Microscopic-Mistral-78k-steps / trainer_state.json
DrNicefellow's picture
Upload 12 files
5aa3164 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.5623226520866496,
"eval_steps": 100000000,
"global_step": 78000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 1.0000000000000001e-07,
"loss": 10.7586,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 5e-05,
"loss": 6.5292,
"step": 500
},
{
"epoch": 0.01,
"learning_rate": 4.99998214600225e-05,
"loss": 4.7262,
"step": 1000
},
{
"epoch": 0.01,
"learning_rate": 4.9999285842640105e-05,
"loss": 4.1997,
"step": 1500
},
{
"epoch": 0.01,
"learning_rate": 4.999839315550315e-05,
"loss": 3.8892,
"step": 2000
},
{
"epoch": 0.02,
"learning_rate": 4.9997143411362066e-05,
"loss": 3.6618,
"step": 2500
},
{
"epoch": 0.02,
"learning_rate": 4.9995536628067196e-05,
"loss": 3.5132,
"step": 3000
},
{
"epoch": 0.03,
"learning_rate": 4.9993572828568536e-05,
"loss": 3.4092,
"step": 3500
},
{
"epoch": 0.03,
"learning_rate": 4.999125204091544e-05,
"loss": 3.3295,
"step": 4000
},
{
"epoch": 0.03,
"learning_rate": 4.998857429825615e-05,
"loss": 3.2596,
"step": 4500
},
{
"epoch": 0.04,
"learning_rate": 4.9985539638837424e-05,
"loss": 3.2068,
"step": 5000
},
{
"epoch": 0.04,
"learning_rate": 4.998214810600389e-05,
"loss": 3.163,
"step": 5500
},
{
"epoch": 0.04,
"learning_rate": 4.997839974819749e-05,
"loss": 3.1256,
"step": 6000
},
{
"epoch": 0.05,
"learning_rate": 4.997429461895675e-05,
"loss": 3.0844,
"step": 6500
},
{
"epoch": 0.05,
"learning_rate": 4.996983277691606e-05,
"loss": 3.0541,
"step": 7000
},
{
"epoch": 0.05,
"learning_rate": 4.996501428580478e-05,
"loss": 3.0293,
"step": 7500
},
{
"epoch": 0.06,
"learning_rate": 4.995983921444638e-05,
"loss": 2.999,
"step": 8000
},
{
"epoch": 0.06,
"learning_rate": 4.9954307636757434e-05,
"loss": 2.9744,
"step": 8500
},
{
"epoch": 0.06,
"learning_rate": 4.994841963174656e-05,
"loss": 2.9525,
"step": 9000
},
{
"epoch": 0.07,
"learning_rate": 4.9942175283513294e-05,
"loss": 2.9321,
"step": 9500
},
{
"epoch": 0.07,
"learning_rate": 4.993557468124691e-05,
"loss": 2.9237,
"step": 10000
},
{
"epoch": 0.08,
"learning_rate": 4.9928617919225106e-05,
"loss": 2.9021,
"step": 10500
},
{
"epoch": 0.08,
"learning_rate": 4.99213050968127e-05,
"loss": 2.8766,
"step": 11000
},
{
"epoch": 0.08,
"learning_rate": 4.991363631846019e-05,
"loss": 2.8691,
"step": 11500
},
{
"epoch": 0.09,
"learning_rate": 4.990561169370224e-05,
"loss": 2.8531,
"step": 12000
},
{
"epoch": 0.09,
"learning_rate": 4.989723133715618e-05,
"loss": 2.8366,
"step": 12500
},
{
"epoch": 0.09,
"learning_rate": 4.988849536852028e-05,
"loss": 2.8265,
"step": 13000
},
{
"epoch": 0.1,
"learning_rate": 4.987940391257212e-05,
"loss": 2.8099,
"step": 13500
},
{
"epoch": 0.1,
"learning_rate": 4.986995709916677e-05,
"loss": 2.8009,
"step": 14000
},
{
"epoch": 0.1,
"learning_rate": 4.9860155063234944e-05,
"loss": 2.7881,
"step": 14500
},
{
"epoch": 0.11,
"learning_rate": 4.9849997944781055e-05,
"loss": 2.7794,
"step": 15000
},
{
"epoch": 0.11,
"learning_rate": 4.9839485888881235e-05,
"loss": 2.7674,
"step": 15500
},
{
"epoch": 0.12,
"learning_rate": 4.982861904568127e-05,
"loss": 2.7559,
"step": 16000
},
{
"epoch": 0.12,
"learning_rate": 4.981739757039443e-05,
"loss": 2.7455,
"step": 16500
},
{
"epoch": 0.12,
"learning_rate": 4.9805821623299285e-05,
"loss": 2.7351,
"step": 17000
},
{
"epoch": 0.13,
"learning_rate": 4.979389136973737e-05,
"loss": 2.7317,
"step": 17500
},
{
"epoch": 0.13,
"learning_rate": 4.978160698011085e-05,
"loss": 2.7192,
"step": 18000
},
{
"epoch": 0.13,
"learning_rate": 4.976896862988012e-05,
"loss": 2.709,
"step": 18500
},
{
"epoch": 0.14,
"learning_rate": 4.975597649956122e-05,
"loss": 2.7043,
"step": 19000
},
{
"epoch": 0.14,
"learning_rate": 4.974263077472334e-05,
"loss": 2.6954,
"step": 19500
},
{
"epoch": 0.14,
"learning_rate": 4.97289316459861e-05,
"loss": 2.6905,
"step": 20000
},
{
"epoch": 0.15,
"learning_rate": 4.9714879309016874e-05,
"loss": 2.6782,
"step": 20500
},
{
"epoch": 0.15,
"learning_rate": 4.970047396452798e-05,
"loss": 2.6751,
"step": 21000
},
{
"epoch": 0.15,
"learning_rate": 4.968571581827381e-05,
"loss": 2.6659,
"step": 21500
},
{
"epoch": 0.16,
"learning_rate": 4.9670605081047886e-05,
"loss": 2.6594,
"step": 22000
},
{
"epoch": 0.16,
"learning_rate": 4.9655141968679866e-05,
"loss": 2.6553,
"step": 22500
},
{
"epoch": 0.17,
"learning_rate": 4.963932670203245e-05,
"loss": 2.6499,
"step": 23000
},
{
"epoch": 0.17,
"learning_rate": 4.962315950699822e-05,
"loss": 2.6391,
"step": 23500
},
{
"epoch": 0.17,
"learning_rate": 4.960664061449643e-05,
"loss": 2.6313,
"step": 24000
},
{
"epoch": 0.18,
"learning_rate": 4.958977026046971e-05,
"loss": 2.6318,
"step": 24500
},
{
"epoch": 0.18,
"learning_rate": 4.957254868588065e-05,
"loss": 2.6275,
"step": 25000
},
{
"epoch": 0.18,
"learning_rate": 4.955497613670842e-05,
"loss": 2.6192,
"step": 25500
},
{
"epoch": 0.19,
"learning_rate": 4.953705286394522e-05,
"loss": 2.6153,
"step": 26000
},
{
"epoch": 0.19,
"learning_rate": 4.9518779123592705e-05,
"loss": 2.6102,
"step": 26500
},
{
"epoch": 0.19,
"learning_rate": 4.9500155176658345e-05,
"loss": 2.601,
"step": 27000
},
{
"epoch": 0.2,
"learning_rate": 4.9481181289151655e-05,
"loss": 2.5991,
"step": 27500
},
{
"epoch": 0.2,
"learning_rate": 4.9461857732080436e-05,
"loss": 2.5946,
"step": 28000
},
{
"epoch": 0.21,
"learning_rate": 4.9442184781446876e-05,
"loss": 2.5911,
"step": 28500
},
{
"epoch": 0.21,
"learning_rate": 4.9422162718243626e-05,
"loss": 2.5827,
"step": 29000
},
{
"epoch": 0.21,
"learning_rate": 4.94017918284498e-05,
"loss": 2.5785,
"step": 29500
},
{
"epoch": 0.22,
"learning_rate": 4.938107240302683e-05,
"loss": 2.5757,
"step": 30000
},
{
"epoch": 0.22,
"learning_rate": 4.93600047379144e-05,
"loss": 2.5705,
"step": 30500
},
{
"epoch": 0.22,
"learning_rate": 4.9338589134026124e-05,
"loss": 2.5703,
"step": 31000
},
{
"epoch": 0.23,
"learning_rate": 4.931682589724534e-05,
"loss": 2.5593,
"step": 31500
},
{
"epoch": 0.23,
"learning_rate": 4.929471533842065e-05,
"loss": 2.5647,
"step": 32000
},
{
"epoch": 0.23,
"learning_rate": 4.9272257773361574e-05,
"loss": 2.5564,
"step": 32500
},
{
"epoch": 0.24,
"learning_rate": 4.924945352283394e-05,
"loss": 2.5492,
"step": 33000
},
{
"epoch": 0.24,
"learning_rate": 4.922630291255539e-05,
"loss": 2.5509,
"step": 33500
},
{
"epoch": 0.25,
"learning_rate": 4.920280627319067e-05,
"loss": 2.5466,
"step": 34000
},
{
"epoch": 0.25,
"learning_rate": 4.917896394034696e-05,
"loss": 2.5412,
"step": 34500
},
{
"epoch": 0.25,
"learning_rate": 4.915477625456899e-05,
"loss": 2.5383,
"step": 35000
},
{
"epoch": 0.26,
"learning_rate": 4.91302435613343e-05,
"loss": 2.5364,
"step": 35500
},
{
"epoch": 0.26,
"learning_rate": 4.91053662110482e-05,
"loss": 2.5292,
"step": 36000
},
{
"epoch": 0.26,
"learning_rate": 4.908014455903881e-05,
"loss": 2.5304,
"step": 36500
},
{
"epoch": 0.27,
"learning_rate": 4.905457896555199e-05,
"loss": 2.5257,
"step": 37000
},
{
"epoch": 0.27,
"learning_rate": 4.902866979574618e-05,
"loss": 2.5189,
"step": 37500
},
{
"epoch": 0.27,
"learning_rate": 4.9002417419687183e-05,
"loss": 2.522,
"step": 38000
},
{
"epoch": 0.28,
"learning_rate": 4.897582221234289e-05,
"loss": 2.5141,
"step": 38500
},
{
"epoch": 0.28,
"learning_rate": 4.8948884553577926e-05,
"loss": 2.517,
"step": 39000
},
{
"epoch": 0.28,
"learning_rate": 4.8921604828148206e-05,
"loss": 2.5077,
"step": 39500
},
{
"epoch": 0.29,
"learning_rate": 4.8893983425695445e-05,
"loss": 2.5068,
"step": 40000
},
{
"epoch": 0.29,
"learning_rate": 4.8866020740741626e-05,
"loss": 2.5048,
"step": 40500
},
{
"epoch": 0.3,
"learning_rate": 4.883771717268331e-05,
"loss": 2.4972,
"step": 41000
},
{
"epoch": 0.3,
"learning_rate": 4.8809073125785964e-05,
"loss": 2.4999,
"step": 41500
},
{
"epoch": 0.3,
"learning_rate": 4.8780089009178206e-05,
"loss": 2.4941,
"step": 42000
},
{
"epoch": 0.31,
"learning_rate": 4.8750765236845894e-05,
"loss": 2.4891,
"step": 42500
},
{
"epoch": 0.31,
"learning_rate": 4.8721102227626304e-05,
"loss": 2.4876,
"step": 43000
},
{
"epoch": 0.31,
"learning_rate": 4.869110040520206e-05,
"loss": 2.4859,
"step": 43500
},
{
"epoch": 0.32,
"learning_rate": 4.8660760198095145e-05,
"loss": 2.4833,
"step": 44000
},
{
"epoch": 0.32,
"learning_rate": 4.8630082039660744e-05,
"loss": 2.4825,
"step": 44500
},
{
"epoch": 0.32,
"learning_rate": 4.859906636808108e-05,
"loss": 2.4801,
"step": 45000
},
{
"epoch": 0.33,
"learning_rate": 4.8567713626359134e-05,
"loss": 2.4763,
"step": 45500
},
{
"epoch": 0.33,
"learning_rate": 4.853602426231234e-05,
"loss": 2.4742,
"step": 46000
},
{
"epoch": 0.34,
"learning_rate": 4.850399872856615e-05,
"loss": 2.4702,
"step": 46500
},
{
"epoch": 0.34,
"learning_rate": 4.847163748254762e-05,
"loss": 2.4679,
"step": 47000
},
{
"epoch": 0.34,
"learning_rate": 4.843894098647884e-05,
"loss": 2.4676,
"step": 47500
},
{
"epoch": 0.35,
"learning_rate": 4.840590970737035e-05,
"loss": 2.4679,
"step": 48000
},
{
"epoch": 0.35,
"learning_rate": 4.8372544117014436e-05,
"loss": 2.4644,
"step": 48500
},
{
"epoch": 0.35,
"learning_rate": 4.8338844691978456e-05,
"loss": 2.4622,
"step": 49000
},
{
"epoch": 0.36,
"learning_rate": 4.830481191359797e-05,
"loss": 2.4605,
"step": 49500
},
{
"epoch": 0.36,
"learning_rate": 4.82704462679699e-05,
"loss": 2.4552,
"step": 50000
},
{
"epoch": 0.36,
"learning_rate": 4.823574824594558e-05,
"loss": 2.4501,
"step": 50500
},
{
"epoch": 0.37,
"learning_rate": 4.820071834312372e-05,
"loss": 2.4563,
"step": 51000
},
{
"epoch": 0.37,
"learning_rate": 4.816535705984337e-05,
"loss": 2.4467,
"step": 51500
},
{
"epoch": 0.37,
"learning_rate": 4.812966490117676e-05,
"loss": 2.4434,
"step": 52000
},
{
"epoch": 0.38,
"learning_rate": 4.8093642376922057e-05,
"loss": 2.4422,
"step": 52500
},
{
"epoch": 0.38,
"learning_rate": 4.8057290001596106e-05,
"loss": 2.4462,
"step": 53000
},
{
"epoch": 0.39,
"learning_rate": 4.802060829442711e-05,
"loss": 2.4424,
"step": 53500
},
{
"epoch": 0.39,
"learning_rate": 4.798359777934714e-05,
"loss": 2.4373,
"step": 54000
},
{
"epoch": 0.39,
"learning_rate": 4.7946258984984735e-05,
"loss": 2.4381,
"step": 54500
},
{
"epoch": 0.4,
"learning_rate": 4.7908592444657287e-05,
"loss": 2.4334,
"step": 55000
},
{
"epoch": 0.4,
"learning_rate": 4.787059869636346e-05,
"loss": 2.4328,
"step": 55500
},
{
"epoch": 0.4,
"learning_rate": 4.783227828277549e-05,
"loss": 2.4284,
"step": 56000
},
{
"epoch": 0.41,
"learning_rate": 4.779363175123144e-05,
"loss": 2.4278,
"step": 56500
},
{
"epoch": 0.41,
"learning_rate": 4.7754659653727384e-05,
"loss": 2.4231,
"step": 57000
},
{
"epoch": 0.41,
"learning_rate": 4.771536254690951e-05,
"loss": 2.4212,
"step": 57500
},
{
"epoch": 0.42,
"learning_rate": 4.767574099206619e-05,
"loss": 2.4247,
"step": 58000
},
{
"epoch": 0.42,
"learning_rate": 4.763579555511993e-05,
"loss": 2.4209,
"step": 58500
},
{
"epoch": 0.43,
"learning_rate": 4.759552680661933e-05,
"loss": 2.4224,
"step": 59000
},
{
"epoch": 0.43,
"learning_rate": 4.755493532173092e-05,
"loss": 2.4204,
"step": 59500
},
{
"epoch": 0.43,
"learning_rate": 4.75140216802309e-05,
"loss": 2.4201,
"step": 60000
},
{
"epoch": 0.44,
"learning_rate": 4.7472786466496944e-05,
"loss": 2.4127,
"step": 60500
},
{
"epoch": 0.44,
"learning_rate": 4.7431230269499764e-05,
"loss": 2.4146,
"step": 61000
},
{
"epoch": 0.44,
"learning_rate": 4.7389353682794765e-05,
"loss": 2.4074,
"step": 61500
},
{
"epoch": 0.45,
"learning_rate": 4.7347157304513544e-05,
"loss": 2.411,
"step": 62000
},
{
"epoch": 0.45,
"learning_rate": 4.7304641737355326e-05,
"loss": 2.411,
"step": 62500
},
{
"epoch": 0.45,
"learning_rate": 4.7261807588578375e-05,
"loss": 2.4068,
"step": 63000
},
{
"epoch": 0.45778831291669553,
"grad_norm": 0.4898933470249176,
"learning_rate": 4.721865546999135e-05,
"loss": 2.3865,
"step": 63500
},
{
"epoch": 0.4613929453018663,
"grad_norm": 0.5031083226203918,
"learning_rate": 4.7175185997944485e-05,
"loss": 2.3853,
"step": 64000
},
{
"epoch": 0.46499757768703714,
"grad_norm": 0.5099691152572632,
"learning_rate": 4.713139979332089e-05,
"loss": 2.3843,
"step": 64500
},
{
"epoch": 0.468602210072208,
"grad_norm": 0.528971791267395,
"learning_rate": 4.708729748152759e-05,
"loss": 2.3867,
"step": 65000
},
{
"epoch": 0.4722068424573788,
"grad_norm": 0.5447410941123962,
"learning_rate": 4.7042879692486655e-05,
"loss": 2.3871,
"step": 65500
},
{
"epoch": 0.47581147484254965,
"grad_norm": 0.5187677145004272,
"learning_rate": 4.699814706062616e-05,
"loss": 2.3821,
"step": 66000
},
{
"epoch": 0.4794161072277205,
"grad_norm": 0.5079197883605957,
"learning_rate": 4.6953100224871163e-05,
"loss": 2.3846,
"step": 66500
},
{
"epoch": 0.4830207396128913,
"grad_norm": 0.5204797387123108,
"learning_rate": 4.6907739828634536e-05,
"loss": 2.3867,
"step": 67000
},
{
"epoch": 0.48662537199806216,
"grad_norm": 0.51589435338974,
"learning_rate": 4.6862066519807816e-05,
"loss": 2.3792,
"step": 67500
},
{
"epoch": 0.490230004383233,
"grad_norm": 0.5342502593994141,
"learning_rate": 4.681608095075192e-05,
"loss": 2.3767,
"step": 68000
},
{
"epoch": 0.49383463676840383,
"grad_norm": 0.5265562534332275,
"learning_rate": 4.676978377828785e-05,
"loss": 2.3839,
"step": 68500
},
{
"epoch": 0.4974392691535746,
"grad_norm": 0.5328825116157532,
"learning_rate": 4.67231756636873e-05,
"loss": 2.3813,
"step": 69000
},
{
"epoch": 0.5010439015387455,
"grad_norm": 0.4992935061454773,
"learning_rate": 4.6676257272663194e-05,
"loss": 2.377,
"step": 69500
},
{
"epoch": 0.5046485339239163,
"grad_norm": 0.5137242674827576,
"learning_rate": 4.662902927536022e-05,
"loss": 2.3798,
"step": 70000
},
{
"epoch": 0.5082531663090871,
"grad_norm": 0.5188072323799133,
"learning_rate": 4.658149234634523e-05,
"loss": 2.3783,
"step": 70500
},
{
"epoch": 0.511857798694258,
"grad_norm": 0.5286343693733215,
"learning_rate": 4.65336471645976e-05,
"loss": 2.3719,
"step": 71000
},
{
"epoch": 0.5154624310794288,
"grad_norm": 0.5098246335983276,
"learning_rate": 4.648549441349953e-05,
"loss": 2.3712,
"step": 71500
},
{
"epoch": 0.5190670634645996,
"grad_norm": 0.528311014175415,
"learning_rate": 4.643703478082632e-05,
"loss": 2.37,
"step": 72000
},
{
"epoch": 0.5226716958497705,
"grad_norm": 0.5106325745582581,
"learning_rate": 4.638826895873651e-05,
"loss": 2.3717,
"step": 72500
},
{
"epoch": 0.5262763282349413,
"grad_norm": 0.5130262970924377,
"learning_rate": 4.6339197643762e-05,
"loss": 2.3669,
"step": 73000
},
{
"epoch": 0.5298809606201121,
"grad_norm": 0.5350182056427002,
"learning_rate": 4.628982153679811e-05,
"loss": 2.3675,
"step": 73500
},
{
"epoch": 0.533485593005283,
"grad_norm": 0.5258482098579407,
"learning_rate": 4.6240141343093545e-05,
"loss": 2.3673,
"step": 74000
},
{
"epoch": 0.5370902253904538,
"grad_norm": 0.5219452977180481,
"learning_rate": 4.6190157772240385e-05,
"loss": 2.3644,
"step": 74500
},
{
"epoch": 0.5406948577756246,
"grad_norm": 0.531566858291626,
"learning_rate": 4.6139871538163856e-05,
"loss": 2.366,
"step": 75000
},
{
"epoch": 0.5442994901607955,
"grad_norm": 0.5164352655410767,
"learning_rate": 4.608928335911222e-05,
"loss": 2.3651,
"step": 75500
},
{
"epoch": 0.5479041225459663,
"grad_norm": 0.558618426322937,
"learning_rate": 4.603839395764646e-05,
"loss": 2.3618,
"step": 76000
},
{
"epoch": 0.5515087549311372,
"grad_norm": 0.5367295145988464,
"learning_rate": 4.598720406062999e-05,
"loss": 2.3581,
"step": 76500
},
{
"epoch": 0.5551133873163079,
"grad_norm": 0.5240716934204102,
"learning_rate": 4.593571439921825e-05,
"loss": 2.3609,
"step": 77000
},
{
"epoch": 0.5587180197014787,
"grad_norm": 0.5218440890312195,
"learning_rate": 4.5883925708848275e-05,
"loss": 2.3583,
"step": 77500
},
{
"epoch": 0.5623226520866496,
"grad_norm": 0.5101985335350037,
"learning_rate": 4.58318387292282e-05,
"loss": 2.3587,
"step": 78000
}
],
"logging_steps": 500,
"max_steps": 416130,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 3000,
"total_flos": 2.8566582460416e+19,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}