xuanxuan / trainer_state.json
Mikivis's picture
commit from mk
ba6ae2b
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"global_step": 7205,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 1.9958362248438587e-05,
"loss": 2.8983,
"step": 20
},
{
"epoch": 0.03,
"learning_rate": 1.99028452463567e-05,
"loss": 2.548,
"step": 40
},
{
"epoch": 0.04,
"learning_rate": 1.984732824427481e-05,
"loss": 2.4594,
"step": 60
},
{
"epoch": 0.06,
"learning_rate": 1.9791811242192923e-05,
"loss": 2.3628,
"step": 80
},
{
"epoch": 0.07,
"learning_rate": 1.9736294240111035e-05,
"loss": 2.3952,
"step": 100
},
{
"epoch": 0.08,
"learning_rate": 1.9680777238029147e-05,
"loss": 2.3894,
"step": 120
},
{
"epoch": 0.1,
"learning_rate": 1.9625260235947262e-05,
"loss": 2.3013,
"step": 140
},
{
"epoch": 0.11,
"learning_rate": 1.9569743233865374e-05,
"loss": 2.3323,
"step": 160
},
{
"epoch": 0.12,
"learning_rate": 1.9514226231783486e-05,
"loss": 2.2816,
"step": 180
},
{
"epoch": 0.14,
"learning_rate": 1.9458709229701598e-05,
"loss": 2.3024,
"step": 200
},
{
"epoch": 0.15,
"learning_rate": 1.940319222761971e-05,
"loss": 2.2957,
"step": 220
},
{
"epoch": 0.17,
"learning_rate": 1.9347675225537822e-05,
"loss": 2.2605,
"step": 240
},
{
"epoch": 0.18,
"learning_rate": 1.9292158223455934e-05,
"loss": 2.2591,
"step": 260
},
{
"epoch": 0.19,
"learning_rate": 1.923664122137405e-05,
"loss": 2.2737,
"step": 280
},
{
"epoch": 0.21,
"learning_rate": 1.918112421929216e-05,
"loss": 2.2776,
"step": 300
},
{
"epoch": 0.22,
"learning_rate": 1.912560721721027e-05,
"loss": 2.2052,
"step": 320
},
{
"epoch": 0.24,
"learning_rate": 1.9070090215128386e-05,
"loss": 2.2548,
"step": 340
},
{
"epoch": 0.25,
"learning_rate": 1.9014573213046498e-05,
"loss": 2.246,
"step": 360
},
{
"epoch": 0.26,
"learning_rate": 1.895905621096461e-05,
"loss": 2.2273,
"step": 380
},
{
"epoch": 0.28,
"learning_rate": 1.8903539208882722e-05,
"loss": 2.2351,
"step": 400
},
{
"epoch": 0.29,
"learning_rate": 1.8848022206800834e-05,
"loss": 2.234,
"step": 420
},
{
"epoch": 0.31,
"learning_rate": 1.8792505204718946e-05,
"loss": 2.2372,
"step": 440
},
{
"epoch": 0.32,
"learning_rate": 1.8736988202637058e-05,
"loss": 2.1749,
"step": 460
},
{
"epoch": 0.33,
"learning_rate": 1.8681471200555174e-05,
"loss": 2.1885,
"step": 480
},
{
"epoch": 0.35,
"learning_rate": 1.8625954198473282e-05,
"loss": 2.1873,
"step": 500
},
{
"epoch": 0.36,
"learning_rate": 1.8570437196391394e-05,
"loss": 2.2094,
"step": 520
},
{
"epoch": 0.37,
"learning_rate": 1.851492019430951e-05,
"loss": 2.1991,
"step": 540
},
{
"epoch": 0.39,
"learning_rate": 1.8459403192227622e-05,
"loss": 2.1826,
"step": 560
},
{
"epoch": 0.4,
"learning_rate": 1.8403886190145734e-05,
"loss": 2.2183,
"step": 580
},
{
"epoch": 0.42,
"learning_rate": 1.8348369188063846e-05,
"loss": 2.1631,
"step": 600
},
{
"epoch": 0.43,
"learning_rate": 1.8292852185981958e-05,
"loss": 2.2094,
"step": 620
},
{
"epoch": 0.44,
"learning_rate": 1.823733518390007e-05,
"loss": 2.2029,
"step": 640
},
{
"epoch": 0.46,
"learning_rate": 1.8181818181818182e-05,
"loss": 2.1485,
"step": 660
},
{
"epoch": 0.47,
"learning_rate": 1.8126301179736298e-05,
"loss": 2.1773,
"step": 680
},
{
"epoch": 0.49,
"learning_rate": 1.807078417765441e-05,
"loss": 2.1284,
"step": 700
},
{
"epoch": 0.5,
"learning_rate": 1.8015267175572518e-05,
"loss": 2.1754,
"step": 720
},
{
"epoch": 0.51,
"learning_rate": 1.7959750173490634e-05,
"loss": 2.208,
"step": 740
},
{
"epoch": 0.53,
"learning_rate": 1.7904233171408746e-05,
"loss": 2.1423,
"step": 760
},
{
"epoch": 0.54,
"learning_rate": 1.7848716169326858e-05,
"loss": 2.1489,
"step": 780
},
{
"epoch": 0.56,
"learning_rate": 1.779319916724497e-05,
"loss": 2.1808,
"step": 800
},
{
"epoch": 0.57,
"learning_rate": 1.7737682165163085e-05,
"loss": 2.1837,
"step": 820
},
{
"epoch": 0.58,
"learning_rate": 1.7682165163081194e-05,
"loss": 2.135,
"step": 840
},
{
"epoch": 0.6,
"learning_rate": 1.7626648160999306e-05,
"loss": 2.1493,
"step": 860
},
{
"epoch": 0.61,
"learning_rate": 1.757113115891742e-05,
"loss": 2.1963,
"step": 880
},
{
"epoch": 0.62,
"learning_rate": 1.7515614156835533e-05,
"loss": 2.1365,
"step": 900
},
{
"epoch": 0.64,
"learning_rate": 1.7460097154753645e-05,
"loss": 2.1524,
"step": 920
},
{
"epoch": 0.65,
"learning_rate": 1.7404580152671757e-05,
"loss": 2.1455,
"step": 940
},
{
"epoch": 0.67,
"learning_rate": 1.734906315058987e-05,
"loss": 2.1359,
"step": 960
},
{
"epoch": 0.68,
"learning_rate": 1.729354614850798e-05,
"loss": 2.1394,
"step": 980
},
{
"epoch": 0.69,
"learning_rate": 1.7238029146426094e-05,
"loss": 2.1559,
"step": 1000
},
{
"epoch": 0.71,
"learning_rate": 1.718251214434421e-05,
"loss": 2.1819,
"step": 1020
},
{
"epoch": 0.72,
"learning_rate": 1.7126995142262318e-05,
"loss": 2.1547,
"step": 1040
},
{
"epoch": 0.74,
"learning_rate": 1.707147814018043e-05,
"loss": 2.1226,
"step": 1060
},
{
"epoch": 0.75,
"learning_rate": 1.7015961138098545e-05,
"loss": 2.0807,
"step": 1080
},
{
"epoch": 0.76,
"learning_rate": 1.6960444136016657e-05,
"loss": 2.1243,
"step": 1100
},
{
"epoch": 0.78,
"learning_rate": 1.690492713393477e-05,
"loss": 2.1298,
"step": 1120
},
{
"epoch": 0.79,
"learning_rate": 1.684941013185288e-05,
"loss": 2.1265,
"step": 1140
},
{
"epoch": 0.8,
"learning_rate": 1.6793893129770993e-05,
"loss": 2.1234,
"step": 1160
},
{
"epoch": 0.82,
"learning_rate": 1.6738376127689105e-05,
"loss": 2.1243,
"step": 1180
},
{
"epoch": 0.83,
"learning_rate": 1.6682859125607217e-05,
"loss": 2.1369,
"step": 1200
},
{
"epoch": 0.85,
"learning_rate": 1.6627342123525333e-05,
"loss": 2.1074,
"step": 1220
},
{
"epoch": 0.86,
"learning_rate": 1.6571825121443445e-05,
"loss": 2.1411,
"step": 1240
},
{
"epoch": 0.87,
"learning_rate": 1.6516308119361554e-05,
"loss": 2.12,
"step": 1260
},
{
"epoch": 0.89,
"learning_rate": 1.646079111727967e-05,
"loss": 2.0964,
"step": 1280
},
{
"epoch": 0.9,
"learning_rate": 1.640527411519778e-05,
"loss": 2.0956,
"step": 1300
},
{
"epoch": 0.92,
"learning_rate": 1.6349757113115893e-05,
"loss": 2.1189,
"step": 1320
},
{
"epoch": 0.93,
"learning_rate": 1.6294240111034005e-05,
"loss": 2.1288,
"step": 1340
},
{
"epoch": 0.94,
"learning_rate": 1.6238723108952117e-05,
"loss": 2.0694,
"step": 1360
},
{
"epoch": 0.96,
"learning_rate": 1.618320610687023e-05,
"loss": 2.155,
"step": 1380
},
{
"epoch": 0.97,
"learning_rate": 1.612768910478834e-05,
"loss": 2.0934,
"step": 1400
},
{
"epoch": 0.99,
"learning_rate": 1.6072172102706457e-05,
"loss": 2.1193,
"step": 1420
},
{
"epoch": 1.0,
"learning_rate": 1.6019430950728662e-05,
"loss": 2.157,
"step": 1440
},
{
"epoch": 1.01,
"learning_rate": 1.5963913948646774e-05,
"loss": 2.0529,
"step": 1460
},
{
"epoch": 1.03,
"learning_rate": 1.5908396946564886e-05,
"loss": 2.0998,
"step": 1480
},
{
"epoch": 1.04,
"learning_rate": 1.5852879944482998e-05,
"loss": 2.0727,
"step": 1500
},
{
"epoch": 1.05,
"learning_rate": 1.579736294240111e-05,
"loss": 2.1137,
"step": 1520
},
{
"epoch": 1.07,
"learning_rate": 1.5741845940319226e-05,
"loss": 2.0415,
"step": 1540
},
{
"epoch": 1.08,
"learning_rate": 1.5686328938237338e-05,
"loss": 2.0903,
"step": 1560
},
{
"epoch": 1.1,
"learning_rate": 1.5630811936155446e-05,
"loss": 2.0646,
"step": 1580
},
{
"epoch": 1.11,
"learning_rate": 1.557529493407356e-05,
"loss": 2.093,
"step": 1600
},
{
"epoch": 1.12,
"learning_rate": 1.5519777931991674e-05,
"loss": 2.0981,
"step": 1620
},
{
"epoch": 1.14,
"learning_rate": 1.5464260929909786e-05,
"loss": 2.0702,
"step": 1640
},
{
"epoch": 1.15,
"learning_rate": 1.5408743927827898e-05,
"loss": 2.0566,
"step": 1660
},
{
"epoch": 1.17,
"learning_rate": 1.5353226925746013e-05,
"loss": 2.1073,
"step": 1680
},
{
"epoch": 1.18,
"learning_rate": 1.5297709923664122e-05,
"loss": 2.0682,
"step": 1700
},
{
"epoch": 1.19,
"learning_rate": 1.5242192921582236e-05,
"loss": 2.0639,
"step": 1720
},
{
"epoch": 1.21,
"learning_rate": 1.518667591950035e-05,
"loss": 2.0975,
"step": 1740
},
{
"epoch": 1.22,
"learning_rate": 1.513115891741846e-05,
"loss": 2.068,
"step": 1760
},
{
"epoch": 1.24,
"learning_rate": 1.5075641915336573e-05,
"loss": 2.0548,
"step": 1780
},
{
"epoch": 1.25,
"learning_rate": 1.5020124913254685e-05,
"loss": 2.0739,
"step": 1800
},
{
"epoch": 1.26,
"learning_rate": 1.4964607911172798e-05,
"loss": 2.009,
"step": 1820
},
{
"epoch": 1.28,
"learning_rate": 1.4909090909090911e-05,
"loss": 2.0673,
"step": 1840
},
{
"epoch": 1.29,
"learning_rate": 1.4853573907009022e-05,
"loss": 2.0708,
"step": 1860
},
{
"epoch": 1.3,
"learning_rate": 1.4798056904927135e-05,
"loss": 2.0457,
"step": 1880
},
{
"epoch": 1.32,
"learning_rate": 1.4742539902845247e-05,
"loss": 2.0952,
"step": 1900
},
{
"epoch": 1.33,
"learning_rate": 1.468702290076336e-05,
"loss": 2.0415,
"step": 1920
},
{
"epoch": 1.35,
"learning_rate": 1.4631505898681473e-05,
"loss": 2.0762,
"step": 1940
},
{
"epoch": 1.36,
"learning_rate": 1.4575988896599585e-05,
"loss": 2.0755,
"step": 1960
},
{
"epoch": 1.37,
"learning_rate": 1.4520471894517699e-05,
"loss": 2.0555,
"step": 1980
},
{
"epoch": 1.39,
"learning_rate": 1.446495489243581e-05,
"loss": 2.0714,
"step": 2000
},
{
"epoch": 1.4,
"learning_rate": 1.4409437890353921e-05,
"loss": 2.055,
"step": 2020
},
{
"epoch": 1.42,
"learning_rate": 1.4353920888272035e-05,
"loss": 2.0359,
"step": 2040
},
{
"epoch": 1.43,
"learning_rate": 1.4298403886190147e-05,
"loss": 2.0178,
"step": 2060
},
{
"epoch": 1.44,
"learning_rate": 1.4242886884108261e-05,
"loss": 1.9921,
"step": 2080
},
{
"epoch": 1.46,
"learning_rate": 1.4187369882026371e-05,
"loss": 2.0643,
"step": 2100
},
{
"epoch": 1.47,
"learning_rate": 1.4131852879944483e-05,
"loss": 2.055,
"step": 2120
},
{
"epoch": 1.49,
"learning_rate": 1.4076335877862597e-05,
"loss": 2.0363,
"step": 2140
},
{
"epoch": 1.5,
"learning_rate": 1.4020818875780709e-05,
"loss": 2.0468,
"step": 2160
},
{
"epoch": 1.51,
"learning_rate": 1.3965301873698823e-05,
"loss": 2.0616,
"step": 2180
},
{
"epoch": 1.53,
"learning_rate": 1.3909784871616933e-05,
"loss": 2.0377,
"step": 2200
},
{
"epoch": 1.54,
"learning_rate": 1.3854267869535045e-05,
"loss": 2.0459,
"step": 2220
},
{
"epoch": 1.55,
"learning_rate": 1.3798750867453159e-05,
"loss": 2.0649,
"step": 2240
},
{
"epoch": 1.57,
"learning_rate": 1.3743233865371271e-05,
"loss": 2.0639,
"step": 2260
},
{
"epoch": 1.58,
"learning_rate": 1.3687716863289385e-05,
"loss": 2.045,
"step": 2280
},
{
"epoch": 1.6,
"learning_rate": 1.3632199861207495e-05,
"loss": 2.0513,
"step": 2300
},
{
"epoch": 1.61,
"learning_rate": 1.3576682859125607e-05,
"loss": 2.0429,
"step": 2320
},
{
"epoch": 1.62,
"learning_rate": 1.3523941707147814e-05,
"loss": 2.0545,
"step": 2340
},
{
"epoch": 1.64,
"learning_rate": 1.3468424705065928e-05,
"loss": 2.0548,
"step": 2360
},
{
"epoch": 1.65,
"learning_rate": 1.341290770298404e-05,
"loss": 2.071,
"step": 2380
},
{
"epoch": 1.67,
"learning_rate": 1.3357390700902154e-05,
"loss": 2.0398,
"step": 2400
},
{
"epoch": 1.68,
"learning_rate": 1.3301873698820264e-05,
"loss": 2.0098,
"step": 2420
},
{
"epoch": 1.69,
"learning_rate": 1.3246356696738378e-05,
"loss": 2.0455,
"step": 2440
},
{
"epoch": 1.71,
"learning_rate": 1.319083969465649e-05,
"loss": 2.0545,
"step": 2460
},
{
"epoch": 1.72,
"learning_rate": 1.3135322692574602e-05,
"loss": 2.0368,
"step": 2480
},
{
"epoch": 1.73,
"learning_rate": 1.3079805690492715e-05,
"loss": 2.0542,
"step": 2500
},
{
"epoch": 1.75,
"learning_rate": 1.3024288688410826e-05,
"loss": 2.0422,
"step": 2520
},
{
"epoch": 1.76,
"learning_rate": 1.296877168632894e-05,
"loss": 2.0206,
"step": 2540
},
{
"epoch": 1.78,
"learning_rate": 1.2913254684247052e-05,
"loss": 2.043,
"step": 2560
},
{
"epoch": 1.79,
"learning_rate": 1.2857737682165164e-05,
"loss": 2.0716,
"step": 2580
},
{
"epoch": 1.8,
"learning_rate": 1.2802220680083277e-05,
"loss": 2.0472,
"step": 2600
},
{
"epoch": 1.82,
"learning_rate": 1.274670367800139e-05,
"loss": 2.0248,
"step": 2620
},
{
"epoch": 1.83,
"learning_rate": 1.2691186675919503e-05,
"loss": 2.0106,
"step": 2640
},
{
"epoch": 1.85,
"learning_rate": 1.2635669673837614e-05,
"loss": 2.0243,
"step": 2660
},
{
"epoch": 1.86,
"learning_rate": 1.2580152671755726e-05,
"loss": 2.0447,
"step": 2680
},
{
"epoch": 1.87,
"learning_rate": 1.252463566967384e-05,
"loss": 2.037,
"step": 2700
},
{
"epoch": 1.89,
"learning_rate": 1.2469118667591951e-05,
"loss": 2.0408,
"step": 2720
},
{
"epoch": 1.9,
"learning_rate": 1.2413601665510065e-05,
"loss": 2.0394,
"step": 2740
},
{
"epoch": 1.92,
"learning_rate": 1.2358084663428175e-05,
"loss": 2.0685,
"step": 2760
},
{
"epoch": 1.93,
"learning_rate": 1.2302567661346287e-05,
"loss": 2.0259,
"step": 2780
},
{
"epoch": 1.94,
"learning_rate": 1.2247050659264401e-05,
"loss": 2.0026,
"step": 2800
},
{
"epoch": 1.96,
"learning_rate": 1.2191533657182513e-05,
"loss": 2.0251,
"step": 2820
},
{
"epoch": 1.97,
"learning_rate": 1.2136016655100627e-05,
"loss": 2.0177,
"step": 2840
},
{
"epoch": 1.98,
"learning_rate": 1.2080499653018737e-05,
"loss": 2.0575,
"step": 2860
},
{
"epoch": 2.0,
"learning_rate": 1.202498265093685e-05,
"loss": 2.0655,
"step": 2880
},
{
"epoch": 2.01,
"learning_rate": 1.1969465648854963e-05,
"loss": 2.0154,
"step": 2900
},
{
"epoch": 2.03,
"learning_rate": 1.1913948646773075e-05,
"loss": 2.0024,
"step": 2920
},
{
"epoch": 2.04,
"learning_rate": 1.1858431644691189e-05,
"loss": 2.0017,
"step": 2940
},
{
"epoch": 2.05,
"learning_rate": 1.18029146426093e-05,
"loss": 1.9705,
"step": 2960
},
{
"epoch": 2.07,
"learning_rate": 1.1747397640527411e-05,
"loss": 2.0569,
"step": 2980
},
{
"epoch": 2.08,
"learning_rate": 1.1691880638445525e-05,
"loss": 2.0121,
"step": 3000
},
{
"epoch": 2.1,
"learning_rate": 1.1636363636363637e-05,
"loss": 1.9899,
"step": 3020
},
{
"epoch": 2.11,
"learning_rate": 1.158084663428175e-05,
"loss": 1.9694,
"step": 3040
},
{
"epoch": 2.12,
"learning_rate": 1.1525329632199863e-05,
"loss": 2.0268,
"step": 3060
},
{
"epoch": 2.14,
"learning_rate": 1.1469812630117973e-05,
"loss": 1.9967,
"step": 3080
},
{
"epoch": 2.15,
"learning_rate": 1.1414295628036087e-05,
"loss": 2.0055,
"step": 3100
},
{
"epoch": 2.17,
"learning_rate": 1.1358778625954199e-05,
"loss": 1.9879,
"step": 3120
},
{
"epoch": 2.18,
"learning_rate": 1.1303261623872313e-05,
"loss": 1.9885,
"step": 3140
},
{
"epoch": 2.19,
"learning_rate": 1.1247744621790425e-05,
"loss": 2.0225,
"step": 3160
},
{
"epoch": 2.21,
"learning_rate": 1.1192227619708535e-05,
"loss": 2.0058,
"step": 3180
},
{
"epoch": 2.22,
"learning_rate": 1.1136710617626649e-05,
"loss": 2.0204,
"step": 3200
},
{
"epoch": 2.23,
"learning_rate": 1.1081193615544761e-05,
"loss": 1.983,
"step": 3220
},
{
"epoch": 2.25,
"learning_rate": 1.1025676613462875e-05,
"loss": 1.9868,
"step": 3240
},
{
"epoch": 2.26,
"learning_rate": 1.0970159611380987e-05,
"loss": 2.0118,
"step": 3260
},
{
"epoch": 2.28,
"learning_rate": 1.09146426092991e-05,
"loss": 2.0158,
"step": 3280
},
{
"epoch": 2.29,
"learning_rate": 1.085912560721721e-05,
"loss": 2.029,
"step": 3300
},
{
"epoch": 2.3,
"learning_rate": 1.0803608605135323e-05,
"loss": 1.9814,
"step": 3320
},
{
"epoch": 2.32,
"learning_rate": 1.0748091603053437e-05,
"loss": 2.0236,
"step": 3340
},
{
"epoch": 2.33,
"learning_rate": 1.0692574600971549e-05,
"loss": 1.9839,
"step": 3360
},
{
"epoch": 2.35,
"learning_rate": 1.0637057598889662e-05,
"loss": 2.0041,
"step": 3380
},
{
"epoch": 2.36,
"learning_rate": 1.0581540596807773e-05,
"loss": 2.0303,
"step": 3400
},
{
"epoch": 2.37,
"learning_rate": 1.0526023594725885e-05,
"loss": 1.9959,
"step": 3420
},
{
"epoch": 2.39,
"learning_rate": 1.0470506592643998e-05,
"loss": 1.9582,
"step": 3440
},
{
"epoch": 2.4,
"learning_rate": 1.041498959056211e-05,
"loss": 2.0094,
"step": 3460
},
{
"epoch": 2.41,
"learning_rate": 1.0359472588480224e-05,
"loss": 2.0021,
"step": 3480
},
{
"epoch": 2.43,
"learning_rate": 1.0303955586398335e-05,
"loss": 1.9736,
"step": 3500
},
{
"epoch": 2.44,
"learning_rate": 1.0248438584316447e-05,
"loss": 2.0042,
"step": 3520
},
{
"epoch": 2.46,
"learning_rate": 1.019292158223456e-05,
"loss": 2.0099,
"step": 3540
},
{
"epoch": 2.47,
"learning_rate": 1.0137404580152672e-05,
"loss": 2.0219,
"step": 3560
},
{
"epoch": 2.48,
"learning_rate": 1.0081887578070786e-05,
"loss": 1.984,
"step": 3580
},
{
"epoch": 2.5,
"learning_rate": 1.0026370575988898e-05,
"loss": 1.9987,
"step": 3600
},
{
"epoch": 2.51,
"learning_rate": 9.97085357390701e-06,
"loss": 1.9931,
"step": 3620
},
{
"epoch": 2.53,
"learning_rate": 9.915336571825122e-06,
"loss": 1.9698,
"step": 3640
},
{
"epoch": 2.54,
"learning_rate": 9.859819569743234e-06,
"loss": 1.9966,
"step": 3660
},
{
"epoch": 2.55,
"learning_rate": 9.804302567661346e-06,
"loss": 1.9893,
"step": 3680
},
{
"epoch": 2.57,
"learning_rate": 9.74878556557946e-06,
"loss": 1.9571,
"step": 3700
},
{
"epoch": 2.58,
"learning_rate": 9.693268563497572e-06,
"loss": 1.9974,
"step": 3720
},
{
"epoch": 2.6,
"learning_rate": 9.637751561415684e-06,
"loss": 1.9828,
"step": 3740
},
{
"epoch": 2.61,
"learning_rate": 9.582234559333796e-06,
"loss": 1.971,
"step": 3760
},
{
"epoch": 2.62,
"learning_rate": 9.526717557251908e-06,
"loss": 1.9799,
"step": 3780
},
{
"epoch": 2.64,
"learning_rate": 9.471200555170022e-06,
"loss": 1.9926,
"step": 3800
},
{
"epoch": 2.65,
"learning_rate": 9.415683553088134e-06,
"loss": 1.9738,
"step": 3820
},
{
"epoch": 2.66,
"learning_rate": 9.360166551006246e-06,
"loss": 1.9917,
"step": 3840
},
{
"epoch": 2.68,
"learning_rate": 9.30464954892436e-06,
"loss": 2.0237,
"step": 3860
},
{
"epoch": 2.69,
"learning_rate": 9.24913254684247e-06,
"loss": 2.0106,
"step": 3880
},
{
"epoch": 2.71,
"learning_rate": 9.193615544760584e-06,
"loss": 2.0059,
"step": 3900
},
{
"epoch": 2.72,
"learning_rate": 9.138098542678696e-06,
"loss": 2.0034,
"step": 3920
},
{
"epoch": 2.73,
"learning_rate": 9.082581540596808e-06,
"loss": 2.0007,
"step": 3940
},
{
"epoch": 2.75,
"learning_rate": 9.027064538514922e-06,
"loss": 1.9945,
"step": 3960
},
{
"epoch": 2.76,
"learning_rate": 8.971547536433032e-06,
"loss": 1.9635,
"step": 3980
},
{
"epoch": 2.78,
"learning_rate": 8.916030534351146e-06,
"loss": 2.0221,
"step": 4000
},
{
"epoch": 2.79,
"learning_rate": 8.860513532269258e-06,
"loss": 1.9817,
"step": 4020
},
{
"epoch": 2.8,
"learning_rate": 8.804996530187372e-06,
"loss": 1.9781,
"step": 4040
},
{
"epoch": 2.82,
"learning_rate": 8.749479528105484e-06,
"loss": 2.0199,
"step": 4060
},
{
"epoch": 2.83,
"learning_rate": 8.693962526023596e-06,
"loss": 1.9616,
"step": 4080
},
{
"epoch": 2.85,
"learning_rate": 8.638445523941708e-06,
"loss": 1.9641,
"step": 4100
},
{
"epoch": 2.86,
"learning_rate": 8.58292852185982e-06,
"loss": 2.0155,
"step": 4120
},
{
"epoch": 2.87,
"learning_rate": 8.527411519777934e-06,
"loss": 1.978,
"step": 4140
},
{
"epoch": 2.89,
"learning_rate": 8.471894517696046e-06,
"loss": 1.9652,
"step": 4160
},
{
"epoch": 2.9,
"learning_rate": 8.416377515614158e-06,
"loss": 1.9862,
"step": 4180
},
{
"epoch": 2.91,
"learning_rate": 8.36086051353227e-06,
"loss": 2.0246,
"step": 4200
},
{
"epoch": 2.93,
"learning_rate": 8.305343511450382e-06,
"loss": 1.9726,
"step": 4220
},
{
"epoch": 2.94,
"learning_rate": 8.249826509368495e-06,
"loss": 2.0039,
"step": 4240
},
{
"epoch": 2.96,
"learning_rate": 8.194309507286607e-06,
"loss": 1.9942,
"step": 4260
},
{
"epoch": 2.97,
"learning_rate": 8.13879250520472e-06,
"loss": 2.0113,
"step": 4280
},
{
"epoch": 2.98,
"learning_rate": 8.083275503122832e-06,
"loss": 1.9691,
"step": 4300
},
{
"epoch": 3.0,
"learning_rate": 8.027758501040944e-06,
"loss": 1.994,
"step": 4320
},
{
"epoch": 3.01,
"learning_rate": 7.972241498959057e-06,
"loss": 2.0017,
"step": 4340
},
{
"epoch": 3.03,
"learning_rate": 7.91672449687717e-06,
"loss": 1.9494,
"step": 4360
},
{
"epoch": 3.04,
"learning_rate": 7.863983344899376e-06,
"loss": 1.9718,
"step": 4380
},
{
"epoch": 3.05,
"learning_rate": 7.808466342817488e-06,
"loss": 1.9305,
"step": 4400
},
{
"epoch": 3.07,
"learning_rate": 7.7529493407356e-06,
"loss": 1.9697,
"step": 4420
},
{
"epoch": 3.08,
"learning_rate": 7.697432338653712e-06,
"loss": 1.9658,
"step": 4440
},
{
"epoch": 3.1,
"learning_rate": 7.641915336571826e-06,
"loss": 1.9424,
"step": 4460
},
{
"epoch": 3.11,
"learning_rate": 7.586398334489938e-06,
"loss": 1.9312,
"step": 4480
},
{
"epoch": 3.12,
"learning_rate": 7.530881332408051e-06,
"loss": 2.0042,
"step": 4500
},
{
"epoch": 3.14,
"learning_rate": 7.475364330326162e-06,
"loss": 1.9868,
"step": 4520
},
{
"epoch": 3.15,
"learning_rate": 7.419847328244275e-06,
"loss": 1.9679,
"step": 4540
},
{
"epoch": 3.16,
"learning_rate": 7.364330326162388e-06,
"loss": 1.9448,
"step": 4560
},
{
"epoch": 3.18,
"learning_rate": 7.3088133240805e-06,
"loss": 1.988,
"step": 4580
},
{
"epoch": 3.19,
"learning_rate": 7.253296321998613e-06,
"loss": 1.9676,
"step": 4600
},
{
"epoch": 3.21,
"learning_rate": 7.197779319916726e-06,
"loss": 1.9859,
"step": 4620
},
{
"epoch": 3.22,
"learning_rate": 7.142262317834837e-06,
"loss": 1.9846,
"step": 4640
},
{
"epoch": 3.23,
"learning_rate": 7.08674531575295e-06,
"loss": 1.9827,
"step": 4660
},
{
"epoch": 3.25,
"learning_rate": 7.031228313671062e-06,
"loss": 1.9648,
"step": 4680
},
{
"epoch": 3.26,
"learning_rate": 6.975711311589175e-06,
"loss": 1.9662,
"step": 4700
},
{
"epoch": 3.28,
"learning_rate": 6.920194309507288e-06,
"loss": 1.9911,
"step": 4720
},
{
"epoch": 3.29,
"learning_rate": 6.864677307425399e-06,
"loss": 2.0097,
"step": 4740
},
{
"epoch": 3.3,
"learning_rate": 6.809160305343512e-06,
"loss": 1.9571,
"step": 4760
},
{
"epoch": 3.32,
"learning_rate": 6.753643303261625e-06,
"loss": 1.935,
"step": 4780
},
{
"epoch": 3.33,
"learning_rate": 6.698126301179737e-06,
"loss": 1.9674,
"step": 4800
},
{
"epoch": 3.34,
"learning_rate": 6.64260929909785e-06,
"loss": 1.9723,
"step": 4820
},
{
"epoch": 3.36,
"learning_rate": 6.587092297015961e-06,
"loss": 1.9424,
"step": 4840
},
{
"epoch": 3.37,
"learning_rate": 6.531575294934074e-06,
"loss": 1.9555,
"step": 4860
},
{
"epoch": 3.39,
"learning_rate": 6.476058292852187e-06,
"loss": 1.9312,
"step": 4880
},
{
"epoch": 3.4,
"learning_rate": 6.420541290770299e-06,
"loss": 1.9737,
"step": 4900
},
{
"epoch": 3.41,
"learning_rate": 6.365024288688412e-06,
"loss": 1.9727,
"step": 4920
},
{
"epoch": 3.43,
"learning_rate": 6.312283136710619e-06,
"loss": 2.0054,
"step": 4940
},
{
"epoch": 3.44,
"learning_rate": 6.25676613462873e-06,
"loss": 1.962,
"step": 4960
},
{
"epoch": 3.46,
"learning_rate": 6.201249132546843e-06,
"loss": 1.9729,
"step": 4980
},
{
"epoch": 3.47,
"learning_rate": 6.145732130464956e-06,
"loss": 1.9911,
"step": 5000
},
{
"epoch": 3.48,
"learning_rate": 6.090215128383068e-06,
"loss": 1.9298,
"step": 5020
},
{
"epoch": 3.5,
"learning_rate": 6.0346981263011805e-06,
"loss": 1.9564,
"step": 5040
},
{
"epoch": 3.51,
"learning_rate": 5.9791811242192934e-06,
"loss": 1.944,
"step": 5060
},
{
"epoch": 3.53,
"learning_rate": 5.923664122137405e-06,
"loss": 1.946,
"step": 5080
},
{
"epoch": 3.54,
"learning_rate": 5.8681471200555175e-06,
"loss": 2.0002,
"step": 5100
},
{
"epoch": 3.55,
"learning_rate": 5.8126301179736296e-06,
"loss": 1.9727,
"step": 5120
},
{
"epoch": 3.57,
"learning_rate": 5.7571131158917425e-06,
"loss": 1.9354,
"step": 5140
},
{
"epoch": 3.58,
"learning_rate": 5.701596113809855e-06,
"loss": 1.9635,
"step": 5160
},
{
"epoch": 3.59,
"learning_rate": 5.6460791117279665e-06,
"loss": 1.9595,
"step": 5180
},
{
"epoch": 3.61,
"learning_rate": 5.5905621096460794e-06,
"loss": 1.9583,
"step": 5200
},
{
"epoch": 3.62,
"learning_rate": 5.535045107564192e-06,
"loss": 1.959,
"step": 5220
},
{
"epoch": 3.64,
"learning_rate": 5.479528105482304e-06,
"loss": 1.9449,
"step": 5240
},
{
"epoch": 3.65,
"learning_rate": 5.424011103400417e-06,
"loss": 1.9484,
"step": 5260
},
{
"epoch": 3.66,
"learning_rate": 5.3684941013185285e-06,
"loss": 1.9413,
"step": 5280
},
{
"epoch": 3.68,
"learning_rate": 5.312977099236641e-06,
"loss": 1.9902,
"step": 5300
},
{
"epoch": 3.69,
"learning_rate": 5.257460097154754e-06,
"loss": 1.9673,
"step": 5320
},
{
"epoch": 3.71,
"learning_rate": 5.201943095072866e-06,
"loss": 1.9432,
"step": 5340
},
{
"epoch": 3.72,
"learning_rate": 5.146426092990979e-06,
"loss": 1.9196,
"step": 5360
},
{
"epoch": 3.73,
"learning_rate": 5.090909090909091e-06,
"loss": 2.0023,
"step": 5380
},
{
"epoch": 3.75,
"learning_rate": 5.035392088827203e-06,
"loss": 1.9332,
"step": 5400
},
{
"epoch": 3.76,
"learning_rate": 4.979875086745316e-06,
"loss": 1.9341,
"step": 5420
},
{
"epoch": 3.78,
"learning_rate": 4.924358084663429e-06,
"loss": 1.9848,
"step": 5440
},
{
"epoch": 3.79,
"learning_rate": 4.868841082581541e-06,
"loss": 1.9639,
"step": 5460
},
{
"epoch": 3.8,
"learning_rate": 4.813324080499653e-06,
"loss": 1.9417,
"step": 5480
},
{
"epoch": 3.82,
"learning_rate": 4.757807078417766e-06,
"loss": 1.9722,
"step": 5500
},
{
"epoch": 3.83,
"learning_rate": 4.702290076335878e-06,
"loss": 1.9817,
"step": 5520
},
{
"epoch": 3.84,
"learning_rate": 4.646773074253991e-06,
"loss": 1.9237,
"step": 5540
},
{
"epoch": 3.86,
"learning_rate": 4.591256072172103e-06,
"loss": 1.9498,
"step": 5560
},
{
"epoch": 3.87,
"learning_rate": 4.535739070090215e-06,
"loss": 1.9873,
"step": 5580
},
{
"epoch": 3.89,
"learning_rate": 4.480222068008328e-06,
"loss": 1.9558,
"step": 5600
},
{
"epoch": 3.9,
"learning_rate": 4.42470506592644e-06,
"loss": 1.9454,
"step": 5620
},
{
"epoch": 3.91,
"learning_rate": 4.369188063844553e-06,
"loss": 1.965,
"step": 5640
},
{
"epoch": 3.93,
"learning_rate": 4.313671061762666e-06,
"loss": 1.9538,
"step": 5660
},
{
"epoch": 3.94,
"learning_rate": 4.258154059680778e-06,
"loss": 1.9655,
"step": 5680
},
{
"epoch": 3.96,
"learning_rate": 4.20263705759889e-06,
"loss": 1.9635,
"step": 5700
},
{
"epoch": 3.97,
"learning_rate": 4.147120055517003e-06,
"loss": 1.9721,
"step": 5720
},
{
"epoch": 3.98,
"learning_rate": 4.091603053435115e-06,
"loss": 2.001,
"step": 5740
},
{
"epoch": 4.0,
"learning_rate": 4.036086051353228e-06,
"loss": 1.9389,
"step": 5760
},
{
"epoch": 4.01,
"learning_rate": 3.98056904927134e-06,
"loss": 1.9536,
"step": 5780
},
{
"epoch": 4.02,
"learning_rate": 3.925052047189452e-06,
"loss": 1.9084,
"step": 5800
},
{
"epoch": 4.04,
"learning_rate": 3.869535045107565e-06,
"loss": 1.9742,
"step": 5820
},
{
"epoch": 4.05,
"learning_rate": 3.8140180430256767e-06,
"loss": 1.9404,
"step": 5840
},
{
"epoch": 4.07,
"learning_rate": 3.7585010409437896e-06,
"loss": 1.9314,
"step": 5860
},
{
"epoch": 4.08,
"learning_rate": 3.7029840388619016e-06,
"loss": 1.9285,
"step": 5880
},
{
"epoch": 4.09,
"learning_rate": 3.647467036780014e-06,
"loss": 1.9508,
"step": 5900
},
{
"epoch": 4.11,
"learning_rate": 3.5919500346981266e-06,
"loss": 1.9263,
"step": 5920
},
{
"epoch": 4.12,
"learning_rate": 3.536433032616239e-06,
"loss": 1.9733,
"step": 5940
},
{
"epoch": 4.14,
"learning_rate": 3.4809160305343515e-06,
"loss": 1.9549,
"step": 5960
},
{
"epoch": 4.15,
"learning_rate": 3.425399028452464e-06,
"loss": 1.9557,
"step": 5980
},
{
"epoch": 4.16,
"learning_rate": 3.369882026370576e-06,
"loss": 1.9279,
"step": 6000
},
{
"epoch": 4.18,
"learning_rate": 3.314365024288689e-06,
"loss": 1.907,
"step": 6020
},
{
"epoch": 4.19,
"learning_rate": 3.258848022206801e-06,
"loss": 1.933,
"step": 6040
},
{
"epoch": 4.21,
"learning_rate": 3.2033310201249134e-06,
"loss": 1.9399,
"step": 6060
},
{
"epoch": 4.22,
"learning_rate": 3.1478140180430263e-06,
"loss": 1.9493,
"step": 6080
},
{
"epoch": 4.23,
"learning_rate": 3.0922970159611383e-06,
"loss": 1.9823,
"step": 6100
},
{
"epoch": 4.25,
"learning_rate": 3.036780013879251e-06,
"loss": 1.9645,
"step": 6120
},
{
"epoch": 4.26,
"learning_rate": 2.9812630117973633e-06,
"loss": 1.9479,
"step": 6140
},
{
"epoch": 4.27,
"learning_rate": 2.9257460097154753e-06,
"loss": 1.9512,
"step": 6160
},
{
"epoch": 4.29,
"learning_rate": 2.870229007633588e-06,
"loss": 1.921,
"step": 6180
},
{
"epoch": 4.3,
"learning_rate": 2.8147120055517002e-06,
"loss": 1.926,
"step": 6200
},
{
"epoch": 4.32,
"learning_rate": 2.7591950034698127e-06,
"loss": 1.9395,
"step": 6220
},
{
"epoch": 4.33,
"learning_rate": 2.7036780013879256e-06,
"loss": 1.9876,
"step": 6240
},
{
"epoch": 4.34,
"learning_rate": 2.6481609993060377e-06,
"loss": 1.9327,
"step": 6260
},
{
"epoch": 4.36,
"learning_rate": 2.595419847328244e-06,
"loss": 1.9638,
"step": 6280
},
{
"epoch": 4.37,
"learning_rate": 2.539902845246357e-06,
"loss": 1.959,
"step": 6300
},
{
"epoch": 4.39,
"learning_rate": 2.4843858431644695e-06,
"loss": 1.9385,
"step": 6320
},
{
"epoch": 4.4,
"learning_rate": 2.431644691186676e-06,
"loss": 1.9325,
"step": 6340
},
{
"epoch": 4.41,
"learning_rate": 2.3761276891047885e-06,
"loss": 1.9624,
"step": 6360
},
{
"epoch": 4.43,
"learning_rate": 2.320610687022901e-06,
"loss": 1.9417,
"step": 6380
},
{
"epoch": 4.44,
"learning_rate": 2.2650936849410135e-06,
"loss": 1.9376,
"step": 6400
},
{
"epoch": 4.46,
"learning_rate": 2.209576682859126e-06,
"loss": 1.9207,
"step": 6420
},
{
"epoch": 4.47,
"learning_rate": 2.1540596807772384e-06,
"loss": 1.9202,
"step": 6440
},
{
"epoch": 4.48,
"learning_rate": 2.0985426786953504e-06,
"loss": 1.9181,
"step": 6460
},
{
"epoch": 4.5,
"learning_rate": 2.043025676613463e-06,
"loss": 1.9712,
"step": 6480
},
{
"epoch": 4.51,
"learning_rate": 1.9875086745315754e-06,
"loss": 1.9433,
"step": 6500
},
{
"epoch": 4.52,
"learning_rate": 1.931991672449688e-06,
"loss": 1.9602,
"step": 6520
},
{
"epoch": 4.54,
"learning_rate": 1.8764746703678003e-06,
"loss": 1.9049,
"step": 6540
},
{
"epoch": 4.55,
"learning_rate": 1.8209576682859126e-06,
"loss": 1.8855,
"step": 6560
},
{
"epoch": 4.57,
"learning_rate": 1.7654406662040252e-06,
"loss": 1.9369,
"step": 6580
},
{
"epoch": 4.58,
"learning_rate": 1.7099236641221375e-06,
"loss": 1.9644,
"step": 6600
},
{
"epoch": 4.59,
"learning_rate": 1.65440666204025e-06,
"loss": 1.9247,
"step": 6620
},
{
"epoch": 4.61,
"learning_rate": 1.5988896599583622e-06,
"loss": 1.9326,
"step": 6640
},
{
"epoch": 4.62,
"learning_rate": 1.543372657876475e-06,
"loss": 1.9486,
"step": 6660
},
{
"epoch": 4.64,
"learning_rate": 1.4878556557945872e-06,
"loss": 1.9217,
"step": 6680
},
{
"epoch": 4.65,
"learning_rate": 1.4323386537126996e-06,
"loss": 1.9574,
"step": 6700
},
{
"epoch": 4.66,
"learning_rate": 1.3768216516308119e-06,
"loss": 1.9716,
"step": 6720
},
{
"epoch": 4.68,
"learning_rate": 1.3213046495489246e-06,
"loss": 1.9439,
"step": 6740
},
{
"epoch": 4.69,
"learning_rate": 1.265787647467037e-06,
"loss": 1.9521,
"step": 6760
},
{
"epoch": 4.71,
"learning_rate": 1.2102706453851493e-06,
"loss": 1.9322,
"step": 6780
},
{
"epoch": 4.72,
"learning_rate": 1.1547536433032617e-06,
"loss": 1.9674,
"step": 6800
},
{
"epoch": 4.73,
"learning_rate": 1.0992366412213742e-06,
"loss": 1.9346,
"step": 6820
},
{
"epoch": 4.75,
"learning_rate": 1.0437196391394867e-06,
"loss": 1.965,
"step": 6840
},
{
"epoch": 4.76,
"learning_rate": 9.88202637057599e-07,
"loss": 1.9198,
"step": 6860
},
{
"epoch": 4.77,
"learning_rate": 9.326856349757114e-07,
"loss": 1.9703,
"step": 6880
},
{
"epoch": 4.79,
"learning_rate": 8.771686328938238e-07,
"loss": 1.9539,
"step": 6900
},
{
"epoch": 4.8,
"learning_rate": 8.216516308119362e-07,
"loss": 1.9268,
"step": 6920
},
{
"epoch": 4.82,
"learning_rate": 7.661346287300486e-07,
"loss": 1.9492,
"step": 6940
},
{
"epoch": 4.83,
"learning_rate": 7.106176266481612e-07,
"loss": 1.9533,
"step": 6960
},
{
"epoch": 4.84,
"learning_rate": 6.551006245662734e-07,
"loss": 1.9536,
"step": 6980
},
{
"epoch": 4.86,
"learning_rate": 5.995836224843859e-07,
"loss": 1.9381,
"step": 7000
},
{
"epoch": 4.87,
"learning_rate": 5.440666204024984e-07,
"loss": 1.9477,
"step": 7020
},
{
"epoch": 4.89,
"learning_rate": 4.885496183206107e-07,
"loss": 1.9502,
"step": 7040
},
{
"epoch": 4.9,
"learning_rate": 4.3303261623872313e-07,
"loss": 1.9621,
"step": 7060
},
{
"epoch": 4.91,
"learning_rate": 3.7751561415683554e-07,
"loss": 1.9642,
"step": 7080
},
{
"epoch": 4.93,
"learning_rate": 3.2199861207494796e-07,
"loss": 1.9513,
"step": 7100
},
{
"epoch": 4.94,
"learning_rate": 2.664816099930604e-07,
"loss": 1.9607,
"step": 7120
},
{
"epoch": 4.95,
"learning_rate": 2.109646079111728e-07,
"loss": 1.943,
"step": 7140
},
{
"epoch": 4.97,
"learning_rate": 1.5544760582928525e-07,
"loss": 1.9397,
"step": 7160
},
{
"epoch": 4.98,
"learning_rate": 9.993060374739764e-08,
"loss": 1.9777,
"step": 7180
},
{
"epoch": 5.0,
"learning_rate": 4.441360166551007e-08,
"loss": 1.9451,
"step": 7200
},
{
"epoch": 5.0,
"step": 7205,
"total_flos": 16112557228032.0,
"train_loss": 2.03250736604541,
"train_runtime": 4701.6865,
"train_samples_per_second": 9.194,
"train_steps_per_second": 1.532
}
],
"max_steps": 7205,
"num_train_epochs": 5,
"total_flos": 16112557228032.0,
"trial_name": null,
"trial_params": null
}