| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.0, | |
| "eval_steps": 500, | |
| "global_step": 930, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08064516129032258, | |
| "grad_norm": 7.9788904562924206, | |
| "learning_rate": 2.1505376344086024e-07, | |
| "loss": 0.7746, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.16129032258064516, | |
| "grad_norm": 9.521156430604261, | |
| "learning_rate": 4.838709677419355e-07, | |
| "loss": 0.7594, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.24193548387096775, | |
| "grad_norm": 6.990008524163482, | |
| "learning_rate": 7.526881720430108e-07, | |
| "loss": 0.7177, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 5.245105441222518, | |
| "learning_rate": 1.021505376344086e-06, | |
| "loss": 0.622, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.4032258064516129, | |
| "grad_norm": 6.050490163797067, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 0.5555, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4838709677419355, | |
| "grad_norm": 4.382694958544679, | |
| "learning_rate": 1.5591397849462367e-06, | |
| "loss": 0.5051, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5645161290322581, | |
| "grad_norm": 3.5106557630691793, | |
| "learning_rate": 1.827956989247312e-06, | |
| "loss": 0.4399, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 3.6811492921662814, | |
| "learning_rate": 2.096774193548387e-06, | |
| "loss": 0.4063, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7258064516129032, | |
| "grad_norm": 3.441163884743535, | |
| "learning_rate": 2.3655913978494625e-06, | |
| "loss": 0.3723, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8064516129032258, | |
| "grad_norm": 3.447581913907039, | |
| "learning_rate": 2.634408602150538e-06, | |
| "loss": 0.4851, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8870967741935484, | |
| "grad_norm": 3.675136115986037, | |
| "learning_rate": 2.903225806451613e-06, | |
| "loss": 0.4024, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 3.0780394937329274, | |
| "learning_rate": 3.1720430107526885e-06, | |
| "loss": 0.413, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0483870967741935, | |
| "grad_norm": 3.2116900427519384, | |
| "learning_rate": 3.440860215053764e-06, | |
| "loss": 0.3071, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.129032258064516, | |
| "grad_norm": 4.009148207400134, | |
| "learning_rate": 3.7096774193548392e-06, | |
| "loss": 0.3085, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2096774193548387, | |
| "grad_norm": 2.8388506596193777, | |
| "learning_rate": 3.978494623655914e-06, | |
| "loss": 0.3009, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.2903225806451613, | |
| "grad_norm": 3.328140626191503, | |
| "learning_rate": 4.2473118279569895e-06, | |
| "loss": 0.3181, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.370967741935484, | |
| "grad_norm": 4.027543468539502, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 0.2963, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.4516129032258065, | |
| "grad_norm": 3.2636669590920286, | |
| "learning_rate": 4.78494623655914e-06, | |
| "loss": 0.2792, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.532258064516129, | |
| "grad_norm": 3.5992606337843487, | |
| "learning_rate": 5.0537634408602155e-06, | |
| "loss": 0.3347, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6129032258064515, | |
| "grad_norm": 3.5694108633818504, | |
| "learning_rate": 5.322580645161291e-06, | |
| "loss": 0.3348, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6935483870967742, | |
| "grad_norm": 3.660221218865697, | |
| "learning_rate": 5.591397849462365e-06, | |
| "loss": 0.3272, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.7741935483870968, | |
| "grad_norm": 3.4279728194623056, | |
| "learning_rate": 5.8602150537634415e-06, | |
| "loss": 0.2729, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8548387096774195, | |
| "grad_norm": 3.3534086258710825, | |
| "learning_rate": 6.129032258064517e-06, | |
| "loss": 0.2944, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.935483870967742, | |
| "grad_norm": 3.040737011508716, | |
| "learning_rate": 6.397849462365592e-06, | |
| "loss": 0.3298, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0161290322580645, | |
| "grad_norm": 3.0276013694676283, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.2396, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.096774193548387, | |
| "grad_norm": 4.83974864448008, | |
| "learning_rate": 6.935483870967743e-06, | |
| "loss": 0.1704, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1774193548387095, | |
| "grad_norm": 3.078350989376756, | |
| "learning_rate": 7.204301075268818e-06, | |
| "loss": 0.1689, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.258064516129032, | |
| "grad_norm": 3.4510008163142754, | |
| "learning_rate": 7.4731182795698935e-06, | |
| "loss": 0.1483, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.338709677419355, | |
| "grad_norm": 3.3742727028380353, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.1847, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.4193548387096775, | |
| "grad_norm": 2.9190658488467767, | |
| "learning_rate": 8.010752688172043e-06, | |
| "loss": 0.181, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 3.2768686821304227, | |
| "learning_rate": 8.279569892473119e-06, | |
| "loss": 0.201, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.5806451612903225, | |
| "grad_norm": 2.8067701037574753, | |
| "learning_rate": 8.548387096774194e-06, | |
| "loss": 0.1922, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.661290322580645, | |
| "grad_norm": 2.951622834663203, | |
| "learning_rate": 8.81720430107527e-06, | |
| "loss": 0.2015, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.741935483870968, | |
| "grad_norm": 3.03623148237881, | |
| "learning_rate": 9.086021505376345e-06, | |
| "loss": 0.1822, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.8225806451612905, | |
| "grad_norm": 3.0772573560886514, | |
| "learning_rate": 9.35483870967742e-06, | |
| "loss": 0.227, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.903225806451613, | |
| "grad_norm": 2.9654885185259827, | |
| "learning_rate": 9.623655913978495e-06, | |
| "loss": 0.2212, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9838709677419355, | |
| "grad_norm": 3.067899631927538, | |
| "learning_rate": 9.89247311827957e-06, | |
| "loss": 0.2277, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.064516129032258, | |
| "grad_norm": 2.7807126927253187, | |
| "learning_rate": 9.999920755303033e-06, | |
| "loss": 0.1288, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.1451612903225805, | |
| "grad_norm": 3.4305790367630777, | |
| "learning_rate": 9.999436491251425e-06, | |
| "loss": 0.1393, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.225806451612903, | |
| "grad_norm": 2.7900737962648043, | |
| "learning_rate": 9.998512030567253e-06, | |
| "loss": 0.1298, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.306451612903226, | |
| "grad_norm": 2.349901750902021, | |
| "learning_rate": 9.99714745464859e-06, | |
| "loss": 0.1306, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.3870967741935485, | |
| "grad_norm": 2.77370394045052, | |
| "learning_rate": 9.995342883645325e-06, | |
| "loss": 0.1521, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.467741935483871, | |
| "grad_norm": 3.0084835646623094, | |
| "learning_rate": 9.993098476448576e-06, | |
| "loss": 0.1314, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 3.5483870967741935, | |
| "grad_norm": 3.093621590181524, | |
| "learning_rate": 9.990414430676716e-06, | |
| "loss": 0.1545, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.629032258064516, | |
| "grad_norm": 3.1528581559810305, | |
| "learning_rate": 9.987290982657961e-06, | |
| "loss": 0.1929, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 3.709677419354839, | |
| "grad_norm": 2.8445198693562816, | |
| "learning_rate": 9.983728407409565e-06, | |
| "loss": 0.1305, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.790322580645161, | |
| "grad_norm": 2.960192414152783, | |
| "learning_rate": 9.979727018613607e-06, | |
| "loss": 0.1425, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.870967741935484, | |
| "grad_norm": 2.6571319169632703, | |
| "learning_rate": 9.975287168589369e-06, | |
| "loss": 0.1491, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.9516129032258065, | |
| "grad_norm": 3.3868215591855253, | |
| "learning_rate": 9.970409248262314e-06, | |
| "loss": 0.1688, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 4.032258064516129, | |
| "grad_norm": 2.091135299801856, | |
| "learning_rate": 9.965093687129669e-06, | |
| "loss": 0.1028, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.112903225806452, | |
| "grad_norm": 3.802878910271049, | |
| "learning_rate": 9.959340953222602e-06, | |
| "loss": 0.0842, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 4.193548387096774, | |
| "grad_norm": 3.7199837682735684, | |
| "learning_rate": 9.953151553065019e-06, | |
| "loss": 0.088, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.274193548387097, | |
| "grad_norm": 2.2960394303716116, | |
| "learning_rate": 9.94652603162896e-06, | |
| "loss": 0.0884, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 4.354838709677419, | |
| "grad_norm": 2.465062363580101, | |
| "learning_rate": 9.939464972286618e-06, | |
| "loss": 0.1006, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.435483870967742, | |
| "grad_norm": 2.0327702748628553, | |
| "learning_rate": 9.931968996758972e-06, | |
| "loss": 0.0899, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 4.516129032258064, | |
| "grad_norm": 2.305360963756021, | |
| "learning_rate": 9.924038765061042e-06, | |
| "loss": 0.079, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.596774193548387, | |
| "grad_norm": 2.4662415593135725, | |
| "learning_rate": 9.915674975443779e-06, | |
| "loss": 0.0875, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 4.67741935483871, | |
| "grad_norm": 2.555687280861228, | |
| "learning_rate": 9.906878364332586e-06, | |
| "loss": 0.0953, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 4.758064516129032, | |
| "grad_norm": 2.3818809469476476, | |
| "learning_rate": 9.897649706262474e-06, | |
| "loss": 0.0855, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 4.838709677419355, | |
| "grad_norm": 2.30924468453585, | |
| "learning_rate": 9.88798981380986e-06, | |
| "loss": 0.1138, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.919354838709677, | |
| "grad_norm": 2.047820547358064, | |
| "learning_rate": 9.877899537521028e-06, | |
| "loss": 0.1256, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.7534037559978417, | |
| "learning_rate": 9.867379765837237e-06, | |
| "loss": 0.1109, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 5.080645161290323, | |
| "grad_norm": 2.1913087837549052, | |
| "learning_rate": 9.85643142501649e-06, | |
| "loss": 0.0544, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 5.161290322580645, | |
| "grad_norm": 2.293004207498763, | |
| "learning_rate": 9.845055479051986e-06, | |
| "loss": 0.0678, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.241935483870968, | |
| "grad_norm": 1.9502100689755515, | |
| "learning_rate": 9.833252929587231e-06, | |
| "loss": 0.0596, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 5.32258064516129, | |
| "grad_norm": 2.377741796012424, | |
| "learning_rate": 9.821024815827854e-06, | |
| "loss": 0.0536, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.403225806451613, | |
| "grad_norm": 1.8788141547729034, | |
| "learning_rate": 9.808372214450093e-06, | |
| "loss": 0.066, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 5.483870967741936, | |
| "grad_norm": 2.18752083985909, | |
| "learning_rate": 9.795296239506011e-06, | |
| "loss": 0.064, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 5.564516129032258, | |
| "grad_norm": 2.1214320464857432, | |
| "learning_rate": 9.781798042325392e-06, | |
| "loss": 0.0718, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 5.645161290322581, | |
| "grad_norm": 2.3419858522206423, | |
| "learning_rate": 9.767878811414373e-06, | |
| "loss": 0.0751, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 5.725806451612903, | |
| "grad_norm": 2.0692559600108846, | |
| "learning_rate": 9.753539772350792e-06, | |
| "loss": 0.0559, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 5.806451612903226, | |
| "grad_norm": 1.7149217664306626, | |
| "learning_rate": 9.738782187676282e-06, | |
| "loss": 0.0615, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 5.887096774193548, | |
| "grad_norm": 2.420112879389103, | |
| "learning_rate": 9.723607356785103e-06, | |
| "loss": 0.0651, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 5.967741935483871, | |
| "grad_norm": 2.0342725263869488, | |
| "learning_rate": 9.70801661580973e-06, | |
| "loss": 0.0657, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 6.048387096774194, | |
| "grad_norm": 1.5676113830493696, | |
| "learning_rate": 9.692011337503212e-06, | |
| "loss": 0.048, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 6.129032258064516, | |
| "grad_norm": 1.4853313640607673, | |
| "learning_rate": 9.675592931118293e-06, | |
| "loss": 0.0395, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.209677419354839, | |
| "grad_norm": 2.076517691069182, | |
| "learning_rate": 9.658762842283343e-06, | |
| "loss": 0.0478, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 6.290322580645161, | |
| "grad_norm": 1.6175069613510986, | |
| "learning_rate": 9.641522552875055e-06, | |
| "loss": 0.0402, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 6.370967741935484, | |
| "grad_norm": 2.536372096562857, | |
| "learning_rate": 9.62387358088798e-06, | |
| "loss": 0.0512, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 6.451612903225806, | |
| "grad_norm": 1.8118088837546005, | |
| "learning_rate": 9.605817480300863e-06, | |
| "loss": 0.0534, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.532258064516129, | |
| "grad_norm": 2.286351385709345, | |
| "learning_rate": 9.587355840939813e-06, | |
| "loss": 0.0478, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 6.612903225806452, | |
| "grad_norm": 1.8901897009499684, | |
| "learning_rate": 9.568490288338324e-06, | |
| "loss": 0.0425, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 6.693548387096774, | |
| "grad_norm": 1.5196329706278056, | |
| "learning_rate": 9.549222483594154e-06, | |
| "loss": 0.0447, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 6.774193548387097, | |
| "grad_norm": 1.4880512934239636, | |
| "learning_rate": 9.529554123223053e-06, | |
| "loss": 0.0433, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 6.854838709677419, | |
| "grad_norm": 1.711263234463656, | |
| "learning_rate": 9.5094869390094e-06, | |
| "loss": 0.0452, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 6.935483870967742, | |
| "grad_norm": 1.8242713772702877, | |
| "learning_rate": 9.48902269785371e-06, | |
| "loss": 0.0417, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 7.016129032258065, | |
| "grad_norm": 1.408406163892193, | |
| "learning_rate": 9.468163201617063e-06, | |
| "loss": 0.05, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 7.096774193548387, | |
| "grad_norm": 1.0961989080978434, | |
| "learning_rate": 9.446910286962453e-06, | |
| "loss": 0.0249, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.17741935483871, | |
| "grad_norm": 1.7189796665774795, | |
| "learning_rate": 9.425265825193077e-06, | |
| "loss": 0.0317, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 7.258064516129032, | |
| "grad_norm": 1.9348300573137698, | |
| "learning_rate": 9.403231722087554e-06, | |
| "loss": 0.0347, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 7.338709677419355, | |
| "grad_norm": 1.7296791252489212, | |
| "learning_rate": 9.380809917732132e-06, | |
| "loss": 0.031, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 7.419354838709677, | |
| "grad_norm": 1.301188666000411, | |
| "learning_rate": 9.358002386349862e-06, | |
| "loss": 0.0337, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.6849307326292067, | |
| "learning_rate": 9.334811136126778e-06, | |
| "loss": 0.034, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 7.580645161290323, | |
| "grad_norm": 1.8341284468308772, | |
| "learning_rate": 9.31123820903506e-06, | |
| "loss": 0.0425, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 7.661290322580645, | |
| "grad_norm": 1.6185716325286443, | |
| "learning_rate": 9.287285680653254e-06, | |
| "loss": 0.0334, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 7.741935483870968, | |
| "grad_norm": 1.4082036883187536, | |
| "learning_rate": 9.262955659983522e-06, | |
| "loss": 0.0365, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 7.82258064516129, | |
| "grad_norm": 1.9810172286312782, | |
| "learning_rate": 9.238250289265921e-06, | |
| "loss": 0.031, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 7.903225806451613, | |
| "grad_norm": 1.8318457886483148, | |
| "learning_rate": 9.21317174378982e-06, | |
| "loss": 0.0354, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 7.983870967741936, | |
| "grad_norm": 1.674905845501415, | |
| "learning_rate": 9.187722231702326e-06, | |
| "loss": 0.0409, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 8.064516129032258, | |
| "grad_norm": 1.206360757615427, | |
| "learning_rate": 9.161903993813892e-06, | |
| "loss": 0.0276, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.14516129032258, | |
| "grad_norm": 0.9995201803814792, | |
| "learning_rate": 9.135719303400995e-06, | |
| "loss": 0.019, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 8.225806451612904, | |
| "grad_norm": 1.77138842254183, | |
| "learning_rate": 9.10917046600598e-06, | |
| "loss": 0.0235, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 8.306451612903226, | |
| "grad_norm": 1.772318685251266, | |
| "learning_rate": 9.082259819234063e-06, | |
| "loss": 0.0262, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 8.387096774193548, | |
| "grad_norm": 1.3694563307357386, | |
| "learning_rate": 9.054989732547507e-06, | |
| "loss": 0.0274, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 8.46774193548387, | |
| "grad_norm": 1.5818168003304018, | |
| "learning_rate": 9.027362607056986e-06, | |
| "loss": 0.0281, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 8.548387096774194, | |
| "grad_norm": 1.477790915380952, | |
| "learning_rate": 8.999380875310176e-06, | |
| "loss": 0.0274, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 8.629032258064516, | |
| "grad_norm": 1.463450411153143, | |
| "learning_rate": 8.971047001077561e-06, | |
| "loss": 0.0305, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 8.709677419354838, | |
| "grad_norm": 1.4651802172060195, | |
| "learning_rate": 8.942363479135516e-06, | |
| "loss": 0.0293, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 8.790322580645162, | |
| "grad_norm": 1.4324189291721474, | |
| "learning_rate": 8.913332835046629e-06, | |
| "loss": 0.0239, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 8.870967741935484, | |
| "grad_norm": 1.5257751575787353, | |
| "learning_rate": 8.883957624937333e-06, | |
| "loss": 0.0261, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 8.951612903225806, | |
| "grad_norm": 1.6993511552941885, | |
| "learning_rate": 8.854240435272842e-06, | |
| "loss": 0.0484, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 9.03225806451613, | |
| "grad_norm": 1.159130365005857, | |
| "learning_rate": 8.824183882629411e-06, | |
| "loss": 0.027, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 9.112903225806452, | |
| "grad_norm": 1.2166909931601617, | |
| "learning_rate": 8.793790613463956e-06, | |
| "loss": 0.0183, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 9.193548387096774, | |
| "grad_norm": 1.160324774431725, | |
| "learning_rate": 8.763063303881021e-06, | |
| "loss": 0.0188, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 9.274193548387096, | |
| "grad_norm": 1.3440474061044185, | |
| "learning_rate": 8.73200465939717e-06, | |
| "loss": 0.0224, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 9.35483870967742, | |
| "grad_norm": 1.1812237336382116, | |
| "learning_rate": 8.700617414702746e-06, | |
| "loss": 0.0222, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 9.435483870967742, | |
| "grad_norm": 1.2040220371499268, | |
| "learning_rate": 8.668904333421098e-06, | |
| "loss": 0.0213, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 9.516129032258064, | |
| "grad_norm": 1.1242316338416976, | |
| "learning_rate": 8.636868207865244e-06, | |
| "loss": 0.026, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 9.596774193548388, | |
| "grad_norm": 1.2056824125041377, | |
| "learning_rate": 8.604511858792006e-06, | |
| "loss": 0.0212, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 9.67741935483871, | |
| "grad_norm": 1.2737111889090427, | |
| "learning_rate": 8.571838135153645e-06, | |
| "loss": 0.0221, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.758064516129032, | |
| "grad_norm": 1.2658665183240156, | |
| "learning_rate": 8.538849913847019e-06, | |
| "loss": 0.0199, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 9.838709677419354, | |
| "grad_norm": 1.444864038120413, | |
| "learning_rate": 8.505550099460264e-06, | |
| "loss": 0.0242, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 9.919354838709678, | |
| "grad_norm": 1.464062720341168, | |
| "learning_rate": 8.471941624017058e-06, | |
| "loss": 0.0241, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.159759434728226, | |
| "learning_rate": 8.43802744671845e-06, | |
| "loss": 0.0262, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 10.080645161290322, | |
| "grad_norm": 1.2344076422984258, | |
| "learning_rate": 8.403810553682307e-06, | |
| "loss": 0.018, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 10.161290322580646, | |
| "grad_norm": 1.1047319831516356, | |
| "learning_rate": 8.369293957680397e-06, | |
| "loss": 0.0156, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 10.241935483870968, | |
| "grad_norm": 0.9725174239581482, | |
| "learning_rate": 8.334480697873101e-06, | |
| "loss": 0.014, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 10.32258064516129, | |
| "grad_norm": 1.0873207313793845, | |
| "learning_rate": 8.299373839541829e-06, | |
| "loss": 0.0184, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 10.403225806451612, | |
| "grad_norm": 1.1713325775584265, | |
| "learning_rate": 8.26397647381912e-06, | |
| "loss": 0.0191, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 10.483870967741936, | |
| "grad_norm": 1.1289992391197494, | |
| "learning_rate": 8.228291717416472e-06, | |
| "loss": 0.0192, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 10.564516129032258, | |
| "grad_norm": 1.4540729880318348, | |
| "learning_rate": 8.192322712349917e-06, | |
| "loss": 0.0209, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 10.64516129032258, | |
| "grad_norm": 1.2131150635320254, | |
| "learning_rate": 8.15607262566337e-06, | |
| "loss": 0.0201, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 10.725806451612904, | |
| "grad_norm": 1.583339620420521, | |
| "learning_rate": 8.119544649149762e-06, | |
| "loss": 0.0207, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 10.806451612903226, | |
| "grad_norm": 1.298849150527335, | |
| "learning_rate": 8.08274199907003e-06, | |
| "loss": 0.017, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 10.887096774193548, | |
| "grad_norm": 1.1682269254944644, | |
| "learning_rate": 8.0456679158699e-06, | |
| "loss": 0.0175, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 10.967741935483872, | |
| "grad_norm": 1.3418845328882663, | |
| "learning_rate": 8.008325663894586e-06, | |
| "loss": 0.0199, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 11.048387096774194, | |
| "grad_norm": 1.0350578214030737, | |
| "learning_rate": 7.970718531101365e-06, | |
| "loss": 0.0173, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 11.129032258064516, | |
| "grad_norm": 0.7751062982939178, | |
| "learning_rate": 7.932849828770062e-06, | |
| "loss": 0.0154, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 11.209677419354838, | |
| "grad_norm": 1.1249340896245428, | |
| "learning_rate": 7.89472289121151e-06, | |
| "loss": 0.0153, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 11.290322580645162, | |
| "grad_norm": 1.069133814090821, | |
| "learning_rate": 7.856341075473963e-06, | |
| "loss": 0.0128, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 11.370967741935484, | |
| "grad_norm": 1.3386909438492063, | |
| "learning_rate": 7.817707761047498e-06, | |
| "loss": 0.0147, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 11.451612903225806, | |
| "grad_norm": 1.2544557604239914, | |
| "learning_rate": 7.77882634956647e-06, | |
| "loss": 0.0137, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 11.532258064516128, | |
| "grad_norm": 1.50608779604205, | |
| "learning_rate": 7.739700264509993e-06, | |
| "loss": 0.0158, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 11.612903225806452, | |
| "grad_norm": 0.8383520545182386, | |
| "learning_rate": 7.700332950900504e-06, | |
| "loss": 0.0115, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 11.693548387096774, | |
| "grad_norm": 1.1528102169152703, | |
| "learning_rate": 7.660727875000432e-06, | |
| "loss": 0.0151, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 11.774193548387096, | |
| "grad_norm": 1.1731217245462982, | |
| "learning_rate": 7.6208885240069995e-06, | |
| "loss": 0.0137, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 11.85483870967742, | |
| "grad_norm": 1.0109598241071842, | |
| "learning_rate": 7.5808184057451765e-06, | |
| "loss": 0.0133, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 11.935483870967742, | |
| "grad_norm": 1.1815937319956697, | |
| "learning_rate": 7.540521048358814e-06, | |
| "loss": 0.0168, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 12.016129032258064, | |
| "grad_norm": 0.6727751980657929, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0145, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 12.096774193548388, | |
| "grad_norm": 0.8287342886061079, | |
| "learning_rate": 7.459258828516645e-06, | |
| "loss": 0.0108, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 12.17741935483871, | |
| "grad_norm": 0.8907982669901318, | |
| "learning_rate": 7.418301121138335e-06, | |
| "loss": 0.0082, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 12.258064516129032, | |
| "grad_norm": 1.08087496747789, | |
| "learning_rate": 7.3771304841604764e-06, | |
| "loss": 0.0121, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 12.338709677419354, | |
| "grad_norm": 0.9242022062218133, | |
| "learning_rate": 7.335750542626772e-06, | |
| "loss": 0.0118, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 12.419354838709678, | |
| "grad_norm": 0.6674173817031188, | |
| "learning_rate": 7.294164940010031e-06, | |
| "loss": 0.0088, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.7427776074770551, | |
| "learning_rate": 7.2523773378913655e-06, | |
| "loss": 0.01, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 12.580645161290322, | |
| "grad_norm": 0.8105255073657276, | |
| "learning_rate": 7.210391415637797e-06, | |
| "loss": 0.0102, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 12.661290322580646, | |
| "grad_norm": 0.9355789095839642, | |
| "learning_rate": 7.168210870078277e-06, | |
| "loss": 0.0112, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 12.741935483870968, | |
| "grad_norm": 0.8096036161441085, | |
| "learning_rate": 7.125839415178204e-06, | |
| "loss": 0.0109, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 12.82258064516129, | |
| "grad_norm": 0.6993400061698866, | |
| "learning_rate": 7.083280781712394e-06, | |
| "loss": 0.0125, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 12.903225806451612, | |
| "grad_norm": 0.9083278507643704, | |
| "learning_rate": 7.0405387169365965e-06, | |
| "loss": 0.0128, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 12.983870967741936, | |
| "grad_norm": 0.8581097245501025, | |
| "learning_rate": 6.9976169842575526e-06, | |
| "loss": 0.0095, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 13.064516129032258, | |
| "grad_norm": 1.1318882973999245, | |
| "learning_rate": 6.9545193629016215e-06, | |
| "loss": 0.0053, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 13.14516129032258, | |
| "grad_norm": 0.6323423097700148, | |
| "learning_rate": 6.911249647582036e-06, | |
| "loss": 0.0071, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 13.225806451612904, | |
| "grad_norm": 0.9759430461350306, | |
| "learning_rate": 6.867811648164769e-06, | |
| "loss": 0.0084, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 13.306451612903226, | |
| "grad_norm": 0.5436428472489789, | |
| "learning_rate": 6.824209189333082e-06, | |
| "loss": 0.0059, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 13.387096774193548, | |
| "grad_norm": 0.8064226300726538, | |
| "learning_rate": 6.780446110250766e-06, | |
| "loss": 0.0083, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 13.46774193548387, | |
| "grad_norm": 0.6692079180896646, | |
| "learning_rate": 6.736526264224101e-06, | |
| "loss": 0.0069, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 13.548387096774194, | |
| "grad_norm": 0.7920667479327235, | |
| "learning_rate": 6.692453518362587e-06, | |
| "loss": 0.0087, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 13.629032258064516, | |
| "grad_norm": 0.9743408271346984, | |
| "learning_rate": 6.648231753238431e-06, | |
| "loss": 0.0108, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 13.709677419354838, | |
| "grad_norm": 1.054329395592257, | |
| "learning_rate": 6.603864862544879e-06, | |
| "loss": 0.008, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 13.790322580645162, | |
| "grad_norm": 0.6890711609162895, | |
| "learning_rate": 6.5593567527533715e-06, | |
| "loss": 0.007, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 13.870967741935484, | |
| "grad_norm": 0.8929770220878291, | |
| "learning_rate": 6.514711342769588e-06, | |
| "loss": 0.0094, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 13.951612903225806, | |
| "grad_norm": 0.8470688740181095, | |
| "learning_rate": 6.469932563588386e-06, | |
| "loss": 0.0074, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 14.03225806451613, | |
| "grad_norm": 0.929920833432491, | |
| "learning_rate": 6.425024357947677e-06, | |
| "loss": 0.0081, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 14.112903225806452, | |
| "grad_norm": 0.5368583000455479, | |
| "learning_rate": 6.3799906799812805e-06, | |
| "loss": 0.0039, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 14.193548387096774, | |
| "grad_norm": 0.5542449665774428, | |
| "learning_rate": 6.334835494870759e-06, | |
| "loss": 0.0058, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 14.274193548387096, | |
| "grad_norm": 1.001191478848351, | |
| "learning_rate": 6.289562778496285e-06, | |
| "loss": 0.0058, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 14.35483870967742, | |
| "grad_norm": 0.4192339189229186, | |
| "learning_rate": 6.244176517086573e-06, | |
| "loss": 0.0033, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 14.435483870967742, | |
| "grad_norm": 0.5368429196948196, | |
| "learning_rate": 6.1986807068678926e-06, | |
| "loss": 0.0052, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 14.516129032258064, | |
| "grad_norm": 0.9333482099983527, | |
| "learning_rate": 6.153079353712201e-06, | |
| "loss": 0.0057, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 14.596774193548388, | |
| "grad_norm": 1.0750824183015748, | |
| "learning_rate": 6.107376472784438e-06, | |
| "loss": 0.0064, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 14.67741935483871, | |
| "grad_norm": 0.46525235985013963, | |
| "learning_rate": 6.061576088188981e-06, | |
| "loss": 0.0039, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 14.758064516129032, | |
| "grad_norm": 0.7126091866734285, | |
| "learning_rate": 6.015682232615336e-06, | |
| "loss": 0.0064, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 14.838709677419354, | |
| "grad_norm": 0.5389766142133813, | |
| "learning_rate": 5.969698946983055e-06, | |
| "loss": 0.0046, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 14.919354838709678, | |
| "grad_norm": 0.7912398368360793, | |
| "learning_rate": 5.923630280085948e-06, | |
| "loss": 0.0066, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.3365541360781848, | |
| "learning_rate": 5.877480288235569e-06, | |
| "loss": 0.0041, | |
| "step": 930 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1860, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 310, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 17705948282880.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |