| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 300, | |
| "global_step": 1089, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0027548209366391185, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1e-06, | |
| "loss": 1.9168, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013774104683195593, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 1.999962411893365e-06, | |
| "loss": 1.9099, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.027548209366391185, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 1.9997327170302815e-06, | |
| "loss": 1.8629, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04132231404958678, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 1.99929425749243e-06, | |
| "loss": 1.8376, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05509641873278237, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 1.998647124839145e-06, | |
| "loss": 1.8137, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06887052341597796, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 1.997791454204984e-06, | |
| "loss": 1.8055, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08264462809917356, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.9967274242715065e-06, | |
| "loss": 1.7913, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09641873278236915, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.995455257229964e-06, | |
| "loss": 1.7897, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.11019283746556474, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.9939752187349e-06, | |
| "loss": 1.7752, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12396694214876033, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9922876178486764e-06, | |
| "loss": 1.7802, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.13774104683195593, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.9903928069769356e-06, | |
| "loss": 1.7686, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.9882911817950105e-06, | |
| "loss": 1.7702, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1652892561983471, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.985983181165299e-06, | |
| "loss": 1.7618, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1790633608815427, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.98346928704562e-06, | |
| "loss": 1.7627, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.1928374655647383, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 1.9807500243885744e-06, | |
| "loss": 1.7665, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2066115702479339, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.9778259610319187e-06, | |
| "loss": 1.755, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.22038567493112948, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 1.9746977075799933e-06, | |
| "loss": 1.7574, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23415977961432508, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 1.9713659172762126e-06, | |
| "loss": 1.7529, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.24793388429752067, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 1.9678312858666578e-06, | |
| "loss": 1.7417, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.26170798898071623, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 1.964094551454788e-06, | |
| "loss": 1.7509, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.27548209366391185, | |
| "grad_norm": 1.5, | |
| "learning_rate": 1.960156494347309e-06, | |
| "loss": 1.7486, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2892561983471074, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 1.9560179368912327e-06, | |
| "loss": 1.7531, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.95167974330215e-06, | |
| "loss": 1.7435, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3168044077134986, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 1.9471428194837667e-06, | |
| "loss": 1.7403, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3305785123966942, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 1.9424081128387337e-06, | |
| "loss": 1.7435, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3443526170798898, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.9374766120708077e-06, | |
| "loss": 1.75, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3581267217630854, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.932349346978389e-06, | |
| "loss": 1.7469, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.371900826446281, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.927027388239482e-06, | |
| "loss": 1.734, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3856749311294766, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.921511847188112e-06, | |
| "loss": 1.7385, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.39944903581267216, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 1.915803875582259e-06, | |
| "loss": 1.736, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4132231404958678, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.9099046653633437e-06, | |
| "loss": 1.7336, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.42699724517906334, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.9038154484073284e-06, | |
| "loss": 1.7324, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.44077134986225897, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.8975374962674753e-06, | |
| "loss": 1.7406, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.8910721199088195e-06, | |
| "loss": 1.7461, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.46831955922865015, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.8844206694344138e-06, | |
| "loss": 1.7296, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4820936639118457, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 1.877584533803398e-06, | |
| "loss": 1.7286, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.49586776859504134, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.8705651405409566e-06, | |
| "loss": 1.7225, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.509641873278237, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.8633639554402234e-06, | |
| "loss": 1.7366, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5234159779614325, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.8559824822561913e-06, | |
| "loss": 1.7353, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5371900826446281, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.8484222623917e-06, | |
| "loss": 1.7223, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5509641873278237, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.8406848745755578e-06, | |
| "loss": 1.7256, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5647382920110193, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 1.832771934532872e-06, | |
| "loss": 1.7288, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5785123966942148, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.8246850946476505e-06, | |
| "loss": 1.7247, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5922865013774105, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.8164260436177524e-06, | |
| "loss": 1.7221, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.8079965061022518e-06, | |
| "loss": 1.7224, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6198347107438017, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.7993982423612941e-06, | |
| "loss": 1.7206, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6336088154269972, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.7906330478885174e-06, | |
| "loss": 1.7252, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6473829201101928, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.7817027530361174e-06, | |
| "loss": 1.725, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6611570247933884, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.7726092226326315e-06, | |
| "loss": 1.7189, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6749311294765841, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.7633543555935245e-06, | |
| "loss": 1.719, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6887052341597796, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.7539400845246564e-06, | |
| "loss": 1.7121, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.7024793388429752, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.744368375318715e-06, | |
| "loss": 1.7198, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.7162534435261708, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.7346412267446958e-06, | |
| "loss": 1.7149, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7300275482093664, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.724760670030521e-06, | |
| "loss": 1.713, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.743801652892562, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.7147287684388738e-06, | |
| "loss": 1.7192, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.7045476168363498e-06, | |
| "loss": 1.721, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7713498622589532, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.6942193412560043e-06, | |
| "loss": 1.7102, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7851239669421488, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.6837460984533934e-06, | |
| "loss": 1.7218, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7988980716253443, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.6731300754562008e-06, | |
| "loss": 1.7107, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.8126721763085399, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.6623734891075385e-06, | |
| "loss": 1.7138, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.8264462809917356, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.6514785856030272e-06, | |
| "loss": 1.7112, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8264462809917356, | |
| "eval_loss": 1.7193405628204346, | |
| "eval_runtime": 8.403, | |
| "eval_samples_per_second": 83.779, | |
| "eval_steps_per_second": 2.618, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8402203856749312, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.640447640021744e-06, | |
| "loss": 1.7211, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.8539944903581267, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.6292829558511376e-06, | |
| "loss": 1.7142, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8677685950413223, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.6179868645060162e-06, | |
| "loss": 1.7028, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8815426997245179, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.6065617248416967e-06, | |
| "loss": 1.7091, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8953168044077136, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 1.59500992266143e-06, | |
| "loss": 1.7143, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.5833338702181959e-06, | |
| "loss": 1.7105, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.9228650137741047, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.5715360057109744e-06, | |
| "loss": 1.7146, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.9366391184573003, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 1.5596187927755993e-06, | |
| "loss": 1.7145, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9504132231404959, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.5475847199703033e-06, | |
| "loss": 1.7099, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.9641873278236914, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.535436300256053e-06, | |
| "loss": 1.7143, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.977961432506887, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.523176070471793e-06, | |
| "loss": 1.7131, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9917355371900827, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.5108065908047014e-06, | |
| "loss": 1.7248, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.0055096418732783, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.4983304442555698e-06, | |
| "loss": 1.7079, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.019283746556474, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 1.4857502360994204e-06, | |
| "loss": 1.712, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.0330578512396693, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.4730685933414714e-06, | |
| "loss": 1.703, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.046831955922865, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.4602881641685643e-06, | |
| "loss": 1.6944, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0606060606060606, | |
| "grad_norm": 1.375, | |
| "learning_rate": 1.4474116173961668e-06, | |
| "loss": 1.7032, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.0743801652892562, | |
| "grad_norm": 1.375, | |
| "learning_rate": 1.4344416419110728e-06, | |
| "loss": 1.7122, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0881542699724518, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.4213809461099033e-06, | |
| "loss": 1.703, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.1019283746556474, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.4082322573335422e-06, | |
| "loss": 1.703, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.115702479338843, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.394998321297608e-06, | |
| "loss": 1.7024, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.1294765840220387, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.3816819015190943e-06, | |
| "loss": 1.7058, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.1432506887052343, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.3682857787392905e-06, | |
| "loss": 1.6987, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.1570247933884297, | |
| "grad_norm": 1.375, | |
| "learning_rate": 1.3548127503431038e-06, | |
| "loss": 1.7029, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.1707988980716253, | |
| "grad_norm": 1.375, | |
| "learning_rate": 1.3412656297749135e-06, | |
| "loss": 1.6998, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.184573002754821, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.327647245951058e-06, | |
| "loss": 1.7051, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.1983471074380165, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.3139604426691072e-06, | |
| "loss": 1.7065, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.300208078014014e-06, | |
| "loss": 1.7019, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.2258953168044078, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.2863930237612896e-06, | |
| "loss": 1.6999, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.2396694214876034, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.2725181647773174e-06, | |
| "loss": 1.7046, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.2534435261707988, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 1.2585863984169343e-06, | |
| "loss": 1.7069, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.2672176308539944, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.2446006339184035e-06, | |
| "loss": 1.7025, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.28099173553719, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.2305637917959058e-06, | |
| "loss": 1.7016, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.2947658402203857, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.2164788032296755e-06, | |
| "loss": 1.6997, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.3085399449035813, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.2023486094539124e-06, | |
| "loss": 1.7003, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.322314049586777, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 1.1881761611425888e-06, | |
| "loss": 1.6885, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.3360881542699725, | |
| "grad_norm": 1.375, | |
| "learning_rate": 1.1739644177932907e-06, | |
| "loss": 1.7, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.3498622589531681, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.159716347109213e-06, | |
| "loss": 1.6989, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 1.1454349243794419e-06, | |
| "loss": 1.7009, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.3774104683195592, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.1311231318576545e-06, | |
| "loss": 1.7007, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.3911845730027548, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.1167839581393628e-06, | |
| "loss": 1.6918, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.4049586776859504, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 1.1024203975378335e-06, | |
| "loss": 1.6975, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.418732782369146, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.0880354494588138e-06, | |
| "loss": 1.6998, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.4325068870523416, | |
| "grad_norm": 1.375, | |
| "learning_rate": 1.073632117774195e-06, | |
| "loss": 1.6973, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.4462809917355373, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 1.0592134101947417e-06, | |
| "loss": 1.7023, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.4600550964187327, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 1.0447823376420206e-06, | |
| "loss": 1.707, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.4738292011019283, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.0303419136196575e-06, | |
| "loss": 1.6916, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.487603305785124, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.0158951535840576e-06, | |
| "loss": 1.6977, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.5013774104683195, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.0014450743147145e-06, | |
| "loss": 1.699, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 9.869946932842466e-07, | |
| "loss": 1.6907, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.5289256198347108, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 9.725470280282855e-07, | |
| "loss": 1.7001, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.5426997245179064, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 9.581050955153545e-07, | |
| "loss": 1.7054, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.556473829201102, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 9.43671911516861e-07, | |
| "loss": 1.687, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.5702479338842976, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 9.292504899773453e-07, | |
| "loss": 1.7055, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.5840220385674932, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 9.148438423851041e-07, | |
| "loss": 1.6975, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.5977961432506889, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 9.00454977143331e-07, | |
| "loss": 1.6997, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.6115702479338843, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 8.860868989419017e-07, | |
| "loss": 1.6983, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.6253443526170799, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 8.717426081299308e-07, | |
| "loss": 1.6995, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.6391184573002755, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 8.574251000892386e-07, | |
| "loss": 1.6948, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.6528925619834711, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 8.431373646088549e-07, | |
| "loss": 1.6961, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.6528925619834711, | |
| "eval_loss": 1.7039618492126465, | |
| "eval_runtime": 8.3736, | |
| "eval_samples_per_second": 84.073, | |
| "eval_steps_per_second": 2.627, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 8.288823852606882e-07, | |
| "loss": 1.6915, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.6804407713498621, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 8.14663138776496e-07, | |
| "loss": 1.7006, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.6942148760330578, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 8.004825944262805e-07, | |
| "loss": 1.7029, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.7079889807162534, | |
| "grad_norm": 1.375, | |
| "learning_rate": 7.863437133982471e-07, | |
| "loss": 1.6942, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.721763085399449, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 7.722494481804445e-07, | |
| "loss": 1.6927, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.7355371900826446, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 7.582027419442268e-07, | |
| "loss": 1.6929, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.7493112947658402, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 7.442065279296578e-07, | |
| "loss": 1.702, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.7630853994490359, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 7.302637288329915e-07, | |
| "loss": 1.6971, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.7768595041322315, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 7.163772561963519e-07, | |
| "loss": 1.6923, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.790633608815427, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 7.02550009799745e-07, | |
| "loss": 1.6883, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.8044077134986227, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 6.887848770555234e-07, | |
| "loss": 1.6991, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 6.750847324054374e-07, | |
| "loss": 1.7069, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.8319559228650137, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 6.614524367203906e-07, | |
| "loss": 1.6979, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.8457300275482094, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 6.478908367030338e-07, | |
| "loss": 1.6829, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.859504132231405, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 6.344027642933128e-07, | |
| "loss": 1.692, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.8732782369146006, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 6.209910360771033e-07, | |
| "loss": 1.703, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.887052341597796, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 6.076584526980484e-07, | |
| "loss": 1.706, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.9008264462809916, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 5.944077982727285e-07, | |
| "loss": 1.6906, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.9146005509641872, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 5.812418398092787e-07, | |
| "loss": 1.6982, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.9283746556473829, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 5.681633266295834e-07, | |
| "loss": 1.6917, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.9421487603305785, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 5.551749897951582e-07, | |
| "loss": 1.7014, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.955922865013774, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 5.422795415368518e-07, | |
| "loss": 1.6861, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.9696969696969697, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 5.294796746884745e-07, | |
| "loss": 1.6953, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.9834710743801653, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 5.167780621244801e-07, | |
| "loss": 1.6973, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.997245179063361, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 5.041773562018135e-07, | |
| "loss": 1.7019, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.0110192837465566, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 4.91680188206047e-07, | |
| "loss": 1.7011, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.024793388429752, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 4.792891678019115e-07, | |
| "loss": 1.7013, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.038567493112948, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 4.6700688248834664e-07, | |
| "loss": 1.6895, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.0523415977961434, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 4.548358970581757e-07, | |
| "loss": 1.7029, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.0661157024793386, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 4.427787530625278e-07, | |
| "loss": 1.6931, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.0798898071625342, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 4.3083796828010675e-07, | |
| "loss": 1.6886, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.09366391184573, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 4.190160361914292e-07, | |
| "loss": 1.6907, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.1074380165289255, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 4.07315425458134e-07, | |
| "loss": 1.6924, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.121212121212121, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 3.9573857940747537e-07, | |
| "loss": 1.7019, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.1349862258953167, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 3.8428791552210594e-07, | |
| "loss": 1.6975, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.1487603305785123, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 3.729658249352563e-07, | |
| "loss": 1.6986, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.162534435261708, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 3.6177467193141886e-07, | |
| "loss": 1.6893, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.1763085399449036, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 3.5071679345263537e-07, | |
| "loss": 1.6833, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.190082644628099, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 3.397944986104968e-07, | |
| "loss": 1.693, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 2.203856749311295, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 3.290100682039516e-07, | |
| "loss": 1.6978, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.2176308539944904, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 3.1836575424303034e-07, | |
| "loss": 1.7019, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 2.231404958677686, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 3.078637794785791e-07, | |
| "loss": 1.6977, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.2451790633608817, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 2.9750633693810224e-07, | |
| "loss": 1.6898, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 2.2589531680440773, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 2.872955894678153e-07, | |
| "loss": 1.6915, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 2.7723366928099754e-07, | |
| "loss": 1.6922, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.2865013774104685, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 2.673226775127422e-07, | |
| "loss": 1.6922, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.3002754820936637, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 2.5756468378119533e-07, | |
| "loss": 1.6873, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.3140495867768593, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 2.4796172575537934e-07, | |
| "loss": 1.7068, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.327823691460055, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 2.3851580872968435e-07, | |
| "loss": 1.6993, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.3415977961432506, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 2.292289052051224e-07, | |
| "loss": 1.6992, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.355371900826446, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 2.2010295447742743e-07, | |
| "loss": 1.6891, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.369146005509642, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 2.111398622320927e-07, | |
| "loss": 1.6968, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.3829201101928374, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 2.0234150014642305e-07, | |
| "loss": 1.6946, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.396694214876033, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.937097054986915e-07, | |
| "loss": 1.6892, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.4104683195592287, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 1.8524628078447602e-07, | |
| "loss": 1.6915, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.769529933402637e-07, | |
| "loss": 1.6946, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.43801652892562, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 1.6883157497439349e-07, | |
| "loss": 1.6975, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 2.4517906336088156, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.6088372160541962e-07, | |
| "loss": 1.6871, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.465564738292011, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.531110929079681e-07, | |
| "loss": 1.6909, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 2.479338842975207, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 1.4551531196616396e-07, | |
| "loss": 1.6908, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.479338842975207, | |
| "eval_loss": 1.7025996446609497, | |
| "eval_runtime": 8.3873, | |
| "eval_samples_per_second": 83.936, | |
| "eval_steps_per_second": 2.623, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.4931129476584024, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.3809796493469728e-07, | |
| "loss": 1.6981, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 2.5068870523415976, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.3086060070760196e-07, | |
| "loss": 1.6902, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.5206611570247937, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.23804730594814e-07, | |
| "loss": 1.6964, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 2.534435261707989, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.1693182800658042e-07, | |
| "loss": 1.6884, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.5482093663911844, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.102433281457802e-07, | |
| "loss": 1.6969, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.56198347107438, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.0374062770822411e-07, | |
| "loss": 1.7003, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.5757575757575757, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 9.742508459099707e-08, | |
| "loss": 1.7095, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.5895316804407713, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 9.129801760890076e-08, | |
| "loss": 1.7026, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.603305785123967, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 8.536070621905811e-08, | |
| "loss": 1.6964, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.6170798898071626, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 7.961439025373617e-08, | |
| "loss": 1.6984, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.630853994490358, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 7.40602696614444e-08, | |
| "loss": 1.7022, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.644628099173554, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 6.869950425636095e-08, | |
| "loss": 1.6955, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.6584022038567494, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 6.353321347613815e-08, | |
| "loss": 1.6962, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.672176308539945, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 5.856247614814292e-08, | |
| "loss": 1.6914, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.6859504132231407, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 5.3788330264174506e-08, | |
| "loss": 1.6934, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.6997245179063363, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 4.921177276371069e-08, | |
| "loss": 1.6947, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.7134986225895315, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 4.483375932572597e-08, | |
| "loss": 1.6929, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 4.0655204169127156e-08, | |
| "loss": 1.6944, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.7410468319559227, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 3.667697986184526e-08, | |
| "loss": 1.6898, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.7548209366391183, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 3.2899917138625055e-08, | |
| "loss": 1.7061, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.768595041322314, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 2.9324804727551055e-08, | |
| "loss": 1.6974, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.7823691460055096, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 2.5952389185344925e-08, | |
| "loss": 1.6892, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.796143250688705, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 2.2783374741469186e-08, | |
| "loss": 1.696, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.809917355371901, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 1.9818423151069406e-08, | |
| "loss": 1.6879, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.8236914600550964, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 1.705815355678619e-08, | |
| "loss": 1.6943, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.837465564738292, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.4503142359465925e-08, | |
| "loss": 1.6919, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.8512396694214877, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.215392309779617e-08, | |
| "loss": 1.6907, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.8650137741046833, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.0010986336891458e-08, | |
| "loss": 1.704, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.878787878787879, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 8.074779565854117e-09, | |
| "loss": 1.691, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.8925619834710745, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 6.34570710432869e-09, | |
| "loss": 1.6975, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.90633608815427, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 4.824130018072026e-09, | |
| "loss": 1.6918, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.9201101928374653, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 3.5103660435551465e-09, | |
| "loss": 1.6933, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.9338842975206614, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 2.4046895216136563e-09, | |
| "loss": 1.6872, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.9476584022038566, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 1.5073313401594568e-09, | |
| "loss": 1.696, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.9614325068870526, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 8.184788859667557e-10, | |
| "loss": 1.6964, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.975206611570248, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 3.3827600554170444e-10, | |
| "loss": 1.6941, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.9889807162534434, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 6.682297508464608e-11, | |
| "loss": 1.6993, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1089, | |
| "total_flos": 6.427401199279931e+18, | |
| "train_loss": 1.711617823146263, | |
| "train_runtime": 5145.0339, | |
| "train_samples_per_second": 13.545, | |
| "train_steps_per_second": 0.212 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1089, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.427401199279931e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |