m2m100_418M_mos_fr_rel_news / trainer_state.json
Davlan's picture
add MT model
4145c96
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 162036,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.98457132982794e-05,
"loss": 3.0748,
"step": 500
},
{
"epoch": 0.02,
"learning_rate": 4.9691426596558794e-05,
"loss": 2.765,
"step": 1000
},
{
"epoch": 0.03,
"learning_rate": 4.953713989483818e-05,
"loss": 2.6724,
"step": 1500
},
{
"epoch": 0.04,
"learning_rate": 4.938285319311758e-05,
"loss": 2.5839,
"step": 2000
},
{
"epoch": 0.05,
"learning_rate": 4.9228566491396974e-05,
"loss": 2.5151,
"step": 2500
},
{
"epoch": 0.06,
"learning_rate": 4.907427978967637e-05,
"loss": 2.4527,
"step": 3000
},
{
"epoch": 0.06,
"learning_rate": 4.8919993087955766e-05,
"loss": 2.4373,
"step": 3500
},
{
"epoch": 0.07,
"learning_rate": 4.876570638623516e-05,
"loss": 2.4024,
"step": 4000
},
{
"epoch": 0.08,
"learning_rate": 4.861141968451455e-05,
"loss": 2.3551,
"step": 4500
},
{
"epoch": 0.09,
"learning_rate": 4.845713298279395e-05,
"loss": 2.2634,
"step": 5000
},
{
"epoch": 0.1,
"learning_rate": 4.830284628107334e-05,
"loss": 2.2679,
"step": 5500
},
{
"epoch": 0.11,
"learning_rate": 4.814855957935274e-05,
"loss": 2.2334,
"step": 6000
},
{
"epoch": 0.12,
"learning_rate": 4.7994272877632133e-05,
"loss": 2.2096,
"step": 6500
},
{
"epoch": 0.13,
"learning_rate": 4.783998617591153e-05,
"loss": 2.2028,
"step": 7000
},
{
"epoch": 0.14,
"learning_rate": 4.768569947419092e-05,
"loss": 2.2095,
"step": 7500
},
{
"epoch": 0.15,
"learning_rate": 4.753141277247032e-05,
"loss": 2.1524,
"step": 8000
},
{
"epoch": 0.16,
"learning_rate": 4.737712607074971e-05,
"loss": 2.2155,
"step": 8500
},
{
"epoch": 0.17,
"learning_rate": 4.722283936902911e-05,
"loss": 2.1692,
"step": 9000
},
{
"epoch": 0.18,
"learning_rate": 4.70685526673085e-05,
"loss": 2.1879,
"step": 9500
},
{
"epoch": 0.19,
"learning_rate": 4.69142659655879e-05,
"loss": 2.146,
"step": 10000
},
{
"epoch": 0.19,
"learning_rate": 4.675997926386729e-05,
"loss": 2.1031,
"step": 10500
},
{
"epoch": 0.2,
"learning_rate": 4.660569256214669e-05,
"loss": 2.0855,
"step": 11000
},
{
"epoch": 0.21,
"learning_rate": 4.645140586042608e-05,
"loss": 2.1089,
"step": 11500
},
{
"epoch": 0.22,
"learning_rate": 4.629711915870548e-05,
"loss": 2.0888,
"step": 12000
},
{
"epoch": 0.23,
"learning_rate": 4.614283245698487e-05,
"loss": 2.067,
"step": 12500
},
{
"epoch": 0.24,
"learning_rate": 4.5988545755264264e-05,
"loss": 2.0437,
"step": 13000
},
{
"epoch": 0.25,
"learning_rate": 4.583425905354366e-05,
"loss": 2.0219,
"step": 13500
},
{
"epoch": 0.26,
"learning_rate": 4.5679972351823056e-05,
"loss": 2.0542,
"step": 14000
},
{
"epoch": 0.27,
"learning_rate": 4.552568565010245e-05,
"loss": 2.0869,
"step": 14500
},
{
"epoch": 0.28,
"learning_rate": 4.537139894838184e-05,
"loss": 2.0638,
"step": 15000
},
{
"epoch": 0.29,
"learning_rate": 4.5217112246661236e-05,
"loss": 1.9949,
"step": 15500
},
{
"epoch": 0.3,
"learning_rate": 4.506282554494063e-05,
"loss": 1.9856,
"step": 16000
},
{
"epoch": 0.31,
"learning_rate": 4.490853884322003e-05,
"loss": 2.0157,
"step": 16500
},
{
"epoch": 0.31,
"learning_rate": 4.475425214149942e-05,
"loss": 2.0186,
"step": 17000
},
{
"epoch": 0.32,
"learning_rate": 4.459996543977882e-05,
"loss": 2.0238,
"step": 17500
},
{
"epoch": 0.33,
"learning_rate": 4.444567873805821e-05,
"loss": 1.9816,
"step": 18000
},
{
"epoch": 0.34,
"learning_rate": 4.4291392036337604e-05,
"loss": 1.9722,
"step": 18500
},
{
"epoch": 0.35,
"learning_rate": 4.4137105334617e-05,
"loss": 1.9828,
"step": 19000
},
{
"epoch": 0.36,
"learning_rate": 4.3982818632896395e-05,
"loss": 1.9801,
"step": 19500
},
{
"epoch": 0.37,
"learning_rate": 4.382853193117579e-05,
"loss": 1.9396,
"step": 20000
},
{
"epoch": 0.38,
"learning_rate": 4.367424522945519e-05,
"loss": 1.9593,
"step": 20500
},
{
"epoch": 0.39,
"learning_rate": 4.3519958527734576e-05,
"loss": 1.9496,
"step": 21000
},
{
"epoch": 0.4,
"learning_rate": 4.336567182601398e-05,
"loss": 1.9351,
"step": 21500
},
{
"epoch": 0.41,
"learning_rate": 4.321138512429337e-05,
"loss": 1.9603,
"step": 22000
},
{
"epoch": 0.42,
"learning_rate": 4.305709842257276e-05,
"loss": 1.918,
"step": 22500
},
{
"epoch": 0.43,
"learning_rate": 4.290281172085216e-05,
"loss": 1.9599,
"step": 23000
},
{
"epoch": 0.44,
"learning_rate": 4.2748525019131555e-05,
"loss": 1.9547,
"step": 23500
},
{
"epoch": 0.44,
"learning_rate": 4.259423831741095e-05,
"loss": 1.9198,
"step": 24000
},
{
"epoch": 0.45,
"learning_rate": 4.2439951615690346e-05,
"loss": 1.9184,
"step": 24500
},
{
"epoch": 0.46,
"learning_rate": 4.2285664913969735e-05,
"loss": 1.9495,
"step": 25000
},
{
"epoch": 0.47,
"learning_rate": 4.213137821224913e-05,
"loss": 1.91,
"step": 25500
},
{
"epoch": 0.48,
"learning_rate": 4.1977091510528526e-05,
"loss": 1.9164,
"step": 26000
},
{
"epoch": 0.49,
"learning_rate": 4.182280480880792e-05,
"loss": 1.9172,
"step": 26500
},
{
"epoch": 0.5,
"learning_rate": 4.166851810708732e-05,
"loss": 1.8811,
"step": 27000
},
{
"epoch": 0.51,
"learning_rate": 4.1514231405366714e-05,
"loss": 1.9077,
"step": 27500
},
{
"epoch": 0.52,
"learning_rate": 4.13599447036461e-05,
"loss": 1.8838,
"step": 28000
},
{
"epoch": 0.53,
"learning_rate": 4.12056580019255e-05,
"loss": 1.9026,
"step": 28500
},
{
"epoch": 0.54,
"learning_rate": 4.1051371300204894e-05,
"loss": 1.8937,
"step": 29000
},
{
"epoch": 0.55,
"learning_rate": 4.089708459848429e-05,
"loss": 1.9169,
"step": 29500
},
{
"epoch": 0.56,
"learning_rate": 4.0742797896763686e-05,
"loss": 1.8863,
"step": 30000
},
{
"epoch": 0.56,
"learning_rate": 4.0588511195043074e-05,
"loss": 1.8467,
"step": 30500
},
{
"epoch": 0.57,
"learning_rate": 4.043422449332248e-05,
"loss": 1.8836,
"step": 31000
},
{
"epoch": 0.58,
"learning_rate": 4.0279937791601866e-05,
"loss": 1.8626,
"step": 31500
},
{
"epoch": 0.59,
"learning_rate": 4.012565108988126e-05,
"loss": 1.8922,
"step": 32000
},
{
"epoch": 0.6,
"learning_rate": 3.997136438816066e-05,
"loss": 1.8939,
"step": 32500
},
{
"epoch": 0.61,
"learning_rate": 3.981707768644005e-05,
"loss": 1.8431,
"step": 33000
},
{
"epoch": 0.62,
"learning_rate": 3.966279098471944e-05,
"loss": 1.8237,
"step": 33500
},
{
"epoch": 0.63,
"learning_rate": 3.9508504282998845e-05,
"loss": 1.8614,
"step": 34000
},
{
"epoch": 0.64,
"learning_rate": 3.9354217581278234e-05,
"loss": 1.8688,
"step": 34500
},
{
"epoch": 0.65,
"learning_rate": 3.9199930879557636e-05,
"loss": 1.8396,
"step": 35000
},
{
"epoch": 0.66,
"learning_rate": 3.9045644177837025e-05,
"loss": 1.8455,
"step": 35500
},
{
"epoch": 0.67,
"learning_rate": 3.889135747611642e-05,
"loss": 1.8384,
"step": 36000
},
{
"epoch": 0.68,
"learning_rate": 3.8737070774395817e-05,
"loss": 1.8301,
"step": 36500
},
{
"epoch": 0.69,
"learning_rate": 3.858278407267521e-05,
"loss": 1.7995,
"step": 37000
},
{
"epoch": 0.69,
"learning_rate": 3.84284973709546e-05,
"loss": 1.8156,
"step": 37500
},
{
"epoch": 0.7,
"learning_rate": 3.8274210669234004e-05,
"loss": 1.8058,
"step": 38000
},
{
"epoch": 0.71,
"learning_rate": 3.811992396751339e-05,
"loss": 1.8164,
"step": 38500
},
{
"epoch": 0.72,
"learning_rate": 3.796563726579279e-05,
"loss": 1.7969,
"step": 39000
},
{
"epoch": 0.73,
"learning_rate": 3.7811350564072184e-05,
"loss": 1.8317,
"step": 39500
},
{
"epoch": 0.74,
"learning_rate": 3.765706386235158e-05,
"loss": 1.8476,
"step": 40000
},
{
"epoch": 0.75,
"learning_rate": 3.7502777160630976e-05,
"loss": 1.8054,
"step": 40500
},
{
"epoch": 0.76,
"learning_rate": 3.7348490458910365e-05,
"loss": 1.8131,
"step": 41000
},
{
"epoch": 0.77,
"learning_rate": 3.719420375718976e-05,
"loss": 1.8101,
"step": 41500
},
{
"epoch": 0.78,
"learning_rate": 3.7039917055469156e-05,
"loss": 1.8161,
"step": 42000
},
{
"epoch": 0.79,
"learning_rate": 3.688563035374855e-05,
"loss": 1.7678,
"step": 42500
},
{
"epoch": 0.8,
"learning_rate": 3.673134365202794e-05,
"loss": 1.793,
"step": 43000
},
{
"epoch": 0.81,
"learning_rate": 3.657705695030734e-05,
"loss": 1.7822,
"step": 43500
},
{
"epoch": 0.81,
"learning_rate": 3.642277024858673e-05,
"loss": 1.8089,
"step": 44000
},
{
"epoch": 0.82,
"learning_rate": 3.626848354686613e-05,
"loss": 1.7637,
"step": 44500
},
{
"epoch": 0.83,
"learning_rate": 3.6114196845145524e-05,
"loss": 1.7801,
"step": 45000
},
{
"epoch": 0.84,
"learning_rate": 3.595991014342492e-05,
"loss": 1.7883,
"step": 45500
},
{
"epoch": 0.85,
"learning_rate": 3.5805623441704315e-05,
"loss": 1.7963,
"step": 46000
},
{
"epoch": 0.86,
"learning_rate": 3.565133673998371e-05,
"loss": 1.7762,
"step": 46500
},
{
"epoch": 0.87,
"learning_rate": 3.54970500382631e-05,
"loss": 1.7566,
"step": 47000
},
{
"epoch": 0.88,
"learning_rate": 3.53427633365425e-05,
"loss": 1.7995,
"step": 47500
},
{
"epoch": 0.89,
"learning_rate": 3.518847663482189e-05,
"loss": 1.7528,
"step": 48000
},
{
"epoch": 0.9,
"learning_rate": 3.503418993310129e-05,
"loss": 1.7332,
"step": 48500
},
{
"epoch": 0.91,
"learning_rate": 3.487990323138068e-05,
"loss": 1.7607,
"step": 49000
},
{
"epoch": 0.92,
"learning_rate": 3.472561652966008e-05,
"loss": 1.7391,
"step": 49500
},
{
"epoch": 0.93,
"learning_rate": 3.457132982793947e-05,
"loss": 1.7475,
"step": 50000
},
{
"epoch": 0.93,
"learning_rate": 3.441704312621887e-05,
"loss": 1.7488,
"step": 50500
},
{
"epoch": 0.94,
"learning_rate": 3.426275642449826e-05,
"loss": 1.7616,
"step": 51000
},
{
"epoch": 0.95,
"learning_rate": 3.410846972277766e-05,
"loss": 1.7395,
"step": 51500
},
{
"epoch": 0.96,
"learning_rate": 3.395418302105705e-05,
"loss": 1.7218,
"step": 52000
},
{
"epoch": 0.97,
"learning_rate": 3.3799896319336446e-05,
"loss": 1.7664,
"step": 52500
},
{
"epoch": 0.98,
"learning_rate": 3.364560961761584e-05,
"loss": 1.7079,
"step": 53000
},
{
"epoch": 0.99,
"learning_rate": 3.349132291589524e-05,
"loss": 1.7082,
"step": 53500
},
{
"epoch": 1.0,
"learning_rate": 3.333703621417463e-05,
"loss": 1.7379,
"step": 54000
},
{
"epoch": 1.01,
"learning_rate": 3.318274951245402e-05,
"loss": 1.5712,
"step": 54500
},
{
"epoch": 1.02,
"learning_rate": 3.302846281073342e-05,
"loss": 1.5532,
"step": 55000
},
{
"epoch": 1.03,
"learning_rate": 3.2874176109012814e-05,
"loss": 1.569,
"step": 55500
},
{
"epoch": 1.04,
"learning_rate": 3.271988940729221e-05,
"loss": 1.5896,
"step": 56000
},
{
"epoch": 1.05,
"learning_rate": 3.25656027055716e-05,
"loss": 1.5641,
"step": 56500
},
{
"epoch": 1.06,
"learning_rate": 3.2411316003851e-05,
"loss": 1.5491,
"step": 57000
},
{
"epoch": 1.06,
"learning_rate": 3.225702930213039e-05,
"loss": 1.6213,
"step": 57500
},
{
"epoch": 1.07,
"learning_rate": 3.2102742600409786e-05,
"loss": 1.5747,
"step": 58000
},
{
"epoch": 1.08,
"learning_rate": 3.194845589868918e-05,
"loss": 1.5745,
"step": 58500
},
{
"epoch": 1.09,
"learning_rate": 3.179416919696858e-05,
"loss": 1.586,
"step": 59000
},
{
"epoch": 1.1,
"learning_rate": 3.1639882495247966e-05,
"loss": 1.5299,
"step": 59500
},
{
"epoch": 1.11,
"learning_rate": 3.148559579352737e-05,
"loss": 1.5755,
"step": 60000
},
{
"epoch": 1.12,
"learning_rate": 3.133130909180676e-05,
"loss": 1.5649,
"step": 60500
},
{
"epoch": 1.13,
"learning_rate": 3.117702239008615e-05,
"loss": 1.5753,
"step": 61000
},
{
"epoch": 1.14,
"learning_rate": 3.102273568836555e-05,
"loss": 1.568,
"step": 61500
},
{
"epoch": 1.15,
"learning_rate": 3.0868448986644945e-05,
"loss": 1.5849,
"step": 62000
},
{
"epoch": 1.16,
"learning_rate": 3.071416228492434e-05,
"loss": 1.5745,
"step": 62500
},
{
"epoch": 1.17,
"learning_rate": 3.0559875583203736e-05,
"loss": 1.5251,
"step": 63000
},
{
"epoch": 1.18,
"learning_rate": 3.0405588881483125e-05,
"loss": 1.5551,
"step": 63500
},
{
"epoch": 1.18,
"learning_rate": 3.0251302179762524e-05,
"loss": 1.5737,
"step": 64000
},
{
"epoch": 1.19,
"learning_rate": 3.0097015478041917e-05,
"loss": 1.5808,
"step": 64500
},
{
"epoch": 1.2,
"learning_rate": 2.994272877632131e-05,
"loss": 1.559,
"step": 65000
},
{
"epoch": 1.21,
"learning_rate": 2.9788442074600708e-05,
"loss": 1.5252,
"step": 65500
},
{
"epoch": 1.22,
"learning_rate": 2.96341553728801e-05,
"loss": 1.5092,
"step": 66000
},
{
"epoch": 1.23,
"learning_rate": 2.94798686711595e-05,
"loss": 1.5717,
"step": 66500
},
{
"epoch": 1.24,
"learning_rate": 2.9325581969438892e-05,
"loss": 1.5577,
"step": 67000
},
{
"epoch": 1.25,
"learning_rate": 2.9171295267718284e-05,
"loss": 1.5429,
"step": 67500
},
{
"epoch": 1.26,
"learning_rate": 2.9017008565997683e-05,
"loss": 1.5632,
"step": 68000
},
{
"epoch": 1.27,
"learning_rate": 2.8862721864277076e-05,
"loss": 1.5468,
"step": 68500
},
{
"epoch": 1.28,
"learning_rate": 2.8708435162556468e-05,
"loss": 1.5373,
"step": 69000
},
{
"epoch": 1.29,
"learning_rate": 2.8554148460835867e-05,
"loss": 1.5475,
"step": 69500
},
{
"epoch": 1.3,
"learning_rate": 2.839986175911526e-05,
"loss": 1.5482,
"step": 70000
},
{
"epoch": 1.31,
"learning_rate": 2.8245575057394652e-05,
"loss": 1.5871,
"step": 70500
},
{
"epoch": 1.31,
"learning_rate": 2.809128835567405e-05,
"loss": 1.5633,
"step": 71000
},
{
"epoch": 1.32,
"learning_rate": 2.7937001653953443e-05,
"loss": 1.5472,
"step": 71500
},
{
"epoch": 1.33,
"learning_rate": 2.7782714952232843e-05,
"loss": 1.5367,
"step": 72000
},
{
"epoch": 1.34,
"learning_rate": 2.7628428250512235e-05,
"loss": 1.5363,
"step": 72500
},
{
"epoch": 1.35,
"learning_rate": 2.7474141548791627e-05,
"loss": 1.5582,
"step": 73000
},
{
"epoch": 1.36,
"learning_rate": 2.7319854847071026e-05,
"loss": 1.5433,
"step": 73500
},
{
"epoch": 1.37,
"learning_rate": 2.716556814535042e-05,
"loss": 1.5969,
"step": 74000
},
{
"epoch": 1.38,
"learning_rate": 2.701128144362981e-05,
"loss": 1.544,
"step": 74500
},
{
"epoch": 1.39,
"learning_rate": 2.6856994741909207e-05,
"loss": 1.5245,
"step": 75000
},
{
"epoch": 1.4,
"learning_rate": 2.6702708040188603e-05,
"loss": 1.4995,
"step": 75500
},
{
"epoch": 1.41,
"learning_rate": 2.6548421338467995e-05,
"loss": 1.5435,
"step": 76000
},
{
"epoch": 1.42,
"learning_rate": 2.639413463674739e-05,
"loss": 1.5175,
"step": 76500
},
{
"epoch": 1.43,
"learning_rate": 2.6239847935026783e-05,
"loss": 1.5681,
"step": 77000
},
{
"epoch": 1.43,
"learning_rate": 2.6085561233306182e-05,
"loss": 1.522,
"step": 77500
},
{
"epoch": 1.44,
"learning_rate": 2.5931274531585574e-05,
"loss": 1.5523,
"step": 78000
},
{
"epoch": 1.45,
"learning_rate": 2.5776987829864967e-05,
"loss": 1.542,
"step": 78500
},
{
"epoch": 1.46,
"learning_rate": 2.5622701128144366e-05,
"loss": 1.5189,
"step": 79000
},
{
"epoch": 1.47,
"learning_rate": 2.5468414426423758e-05,
"loss": 1.5307,
"step": 79500
},
{
"epoch": 1.48,
"learning_rate": 2.531412772470315e-05,
"loss": 1.5297,
"step": 80000
},
{
"epoch": 1.49,
"learning_rate": 2.515984102298255e-05,
"loss": 1.5736,
"step": 80500
},
{
"epoch": 1.5,
"learning_rate": 2.5005554321261942e-05,
"loss": 1.5151,
"step": 81000
},
{
"epoch": 1.51,
"learning_rate": 2.4851267619541338e-05,
"loss": 1.4902,
"step": 81500
},
{
"epoch": 1.52,
"learning_rate": 2.4696980917820734e-05,
"loss": 1.5353,
"step": 82000
},
{
"epoch": 1.53,
"learning_rate": 2.4542694216100126e-05,
"loss": 1.5022,
"step": 82500
},
{
"epoch": 1.54,
"learning_rate": 2.438840751437952e-05,
"loss": 1.533,
"step": 83000
},
{
"epoch": 1.55,
"learning_rate": 2.4234120812658917e-05,
"loss": 1.5344,
"step": 83500
},
{
"epoch": 1.56,
"learning_rate": 2.4079834110938313e-05,
"loss": 1.5378,
"step": 84000
},
{
"epoch": 1.56,
"learning_rate": 2.3925547409217705e-05,
"loss": 1.5323,
"step": 84500
},
{
"epoch": 1.57,
"learning_rate": 2.37712607074971e-05,
"loss": 1.5344,
"step": 85000
},
{
"epoch": 1.58,
"learning_rate": 2.3616974005776497e-05,
"loss": 1.4994,
"step": 85500
},
{
"epoch": 1.59,
"learning_rate": 2.346268730405589e-05,
"loss": 1.546,
"step": 86000
},
{
"epoch": 1.6,
"learning_rate": 2.3308400602335285e-05,
"loss": 1.4968,
"step": 86500
},
{
"epoch": 1.61,
"learning_rate": 2.315411390061468e-05,
"loss": 1.4917,
"step": 87000
},
{
"epoch": 1.62,
"learning_rate": 2.2999827198894077e-05,
"loss": 1.5162,
"step": 87500
},
{
"epoch": 1.63,
"learning_rate": 2.284554049717347e-05,
"loss": 1.5228,
"step": 88000
},
{
"epoch": 1.64,
"learning_rate": 2.2691253795452865e-05,
"loss": 1.5383,
"step": 88500
},
{
"epoch": 1.65,
"learning_rate": 2.2536967093732257e-05,
"loss": 1.5145,
"step": 89000
},
{
"epoch": 1.66,
"learning_rate": 2.2382680392011653e-05,
"loss": 1.483,
"step": 89500
},
{
"epoch": 1.67,
"learning_rate": 2.2228393690291045e-05,
"loss": 1.4963,
"step": 90000
},
{
"epoch": 1.68,
"learning_rate": 2.207410698857044e-05,
"loss": 1.5442,
"step": 90500
},
{
"epoch": 1.68,
"learning_rate": 2.1919820286849836e-05,
"loss": 1.5112,
"step": 91000
},
{
"epoch": 1.69,
"learning_rate": 2.176553358512923e-05,
"loss": 1.524,
"step": 91500
},
{
"epoch": 1.7,
"learning_rate": 2.1611246883408625e-05,
"loss": 1.4932,
"step": 92000
},
{
"epoch": 1.71,
"learning_rate": 2.145696018168802e-05,
"loss": 1.5076,
"step": 92500
},
{
"epoch": 1.72,
"learning_rate": 2.1302673479967416e-05,
"loss": 1.5177,
"step": 93000
},
{
"epoch": 1.73,
"learning_rate": 2.114838677824681e-05,
"loss": 1.5127,
"step": 93500
},
{
"epoch": 1.74,
"learning_rate": 2.0994100076526204e-05,
"loss": 1.4766,
"step": 94000
},
{
"epoch": 1.75,
"learning_rate": 2.08398133748056e-05,
"loss": 1.5082,
"step": 94500
},
{
"epoch": 1.76,
"learning_rate": 2.0685526673084996e-05,
"loss": 1.4943,
"step": 95000
},
{
"epoch": 1.77,
"learning_rate": 2.0531239971364388e-05,
"loss": 1.5166,
"step": 95500
},
{
"epoch": 1.78,
"learning_rate": 2.0376953269643784e-05,
"loss": 1.4579,
"step": 96000
},
{
"epoch": 1.79,
"learning_rate": 2.022266656792318e-05,
"loss": 1.4811,
"step": 96500
},
{
"epoch": 1.8,
"learning_rate": 2.0068379866202575e-05,
"loss": 1.445,
"step": 97000
},
{
"epoch": 1.81,
"learning_rate": 1.9914093164481967e-05,
"loss": 1.4655,
"step": 97500
},
{
"epoch": 1.81,
"learning_rate": 1.9759806462761363e-05,
"loss": 1.524,
"step": 98000
},
{
"epoch": 1.82,
"learning_rate": 1.960551976104076e-05,
"loss": 1.4791,
"step": 98500
},
{
"epoch": 1.83,
"learning_rate": 1.945123305932015e-05,
"loss": 1.4701,
"step": 99000
},
{
"epoch": 1.84,
"learning_rate": 1.9296946357599547e-05,
"loss": 1.5028,
"step": 99500
},
{
"epoch": 1.85,
"learning_rate": 1.9142659655878943e-05,
"loss": 1.5011,
"step": 100000
},
{
"epoch": 1.86,
"learning_rate": 1.898837295415834e-05,
"loss": 1.4333,
"step": 100500
},
{
"epoch": 1.87,
"learning_rate": 1.883408625243773e-05,
"loss": 1.5043,
"step": 101000
},
{
"epoch": 1.88,
"learning_rate": 1.8679799550717127e-05,
"loss": 1.4852,
"step": 101500
},
{
"epoch": 1.89,
"learning_rate": 1.8525512848996522e-05,
"loss": 1.4806,
"step": 102000
},
{
"epoch": 1.9,
"learning_rate": 1.8371226147275915e-05,
"loss": 1.4943,
"step": 102500
},
{
"epoch": 1.91,
"learning_rate": 1.821693944555531e-05,
"loss": 1.4541,
"step": 103000
},
{
"epoch": 1.92,
"learning_rate": 1.8062652743834703e-05,
"loss": 1.4638,
"step": 103500
},
{
"epoch": 1.93,
"learning_rate": 1.79083660421141e-05,
"loss": 1.5026,
"step": 104000
},
{
"epoch": 1.93,
"learning_rate": 1.775407934039349e-05,
"loss": 1.4907,
"step": 104500
},
{
"epoch": 1.94,
"learning_rate": 1.7599792638672887e-05,
"loss": 1.4866,
"step": 105000
},
{
"epoch": 1.95,
"learning_rate": 1.7445505936952282e-05,
"loss": 1.488,
"step": 105500
},
{
"epoch": 1.96,
"learning_rate": 1.7291219235231678e-05,
"loss": 1.4508,
"step": 106000
},
{
"epoch": 1.97,
"learning_rate": 1.713693253351107e-05,
"loss": 1.4706,
"step": 106500
},
{
"epoch": 1.98,
"learning_rate": 1.6982645831790466e-05,
"loss": 1.4555,
"step": 107000
},
{
"epoch": 1.99,
"learning_rate": 1.6828359130069862e-05,
"loss": 1.4772,
"step": 107500
},
{
"epoch": 2.0,
"learning_rate": 1.6674072428349258e-05,
"loss": 1.479,
"step": 108000
},
{
"epoch": 2.01,
"learning_rate": 1.651978572662865e-05,
"loss": 1.2942,
"step": 108500
},
{
"epoch": 2.02,
"learning_rate": 1.6365499024908046e-05,
"loss": 1.3319,
"step": 109000
},
{
"epoch": 2.03,
"learning_rate": 1.621121232318744e-05,
"loss": 1.279,
"step": 109500
},
{
"epoch": 2.04,
"learning_rate": 1.6056925621466834e-05,
"loss": 1.2887,
"step": 110000
},
{
"epoch": 2.05,
"learning_rate": 1.590263891974623e-05,
"loss": 1.2943,
"step": 110500
},
{
"epoch": 2.06,
"learning_rate": 1.5748352218025625e-05,
"loss": 1.2976,
"step": 111000
},
{
"epoch": 2.06,
"learning_rate": 1.559406551630502e-05,
"loss": 1.303,
"step": 111500
},
{
"epoch": 2.07,
"learning_rate": 1.5439778814584413e-05,
"loss": 1.2693,
"step": 112000
},
{
"epoch": 2.08,
"learning_rate": 1.528549211286381e-05,
"loss": 1.262,
"step": 112500
},
{
"epoch": 2.09,
"learning_rate": 1.5131205411143203e-05,
"loss": 1.3304,
"step": 113000
},
{
"epoch": 2.1,
"learning_rate": 1.4976918709422599e-05,
"loss": 1.3131,
"step": 113500
},
{
"epoch": 2.11,
"learning_rate": 1.4822632007701991e-05,
"loss": 1.2838,
"step": 114000
},
{
"epoch": 2.12,
"learning_rate": 1.4668345305981387e-05,
"loss": 1.2936,
"step": 114500
},
{
"epoch": 2.13,
"learning_rate": 1.4514058604260783e-05,
"loss": 1.2762,
"step": 115000
},
{
"epoch": 2.14,
"learning_rate": 1.4359771902540175e-05,
"loss": 1.2612,
"step": 115500
},
{
"epoch": 2.15,
"learning_rate": 1.420548520081957e-05,
"loss": 1.3006,
"step": 116000
},
{
"epoch": 2.16,
"learning_rate": 1.4051198499098966e-05,
"loss": 1.2681,
"step": 116500
},
{
"epoch": 2.17,
"learning_rate": 1.3896911797378362e-05,
"loss": 1.2798,
"step": 117000
},
{
"epoch": 2.18,
"learning_rate": 1.3742625095657755e-05,
"loss": 1.2849,
"step": 117500
},
{
"epoch": 2.18,
"learning_rate": 1.358833839393715e-05,
"loss": 1.2663,
"step": 118000
},
{
"epoch": 2.19,
"learning_rate": 1.3434051692216546e-05,
"loss": 1.2937,
"step": 118500
},
{
"epoch": 2.2,
"learning_rate": 1.3279764990495942e-05,
"loss": 1.2697,
"step": 119000
},
{
"epoch": 2.21,
"learning_rate": 1.3125478288775334e-05,
"loss": 1.3,
"step": 119500
},
{
"epoch": 2.22,
"learning_rate": 1.297119158705473e-05,
"loss": 1.2904,
"step": 120000
},
{
"epoch": 2.23,
"learning_rate": 1.2816904885334124e-05,
"loss": 1.246,
"step": 120500
},
{
"epoch": 2.24,
"learning_rate": 1.266261818361352e-05,
"loss": 1.2941,
"step": 121000
},
{
"epoch": 2.25,
"learning_rate": 1.2508331481892912e-05,
"loss": 1.2697,
"step": 121500
},
{
"epoch": 2.26,
"learning_rate": 1.2354044780172308e-05,
"loss": 1.2893,
"step": 122000
},
{
"epoch": 2.27,
"learning_rate": 1.2199758078451702e-05,
"loss": 1.2655,
"step": 122500
},
{
"epoch": 2.28,
"learning_rate": 1.2045471376731097e-05,
"loss": 1.2802,
"step": 123000
},
{
"epoch": 2.29,
"learning_rate": 1.1891184675010492e-05,
"loss": 1.3079,
"step": 123500
},
{
"epoch": 2.3,
"learning_rate": 1.1736897973289887e-05,
"loss": 1.3249,
"step": 124000
},
{
"epoch": 2.31,
"learning_rate": 1.1582611271569281e-05,
"loss": 1.2662,
"step": 124500
},
{
"epoch": 2.31,
"learning_rate": 1.1428324569848677e-05,
"loss": 1.2614,
"step": 125000
},
{
"epoch": 2.32,
"learning_rate": 1.1274037868128071e-05,
"loss": 1.2539,
"step": 125500
},
{
"epoch": 2.33,
"learning_rate": 1.1119751166407467e-05,
"loss": 1.2666,
"step": 126000
},
{
"epoch": 2.34,
"learning_rate": 1.0965464464686861e-05,
"loss": 1.2837,
"step": 126500
},
{
"epoch": 2.35,
"learning_rate": 1.0811177762966255e-05,
"loss": 1.2923,
"step": 127000
},
{
"epoch": 2.36,
"learning_rate": 1.0656891061245649e-05,
"loss": 1.2983,
"step": 127500
},
{
"epoch": 2.37,
"learning_rate": 1.0502604359525043e-05,
"loss": 1.2517,
"step": 128000
},
{
"epoch": 2.38,
"learning_rate": 1.0348317657804439e-05,
"loss": 1.2587,
"step": 128500
},
{
"epoch": 2.39,
"learning_rate": 1.0194030956083833e-05,
"loss": 1.237,
"step": 129000
},
{
"epoch": 2.4,
"learning_rate": 1.0039744254363228e-05,
"loss": 1.2912,
"step": 129500
},
{
"epoch": 2.41,
"learning_rate": 9.885457552642623e-06,
"loss": 1.2989,
"step": 130000
},
{
"epoch": 2.42,
"learning_rate": 9.731170850922018e-06,
"loss": 1.2758,
"step": 130500
},
{
"epoch": 2.43,
"learning_rate": 9.576884149201412e-06,
"loss": 1.2666,
"step": 131000
},
{
"epoch": 2.43,
"learning_rate": 9.422597447480808e-06,
"loss": 1.2811,
"step": 131500
},
{
"epoch": 2.44,
"learning_rate": 9.268310745760202e-06,
"loss": 1.2316,
"step": 132000
},
{
"epoch": 2.45,
"learning_rate": 9.114024044039598e-06,
"loss": 1.2884,
"step": 132500
},
{
"epoch": 2.46,
"learning_rate": 8.959737342318992e-06,
"loss": 1.2721,
"step": 133000
},
{
"epoch": 2.47,
"learning_rate": 8.805450640598386e-06,
"loss": 1.2255,
"step": 133500
},
{
"epoch": 2.48,
"learning_rate": 8.65116393887778e-06,
"loss": 1.2724,
"step": 134000
},
{
"epoch": 2.49,
"learning_rate": 8.496877237157174e-06,
"loss": 1.2718,
"step": 134500
},
{
"epoch": 2.5,
"learning_rate": 8.34259053543657e-06,
"loss": 1.2656,
"step": 135000
},
{
"epoch": 2.51,
"learning_rate": 8.188303833715964e-06,
"loss": 1.2544,
"step": 135500
},
{
"epoch": 2.52,
"learning_rate": 8.03401713199536e-06,
"loss": 1.2826,
"step": 136000
},
{
"epoch": 2.53,
"learning_rate": 7.879730430274754e-06,
"loss": 1.2579,
"step": 136500
},
{
"epoch": 2.54,
"learning_rate": 7.72544372855415e-06,
"loss": 1.2728,
"step": 137000
},
{
"epoch": 2.55,
"learning_rate": 7.571157026833543e-06,
"loss": 1.2674,
"step": 137500
},
{
"epoch": 2.55,
"learning_rate": 7.416870325112939e-06,
"loss": 1.2948,
"step": 138000
},
{
"epoch": 2.56,
"learning_rate": 7.262583623392333e-06,
"loss": 1.3085,
"step": 138500
},
{
"epoch": 2.57,
"learning_rate": 7.108296921671728e-06,
"loss": 1.2522,
"step": 139000
},
{
"epoch": 2.58,
"learning_rate": 6.954010219951122e-06,
"loss": 1.2579,
"step": 139500
},
{
"epoch": 2.59,
"learning_rate": 6.799723518230516e-06,
"loss": 1.2471,
"step": 140000
},
{
"epoch": 2.6,
"learning_rate": 6.645436816509912e-06,
"loss": 1.2436,
"step": 140500
},
{
"epoch": 2.61,
"learning_rate": 6.491150114789306e-06,
"loss": 1.2643,
"step": 141000
},
{
"epoch": 2.62,
"learning_rate": 6.3368634130687016e-06,
"loss": 1.2587,
"step": 141500
},
{
"epoch": 2.63,
"learning_rate": 6.1825767113480965e-06,
"loss": 1.2793,
"step": 142000
},
{
"epoch": 2.64,
"learning_rate": 6.0282900096274905e-06,
"loss": 1.2608,
"step": 142500
},
{
"epoch": 2.65,
"learning_rate": 5.8740033079068845e-06,
"loss": 1.2608,
"step": 143000
},
{
"epoch": 2.66,
"learning_rate": 5.7197166061862794e-06,
"loss": 1.2405,
"step": 143500
},
{
"epoch": 2.67,
"learning_rate": 5.565429904465674e-06,
"loss": 1.2692,
"step": 144000
},
{
"epoch": 2.68,
"learning_rate": 5.411143202745069e-06,
"loss": 1.2318,
"step": 144500
},
{
"epoch": 2.68,
"learning_rate": 5.256856501024464e-06,
"loss": 1.2593,
"step": 145000
},
{
"epoch": 2.69,
"learning_rate": 5.102569799303859e-06,
"loss": 1.2494,
"step": 145500
},
{
"epoch": 2.7,
"learning_rate": 4.948283097583253e-06,
"loss": 1.2337,
"step": 146000
},
{
"epoch": 2.71,
"learning_rate": 4.793996395862648e-06,
"loss": 1.2257,
"step": 146500
},
{
"epoch": 2.72,
"learning_rate": 4.639709694142043e-06,
"loss": 1.2534,
"step": 147000
},
{
"epoch": 2.73,
"learning_rate": 4.485422992421438e-06,
"loss": 1.2547,
"step": 147500
},
{
"epoch": 2.74,
"learning_rate": 4.331136290700833e-06,
"loss": 1.23,
"step": 148000
},
{
"epoch": 2.75,
"learning_rate": 4.1768495889802275e-06,
"loss": 1.2586,
"step": 148500
},
{
"epoch": 2.76,
"learning_rate": 4.0225628872596215e-06,
"loss": 1.218,
"step": 149000
},
{
"epoch": 2.77,
"learning_rate": 3.8682761855390155e-06,
"loss": 1.255,
"step": 149500
},
{
"epoch": 2.78,
"learning_rate": 3.713989483818411e-06,
"loss": 1.2545,
"step": 150000
},
{
"epoch": 2.79,
"learning_rate": 3.5597027820978053e-06,
"loss": 1.2273,
"step": 150500
},
{
"epoch": 2.8,
"learning_rate": 3.4054160803772002e-06,
"loss": 1.2631,
"step": 151000
},
{
"epoch": 2.8,
"learning_rate": 3.251129378656595e-06,
"loss": 1.254,
"step": 151500
},
{
"epoch": 2.81,
"learning_rate": 3.0968426769359896e-06,
"loss": 1.2571,
"step": 152000
},
{
"epoch": 2.82,
"learning_rate": 2.9425559752153845e-06,
"loss": 1.2281,
"step": 152500
},
{
"epoch": 2.83,
"learning_rate": 2.7882692734947793e-06,
"loss": 1.2298,
"step": 153000
},
{
"epoch": 2.84,
"learning_rate": 2.633982571774174e-06,
"loss": 1.2516,
"step": 153500
},
{
"epoch": 2.85,
"learning_rate": 2.4796958700535683e-06,
"loss": 1.2531,
"step": 154000
},
{
"epoch": 2.86,
"learning_rate": 2.325409168332963e-06,
"loss": 1.2439,
"step": 154500
},
{
"epoch": 2.87,
"learning_rate": 2.171122466612358e-06,
"loss": 1.2582,
"step": 155000
},
{
"epoch": 2.88,
"learning_rate": 2.0168357648917525e-06,
"loss": 1.2494,
"step": 155500
},
{
"epoch": 2.89,
"learning_rate": 1.8625490631711474e-06,
"loss": 1.2354,
"step": 156000
},
{
"epoch": 2.9,
"learning_rate": 1.7082623614505419e-06,
"loss": 1.2444,
"step": 156500
},
{
"epoch": 2.91,
"learning_rate": 1.5539756597299368e-06,
"loss": 1.2313,
"step": 157000
},
{
"epoch": 2.92,
"learning_rate": 1.3996889580093312e-06,
"loss": 1.2268,
"step": 157500
},
{
"epoch": 2.93,
"learning_rate": 1.2454022562887261e-06,
"loss": 1.2137,
"step": 158000
},
{
"epoch": 2.93,
"learning_rate": 1.0911155545681208e-06,
"loss": 1.2122,
"step": 158500
},
{
"epoch": 2.94,
"learning_rate": 9.368288528475154e-07,
"loss": 1.2267,
"step": 159000
},
{
"epoch": 2.95,
"learning_rate": 7.825421511269101e-07,
"loss": 1.2157,
"step": 159500
},
{
"epoch": 2.96,
"learning_rate": 6.282554494063048e-07,
"loss": 1.2435,
"step": 160000
},
{
"epoch": 2.97,
"learning_rate": 4.7396874768569954e-07,
"loss": 1.2441,
"step": 160500
},
{
"epoch": 2.98,
"learning_rate": 3.196820459650942e-07,
"loss": 1.2458,
"step": 161000
},
{
"epoch": 2.99,
"learning_rate": 1.653953442444889e-07,
"loss": 1.2468,
"step": 161500
},
{
"epoch": 3.0,
"learning_rate": 1.1108642523883582e-08,
"loss": 1.2109,
"step": 162000
},
{
"epoch": 3.0,
"step": 162036,
"total_flos": 1.028572903536722e+17,
"train_loss": 1.5869738945279097,
"train_runtime": 29529.1551,
"train_samples_per_second": 21.949,
"train_steps_per_second": 5.487
}
],
"max_steps": 162036,
"num_train_epochs": 3,
"total_flos": 1.028572903536722e+17,
"trial_name": null,
"trial_params": null
}