339019189765456e-05, | |
| "loss": 0.0025, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0767590618336886, | |
| "grad_norm": 0.014066731557250023, | |
| "learning_rate": 3.923240938166312e-05, | |
| "loss": 0.0067, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0874200426439233, | |
| "grad_norm": 0.05760305002331734, | |
| "learning_rate": 3.9125799573560765e-05, | |
| "loss": 0.014, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0980810234541578, | |
| "grad_norm": 0.48546120524406433, | |
| "learning_rate": 3.901918976545843e-05, | |
| "loss": 0.0422, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.1087420042643923, | |
| "grad_norm": 0.12980687618255615, | |
| "learning_rate": 3.8912579957356075e-05, | |
| "loss": 0.004, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1194029850746268, | |
| "grad_norm": 0.019840959459543228, | |
| "learning_rate": 3.8805970149253736e-05, | |
| "loss": 0.0018, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.1300639658848615, | |
| "grad_norm": 0.03178727626800537, | |
| "learning_rate": 3.8699360341151384e-05, | |
| "loss": 0.0012, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.140724946695096, | |
| "grad_norm": 0.024738993495702744, | |
| "learning_rate": 3.8592750533049046e-05, | |
| "loss": 0.0007, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1513859275053304, | |
| "grad_norm": 0.04438888281583786, | |
| "learning_rate": 3.8486140724946694e-05, | |
| "loss": 0.0385, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.1620469083155651, | |
| "grad_norm": 0.08628378063440323, | |
| "learning_rate": 3.8379530916844355e-05, | |
| "loss": 0.0037, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.1727078891257996, | |
| "grad_norm": 0.03334299102425575, | |
| "learning_rate": 3.8272921108742e-05, | |
| "loss": 0.0265, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.183368869936034, | |
| "grad_norm": 0.16531717777252197, | |
| "learning_rate": 3.8166311300639665e-05, | |
| "loss": 0.0075, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1940298507462686, | |
| "grad_norm": 0.006571681704372168, | |
| "learning_rate": 3.805970149253731e-05, | |
| "loss": 0.0006, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.2046908315565032, | |
| "grad_norm": 4.12578010559082, | |
| "learning_rate": 3.7953091684434974e-05, | |
| "loss": 0.0178, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.2153518123667377, | |
| "grad_norm": 0.0815979465842247, | |
| "learning_rate": 3.784648187633262e-05, | |
| "loss": 0.0037, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.2260127931769722, | |
| "grad_norm": 3.699103593826294, | |
| "learning_rate": 3.7739872068230284e-05, | |
| "loss": 0.0055, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.236673773987207, | |
| "grad_norm": 0.012649956159293652, | |
| "learning_rate": 3.763326226012793e-05, | |
| "loss": 0.0021, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2473347547974414, | |
| "grad_norm": 4.331559658050537, | |
| "learning_rate": 3.752665245202559e-05, | |
| "loss": 0.0329, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.2579957356076759, | |
| "grad_norm": 0.42186281085014343, | |
| "learning_rate": 3.742004264392324e-05, | |
| "loss": 0.0013, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2686567164179103, | |
| "grad_norm": 0.011500491760671139, | |
| "learning_rate": 3.73134328358209e-05, | |
| "loss": 0.001, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.279317697228145, | |
| "grad_norm": 0.042148228734731674, | |
| "learning_rate": 3.720682302771855e-05, | |
| "loss": 0.0414, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2899786780383795, | |
| "grad_norm": 0.026529042050242424, | |
| "learning_rate": 3.710021321961621e-05, | |
| "loss": 0.0092, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.3006396588486142, | |
| "grad_norm": 2.9121885299682617, | |
| "learning_rate": 3.699360341151386e-05, | |
| "loss": 0.0701, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3113006396588487, | |
| "grad_norm": 0.00463957991451025, | |
| "learning_rate": 3.6886993603411515e-05, | |
| "loss": 0.0016, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.3219616204690832, | |
| "grad_norm": 0.010834756307303905, | |
| "learning_rate": 3.678038379530917e-05, | |
| "loss": 0.0024, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.3326226012793176, | |
| "grad_norm": 0.03318125382065773, | |
| "learning_rate": 3.6673773987206824e-05, | |
| "loss": 0.0016, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.3432835820895521, | |
| "grad_norm": 0.02642156556248665, | |
| "learning_rate": 3.656716417910448e-05, | |
| "loss": 0.0232, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.3539445628997868, | |
| "grad_norm": 0.0203594621270895, | |
| "learning_rate": 3.6460554371002134e-05, | |
| "loss": 0.019, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.3646055437100213, | |
| "grad_norm": 0.1975228190422058, | |
| "learning_rate": 3.635394456289979e-05, | |
| "loss": 0.0345, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.375266524520256, | |
| "grad_norm": 1.1461132764816284, | |
| "learning_rate": 3.624733475479744e-05, | |
| "loss": 0.0025, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.3859275053304905, | |
| "grad_norm": 3.93400502204895, | |
| "learning_rate": 3.61407249466951e-05, | |
| "loss": 0.028, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.396588486140725, | |
| "grad_norm": 0.2767792344093323, | |
| "learning_rate": 3.603411513859275e-05, | |
| "loss": 0.0029, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.4072494669509594, | |
| "grad_norm": 0.02454341948032379, | |
| "learning_rate": 3.592750533049041e-05, | |
| "loss": 0.0079, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.417910447761194, | |
| "grad_norm": 0.5461854338645935, | |
| "learning_rate": 3.582089552238806e-05, | |
| "loss": 0.0021, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.007815384306013584, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.0167, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.439232409381663, | |
| "grad_norm": 0.030073419213294983, | |
| "learning_rate": 3.560767590618337e-05, | |
| "loss": 0.0174, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.4498933901918978, | |
| "grad_norm": 0.04773944616317749, | |
| "learning_rate": 3.5501066098081026e-05, | |
| "loss": 0.007, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4605543710021323, | |
| "grad_norm": 0.012186390347778797, | |
| "learning_rate": 3.539445628997868e-05, | |
| "loss": 0.0007, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.4712153518123667, | |
| "grad_norm": 0.02012632042169571, | |
| "learning_rate": 3.5287846481876336e-05, | |
| "loss": 0.0005, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4818763326226012, | |
| "grad_norm": 0.006190048530697823, | |
| "learning_rate": 3.518123667377399e-05, | |
| "loss": 0.0037, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "grad_norm": 0.00781849306076765, | |
| "learning_rate": 3.5074626865671645e-05, | |
| "loss": 0.0006, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5031982942430704, | |
| "grad_norm": 0.012595439329743385, | |
| "learning_rate": 3.496801705756929e-05, | |
| "loss": 0.0007, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.5138592750533049, | |
| "grad_norm": 5.092950820922852, | |
| "learning_rate": 3.4861407249466955e-05, | |
| "loss": 0.0254, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.5245202558635396, | |
| "grad_norm": 0.013176420703530312, | |
| "learning_rate": 3.47547974413646e-05, | |
| "loss": 0.0005, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.535181236673774, | |
| "grad_norm": 0.002772344509139657, | |
| "learning_rate": 3.4648187633262264e-05, | |
| "loss": 0.0011, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.5458422174840085, | |
| "grad_norm": 2.556558847427368, | |
| "learning_rate": 3.454157782515991e-05, | |
| "loss": 0.0056, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.556503198294243, | |
| "grad_norm": 0.0028016751166433096, | |
| "learning_rate": 3.4434968017057574e-05, | |
| "loss": 0.0006, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5671641791044775, | |
| "grad_norm": 0.002541080117225647, | |
| "learning_rate": 3.432835820895522e-05, | |
| "loss": 0.0005, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.5778251599147122, | |
| "grad_norm": 0.01701749488711357, | |
| "learning_rate": 3.422174840085288e-05, | |
| "loss": 0.0005, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5884861407249466, | |
| "grad_norm": 0.0032772503327578306, | |
| "learning_rate": 3.411513859275053e-05, | |
| "loss": 0.0015, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.5991471215351813, | |
| "grad_norm": 0.4739973247051239, | |
| "learning_rate": 3.400852878464819e-05, | |
| "loss": 0.0255, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.6098081023454158, | |
| "grad_norm": 0.031100528314709663, | |
| "learning_rate": 3.390191897654584e-05, | |
| "loss": 0.0094, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.6204690831556503, | |
| "grad_norm": 0.016540905460715294, | |
| "learning_rate": 3.37953091684435e-05, | |
| "loss": 0.0103, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.6311300639658848, | |
| "grad_norm": 0.0029734622221440077, | |
| "learning_rate": 3.368869936034115e-05, | |
| "loss": 0.0003, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.6417910447761193, | |
| "grad_norm": 3.637174367904663, | |
| "learning_rate": 3.358208955223881e-05, | |
| "loss": 0.0155, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.652452025586354, | |
| "grad_norm": 0.002992335008457303, | |
| "learning_rate": 3.347547974413646e-05, | |
| "loss": 0.0082, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.6631130063965884, | |
| "grad_norm": 0.0026899639051407576, | |
| "learning_rate": 3.336886993603412e-05, | |
| "loss": 0.0015, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.6737739872068231, | |
| "grad_norm": 0.02699490822851658, | |
| "learning_rate": 3.326226012793177e-05, | |
| "loss": 0.0007, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.6844349680170576, | |
| "grad_norm": 0.007621058728545904, | |
| "learning_rate": 3.3155650319829424e-05, | |
| "loss": 0.0246, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.695095948827292, | |
| "grad_norm": 0.012562038376927376, | |
| "learning_rate": 3.304904051172708e-05, | |
| "loss": 0.0031, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.7057569296375266, | |
| "grad_norm": 0.002219436690211296, | |
| "learning_rate": 3.294243070362473e-05, | |
| "loss": 0.0005, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.716417910447761, | |
| "grad_norm": 0.09879221022129059, | |
| "learning_rate": 3.283582089552239e-05, | |
| "loss": 0.0015, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.7270788912579957, | |
| "grad_norm": 0.07470264285802841, | |
| "learning_rate": 3.272921108742004e-05, | |
| "loss": 0.0136, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.7377398720682304, | |
| "grad_norm": 0.002869713120162487, | |
| "learning_rate": 3.26226012793177e-05, | |
| "loss": 0.0006, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.748400852878465, | |
| "grad_norm": 3.1059861183166504, | |
| "learning_rate": 3.251599147121535e-05, | |
| "loss": 0.0351, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7590618336886994, | |
| "grad_norm": 2.2680935859680176, | |
| "learning_rate": 3.240938166311301e-05, | |
| "loss": 0.0479, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.7697228144989339, | |
| "grad_norm": 0.0024357482325285673, | |
| "learning_rate": 3.230277185501066e-05, | |
| "loss": 0.0373, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.7803837953091683, | |
| "grad_norm": 0.010801059193909168, | |
| "learning_rate": 3.2196162046908317e-05, | |
| "loss": 0.005, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.7910447761194028, | |
| "grad_norm": 0.0023505063727498055, | |
| "learning_rate": 3.208955223880597e-05, | |
| "loss": 0.0068, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.8017057569296375, | |
| "grad_norm": 0.004954994190484285, | |
| "learning_rate": 3.1982942430703626e-05, | |
| "loss": 0.0006, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.8123667377398722, | |
| "grad_norm": 0.005325252655893564, | |
| "learning_rate": 3.187633262260128e-05, | |
| "loss": 0.0013, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.8230277185501067, | |
| "grad_norm": 0.0022678172681480646, | |
| "learning_rate": 3.1769722814498935e-05, | |
| "loss": 0.0451, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.8336886993603412, | |
| "grad_norm": 0.05244056135416031, | |
| "learning_rate": 3.166311300639659e-05, | |
| "loss": 0.0004, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.8443496801705757, | |
| "grad_norm": 6.828389644622803, | |
| "learning_rate": 3.1556503198294245e-05, | |
| "loss": 0.0543, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.8550106609808101, | |
| "grad_norm": 0.0025951359421014786, | |
| "learning_rate": 3.14498933901919e-05, | |
| "loss": 0.0009, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.8656716417910446, | |
| "grad_norm": 0.025431588292121887, | |
| "learning_rate": 3.1343283582089554e-05, | |
| "loss": 0.0005, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.8763326226012793, | |
| "grad_norm": 0.5160558819770813, | |
| "learning_rate": 3.123667377398721e-05, | |
| "loss": 0.0451, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.886993603411514, | |
| "grad_norm": 0.005602386314421892, | |
| "learning_rate": 3.1130063965884864e-05, | |
| "loss": 0.0013, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.8976545842217485, | |
| "grad_norm": 0.008902816101908684, | |
| "learning_rate": 3.102345415778252e-05, | |
| "loss": 0.0005, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.908315565031983, | |
| "grad_norm": 0.008681601844727993, | |
| "learning_rate": 3.0916844349680173e-05, | |
| "loss": 0.0163, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.9189765458422174, | |
| "grad_norm": 0.0028283647261559963, | |
| "learning_rate": 3.081023454157783e-05, | |
| "loss": 0.0006, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.929637526652452, | |
| "grad_norm": 0.013242118060588837, | |
| "learning_rate": 3.070362473347548e-05, | |
| "loss": 0.0009, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.9402985074626866, | |
| "grad_norm": 3.295297145843506, | |
| "learning_rate": 3.059701492537314e-05, | |
| "loss": 0.0144, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.950959488272921, | |
| "grad_norm": 0.005470735486596823, | |
| "learning_rate": 3.0490405117270792e-05, | |
| "loss": 0.0003, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.9616204690831558, | |
| "grad_norm": 0.0032625121530145407, | |
| "learning_rate": 3.0383795309168444e-05, | |
| "loss": 0.0003, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.9722814498933903, | |
| "grad_norm": 0.003790609771385789, | |
| "learning_rate": 3.0277185501066102e-05, | |
| "loss": 0.0666, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.9829424307036247, | |
| "grad_norm": 0.0024929724168032408, | |
| "learning_rate": 3.0170575692963753e-05, | |
| "loss": 0.0003, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.9936034115138592, | |
| "grad_norm": 0.05972645431756973, | |
| "learning_rate": 3.006396588486141e-05, | |
| "loss": 0.0005, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9829333333333333, | |
| "eval_loss": 0.059450652450323105, | |
| "eval_runtime": 17.4029, | |
| "eval_samples_per_second": 215.482, | |
| "eval_steps_per_second": 6.78, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 2.0042643923240937, | |
| "grad_norm": 0.011405990459024906, | |
| "learning_rate": 2.9957356076759063e-05, | |
| "loss": 0.0008, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.014925373134328, | |
| "grad_norm": 0.008791730739176273, | |
| "learning_rate": 2.9850746268656714e-05, | |
| "loss": 0.0005, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.025586353944563, | |
| "grad_norm": 0.21560530364513397, | |
| "learning_rate": 2.9744136460554372e-05, | |
| "loss": 0.0009, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.0362473347547976, | |
| "grad_norm": 0.011877737939357758, | |
| "learning_rate": 2.9637526652452023e-05, | |
| "loss": 0.0004, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.046908315565032, | |
| "grad_norm": 0.0031844661571085453, | |
| "learning_rate": 2.953091684434968e-05, | |
| "loss": 0.0004, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.0575692963752665, | |
| "grad_norm": 0.01256948709487915, | |
| "learning_rate": 2.9424307036247333e-05, | |
| "loss": 0.0004, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.068230277185501, | |
| "grad_norm": 0.003664059331640601, | |
| "learning_rate": 2.931769722814499e-05, | |
| "loss": 0.0005, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.0788912579957355, | |
| "grad_norm": 0.009233221411705017, | |
| "learning_rate": 2.9211087420042642e-05, | |
| "loss": 0.0005, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.08955223880597, | |
| "grad_norm": 0.002802032744511962, | |
| "learning_rate": 2.91044776119403e-05, | |
| "loss": 0.0004, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.100213219616205, | |
| "grad_norm": 0.0022706035524606705, | |
| "learning_rate": 2.8997867803837952e-05, | |
| "loss": 0.0007, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.1108742004264394, | |
| "grad_norm": 0.002773319836705923, | |
| "learning_rate": 2.889125799573561e-05, | |
| "loss": 0.0006, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.121535181236674, | |
| "grad_norm": 0.0067685265094041824, | |
| "learning_rate": 2.878464818763326e-05, | |
| "loss": 0.0003, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.1321961620469083, | |
| "grad_norm": 0.014644935727119446, | |
| "learning_rate": 2.867803837953092e-05, | |
| "loss": 0.0005, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.003831750713288784, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.0004, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.1535181236673773, | |
| "grad_norm": 0.013705959543585777, | |
| "learning_rate": 2.846481876332623e-05, | |
| "loss": 0.0026, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.1641791044776117, | |
| "grad_norm": 0.0025688319001346827, | |
| "learning_rate": 2.835820895522388e-05, | |
| "loss": 0.0004, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.1748400852878467, | |
| "grad_norm": 0.2528219223022461, | |
| "learning_rate": 2.825159914712154e-05, | |
| "loss": 0.0014, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.185501066098081, | |
| "grad_norm": 0.0019451375119388103, | |
| "learning_rate": 2.814498933901919e-05, | |
| "loss": 0.0005, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.1961620469083156, | |
| "grad_norm": 0.0054586827754974365, | |
| "learning_rate": 2.8038379530916848e-05, | |
| "loss": 0.0003, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.20682302771855, | |
| "grad_norm": 0.0018325847340747714, | |
| "learning_rate": 2.79317697228145e-05, | |
| "loss": 0.0031, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.2174840085287846, | |
| "grad_norm": 0.030601533129811287, | |
| "learning_rate": 2.7825159914712157e-05, | |
| "loss": 0.0004, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.228144989339019, | |
| "grad_norm": 0.002786671044304967, | |
| "learning_rate": 2.771855010660981e-05, | |
| "loss": 0.0065, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.2388059701492535, | |
| "grad_norm": 0.0025517321191728115, | |
| "learning_rate": 2.7611940298507467e-05, | |
| "loss": 0.0002, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.2494669509594885, | |
| "grad_norm": 0.0022363627795130014, | |
| "learning_rate": 2.7505330490405118e-05, | |
| "loss": 0.0003, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.260127931769723, | |
| "grad_norm": 0.002015487290918827, | |
| "learning_rate": 2.7398720682302776e-05, | |
| "loss": 0.0002, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.2707889125799574, | |
| "grad_norm": 0.002033746801316738, | |
| "learning_rate": 2.7292110874200428e-05, | |
| "loss": 0.0003, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.281449893390192, | |
| "grad_norm": 0.23982590436935425, | |
| "learning_rate": 2.7185501066098086e-05, | |
| "loss": 0.0007, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.2921108742004264, | |
| "grad_norm": 0.0019197396468371153, | |
| "learning_rate": 2.7078891257995737e-05, | |
| "loss": 0.0111, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.302771855010661, | |
| "grad_norm": 0.001743512460961938, | |
| "learning_rate": 2.6972281449893395e-05, | |
| "loss": 0.0004, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.3134328358208958, | |
| "grad_norm": 0.011543733067810535, | |
| "learning_rate": 2.6865671641791047e-05, | |
| "loss": 0.0002, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.3240938166311302, | |
| "grad_norm": 0.001960090361535549, | |
| "learning_rate": 2.6759061833688705e-05, | |
| "loss": 0.0002, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.3347547974413647, | |
| "grad_norm": 0.0021510159131139517, | |
| "learning_rate": 2.6652452025586356e-05, | |
| "loss": 0.0003, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.345415778251599, | |
| "grad_norm": 0.001994698541238904, | |
| "learning_rate": 2.6545842217484007e-05, | |
| "loss": 0.0051, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.3560767590618337, | |
| "grad_norm": 0.0020854491740465164, | |
| "learning_rate": 2.6439232409381666e-05, | |
| "loss": 0.0002, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.366737739872068, | |
| "grad_norm": 0.04646117985248566, | |
| "learning_rate": 2.6332622601279317e-05, | |
| "loss": 0.0033, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.3773987206823026, | |
| "grad_norm": 0.0025137723423540592, | |
| "learning_rate": 2.6226012793176975e-05, | |
| "loss": 0.0002, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.388059701492537, | |
| "grad_norm": 2.028430461883545, | |
| "learning_rate": 2.6119402985074626e-05, | |
| "loss": 0.0008, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.398720682302772, | |
| "grad_norm": 0.00535277696326375, | |
| "learning_rate": 2.6012793176972285e-05, | |
| "loss": 0.002, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.4093816631130065, | |
| "grad_norm": 15.36055850982666, | |
| "learning_rate": 2.5906183368869936e-05, | |
| "loss": 0.0089, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.420042643923241, | |
| "grad_norm": 0.001888532075099647, | |
| "learning_rate": 2.5799573560767594e-05, | |
| "loss": 0.0005, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.4307036247334755, | |
| "grad_norm": 0.002072590170428157, | |
| "learning_rate": 2.5692963752665245e-05, | |
| "loss": 0.0017, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.44136460554371, | |
| "grad_norm": 0.006232311949133873, | |
| "learning_rate": 2.5586353944562904e-05, | |
| "loss": 0.0004, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.4520255863539444, | |
| "grad_norm": 0.03075665421783924, | |
| "learning_rate": 2.5479744136460555e-05, | |
| "loss": 0.0005, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.4626865671641793, | |
| "grad_norm": 0.0019579094368964434, | |
| "learning_rate": 2.537313432835821e-05, | |
| "loss": 0.0002, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.473347547974414, | |
| "grad_norm": 0.00516589917242527, | |
| "learning_rate": 2.5266524520255864e-05, | |
| "loss": 0.0129, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.4840085287846483, | |
| "grad_norm": 0.0019142766250297427, | |
| "learning_rate": 2.515991471215352e-05, | |
| "loss": 0.002, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.4946695095948828, | |
| "grad_norm": 0.001729660900309682, | |
| "learning_rate": 2.5053304904051174e-05, | |
| "loss": 0.0002, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.5053304904051172, | |
| "grad_norm": 0.004812856670469046, | |
| "learning_rate": 2.494669509594883e-05, | |
| "loss": 0.0124, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.5159914712153517, | |
| "grad_norm": 0.002255210420116782, | |
| "learning_rate": 2.4840085287846483e-05, | |
| "loss": 0.0018, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.526652452025586, | |
| "grad_norm": 0.0042625595815479755, | |
| "learning_rate": 2.4733475479744138e-05, | |
| "loss": 0.0022, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.5373134328358207, | |
| "grad_norm": 0.003317923052236438, | |
| "learning_rate": 2.4626865671641793e-05, | |
| "loss": 0.0002, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.5479744136460556, | |
| "grad_norm": 0.0021533577237278223, | |
| "learning_rate": 2.4520255863539444e-05, | |
| "loss": 0.0006, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.55863539445629, | |
| "grad_norm": 0.0016170875169336796, | |
| "learning_rate": 2.44136460554371e-05, | |
| "loss": 0.0002, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.5692963752665245, | |
| "grad_norm": 0.0032187297474592924, | |
| "learning_rate": 2.4307036247334754e-05, | |
| "loss": 0.0003, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.579957356076759, | |
| "grad_norm": 0.0017038070363923907, | |
| "learning_rate": 2.420042643923241e-05, | |
| "loss": 0.0007, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.5906183368869935, | |
| "grad_norm": 0.0022185908164829016, | |
| "learning_rate": 2.4093816631130063e-05, | |
| "loss": 0.0003, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.6012793176972284, | |
| "grad_norm": 0.011953674256801605, | |
| "learning_rate": 2.3987206823027718e-05, | |
| "loss": 0.0004, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.611940298507463, | |
| "grad_norm": 0.0017118758987635374, | |
| "learning_rate": 2.3880597014925373e-05, | |
| "loss": 0.0002, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.6226012793176974, | |
| "grad_norm": 0.002300629625096917, | |
| "learning_rate": 2.3773987206823027e-05, | |
| "loss": 0.0002, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.633262260127932, | |
| "grad_norm": 0.0016920581692829728, | |
| "learning_rate": 2.3667377398720682e-05, | |
| "loss": 0.0062, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.6439232409381663, | |
| "grad_norm": 0.24134622514247894, | |
| "learning_rate": 2.3560767590618337e-05, | |
| "loss": 0.0056, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.654584221748401, | |
| "grad_norm": 0.0015582832274958491, | |
| "learning_rate": 2.345415778251599e-05, | |
| "loss": 0.0012, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.6652452025586353, | |
| "grad_norm": 0.022430991753935814, | |
| "learning_rate": 2.3347547974413646e-05, | |
| "loss": 0.0003, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.6759061833688698, | |
| "grad_norm": 0.13407553732395172, | |
| "learning_rate": 2.32409381663113e-05, | |
| "loss": 0.0003, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.6865671641791042, | |
| "grad_norm": 0.001633663778193295, | |
| "learning_rate": 2.3134328358208956e-05, | |
| "loss": 0.0003, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.697228144989339, | |
| "grad_norm": 0.002987906336784363, | |
| "learning_rate": 2.302771855010661e-05, | |
| "loss": 0.0002, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.7078891257995736, | |
| "grad_norm": 0.001594817265868187, | |
| "learning_rate": 2.2921108742004265e-05, | |
| "loss": 0.0002, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.718550106609808, | |
| "grad_norm": 0.07639887928962708, | |
| "learning_rate": 2.281449893390192e-05, | |
| "loss": 0.0003, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.7292110874200426, | |
| "grad_norm": 0.002753237960860133, | |
| "learning_rate": 2.2707889125799575e-05, | |
| "loss": 0.0003, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.739872068230277, | |
| "grad_norm": 0.00843032356351614, | |
| "learning_rate": 2.260127931769723e-05, | |
| "loss": 0.0003, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.750533049040512, | |
| "grad_norm": 0.0017697943840175867, | |
| "learning_rate": 2.2494669509594884e-05, | |
| "loss": 0.0005, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.7611940298507465, | |
| "grad_norm": 0.012765592895448208, | |
| "learning_rate": 2.238805970149254e-05, | |
| "loss": 0.0003, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.771855010660981, | |
| "grad_norm": 0.001516074757091701, | |
| "learning_rate": 2.2281449893390194e-05, | |
| "loss": 0.0002, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.7825159914712154, | |
| "grad_norm": 0.0016697804676368833, | |
| "learning_rate": 2.217484008528785e-05, | |
| "loss": 0.0002, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.79317697228145, | |
| "grad_norm": 0.0018960988381877542, | |
| "learning_rate": 2.2068230277185503e-05, | |
| "loss": 0.0002, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.8038379530916844, | |
| "grad_norm": 0.002706008730456233, | |
| "learning_rate": 2.1961620469083158e-05, | |
| "loss": 0.0002, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.814498933901919, | |
| "grad_norm": 0.0015761232934892178, | |
| "learning_rate": 2.1855010660980813e-05, | |
| "loss": 0.0002, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.8251599147121533, | |
| "grad_norm": 0.04365153983235359, | |
| "learning_rate": 2.1748400852878467e-05, | |
| "loss": 0.0054, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.835820895522388, | |
| "grad_norm": 0.0030039746779948473, | |
| "learning_rate": 2.164179104477612e-05, | |
| "loss": 0.0003, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.8464818763326227, | |
| "grad_norm": 0.0018323756521567702, | |
| "learning_rate": 2.1535181236673773e-05, | |
| "loss": 0.0007, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.0018739423248916864, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.0002, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.8678038379530917, | |
| "grad_norm": 0.0016587284626439214, | |
| "learning_rate": 2.1321961620469083e-05, | |
| "loss": 0.0002, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.878464818763326, | |
| "grad_norm": 0.0021149320527911186, | |
| "learning_rate": 2.1215351812366738e-05, | |
| "loss": 0.0002, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.8891257995735606, | |
| "grad_norm": 0.001595294801518321, | |
| "learning_rate": 2.1108742004264392e-05, | |
| "loss": 0.0002, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.8997867803837956, | |
| "grad_norm": 0.003828309243544936, | |
| "learning_rate": 2.1002132196162047e-05, | |
| "loss": 0.0089, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.91044776119403, | |
| "grad_norm": 0.004440088756382465, | |
| "learning_rate": 2.0895522388059702e-05, | |
| "loss": 0.0003, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.9211087420042645, | |
| "grad_norm": 0.0017082487465813756, | |
| "learning_rate": 2.0788912579957357e-05, | |
| "loss": 0.0089, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.931769722814499, | |
| "grad_norm": 0.17494313418865204, | |
| "learning_rate": 2.068230277185501e-05, | |
| "loss": 0.0079, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.9424307036247335, | |
| "grad_norm": 0.001594117609784007, | |
| "learning_rate": 2.0575692963752666e-05, | |
| "loss": 0.0002, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.953091684434968, | |
| "grad_norm": 0.008118892088532448, | |
| "learning_rate": 2.046908315565032e-05, | |
| "loss": 0.0004, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.9637526652452024, | |
| "grad_norm": 0.0024735559709370136, | |
| "learning_rate": 2.0362473347547976e-05, | |
| "loss": 0.0002, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.974413646055437, | |
| "grad_norm": 0.0014542067656293511, | |
| "learning_rate": 2.025586353944563e-05, | |
| "loss": 0.0002, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "grad_norm": 0.0031547469552606344, | |
| "learning_rate": 2.0149253731343285e-05, | |
| "loss": 0.0002, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.9957356076759063, | |
| "grad_norm": 0.0015797675587236881, | |
| "learning_rate": 2.004264392324094e-05, | |
| "loss": 0.0002, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9861333333333333, | |
| "eval_loss": 0.06201761215925217, | |
| "eval_runtime": 17.375, | |
| "eval_samples_per_second": 215.827, | |
| "eval_steps_per_second": 6.791, | |
| "step": 1407 | |
| }, | |
| { | |
| "epoch": 3.0063965884861408, | |
| "grad_norm": 1.461633563041687, | |
| "learning_rate": 1.9936034115138594e-05, | |
| "loss": 0.0007, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.0170575692963753, | |
| "grad_norm": 0.008623131550848484, | |
| "learning_rate": 1.982942430703625e-05, | |
| "loss": 0.0002, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 3.0277185501066097, | |
| "grad_norm": 0.0022685376461595297, | |
| "learning_rate": 1.9722814498933904e-05, | |
| "loss": 0.0002, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.038379530916844, | |
| "grad_norm": 0.0021228427067399025, | |
| "learning_rate": 1.961620469083156e-05, | |
| "loss": 0.0002, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 3.0490405117270787, | |
| "grad_norm": 0.0026628009509295225, | |
| "learning_rate": 1.9509594882729213e-05, | |
| "loss": 0.0002, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.0597014925373136, | |
| "grad_norm": 0.0015997429145500064, | |
| "learning_rate": 1.9402985074626868e-05, | |
| "loss": 0.0065, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 3.070362473347548, | |
| "grad_norm": 0.0015492772217839956, | |
| "learning_rate": 1.9296375266524523e-05, | |
| "loss": 0.0002, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.0810234541577826, | |
| "grad_norm": 0.0014074740465730429, | |
| "learning_rate": 1.9189765458422178e-05, | |
| "loss": 0.0002, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 3.091684434968017, | |
| "grad_norm": 0.0014764177612960339, | |
| "learning_rate": 1.9083155650319832e-05, | |
| "loss": 0.0002, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.1023454157782515, | |
| "grad_norm": 0.0015944272745400667, | |
| "learning_rate": 1.8976545842217487e-05, | |
| "loss": 0.0002, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 3.113006396588486, | |
| "grad_norm": 0.002632107585668564, | |
| "learning_rate": 1.8869936034115142e-05, | |
| "loss": 0.0002, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.1236673773987205, | |
| "grad_norm": 0.0015126124490052462, | |
| "learning_rate": 1.8763326226012797e-05, | |
| "loss": 0.0002, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 3.1343283582089554, | |
| "grad_norm": 0.024501264095306396, | |
| "learning_rate": 1.865671641791045e-05, | |
| "loss": 0.0002, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.14498933901919, | |
| "grad_norm": 0.014305677264928818, | |
| "learning_rate": 1.8550106609808106e-05, | |
| "loss": 0.0002, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 3.1556503198294243, | |
| "grad_norm": 0.0020725333597511053, | |
| "learning_rate": 1.8443496801705757e-05, | |
| "loss": 0.0032, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.166311300639659, | |
| "grad_norm": 0.0038152721244841814, | |
| "learning_rate": 1.8336886993603412e-05, | |
| "loss": 0.0002, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 3.1769722814498933, | |
| "grad_norm": 0.001414337893947959, | |
| "learning_rate": 1.8230277185501067e-05, | |
| "loss": 0.0002, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.1876332622601278, | |
| "grad_norm": 0.002374679781496525, | |
| "learning_rate": 1.812366737739872e-05, | |
| "loss": 0.0002, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 3.1982942430703627, | |
| "grad_norm": 0.002602379070594907, | |
| "learning_rate": 1.8017057569296376e-05, | |
| "loss": 0.0002, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.208955223880597, | |
| "grad_norm": 0.001584042445756495, | |
| "learning_rate": 1.791044776119403e-05, | |
| "loss": 0.0003, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 3.2196162046908317, | |
| "grad_norm": 0.0019072901923209429, | |
| "learning_rate": 1.7803837953091686e-05, | |
| "loss": 0.0002, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.230277185501066, | |
| "grad_norm": 0.0014707124792039394, | |
| "learning_rate": 1.769722814498934e-05, | |
| "loss": 0.0002, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 3.2409381663113006, | |
| "grad_norm": 0.0016866616206243634, | |
| "learning_rate": 1.7590618336886995e-05, | |
| "loss": 0.0002, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.251599147121535, | |
| "grad_norm": 0.0013746339827775955, | |
| "learning_rate": 1.7484008528784647e-05, | |
| "loss": 0.0002, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 3.2622601279317696, | |
| "grad_norm": 0.0014366420218721032, | |
| "learning_rate": 1.73773987206823e-05, | |
| "loss": 0.0002, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.272921108742004, | |
| "grad_norm": 21.536617279052734, | |
| "learning_rate": 1.7270788912579956e-05, | |
| "loss": 0.0113, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 3.283582089552239, | |
| "grad_norm": 0.030154358595609665, | |
| "learning_rate": 1.716417910447761e-05, | |
| "loss": 0.0002, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.2942430703624734, | |
| "grad_norm": 0.0018184144282713532, | |
| "learning_rate": 1.7057569296375266e-05, | |
| "loss": 0.0004, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 3.304904051172708, | |
| "grad_norm": 0.0014812969602644444, | |
| "learning_rate": 1.695095948827292e-05, | |
| "loss": 0.0002, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.3155650319829424, | |
| "grad_norm": 0.0013438789173960686, | |
| "learning_rate": 1.6844349680170575e-05, | |
| "loss": 0.0005, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 3.326226012793177, | |
| "grad_norm": 0.003677763743326068, | |
| "learning_rate": 1.673773987206823e-05, | |
| "loss": 0.0008, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.3368869936034113, | |
| "grad_norm": 0.10813038796186447, | |
| "learning_rate": 1.6631130063965885e-05, | |
| "loss": 0.0003, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 3.3475479744136463, | |
| "grad_norm": 0.007127894088625908, | |
| "learning_rate": 1.652452025586354e-05, | |
| "loss": 0.0002, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.3582089552238807, | |
| "grad_norm": 0.001401298213750124, | |
| "learning_rate": 1.6417910447761194e-05, | |
| "loss": 0.0002, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 3.368869936034115, | |
| "grad_norm": 0.0013577801873907447, | |
| "learning_rate": 1.631130063965885e-05, | |
| "loss": 0.0002, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.3795309168443497, | |
| "grad_norm": 0.0017308747628703713, | |
| "learning_rate": 1.6204690831556504e-05, | |
| "loss": 0.0002, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 3.390191897654584, | |
| "grad_norm": 0.0013733214000239968, | |
| "learning_rate": 1.6098081023454158e-05, | |
| "loss": 0.0002, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 3.4008528784648187, | |
| "grad_norm": 0.011999400332570076, | |
| "learning_rate": 1.5991471215351813e-05, | |
| "loss": 0.0002, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 3.411513859275053, | |
| "grad_norm": 0.0013399006566032767, | |
| "learning_rate": 1.5884861407249468e-05, | |
| "loss": 0.0003, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.4221748400852876, | |
| "grad_norm": 0.0014706465881317854, | |
| "learning_rate": 1.5778251599147122e-05, | |
| "loss": 0.0002, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 3.4328358208955225, | |
| "grad_norm": 0.0013743231538683176, | |
| "learning_rate": 1.5671641791044777e-05, | |
| "loss": 0.0002, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 3.443496801705757, | |
| "grad_norm": 0.0013075482565909624, | |
| "learning_rate": 1.5565031982942432e-05, | |
| "loss": 0.0002, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 3.4541577825159915, | |
| "grad_norm": 0.0014712231932207942, | |
| "learning_rate": 1.5458422174840087e-05, | |
| "loss": 0.0002, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.464818763326226, | |
| "grad_norm": 0.0013278430560603738, | |
| "learning_rate": 1.535181236673774e-05, | |
| "loss": 0.0009, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 3.4754797441364604, | |
| "grad_norm": 0.0014948764583095908, | |
| "learning_rate": 1.5245202558635396e-05, | |
| "loss": 0.0002, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 3.486140724946695, | |
| "grad_norm": 0.0016076472820714116, | |
| "learning_rate": 1.5138592750533051e-05, | |
| "loss": 0.0002, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 3.49680170575693, | |
| "grad_norm": 0.046660102903842926, | |
| "learning_rate": 1.5031982942430706e-05, | |
| "loss": 0.0003, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.5074626865671643, | |
| "grad_norm": 0.001787811517715454, | |
| "learning_rate": 1.4925373134328357e-05, | |
| "loss": 0.0002, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 3.518123667377399, | |
| "grad_norm": 0.019754258915781975, | |
| "learning_rate": 1.4818763326226012e-05, | |
| "loss": 0.0002, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.5287846481876333, | |
| "grad_norm": 0.0013567762216553092, | |
| "learning_rate": 1.4712153518123666e-05, | |
| "loss": 0.0002, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 3.5394456289978677, | |
| "grad_norm": 0.0017151799984276295, | |
| "learning_rate": 1.4605543710021321e-05, | |
| "loss": 0.0002, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 3.550106609808102, | |
| "grad_norm": 0.0014122892171144485, | |
| "learning_rate": 1.4498933901918976e-05, | |
| "loss": 0.0001, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 3.5607675906183367, | |
| "grad_norm": 0.001381227164529264, | |
| "learning_rate": 1.439232409381663e-05, | |
| "loss": 0.0002, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.001501563936471939, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.0002, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 3.582089552238806, | |
| "grad_norm": 10.50568962097168, | |
| "learning_rate": 1.417910447761194e-05, | |
| "loss": 0.0155, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 3.5927505330490406, | |
| "grad_norm": 0.0035982050467282534, | |
| "learning_rate": 1.4072494669509595e-05, | |
| "loss": 0.004, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 3.603411513859275, | |
| "grad_norm": 0.0030845210421830416, | |
| "learning_rate": 1.396588486140725e-05, | |
| "loss": 0.0002, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 3.6140724946695095, | |
| "grad_norm": 0.0024272853042930365, | |
| "learning_rate": 1.3859275053304904e-05, | |
| "loss": 0.0002, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 3.624733475479744, | |
| "grad_norm": 0.001277498435229063, | |
| "learning_rate": 1.3752665245202559e-05, | |
| "loss": 0.0001, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.635394456289979, | |
| "grad_norm": 0.001351995742879808, | |
| "learning_rate": 1.3646055437100214e-05, | |
| "loss": 0.0002, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 3.6460554371002134, | |
| "grad_norm": 0.0013289814814925194, | |
| "learning_rate": 1.3539445628997869e-05, | |
| "loss": 0.0001, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 3.656716417910448, | |
| "grad_norm": 0.004144403617829084, | |
| "learning_rate": 1.3432835820895523e-05, | |
| "loss": 0.0002, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 3.6673773987206824, | |
| "grad_norm": 0.0013239245163276792, | |
| "learning_rate": 1.3326226012793178e-05, | |
| "loss": 0.0002, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.678038379530917, | |
| "grad_norm": 0.003802697407081723, | |
| "learning_rate": 1.3219616204690833e-05, | |
| "loss": 0.0002, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 3.6886993603411513, | |
| "grad_norm": 0.0013056733878329396, | |
| "learning_rate": 1.3113006396588488e-05, | |
| "loss": 0.0001, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.699360341151386, | |
| "grad_norm": 0.0013066122774034739, | |
| "learning_rate": 1.3006396588486142e-05, | |
| "loss": 0.0068, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 3.7100213219616203, | |
| "grad_norm": 0.001307385740801692, | |
| "learning_rate": 1.2899786780383797e-05, | |
| "loss": 0.0002, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.7206823027718547, | |
| "grad_norm": 0.0012692222371697426, | |
| "learning_rate": 1.2793176972281452e-05, | |
| "loss": 0.0001, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 3.7313432835820897, | |
| "grad_norm": 0.0012794553767889738, | |
| "learning_rate": 1.2686567164179105e-05, | |
| "loss": 0.0002, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.742004264392324, | |
| "grad_norm": 0.0016664585564285517, | |
| "learning_rate": 1.257995735607676e-05, | |
| "loss": 0.0001, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 3.7526652452025586, | |
| "grad_norm": 0.0013751591322943568, | |
| "learning_rate": 1.2473347547974414e-05, | |
| "loss": 0.0002, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.763326226012793, | |
| "grad_norm": 0.0017565148882567883, | |
| "learning_rate": 1.2366737739872069e-05, | |
| "loss": 0.0002, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 3.7739872068230276, | |
| "grad_norm": 0.001302393269725144, | |
| "learning_rate": 1.2260127931769722e-05, | |
| "loss": 0.0001, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.7846481876332625, | |
| "grad_norm": 0.0031294100917875767, | |
| "learning_rate": 1.2153518123667377e-05, | |
| "loss": 0.0002, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 3.795309168443497, | |
| "grad_norm": 0.002873555989935994, | |
| "learning_rate": 1.2046908315565032e-05, | |
| "loss": 0.0002, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.8059701492537314, | |
| "grad_norm": 0.0014639534056186676, | |
| "learning_rate": 1.1940298507462686e-05, | |
| "loss": 0.0001, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 3.816631130063966, | |
| "grad_norm": 0.0024766058195382357, | |
| "learning_rate": 1.1833688699360341e-05, | |
| "loss": 0.0001, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.8272921108742004, | |
| "grad_norm": 0.0016069738194346428, | |
| "learning_rate": 1.1727078891257996e-05, | |
| "loss": 0.0001, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 3.837953091684435, | |
| "grad_norm": 0.0014199650613591075, | |
| "learning_rate": 1.162046908315565e-05, | |
| "loss": 0.0001, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.8486140724946694, | |
| "grad_norm": 0.002102483995258808, | |
| "learning_rate": 1.1513859275053305e-05, | |
| "loss": 0.0039, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 3.859275053304904, | |
| "grad_norm": 0.0018299691146239638, | |
| "learning_rate": 1.140724946695096e-05, | |
| "loss": 0.0002, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.8699360341151388, | |
| "grad_norm": 0.00693817762658, | |
| "learning_rate": 1.1300639658848615e-05, | |
| "loss": 0.0002, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 3.8805970149253732, | |
| "grad_norm": 0.005339950788766146, | |
| "learning_rate": 1.119402985074627e-05, | |
| "loss": 0.0002, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.8912579957356077, | |
| "grad_norm": 0.001799010206013918, | |
| "learning_rate": 1.1087420042643924e-05, | |
| "loss": 0.0002, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 3.901918976545842, | |
| "grad_norm": 0.0032261726446449757, | |
| "learning_rate": 1.0980810234541579e-05, | |
| "loss": 0.0002, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.9125799573560767, | |
| "grad_norm": 0.01785244606435299, | |
| "learning_rate": 1.0874200426439234e-05, | |
| "loss": 0.0002, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 3.923240938166311, | |
| "grad_norm": 0.0012020582798868418, | |
| "learning_rate": 1.0767590618336887e-05, | |
| "loss": 0.0002, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.933901918976546, | |
| "grad_norm": 0.0012705688131973147, | |
| "learning_rate": 1.0660980810234541e-05, | |
| "loss": 0.0001, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 3.9445628997867805, | |
| "grad_norm": 0.0014653359539806843, | |
| "learning_rate": 1.0554371002132196e-05, | |
| "loss": 0.0002, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.955223880597015, | |
| "grad_norm": 0.0015560160391032696, | |
| "learning_rate": 1.0447761194029851e-05, | |
| "loss": 0.0002, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 3.9658848614072495, | |
| "grad_norm": 0.0036310593131929636, | |
| "learning_rate": 1.0341151385927506e-05, | |
| "loss": 0.0001, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.976545842217484, | |
| "grad_norm": 0.0012366551673039794, | |
| "learning_rate": 1.023454157782516e-05, | |
| "loss": 0.0001, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 3.9872068230277184, | |
| "grad_norm": 0.0017698772717267275, | |
| "learning_rate": 1.0127931769722815e-05, | |
| "loss": 0.0002, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.997867803837953, | |
| "grad_norm": 0.0013641887344419956, | |
| "learning_rate": 1.002132196162047e-05, | |
| "loss": 0.0001, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9861333333333333, | |
| "eval_loss": 0.06454310566186905, | |
| "eval_runtime": 17.2675, | |
| "eval_samples_per_second": 217.171, | |
| "eval_steps_per_second": 6.834, | |
| "step": 1876 | |
| }, | |
| { | |
| "epoch": 4.008528784648187, | |
| "grad_norm": 0.0012424176093190908, | |
| "learning_rate": 9.914712153518125e-06, | |
| "loss": 0.0001, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.019189765458422, | |
| "grad_norm": 0.0013428545789793134, | |
| "learning_rate": 9.80810234541578e-06, | |
| "loss": 0.0002, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 4.029850746268656, | |
| "grad_norm": 0.002022024942561984, | |
| "learning_rate": 9.701492537313434e-06, | |
| "loss": 0.0001, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.040511727078891, | |
| "grad_norm": 0.004463700577616692, | |
| "learning_rate": 9.594882729211089e-06, | |
| "loss": 0.0001, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 4.051172707889126, | |
| "grad_norm": 0.0013927206164225936, | |
| "learning_rate": 9.488272921108744e-06, | |
| "loss": 0.0002, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.061833688699361, | |
| "grad_norm": 0.001286355429328978, | |
| "learning_rate": 9.381663113006398e-06, | |
| "loss": 0.0002, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 4.072494669509595, | |
| "grad_norm": 0.004472631961107254, | |
| "learning_rate": 9.275053304904053e-06, | |
| "loss": 0.0002, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.08315565031983, | |
| "grad_norm": 0.001554334769025445, | |
| "learning_rate": 9.168443496801706e-06, | |
| "loss": 0.0001, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 4.093816631130064, | |
| "grad_norm": 0.0013858502497896552, | |
| "learning_rate": 9.06183368869936e-06, | |
| "loss": 0.0002, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.104477611940299, | |
| "grad_norm": 0.0013580528320744634, | |
| "learning_rate": 8.955223880597016e-06, | |
| "loss": 0.0001, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 4.115138592750533, | |
| "grad_norm": 0.0011945017613470554, | |
| "learning_rate": 8.84861407249467e-06, | |
| "loss": 0.0002, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.1257995735607675, | |
| "grad_norm": 0.0015058082062751055, | |
| "learning_rate": 8.742004264392323e-06, | |
| "loss": 0.0001, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 4.136460554371002, | |
| "grad_norm": 0.0018504057079553604, | |
| "learning_rate": 8.635394456289978e-06, | |
| "loss": 0.0001, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.1471215351812365, | |
| "grad_norm": 0.001987944357097149, | |
| "learning_rate": 8.528784648187633e-06, | |
| "loss": 0.0002, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 4.157782515991471, | |
| "grad_norm": 0.001231278758496046, | |
| "learning_rate": 8.422174840085288e-06, | |
| "loss": 0.0001, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.1684434968017055, | |
| "grad_norm": 0.0014915940118953586, | |
| "learning_rate": 8.315565031982942e-06, | |
| "loss": 0.0029, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 4.17910447761194, | |
| "grad_norm": 0.0017193792155012488, | |
| "learning_rate": 8.208955223880597e-06, | |
| "loss": 0.0001, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.189765458422174, | |
| "grad_norm": 0.0013483568327501416, | |
| "learning_rate": 8.102345415778252e-06, | |
| "loss": 0.0001, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 4.20042643923241, | |
| "grad_norm": 0.0011997171677649021, | |
| "learning_rate": 7.995735607675907e-06, | |
| "loss": 0.0001, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 4.211087420042644, | |
| "grad_norm": 0.0017619481077417731, | |
| "learning_rate": 7.889125799573561e-06, | |
| "loss": 0.0002, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 4.221748400852879, | |
| "grad_norm": 0.001219351775944233, | |
| "learning_rate": 7.782515991471216e-06, | |
| "loss": 0.0001, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.232409381663113, | |
| "grad_norm": 0.16717660427093506, | |
| "learning_rate": 7.67590618336887e-06, | |
| "loss": 0.0063, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 4.243070362473348, | |
| "grad_norm": 0.0017397510819137096, | |
| "learning_rate": 7.5692963752665255e-06, | |
| "loss": 0.0001, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 4.253731343283582, | |
| "grad_norm": 0.0011504489229992032, | |
| "learning_rate": 7.4626865671641785e-06, | |
| "loss": 0.0001, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 4.264392324093817, | |
| "grad_norm": 0.0012177824974060059, | |
| "learning_rate": 7.356076759061833e-06, | |
| "loss": 0.0001, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.275053304904051, | |
| "grad_norm": 0.001127021387219429, | |
| "learning_rate": 7.249466950959488e-06, | |
| "loss": 0.0002, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.004080480430275202, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.0001, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 4.29637526652452, | |
| "grad_norm": 0.0011779138585552573, | |
| "learning_rate": 7.0362473347547975e-06, | |
| "loss": 0.0002, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 4.3070362473347545, | |
| "grad_norm": 0.0012129482347518206, | |
| "learning_rate": 6.929637526652452e-06, | |
| "loss": 0.0002, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 4.317697228144989, | |
| "grad_norm": 0.08944092690944672, | |
| "learning_rate": 6.823027718550107e-06, | |
| "loss": 0.0025, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 4.3283582089552235, | |
| "grad_norm": 0.0011878637596964836, | |
| "learning_rate": 6.716417910447762e-06, | |
| "loss": 0.0001, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 4.339019189765459, | |
| "grad_norm": 0.007837583310902119, | |
| "learning_rate": 6.609808102345416e-06, | |
| "loss": 0.0002, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 4.349680170575693, | |
| "grad_norm": 0.0012950020609423518, | |
| "learning_rate": 6.503198294243071e-06, | |
| "loss": 0.0001, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 4.360341151385928, | |
| "grad_norm": 0.0011917536612600088, | |
| "learning_rate": 6.396588486140726e-06, | |
| "loss": 0.0001, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 4.371002132196162, | |
| "grad_norm": 0.0013600134989246726, | |
| "learning_rate": 6.28997867803838e-06, | |
| "loss": 0.0001, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.381663113006397, | |
| "grad_norm": 0.0013914266601204872, | |
| "learning_rate": 6.1833688699360345e-06, | |
| "loss": 0.0001, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 4.392324093816631, | |
| "grad_norm": 0.0011868373258039355, | |
| "learning_rate": 6.076759061833688e-06, | |
| "loss": 0.0001, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 4.402985074626866, | |
| "grad_norm": 0.009749215096235275, | |
| "learning_rate": 5.970149253731343e-06, | |
| "loss": 0.0002, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 4.4136460554371, | |
| "grad_norm": 0.0011316449381411076, | |
| "learning_rate": 5.863539445628998e-06, | |
| "loss": 0.0001, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 4.424307036247335, | |
| "grad_norm": 0.0033109618816524744, | |
| "learning_rate": 5.756929637526653e-06, | |
| "loss": 0.0002, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 4.434968017057569, | |
| "grad_norm": 0.0012717369245365262, | |
| "learning_rate": 5.650319829424307e-06, | |
| "loss": 0.0001, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 4.445628997867804, | |
| "grad_norm": 0.0011261467589065433, | |
| "learning_rate": 5.543710021321962e-06, | |
| "loss": 0.0001, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 4.456289978678038, | |
| "grad_norm": 0.0015071145026013255, | |
| "learning_rate": 5.437100213219617e-06, | |
| "loss": 0.0001, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 4.466950959488273, | |
| "grad_norm": 0.0013715195236727595, | |
| "learning_rate": 5.330490405117271e-06, | |
| "loss": 0.0001, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 4.477611940298507, | |
| "grad_norm": 0.0011400962248444557, | |
| "learning_rate": 5.2238805970149255e-06, | |
| "loss": 0.0001, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.4882729211087415, | |
| "grad_norm": 0.001285375445149839, | |
| "learning_rate": 5.11727078891258e-06, | |
| "loss": 0.0001, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 4.498933901918977, | |
| "grad_norm": 0.0013375984271988273, | |
| "learning_rate": 5.010660980810235e-06, | |
| "loss": 0.0001, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 4.509594882729211, | |
| "grad_norm": 0.0012032234808430076, | |
| "learning_rate": 4.90405117270789e-06, | |
| "loss": 0.0001, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 4.520255863539446, | |
| "grad_norm": 0.0013014287687838078, | |
| "learning_rate": 4.797441364605544e-06, | |
| "loss": 0.0002, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 4.53091684434968, | |
| "grad_norm": 0.0020497296936810017, | |
| "learning_rate": 4.690831556503199e-06, | |
| "loss": 0.0001, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 4.541577825159915, | |
| "grad_norm": 0.004726668354123831, | |
| "learning_rate": 4.584221748400853e-06, | |
| "loss": 0.0001, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 4.552238805970149, | |
| "grad_norm": 0.0030003113206475973, | |
| "learning_rate": 4.477611940298508e-06, | |
| "loss": 0.0002, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 4.562899786780384, | |
| "grad_norm": 0.0013835778227075934, | |
| "learning_rate": 4.371002132196162e-06, | |
| "loss": 0.0001, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 4.573560767590618, | |
| "grad_norm": 0.0012324253330007195, | |
| "learning_rate": 4.264392324093816e-06, | |
| "loss": 0.0001, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 4.584221748400853, | |
| "grad_norm": 0.001273977686651051, | |
| "learning_rate": 4.157782515991471e-06, | |
| "loss": 0.0001, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 4.594882729211087, | |
| "grad_norm": 0.0011130374623462558, | |
| "learning_rate": 4.051172707889126e-06, | |
| "loss": 0.0001, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 4.605543710021322, | |
| "grad_norm": 0.001868380350060761, | |
| "learning_rate": 3.944562899786781e-06, | |
| "loss": 0.0002, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 4.616204690831556, | |
| "grad_norm": 0.0011330582201480865, | |
| "learning_rate": 3.837953091684435e-06, | |
| "loss": 0.0001, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 4.6268656716417915, | |
| "grad_norm": 0.0012200495693832636, | |
| "learning_rate": 3.7313432835820893e-06, | |
| "loss": 0.0001, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 4.637526652452026, | |
| "grad_norm": 0.001197567442432046, | |
| "learning_rate": 3.624733475479744e-06, | |
| "loss": 0.0001, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 4.6481876332622605, | |
| "grad_norm": 0.0012370103504508734, | |
| "learning_rate": 3.5181236673773987e-06, | |
| "loss": 0.0001, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 4.658848614072495, | |
| "grad_norm": 0.001441104570403695, | |
| "learning_rate": 3.4115138592750535e-06, | |
| "loss": 0.0004, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 4.669509594882729, | |
| "grad_norm": 0.005181392654776573, | |
| "learning_rate": 3.304904051172708e-06, | |
| "loss": 0.0002, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 4.680170575692964, | |
| "grad_norm": 0.0030894805677235126, | |
| "learning_rate": 3.198294243070363e-06, | |
| "loss": 0.0001, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 4.690831556503198, | |
| "grad_norm": 0.0029679632280021906, | |
| "learning_rate": 3.0916844349680173e-06, | |
| "loss": 0.0001, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.701492537313433, | |
| "grad_norm": 0.016101934015750885, | |
| "learning_rate": 2.9850746268656716e-06, | |
| "loss": 0.0001, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 4.712153518123667, | |
| "grad_norm": 0.0016056873137131333, | |
| "learning_rate": 2.8784648187633263e-06, | |
| "loss": 0.0001, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 4.722814498933902, | |
| "grad_norm": 0.3704182505607605, | |
| "learning_rate": 2.771855010660981e-06, | |
| "loss": 0.0054, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 4.733475479744136, | |
| "grad_norm": 0.0011979964328929782, | |
| "learning_rate": 2.6652452025586354e-06, | |
| "loss": 0.0001, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 4.744136460554371, | |
| "grad_norm": 0.0012237406335771084, | |
| "learning_rate": 2.55863539445629e-06, | |
| "loss": 0.0001, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 4.754797441364605, | |
| "grad_norm": 0.0016670930199325085, | |
| "learning_rate": 2.452025586353945e-06, | |
| "loss": 0.0001, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 4.76545842217484, | |
| "grad_norm": 0.0016099626664072275, | |
| "learning_rate": 2.3454157782515996e-06, | |
| "loss": 0.0001, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 4.776119402985074, | |
| "grad_norm": 0.0012452139053493738, | |
| "learning_rate": 2.238805970149254e-06, | |
| "loss": 0.0001, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 4.786780383795309, | |
| "grad_norm": 0.0012234573950991035, | |
| "learning_rate": 2.132196162046908e-06, | |
| "loss": 0.0001, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 4.797441364605544, | |
| "grad_norm": 0.001381521113216877, | |
| "learning_rate": 2.025586353944563e-06, | |
| "loss": 0.0001, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 4.8081023454157785, | |
| "grad_norm": 0.0011367109837010503, | |
| "learning_rate": 1.9189765458422177e-06, | |
| "loss": 0.0001, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 4.818763326226013, | |
| "grad_norm": 0.0013733627274632454, | |
| "learning_rate": 1.812366737739872e-06, | |
| "loss": 0.0001, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 4.8294243070362475, | |
| "grad_norm": 0.0011499322718009353, | |
| "learning_rate": 1.7057569296375267e-06, | |
| "loss": 0.0001, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 4.840085287846482, | |
| "grad_norm": 0.0013391554821282625, | |
| "learning_rate": 1.5991471215351815e-06, | |
| "loss": 0.0001, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 4.850746268656716, | |
| "grad_norm": 0.0013704358134418726, | |
| "learning_rate": 1.4925373134328358e-06, | |
| "loss": 0.0001, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 4.861407249466951, | |
| "grad_norm": 0.0012855156091973186, | |
| "learning_rate": 1.3859275053304905e-06, | |
| "loss": 0.0001, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 4.872068230277185, | |
| "grad_norm": 0.0011203851317986846, | |
| "learning_rate": 1.279317697228145e-06, | |
| "loss": 0.0001, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 4.88272921108742, | |
| "grad_norm": 0.0011925158323720098, | |
| "learning_rate": 1.1727078891257998e-06, | |
| "loss": 0.0001, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 4.893390191897654, | |
| "grad_norm": 0.0013797224964946508, | |
| "learning_rate": 1.066098081023454e-06, | |
| "loss": 0.0024, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 4.904051172707889, | |
| "grad_norm": 0.001322761527262628, | |
| "learning_rate": 9.594882729211088e-07, | |
| "loss": 0.0001, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.914712153518123, | |
| "grad_norm": 0.0011309119872748852, | |
| "learning_rate": 8.528784648187634e-07, | |
| "loss": 0.0001, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 4.925373134328359, | |
| "grad_norm": 0.0011366803664714098, | |
| "learning_rate": 7.462686567164179e-07, | |
| "loss": 0.0001, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 4.936034115138593, | |
| "grad_norm": 0.0017129272455349565, | |
| "learning_rate": 6.396588486140725e-07, | |
| "loss": 0.0002, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 4.946695095948828, | |
| "grad_norm": 0.002514494815841317, | |
| "learning_rate": 5.33049040511727e-07, | |
| "loss": 0.0001, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 4.957356076759062, | |
| "grad_norm": 0.0012437364785000682, | |
| "learning_rate": 4.264392324093817e-07, | |
| "loss": 0.0001, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 4.968017057569297, | |
| "grad_norm": 0.0011664883932098746, | |
| "learning_rate": 3.1982942430703626e-07, | |
| "loss": 0.0001, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 4.978678038379531, | |
| "grad_norm": 0.005681448150426149, | |
| "learning_rate": 2.1321961620469084e-07, | |
| "loss": 0.0002, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 4.9893390191897655, | |
| "grad_norm": 0.0011803240049630404, | |
| "learning_rate": 1.0660980810234542e-07, | |
| "loss": 0.0001, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.0018254140159115195, | |
| "learning_rate": 0.0, | |
| "loss": 0.0001, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9874666666666667, | |
| "eval_loss": 0.06246929615736008, | |
| "eval_runtime": 17.414, | |
| "eval_samples_per_second": 215.344, | |
| "eval_steps_per_second": 6.776, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 2345, | |
| "total_flos": 5.8118992210944e+18, | |
| "train_loss": 0.013133438363490598, | |
| "train_runtime": 809.9301, | |
| "train_samples_per_second": 92.601, | |
| "train_steps_per_second": 2.895 | |
| } | |
| ], | |
| "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 | |
| } | |