m2m100_418M_sna_en_rel / trainer_state.json
Davlan's picture
Upload trainer_state.json
e4c20d6
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 228315,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.989050215710751e-05,
"loss": 2.8422,
"step": 500
},
{
"epoch": 0.01,
"learning_rate": 4.978100431421501e-05,
"loss": 2.2565,
"step": 1000
},
{
"epoch": 0.02,
"learning_rate": 4.9671506471322515e-05,
"loss": 2.0794,
"step": 1500
},
{
"epoch": 0.03,
"learning_rate": 4.9562008628430026e-05,
"loss": 1.9345,
"step": 2000
},
{
"epoch": 0.03,
"learning_rate": 4.9452510785537524e-05,
"loss": 1.8509,
"step": 2500
},
{
"epoch": 0.04,
"learning_rate": 4.9343012942645035e-05,
"loss": 1.8402,
"step": 3000
},
{
"epoch": 0.05,
"learning_rate": 4.923351509975254e-05,
"loss": 1.756,
"step": 3500
},
{
"epoch": 0.05,
"learning_rate": 4.9124017256860043e-05,
"loss": 1.7107,
"step": 4000
},
{
"epoch": 0.06,
"learning_rate": 4.901451941396755e-05,
"loss": 1.6741,
"step": 4500
},
{
"epoch": 0.07,
"learning_rate": 4.890502157107505e-05,
"loss": 1.6439,
"step": 5000
},
{
"epoch": 0.07,
"learning_rate": 4.8795523728182556e-05,
"loss": 1.6048,
"step": 5500
},
{
"epoch": 0.08,
"learning_rate": 4.868602588529006e-05,
"loss": 1.6165,
"step": 6000
},
{
"epoch": 0.09,
"learning_rate": 4.8576528042397565e-05,
"loss": 1.5696,
"step": 6500
},
{
"epoch": 0.09,
"learning_rate": 4.846703019950507e-05,
"loss": 1.5535,
"step": 7000
},
{
"epoch": 0.1,
"learning_rate": 4.835753235661258e-05,
"loss": 1.5339,
"step": 7500
},
{
"epoch": 0.11,
"learning_rate": 4.824803451372008e-05,
"loss": 1.5121,
"step": 8000
},
{
"epoch": 0.11,
"learning_rate": 4.813853667082759e-05,
"loss": 1.474,
"step": 8500
},
{
"epoch": 0.12,
"learning_rate": 4.802903882793509e-05,
"loss": 1.4876,
"step": 9000
},
{
"epoch": 0.12,
"learning_rate": 4.79195409850426e-05,
"loss": 1.4749,
"step": 9500
},
{
"epoch": 0.13,
"learning_rate": 4.78100431421501e-05,
"loss": 1.4663,
"step": 10000
},
{
"epoch": 0.14,
"learning_rate": 4.770054529925761e-05,
"loss": 1.4597,
"step": 10500
},
{
"epoch": 0.14,
"learning_rate": 4.759104745636511e-05,
"loss": 1.4437,
"step": 11000
},
{
"epoch": 0.15,
"learning_rate": 4.7481549613472614e-05,
"loss": 1.427,
"step": 11500
},
{
"epoch": 0.16,
"learning_rate": 4.7372051770580125e-05,
"loss": 1.4302,
"step": 12000
},
{
"epoch": 0.16,
"learning_rate": 4.726255392768762e-05,
"loss": 1.4336,
"step": 12500
},
{
"epoch": 0.17,
"learning_rate": 4.7153056084795134e-05,
"loss": 1.408,
"step": 13000
},
{
"epoch": 0.18,
"learning_rate": 4.704355824190264e-05,
"loss": 1.4198,
"step": 13500
},
{
"epoch": 0.18,
"learning_rate": 4.693406039901014e-05,
"loss": 1.3796,
"step": 14000
},
{
"epoch": 0.19,
"learning_rate": 4.682456255611765e-05,
"loss": 1.3649,
"step": 14500
},
{
"epoch": 0.2,
"learning_rate": 4.671506471322515e-05,
"loss": 1.3605,
"step": 15000
},
{
"epoch": 0.2,
"learning_rate": 4.6605566870332655e-05,
"loss": 1.3482,
"step": 15500
},
{
"epoch": 0.21,
"learning_rate": 4.649606902744016e-05,
"loss": 1.3444,
"step": 16000
},
{
"epoch": 0.22,
"learning_rate": 4.6386571184547664e-05,
"loss": 1.349,
"step": 16500
},
{
"epoch": 0.22,
"learning_rate": 4.627707334165517e-05,
"loss": 1.3299,
"step": 17000
},
{
"epoch": 0.23,
"learning_rate": 4.616757549876268e-05,
"loss": 1.3377,
"step": 17500
},
{
"epoch": 0.24,
"learning_rate": 4.605807765587018e-05,
"loss": 1.3175,
"step": 18000
},
{
"epoch": 0.24,
"learning_rate": 4.594857981297769e-05,
"loss": 1.3205,
"step": 18500
},
{
"epoch": 0.25,
"learning_rate": 4.583908197008519e-05,
"loss": 1.3187,
"step": 19000
},
{
"epoch": 0.26,
"learning_rate": 4.5729584127192696e-05,
"loss": 1.3051,
"step": 19500
},
{
"epoch": 0.26,
"learning_rate": 4.56200862843002e-05,
"loss": 1.311,
"step": 20000
},
{
"epoch": 0.27,
"learning_rate": 4.551058844140771e-05,
"loss": 1.3058,
"step": 20500
},
{
"epoch": 0.28,
"learning_rate": 4.540109059851521e-05,
"loss": 1.2713,
"step": 21000
},
{
"epoch": 0.28,
"learning_rate": 4.5291592755622713e-05,
"loss": 1.2873,
"step": 21500
},
{
"epoch": 0.29,
"learning_rate": 4.5182094912730225e-05,
"loss": 1.2967,
"step": 22000
},
{
"epoch": 0.3,
"learning_rate": 4.507259706983772e-05,
"loss": 1.3029,
"step": 22500
},
{
"epoch": 0.3,
"learning_rate": 4.496309922694523e-05,
"loss": 1.2662,
"step": 23000
},
{
"epoch": 0.31,
"learning_rate": 4.485360138405274e-05,
"loss": 1.2587,
"step": 23500
},
{
"epoch": 0.32,
"learning_rate": 4.474410354116024e-05,
"loss": 1.2504,
"step": 24000
},
{
"epoch": 0.32,
"learning_rate": 4.4634605698267746e-05,
"loss": 1.2654,
"step": 24500
},
{
"epoch": 0.33,
"learning_rate": 4.452510785537525e-05,
"loss": 1.2508,
"step": 25000
},
{
"epoch": 0.34,
"learning_rate": 4.4415610012482754e-05,
"loss": 1.2468,
"step": 25500
},
{
"epoch": 0.34,
"learning_rate": 4.4306112169590266e-05,
"loss": 1.2379,
"step": 26000
},
{
"epoch": 0.35,
"learning_rate": 4.419661432669776e-05,
"loss": 1.2461,
"step": 26500
},
{
"epoch": 0.35,
"learning_rate": 4.408711648380527e-05,
"loss": 1.2371,
"step": 27000
},
{
"epoch": 0.36,
"learning_rate": 4.397761864091278e-05,
"loss": 1.2235,
"step": 27500
},
{
"epoch": 0.37,
"learning_rate": 4.3868120798020276e-05,
"loss": 1.2331,
"step": 28000
},
{
"epoch": 0.37,
"learning_rate": 4.375862295512779e-05,
"loss": 1.2419,
"step": 28500
},
{
"epoch": 0.38,
"learning_rate": 4.364912511223529e-05,
"loss": 1.2349,
"step": 29000
},
{
"epoch": 0.39,
"learning_rate": 4.3539627269342795e-05,
"loss": 1.2146,
"step": 29500
},
{
"epoch": 0.39,
"learning_rate": 4.34301294264503e-05,
"loss": 1.2117,
"step": 30000
},
{
"epoch": 0.4,
"learning_rate": 4.332063158355781e-05,
"loss": 1.2347,
"step": 30500
},
{
"epoch": 0.41,
"learning_rate": 4.321113374066531e-05,
"loss": 1.2084,
"step": 31000
},
{
"epoch": 0.41,
"learning_rate": 4.310163589777282e-05,
"loss": 1.2006,
"step": 31500
},
{
"epoch": 0.42,
"learning_rate": 4.2992138054880324e-05,
"loss": 1.2135,
"step": 32000
},
{
"epoch": 0.43,
"learning_rate": 4.288264021198782e-05,
"loss": 1.1968,
"step": 32500
},
{
"epoch": 0.43,
"learning_rate": 4.277314236909533e-05,
"loss": 1.1877,
"step": 33000
},
{
"epoch": 0.44,
"learning_rate": 4.2663644526202836e-05,
"loss": 1.1946,
"step": 33500
},
{
"epoch": 0.45,
"learning_rate": 4.255414668331034e-05,
"loss": 1.2048,
"step": 34000
},
{
"epoch": 0.45,
"learning_rate": 4.2444648840417845e-05,
"loss": 1.2142,
"step": 34500
},
{
"epoch": 0.46,
"learning_rate": 4.233515099752535e-05,
"loss": 1.1975,
"step": 35000
},
{
"epoch": 0.47,
"learning_rate": 4.2225653154632854e-05,
"loss": 1.2019,
"step": 35500
},
{
"epoch": 0.47,
"learning_rate": 4.2116155311740365e-05,
"loss": 1.1783,
"step": 36000
},
{
"epoch": 0.48,
"learning_rate": 4.200665746884786e-05,
"loss": 1.1925,
"step": 36500
},
{
"epoch": 0.49,
"learning_rate": 4.189715962595537e-05,
"loss": 1.1773,
"step": 37000
},
{
"epoch": 0.49,
"learning_rate": 4.178766178306288e-05,
"loss": 1.2001,
"step": 37500
},
{
"epoch": 0.5,
"learning_rate": 4.1678163940170375e-05,
"loss": 1.165,
"step": 38000
},
{
"epoch": 0.51,
"learning_rate": 4.1568666097277886e-05,
"loss": 1.1679,
"step": 38500
},
{
"epoch": 0.51,
"learning_rate": 4.145916825438539e-05,
"loss": 1.1633,
"step": 39000
},
{
"epoch": 0.52,
"learning_rate": 4.1349670411492895e-05,
"loss": 1.1704,
"step": 39500
},
{
"epoch": 0.53,
"learning_rate": 4.12401725686004e-05,
"loss": 1.1789,
"step": 40000
},
{
"epoch": 0.53,
"learning_rate": 4.113067472570791e-05,
"loss": 1.1635,
"step": 40500
},
{
"epoch": 0.54,
"learning_rate": 4.102117688281541e-05,
"loss": 1.169,
"step": 41000
},
{
"epoch": 0.55,
"learning_rate": 4.091167903992292e-05,
"loss": 1.1543,
"step": 41500
},
{
"epoch": 0.55,
"learning_rate": 4.080218119703042e-05,
"loss": 1.147,
"step": 42000
},
{
"epoch": 0.56,
"learning_rate": 4.069268335413793e-05,
"loss": 1.1631,
"step": 42500
},
{
"epoch": 0.57,
"learning_rate": 4.058318551124543e-05,
"loss": 1.1453,
"step": 43000
},
{
"epoch": 0.57,
"learning_rate": 4.0473687668352936e-05,
"loss": 1.1397,
"step": 43500
},
{
"epoch": 0.58,
"learning_rate": 4.036418982546044e-05,
"loss": 1.155,
"step": 44000
},
{
"epoch": 0.58,
"learning_rate": 4.0254691982567944e-05,
"loss": 1.1369,
"step": 44500
},
{
"epoch": 0.59,
"learning_rate": 4.014519413967545e-05,
"loss": 1.1549,
"step": 45000
},
{
"epoch": 0.6,
"learning_rate": 4.003569629678295e-05,
"loss": 1.1291,
"step": 45500
},
{
"epoch": 0.6,
"learning_rate": 3.9926198453890464e-05,
"loss": 1.1352,
"step": 46000
},
{
"epoch": 0.61,
"learning_rate": 3.981670061099796e-05,
"loss": 1.1519,
"step": 46500
},
{
"epoch": 0.62,
"learning_rate": 3.970720276810547e-05,
"loss": 1.1513,
"step": 47000
},
{
"epoch": 0.62,
"learning_rate": 3.9597704925212977e-05,
"loss": 1.1278,
"step": 47500
},
{
"epoch": 0.63,
"learning_rate": 3.948820708232048e-05,
"loss": 1.1249,
"step": 48000
},
{
"epoch": 0.64,
"learning_rate": 3.9378709239427985e-05,
"loss": 1.1239,
"step": 48500
},
{
"epoch": 0.64,
"learning_rate": 3.926921139653549e-05,
"loss": 1.1055,
"step": 49000
},
{
"epoch": 0.65,
"learning_rate": 3.9159713553642994e-05,
"loss": 1.1311,
"step": 49500
},
{
"epoch": 0.66,
"learning_rate": 3.90502157107505e-05,
"loss": 1.0933,
"step": 50000
},
{
"epoch": 0.66,
"learning_rate": 3.894071786785801e-05,
"loss": 1.1217,
"step": 50500
},
{
"epoch": 0.67,
"learning_rate": 3.8831220024965507e-05,
"loss": 1.0915,
"step": 51000
},
{
"epoch": 0.68,
"learning_rate": 3.872172218207302e-05,
"loss": 1.1213,
"step": 51500
},
{
"epoch": 0.68,
"learning_rate": 3.861222433918052e-05,
"loss": 1.1248,
"step": 52000
},
{
"epoch": 0.69,
"learning_rate": 3.8502726496288026e-05,
"loss": 1.1291,
"step": 52500
},
{
"epoch": 0.7,
"learning_rate": 3.839322865339553e-05,
"loss": 1.1115,
"step": 53000
},
{
"epoch": 0.7,
"learning_rate": 3.8283730810503035e-05,
"loss": 1.0993,
"step": 53500
},
{
"epoch": 0.71,
"learning_rate": 3.817423296761054e-05,
"loss": 1.0849,
"step": 54000
},
{
"epoch": 0.72,
"learning_rate": 3.806473512471804e-05,
"loss": 1.1047,
"step": 54500
},
{
"epoch": 0.72,
"learning_rate": 3.795523728182555e-05,
"loss": 1.1082,
"step": 55000
},
{
"epoch": 0.73,
"learning_rate": 3.784573943893305e-05,
"loss": 1.0974,
"step": 55500
},
{
"epoch": 0.74,
"learning_rate": 3.773624159604056e-05,
"loss": 1.0883,
"step": 56000
},
{
"epoch": 0.74,
"learning_rate": 3.762674375314806e-05,
"loss": 1.1033,
"step": 56500
},
{
"epoch": 0.75,
"learning_rate": 3.751724591025557e-05,
"loss": 1.0903,
"step": 57000
},
{
"epoch": 0.76,
"learning_rate": 3.7407748067363076e-05,
"loss": 1.0873,
"step": 57500
},
{
"epoch": 0.76,
"learning_rate": 3.729825022447058e-05,
"loss": 1.0941,
"step": 58000
},
{
"epoch": 0.77,
"learning_rate": 3.7188752381578084e-05,
"loss": 1.0842,
"step": 58500
},
{
"epoch": 0.78,
"learning_rate": 3.707925453868559e-05,
"loss": 1.0919,
"step": 59000
},
{
"epoch": 0.78,
"learning_rate": 3.696975669579309e-05,
"loss": 1.0829,
"step": 59500
},
{
"epoch": 0.79,
"learning_rate": 3.68602588529006e-05,
"loss": 1.0863,
"step": 60000
},
{
"epoch": 0.79,
"learning_rate": 3.675076101000811e-05,
"loss": 1.0768,
"step": 60500
},
{
"epoch": 0.8,
"learning_rate": 3.6641263167115606e-05,
"loss": 1.0685,
"step": 61000
},
{
"epoch": 0.81,
"learning_rate": 3.653176532422312e-05,
"loss": 1.0648,
"step": 61500
},
{
"epoch": 0.81,
"learning_rate": 3.642226748133062e-05,
"loss": 1.0705,
"step": 62000
},
{
"epoch": 0.82,
"learning_rate": 3.6312769638438125e-05,
"loss": 1.0541,
"step": 62500
},
{
"epoch": 0.83,
"learning_rate": 3.620327179554563e-05,
"loss": 1.0677,
"step": 63000
},
{
"epoch": 0.83,
"learning_rate": 3.6093773952653134e-05,
"loss": 1.0758,
"step": 63500
},
{
"epoch": 0.84,
"learning_rate": 3.598427610976064e-05,
"loss": 1.0784,
"step": 64000
},
{
"epoch": 0.85,
"learning_rate": 3.587477826686814e-05,
"loss": 1.0801,
"step": 64500
},
{
"epoch": 0.85,
"learning_rate": 3.576528042397565e-05,
"loss": 1.0716,
"step": 65000
},
{
"epoch": 0.86,
"learning_rate": 3.565578258108315e-05,
"loss": 1.0723,
"step": 65500
},
{
"epoch": 0.87,
"learning_rate": 3.554628473819066e-05,
"loss": 1.0603,
"step": 66000
},
{
"epoch": 0.87,
"learning_rate": 3.5436786895298166e-05,
"loss": 1.0468,
"step": 66500
},
{
"epoch": 0.88,
"learning_rate": 3.532728905240567e-05,
"loss": 1.064,
"step": 67000
},
{
"epoch": 0.89,
"learning_rate": 3.5217791209513175e-05,
"loss": 1.0552,
"step": 67500
},
{
"epoch": 0.89,
"learning_rate": 3.510829336662068e-05,
"loss": 1.0592,
"step": 68000
},
{
"epoch": 0.9,
"learning_rate": 3.4998795523728183e-05,
"loss": 1.0435,
"step": 68500
},
{
"epoch": 0.91,
"learning_rate": 3.4889297680835694e-05,
"loss": 1.0582,
"step": 69000
},
{
"epoch": 0.91,
"learning_rate": 3.477979983794319e-05,
"loss": 1.0469,
"step": 69500
},
{
"epoch": 0.92,
"learning_rate": 3.4670301995050696e-05,
"loss": 1.0407,
"step": 70000
},
{
"epoch": 0.93,
"learning_rate": 3.456080415215821e-05,
"loss": 1.0443,
"step": 70500
},
{
"epoch": 0.93,
"learning_rate": 3.4451306309265705e-05,
"loss": 1.0444,
"step": 71000
},
{
"epoch": 0.94,
"learning_rate": 3.4341808466373216e-05,
"loss": 1.0123,
"step": 71500
},
{
"epoch": 0.95,
"learning_rate": 3.423231062348072e-05,
"loss": 1.0453,
"step": 72000
},
{
"epoch": 0.95,
"learning_rate": 3.4122812780588224e-05,
"loss": 1.0656,
"step": 72500
},
{
"epoch": 0.96,
"learning_rate": 3.401331493769573e-05,
"loss": 1.0242,
"step": 73000
},
{
"epoch": 0.97,
"learning_rate": 3.390381709480323e-05,
"loss": 1.0239,
"step": 73500
},
{
"epoch": 0.97,
"learning_rate": 3.379431925191074e-05,
"loss": 1.0326,
"step": 74000
},
{
"epoch": 0.98,
"learning_rate": 3.368482140901824e-05,
"loss": 1.0137,
"step": 74500
},
{
"epoch": 0.99,
"learning_rate": 3.357532356612575e-05,
"loss": 1.0391,
"step": 75000
},
{
"epoch": 0.99,
"learning_rate": 3.346582572323325e-05,
"loss": 1.0206,
"step": 75500
},
{
"epoch": 1.0,
"learning_rate": 3.335632788034076e-05,
"loss": 1.0391,
"step": 76000
},
{
"epoch": 1.01,
"learning_rate": 3.3246830037448265e-05,
"loss": 0.943,
"step": 76500
},
{
"epoch": 1.01,
"learning_rate": 3.313733219455577e-05,
"loss": 0.942,
"step": 77000
},
{
"epoch": 1.02,
"learning_rate": 3.3027834351663274e-05,
"loss": 0.9269,
"step": 77500
},
{
"epoch": 1.02,
"learning_rate": 3.291833650877078e-05,
"loss": 0.9384,
"step": 78000
},
{
"epoch": 1.03,
"learning_rate": 3.280883866587828e-05,
"loss": 0.9318,
"step": 78500
},
{
"epoch": 1.04,
"learning_rate": 3.2699340822985794e-05,
"loss": 0.9223,
"step": 79000
},
{
"epoch": 1.04,
"learning_rate": 3.258984298009329e-05,
"loss": 0.9235,
"step": 79500
},
{
"epoch": 1.05,
"learning_rate": 3.2480345137200795e-05,
"loss": 0.9259,
"step": 80000
},
{
"epoch": 1.06,
"learning_rate": 3.2370847294308306e-05,
"loss": 0.9197,
"step": 80500
},
{
"epoch": 1.06,
"learning_rate": 3.2261349451415804e-05,
"loss": 0.9181,
"step": 81000
},
{
"epoch": 1.07,
"learning_rate": 3.2151851608523315e-05,
"loss": 0.9107,
"step": 81500
},
{
"epoch": 1.08,
"learning_rate": 3.204235376563082e-05,
"loss": 0.9156,
"step": 82000
},
{
"epoch": 1.08,
"learning_rate": 3.1932855922738324e-05,
"loss": 0.9345,
"step": 82500
},
{
"epoch": 1.09,
"learning_rate": 3.182335807984583e-05,
"loss": 0.9375,
"step": 83000
},
{
"epoch": 1.1,
"learning_rate": 3.171386023695334e-05,
"loss": 0.9446,
"step": 83500
},
{
"epoch": 1.1,
"learning_rate": 3.1604362394060836e-05,
"loss": 0.9237,
"step": 84000
},
{
"epoch": 1.11,
"learning_rate": 3.149486455116835e-05,
"loss": 0.9282,
"step": 84500
},
{
"epoch": 1.12,
"learning_rate": 3.138536670827585e-05,
"loss": 0.9107,
"step": 85000
},
{
"epoch": 1.12,
"learning_rate": 3.127586886538335e-05,
"loss": 0.9396,
"step": 85500
},
{
"epoch": 1.13,
"learning_rate": 3.116637102249086e-05,
"loss": 0.9044,
"step": 86000
},
{
"epoch": 1.14,
"learning_rate": 3.1056873179598365e-05,
"loss": 0.922,
"step": 86500
},
{
"epoch": 1.14,
"learning_rate": 3.094737533670587e-05,
"loss": 0.909,
"step": 87000
},
{
"epoch": 1.15,
"learning_rate": 3.083787749381337e-05,
"loss": 0.9387,
"step": 87500
},
{
"epoch": 1.16,
"learning_rate": 3.072837965092088e-05,
"loss": 0.9041,
"step": 88000
},
{
"epoch": 1.16,
"learning_rate": 3.061888180802838e-05,
"loss": 0.9331,
"step": 88500
},
{
"epoch": 1.17,
"learning_rate": 3.050938396513589e-05,
"loss": 0.9157,
"step": 89000
},
{
"epoch": 1.18,
"learning_rate": 3.0399886122243394e-05,
"loss": 0.9012,
"step": 89500
},
{
"epoch": 1.18,
"learning_rate": 3.02903882793509e-05,
"loss": 0.9209,
"step": 90000
},
{
"epoch": 1.19,
"learning_rate": 3.0180890436458402e-05,
"loss": 0.9098,
"step": 90500
},
{
"epoch": 1.2,
"learning_rate": 3.0071392593565906e-05,
"loss": 0.9289,
"step": 91000
},
{
"epoch": 1.2,
"learning_rate": 2.9961894750673414e-05,
"loss": 0.926,
"step": 91500
},
{
"epoch": 1.21,
"learning_rate": 2.9852396907780915e-05,
"loss": 0.9085,
"step": 92000
},
{
"epoch": 1.22,
"learning_rate": 2.9742899064888423e-05,
"loss": 0.9116,
"step": 92500
},
{
"epoch": 1.22,
"learning_rate": 2.9633401221995927e-05,
"loss": 0.9176,
"step": 93000
},
{
"epoch": 1.23,
"learning_rate": 2.9523903379103435e-05,
"loss": 0.9109,
"step": 93500
},
{
"epoch": 1.24,
"learning_rate": 2.941440553621094e-05,
"loss": 0.9104,
"step": 94000
},
{
"epoch": 1.24,
"learning_rate": 2.9304907693318447e-05,
"loss": 0.9206,
"step": 94500
},
{
"epoch": 1.25,
"learning_rate": 2.9195409850425947e-05,
"loss": 0.9311,
"step": 95000
},
{
"epoch": 1.25,
"learning_rate": 2.9085912007533455e-05,
"loss": 0.9071,
"step": 95500
},
{
"epoch": 1.26,
"learning_rate": 2.897641416464096e-05,
"loss": 0.9096,
"step": 96000
},
{
"epoch": 1.27,
"learning_rate": 2.886691632174846e-05,
"loss": 0.9062,
"step": 96500
},
{
"epoch": 1.27,
"learning_rate": 2.8757418478855968e-05,
"loss": 0.8921,
"step": 97000
},
{
"epoch": 1.28,
"learning_rate": 2.8647920635963472e-05,
"loss": 0.9056,
"step": 97500
},
{
"epoch": 1.29,
"learning_rate": 2.853842279307098e-05,
"loss": 0.9298,
"step": 98000
},
{
"epoch": 1.29,
"learning_rate": 2.842892495017848e-05,
"loss": 0.8994,
"step": 98500
},
{
"epoch": 1.3,
"learning_rate": 2.831942710728599e-05,
"loss": 0.9263,
"step": 99000
},
{
"epoch": 1.31,
"learning_rate": 2.8209929264393493e-05,
"loss": 0.918,
"step": 99500
},
{
"epoch": 1.31,
"learning_rate": 2.8100431421501e-05,
"loss": 0.8925,
"step": 100000
},
{
"epoch": 1.32,
"learning_rate": 2.79909335786085e-05,
"loss": 0.8952,
"step": 100500
},
{
"epoch": 1.33,
"learning_rate": 2.7881435735716006e-05,
"loss": 0.8991,
"step": 101000
},
{
"epoch": 1.33,
"learning_rate": 2.7771937892823513e-05,
"loss": 0.916,
"step": 101500
},
{
"epoch": 1.34,
"learning_rate": 2.7662440049931014e-05,
"loss": 0.9108,
"step": 102000
},
{
"epoch": 1.35,
"learning_rate": 2.7552942207038525e-05,
"loss": 0.9022,
"step": 102500
},
{
"epoch": 1.35,
"learning_rate": 2.7443444364146026e-05,
"loss": 0.9178,
"step": 103000
},
{
"epoch": 1.36,
"learning_rate": 2.7333946521253534e-05,
"loss": 0.9068,
"step": 103500
},
{
"epoch": 1.37,
"learning_rate": 2.7224448678361038e-05,
"loss": 0.9017,
"step": 104000
},
{
"epoch": 1.37,
"learning_rate": 2.7114950835468546e-05,
"loss": 0.9017,
"step": 104500
},
{
"epoch": 1.38,
"learning_rate": 2.7005452992576047e-05,
"loss": 0.8982,
"step": 105000
},
{
"epoch": 1.39,
"learning_rate": 2.6895955149683554e-05,
"loss": 0.902,
"step": 105500
},
{
"epoch": 1.39,
"learning_rate": 2.678645730679106e-05,
"loss": 0.9025,
"step": 106000
},
{
"epoch": 1.4,
"learning_rate": 2.667695946389856e-05,
"loss": 0.8889,
"step": 106500
},
{
"epoch": 1.41,
"learning_rate": 2.6567461621006067e-05,
"loss": 0.8943,
"step": 107000
},
{
"epoch": 1.41,
"learning_rate": 2.645796377811357e-05,
"loss": 0.8982,
"step": 107500
},
{
"epoch": 1.42,
"learning_rate": 2.634846593522108e-05,
"loss": 0.8975,
"step": 108000
},
{
"epoch": 1.43,
"learning_rate": 2.623896809232858e-05,
"loss": 0.9069,
"step": 108500
},
{
"epoch": 1.43,
"learning_rate": 2.6129470249436088e-05,
"loss": 0.8996,
"step": 109000
},
{
"epoch": 1.44,
"learning_rate": 2.6019972406543592e-05,
"loss": 0.8928,
"step": 109500
},
{
"epoch": 1.45,
"learning_rate": 2.59104745636511e-05,
"loss": 0.8819,
"step": 110000
},
{
"epoch": 1.45,
"learning_rate": 2.58009767207586e-05,
"loss": 0.8913,
"step": 110500
},
{
"epoch": 1.46,
"learning_rate": 2.5691478877866108e-05,
"loss": 0.8766,
"step": 111000
},
{
"epoch": 1.47,
"learning_rate": 2.5581981034973612e-05,
"loss": 0.8924,
"step": 111500
},
{
"epoch": 1.47,
"learning_rate": 2.5472483192081113e-05,
"loss": 0.8897,
"step": 112000
},
{
"epoch": 1.48,
"learning_rate": 2.5362985349188624e-05,
"loss": 0.884,
"step": 112500
},
{
"epoch": 1.48,
"learning_rate": 2.5253487506296125e-05,
"loss": 0.8877,
"step": 113000
},
{
"epoch": 1.49,
"learning_rate": 2.5143989663403633e-05,
"loss": 0.8877,
"step": 113500
},
{
"epoch": 1.5,
"learning_rate": 2.5034491820511137e-05,
"loss": 0.8912,
"step": 114000
},
{
"epoch": 1.5,
"learning_rate": 2.492499397761864e-05,
"loss": 0.8964,
"step": 114500
},
{
"epoch": 1.51,
"learning_rate": 2.4815496134726146e-05,
"loss": 0.8984,
"step": 115000
},
{
"epoch": 1.52,
"learning_rate": 2.470599829183365e-05,
"loss": 0.8714,
"step": 115500
},
{
"epoch": 1.52,
"learning_rate": 2.4596500448941158e-05,
"loss": 0.8903,
"step": 116000
},
{
"epoch": 1.53,
"learning_rate": 2.4487002606048662e-05,
"loss": 0.8807,
"step": 116500
},
{
"epoch": 1.54,
"learning_rate": 2.4377504763156166e-05,
"loss": 0.8797,
"step": 117000
},
{
"epoch": 1.54,
"learning_rate": 2.4268006920263674e-05,
"loss": 0.8831,
"step": 117500
},
{
"epoch": 1.55,
"learning_rate": 2.4158509077371178e-05,
"loss": 0.8926,
"step": 118000
},
{
"epoch": 1.56,
"learning_rate": 2.4049011234478682e-05,
"loss": 0.892,
"step": 118500
},
{
"epoch": 1.56,
"learning_rate": 2.3939513391586187e-05,
"loss": 0.8858,
"step": 119000
},
{
"epoch": 1.57,
"learning_rate": 2.3830015548693694e-05,
"loss": 0.8789,
"step": 119500
},
{
"epoch": 1.58,
"learning_rate": 2.3720517705801195e-05,
"loss": 0.8924,
"step": 120000
},
{
"epoch": 1.58,
"learning_rate": 2.36110198629087e-05,
"loss": 0.8641,
"step": 120500
},
{
"epoch": 1.59,
"learning_rate": 2.3501522020016207e-05,
"loss": 0.8947,
"step": 121000
},
{
"epoch": 1.6,
"learning_rate": 2.339202417712371e-05,
"loss": 0.8742,
"step": 121500
},
{
"epoch": 1.6,
"learning_rate": 2.3282526334231216e-05,
"loss": 0.8714,
"step": 122000
},
{
"epoch": 1.61,
"learning_rate": 2.3173028491338723e-05,
"loss": 0.848,
"step": 122500
},
{
"epoch": 1.62,
"learning_rate": 2.3063530648446228e-05,
"loss": 0.8761,
"step": 123000
},
{
"epoch": 1.62,
"learning_rate": 2.2954032805553732e-05,
"loss": 0.8801,
"step": 123500
},
{
"epoch": 1.63,
"learning_rate": 2.2844534962661236e-05,
"loss": 0.8737,
"step": 124000
},
{
"epoch": 1.64,
"learning_rate": 2.2735037119768744e-05,
"loss": 0.8847,
"step": 124500
},
{
"epoch": 1.64,
"learning_rate": 2.2625539276876245e-05,
"loss": 0.8602,
"step": 125000
},
{
"epoch": 1.65,
"learning_rate": 2.251604143398375e-05,
"loss": 0.8615,
"step": 125500
},
{
"epoch": 1.66,
"learning_rate": 2.2406543591091257e-05,
"loss": 0.8646,
"step": 126000
},
{
"epoch": 1.66,
"learning_rate": 2.229704574819876e-05,
"loss": 0.8604,
"step": 126500
},
{
"epoch": 1.67,
"learning_rate": 2.2187547905306265e-05,
"loss": 0.8668,
"step": 127000
},
{
"epoch": 1.68,
"learning_rate": 2.2078050062413773e-05,
"loss": 0.8758,
"step": 127500
},
{
"epoch": 1.68,
"learning_rate": 2.1968552219521277e-05,
"loss": 0.8695,
"step": 128000
},
{
"epoch": 1.69,
"learning_rate": 2.185905437662878e-05,
"loss": 0.8729,
"step": 128500
},
{
"epoch": 1.7,
"learning_rate": 2.1749556533736286e-05,
"loss": 0.8571,
"step": 129000
},
{
"epoch": 1.7,
"learning_rate": 2.1640058690843793e-05,
"loss": 0.8729,
"step": 129500
},
{
"epoch": 1.71,
"learning_rate": 2.1530560847951298e-05,
"loss": 0.8689,
"step": 130000
},
{
"epoch": 1.71,
"learning_rate": 2.14210630050588e-05,
"loss": 0.8732,
"step": 130500
},
{
"epoch": 1.72,
"learning_rate": 2.1311565162166306e-05,
"loss": 0.8606,
"step": 131000
},
{
"epoch": 1.73,
"learning_rate": 2.120206731927381e-05,
"loss": 0.8642,
"step": 131500
},
{
"epoch": 1.73,
"learning_rate": 2.1092569476381315e-05,
"loss": 0.8395,
"step": 132000
},
{
"epoch": 1.74,
"learning_rate": 2.0983071633488823e-05,
"loss": 0.8549,
"step": 132500
},
{
"epoch": 1.75,
"learning_rate": 2.0873573790596327e-05,
"loss": 0.8683,
"step": 133000
},
{
"epoch": 1.75,
"learning_rate": 2.076407594770383e-05,
"loss": 0.8402,
"step": 133500
},
{
"epoch": 1.76,
"learning_rate": 2.0654578104811335e-05,
"loss": 0.8503,
"step": 134000
},
{
"epoch": 1.77,
"learning_rate": 2.0545080261918843e-05,
"loss": 0.8514,
"step": 134500
},
{
"epoch": 1.77,
"learning_rate": 2.0435582419026347e-05,
"loss": 0.8441,
"step": 135000
},
{
"epoch": 1.78,
"learning_rate": 2.0326084576133848e-05,
"loss": 0.8698,
"step": 135500
},
{
"epoch": 1.79,
"learning_rate": 2.0216586733241356e-05,
"loss": 0.8523,
"step": 136000
},
{
"epoch": 1.79,
"learning_rate": 2.010708889034886e-05,
"loss": 0.8537,
"step": 136500
},
{
"epoch": 1.8,
"learning_rate": 1.9997591047456364e-05,
"loss": 0.8606,
"step": 137000
},
{
"epoch": 1.81,
"learning_rate": 1.9888093204563872e-05,
"loss": 0.842,
"step": 137500
},
{
"epoch": 1.81,
"learning_rate": 1.9778595361671376e-05,
"loss": 0.8611,
"step": 138000
},
{
"epoch": 1.82,
"learning_rate": 1.966909751877888e-05,
"loss": 0.8503,
"step": 138500
},
{
"epoch": 1.83,
"learning_rate": 1.9559599675886385e-05,
"loss": 0.8514,
"step": 139000
},
{
"epoch": 1.83,
"learning_rate": 1.9450101832993893e-05,
"loss": 0.8432,
"step": 139500
},
{
"epoch": 1.84,
"learning_rate": 1.9340603990101397e-05,
"loss": 0.8503,
"step": 140000
},
{
"epoch": 1.85,
"learning_rate": 1.92311061472089e-05,
"loss": 0.8418,
"step": 140500
},
{
"epoch": 1.85,
"learning_rate": 1.9121608304316405e-05,
"loss": 0.8479,
"step": 141000
},
{
"epoch": 1.86,
"learning_rate": 1.901211046142391e-05,
"loss": 0.8356,
"step": 141500
},
{
"epoch": 1.87,
"learning_rate": 1.8902612618531414e-05,
"loss": 0.8289,
"step": 142000
},
{
"epoch": 1.87,
"learning_rate": 1.879311477563892e-05,
"loss": 0.8479,
"step": 142500
},
{
"epoch": 1.88,
"learning_rate": 1.8683616932746426e-05,
"loss": 0.8406,
"step": 143000
},
{
"epoch": 1.89,
"learning_rate": 1.857411908985393e-05,
"loss": 0.8468,
"step": 143500
},
{
"epoch": 1.89,
"learning_rate": 1.8464621246961434e-05,
"loss": 0.8227,
"step": 144000
},
{
"epoch": 1.9,
"learning_rate": 1.8355123404068942e-05,
"loss": 0.8484,
"step": 144500
},
{
"epoch": 1.91,
"learning_rate": 1.8245625561176446e-05,
"loss": 0.8234,
"step": 145000
},
{
"epoch": 1.91,
"learning_rate": 1.813612771828395e-05,
"loss": 0.8352,
"step": 145500
},
{
"epoch": 1.92,
"learning_rate": 1.802662987539146e-05,
"loss": 0.8571,
"step": 146000
},
{
"epoch": 1.92,
"learning_rate": 1.791713203249896e-05,
"loss": 0.8444,
"step": 146500
},
{
"epoch": 1.93,
"learning_rate": 1.7807634189606464e-05,
"loss": 0.8566,
"step": 147000
},
{
"epoch": 1.94,
"learning_rate": 1.769813634671397e-05,
"loss": 0.849,
"step": 147500
},
{
"epoch": 1.94,
"learning_rate": 1.7588638503821475e-05,
"loss": 0.8368,
"step": 148000
},
{
"epoch": 1.95,
"learning_rate": 1.747914066092898e-05,
"loss": 0.8268,
"step": 148500
},
{
"epoch": 1.96,
"learning_rate": 1.7369642818036484e-05,
"loss": 0.8391,
"step": 149000
},
{
"epoch": 1.96,
"learning_rate": 1.7260144975143992e-05,
"loss": 0.8255,
"step": 149500
},
{
"epoch": 1.97,
"learning_rate": 1.7150647132251496e-05,
"loss": 0.8388,
"step": 150000
},
{
"epoch": 1.98,
"learning_rate": 1.7041149289359e-05,
"loss": 0.8416,
"step": 150500
},
{
"epoch": 1.98,
"learning_rate": 1.6931651446466508e-05,
"loss": 0.8199,
"step": 151000
},
{
"epoch": 1.99,
"learning_rate": 1.682215360357401e-05,
"loss": 0.8417,
"step": 151500
},
{
"epoch": 2.0,
"learning_rate": 1.6712655760681513e-05,
"loss": 0.8189,
"step": 152000
},
{
"epoch": 2.0,
"learning_rate": 1.660315791778902e-05,
"loss": 0.7702,
"step": 152500
},
{
"epoch": 2.01,
"learning_rate": 1.6493660074896525e-05,
"loss": 0.724,
"step": 153000
},
{
"epoch": 2.02,
"learning_rate": 1.638416223200403e-05,
"loss": 0.7192,
"step": 153500
},
{
"epoch": 2.02,
"learning_rate": 1.6274664389111534e-05,
"loss": 0.7164,
"step": 154000
},
{
"epoch": 2.03,
"learning_rate": 1.616516654621904e-05,
"loss": 0.741,
"step": 154500
},
{
"epoch": 2.04,
"learning_rate": 1.6055668703326546e-05,
"loss": 0.721,
"step": 155000
},
{
"epoch": 2.04,
"learning_rate": 1.594617086043405e-05,
"loss": 0.7184,
"step": 155500
},
{
"epoch": 2.05,
"learning_rate": 1.5836673017541557e-05,
"loss": 0.7277,
"step": 156000
},
{
"epoch": 2.06,
"learning_rate": 1.5727175174649062e-05,
"loss": 0.7188,
"step": 156500
},
{
"epoch": 2.06,
"learning_rate": 1.5617677331756563e-05,
"loss": 0.7466,
"step": 157000
},
{
"epoch": 2.07,
"learning_rate": 1.550817948886407e-05,
"loss": 0.7235,
"step": 157500
},
{
"epoch": 2.08,
"learning_rate": 1.5398681645971575e-05,
"loss": 0.7307,
"step": 158000
},
{
"epoch": 2.08,
"learning_rate": 1.528918380307908e-05,
"loss": 0.7292,
"step": 158500
},
{
"epoch": 2.09,
"learning_rate": 1.5179685960186585e-05,
"loss": 0.7335,
"step": 159000
},
{
"epoch": 2.1,
"learning_rate": 1.507018811729409e-05,
"loss": 0.7284,
"step": 159500
},
{
"epoch": 2.1,
"learning_rate": 1.4960690274401595e-05,
"loss": 0.7281,
"step": 160000
},
{
"epoch": 2.11,
"learning_rate": 1.4851192431509101e-05,
"loss": 0.7312,
"step": 160500
},
{
"epoch": 2.12,
"learning_rate": 1.4741694588616605e-05,
"loss": 0.7172,
"step": 161000
},
{
"epoch": 2.12,
"learning_rate": 1.4632196745724111e-05,
"loss": 0.7143,
"step": 161500
},
{
"epoch": 2.13,
"learning_rate": 1.4522698902831616e-05,
"loss": 0.7239,
"step": 162000
},
{
"epoch": 2.14,
"learning_rate": 1.4413201059939118e-05,
"loss": 0.7264,
"step": 162500
},
{
"epoch": 2.14,
"learning_rate": 1.4303703217046624e-05,
"loss": 0.7232,
"step": 163000
},
{
"epoch": 2.15,
"learning_rate": 1.4194205374154128e-05,
"loss": 0.7229,
"step": 163500
},
{
"epoch": 2.15,
"learning_rate": 1.4084707531261634e-05,
"loss": 0.7205,
"step": 164000
},
{
"epoch": 2.16,
"learning_rate": 1.397520968836914e-05,
"loss": 0.7313,
"step": 164500
},
{
"epoch": 2.17,
"learning_rate": 1.3865711845476645e-05,
"loss": 0.7188,
"step": 165000
},
{
"epoch": 2.17,
"learning_rate": 1.375621400258415e-05,
"loss": 0.7225,
"step": 165500
},
{
"epoch": 2.18,
"learning_rate": 1.3646716159691655e-05,
"loss": 0.7372,
"step": 166000
},
{
"epoch": 2.19,
"learning_rate": 1.3537218316799161e-05,
"loss": 0.7259,
"step": 166500
},
{
"epoch": 2.19,
"learning_rate": 1.3427720473906665e-05,
"loss": 0.7172,
"step": 167000
},
{
"epoch": 2.2,
"learning_rate": 1.3318222631014168e-05,
"loss": 0.7245,
"step": 167500
},
{
"epoch": 2.21,
"learning_rate": 1.3208724788121674e-05,
"loss": 0.7322,
"step": 168000
},
{
"epoch": 2.21,
"learning_rate": 1.3099226945229178e-05,
"loss": 0.7172,
"step": 168500
},
{
"epoch": 2.22,
"learning_rate": 1.2989729102336684e-05,
"loss": 0.7294,
"step": 169000
},
{
"epoch": 2.23,
"learning_rate": 1.288023125944419e-05,
"loss": 0.7152,
"step": 169500
},
{
"epoch": 2.23,
"learning_rate": 1.2770733416551694e-05,
"loss": 0.7217,
"step": 170000
},
{
"epoch": 2.24,
"learning_rate": 1.26612355736592e-05,
"loss": 0.7383,
"step": 170500
},
{
"epoch": 2.25,
"learning_rate": 1.2551737730766704e-05,
"loss": 0.7229,
"step": 171000
},
{
"epoch": 2.25,
"learning_rate": 1.2442239887874209e-05,
"loss": 0.7278,
"step": 171500
},
{
"epoch": 2.26,
"learning_rate": 1.2332742044981715e-05,
"loss": 0.7267,
"step": 172000
},
{
"epoch": 2.27,
"learning_rate": 1.2223244202089219e-05,
"loss": 0.7088,
"step": 172500
},
{
"epoch": 2.27,
"learning_rate": 1.2113746359196725e-05,
"loss": 0.7105,
"step": 173000
},
{
"epoch": 2.28,
"learning_rate": 1.200424851630423e-05,
"loss": 0.7111,
"step": 173500
},
{
"epoch": 2.29,
"learning_rate": 1.1894750673411735e-05,
"loss": 0.7171,
"step": 174000
},
{
"epoch": 2.29,
"learning_rate": 1.178525283051924e-05,
"loss": 0.7179,
"step": 174500
},
{
"epoch": 2.3,
"learning_rate": 1.1675754987626744e-05,
"loss": 0.7239,
"step": 175000
},
{
"epoch": 2.31,
"learning_rate": 1.156625714473425e-05,
"loss": 0.7183,
"step": 175500
},
{
"epoch": 2.31,
"learning_rate": 1.1456759301841754e-05,
"loss": 0.7126,
"step": 176000
},
{
"epoch": 2.32,
"learning_rate": 1.134726145894926e-05,
"loss": 0.6983,
"step": 176500
},
{
"epoch": 2.33,
"learning_rate": 1.1237763616056764e-05,
"loss": 0.7036,
"step": 177000
},
{
"epoch": 2.33,
"learning_rate": 1.1128265773164269e-05,
"loss": 0.7152,
"step": 177500
},
{
"epoch": 2.34,
"learning_rate": 1.1018767930271775e-05,
"loss": 0.7134,
"step": 178000
},
{
"epoch": 2.35,
"learning_rate": 1.0909270087379279e-05,
"loss": 0.7066,
"step": 178500
},
{
"epoch": 2.35,
"learning_rate": 1.0799772244486785e-05,
"loss": 0.7004,
"step": 179000
},
{
"epoch": 2.36,
"learning_rate": 1.0690274401594289e-05,
"loss": 0.717,
"step": 179500
},
{
"epoch": 2.37,
"learning_rate": 1.0580776558701793e-05,
"loss": 0.7036,
"step": 180000
},
{
"epoch": 2.37,
"learning_rate": 1.04712787158093e-05,
"loss": 0.7202,
"step": 180500
},
{
"epoch": 2.38,
"learning_rate": 1.0361780872916804e-05,
"loss": 0.7273,
"step": 181000
},
{
"epoch": 2.38,
"learning_rate": 1.025228303002431e-05,
"loss": 0.7085,
"step": 181500
},
{
"epoch": 2.39,
"learning_rate": 1.0142785187131814e-05,
"loss": 0.7042,
"step": 182000
},
{
"epoch": 2.4,
"learning_rate": 1.0033287344239318e-05,
"loss": 0.7001,
"step": 182500
},
{
"epoch": 2.4,
"learning_rate": 9.923789501346824e-06,
"loss": 0.7195,
"step": 183000
},
{
"epoch": 2.41,
"learning_rate": 9.814291658454328e-06,
"loss": 0.7028,
"step": 183500
},
{
"epoch": 2.42,
"learning_rate": 9.704793815561834e-06,
"loss": 0.7031,
"step": 184000
},
{
"epoch": 2.42,
"learning_rate": 9.59529597266934e-06,
"loss": 0.6977,
"step": 184500
},
{
"epoch": 2.43,
"learning_rate": 9.485798129776843e-06,
"loss": 0.7163,
"step": 185000
},
{
"epoch": 2.44,
"learning_rate": 9.376300286884349e-06,
"loss": 0.7157,
"step": 185500
},
{
"epoch": 2.44,
"learning_rate": 9.266802443991853e-06,
"loss": 0.7178,
"step": 186000
},
{
"epoch": 2.45,
"learning_rate": 9.157304601099359e-06,
"loss": 0.7052,
"step": 186500
},
{
"epoch": 2.46,
"learning_rate": 9.047806758206865e-06,
"loss": 0.712,
"step": 187000
},
{
"epoch": 2.46,
"learning_rate": 8.938308915314368e-06,
"loss": 0.7119,
"step": 187500
},
{
"epoch": 2.47,
"learning_rate": 8.828811072421874e-06,
"loss": 0.7064,
"step": 188000
},
{
"epoch": 2.48,
"learning_rate": 8.719313229529378e-06,
"loss": 0.7048,
"step": 188500
},
{
"epoch": 2.48,
"learning_rate": 8.609815386636884e-06,
"loss": 0.7077,
"step": 189000
},
{
"epoch": 2.49,
"learning_rate": 8.50031754374439e-06,
"loss": 0.7187,
"step": 189500
},
{
"epoch": 2.5,
"learning_rate": 8.390819700851892e-06,
"loss": 0.7167,
"step": 190000
},
{
"epoch": 2.5,
"learning_rate": 8.281321857959398e-06,
"loss": 0.705,
"step": 190500
},
{
"epoch": 2.51,
"learning_rate": 8.171824015066903e-06,
"loss": 0.7016,
"step": 191000
},
{
"epoch": 2.52,
"learning_rate": 8.062326172174409e-06,
"loss": 0.72,
"step": 191500
},
{
"epoch": 2.52,
"learning_rate": 7.952828329281915e-06,
"loss": 0.7226,
"step": 192000
},
{
"epoch": 2.53,
"learning_rate": 7.843330486389419e-06,
"loss": 0.7012,
"step": 192500
},
{
"epoch": 2.54,
"learning_rate": 7.733832643496923e-06,
"loss": 0.7136,
"step": 193000
},
{
"epoch": 2.54,
"learning_rate": 7.624334800604428e-06,
"loss": 0.7028,
"step": 193500
},
{
"epoch": 2.55,
"learning_rate": 7.5148369577119335e-06,
"loss": 0.7074,
"step": 194000
},
{
"epoch": 2.56,
"learning_rate": 7.405339114819439e-06,
"loss": 0.6994,
"step": 194500
},
{
"epoch": 2.56,
"learning_rate": 7.295841271926944e-06,
"loss": 0.7148,
"step": 195000
},
{
"epoch": 2.57,
"learning_rate": 7.186343429034448e-06,
"loss": 0.7164,
"step": 195500
},
{
"epoch": 2.58,
"learning_rate": 7.076845586141953e-06,
"loss": 0.7094,
"step": 196000
},
{
"epoch": 2.58,
"learning_rate": 6.967347743249458e-06,
"loss": 0.705,
"step": 196500
},
{
"epoch": 2.59,
"learning_rate": 6.857849900356963e-06,
"loss": 0.7101,
"step": 197000
},
{
"epoch": 2.6,
"learning_rate": 6.7483520574644685e-06,
"loss": 0.7195,
"step": 197500
},
{
"epoch": 2.6,
"learning_rate": 6.638854214571973e-06,
"loss": 0.6933,
"step": 198000
},
{
"epoch": 2.61,
"learning_rate": 6.529356371679478e-06,
"loss": 0.6898,
"step": 198500
},
{
"epoch": 2.61,
"learning_rate": 6.419858528786983e-06,
"loss": 0.7002,
"step": 199000
},
{
"epoch": 2.62,
"learning_rate": 6.310360685894488e-06,
"loss": 0.7082,
"step": 199500
},
{
"epoch": 2.63,
"learning_rate": 6.200862843001993e-06,
"loss": 0.7006,
"step": 200000
},
{
"epoch": 2.63,
"learning_rate": 6.091365000109498e-06,
"loss": 0.6882,
"step": 200500
},
{
"epoch": 2.64,
"learning_rate": 5.9818671572170035e-06,
"loss": 0.6964,
"step": 201000
},
{
"epoch": 2.65,
"learning_rate": 5.872369314324508e-06,
"loss": 0.7153,
"step": 201500
},
{
"epoch": 2.65,
"learning_rate": 5.762871471432013e-06,
"loss": 0.6976,
"step": 202000
},
{
"epoch": 2.66,
"learning_rate": 5.653373628539518e-06,
"loss": 0.7139,
"step": 202500
},
{
"epoch": 2.67,
"learning_rate": 5.543875785647023e-06,
"loss": 0.6967,
"step": 203000
},
{
"epoch": 2.67,
"learning_rate": 5.434377942754528e-06,
"loss": 0.6938,
"step": 203500
},
{
"epoch": 2.68,
"learning_rate": 5.324880099862033e-06,
"loss": 0.7028,
"step": 204000
},
{
"epoch": 2.69,
"learning_rate": 5.215382256969538e-06,
"loss": 0.6873,
"step": 204500
},
{
"epoch": 2.69,
"learning_rate": 5.105884414077043e-06,
"loss": 0.6845,
"step": 205000
},
{
"epoch": 2.7,
"learning_rate": 4.996386571184548e-06,
"loss": 0.6983,
"step": 205500
},
{
"epoch": 2.71,
"learning_rate": 4.886888728292053e-06,
"loss": 0.6852,
"step": 206000
},
{
"epoch": 2.71,
"learning_rate": 4.777390885399558e-06,
"loss": 0.6945,
"step": 206500
},
{
"epoch": 2.72,
"learning_rate": 4.6678930425070625e-06,
"loss": 0.7116,
"step": 207000
},
{
"epoch": 2.73,
"learning_rate": 4.558395199614568e-06,
"loss": 0.6896,
"step": 207500
},
{
"epoch": 2.73,
"learning_rate": 4.448897356722073e-06,
"loss": 0.6987,
"step": 208000
},
{
"epoch": 2.74,
"learning_rate": 4.339399513829578e-06,
"loss": 0.684,
"step": 208500
},
{
"epoch": 2.75,
"learning_rate": 4.229901670937083e-06,
"loss": 0.6853,
"step": 209000
},
{
"epoch": 2.75,
"learning_rate": 4.120403828044587e-06,
"loss": 0.6847,
"step": 209500
},
{
"epoch": 2.76,
"learning_rate": 4.010905985152092e-06,
"loss": 0.699,
"step": 210000
},
{
"epoch": 2.77,
"learning_rate": 3.901408142259598e-06,
"loss": 0.696,
"step": 210500
},
{
"epoch": 2.77,
"learning_rate": 3.7919102993671026e-06,
"loss": 0.692,
"step": 211000
},
{
"epoch": 2.78,
"learning_rate": 3.6824124564746078e-06,
"loss": 0.6732,
"step": 211500
},
{
"epoch": 2.79,
"learning_rate": 3.5729146135821125e-06,
"loss": 0.6932,
"step": 212000
},
{
"epoch": 2.79,
"learning_rate": 3.4634167706896176e-06,
"loss": 0.6969,
"step": 212500
},
{
"epoch": 2.8,
"learning_rate": 3.3539189277971227e-06,
"loss": 0.6961,
"step": 213000
},
{
"epoch": 2.81,
"learning_rate": 3.2444210849046274e-06,
"loss": 0.7018,
"step": 213500
},
{
"epoch": 2.81,
"learning_rate": 3.1349232420121326e-06,
"loss": 0.6883,
"step": 214000
},
{
"epoch": 2.82,
"learning_rate": 3.0254253991196377e-06,
"loss": 0.7013,
"step": 214500
},
{
"epoch": 2.83,
"learning_rate": 2.9159275562271424e-06,
"loss": 0.6897,
"step": 215000
},
{
"epoch": 2.83,
"learning_rate": 2.8064297133346475e-06,
"loss": 0.691,
"step": 215500
},
{
"epoch": 2.84,
"learning_rate": 2.6969318704421526e-06,
"loss": 0.6923,
"step": 216000
},
{
"epoch": 2.84,
"learning_rate": 2.5874340275496573e-06,
"loss": 0.6998,
"step": 216500
},
{
"epoch": 2.85,
"learning_rate": 2.4779361846571625e-06,
"loss": 0.7017,
"step": 217000
},
{
"epoch": 2.86,
"learning_rate": 2.3684383417646676e-06,
"loss": 0.6945,
"step": 217500
},
{
"epoch": 2.86,
"learning_rate": 2.2589404988721723e-06,
"loss": 0.6952,
"step": 218000
},
{
"epoch": 2.87,
"learning_rate": 2.1494426559796774e-06,
"loss": 0.6965,
"step": 218500
},
{
"epoch": 2.88,
"learning_rate": 2.039944813087182e-06,
"loss": 0.6875,
"step": 219000
},
{
"epoch": 2.88,
"learning_rate": 1.9304469701946872e-06,
"loss": 0.7007,
"step": 219500
},
{
"epoch": 2.89,
"learning_rate": 1.8209491273021924e-06,
"loss": 0.6745,
"step": 220000
},
{
"epoch": 2.9,
"learning_rate": 1.7114512844096973e-06,
"loss": 0.6927,
"step": 220500
},
{
"epoch": 2.9,
"learning_rate": 1.6019534415172022e-06,
"loss": 0.6987,
"step": 221000
},
{
"epoch": 2.91,
"learning_rate": 1.492455598624707e-06,
"loss": 0.6906,
"step": 221500
},
{
"epoch": 2.92,
"learning_rate": 1.382957755732212e-06,
"loss": 0.6886,
"step": 222000
},
{
"epoch": 2.92,
"learning_rate": 1.2734599128397171e-06,
"loss": 0.6876,
"step": 222500
},
{
"epoch": 2.93,
"learning_rate": 1.163962069947222e-06,
"loss": 0.6687,
"step": 223000
},
{
"epoch": 2.94,
"learning_rate": 1.054464227054727e-06,
"loss": 0.69,
"step": 223500
},
{
"epoch": 2.94,
"learning_rate": 9.44966384162232e-07,
"loss": 0.685,
"step": 224000
},
{
"epoch": 2.95,
"learning_rate": 8.35468541269737e-07,
"loss": 0.6765,
"step": 224500
},
{
"epoch": 2.96,
"learning_rate": 7.25970698377242e-07,
"loss": 0.6779,
"step": 225000
},
{
"epoch": 2.96,
"learning_rate": 6.164728554847469e-07,
"loss": 0.6859,
"step": 225500
},
{
"epoch": 2.97,
"learning_rate": 5.06975012592252e-07,
"loss": 0.6783,
"step": 226000
},
{
"epoch": 2.98,
"learning_rate": 3.9747716969975693e-07,
"loss": 0.6843,
"step": 226500
},
{
"epoch": 2.98,
"learning_rate": 2.879793268072619e-07,
"loss": 0.6843,
"step": 227000
},
{
"epoch": 2.99,
"learning_rate": 1.784814839147669e-07,
"loss": 0.6965,
"step": 227500
},
{
"epoch": 3.0,
"learning_rate": 6.898364102227187e-08,
"loss": 0.689,
"step": 228000
},
{
"epoch": 3.0,
"step": 228315,
"total_flos": 3.558682090165371e+17,
"train_loss": 0.946421772361812,
"train_runtime": 62113.3906,
"train_samples_per_second": 36.758,
"train_steps_per_second": 3.676
}
],
"max_steps": 228315,
"num_train_epochs": 3,
"total_flos": 3.558682090165371e+17,
"trial_name": null,
"trial_params": null
}