| { | |
| "best_metric": 0.21690432727336884, | |
| "best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles10_seed3_q3_DA/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": 1.2761040925979614, | |
| "learning_rate": 4.989339019189766e-05, | |
| "loss": 0.0308, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021321961620469083, | |
| "grad_norm": 0.7120761275291443, | |
| "learning_rate": 4.978678038379531e-05, | |
| "loss": 0.129, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031982942430703626, | |
| "grad_norm": 5.60127592086792, | |
| "learning_rate": 4.9680170575692967e-05, | |
| "loss": 0.068, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042643923240938165, | |
| "grad_norm": 6.488447189331055, | |
| "learning_rate": 4.957356076759062e-05, | |
| "loss": 0.0945, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053304904051172705, | |
| "grad_norm": 0.754040539264679, | |
| "learning_rate": 4.9466950959488276e-05, | |
| "loss": 0.0877, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06396588486140725, | |
| "grad_norm": 8.383085250854492, | |
| "learning_rate": 4.936034115138593e-05, | |
| "loss": 0.1354, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07462686567164178, | |
| "grad_norm": 3.192227840423584, | |
| "learning_rate": 4.9253731343283586e-05, | |
| "loss": 0.0889, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08528784648187633, | |
| "grad_norm": 0.5611108541488647, | |
| "learning_rate": 4.914712153518124e-05, | |
| "loss": 0.095, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09594882729211088, | |
| "grad_norm": 0.4259551167488098, | |
| "learning_rate": 4.904051172707889e-05, | |
| "loss": 0.0996, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10660980810234541, | |
| "grad_norm": 5.680491924285889, | |
| "learning_rate": 4.893390191897655e-05, | |
| "loss": 0.0673, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11727078891257996, | |
| "grad_norm": 3.350114583969116, | |
| "learning_rate": 4.88272921108742e-05, | |
| "loss": 0.0756, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1279317697228145, | |
| "grad_norm": 2.8575639724731445, | |
| "learning_rate": 4.872068230277186e-05, | |
| "loss": 0.0994, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13859275053304904, | |
| "grad_norm": 4.802866458892822, | |
| "learning_rate": 4.861407249466951e-05, | |
| "loss": 0.035, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14925373134328357, | |
| "grad_norm": 2.742772340774536, | |
| "learning_rate": 4.850746268656717e-05, | |
| "loss": 0.0626, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15991471215351813, | |
| "grad_norm": 7.059576511383057, | |
| "learning_rate": 4.840085287846482e-05, | |
| "loss": 0.0433, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17057569296375266, | |
| "grad_norm": 1.340943455696106, | |
| "learning_rate": 4.829424307036248e-05, | |
| "loss": 0.0352, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1812366737739872, | |
| "grad_norm": 6.179232120513916, | |
| "learning_rate": 4.8187633262260126e-05, | |
| "loss": 0.1597, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19189765458422176, | |
| "grad_norm": 0.4416646957397461, | |
| "learning_rate": 4.808102345415779e-05, | |
| "loss": 0.0196, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2025586353944563, | |
| "grad_norm": 4.4928154945373535, | |
| "learning_rate": 4.7974413646055436e-05, | |
| "loss": 0.0581, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21321961620469082, | |
| "grad_norm": 6.689644813537598, | |
| "learning_rate": 4.78678038379531e-05, | |
| "loss": 0.1373, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22388059701492538, | |
| "grad_norm": 5.033041477203369, | |
| "learning_rate": 4.7761194029850745e-05, | |
| "loss": 0.1458, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2345415778251599, | |
| "grad_norm": 1.3557605743408203, | |
| "learning_rate": 4.765458422174841e-05, | |
| "loss": 0.0678, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24520255863539445, | |
| "grad_norm": 0.7837586998939514, | |
| "learning_rate": 4.7547974413646055e-05, | |
| "loss": 0.0163, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.255863539445629, | |
| "grad_norm": 6.083350658416748, | |
| "learning_rate": 4.7441364605543716e-05, | |
| "loss": 0.083, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26652452025586354, | |
| "grad_norm": 1.0917638540267944, | |
| "learning_rate": 4.7334754797441364e-05, | |
| "loss": 0.0818, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2771855010660981, | |
| "grad_norm": 4.8440985679626465, | |
| "learning_rate": 4.7228144989339026e-05, | |
| "loss": 0.0586, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2878464818763326, | |
| "grad_norm": 1.3990978002548218, | |
| "learning_rate": 4.7121535181236674e-05, | |
| "loss": 0.0617, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 3.574483871459961, | |
| "learning_rate": 4.7014925373134335e-05, | |
| "loss": 0.0796, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3091684434968017, | |
| "grad_norm": 1.9401699304580688, | |
| "learning_rate": 4.690831556503198e-05, | |
| "loss": 0.0529, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31982942430703626, | |
| "grad_norm": 6.813943862915039, | |
| "learning_rate": 4.6801705756929645e-05, | |
| "loss": 0.0445, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3304904051172708, | |
| "grad_norm": 3.839949369430542, | |
| "learning_rate": 4.669509594882729e-05, | |
| "loss": 0.0812, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3411513859275053, | |
| "grad_norm": 4.410350799560547, | |
| "learning_rate": 4.658848614072495e-05, | |
| "loss": 0.0975, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.35181236673773986, | |
| "grad_norm": 4.987349510192871, | |
| "learning_rate": 4.64818763326226e-05, | |
| "loss": 0.1128, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3624733475479744, | |
| "grad_norm": 4.065298080444336, | |
| "learning_rate": 4.637526652452026e-05, | |
| "loss": 0.0792, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.373134328358209, | |
| "grad_norm": 2.535940647125244, | |
| "learning_rate": 4.626865671641791e-05, | |
| "loss": 0.057, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3837953091684435, | |
| "grad_norm": 0.30488625168800354, | |
| "learning_rate": 4.6162046908315566e-05, | |
| "loss": 0.0364, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39445628997867804, | |
| "grad_norm": 2.0524277687072754, | |
| "learning_rate": 4.605543710021322e-05, | |
| "loss": 0.0477, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4051172707889126, | |
| "grad_norm": 0.5935394167900085, | |
| "learning_rate": 4.5948827292110876e-05, | |
| "loss": 0.0479, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4157782515991471, | |
| "grad_norm": 0.06417328119277954, | |
| "learning_rate": 4.584221748400853e-05, | |
| "loss": 0.0514, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 7.112630367279053, | |
| "learning_rate": 4.5735607675906185e-05, | |
| "loss": 0.0715, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43710021321961623, | |
| "grad_norm": 3.356646776199341, | |
| "learning_rate": 4.562899786780384e-05, | |
| "loss": 0.0837, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44776119402985076, | |
| "grad_norm": 0.06478744745254517, | |
| "learning_rate": 4.5522388059701495e-05, | |
| "loss": 0.0815, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4584221748400853, | |
| "grad_norm": 2.573807954788208, | |
| "learning_rate": 4.541577825159915e-05, | |
| "loss": 0.1046, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4690831556503198, | |
| "grad_norm": 1.842407464981079, | |
| "learning_rate": 4.5309168443496804e-05, | |
| "loss": 0.076, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.47974413646055436, | |
| "grad_norm": 8.195022583007812, | |
| "learning_rate": 4.520255863539446e-05, | |
| "loss": 0.0964, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4904051172707889, | |
| "grad_norm": 7.50063419342041, | |
| "learning_rate": 4.5095948827292114e-05, | |
| "loss": 0.0689, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5010660980810234, | |
| "grad_norm": 1.9222331047058105, | |
| "learning_rate": 4.498933901918977e-05, | |
| "loss": 0.0729, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.511727078891258, | |
| "grad_norm": 5.999495029449463, | |
| "learning_rate": 4.488272921108742e-05, | |
| "loss": 0.0558, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5223880597014925, | |
| "grad_norm": 0.05648266524076462, | |
| "learning_rate": 4.477611940298508e-05, | |
| "loss": 0.0697, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5330490405117271, | |
| "grad_norm": 0.11434520035982132, | |
| "learning_rate": 4.466950959488273e-05, | |
| "loss": 0.0478, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5437100213219617, | |
| "grad_norm": 1.511038064956665, | |
| "learning_rate": 4.456289978678039e-05, | |
| "loss": 0.0854, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5543710021321961, | |
| "grad_norm": 2.914787769317627, | |
| "learning_rate": 4.445628997867804e-05, | |
| "loss": 0.1109, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5650319829424307, | |
| "grad_norm": 2.3214447498321533, | |
| "learning_rate": 4.43496801705757e-05, | |
| "loss": 0.0468, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5756929637526652, | |
| "grad_norm": 3.5576605796813965, | |
| "learning_rate": 4.424307036247335e-05, | |
| "loss": 0.0898, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5863539445628998, | |
| "grad_norm": 0.2248571217060089, | |
| "learning_rate": 4.4136460554371006e-05, | |
| "loss": 0.0506, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 5.306698322296143, | |
| "learning_rate": 4.402985074626866e-05, | |
| "loss": 0.1244, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6076759061833689, | |
| "grad_norm": 3.1506476402282715, | |
| "learning_rate": 4.3923240938166316e-05, | |
| "loss": 0.0754, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6183368869936035, | |
| "grad_norm": 9.396873474121094, | |
| "learning_rate": 4.381663113006397e-05, | |
| "loss": 0.1772, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6289978678038379, | |
| "grad_norm": 6.985416412353516, | |
| "learning_rate": 4.3710021321961625e-05, | |
| "loss": 0.1688, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6396588486140725, | |
| "grad_norm": 2.2399210929870605, | |
| "learning_rate": 4.360341151385928e-05, | |
| "loss": 0.0418, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.650319829424307, | |
| "grad_norm": 0.4065213203430176, | |
| "learning_rate": 4.3496801705756935e-05, | |
| "loss": 0.0554, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6609808102345416, | |
| "grad_norm": 3.223773241043091, | |
| "learning_rate": 4.339019189765459e-05, | |
| "loss": 0.0629, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6716417910447762, | |
| "grad_norm": 2.0158231258392334, | |
| "learning_rate": 4.328358208955224e-05, | |
| "loss": 0.0827, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6823027718550106, | |
| "grad_norm": 0.5173208117485046, | |
| "learning_rate": 4.31769722814499e-05, | |
| "loss": 0.0585, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6929637526652452, | |
| "grad_norm": 5.735258102416992, | |
| "learning_rate": 4.307036247334755e-05, | |
| "loss": 0.1071, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7036247334754797, | |
| "grad_norm": 5.027115821838379, | |
| "learning_rate": 4.29637526652452e-05, | |
| "loss": 0.0913, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 4.075342655181885, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.043, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7249466950959488, | |
| "grad_norm": 4.211130142211914, | |
| "learning_rate": 4.275053304904051e-05, | |
| "loss": 0.1028, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7356076759061834, | |
| "grad_norm": 4.21316385269165, | |
| "learning_rate": 4.2643923240938166e-05, | |
| "loss": 0.0598, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "grad_norm": 1.3392406702041626, | |
| "learning_rate": 4.253731343283582e-05, | |
| "loss": 0.1004, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7569296375266524, | |
| "grad_norm": 1.3003551959991455, | |
| "learning_rate": 4.2430703624733475e-05, | |
| "loss": 0.0311, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.767590618336887, | |
| "grad_norm": 3.296175003051758, | |
| "learning_rate": 4.232409381663113e-05, | |
| "loss": 0.0661, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7782515991471215, | |
| "grad_norm": 1.0256417989730835, | |
| "learning_rate": 4.2217484008528785e-05, | |
| "loss": 0.0675, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7889125799573561, | |
| "grad_norm": 4.519186973571777, | |
| "learning_rate": 4.211087420042644e-05, | |
| "loss": 0.105, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7995735607675906, | |
| "grad_norm": 5.849189281463623, | |
| "learning_rate": 4.2004264392324094e-05, | |
| "loss": 0.0612, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8102345415778252, | |
| "grad_norm": 0.7139642238616943, | |
| "learning_rate": 4.189765458422175e-05, | |
| "loss": 0.0953, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8208955223880597, | |
| "grad_norm": 3.0020992755889893, | |
| "learning_rate": 4.1791044776119404e-05, | |
| "loss": 0.1173, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8315565031982942, | |
| "grad_norm": 4.255969524383545, | |
| "learning_rate": 4.168443496801706e-05, | |
| "loss": 0.1892, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8422174840085288, | |
| "grad_norm": 2.3645100593566895, | |
| "learning_rate": 4.157782515991471e-05, | |
| "loss": 0.1455, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 0.5539878010749817, | |
| "learning_rate": 4.147121535181237e-05, | |
| "loss": 0.0181, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8635394456289979, | |
| "grad_norm": 5.9484076499938965, | |
| "learning_rate": 4.136460554371002e-05, | |
| "loss": 0.0668, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8742004264392325, | |
| "grad_norm": 0.10675335675477982, | |
| "learning_rate": 4.125799573560768e-05, | |
| "loss": 0.0836, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8848614072494669, | |
| "grad_norm": 3.036376476287842, | |
| "learning_rate": 4.115138592750533e-05, | |
| "loss": 0.1241, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8955223880597015, | |
| "grad_norm": 1.6516715288162231, | |
| "learning_rate": 4.104477611940299e-05, | |
| "loss": 0.0694, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.906183368869936, | |
| "grad_norm": 2.471933126449585, | |
| "learning_rate": 4.093816631130064e-05, | |
| "loss": 0.0747, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9168443496801706, | |
| "grad_norm": 0.31966668367385864, | |
| "learning_rate": 4.0831556503198296e-05, | |
| "loss": 0.0946, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9275053304904051, | |
| "grad_norm": 1.2862030267715454, | |
| "learning_rate": 4.072494669509595e-05, | |
| "loss": 0.049, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9381663113006397, | |
| "grad_norm": 3.1551568508148193, | |
| "learning_rate": 4.0618336886993606e-05, | |
| "loss": 0.0465, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9488272921108742, | |
| "grad_norm": 1.5859147310256958, | |
| "learning_rate": 4.051172707889126e-05, | |
| "loss": 0.0406, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9594882729211087, | |
| "grad_norm": 1.7494618892669678, | |
| "learning_rate": 4.0405117270788915e-05, | |
| "loss": 0.0779, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9701492537313433, | |
| "grad_norm": 8.09626579284668, | |
| "learning_rate": 4.029850746268657e-05, | |
| "loss": 0.0979, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9808102345415778, | |
| "grad_norm": 3.822662830352783, | |
| "learning_rate": 4.0191897654584225e-05, | |
| "loss": 0.0814, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9914712153518124, | |
| "grad_norm": 1.1835472583770752, | |
| "learning_rate": 4.008528784648188e-05, | |
| "loss": 0.0424, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9344, | |
| "eval_loss": 0.21690432727336884, | |
| "eval_runtime": 32.2124, | |
| "eval_samples_per_second": 116.415, | |
| "eval_steps_per_second": 3.663, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.0021321961620469, | |
| "grad_norm": 1.3520677089691162, | |
| "learning_rate": 3.997867803837953e-05, | |
| "loss": 0.1682, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0127931769722816, | |
| "grad_norm": 0.47257593274116516, | |
| "learning_rate": 3.987206823027719e-05, | |
| "loss": 0.0555, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.023454157782516, | |
| "grad_norm": 3.6214115619659424, | |
| "learning_rate": 3.976545842217484e-05, | |
| "loss": 0.0416, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0341151385927505, | |
| "grad_norm": 3.082700252532959, | |
| "learning_rate": 3.96588486140725e-05, | |
| "loss": 0.1115, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.044776119402985, | |
| "grad_norm": 0.8865631818771362, | |
| "learning_rate": 3.9552238805970146e-05, | |
| "loss": 0.0676, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0554371002132197, | |
| "grad_norm": 6.059133529663086, | |
| "learning_rate": 3.944562899786781e-05, | |
| "loss": 0.1449, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.0660980810234542, | |
| "grad_norm": 1.805566668510437, | |
| "learning_rate": 3.9339019189765456e-05, | |
| "loss": 0.0805, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0767590618336886, | |
| "grad_norm": 7.3406829833984375, | |
| "learning_rate": 3.923240938166312e-05, | |
| "loss": 0.0789, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0874200426439233, | |
| "grad_norm": 5.967226982116699, | |
| "learning_rate": 3.9125799573560765e-05, | |
| "loss": 0.0599, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0980810234541578, | |
| "grad_norm": 4.318120956420898, | |
| "learning_rate": 3.901918976545843e-05, | |
| "loss": 0.107, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.1087420042643923, | |
| "grad_norm": 2.8982036113739014, | |
| "learning_rate": 3.8912579957356075e-05, | |
| "loss": 0.0528, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1194029850746268, | |
| "grad_norm": 2.773726224899292, | |
| "learning_rate": 3.8805970149253736e-05, | |
| "loss": 0.094, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.1300639658848615, | |
| "grad_norm": 3.3290534019470215, | |
| "learning_rate": 3.8699360341151384e-05, | |
| "loss": 0.0682, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.140724946695096, | |
| "grad_norm": 2.5923445224761963, | |
| "learning_rate": 3.8592750533049046e-05, | |
| "loss": 0.1102, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1513859275053304, | |
| "grad_norm": 1.3364297151565552, | |
| "learning_rate": 3.8486140724946694e-05, | |
| "loss": 0.131, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.1620469083155651, | |
| "grad_norm": 2.6128759384155273, | |
| "learning_rate": 3.8379530916844355e-05, | |
| "loss": 0.0836, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.1727078891257996, | |
| "grad_norm": 0.9515971541404724, | |
| "learning_rate": 3.8272921108742e-05, | |
| "loss": 0.011, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.183368869936034, | |
| "grad_norm": 3.3442306518554688, | |
| "learning_rate": 3.8166311300639665e-05, | |
| "loss": 0.0432, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1940298507462686, | |
| "grad_norm": 2.9068615436553955, | |
| "learning_rate": 3.805970149253731e-05, | |
| "loss": 0.1285, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.2046908315565032, | |
| "grad_norm": 7.473453044891357, | |
| "learning_rate": 3.7953091684434974e-05, | |
| "loss": 0.0794, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.2153518123667377, | |
| "grad_norm": 3.618746042251587, | |
| "learning_rate": 3.784648187633262e-05, | |
| "loss": 0.0215, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.2260127931769722, | |
| "grad_norm": 2.0968689918518066, | |
| "learning_rate": 3.7739872068230284e-05, | |
| "loss": 0.017, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.236673773987207, | |
| "grad_norm": 0.2722117304801941, | |
| "learning_rate": 3.763326226012793e-05, | |
| "loss": 0.0261, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2473347547974414, | |
| "grad_norm": 3.4401211738586426, | |
| "learning_rate": 3.752665245202559e-05, | |
| "loss": 0.0662, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.2579957356076759, | |
| "grad_norm": 1.6000653505325317, | |
| "learning_rate": 3.742004264392324e-05, | |
| "loss": 0.0509, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2686567164179103, | |
| "grad_norm": 3.0836596488952637, | |
| "learning_rate": 3.73134328358209e-05, | |
| "loss": 0.0328, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.279317697228145, | |
| "grad_norm": 6.320271015167236, | |
| "learning_rate": 3.720682302771855e-05, | |
| "loss": 0.0518, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2899786780383795, | |
| "grad_norm": 5.865463733673096, | |
| "learning_rate": 3.710021321961621e-05, | |
| "loss": 0.0467, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.3006396588486142, | |
| "grad_norm": 5.013850688934326, | |
| "learning_rate": 3.699360341151386e-05, | |
| "loss": 0.1034, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3113006396588487, | |
| "grad_norm": 1.482189655303955, | |
| "learning_rate": 3.6886993603411515e-05, | |
| "loss": 0.0176, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.3219616204690832, | |
| "grad_norm": 3.8735485076904297, | |
| "learning_rate": 3.678038379530917e-05, | |
| "loss": 0.0829, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.3326226012793176, | |
| "grad_norm": 4.566304683685303, | |
| "learning_rate": 3.6673773987206824e-05, | |
| "loss": 0.0903, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.3432835820895521, | |
| "grad_norm": 5.240992546081543, | |
| "learning_rate": 3.656716417910448e-05, | |
| "loss": 0.0873, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.3539445628997868, | |
| "grad_norm": 3.7858400344848633, | |
| "learning_rate": 3.6460554371002134e-05, | |
| "loss": 0.0748, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.3646055437100213, | |
| "grad_norm": 6.0873942375183105, | |
| "learning_rate": 3.635394456289979e-05, | |
| "loss": 0.1112, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.375266524520256, | |
| "grad_norm": 0.48490238189697266, | |
| "learning_rate": 3.624733475479744e-05, | |
| "loss": 0.0646, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.3859275053304905, | |
| "grad_norm": 3.852931261062622, | |
| "learning_rate": 3.61407249466951e-05, | |
| "loss": 0.1313, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.396588486140725, | |
| "grad_norm": 1.9783458709716797, | |
| "learning_rate": 3.603411513859275e-05, | |
| "loss": 0.0564, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.4072494669509594, | |
| "grad_norm": 0.8096998333930969, | |
| "learning_rate": 3.592750533049041e-05, | |
| "loss": 0.0608, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.417910447761194, | |
| "grad_norm": 3.065155029296875, | |
| "learning_rate": 3.582089552238806e-05, | |
| "loss": 0.1055, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.8327421545982361, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.0247, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.439232409381663, | |
| "grad_norm": 3.2900428771972656, | |
| "learning_rate": 3.560767590618337e-05, | |
| "loss": 0.0551, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.4498933901918978, | |
| "grad_norm": 0.3575810194015503, | |
| "learning_rate": 3.5501066098081026e-05, | |
| "loss": 0.0291, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4605543710021323, | |
| "grad_norm": 6.238643169403076, | |
| "learning_rate": 3.539445628997868e-05, | |
| "loss": 0.0763, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.4712153518123667, | |
| "grad_norm": 3.4827845096588135, | |
| "learning_rate": 3.5287846481876336e-05, | |
| "loss": 0.067, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4818763326226012, | |
| "grad_norm": 4.054008483886719, | |
| "learning_rate": 3.518123667377399e-05, | |
| "loss": 0.0801, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "grad_norm": 0.08711248636245728, | |
| "learning_rate": 3.5074626865671645e-05, | |
| "loss": 0.0926, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5031982942430704, | |
| "grad_norm": 2.605539321899414, | |
| "learning_rate": 3.496801705756929e-05, | |
| "loss": 0.0759, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.5138592750533049, | |
| "grad_norm": 1.505613923072815, | |
| "learning_rate": 3.4861407249466955e-05, | |
| "loss": 0.0594, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.5245202558635396, | |
| "grad_norm": 1.2791396379470825, | |
| "learning_rate": 3.47547974413646e-05, | |
| "loss": 0.0916, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.535181236673774, | |
| "grad_norm": 3.1148681640625, | |
| "learning_rate": 3.4648187633262264e-05, | |
| "loss": 0.106, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.5458422174840085, | |
| "grad_norm": 3.3148257732391357, | |
| "learning_rate": 3.454157782515991e-05, | |
| "loss": 0.0338, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.556503198294243, | |
| "grad_norm": 1.446936845779419, | |
| "learning_rate": 3.4434968017057574e-05, | |
| "loss": 0.0178, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5671641791044775, | |
| "grad_norm": 1.2113245725631714, | |
| "learning_rate": 3.432835820895522e-05, | |
| "loss": 0.0849, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.5778251599147122, | |
| "grad_norm": 0.0070991660468280315, | |
| "learning_rate": 3.422174840085288e-05, | |
| "loss": 0.0399, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5884861407249466, | |
| "grad_norm": 0.6842405200004578, | |
| "learning_rate": 3.411513859275053e-05, | |
| "loss": 0.0995, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.5991471215351813, | |
| "grad_norm": 0.3559590280056, | |
| "learning_rate": 3.400852878464819e-05, | |
| "loss": 0.0326, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.6098081023454158, | |
| "grad_norm": 2.242896318435669, | |
| "learning_rate": 3.390191897654584e-05, | |
| "loss": 0.0523, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.6204690831556503, | |
| "grad_norm": 3.0037975311279297, | |
| "learning_rate": 3.37953091684435e-05, | |
| "loss": 0.0589, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.6311300639658848, | |
| "grad_norm": 1.2411649227142334, | |
| "learning_rate": 3.368869936034115e-05, | |
| "loss": 0.0953, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.6417910447761193, | |
| "grad_norm": 2.926041841506958, | |
| "learning_rate": 3.358208955223881e-05, | |
| "loss": 0.0769, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.652452025586354, | |
| "grad_norm": 4.137303352355957, | |
| "learning_rate": 3.347547974413646e-05, | |
| "loss": 0.0358, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.6631130063965884, | |
| "grad_norm": 1.1193041801452637, | |
| "learning_rate": 3.336886993603412e-05, | |
| "loss": 0.0489, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.6737739872068231, | |
| "grad_norm": 0.3301200866699219, | |
| "learning_rate": 3.326226012793177e-05, | |
| "loss": 0.0863, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.6844349680170576, | |
| "grad_norm": 2.704554319381714, | |
| "learning_rate": 3.3155650319829424e-05, | |
| "loss": 0.0535, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.695095948827292, | |
| "grad_norm": 5.649402141571045, | |
| "learning_rate": 3.304904051172708e-05, | |
| "loss": 0.0774, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.7057569296375266, | |
| "grad_norm": 4.148595809936523, | |
| "learning_rate": 3.294243070362473e-05, | |
| "loss": 0.0755, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.716417910447761, | |
| "grad_norm": 1.9841556549072266, | |
| "learning_rate": 3.283582089552239e-05, | |
| "loss": 0.0991, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.7270788912579957, | |
| "grad_norm": 1.308912992477417, | |
| "learning_rate": 3.272921108742004e-05, | |
| "loss": 0.1002, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.7377398720682304, | |
| "grad_norm": 3.9047133922576904, | |
| "learning_rate": 3.26226012793177e-05, | |
| "loss": 0.0586, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.748400852878465, | |
| "grad_norm": 3.4976601600646973, | |
| "learning_rate": 3.251599147121535e-05, | |
| "loss": 0.0192, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7590618336886994, | |
| "grad_norm": 3.6482577323913574, | |
| "learning_rate": 3.240938166311301e-05, | |
| "loss": 0.0438, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.7697228144989339, | |
| "grad_norm": 0.8176335692405701, | |
| "learning_rate": 3.230277185501066e-05, | |
| "loss": 0.0235, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.7803837953091683, | |
| "grad_norm": 3.829963207244873, | |
| "learning_rate": 3.2196162046908317e-05, | |
| "loss": 0.0456, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.7910447761194028, | |
| "grad_norm": 0.31196293234825134, | |
| "learning_rate": 3.208955223880597e-05, | |
| "loss": 0.0982, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.8017057569296375, | |
| "grad_norm": 7.491602420806885, | |
| "learning_rate": 3.1982942430703626e-05, | |
| "loss": 0.0181, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.8123667377398722, | |
| "grad_norm": 6.805478572845459, | |
| "learning_rate": 3.187633262260128e-05, | |
| "loss": 0.0503, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.8230277185501067, | |
| "grad_norm": 1.0252180099487305, | |
| "learning_rate": 3.1769722814498935e-05, | |
| "loss": 0.0568, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.8336886993603412, | |
| "grad_norm": 0.4379175007343292, | |
| "learning_rate": 3.166311300639659e-05, | |
| "loss": 0.0772, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.8443496801705757, | |
| "grad_norm": 3.555905818939209, | |
| "learning_rate": 3.1556503198294245e-05, | |
| "loss": 0.0462, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.8550106609808101, | |
| "grad_norm": 4.489678382873535, | |
| "learning_rate": 3.14498933901919e-05, | |
| "loss": 0.0287, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.8656716417910446, | |
| "grad_norm": 4.72069787979126, | |
| "learning_rate": 3.1343283582089554e-05, | |
| "loss": 0.0366, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.8763326226012793, | |
| "grad_norm": 0.8965930938720703, | |
| "learning_rate": 3.123667377398721e-05, | |
| "loss": 0.0541, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.886993603411514, | |
| "grad_norm": 3.1170096397399902, | |
| "learning_rate": 3.1130063965884864e-05, | |
| "loss": 0.1147, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.8976545842217485, | |
| "grad_norm": 0.2730638384819031, | |
| "learning_rate": 3.102345415778252e-05, | |
| "loss": 0.0537, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.908315565031983, | |
| "grad_norm": 2.117879867553711, | |
| "learning_rate": 3.0916844349680173e-05, | |
| "loss": 0.1638, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.9189765458422174, | |
| "grad_norm": 3.1977241039276123, | |
| "learning_rate": 3.081023454157783e-05, | |
| "loss": 0.0662, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.929637526652452, | |
| "grad_norm": 3.2297840118408203, | |
| "learning_rate": 3.070362473347548e-05, | |
| "loss": 0.0684, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.9402985074626866, | |
| "grad_norm": 3.7868151664733887, | |
| "learning_rate": 3.059701492537314e-05, | |
| "loss": 0.0997, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.950959488272921, | |
| "grad_norm": 0.1728767305612564, | |
| "learning_rate": 3.0490405117270792e-05, | |
| "loss": 0.0406, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.9616204690831558, | |
| "grad_norm": 2.2640936374664307, | |
| "learning_rate": 3.0383795309168444e-05, | |
| "loss": 0.0644, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.9722814498933903, | |
| "grad_norm": 0.07849620282649994, | |
| "learning_rate": 3.0277185501066102e-05, | |
| "loss": 0.0334, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.9829424307036247, | |
| "grad_norm": 3.5660316944122314, | |
| "learning_rate": 3.0170575692963753e-05, | |
| "loss": 0.08, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.9936034115138592, | |
| "grad_norm": 4.534877300262451, | |
| "learning_rate": 3.006396588486141e-05, | |
| "loss": 0.0932, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9408, | |
| "eval_loss": 0.2174207866191864, | |
| "eval_runtime": 30.128, | |
| "eval_samples_per_second": 124.469, | |
| "eval_steps_per_second": 3.917, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 2.0042643923240937, | |
| "grad_norm": 0.34220871329307556, | |
| "learning_rate": 2.9957356076759063e-05, | |
| "loss": 0.0534, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.014925373134328, | |
| "grad_norm": 2.9265432357788086, | |
| "learning_rate": 2.9850746268656714e-05, | |
| "loss": 0.0352, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.025586353944563, | |
| "grad_norm": 0.5136100053787231, | |
| "learning_rate": 2.9744136460554372e-05, | |
| "loss": 0.1112, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.0362473347547976, | |
| "grad_norm": 1.8729381561279297, | |
| "learning_rate": 2.9637526652452023e-05, | |
| "loss": 0.1049, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.046908315565032, | |
| "grad_norm": 0.6492112278938293, | |
| "learning_rate": 2.953091684434968e-05, | |
| "loss": 0.005, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.0575692963752665, | |
| "grad_norm": 8.696846008300781, | |
| "learning_rate": 2.9424307036247333e-05, | |
| "loss": 0.0571, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.068230277185501, | |
| "grad_norm": 0.47494128346443176, | |
| "learning_rate": 2.931769722814499e-05, | |
| "loss": 0.0595, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.0788912579957355, | |
| "grad_norm": 2.162015199661255, | |
| "learning_rate": 2.9211087420042642e-05, | |
| "loss": 0.0424, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.08955223880597, | |
| "grad_norm": 2.1177449226379395, | |
| "learning_rate": 2.91044776119403e-05, | |
| "loss": 0.0468, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.100213219616205, | |
| "grad_norm": 0.01072877086699009, | |
| "learning_rate": 2.8997867803837952e-05, | |
| "loss": 0.0117, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.1108742004264394, | |
| "grad_norm": 3.3539516925811768, | |
| "learning_rate": 2.889125799573561e-05, | |
| "loss": 0.0874, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.121535181236674, | |
| "grad_norm": 6.6221418380737305, | |
| "learning_rate": 2.878464818763326e-05, | |
| "loss": 0.0581, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.1321961620469083, | |
| "grad_norm": 0.4532221853733063, | |
| "learning_rate": 2.867803837953092e-05, | |
| "loss": 0.1169, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.7453476786613464, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.0261, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.1535181236673773, | |
| "grad_norm": 3.842346429824829, | |
| "learning_rate": 2.846481876332623e-05, | |
| "loss": 0.0722, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.1641791044776117, | |
| "grad_norm": 3.4456448554992676, | |
| "learning_rate": 2.835820895522388e-05, | |
| "loss": 0.0536, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.1748400852878467, | |
| "grad_norm": 0.7244068384170532, | |
| "learning_rate": 2.825159914712154e-05, | |
| "loss": 0.0172, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.185501066098081, | |
| "grad_norm": 0.14325617253780365, | |
| "learning_rate": 2.814498933901919e-05, | |
| "loss": 0.0196, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.1961620469083156, | |
| "grad_norm": 0.062447935342788696, | |
| "learning_rate": 2.8038379530916848e-05, | |
| "loss": 0.0177, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.20682302771855, | |
| "grad_norm": 2.0038044452667236, | |
| "learning_rate": 2.79317697228145e-05, | |
| "loss": 0.054, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.2174840085287846, | |
| "grad_norm": 3.240445137023926, | |
| "learning_rate": 2.7825159914712157e-05, | |
| "loss": 0.0936, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.228144989339019, | |
| "grad_norm": 2.3319170475006104, | |
| "learning_rate": 2.771855010660981e-05, | |
| "loss": 0.0601, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.2388059701492535, | |
| "grad_norm": 6.946221828460693, | |
| "learning_rate": 2.7611940298507467e-05, | |
| "loss": 0.1142, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.2494669509594885, | |
| "grad_norm": 3.61859130859375, | |
| "learning_rate": 2.7505330490405118e-05, | |
| "loss": 0.0894, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.260127931769723, | |
| "grad_norm": 5.21867561340332, | |
| "learning_rate": 2.7398720682302776e-05, | |
| "loss": 0.0615, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.2707889125799574, | |
| "grad_norm": 1.7304450273513794, | |
| "learning_rate": 2.7292110874200428e-05, | |
| "loss": 0.0608, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.281449893390192, | |
| "grad_norm": 0.5409643650054932, | |
| "learning_rate": 2.7185501066098086e-05, | |
| "loss": 0.0419, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.2921108742004264, | |
| "grad_norm": 4.42728328704834, | |
| "learning_rate": 2.7078891257995737e-05, | |
| "loss": 0.0359, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.302771855010661, | |
| "grad_norm": 0.14021319150924683, | |
| "learning_rate": 2.6972281449893395e-05, | |
| "loss": 0.0572, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.3134328358208958, | |
| "grad_norm": 4.749971866607666, | |
| "learning_rate": 2.6865671641791047e-05, | |
| "loss": 0.0236, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.3240938166311302, | |
| "grad_norm": 5.575937747955322, | |
| "learning_rate": 2.6759061833688705e-05, | |
| "loss": 0.091, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.3347547974413647, | |
| "grad_norm": 3.336470603942871, | |
| "learning_rate": 2.6652452025586356e-05, | |
| "loss": 0.079, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.345415778251599, | |
| "grad_norm": 1.8877921104431152, | |
| "learning_rate": 2.6545842217484007e-05, | |
| "loss": 0.0246, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.3560767590618337, | |
| "grad_norm": 0.056747760623693466, | |
| "learning_rate": 2.6439232409381666e-05, | |
| "loss": 0.0505, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.366737739872068, | |
| "grad_norm": 1.290672779083252, | |
| "learning_rate": 2.6332622601279317e-05, | |
| "loss": 0.0371, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.3773987206823026, | |
| "grad_norm": 0.17507269978523254, | |
| "learning_rate": 2.6226012793176975e-05, | |
| "loss": 0.017, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.388059701492537, | |
| "grad_norm": 1.9645771980285645, | |
| "learning_rate": 2.6119402985074626e-05, | |
| "loss": 0.0122, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.398720682302772, | |
| "grad_norm": 0.5594134330749512, | |
| "learning_rate": 2.6012793176972285e-05, | |
| "loss": 0.0378, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.4093816631130065, | |
| "grad_norm": 0.10119452327489853, | |
| "learning_rate": 2.5906183368869936e-05, | |
| "loss": 0.054, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.420042643923241, | |
| "grad_norm": 3.3604164123535156, | |
| "learning_rate": 2.5799573560767594e-05, | |
| "loss": 0.0877, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.4307036247334755, | |
| "grad_norm": 0.5324254035949707, | |
| "learning_rate": 2.5692963752665245e-05, | |
| "loss": 0.0487, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.44136460554371, | |
| "grad_norm": 0.34250810742378235, | |
| "learning_rate": 2.5586353944562904e-05, | |
| "loss": 0.1339, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.4520255863539444, | |
| "grad_norm": 3.1593129634857178, | |
| "learning_rate": 2.5479744136460555e-05, | |
| "loss": 0.0475, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.4626865671641793, | |
| "grad_norm": 5.577476501464844, | |
| "learning_rate": 2.537313432835821e-05, | |
| "loss": 0.0411, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.473347547974414, | |
| "grad_norm": 1.2174274921417236, | |
| "learning_rate": 2.5266524520255864e-05, | |
| "loss": 0.0392, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.4840085287846483, | |
| "grad_norm": 1.5366300344467163, | |
| "learning_rate": 2.515991471215352e-05, | |
| "loss": 0.0617, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.4946695095948828, | |
| "grad_norm": 1.4823065996170044, | |
| "learning_rate": 2.5053304904051174e-05, | |
| "loss": 0.091, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.5053304904051172, | |
| "grad_norm": 0.028011556714773178, | |
| "learning_rate": 2.494669509594883e-05, | |
| "loss": 0.0419, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.5159914712153517, | |
| "grad_norm": 0.28680238127708435, | |
| "learning_rate": 2.4840085287846483e-05, | |
| "loss": 0.0179, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.526652452025586, | |
| "grad_norm": 0.17218799889087677, | |
| "learning_rate": 2.4733475479744138e-05, | |
| "loss": 0.0107, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.5373134328358207, | |
| "grad_norm": 5.200352191925049, | |
| "learning_rate": 2.4626865671641793e-05, | |
| "loss": 0.0759, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.5479744136460556, | |
| "grad_norm": 3.023249387741089, | |
| "learning_rate": 2.4520255863539444e-05, | |
| "loss": 0.0157, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.55863539445629, | |
| "grad_norm": 3.2880547046661377, | |
| "learning_rate": 2.44136460554371e-05, | |
| "loss": 0.0552, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.5692963752665245, | |
| "grad_norm": 3.7705929279327393, | |
| "learning_rate": 2.4307036247334754e-05, | |
| "loss": 0.0979, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.579957356076759, | |
| "grad_norm": 0.043199606239795685, | |
| "learning_rate": 2.420042643923241e-05, | |
| "loss": 0.0908, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.5906183368869935, | |
| "grad_norm": 0.04789150133728981, | |
| "learning_rate": 2.4093816631130063e-05, | |
| "loss": 0.0518, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.6012793176972284, | |
| "grad_norm": 0.09517115354537964, | |
| "learning_rate": 2.3987206823027718e-05, | |
| "loss": 0.0387, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.611940298507463, | |
| "grad_norm": 4.101116180419922, | |
| "learning_rate": 2.3880597014925373e-05, | |
| "loss": 0.044, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.6226012793176974, | |
| "grad_norm": 0.03614508733153343, | |
| "learning_rate": 2.3773987206823027e-05, | |
| "loss": 0.0269, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.633262260127932, | |
| "grad_norm": 1.8597288131713867, | |
| "learning_rate": 2.3667377398720682e-05, | |
| "loss": 0.0276, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.6439232409381663, | |
| "grad_norm": 0.567395031452179, | |
| "learning_rate": 2.3560767590618337e-05, | |
| "loss": 0.0143, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.654584221748401, | |
| "grad_norm": 0.8959960341453552, | |
| "learning_rate": 2.345415778251599e-05, | |
| "loss": 0.0836, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.6652452025586353, | |
| "grad_norm": 0.14519836008548737, | |
| "learning_rate": 2.3347547974413646e-05, | |
| "loss": 0.0081, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.6759061833688698, | |
| "grad_norm": 4.262569427490234, | |
| "learning_rate": 2.32409381663113e-05, | |
| "loss": 0.0554, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.6865671641791042, | |
| "grad_norm": 4.762922763824463, | |
| "learning_rate": 2.3134328358208956e-05, | |
| "loss": 0.0823, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.697228144989339, | |
| "grad_norm": 1.29306960105896, | |
| "learning_rate": 2.302771855010661e-05, | |
| "loss": 0.0185, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.7078891257995736, | |
| "grad_norm": 4.81442403793335, | |
| "learning_rate": 2.2921108742004265e-05, | |
| "loss": 0.0292, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.718550106609808, | |
| "grad_norm": 1.9069541692733765, | |
| "learning_rate": 2.281449893390192e-05, | |
| "loss": 0.0292, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.7292110874200426, | |
| "grad_norm": 1.2942973375320435, | |
| "learning_rate": 2.2707889125799575e-05, | |
| "loss": 0.0891, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.739872068230277, | |
| "grad_norm": 2.6902883052825928, | |
| "learning_rate": 2.260127931769723e-05, | |
| "loss": 0.0188, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.750533049040512, | |
| "grad_norm": 6.201111793518066, | |
| "learning_rate": 2.2494669509594884e-05, | |
| "loss": 0.0645, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.7611940298507465, | |
| "grad_norm": 1.4502270221710205, | |
| "learning_rate": 2.238805970149254e-05, | |
| "loss": 0.1304, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.771855010660981, | |
| "grad_norm": 4.815301418304443, | |
| "learning_rate": 2.2281449893390194e-05, | |
| "loss": 0.0515, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.7825159914712154, | |
| "grad_norm": 3.6600866317749023, | |
| "learning_rate": 2.217484008528785e-05, | |
| "loss": 0.0374, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.79317697228145, | |
| "grad_norm": 1.428259015083313, | |
| "learning_rate": 2.2068230277185503e-05, | |
| "loss": 0.011, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.8038379530916844, | |
| "grad_norm": 0.9658362865447998, | |
| "learning_rate": 2.1961620469083158e-05, | |
| "loss": 0.0467, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.814498933901919, | |
| "grad_norm": 3.706225872039795, | |
| "learning_rate": 2.1855010660980813e-05, | |
| "loss": 0.1018, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.8251599147121533, | |
| "grad_norm": 0.08485915511846542, | |
| "learning_rate": 2.1748400852878467e-05, | |
| "loss": 0.032, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.835820895522388, | |
| "grad_norm": 0.23524874448776245, | |
| "learning_rate": 2.164179104477612e-05, | |
| "loss": 0.019, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.8464818763326227, | |
| "grad_norm": 3.4048266410827637, | |
| "learning_rate": 2.1535181236673773e-05, | |
| "loss": 0.0087, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 2.7785627841949463, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.0281, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.8678038379530917, | |
| "grad_norm": 0.6112996935844421, | |
| "learning_rate": 2.1321961620469083e-05, | |
| "loss": 0.039, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.878464818763326, | |
| "grad_norm": 0.16225238144397736, | |
| "learning_rate": 2.1215351812366738e-05, | |
| "loss": 0.0094, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.8891257995735606, | |
| "grad_norm": 4.850905418395996, | |
| "learning_rate": 2.1108742004264392e-05, | |
| "loss": 0.0438, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.8997867803837956, | |
| "grad_norm": 5.391885280609131, | |
| "learning_rate": 2.1002132196162047e-05, | |
| "loss": 0.0315, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.91044776119403, | |
| "grad_norm": 3.315335512161255, | |
| "learning_rate": 2.0895522388059702e-05, | |
| "loss": 0.0809, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.9211087420042645, | |
| "grad_norm": 4.729955673217773, | |
| "learning_rate": 2.0788912579957357e-05, | |
| "loss": 0.0619, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.931769722814499, | |
| "grad_norm": 10.124401092529297, | |
| "learning_rate": 2.068230277185501e-05, | |
| "loss": 0.0681, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.9424307036247335, | |
| "grad_norm": 1.885017991065979, | |
| "learning_rate": 2.0575692963752666e-05, | |
| "loss": 0.0361, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.953091684434968, | |
| "grad_norm": 0.060406193137168884, | |
| "learning_rate": 2.046908315565032e-05, | |
| "loss": 0.0419, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.9637526652452024, | |
| "grad_norm": 5.123465061187744, | |
| "learning_rate": 2.0362473347547976e-05, | |
| "loss": 0.0813, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.974413646055437, | |
| "grad_norm": 7.036966323852539, | |
| "learning_rate": 2.025586353944563e-05, | |
| "loss": 0.0265, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "grad_norm": 0.5138486623764038, | |
| "learning_rate": 2.0149253731343285e-05, | |
| "loss": 0.0252, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.9957356076759063, | |
| "grad_norm": 1.0654997825622559, | |
| "learning_rate": 2.004264392324094e-05, | |
| "loss": 0.0517, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9429333333333333, | |
| "eval_loss": 0.2282038927078247, | |
| "eval_runtime": 30.4898, | |
| "eval_samples_per_second": 122.992, | |
| "eval_steps_per_second": 3.87, | |
| "step": 1407 | |
| }, | |
| { | |
| "epoch": 3.0063965884861408, | |
| "grad_norm": 6.2994704246521, | |
| "learning_rate": 1.9936034115138594e-05, | |
| "loss": 0.027, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.0170575692963753, | |
| "grad_norm": 2.811976671218872, | |
| "learning_rate": 1.982942430703625e-05, | |
| "loss": 0.0386, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 3.0277185501066097, | |
| "grad_norm": 3.222944498062134, | |
| "learning_rate": 1.9722814498933904e-05, | |
| "loss": 0.047, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.038379530916844, | |
| "grad_norm": 0.5290930867195129, | |
| "learning_rate": 1.961620469083156e-05, | |
| "loss": 0.1005, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 3.0490405117270787, | |
| "grad_norm": 1.3705010414123535, | |
| "learning_rate": 1.9509594882729213e-05, | |
| "loss": 0.081, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.0597014925373136, | |
| "grad_norm": 0.2586652934551239, | |
| "learning_rate": 1.9402985074626868e-05, | |
| "loss": 0.0452, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 3.070362473347548, | |
| "grad_norm": 1.2062602043151855, | |
| "learning_rate": 1.9296375266524523e-05, | |
| "loss": 0.0162, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.0810234541577826, | |
| "grad_norm": 3.9019551277160645, | |
| "learning_rate": 1.9189765458422178e-05, | |
| "loss": 0.0394, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 3.091684434968017, | |
| "grad_norm": 4.043559551239014, | |
| "learning_rate": 1.9083155650319832e-05, | |
| "loss": 0.0905, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.1023454157782515, | |
| "grad_norm": 3.7266945838928223, | |
| "learning_rate": 1.8976545842217487e-05, | |
| "loss": 0.0254, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 3.113006396588486, | |
| "grad_norm": 5.606441497802734, | |
| "learning_rate": 1.8869936034115142e-05, | |
| "loss": 0.0704, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.1236673773987205, | |
| "grad_norm": 3.7815864086151123, | |
| "learning_rate": 1.8763326226012797e-05, | |
| "loss": 0.0315, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 3.1343283582089554, | |
| "grad_norm": 2.4763667583465576, | |
| "learning_rate": 1.865671641791045e-05, | |
| "loss": 0.044, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.14498933901919, | |
| "grad_norm": 3.784590482711792, | |
| "learning_rate": 1.8550106609808106e-05, | |
| "loss": 0.0774, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 3.1556503198294243, | |
| "grad_norm": 4.494934558868408, | |
| "learning_rate": 1.8443496801705757e-05, | |
| "loss": 0.1083, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.166311300639659, | |
| "grad_norm": 6.29168701171875, | |
| "learning_rate": 1.8336886993603412e-05, | |
| "loss": 0.0299, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 3.1769722814498933, | |
| "grad_norm": 0.40900638699531555, | |
| "learning_rate": 1.8230277185501067e-05, | |
| "loss": 0.0507, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.1876332622601278, | |
| "grad_norm": 0.007251468952745199, | |
| "learning_rate": 1.812366737739872e-05, | |
| "loss": 0.0085, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 3.1982942430703627, | |
| "grad_norm": 0.4380704164505005, | |
| "learning_rate": 1.8017057569296376e-05, | |
| "loss": 0.0448, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.208955223880597, | |
| "grad_norm": 1.5679435729980469, | |
| "learning_rate": 1.791044776119403e-05, | |
| "loss": 0.0532, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 3.2196162046908317, | |
| "grad_norm": 4.718801498413086, | |
| "learning_rate": 1.7803837953091686e-05, | |
| "loss": 0.0482, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.230277185501066, | |
| "grad_norm": 3.951225519180298, | |
| "learning_rate": 1.769722814498934e-05, | |
| "loss": 0.0358, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 3.2409381663113006, | |
| "grad_norm": 3.9249770641326904, | |
| "learning_rate": 1.7590618336886995e-05, | |
| "loss": 0.0183, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.251599147121535, | |
| "grad_norm": 0.36235642433166504, | |
| "learning_rate": 1.7484008528784647e-05, | |
| "loss": 0.0228, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 3.2622601279317696, | |
| "grad_norm": 2.9806249141693115, | |
| "learning_rate": 1.73773987206823e-05, | |
| "loss": 0.0253, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.272921108742004, | |
| "grad_norm": 3.8597230911254883, | |
| "learning_rate": 1.7270788912579956e-05, | |
| "loss": 0.0622, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 3.283582089552239, | |
| "grad_norm": 0.09389183670282364, | |
| "learning_rate": 1.716417910447761e-05, | |
| "loss": 0.0461, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.2942430703624734, | |
| "grad_norm": 13.272943496704102, | |
| "learning_rate": 1.7057569296375266e-05, | |
| "loss": 0.0377, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 3.304904051172708, | |
| "grad_norm": 0.330132395029068, | |
| "learning_rate": 1.695095948827292e-05, | |
| "loss": 0.0118, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.3155650319829424, | |
| "grad_norm": 0.8517094850540161, | |
| "learning_rate": 1.6844349680170575e-05, | |
| "loss": 0.0146, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 3.326226012793177, | |
| "grad_norm": 3.9925179481506348, | |
| "learning_rate": 1.673773987206823e-05, | |
| "loss": 0.0309, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.3368869936034113, | |
| "grad_norm": 9.027678489685059, | |
| "learning_rate": 1.6631130063965885e-05, | |
| "loss": 0.0132, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 3.3475479744136463, | |
| "grad_norm": 1.5518391132354736, | |
| "learning_rate": 1.652452025586354e-05, | |
| "loss": 0.1057, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.3582089552238807, | |
| "grad_norm": 1.9203189611434937, | |
| "learning_rate": 1.6417910447761194e-05, | |
| "loss": 0.0399, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 3.368869936034115, | |
| "grad_norm": 1.2175391912460327, | |
| "learning_rate": 1.631130063965885e-05, | |
| "loss": 0.0186, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.3795309168443497, | |
| "grad_norm": 0.017526239156723022, | |
| "learning_rate": 1.6204690831556504e-05, | |
| "loss": 0.0028, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 3.390191897654584, | |
| "grad_norm": 0.017075370997190475, | |
| "learning_rate": 1.6098081023454158e-05, | |
| "loss": 0.0386, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 3.4008528784648187, | |
| "grad_norm": 4.834423542022705, | |
| "learning_rate": 1.5991471215351813e-05, | |
| "loss": 0.0793, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 3.411513859275053, | |
| "grad_norm": 6.423723220825195, | |
| "learning_rate": 1.5884861407249468e-05, | |
| "loss": 0.0574, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.4221748400852876, | |
| "grad_norm": 0.04012315720319748, | |
| "learning_rate": 1.5778251599147122e-05, | |
| "loss": 0.0489, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 3.4328358208955225, | |
| "grad_norm": 0.03658470883965492, | |
| "learning_rate": 1.5671641791044777e-05, | |
| "loss": 0.0359, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 3.443496801705757, | |
| "grad_norm": 0.22648349404335022, | |
| "learning_rate": 1.5565031982942432e-05, | |
| "loss": 0.0773, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 3.4541577825159915, | |
| "grad_norm": 4.281680107116699, | |
| "learning_rate": 1.5458422174840087e-05, | |
| "loss": 0.008, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.464818763326226, | |
| "grad_norm": 0.1702335923910141, | |
| "learning_rate": 1.535181236673774e-05, | |
| "loss": 0.0057, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 3.4754797441364604, | |
| "grad_norm": 0.7841653823852539, | |
| "learning_rate": 1.5245202558635396e-05, | |
| "loss": 0.0748, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 3.486140724946695, | |
| "grad_norm": 6.228107452392578, | |
| "learning_rate": 1.5138592750533051e-05, | |
| "loss": 0.0968, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 3.49680170575693, | |
| "grad_norm": 0.5722008347511292, | |
| "learning_rate": 1.5031982942430706e-05, | |
| "loss": 0.0283, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.5074626865671643, | |
| "grad_norm": 4.929506301879883, | |
| "learning_rate": 1.4925373134328357e-05, | |
| "loss": 0.0517, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 3.518123667377399, | |
| "grad_norm": 7.231541156768799, | |
| "learning_rate": 1.4818763326226012e-05, | |
| "loss": 0.0633, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.5287846481876333, | |
| "grad_norm": 0.042003732174634933, | |
| "learning_rate": 1.4712153518123666e-05, | |
| "loss": 0.0379, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 3.5394456289978677, | |
| "grad_norm": 1.0117156505584717, | |
| "learning_rate": 1.4605543710021321e-05, | |
| "loss": 0.0367, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 3.550106609808102, | |
| "grad_norm": 6.533594608306885, | |
| "learning_rate": 1.4498933901918976e-05, | |
| "loss": 0.0716, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 3.5607675906183367, | |
| "grad_norm": 2.0056521892547607, | |
| "learning_rate": 1.439232409381663e-05, | |
| "loss": 0.0227, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 5.68073034286499, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.1039, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 3.582089552238806, | |
| "grad_norm": 3.231276512145996, | |
| "learning_rate": 1.417910447761194e-05, | |
| "loss": 0.0637, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 3.5927505330490406, | |
| "grad_norm": 0.1591590791940689, | |
| "learning_rate": 1.4072494669509595e-05, | |
| "loss": 0.0273, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 3.603411513859275, | |
| "grad_norm": 3.1214847564697266, | |
| "learning_rate": 1.396588486140725e-05, | |
| "loss": 0.0252, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 3.6140724946695095, | |
| "grad_norm": 0.026124611496925354, | |
| "learning_rate": 1.3859275053304904e-05, | |
| "loss": 0.0352, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 3.624733475479744, | |
| "grad_norm": 3.268014669418335, | |
| "learning_rate": 1.3752665245202559e-05, | |
| "loss": 0.0644, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.635394456289979, | |
| "grad_norm": 9.879298210144043, | |
| "learning_rate": 1.3646055437100214e-05, | |
| "loss": 0.0609, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 3.6460554371002134, | |
| "grad_norm": 4.757223129272461, | |
| "learning_rate": 1.3539445628997869e-05, | |
| "loss": 0.1102, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 3.656716417910448, | |
| "grad_norm": 0.08892199397087097, | |
| "learning_rate": 1.3432835820895523e-05, | |
| "loss": 0.0415, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 3.6673773987206824, | |
| "grad_norm": 4.834870338439941, | |
| "learning_rate": 1.3326226012793178e-05, | |
| "loss": 0.047, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.678038379530917, | |
| "grad_norm": 5.872807502746582, | |
| "learning_rate": 1.3219616204690833e-05, | |
| "loss": 0.0566, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 3.6886993603411513, | |
| "grad_norm": 0.25904572010040283, | |
| "learning_rate": 1.3113006396588488e-05, | |
| "loss": 0.0395, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.699360341151386, | |
| "grad_norm": 0.14166103303432465, | |
| "learning_rate": 1.3006396588486142e-05, | |
| "loss": 0.0381, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 3.7100213219616203, | |
| "grad_norm": 5.544002056121826, | |
| "learning_rate": 1.2899786780383797e-05, | |
| "loss": 0.0284, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.7206823027718547, | |
| "grad_norm": 0.1531132161617279, | |
| "learning_rate": 1.2793176972281452e-05, | |
| "loss": 0.0322, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 3.7313432835820897, | |
| "grad_norm": 0.17648270726203918, | |
| "learning_rate": 1.2686567164179105e-05, | |
| "loss": 0.0417, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.742004264392324, | |
| "grad_norm": 4.241667747497559, | |
| "learning_rate": 1.257995735607676e-05, | |
| "loss": 0.0785, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 3.7526652452025586, | |
| "grad_norm": 0.14803919196128845, | |
| "learning_rate": 1.2473347547974414e-05, | |
| "loss": 0.0253, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.763326226012793, | |
| "grad_norm": 0.034351613372564316, | |
| "learning_rate": 1.2366737739872069e-05, | |
| "loss": 0.007, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 3.7739872068230276, | |
| "grad_norm": 0.20067091286182404, | |
| "learning_rate": 1.2260127931769722e-05, | |
| "loss": 0.0273, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.7846481876332625, | |
| "grad_norm": 3.871537446975708, | |
| "learning_rate": 1.2153518123667377e-05, | |
| "loss": 0.0153, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 3.795309168443497, | |
| "grad_norm": 0.03329772502183914, | |
| "learning_rate": 1.2046908315565032e-05, | |
| "loss": 0.0487, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.8059701492537314, | |
| "grad_norm": 0.3220955431461334, | |
| "learning_rate": 1.1940298507462686e-05, | |
| "loss": 0.0309, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 3.816631130063966, | |
| "grad_norm": 0.21222761273384094, | |
| "learning_rate": 1.1833688699360341e-05, | |
| "loss": 0.1011, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.8272921108742004, | |
| "grad_norm": 1.2007464170455933, | |
| "learning_rate": 1.1727078891257996e-05, | |
| "loss": 0.0415, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 3.837953091684435, | |
| "grad_norm": 0.031086280941963196, | |
| "learning_rate": 1.162046908315565e-05, | |
| "loss": 0.0502, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.8486140724946694, | |
| "grad_norm": 0.13002870976924896, | |
| "learning_rate": 1.1513859275053305e-05, | |
| "loss": 0.0043, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 3.859275053304904, | |
| "grad_norm": 3.307753324508667, | |
| "learning_rate": 1.140724946695096e-05, | |
| "loss": 0.0575, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.8699360341151388, | |
| "grad_norm": 4.746096134185791, | |
| "learning_rate": 1.1300639658848615e-05, | |
| "loss": 0.0174, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 3.8805970149253732, | |
| "grad_norm": 0.34301742911338806, | |
| "learning_rate": 1.119402985074627e-05, | |
| "loss": 0.0091, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.8912579957356077, | |
| "grad_norm": 0.145485058426857, | |
| "learning_rate": 1.1087420042643924e-05, | |
| "loss": 0.0021, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 3.901918976545842, | |
| "grad_norm": 0.8632763624191284, | |
| "learning_rate": 1.0980810234541579e-05, | |
| "loss": 0.0787, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.9125799573560767, | |
| "grad_norm": 2.403759002685547, | |
| "learning_rate": 1.0874200426439234e-05, | |
| "loss": 0.0351, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 3.923240938166311, | |
| "grad_norm": 2.637483835220337, | |
| "learning_rate": 1.0767590618336887e-05, | |
| "loss": 0.0577, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.933901918976546, | |
| "grad_norm": 0.49173247814178467, | |
| "learning_rate": 1.0660980810234541e-05, | |
| "loss": 0.0213, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 3.9445628997867805, | |
| "grad_norm": 2.3356168270111084, | |
| "learning_rate": 1.0554371002132196e-05, | |
| "loss": 0.0314, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.955223880597015, | |
| "grad_norm": 0.24166513979434967, | |
| "learning_rate": 1.0447761194029851e-05, | |
| "loss": 0.0098, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 3.9658848614072495, | |
| "grad_norm": 0.030654439702630043, | |
| "learning_rate": 1.0341151385927506e-05, | |
| "loss": 0.0073, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.976545842217484, | |
| "grad_norm": 0.6994831562042236, | |
| "learning_rate": 1.023454157782516e-05, | |
| "loss": 0.0385, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 3.9872068230277184, | |
| "grad_norm": 1.5712538957595825, | |
| "learning_rate": 1.0127931769722815e-05, | |
| "loss": 0.0695, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.997867803837953, | |
| "grad_norm": 0.3161895275115967, | |
| "learning_rate": 1.002132196162047e-05, | |
| "loss": 0.0457, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9405333333333333, | |
| "eval_loss": 0.248938649892807, | |
| "eval_runtime": 30.2343, | |
| "eval_samples_per_second": 124.032, | |
| "eval_steps_per_second": 3.903, | |
| "step": 1876 | |
| }, | |
| { | |
| "epoch": 4.008528784648187, | |
| "grad_norm": 0.08230875432491302, | |
| "learning_rate": 9.914712153518125e-06, | |
| "loss": 0.065, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.019189765458422, | |
| "grad_norm": 0.4110512435436249, | |
| "learning_rate": 9.80810234541578e-06, | |
| "loss": 0.027, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 4.029850746268656, | |
| "grad_norm": 0.3906515836715698, | |
| "learning_rate": 9.701492537313434e-06, | |
| "loss": 0.0155, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.040511727078891, | |
| "grad_norm": 7.8689703941345215, | |
| "learning_rate": 9.594882729211089e-06, | |
| "loss": 0.0087, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 4.051172707889126, | |
| "grad_norm": 0.07380155473947525, | |
| "learning_rate": 9.488272921108744e-06, | |
| "loss": 0.0264, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.061833688699361, | |
| "grad_norm": 0.10233091562986374, | |
| "learning_rate": 9.381663113006398e-06, | |
| "loss": 0.0257, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 4.072494669509595, | |
| "grad_norm": 0.007455846760421991, | |
| "learning_rate": 9.275053304904053e-06, | |
| "loss": 0.0397, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.08315565031983, | |
| "grad_norm": 0.07820025086402893, | |
| "learning_rate": 9.168443496801706e-06, | |
| "loss": 0.0206, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 4.093816631130064, | |
| "grad_norm": 0.3902164399623871, | |
| "learning_rate": 9.06183368869936e-06, | |
| "loss": 0.0327, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.104477611940299, | |
| "grad_norm": 1.3601645231246948, | |
| "learning_rate": 8.955223880597016e-06, | |
| "loss": 0.023, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 4.115138592750533, | |
| "grad_norm": 4.218597412109375, | |
| "learning_rate": 8.84861407249467e-06, | |
| "loss": 0.0855, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.1257995735607675, | |
| "grad_norm": 1.873428225517273, | |
| "learning_rate": 8.742004264392323e-06, | |
| "loss": 0.031, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 4.136460554371002, | |
| "grad_norm": 0.011160059832036495, | |
| "learning_rate": 8.635394456289978e-06, | |
| "loss": 0.0146, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.1471215351812365, | |
| "grad_norm": 2.47340989112854, | |
| "learning_rate": 8.528784648187633e-06, | |
| "loss": 0.0407, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 4.157782515991471, | |
| "grad_norm": 0.11140111088752747, | |
| "learning_rate": 8.422174840085288e-06, | |
| "loss": 0.0606, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.1684434968017055, | |
| "grad_norm": 0.38703852891921997, | |
| "learning_rate": 8.315565031982942e-06, | |
| "loss": 0.0389, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 4.17910447761194, | |
| "grad_norm": 4.00616455078125, | |
| "learning_rate": 8.208955223880597e-06, | |
| "loss": 0.11, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.189765458422174, | |
| "grad_norm": 5.303103446960449, | |
| "learning_rate": 8.102345415778252e-06, | |
| "loss": 0.0345, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 4.20042643923241, | |
| "grad_norm": 0.08131235092878342, | |
| "learning_rate": 7.995735607675907e-06, | |
| "loss": 0.0264, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 4.211087420042644, | |
| "grad_norm": 1.8881417512893677, | |
| "learning_rate": 7.889125799573561e-06, | |
| "loss": 0.0313, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 4.221748400852879, | |
| "grad_norm": 4.5652174949646, | |
| "learning_rate": 7.782515991471216e-06, | |
| "loss": 0.0157, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.232409381663113, | |
| "grad_norm": 5.838714599609375, | |
| "learning_rate": 7.67590618336887e-06, | |
| "loss": 0.0518, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 4.243070362473348, | |
| "grad_norm": 4.987056732177734, | |
| "learning_rate": 7.5692963752665255e-06, | |
| "loss": 0.053, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 4.253731343283582, | |
| "grad_norm": 1.3479896783828735, | |
| "learning_rate": 7.4626865671641785e-06, | |
| "loss": 0.0162, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 4.264392324093817, | |
| "grad_norm": 0.007616422604769468, | |
| "learning_rate": 7.356076759061833e-06, | |
| "loss": 0.0197, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.275053304904051, | |
| "grad_norm": 0.15424099564552307, | |
| "learning_rate": 7.249466950959488e-06, | |
| "loss": 0.0567, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 1.3741356134414673, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.0369, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 4.29637526652452, | |
| "grad_norm": 7.657044410705566, | |
| "learning_rate": 7.0362473347547975e-06, | |
| "loss": 0.0171, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 4.3070362473347545, | |
| "grad_norm": 1.2116847038269043, | |
| "learning_rate": 6.929637526652452e-06, | |
| "loss": 0.0277, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 4.317697228144989, | |
| "grad_norm": 1.8645787239074707, | |
| "learning_rate": 6.823027718550107e-06, | |
| "loss": 0.0413, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 4.3283582089552235, | |
| "grad_norm": 0.04453587904572487, | |
| "learning_rate": 6.716417910447762e-06, | |
| "loss": 0.0045, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 4.339019189765459, | |
| "grad_norm": 5.139947414398193, | |
| "learning_rate": 6.609808102345416e-06, | |
| "loss": 0.0667, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 4.349680170575693, | |
| "grad_norm": 0.1250516176223755, | |
| "learning_rate": 6.503198294243071e-06, | |
| "loss": 0.0088, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 4.360341151385928, | |
| "grad_norm": 1.8522542715072632, | |
| "learning_rate": 6.396588486140726e-06, | |
| "loss": 0.0343, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 4.371002132196162, | |
| "grad_norm": 9.212394714355469, | |
| "learning_rate": 6.28997867803838e-06, | |
| "loss": 0.0391, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.381663113006397, | |
| "grad_norm": 0.1429029256105423, | |
| "learning_rate": 6.1833688699360345e-06, | |
| "loss": 0.0777, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 4.392324093816631, | |
| "grad_norm": 2.789438486099243, | |
| "learning_rate": 6.076759061833688e-06, | |
| "loss": 0.0071, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 4.402985074626866, | |
| "grad_norm": 0.21160826086997986, | |
| "learning_rate": 5.970149253731343e-06, | |
| "loss": 0.0015, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 4.4136460554371, | |
| "grad_norm": 0.10126373916864395, | |
| "learning_rate": 5.863539445628998e-06, | |
| "loss": 0.0184, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 4.424307036247335, | |
| "grad_norm": 0.2786445617675781, | |
| "learning_rate": 5.756929637526653e-06, | |
| "loss": 0.0278, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 4.434968017057569, | |
| "grad_norm": 5.563600540161133, | |
| "learning_rate": 5.650319829424307e-06, | |
| "loss": 0.0416, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 4.445628997867804, | |
| "grad_norm": 1.6594213247299194, | |
| "learning_rate": 5.543710021321962e-06, | |
| "loss": 0.0204, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 4.456289978678038, | |
| "grad_norm": 1.33035147190094, | |
| "learning_rate": 5.437100213219617e-06, | |
| "loss": 0.0063, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 4.466950959488273, | |
| "grad_norm": 0.23442961275577545, | |
| "learning_rate": 5.330490405117271e-06, | |
| "loss": 0.0138, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 4.477611940298507, | |
| "grad_norm": 0.18316130340099335, | |
| "learning_rate": 5.2238805970149255e-06, | |
| "loss": 0.0043, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.4882729211087415, | |
| "grad_norm": 0.04730527475476265, | |
| "learning_rate": 5.11727078891258e-06, | |
| "loss": 0.0283, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 4.498933901918977, | |
| "grad_norm": 6.142653465270996, | |
| "learning_rate": 5.010660980810235e-06, | |
| "loss": 0.0531, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 4.509594882729211, | |
| "grad_norm": 0.4025548994541168, | |
| "learning_rate": 4.90405117270789e-06, | |
| "loss": 0.0302, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 4.520255863539446, | |
| "grad_norm": 0.3857446312904358, | |
| "learning_rate": 4.797441364605544e-06, | |
| "loss": 0.0327, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 4.53091684434968, | |
| "grad_norm": 0.35492250323295593, | |
| "learning_rate": 4.690831556503199e-06, | |
| "loss": 0.0334, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 4.541577825159915, | |
| "grad_norm": 1.0641695261001587, | |
| "learning_rate": 4.584221748400853e-06, | |
| "loss": 0.053, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 4.552238805970149, | |
| "grad_norm": 0.23556968569755554, | |
| "learning_rate": 4.477611940298508e-06, | |
| "loss": 0.0023, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 4.562899786780384, | |
| "grad_norm": 0.7794469594955444, | |
| "learning_rate": 4.371002132196162e-06, | |
| "loss": 0.0115, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 4.573560767590618, | |
| "grad_norm": 3.8487508296966553, | |
| "learning_rate": 4.264392324093816e-06, | |
| "loss": 0.0555, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 4.584221748400853, | |
| "grad_norm": 2.9853155612945557, | |
| "learning_rate": 4.157782515991471e-06, | |
| "loss": 0.0303, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 4.594882729211087, | |
| "grad_norm": 0.07582591474056244, | |
| "learning_rate": 4.051172707889126e-06, | |
| "loss": 0.0789, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 4.605543710021322, | |
| "grad_norm": 4.918933391571045, | |
| "learning_rate": 3.944562899786781e-06, | |
| "loss": 0.0281, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 4.616204690831556, | |
| "grad_norm": 0.03045099973678589, | |
| "learning_rate": 3.837953091684435e-06, | |
| "loss": 0.0445, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 4.6268656716417915, | |
| "grad_norm": 4.143336296081543, | |
| "learning_rate": 3.7313432835820893e-06, | |
| "loss": 0.0076, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 4.637526652452026, | |
| "grad_norm": 0.7964608669281006, | |
| "learning_rate": 3.624733475479744e-06, | |
| "loss": 0.0472, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 4.6481876332622605, | |
| "grad_norm": 8.989070892333984, | |
| "learning_rate": 3.5181236673773987e-06, | |
| "loss": 0.0702, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 4.658848614072495, | |
| "grad_norm": 4.960216045379639, | |
| "learning_rate": 3.4115138592750535e-06, | |
| "loss": 0.0376, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 4.669509594882729, | |
| "grad_norm": 0.028357909992337227, | |
| "learning_rate": 3.304904051172708e-06, | |
| "loss": 0.037, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 4.680170575692964, | |
| "grad_norm": 4.005814075469971, | |
| "learning_rate": 3.198294243070363e-06, | |
| "loss": 0.0412, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 4.690831556503198, | |
| "grad_norm": 1.183930516242981, | |
| "learning_rate": 3.0916844349680173e-06, | |
| "loss": 0.0219, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.701492537313433, | |
| "grad_norm": 3.5625252723693848, | |
| "learning_rate": 2.9850746268656716e-06, | |
| "loss": 0.0679, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 4.712153518123667, | |
| "grad_norm": 9.982269287109375, | |
| "learning_rate": 2.8784648187633263e-06, | |
| "loss": 0.0302, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 4.722814498933902, | |
| "grad_norm": 0.27800649404525757, | |
| "learning_rate": 2.771855010660981e-06, | |
| "loss": 0.0038, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 4.733475479744136, | |
| "grad_norm": 10.753087043762207, | |
| "learning_rate": 2.6652452025586354e-06, | |
| "loss": 0.0462, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 4.744136460554371, | |
| "grad_norm": 0.37115585803985596, | |
| "learning_rate": 2.55863539445629e-06, | |
| "loss": 0.0384, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 4.754797441364605, | |
| "grad_norm": 5.869413375854492, | |
| "learning_rate": 2.452025586353945e-06, | |
| "loss": 0.102, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 4.76545842217484, | |
| "grad_norm": 0.03743444383144379, | |
| "learning_rate": 2.3454157782515996e-06, | |
| "loss": 0.0141, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 4.776119402985074, | |
| "grad_norm": 0.22011743485927582, | |
| "learning_rate": 2.238805970149254e-06, | |
| "loss": 0.0016, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 4.786780383795309, | |
| "grad_norm": 1.3129013776779175, | |
| "learning_rate": 2.132196162046908e-06, | |
| "loss": 0.0335, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 4.797441364605544, | |
| "grad_norm": 0.3285354673862457, | |
| "learning_rate": 2.025586353944563e-06, | |
| "loss": 0.0341, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 4.8081023454157785, | |
| "grad_norm": 0.647816002368927, | |
| "learning_rate": 1.9189765458422177e-06, | |
| "loss": 0.0644, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 4.818763326226013, | |
| "grad_norm": 0.5337867140769958, | |
| "learning_rate": 1.812366737739872e-06, | |
| "loss": 0.0259, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 4.8294243070362475, | |
| "grad_norm": 0.8408359289169312, | |
| "learning_rate": 1.7057569296375267e-06, | |
| "loss": 0.0341, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 4.840085287846482, | |
| "grad_norm": 1.0795294046401978, | |
| "learning_rate": 1.5991471215351815e-06, | |
| "loss": 0.058, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 4.850746268656716, | |
| "grad_norm": 0.26184937357902527, | |
| "learning_rate": 1.4925373134328358e-06, | |
| "loss": 0.0354, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 4.861407249466951, | |
| "grad_norm": 4.989986896514893, | |
| "learning_rate": 1.3859275053304905e-06, | |
| "loss": 0.0631, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 4.872068230277185, | |
| "grad_norm": 0.915327250957489, | |
| "learning_rate": 1.279317697228145e-06, | |
| "loss": 0.0753, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 4.88272921108742, | |
| "grad_norm": 5.43763542175293, | |
| "learning_rate": 1.1727078891257998e-06, | |
| "loss": 0.0314, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 4.893390191897654, | |
| "grad_norm": 4.289970397949219, | |
| "learning_rate": 1.066098081023454e-06, | |
| "loss": 0.0697, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 4.904051172707889, | |
| "grad_norm": 3.189176321029663, | |
| "learning_rate": 9.594882729211088e-07, | |
| "loss": 0.0322, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.914712153518123, | |
| "grad_norm": 4.036993026733398, | |
| "learning_rate": 8.528784648187634e-07, | |
| "loss": 0.0548, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 4.925373134328359, | |
| "grad_norm": 0.1266753226518631, | |
| "learning_rate": 7.462686567164179e-07, | |
| "loss": 0.0592, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 4.936034115138593, | |
| "grad_norm": 0.2614297568798065, | |
| "learning_rate": 6.396588486140725e-07, | |
| "loss": 0.0352, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 4.946695095948828, | |
| "grad_norm": 6.704292297363281, | |
| "learning_rate": 5.33049040511727e-07, | |
| "loss": 0.0324, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 4.957356076759062, | |
| "grad_norm": 0.07638683915138245, | |
| "learning_rate": 4.264392324093817e-07, | |
| "loss": 0.0521, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 4.968017057569297, | |
| "grad_norm": 0.26516252756118774, | |
| "learning_rate": 3.1982942430703626e-07, | |
| "loss": 0.0019, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 4.978678038379531, | |
| "grad_norm": 0.18425673246383667, | |
| "learning_rate": 2.1321961620469084e-07, | |
| "loss": 0.029, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 4.9893390191897655, | |
| "grad_norm": 0.11901337653398514, | |
| "learning_rate": 1.0660980810234542e-07, | |
| "loss": 0.0018, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.99820876121521, | |
| "learning_rate": 0.0, | |
| "loss": 0.017, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9469333333333333, | |
| "eval_loss": 0.2371838539838791, | |
| "eval_runtime": 30.1915, | |
| "eval_samples_per_second": 124.207, | |
| "eval_steps_per_second": 3.908, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 2345, | |
| "total_flos": 5.8118992210944e+18, | |
| "train_loss": 0.05562105006290111, | |
| "train_runtime": 1504.1424, | |
| "train_samples_per_second": 49.862, | |
| "train_steps_per_second": 1.559 | |
| } | |
| ], | |
| "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 | |
| } | |