m2m100_418M_kin_en_rel / trainer_state.json
Davlan's picture
Upload trainer_state.json
db33061
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 145677,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.982838745992847e-05,
"loss": 3.161,
"step": 500
},
{
"epoch": 0.02,
"learning_rate": 4.965677491985695e-05,
"loss": 2.686,
"step": 1000
},
{
"epoch": 0.03,
"learning_rate": 4.948516237978542e-05,
"loss": 2.5,
"step": 1500
},
{
"epoch": 0.04,
"learning_rate": 4.9313549839713894e-05,
"loss": 2.3421,
"step": 2000
},
{
"epoch": 0.05,
"learning_rate": 4.914193729964236e-05,
"loss": 2.3123,
"step": 2500
},
{
"epoch": 0.06,
"learning_rate": 4.897032475957083e-05,
"loss": 2.2321,
"step": 3000
},
{
"epoch": 0.07,
"learning_rate": 4.87987122194993e-05,
"loss": 2.1761,
"step": 3500
},
{
"epoch": 0.08,
"learning_rate": 4.862709967942778e-05,
"loss": 2.1261,
"step": 4000
},
{
"epoch": 0.09,
"learning_rate": 4.845548713935625e-05,
"loss": 2.109,
"step": 4500
},
{
"epoch": 0.1,
"learning_rate": 4.8283874599284724e-05,
"loss": 2.0872,
"step": 5000
},
{
"epoch": 0.11,
"learning_rate": 4.811226205921319e-05,
"loss": 2.036,
"step": 5500
},
{
"epoch": 0.12,
"learning_rate": 4.794064951914167e-05,
"loss": 2.0286,
"step": 6000
},
{
"epoch": 0.13,
"learning_rate": 4.776903697907014e-05,
"loss": 1.9971,
"step": 6500
},
{
"epoch": 0.14,
"learning_rate": 4.759742443899861e-05,
"loss": 1.9656,
"step": 7000
},
{
"epoch": 0.15,
"learning_rate": 4.742581189892708e-05,
"loss": 1.9784,
"step": 7500
},
{
"epoch": 0.16,
"learning_rate": 4.725419935885555e-05,
"loss": 1.924,
"step": 8000
},
{
"epoch": 0.18,
"learning_rate": 4.708258681878402e-05,
"loss": 1.9137,
"step": 8500
},
{
"epoch": 0.19,
"learning_rate": 4.691097427871249e-05,
"loss": 1.9194,
"step": 9000
},
{
"epoch": 0.2,
"learning_rate": 4.673936173864097e-05,
"loss": 1.8744,
"step": 9500
},
{
"epoch": 0.21,
"learning_rate": 4.656774919856944e-05,
"loss": 1.8964,
"step": 10000
},
{
"epoch": 0.22,
"learning_rate": 4.6396136658497914e-05,
"loss": 1.8515,
"step": 10500
},
{
"epoch": 0.23,
"learning_rate": 4.6224524118426383e-05,
"loss": 1.842,
"step": 11000
},
{
"epoch": 0.24,
"learning_rate": 4.605291157835486e-05,
"loss": 1.8577,
"step": 11500
},
{
"epoch": 0.25,
"learning_rate": 4.588129903828333e-05,
"loss": 1.8402,
"step": 12000
},
{
"epoch": 0.26,
"learning_rate": 4.57096864982118e-05,
"loss": 1.8155,
"step": 12500
},
{
"epoch": 0.27,
"learning_rate": 4.553807395814027e-05,
"loss": 1.8154,
"step": 13000
},
{
"epoch": 0.28,
"learning_rate": 4.5366461418068744e-05,
"loss": 1.796,
"step": 13500
},
{
"epoch": 0.29,
"learning_rate": 4.5194848877997213e-05,
"loss": 1.7842,
"step": 14000
},
{
"epoch": 0.3,
"learning_rate": 4.502323633792569e-05,
"loss": 1.7819,
"step": 14500
},
{
"epoch": 0.31,
"learning_rate": 4.485162379785416e-05,
"loss": 1.7769,
"step": 15000
},
{
"epoch": 0.32,
"learning_rate": 4.4680011257782635e-05,
"loss": 1.7649,
"step": 15500
},
{
"epoch": 0.33,
"learning_rate": 4.4508398717711105e-05,
"loss": 1.7653,
"step": 16000
},
{
"epoch": 0.34,
"learning_rate": 4.4336786177639574e-05,
"loss": 1.7418,
"step": 16500
},
{
"epoch": 0.35,
"learning_rate": 4.416517363756805e-05,
"loss": 1.7284,
"step": 17000
},
{
"epoch": 0.36,
"learning_rate": 4.399356109749652e-05,
"loss": 1.7218,
"step": 17500
},
{
"epoch": 0.37,
"learning_rate": 4.382194855742499e-05,
"loss": 1.719,
"step": 18000
},
{
"epoch": 0.38,
"learning_rate": 4.365033601735346e-05,
"loss": 1.6981,
"step": 18500
},
{
"epoch": 0.39,
"learning_rate": 4.3478723477281935e-05,
"loss": 1.702,
"step": 19000
},
{
"epoch": 0.4,
"learning_rate": 4.3307110937210404e-05,
"loss": 1.7055,
"step": 19500
},
{
"epoch": 0.41,
"learning_rate": 4.313549839713888e-05,
"loss": 1.7072,
"step": 20000
},
{
"epoch": 0.42,
"learning_rate": 4.296388585706735e-05,
"loss": 1.6702,
"step": 20500
},
{
"epoch": 0.43,
"learning_rate": 4.2792273316995826e-05,
"loss": 1.6595,
"step": 21000
},
{
"epoch": 0.44,
"learning_rate": 4.2620660776924295e-05,
"loss": 1.6889,
"step": 21500
},
{
"epoch": 0.45,
"learning_rate": 4.2449048236852765e-05,
"loss": 1.6576,
"step": 22000
},
{
"epoch": 0.46,
"learning_rate": 4.2277435696781234e-05,
"loss": 1.6732,
"step": 22500
},
{
"epoch": 0.47,
"learning_rate": 4.210582315670971e-05,
"loss": 1.6441,
"step": 23000
},
{
"epoch": 0.48,
"learning_rate": 4.193421061663818e-05,
"loss": 1.6511,
"step": 23500
},
{
"epoch": 0.49,
"learning_rate": 4.176259807656665e-05,
"loss": 1.6529,
"step": 24000
},
{
"epoch": 0.5,
"learning_rate": 4.1590985536495125e-05,
"loss": 1.6562,
"step": 24500
},
{
"epoch": 0.51,
"learning_rate": 4.1419372996423594e-05,
"loss": 1.6359,
"step": 25000
},
{
"epoch": 0.53,
"learning_rate": 4.124776045635207e-05,
"loss": 1.6404,
"step": 25500
},
{
"epoch": 0.54,
"learning_rate": 4.107614791628054e-05,
"loss": 1.6256,
"step": 26000
},
{
"epoch": 0.55,
"learning_rate": 4.0904535376209016e-05,
"loss": 1.6337,
"step": 26500
},
{
"epoch": 0.56,
"learning_rate": 4.0732922836137486e-05,
"loss": 1.6198,
"step": 27000
},
{
"epoch": 0.57,
"learning_rate": 4.0561310296065955e-05,
"loss": 1.616,
"step": 27500
},
{
"epoch": 0.58,
"learning_rate": 4.0389697755994424e-05,
"loss": 1.6317,
"step": 28000
},
{
"epoch": 0.59,
"learning_rate": 4.02180852159229e-05,
"loss": 1.6145,
"step": 28500
},
{
"epoch": 0.6,
"learning_rate": 4.004647267585137e-05,
"loss": 1.6096,
"step": 29000
},
{
"epoch": 0.61,
"learning_rate": 3.9874860135779846e-05,
"loss": 1.5697,
"step": 29500
},
{
"epoch": 0.62,
"learning_rate": 3.9703247595708316e-05,
"loss": 1.5838,
"step": 30000
},
{
"epoch": 0.63,
"learning_rate": 3.953163505563679e-05,
"loss": 1.5767,
"step": 30500
},
{
"epoch": 0.64,
"learning_rate": 3.936002251556526e-05,
"loss": 1.5776,
"step": 31000
},
{
"epoch": 0.65,
"learning_rate": 3.918840997549373e-05,
"loss": 1.5802,
"step": 31500
},
{
"epoch": 0.66,
"learning_rate": 3.90167974354222e-05,
"loss": 1.5711,
"step": 32000
},
{
"epoch": 0.67,
"learning_rate": 3.884518489535067e-05,
"loss": 1.5662,
"step": 32500
},
{
"epoch": 0.68,
"learning_rate": 3.8673572355279146e-05,
"loss": 1.592,
"step": 33000
},
{
"epoch": 0.69,
"learning_rate": 3.8501959815207615e-05,
"loss": 1.5665,
"step": 33500
},
{
"epoch": 0.7,
"learning_rate": 3.833034727513609e-05,
"loss": 1.5698,
"step": 34000
},
{
"epoch": 0.71,
"learning_rate": 3.815873473506456e-05,
"loss": 1.5548,
"step": 34500
},
{
"epoch": 0.72,
"learning_rate": 3.798712219499304e-05,
"loss": 1.5993,
"step": 35000
},
{
"epoch": 0.73,
"learning_rate": 3.7815509654921506e-05,
"loss": 1.5348,
"step": 35500
},
{
"epoch": 0.74,
"learning_rate": 3.764389711484998e-05,
"loss": 1.5567,
"step": 36000
},
{
"epoch": 0.75,
"learning_rate": 3.747228457477845e-05,
"loss": 1.566,
"step": 36500
},
{
"epoch": 0.76,
"learning_rate": 3.730067203470692e-05,
"loss": 1.5338,
"step": 37000
},
{
"epoch": 0.77,
"learning_rate": 3.712905949463539e-05,
"loss": 1.5232,
"step": 37500
},
{
"epoch": 0.78,
"learning_rate": 3.695744695456387e-05,
"loss": 1.5297,
"step": 38000
},
{
"epoch": 0.79,
"learning_rate": 3.6785834414492336e-05,
"loss": 1.5412,
"step": 38500
},
{
"epoch": 0.8,
"learning_rate": 3.661422187442081e-05,
"loss": 1.5383,
"step": 39000
},
{
"epoch": 0.81,
"learning_rate": 3.644260933434928e-05,
"loss": 1.5126,
"step": 39500
},
{
"epoch": 0.82,
"learning_rate": 3.627099679427775e-05,
"loss": 1.4971,
"step": 40000
},
{
"epoch": 0.83,
"learning_rate": 3.609938425420623e-05,
"loss": 1.4952,
"step": 40500
},
{
"epoch": 0.84,
"learning_rate": 3.59277717141347e-05,
"loss": 1.5381,
"step": 41000
},
{
"epoch": 0.85,
"learning_rate": 3.575615917406317e-05,
"loss": 1.5049,
"step": 41500
},
{
"epoch": 0.86,
"learning_rate": 3.558454663399164e-05,
"loss": 1.499,
"step": 42000
},
{
"epoch": 0.88,
"learning_rate": 3.541293409392011e-05,
"loss": 1.5128,
"step": 42500
},
{
"epoch": 0.89,
"learning_rate": 3.524132155384858e-05,
"loss": 1.4833,
"step": 43000
},
{
"epoch": 0.9,
"learning_rate": 3.506970901377706e-05,
"loss": 1.5054,
"step": 43500
},
{
"epoch": 0.91,
"learning_rate": 3.489809647370553e-05,
"loss": 1.4875,
"step": 44000
},
{
"epoch": 0.92,
"learning_rate": 3.4726483933634e-05,
"loss": 1.4746,
"step": 44500
},
{
"epoch": 0.93,
"learning_rate": 3.455487139356247e-05,
"loss": 1.5014,
"step": 45000
},
{
"epoch": 0.94,
"learning_rate": 3.438325885349095e-05,
"loss": 1.4916,
"step": 45500
},
{
"epoch": 0.95,
"learning_rate": 3.421164631341942e-05,
"loss": 1.5108,
"step": 46000
},
{
"epoch": 0.96,
"learning_rate": 3.404003377334789e-05,
"loss": 1.4931,
"step": 46500
},
{
"epoch": 0.97,
"learning_rate": 3.386842123327636e-05,
"loss": 1.4794,
"step": 47000
},
{
"epoch": 0.98,
"learning_rate": 3.3696808693204826e-05,
"loss": 1.482,
"step": 47500
},
{
"epoch": 0.99,
"learning_rate": 3.35251961531333e-05,
"loss": 1.454,
"step": 48000
},
{
"epoch": 1.0,
"learning_rate": 3.335358361306177e-05,
"loss": 1.4703,
"step": 48500
},
{
"epoch": 1.01,
"learning_rate": 3.318197107299025e-05,
"loss": 1.3302,
"step": 49000
},
{
"epoch": 1.02,
"learning_rate": 3.301035853291872e-05,
"loss": 1.3441,
"step": 49500
},
{
"epoch": 1.03,
"learning_rate": 3.283874599284719e-05,
"loss": 1.311,
"step": 50000
},
{
"epoch": 1.04,
"learning_rate": 3.266713345277566e-05,
"loss": 1.3116,
"step": 50500
},
{
"epoch": 1.05,
"learning_rate": 3.249552091270414e-05,
"loss": 1.3205,
"step": 51000
},
{
"epoch": 1.06,
"learning_rate": 3.232390837263261e-05,
"loss": 1.3269,
"step": 51500
},
{
"epoch": 1.07,
"learning_rate": 3.215229583256108e-05,
"loss": 1.3125,
"step": 52000
},
{
"epoch": 1.08,
"learning_rate": 3.198068329248955e-05,
"loss": 1.3155,
"step": 52500
},
{
"epoch": 1.09,
"learning_rate": 3.180907075241802e-05,
"loss": 1.2983,
"step": 53000
},
{
"epoch": 1.1,
"learning_rate": 3.163745821234649e-05,
"loss": 1.3261,
"step": 53500
},
{
"epoch": 1.11,
"learning_rate": 3.146584567227497e-05,
"loss": 1.3233,
"step": 54000
},
{
"epoch": 1.12,
"learning_rate": 3.129423313220344e-05,
"loss": 1.3272,
"step": 54500
},
{
"epoch": 1.13,
"learning_rate": 3.1122620592131915e-05,
"loss": 1.3353,
"step": 55000
},
{
"epoch": 1.14,
"learning_rate": 3.0951008052060384e-05,
"loss": 1.3086,
"step": 55500
},
{
"epoch": 1.15,
"learning_rate": 3.077939551198885e-05,
"loss": 1.3211,
"step": 56000
},
{
"epoch": 1.16,
"learning_rate": 3.060778297191732e-05,
"loss": 1.2961,
"step": 56500
},
{
"epoch": 1.17,
"learning_rate": 3.0436170431845795e-05,
"loss": 1.3105,
"step": 57000
},
{
"epoch": 1.18,
"learning_rate": 3.026455789177427e-05,
"loss": 1.3043,
"step": 57500
},
{
"epoch": 1.19,
"learning_rate": 3.0092945351702738e-05,
"loss": 1.2852,
"step": 58000
},
{
"epoch": 1.2,
"learning_rate": 2.9921332811631214e-05,
"loss": 1.3173,
"step": 58500
},
{
"epoch": 1.22,
"learning_rate": 2.9749720271559683e-05,
"loss": 1.2992,
"step": 59000
},
{
"epoch": 1.23,
"learning_rate": 2.957810773148816e-05,
"loss": 1.3263,
"step": 59500
},
{
"epoch": 1.24,
"learning_rate": 2.940649519141663e-05,
"loss": 1.3091,
"step": 60000
},
{
"epoch": 1.25,
"learning_rate": 2.92348826513451e-05,
"loss": 1.2984,
"step": 60500
},
{
"epoch": 1.26,
"learning_rate": 2.906327011127357e-05,
"loss": 1.3037,
"step": 61000
},
{
"epoch": 1.27,
"learning_rate": 2.8891657571202047e-05,
"loss": 1.3045,
"step": 61500
},
{
"epoch": 1.28,
"learning_rate": 2.8720045031130517e-05,
"loss": 1.2961,
"step": 62000
},
{
"epoch": 1.29,
"learning_rate": 2.854843249105899e-05,
"loss": 1.306,
"step": 62500
},
{
"epoch": 1.3,
"learning_rate": 2.837681995098746e-05,
"loss": 1.3032,
"step": 63000
},
{
"epoch": 1.31,
"learning_rate": 2.8205207410915928e-05,
"loss": 1.2992,
"step": 63500
},
{
"epoch": 1.32,
"learning_rate": 2.8033594870844404e-05,
"loss": 1.3061,
"step": 64000
},
{
"epoch": 1.33,
"learning_rate": 2.7861982330772874e-05,
"loss": 1.3085,
"step": 64500
},
{
"epoch": 1.34,
"learning_rate": 2.7690369790701347e-05,
"loss": 1.312,
"step": 65000
},
{
"epoch": 1.35,
"learning_rate": 2.7518757250629816e-05,
"loss": 1.2912,
"step": 65500
},
{
"epoch": 1.36,
"learning_rate": 2.7347144710558292e-05,
"loss": 1.3026,
"step": 66000
},
{
"epoch": 1.37,
"learning_rate": 2.717553217048676e-05,
"loss": 1.2848,
"step": 66500
},
{
"epoch": 1.38,
"learning_rate": 2.7003919630415238e-05,
"loss": 1.3018,
"step": 67000
},
{
"epoch": 1.39,
"learning_rate": 2.6832307090343707e-05,
"loss": 1.2972,
"step": 67500
},
{
"epoch": 1.4,
"learning_rate": 2.666069455027218e-05,
"loss": 1.2736,
"step": 68000
},
{
"epoch": 1.41,
"learning_rate": 2.648908201020065e-05,
"loss": 1.2721,
"step": 68500
},
{
"epoch": 1.42,
"learning_rate": 2.6317469470129126e-05,
"loss": 1.2921,
"step": 69000
},
{
"epoch": 1.43,
"learning_rate": 2.6145856930057595e-05,
"loss": 1.287,
"step": 69500
},
{
"epoch": 1.44,
"learning_rate": 2.5974244389986068e-05,
"loss": 1.286,
"step": 70000
},
{
"epoch": 1.45,
"learning_rate": 2.5802631849914537e-05,
"loss": 1.2839,
"step": 70500
},
{
"epoch": 1.46,
"learning_rate": 2.5631019309843007e-05,
"loss": 1.262,
"step": 71000
},
{
"epoch": 1.47,
"learning_rate": 2.5459406769771483e-05,
"loss": 1.2712,
"step": 71500
},
{
"epoch": 1.48,
"learning_rate": 2.5287794229699952e-05,
"loss": 1.2876,
"step": 72000
},
{
"epoch": 1.49,
"learning_rate": 2.5116181689628425e-05,
"loss": 1.2771,
"step": 72500
},
{
"epoch": 1.5,
"learning_rate": 2.4944569149556898e-05,
"loss": 1.2536,
"step": 73000
},
{
"epoch": 1.51,
"learning_rate": 2.477295660948537e-05,
"loss": 1.2693,
"step": 73500
},
{
"epoch": 1.52,
"learning_rate": 2.4601344069413843e-05,
"loss": 1.2467,
"step": 74000
},
{
"epoch": 1.53,
"learning_rate": 2.4429731529342316e-05,
"loss": 1.2886,
"step": 74500
},
{
"epoch": 1.54,
"learning_rate": 2.4258118989270785e-05,
"loss": 1.2981,
"step": 75000
},
{
"epoch": 1.55,
"learning_rate": 2.4086506449199255e-05,
"loss": 1.2729,
"step": 75500
},
{
"epoch": 1.57,
"learning_rate": 2.3914893909127728e-05,
"loss": 1.2682,
"step": 76000
},
{
"epoch": 1.58,
"learning_rate": 2.37432813690562e-05,
"loss": 1.2577,
"step": 76500
},
{
"epoch": 1.59,
"learning_rate": 2.3571668828984673e-05,
"loss": 1.2735,
"step": 77000
},
{
"epoch": 1.6,
"learning_rate": 2.3400056288913143e-05,
"loss": 1.2663,
"step": 77500
},
{
"epoch": 1.61,
"learning_rate": 2.3228443748841615e-05,
"loss": 1.272,
"step": 78000
},
{
"epoch": 1.62,
"learning_rate": 2.3056831208770088e-05,
"loss": 1.2679,
"step": 78500
},
{
"epoch": 1.63,
"learning_rate": 2.288521866869856e-05,
"loss": 1.2369,
"step": 79000
},
{
"epoch": 1.64,
"learning_rate": 2.2713606128627034e-05,
"loss": 1.2633,
"step": 79500
},
{
"epoch": 1.65,
"learning_rate": 2.2541993588555503e-05,
"loss": 1.2504,
"step": 80000
},
{
"epoch": 1.66,
"learning_rate": 2.2370381048483976e-05,
"loss": 1.2702,
"step": 80500
},
{
"epoch": 1.67,
"learning_rate": 2.219876850841245e-05,
"loss": 1.2467,
"step": 81000
},
{
"epoch": 1.68,
"learning_rate": 2.202715596834092e-05,
"loss": 1.2392,
"step": 81500
},
{
"epoch": 1.69,
"learning_rate": 2.1855543428269394e-05,
"loss": 1.257,
"step": 82000
},
{
"epoch": 1.7,
"learning_rate": 2.1683930888197864e-05,
"loss": 1.2431,
"step": 82500
},
{
"epoch": 1.71,
"learning_rate": 2.1512318348126333e-05,
"loss": 1.2486,
"step": 83000
},
{
"epoch": 1.72,
"learning_rate": 2.1340705808054806e-05,
"loss": 1.212,
"step": 83500
},
{
"epoch": 1.73,
"learning_rate": 2.116909326798328e-05,
"loss": 1.2743,
"step": 84000
},
{
"epoch": 1.74,
"learning_rate": 2.099748072791175e-05,
"loss": 1.2249,
"step": 84500
},
{
"epoch": 1.75,
"learning_rate": 2.082586818784022e-05,
"loss": 1.2234,
"step": 85000
},
{
"epoch": 1.76,
"learning_rate": 2.0654255647768694e-05,
"loss": 1.2778,
"step": 85500
},
{
"epoch": 1.77,
"learning_rate": 2.0482643107697167e-05,
"loss": 1.2382,
"step": 86000
},
{
"epoch": 1.78,
"learning_rate": 2.031103056762564e-05,
"loss": 1.2764,
"step": 86500
},
{
"epoch": 1.79,
"learning_rate": 2.0139418027554112e-05,
"loss": 1.2352,
"step": 87000
},
{
"epoch": 1.8,
"learning_rate": 1.996780548748258e-05,
"loss": 1.2464,
"step": 87500
},
{
"epoch": 1.81,
"learning_rate": 1.9796192947411054e-05,
"loss": 1.2257,
"step": 88000
},
{
"epoch": 1.82,
"learning_rate": 1.9624580407339527e-05,
"loss": 1.2331,
"step": 88500
},
{
"epoch": 1.83,
"learning_rate": 1.9452967867268e-05,
"loss": 1.2404,
"step": 89000
},
{
"epoch": 1.84,
"learning_rate": 1.928135532719647e-05,
"loss": 1.2406,
"step": 89500
},
{
"epoch": 1.85,
"learning_rate": 1.9109742787124942e-05,
"loss": 1.2287,
"step": 90000
},
{
"epoch": 1.86,
"learning_rate": 1.893813024705341e-05,
"loss": 1.2386,
"step": 90500
},
{
"epoch": 1.87,
"learning_rate": 1.8766517706981884e-05,
"loss": 1.2105,
"step": 91000
},
{
"epoch": 1.88,
"learning_rate": 1.8594905166910357e-05,
"loss": 1.2354,
"step": 91500
},
{
"epoch": 1.89,
"learning_rate": 1.842329262683883e-05,
"loss": 1.2233,
"step": 92000
},
{
"epoch": 1.9,
"learning_rate": 1.82516800867673e-05,
"loss": 1.2254,
"step": 92500
},
{
"epoch": 1.92,
"learning_rate": 1.8080067546695772e-05,
"loss": 1.218,
"step": 93000
},
{
"epoch": 1.93,
"learning_rate": 1.7908455006624245e-05,
"loss": 1.2331,
"step": 93500
},
{
"epoch": 1.94,
"learning_rate": 1.7736842466552718e-05,
"loss": 1.2435,
"step": 94000
},
{
"epoch": 1.95,
"learning_rate": 1.7565229926481187e-05,
"loss": 1.2365,
"step": 94500
},
{
"epoch": 1.96,
"learning_rate": 1.739361738640966e-05,
"loss": 1.2457,
"step": 95000
},
{
"epoch": 1.97,
"learning_rate": 1.7222004846338133e-05,
"loss": 1.2267,
"step": 95500
},
{
"epoch": 1.98,
"learning_rate": 1.7050392306266605e-05,
"loss": 1.2258,
"step": 96000
},
{
"epoch": 1.99,
"learning_rate": 1.6878779766195078e-05,
"loss": 1.2074,
"step": 96500
},
{
"epoch": 2.0,
"learning_rate": 1.6707167226123548e-05,
"loss": 1.2509,
"step": 97000
},
{
"epoch": 2.01,
"learning_rate": 1.653555468605202e-05,
"loss": 1.1091,
"step": 97500
},
{
"epoch": 2.02,
"learning_rate": 1.6363942145980493e-05,
"loss": 1.0773,
"step": 98000
},
{
"epoch": 2.03,
"learning_rate": 1.6192329605908963e-05,
"loss": 1.0776,
"step": 98500
},
{
"epoch": 2.04,
"learning_rate": 1.6020717065837435e-05,
"loss": 1.0561,
"step": 99000
},
{
"epoch": 2.05,
"learning_rate": 1.5849104525765908e-05,
"loss": 1.0627,
"step": 99500
},
{
"epoch": 2.06,
"learning_rate": 1.5677491985694378e-05,
"loss": 1.0586,
"step": 100000
},
{
"epoch": 2.07,
"learning_rate": 1.550587944562285e-05,
"loss": 1.0628,
"step": 100500
},
{
"epoch": 2.08,
"learning_rate": 1.5334266905551323e-05,
"loss": 1.0768,
"step": 101000
},
{
"epoch": 2.09,
"learning_rate": 1.5162654365479794e-05,
"loss": 1.0788,
"step": 101500
},
{
"epoch": 2.1,
"learning_rate": 1.4991041825408267e-05,
"loss": 1.0629,
"step": 102000
},
{
"epoch": 2.11,
"learning_rate": 1.4819429285336738e-05,
"loss": 1.0641,
"step": 102500
},
{
"epoch": 2.12,
"learning_rate": 1.4647816745265211e-05,
"loss": 1.0654,
"step": 103000
},
{
"epoch": 2.13,
"learning_rate": 1.4476204205193684e-05,
"loss": 1.0584,
"step": 103500
},
{
"epoch": 2.14,
"learning_rate": 1.4304591665122155e-05,
"loss": 1.0606,
"step": 104000
},
{
"epoch": 2.15,
"learning_rate": 1.4132979125050628e-05,
"loss": 1.0814,
"step": 104500
},
{
"epoch": 2.16,
"learning_rate": 1.3961366584979099e-05,
"loss": 1.0753,
"step": 105000
},
{
"epoch": 2.17,
"learning_rate": 1.3789754044907571e-05,
"loss": 1.0789,
"step": 105500
},
{
"epoch": 2.18,
"learning_rate": 1.3618141504836043e-05,
"loss": 1.0673,
"step": 106000
},
{
"epoch": 2.19,
"learning_rate": 1.3446528964764512e-05,
"loss": 1.0591,
"step": 106500
},
{
"epoch": 2.2,
"learning_rate": 1.3274916424692985e-05,
"loss": 1.0724,
"step": 107000
},
{
"epoch": 2.21,
"learning_rate": 1.3103303884621458e-05,
"loss": 1.0467,
"step": 107500
},
{
"epoch": 2.22,
"learning_rate": 1.2931691344549929e-05,
"loss": 1.048,
"step": 108000
},
{
"epoch": 2.23,
"learning_rate": 1.2760078804478401e-05,
"loss": 1.0669,
"step": 108500
},
{
"epoch": 2.24,
"learning_rate": 1.2588466264406873e-05,
"loss": 1.0526,
"step": 109000
},
{
"epoch": 2.25,
"learning_rate": 1.2416853724335345e-05,
"loss": 1.0604,
"step": 109500
},
{
"epoch": 2.27,
"learning_rate": 1.2245241184263816e-05,
"loss": 1.0689,
"step": 110000
},
{
"epoch": 2.28,
"learning_rate": 1.207362864419229e-05,
"loss": 1.0626,
"step": 110500
},
{
"epoch": 2.29,
"learning_rate": 1.1902016104120762e-05,
"loss": 1.0637,
"step": 111000
},
{
"epoch": 2.3,
"learning_rate": 1.1730403564049233e-05,
"loss": 1.0511,
"step": 111500
},
{
"epoch": 2.31,
"learning_rate": 1.1558791023977706e-05,
"loss": 1.0449,
"step": 112000
},
{
"epoch": 2.32,
"learning_rate": 1.1387178483906175e-05,
"loss": 1.0675,
"step": 112500
},
{
"epoch": 2.33,
"learning_rate": 1.1215565943834648e-05,
"loss": 1.0524,
"step": 113000
},
{
"epoch": 2.34,
"learning_rate": 1.1043953403763121e-05,
"loss": 1.0554,
"step": 113500
},
{
"epoch": 2.35,
"learning_rate": 1.0872340863691592e-05,
"loss": 1.0476,
"step": 114000
},
{
"epoch": 2.36,
"learning_rate": 1.0700728323620065e-05,
"loss": 1.0516,
"step": 114500
},
{
"epoch": 2.37,
"learning_rate": 1.0529115783548536e-05,
"loss": 1.0472,
"step": 115000
},
{
"epoch": 2.38,
"learning_rate": 1.0357503243477009e-05,
"loss": 1.0581,
"step": 115500
},
{
"epoch": 2.39,
"learning_rate": 1.018589070340548e-05,
"loss": 1.0515,
"step": 116000
},
{
"epoch": 2.4,
"learning_rate": 1.0014278163333951e-05,
"loss": 1.0549,
"step": 116500
},
{
"epoch": 2.41,
"learning_rate": 9.842665623262424e-06,
"loss": 1.0411,
"step": 117000
},
{
"epoch": 2.42,
"learning_rate": 9.671053083190895e-06,
"loss": 1.0534,
"step": 117500
},
{
"epoch": 2.43,
"learning_rate": 9.499440543119368e-06,
"loss": 1.0509,
"step": 118000
},
{
"epoch": 2.44,
"learning_rate": 9.327828003047839e-06,
"loss": 1.0508,
"step": 118500
},
{
"epoch": 2.45,
"learning_rate": 9.156215462976311e-06,
"loss": 1.0392,
"step": 119000
},
{
"epoch": 2.46,
"learning_rate": 8.984602922904784e-06,
"loss": 1.0254,
"step": 119500
},
{
"epoch": 2.47,
"learning_rate": 8.812990382833254e-06,
"loss": 1.0562,
"step": 120000
},
{
"epoch": 2.48,
"learning_rate": 8.641377842761726e-06,
"loss": 1.0678,
"step": 120500
},
{
"epoch": 2.49,
"learning_rate": 8.469765302690198e-06,
"loss": 1.0201,
"step": 121000
},
{
"epoch": 2.5,
"learning_rate": 8.29815276261867e-06,
"loss": 1.0377,
"step": 121500
},
{
"epoch": 2.51,
"learning_rate": 8.126540222547143e-06,
"loss": 1.056,
"step": 122000
},
{
"epoch": 2.52,
"learning_rate": 7.954927682475614e-06,
"loss": 1.0411,
"step": 122500
},
{
"epoch": 2.53,
"learning_rate": 7.783315142404087e-06,
"loss": 1.0324,
"step": 123000
},
{
"epoch": 2.54,
"learning_rate": 7.611702602332559e-06,
"loss": 1.0235,
"step": 123500
},
{
"epoch": 2.55,
"learning_rate": 7.440090062261029e-06,
"loss": 1.0511,
"step": 124000
},
{
"epoch": 2.56,
"learning_rate": 7.268477522189501e-06,
"loss": 1.0382,
"step": 124500
},
{
"epoch": 2.57,
"learning_rate": 7.096864982117973e-06,
"loss": 1.0166,
"step": 125000
},
{
"epoch": 2.58,
"learning_rate": 6.925252442046445e-06,
"loss": 1.0437,
"step": 125500
},
{
"epoch": 2.59,
"learning_rate": 6.753639901974918e-06,
"loss": 1.0575,
"step": 126000
},
{
"epoch": 2.61,
"learning_rate": 6.58202736190339e-06,
"loss": 1.0355,
"step": 126500
},
{
"epoch": 2.62,
"learning_rate": 6.410414821831862e-06,
"loss": 1.0414,
"step": 127000
},
{
"epoch": 2.63,
"learning_rate": 6.238802281760333e-06,
"loss": 1.0213,
"step": 127500
},
{
"epoch": 2.64,
"learning_rate": 6.067189741688805e-06,
"loss": 1.0432,
"step": 128000
},
{
"epoch": 2.65,
"learning_rate": 5.895577201617277e-06,
"loss": 1.0309,
"step": 128500
},
{
"epoch": 2.66,
"learning_rate": 5.723964661545749e-06,
"loss": 1.0199,
"step": 129000
},
{
"epoch": 2.67,
"learning_rate": 5.5523521214742206e-06,
"loss": 1.0388,
"step": 129500
},
{
"epoch": 2.68,
"learning_rate": 5.3807395814026925e-06,
"loss": 1.0346,
"step": 130000
},
{
"epoch": 2.69,
"learning_rate": 5.2091270413311644e-06,
"loss": 1.0467,
"step": 130500
},
{
"epoch": 2.7,
"learning_rate": 5.037514501259636e-06,
"loss": 1.035,
"step": 131000
},
{
"epoch": 2.71,
"learning_rate": 4.865901961188108e-06,
"loss": 1.0314,
"step": 131500
},
{
"epoch": 2.72,
"learning_rate": 4.69428942111658e-06,
"loss": 1.0416,
"step": 132000
},
{
"epoch": 2.73,
"learning_rate": 4.522676881045052e-06,
"loss": 1.0085,
"step": 132500
},
{
"epoch": 2.74,
"learning_rate": 4.351064340973523e-06,
"loss": 1.043,
"step": 133000
},
{
"epoch": 2.75,
"learning_rate": 4.179451800901995e-06,
"loss": 1.0191,
"step": 133500
},
{
"epoch": 2.76,
"learning_rate": 4.007839260830468e-06,
"loss": 1.0402,
"step": 134000
},
{
"epoch": 2.77,
"learning_rate": 3.83622672075894e-06,
"loss": 1.0454,
"step": 134500
},
{
"epoch": 2.78,
"learning_rate": 3.664614180687411e-06,
"loss": 1.019,
"step": 135000
},
{
"epoch": 2.79,
"learning_rate": 3.493001640615883e-06,
"loss": 1.0326,
"step": 135500
},
{
"epoch": 2.8,
"learning_rate": 3.3213891005443554e-06,
"loss": 1.033,
"step": 136000
},
{
"epoch": 2.81,
"learning_rate": 3.1497765604728273e-06,
"loss": 1.0224,
"step": 136500
},
{
"epoch": 2.82,
"learning_rate": 2.978164020401299e-06,
"loss": 1.0207,
"step": 137000
},
{
"epoch": 2.83,
"learning_rate": 2.806551480329771e-06,
"loss": 1.0219,
"step": 137500
},
{
"epoch": 2.84,
"learning_rate": 2.6349389402582427e-06,
"loss": 1.0244,
"step": 138000
},
{
"epoch": 2.85,
"learning_rate": 2.4633264001867147e-06,
"loss": 1.0157,
"step": 138500
},
{
"epoch": 2.86,
"learning_rate": 2.2917138601151866e-06,
"loss": 1.008,
"step": 139000
},
{
"epoch": 2.87,
"learning_rate": 2.120101320043658e-06,
"loss": 1.0381,
"step": 139500
},
{
"epoch": 2.88,
"learning_rate": 1.9484887799721305e-06,
"loss": 1.0198,
"step": 140000
},
{
"epoch": 2.89,
"learning_rate": 1.776876239900602e-06,
"loss": 1.0076,
"step": 140500
},
{
"epoch": 2.9,
"learning_rate": 1.6052636998290742e-06,
"loss": 1.0175,
"step": 141000
},
{
"epoch": 2.91,
"learning_rate": 1.433651159757546e-06,
"loss": 1.0199,
"step": 141500
},
{
"epoch": 2.92,
"learning_rate": 1.2620386196860177e-06,
"loss": 1.026,
"step": 142000
},
{
"epoch": 2.93,
"learning_rate": 1.0904260796144896e-06,
"loss": 1.0359,
"step": 142500
},
{
"epoch": 2.94,
"learning_rate": 9.188135395429614e-07,
"loss": 1.0409,
"step": 143000
},
{
"epoch": 2.96,
"learning_rate": 7.472009994714334e-07,
"loss": 1.0028,
"step": 143500
},
{
"epoch": 2.97,
"learning_rate": 5.755884593999052e-07,
"loss": 1.0147,
"step": 144000
},
{
"epoch": 2.98,
"learning_rate": 4.0397591932837716e-07,
"loss": 1.0111,
"step": 144500
},
{
"epoch": 2.99,
"learning_rate": 2.3236337925684908e-07,
"loss": 1.0241,
"step": 145000
},
{
"epoch": 3.0,
"learning_rate": 6.075083918532095e-08,
"loss": 1.0178,
"step": 145500
},
{
"epoch": 3.0,
"step": 145677,
"total_flos": 2.385792939319296e+17,
"train_loss": 1.349638056625609,
"train_runtime": 40215.8311,
"train_samples_per_second": 36.223,
"train_steps_per_second": 3.622
}
],
"max_steps": 145677,
"num_train_epochs": 3,
"total_flos": 2.385792939319296e+17,
"trial_name": null,
"trial_params": null
}