| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9905341446923597, | |
| "eval_steps": 46, | |
| "global_step": 368, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005409060175794456, | |
| "grad_norm": 44.9358828224533, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 3.0416, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005409060175794456, | |
| "eval_loss": 3.2299506664276123, | |
| "eval_runtime": 80.3263, | |
| "eval_samples_per_second": 15.512, | |
| "eval_steps_per_second": 1.942, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.010818120351588911, | |
| "grad_norm": 43.82162868015869, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 3.0979, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.016227180527383367, | |
| "grad_norm": 49.66354655578431, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 3.229, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.021636240703177823, | |
| "grad_norm": 44.90350470119676, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 3.1823, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.027045300878972278, | |
| "grad_norm": 45.64225015268925, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 3.1503, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.032454361054766734, | |
| "grad_norm": 43.46015056839171, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 2.9786, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03786342123056119, | |
| "grad_norm": 39.8236071706638, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 2.7861, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.043272481406355645, | |
| "grad_norm": 18.849036433644883, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 2.6546, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0486815415821501, | |
| "grad_norm": 18.842114481944215, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 2.685, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.054090601757944556, | |
| "grad_norm": 19.369280399976688, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.6052, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05949966193373901, | |
| "grad_norm": 16.167411037120388, | |
| "learning_rate": 2.2e-06, | |
| "loss": 2.4654, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06490872210953347, | |
| "grad_norm": 14.946675935925768, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 2.4427, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07031778228532792, | |
| "grad_norm": 13.586548911059968, | |
| "learning_rate": 2.6e-06, | |
| "loss": 2.5375, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07572684246112238, | |
| "grad_norm": 9.968943698698908, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 2.316, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08113590263691683, | |
| "grad_norm": 10.416045938582544, | |
| "learning_rate": 3e-06, | |
| "loss": 2.4094, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08654496281271129, | |
| "grad_norm": 9.298587355474794, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 2.3988, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.09195402298850575, | |
| "grad_norm": 8.609391957088402, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 2.3259, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0973630831643002, | |
| "grad_norm": 7.68261305571874, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 2.2034, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10277214334009466, | |
| "grad_norm": 8.199455018530278, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 2.1983, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.10818120351588911, | |
| "grad_norm": 7.229515373269984, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.3398, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11359026369168357, | |
| "grad_norm": 8.94152448817592, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 2.2895, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.11899932386747802, | |
| "grad_norm": 7.678296028460811, | |
| "learning_rate": 4.4e-06, | |
| "loss": 2.2891, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.12440838404327248, | |
| "grad_norm": 8.007524747470107, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 2.2389, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.12981744421906694, | |
| "grad_norm": 7.352151792895566, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 2.2593, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1352265043948614, | |
| "grad_norm": 7.103281407665676, | |
| "learning_rate": 5e-06, | |
| "loss": 2.1577, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14063556457065585, | |
| "grad_norm": 6.919318449883582, | |
| "learning_rate": 4.999955579111413e-06, | |
| "loss": 2.2622, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1460446247464503, | |
| "grad_norm": 7.072321363589398, | |
| "learning_rate": 4.999822318024222e-06, | |
| "loss": 2.1805, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.15145368492224476, | |
| "grad_norm": 8.556428262133437, | |
| "learning_rate": 4.999600221474089e-06, | |
| "loss": 2.2209, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1568627450980392, | |
| "grad_norm": 6.318203548443499, | |
| "learning_rate": 4.999289297353593e-06, | |
| "loss": 2.1821, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.16227180527383367, | |
| "grad_norm": 7.959416954439196, | |
| "learning_rate": 4.998889556711958e-06, | |
| "loss": 2.1423, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.16768086544962812, | |
| "grad_norm": 7.493305387455565, | |
| "learning_rate": 4.9984010137546475e-06, | |
| "loss": 2.1499, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.17308992562542258, | |
| "grad_norm": 7.399157248009571, | |
| "learning_rate": 4.997823685842875e-06, | |
| "loss": 2.1654, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.17849898580121704, | |
| "grad_norm": 7.095225049901362, | |
| "learning_rate": 4.997157593492974e-06, | |
| "loss": 2.2062, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1839080459770115, | |
| "grad_norm": 6.544205100152826, | |
| "learning_rate": 4.996402760375676e-06, | |
| "loss": 2.1302, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.18931710615280595, | |
| "grad_norm": 8.497617899265414, | |
| "learning_rate": 4.995559213315267e-06, | |
| "loss": 2.2452, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1947261663286004, | |
| "grad_norm": 7.711468079457363, | |
| "learning_rate": 4.9946269822886335e-06, | |
| "loss": 2.1562, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.20013522650439486, | |
| "grad_norm": 6.828572393494306, | |
| "learning_rate": 4.993606100424202e-06, | |
| "loss": 2.0838, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.20554428668018931, | |
| "grad_norm": 7.829101006960679, | |
| "learning_rate": 4.992496604000756e-06, | |
| "loss": 2.1608, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.21095334685598377, | |
| "grad_norm": 7.0790174546853954, | |
| "learning_rate": 4.991298532446149e-06, | |
| "loss": 2.1385, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.21636240703177823, | |
| "grad_norm": 6.4583828883763825, | |
| "learning_rate": 4.9900119283359025e-06, | |
| "loss": 2.1094, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.22177146720757268, | |
| "grad_norm": 6.919640648587095, | |
| "learning_rate": 4.988636837391696e-06, | |
| "loss": 2.2184, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.22718052738336714, | |
| "grad_norm": 7.054119878110095, | |
| "learning_rate": 4.987173308479738e-06, | |
| "loss": 2.1584, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2325895875591616, | |
| "grad_norm": 8.011842535646375, | |
| "learning_rate": 4.985621393609032e-06, | |
| "loss": 2.2022, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.23799864773495605, | |
| "grad_norm": 7.872258963246048, | |
| "learning_rate": 4.98398114792953e-06, | |
| "loss": 2.0779, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2434077079107505, | |
| "grad_norm": 7.5559333876351085, | |
| "learning_rate": 4.982252629730167e-06, | |
| "loss": 2.0195, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.24881676808654496, | |
| "grad_norm": 7.7955956497680985, | |
| "learning_rate": 4.980435900436793e-06, | |
| "loss": 2.0475, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.24881676808654496, | |
| "eval_loss": 2.1143910884857178, | |
| "eval_runtime": 80.1839, | |
| "eval_samples_per_second": 15.539, | |
| "eval_steps_per_second": 1.946, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.2542258282623394, | |
| "grad_norm": 6.802229063375309, | |
| "learning_rate": 4.978531024609994e-06, | |
| "loss": 2.0305, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.25963488843813387, | |
| "grad_norm": 7.592805462348185, | |
| "learning_rate": 4.9765380699427905e-06, | |
| "loss": 2.0484, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.2650439486139283, | |
| "grad_norm": 7.243742353954717, | |
| "learning_rate": 4.9744571072582365e-06, | |
| "loss": 2.1142, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2704530087897228, | |
| "grad_norm": 7.924276414192629, | |
| "learning_rate": 4.972288210506902e-06, | |
| "loss": 2.0623, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 7.407311330367973, | |
| "learning_rate": 4.970031456764242e-06, | |
| "loss": 2.032, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.2812711291413117, | |
| "grad_norm": 7.756849919749466, | |
| "learning_rate": 4.967686926227862e-06, | |
| "loss": 1.9791, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.28668018931710615, | |
| "grad_norm": 6.01269466571588, | |
| "learning_rate": 4.965254702214668e-06, | |
| "loss": 2.0046, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.2920892494929006, | |
| "grad_norm": 6.9870832241271685, | |
| "learning_rate": 4.9627348711578996e-06, | |
| "loss": 2.1095, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.29749830966869506, | |
| "grad_norm": 9.083794116875618, | |
| "learning_rate": 4.960127522604065e-06, | |
| "loss": 2.0736, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3029073698444895, | |
| "grad_norm": 7.096143715762253, | |
| "learning_rate": 4.957432749209755e-06, | |
| "loss": 1.9896, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.30831643002028397, | |
| "grad_norm": 5.7983259418594715, | |
| "learning_rate": 4.954650646738354e-06, | |
| "loss": 1.944, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3137254901960784, | |
| "grad_norm": 6.984280635078145, | |
| "learning_rate": 4.951781314056633e-06, | |
| "loss": 2.0276, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3191345503718729, | |
| "grad_norm": 6.88721515440217, | |
| "learning_rate": 4.948824853131237e-06, | |
| "loss": 2.0442, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.32454361054766734, | |
| "grad_norm": 7.821091549898941, | |
| "learning_rate": 4.9457813690250635e-06, | |
| "loss": 2.0072, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3299526707234618, | |
| "grad_norm": 5.64425372355818, | |
| "learning_rate": 4.942650969893527e-06, | |
| "loss": 1.994, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.33536173089925625, | |
| "grad_norm": 6.2844987961209675, | |
| "learning_rate": 4.939433766980717e-06, | |
| "loss": 2.0615, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.3407707910750507, | |
| "grad_norm": 7.116345411158451, | |
| "learning_rate": 4.936129874615443e-06, | |
| "loss": 2.048, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.34617985125084516, | |
| "grad_norm": 6.7059981437853375, | |
| "learning_rate": 4.932739410207172e-06, | |
| "loss": 1.899, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.3515889114266396, | |
| "grad_norm": 6.681993008670398, | |
| "learning_rate": 4.929262494241859e-06, | |
| "loss": 1.9566, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.35699797160243407, | |
| "grad_norm": 6.496917926567058, | |
| "learning_rate": 4.9256992502776605e-06, | |
| "loss": 1.956, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.36240703177822853, | |
| "grad_norm": 6.708931079572326, | |
| "learning_rate": 4.922049804940546e-06, | |
| "loss": 1.9967, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.367816091954023, | |
| "grad_norm": 5.620040634218876, | |
| "learning_rate": 4.9183142879198e-06, | |
| "loss": 1.916, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.37322515212981744, | |
| "grad_norm": 7.850717902889795, | |
| "learning_rate": 4.914492831963411e-06, | |
| "loss": 1.941, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3786342123056119, | |
| "grad_norm": 7.497086645607299, | |
| "learning_rate": 4.910585572873355e-06, | |
| "loss": 1.9286, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.38404327248140635, | |
| "grad_norm": 10.987140598022854, | |
| "learning_rate": 4.906592649500767e-06, | |
| "loss": 1.8511, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.3894523326572008, | |
| "grad_norm": 6.780385991870079, | |
| "learning_rate": 4.902514203741013e-06, | |
| "loss": 1.9953, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.39486139283299526, | |
| "grad_norm": 6.69524986070073, | |
| "learning_rate": 4.898350380528638e-06, | |
| "loss": 1.9524, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4002704530087897, | |
| "grad_norm": 6.1699263921182474, | |
| "learning_rate": 4.894101327832225e-06, | |
| "loss": 1.8258, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4056795131845842, | |
| "grad_norm": 6.732182374500395, | |
| "learning_rate": 4.8897671966491315e-06, | |
| "loss": 1.9725, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.41108857336037863, | |
| "grad_norm": 6.598471225386019, | |
| "learning_rate": 4.8853481410001225e-06, | |
| "loss": 1.8993, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.4164976335361731, | |
| "grad_norm": 6.93174108718265, | |
| "learning_rate": 4.8808443179239025e-06, | |
| "loss": 1.9447, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.42190669371196754, | |
| "grad_norm": 6.726165747858891, | |
| "learning_rate": 4.87625588747153e-06, | |
| "loss": 1.9723, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.427315753887762, | |
| "grad_norm": 5.798399986576307, | |
| "learning_rate": 4.87158301270073e-06, | |
| "loss": 1.9677, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.43272481406355645, | |
| "grad_norm": 6.420619711799955, | |
| "learning_rate": 4.8668258596701035e-06, | |
| "loss": 1.9731, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4381338742393509, | |
| "grad_norm": 8.374770038931102, | |
| "learning_rate": 4.861984597433223e-06, | |
| "loss": 1.9635, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.44354293441514536, | |
| "grad_norm": 6.2592251917152595, | |
| "learning_rate": 4.857059398032622e-06, | |
| "loss": 1.9228, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4489519945909398, | |
| "grad_norm": 6.759714436705178, | |
| "learning_rate": 4.85205043649369e-06, | |
| "loss": 1.9456, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4543610547667343, | |
| "grad_norm": 7.0963936709641615, | |
| "learning_rate": 4.846957890818444e-06, | |
| "loss": 1.9217, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.45977011494252873, | |
| "grad_norm": 6.941342449698633, | |
| "learning_rate": 4.841781941979207e-06, | |
| "loss": 1.7469, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.4651791751183232, | |
| "grad_norm": 7.445994149780175, | |
| "learning_rate": 4.836522773912178e-06, | |
| "loss": 1.9254, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 6.588581579766995, | |
| "learning_rate": 4.83118057351089e-06, | |
| "loss": 1.8967, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.4759972954699121, | |
| "grad_norm": 7.117087421099954, | |
| "learning_rate": 4.825755530619576e-06, | |
| "loss": 1.8865, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.48140635564570655, | |
| "grad_norm": 6.724049881659424, | |
| "learning_rate": 4.820247838026414e-06, | |
| "loss": 1.9435, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.486815415821501, | |
| "grad_norm": 5.6973534950025595, | |
| "learning_rate": 4.814657691456685e-06, | |
| "loss": 1.7882, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.49222447599729546, | |
| "grad_norm": 6.388481680244839, | |
| "learning_rate": 4.808985289565813e-06, | |
| "loss": 1.8173, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.4976335361730899, | |
| "grad_norm": 6.3977034813469835, | |
| "learning_rate": 4.803230833932302e-06, | |
| "loss": 1.8489, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.4976335361730899, | |
| "eval_loss": 1.89235520362854, | |
| "eval_runtime": 80.2557, | |
| "eval_samples_per_second": 15.525, | |
| "eval_steps_per_second": 1.944, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5030425963488844, | |
| "grad_norm": 7.046151003888674, | |
| "learning_rate": 4.797394529050577e-06, | |
| "loss": 1.9349, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5084516565246788, | |
| "grad_norm": 6.590771586445206, | |
| "learning_rate": 4.791476582323719e-06, | |
| "loss": 1.9962, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5138607167004733, | |
| "grad_norm": 7.166866163450392, | |
| "learning_rate": 4.785477204056089e-06, | |
| "loss": 1.8174, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5192697768762677, | |
| "grad_norm": 6.716770374519456, | |
| "learning_rate": 4.779396607445858e-06, | |
| "loss": 1.9214, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5246788370520622, | |
| "grad_norm": 7.244507141710972, | |
| "learning_rate": 4.77323500857743e-06, | |
| "loss": 1.8681, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.5300878972278567, | |
| "grad_norm": 6.3217358598893165, | |
| "learning_rate": 4.7669926264137625e-06, | |
| "loss": 1.8825, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5354969574036511, | |
| "grad_norm": 5.77085923138114, | |
| "learning_rate": 4.760669682788584e-06, | |
| "loss": 1.7838, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5409060175794456, | |
| "grad_norm": 5.638989379158351, | |
| "learning_rate": 4.754266402398517e-06, | |
| "loss": 1.7818, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.54631507775524, | |
| "grad_norm": 6.283880096623512, | |
| "learning_rate": 4.747783012795083e-06, | |
| "loss": 1.7555, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 7.08060002757244, | |
| "learning_rate": 4.741219744376624e-06, | |
| "loss": 1.8585, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5571331981068289, | |
| "grad_norm": 7.077388397562155, | |
| "learning_rate": 4.734576830380113e-06, | |
| "loss": 1.7837, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5625422582826234, | |
| "grad_norm": 6.093542681362889, | |
| "learning_rate": 4.727854506872863e-06, | |
| "loss": 1.7841, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5679513184584178, | |
| "grad_norm": 6.250665388589361, | |
| "learning_rate": 4.721053012744142e-06, | |
| "loss": 1.8065, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5733603786342123, | |
| "grad_norm": 7.137442110609306, | |
| "learning_rate": 4.71417258969668e-06, | |
| "loss": 1.8022, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.5787694388100068, | |
| "grad_norm": 6.025245815328945, | |
| "learning_rate": 4.70721348223808e-06, | |
| "loss": 1.8129, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.5841784989858012, | |
| "grad_norm": 6.920748676536336, | |
| "learning_rate": 4.700175937672134e-06, | |
| "loss": 1.8498, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.5895875591615957, | |
| "grad_norm": 7.6481497850706415, | |
| "learning_rate": 4.693060206090028e-06, | |
| "loss": 1.7777, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.5949966193373901, | |
| "grad_norm": 7.666186710495507, | |
| "learning_rate": 4.685866540361456e-06, | |
| "loss": 1.834, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6004056795131846, | |
| "grad_norm": 7.021331030038572, | |
| "learning_rate": 4.678595196125638e-06, | |
| "loss": 1.8563, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.605814739688979, | |
| "grad_norm": 6.487310124057107, | |
| "learning_rate": 4.671246431782234e-06, | |
| "loss": 1.7152, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6112237998647735, | |
| "grad_norm": 6.492879712741564, | |
| "learning_rate": 4.6638205084821544e-06, | |
| "loss": 1.7396, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6166328600405679, | |
| "grad_norm": 6.932406364697437, | |
| "learning_rate": 4.656317690118291e-06, | |
| "loss": 1.6753, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6220419202163624, | |
| "grad_norm": 6.635109301432599, | |
| "learning_rate": 4.648738243316128e-06, | |
| "loss": 1.7832, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6274509803921569, | |
| "grad_norm": 6.6861559036147975, | |
| "learning_rate": 4.641082437424277e-06, | |
| "loss": 1.7703, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6328600405679513, | |
| "grad_norm": 6.9646373816737075, | |
| "learning_rate": 4.633350544504899e-06, | |
| "loss": 1.7764, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6382691007437458, | |
| "grad_norm": 6.144269305268074, | |
| "learning_rate": 4.625542839324036e-06, | |
| "loss": 1.8115, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.6436781609195402, | |
| "grad_norm": 6.489513509921581, | |
| "learning_rate": 4.617659599341849e-06, | |
| "loss": 1.7991, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6490872210953347, | |
| "grad_norm": 6.883770839388401, | |
| "learning_rate": 4.609701104702759e-06, | |
| "loss": 1.6693, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6544962812711291, | |
| "grad_norm": 7.5470889656860605, | |
| "learning_rate": 4.6016676382254895e-06, | |
| "loss": 1.7072, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6599053414469236, | |
| "grad_norm": 7.546505121169261, | |
| "learning_rate": 4.593559485393015e-06, | |
| "loss": 1.7841, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.665314401622718, | |
| "grad_norm": 7.028634956829483, | |
| "learning_rate": 4.585376934342418e-06, | |
| "loss": 1.8216, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6707234617985125, | |
| "grad_norm": 6.173100762882487, | |
| "learning_rate": 4.577120275854649e-06, | |
| "loss": 1.7126, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.676132521974307, | |
| "grad_norm": 7.3755207955713, | |
| "learning_rate": 4.568789803344196e-06, | |
| "loss": 1.6914, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.6815415821501014, | |
| "grad_norm": 7.110847605538237, | |
| "learning_rate": 4.56038581284865e-06, | |
| "loss": 1.7385, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.6869506423258959, | |
| "grad_norm": 6.361787130642339, | |
| "learning_rate": 4.551908603018191e-06, | |
| "loss": 1.6135, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.6923597025016903, | |
| "grad_norm": 6.445020674626523, | |
| "learning_rate": 4.543358475104975e-06, | |
| "loss": 1.709, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.6977687626774848, | |
| "grad_norm": 9.06412704312379, | |
| "learning_rate": 4.5347357329524254e-06, | |
| "loss": 1.846, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7031778228532792, | |
| "grad_norm": 6.7066499384595675, | |
| "learning_rate": 4.5260406829844364e-06, | |
| "loss": 1.7074, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7085868830290737, | |
| "grad_norm": 5.861118356970615, | |
| "learning_rate": 4.5172736341944845e-06, | |
| "loss": 1.7179, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7139959432048681, | |
| "grad_norm": 7.102544513000641, | |
| "learning_rate": 4.5084348981346495e-06, | |
| "loss": 1.7577, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.7194050033806626, | |
| "grad_norm": 10.027416022472378, | |
| "learning_rate": 4.499524788904537e-06, | |
| "loss": 1.6513, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7248140635564571, | |
| "grad_norm": 6.648158309890374, | |
| "learning_rate": 4.490543623140123e-06, | |
| "loss": 1.6589, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.7302231237322515, | |
| "grad_norm": 6.662339022641756, | |
| "learning_rate": 4.481491720002499e-06, | |
| "loss": 1.6705, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.735632183908046, | |
| "grad_norm": 6.0555935222186665, | |
| "learning_rate": 4.472369401166531e-06, | |
| "loss": 1.7424, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7410412440838404, | |
| "grad_norm": 8.895889108381375, | |
| "learning_rate": 4.463176990809423e-06, | |
| "loss": 1.7386, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.7464503042596349, | |
| "grad_norm": 7.003045853961112, | |
| "learning_rate": 4.453914815599206e-06, | |
| "loss": 1.7036, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7464503042596349, | |
| "eval_loss": 1.716886281967163, | |
| "eval_runtime": 80.2672, | |
| "eval_samples_per_second": 15.523, | |
| "eval_steps_per_second": 1.944, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7518593644354293, | |
| "grad_norm": 6.693085585003622, | |
| "learning_rate": 4.444583204683123e-06, | |
| "loss": 1.7129, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7572684246112238, | |
| "grad_norm": 7.502871529343553, | |
| "learning_rate": 4.435182489675931e-06, | |
| "loss": 1.6455, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7626774847870182, | |
| "grad_norm": 6.062617524551209, | |
| "learning_rate": 4.425713004648123e-06, | |
| "loss": 1.5962, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7680865449628127, | |
| "grad_norm": 7.073651768871659, | |
| "learning_rate": 4.416175086114049e-06, | |
| "loss": 1.6784, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.7734956051386072, | |
| "grad_norm": 5.885299008267407, | |
| "learning_rate": 4.406569073019965e-06, | |
| "loss": 1.7525, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.7789046653144016, | |
| "grad_norm": 6.313009716756996, | |
| "learning_rate": 4.396895306731978e-06, | |
| "loss": 1.7347, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "grad_norm": 7.009990810809474, | |
| "learning_rate": 4.387154131023924e-06, | |
| "loss": 1.691, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7897227856659905, | |
| "grad_norm": 6.900733353908871, | |
| "learning_rate": 4.377345892065149e-06, | |
| "loss": 1.6789, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.795131845841785, | |
| "grad_norm": 6.705414309211519, | |
| "learning_rate": 4.367470938408204e-06, | |
| "loss": 1.6024, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.8005409060175794, | |
| "grad_norm": 5.746533610111662, | |
| "learning_rate": 4.357529620976463e-06, | |
| "loss": 1.5715, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.8059499661933739, | |
| "grad_norm": 6.344927493402156, | |
| "learning_rate": 4.3475222930516484e-06, | |
| "loss": 1.6393, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8113590263691683, | |
| "grad_norm": 7.177833196869734, | |
| "learning_rate": 4.337449310261279e-06, | |
| "loss": 1.7165, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8167680865449628, | |
| "grad_norm": 7.330166717470119, | |
| "learning_rate": 4.327311030566033e-06, | |
| "loss": 1.5549, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8221771467207573, | |
| "grad_norm": 6.615833346853653, | |
| "learning_rate": 4.317107814247022e-06, | |
| "loss": 1.5402, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 6.435680171929204, | |
| "learning_rate": 4.306840023892998e-06, | |
| "loss": 1.6594, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.8329952670723462, | |
| "grad_norm": 8.276922490588758, | |
| "learning_rate": 4.2965080243874555e-06, | |
| "loss": 1.7175, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.8384043272481406, | |
| "grad_norm": 6.944104925297799, | |
| "learning_rate": 4.2861121828956745e-06, | |
| "loss": 1.6139, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.8438133874239351, | |
| "grad_norm": 6.95913091453491, | |
| "learning_rate": 4.275652868851669e-06, | |
| "loss": 1.6509, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8492224475997295, | |
| "grad_norm": 6.520552325908588, | |
| "learning_rate": 4.265130453945056e-06, | |
| "loss": 1.6153, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.854631507775524, | |
| "grad_norm": 5.793083719201756, | |
| "learning_rate": 4.254545312107854e-06, | |
| "loss": 1.6652, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.8600405679513184, | |
| "grad_norm": 7.087661599168472, | |
| "learning_rate": 4.243897819501187e-06, | |
| "loss": 1.5381, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.8654496281271129, | |
| "grad_norm": 7.094455948395506, | |
| "learning_rate": 4.233188354501921e-06, | |
| "loss": 1.5494, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8708586883029074, | |
| "grad_norm": 8.203152015742518, | |
| "learning_rate": 4.222417297689217e-06, | |
| "loss": 1.6491, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.8762677484787018, | |
| "grad_norm": 6.453682013406382, | |
| "learning_rate": 4.211585031831007e-06, | |
| "loss": 1.6017, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.8816768086544963, | |
| "grad_norm": 5.936146356941861, | |
| "learning_rate": 4.200691941870392e-06, | |
| "loss": 1.578, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.8870858688302907, | |
| "grad_norm": 6.939707441726559, | |
| "learning_rate": 4.189738414911959e-06, | |
| "loss": 1.6267, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.8924949290060852, | |
| "grad_norm": 5.410313740063528, | |
| "learning_rate": 4.178724840208029e-06, | |
| "loss": 1.4958, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.8979039891818796, | |
| "grad_norm": 6.696136302418629, | |
| "learning_rate": 4.167651609144822e-06, | |
| "loss": 1.5497, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.9033130493576741, | |
| "grad_norm": 7.906312937190654, | |
| "learning_rate": 4.15651911522855e-06, | |
| "loss": 1.5839, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.9087221095334685, | |
| "grad_norm": 6.535495007076388, | |
| "learning_rate": 4.145327754071427e-06, | |
| "loss": 1.5472, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.914131169709263, | |
| "grad_norm": 5.836290315234971, | |
| "learning_rate": 4.134077923377622e-06, | |
| "loss": 1.552, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.9195402298850575, | |
| "grad_norm": 6.8226833628498476, | |
| "learning_rate": 4.122770022929114e-06, | |
| "loss": 1.5789, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9249492900608519, | |
| "grad_norm": 6.636708734597821, | |
| "learning_rate": 4.1114044545714935e-06, | |
| "loss": 1.5297, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9303583502366464, | |
| "grad_norm": 6.480166160413371, | |
| "learning_rate": 4.0999816221996755e-06, | |
| "loss": 1.64, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9357674104124408, | |
| "grad_norm": 6.305918232446464, | |
| "learning_rate": 4.088501931743551e-06, | |
| "loss": 1.6244, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 6.244739864099848, | |
| "learning_rate": 4.076965791153562e-06, | |
| "loss": 1.4548, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.9465855307640297, | |
| "grad_norm": 6.652315653413243, | |
| "learning_rate": 4.065373610386201e-06, | |
| "loss": 1.6279, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.9519945909398242, | |
| "grad_norm": 7.0369918458505, | |
| "learning_rate": 4.0537258013894434e-06, | |
| "loss": 1.5423, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.9574036511156186, | |
| "grad_norm": 6.219650042475582, | |
| "learning_rate": 4.042022778088111e-06, | |
| "loss": 1.6608, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.9628127112914131, | |
| "grad_norm": 6.876925788244563, | |
| "learning_rate": 4.030264956369158e-06, | |
| "loss": 1.5072, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.9682217714672076, | |
| "grad_norm": 6.90949169776309, | |
| "learning_rate": 4.018452754066895e-06, | |
| "loss": 1.6312, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.973630831643002, | |
| "grad_norm": 6.8575340845851604, | |
| "learning_rate": 4.006586590948141e-06, | |
| "loss": 1.5603, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.9790398918187965, | |
| "grad_norm": 6.399141972511512, | |
| "learning_rate": 3.994666888697304e-06, | |
| "loss": 1.4676, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.9844489519945909, | |
| "grad_norm": 6.8804715530154255, | |
| "learning_rate": 3.982694070901396e-06, | |
| "loss": 1.5238, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.9898580121703854, | |
| "grad_norm": 6.143146424939786, | |
| "learning_rate": 3.970668563034982e-06, | |
| "loss": 1.5458, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.9952670723461798, | |
| "grad_norm": 7.1826166940238005, | |
| "learning_rate": 3.958590792445057e-06, | |
| "loss": 1.5179, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.9952670723461798, | |
| "eval_loss": 1.555787205696106, | |
| "eval_runtime": 80.2289, | |
| "eval_samples_per_second": 15.531, | |
| "eval_steps_per_second": 1.944, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.0006761325219744, | |
| "grad_norm": 6.359754610852221, | |
| "learning_rate": 3.946461188335863e-06, | |
| "loss": 1.4166, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.0060851926977687, | |
| "grad_norm": 6.558163414784584, | |
| "learning_rate": 3.934280181753634e-06, | |
| "loss": 1.2751, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.0114942528735633, | |
| "grad_norm": 7.219629642258684, | |
| "learning_rate": 3.922048205571279e-06, | |
| "loss": 1.1886, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.0169033130493577, | |
| "grad_norm": 5.641940967289977, | |
| "learning_rate": 3.909765694473e-06, | |
| "loss": 1.162, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.0223123732251522, | |
| "grad_norm": 5.146358119235958, | |
| "learning_rate": 3.897433084938841e-06, | |
| "loss": 1.1985, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.0277214334009466, | |
| "grad_norm": 6.030205789253815, | |
| "learning_rate": 3.885050815229182e-06, | |
| "loss": 1.2081, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0331304935767411, | |
| "grad_norm": 7.022365747053938, | |
| "learning_rate": 3.872619325369162e-06, | |
| "loss": 1.2893, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.0385395537525355, | |
| "grad_norm": 5.333726155190956, | |
| "learning_rate": 3.860139057133042e-06, | |
| "loss": 1.0908, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.04394861392833, | |
| "grad_norm": 5.45665767931003, | |
| "learning_rate": 3.8476104540285054e-06, | |
| "loss": 1.1623, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.0493576741041244, | |
| "grad_norm": 5.712839244673719, | |
| "learning_rate": 3.835033961280898e-06, | |
| "loss": 1.2006, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.054766734279919, | |
| "grad_norm": 5.034751006183729, | |
| "learning_rate": 3.8224100258174066e-06, | |
| "loss": 1.1717, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.0601757944557133, | |
| "grad_norm": 5.683598056495194, | |
| "learning_rate": 3.809739096251176e-06, | |
| "loss": 1.0888, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.0655848546315079, | |
| "grad_norm": 5.519253752312427, | |
| "learning_rate": 3.7970216228653667e-06, | |
| "loss": 1.1504, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.0709939148073022, | |
| "grad_norm": 5.447708800800649, | |
| "learning_rate": 3.7842580575971533e-06, | |
| "loss": 1.1493, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.0764029749830968, | |
| "grad_norm": 5.433128433416691, | |
| "learning_rate": 3.7714488540216637e-06, | |
| "loss": 1.2068, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.0818120351588911, | |
| "grad_norm": 5.161001564479917, | |
| "learning_rate": 3.7585944673358632e-06, | |
| "loss": 1.061, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0872210953346857, | |
| "grad_norm": 6.0824539385903895, | |
| "learning_rate": 3.745695354342374e-06, | |
| "loss": 1.0569, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.09263015551048, | |
| "grad_norm": 5.834270627588388, | |
| "learning_rate": 3.7327519734332453e-06, | |
| "loss": 1.1536, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.0980392156862746, | |
| "grad_norm": 5.775803225686334, | |
| "learning_rate": 3.7197647845736616e-06, | |
| "loss": 1.0768, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.103448275862069, | |
| "grad_norm": 5.875985815161611, | |
| "learning_rate": 3.7067342492855997e-06, | |
| "loss": 1.0848, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.1088573360378635, | |
| "grad_norm": 5.252887507709333, | |
| "learning_rate": 3.6936608306314227e-06, | |
| "loss": 1.0704, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.1142663962136579, | |
| "grad_norm": 5.6931233725933375, | |
| "learning_rate": 3.6805449931974313e-06, | |
| "loss": 1.0765, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.1196754563894524, | |
| "grad_norm": 5.48726034244755, | |
| "learning_rate": 3.6673872030773473e-06, | |
| "loss": 1.1818, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.1250845165652468, | |
| "grad_norm": 6.429761402145819, | |
| "learning_rate": 3.654187927855754e-06, | |
| "loss": 1.0093, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.1304935767410413, | |
| "grad_norm": 6.370439411738345, | |
| "learning_rate": 3.6409476365914786e-06, | |
| "loss": 1.1248, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.1359026369168357, | |
| "grad_norm": 5.059354487980713, | |
| "learning_rate": 3.6276667998009242e-06, | |
| "loss": 1.1993, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.1413116970926303, | |
| "grad_norm": 5.240968702755266, | |
| "learning_rate": 3.6143458894413463e-06, | |
| "loss": 1.1065, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.1467207572684246, | |
| "grad_norm": 6.6242597409867425, | |
| "learning_rate": 3.600985378894086e-06, | |
| "loss": 1.1479, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.1521298174442192, | |
| "grad_norm": 5.568795571449795, | |
| "learning_rate": 3.5875857429477447e-06, | |
| "loss": 1.1048, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.1575388776200135, | |
| "grad_norm": 5.421352959165325, | |
| "learning_rate": 3.5741474577813086e-06, | |
| "loss": 1.1126, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.162947937795808, | |
| "grad_norm": 4.977758792843085, | |
| "learning_rate": 3.5606710009472335e-06, | |
| "loss": 1.0064, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.1683569979716024, | |
| "grad_norm": 5.911407929293024, | |
| "learning_rate": 3.54715685135447e-06, | |
| "loss": 1.1859, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.173766058147397, | |
| "grad_norm": 5.557031565026226, | |
| "learning_rate": 3.5336054892514437e-06, | |
| "loss": 1.1526, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.1791751183231913, | |
| "grad_norm": 5.900696316987968, | |
| "learning_rate": 3.520017396208993e-06, | |
| "loss": 1.0158, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.184584178498986, | |
| "grad_norm": 5.287116454496747, | |
| "learning_rate": 3.5063930551032494e-06, | |
| "loss": 1.0677, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.1899932386747802, | |
| "grad_norm": 5.106236012898038, | |
| "learning_rate": 3.4927329500984857e-06, | |
| "loss": 1.0827, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.1954022988505748, | |
| "grad_norm": 4.8246900511041595, | |
| "learning_rate": 3.4790375666299026e-06, | |
| "loss": 1.1788, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.2008113590263692, | |
| "grad_norm": 5.545181164784277, | |
| "learning_rate": 3.465307391386383e-06, | |
| "loss": 1.1896, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.2062204192021637, | |
| "grad_norm": 5.427546334067872, | |
| "learning_rate": 3.4515429122931955e-06, | |
| "loss": 1.0658, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.211629479377958, | |
| "grad_norm": 5.245977350973333, | |
| "learning_rate": 3.437744618494653e-06, | |
| "loss": 1.1658, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.2170385395537526, | |
| "grad_norm": 5.043336556534448, | |
| "learning_rate": 3.423913000336732e-06, | |
| "loss": 1.098, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.222447599729547, | |
| "grad_norm": 5.579610380422948, | |
| "learning_rate": 3.41004854934965e-06, | |
| "loss": 1.0795, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.2278566599053415, | |
| "grad_norm": 6.096777208300464, | |
| "learning_rate": 3.3961517582303916e-06, | |
| "loss": 1.0956, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.2332657200811359, | |
| "grad_norm": 5.2061771582860645, | |
| "learning_rate": 3.3822231208252053e-06, | |
| "loss": 1.1306, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.2386747802569305, | |
| "grad_norm": 5.182819955042982, | |
| "learning_rate": 3.3682631321120507e-06, | |
| "loss": 1.1009, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.2440838404327248, | |
| "grad_norm": 5.625028453640487, | |
| "learning_rate": 3.354272288183012e-06, | |
| "loss": 1.1819, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.2440838404327248, | |
| "eval_loss": 1.463145136833191, | |
| "eval_runtime": 80.1754, | |
| "eval_samples_per_second": 15.541, | |
| "eval_steps_per_second": 1.946, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.2494929006085194, | |
| "grad_norm": 5.590526458310239, | |
| "learning_rate": 3.340251086226663e-06, | |
| "loss": 0.9821, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.2549019607843137, | |
| "grad_norm": 11.81361368808831, | |
| "learning_rate": 3.326200024510405e-06, | |
| "loss": 1.1092, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.260311020960108, | |
| "grad_norm": 5.093824480458287, | |
| "learning_rate": 3.3121196023627543e-06, | |
| "loss": 1.1585, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.2657200811359026, | |
| "grad_norm": 5.166324447334581, | |
| "learning_rate": 3.2980103201556023e-06, | |
| "loss": 1.157, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.2711291413116972, | |
| "grad_norm": 5.637337054230702, | |
| "learning_rate": 3.2838726792864315e-06, | |
| "loss": 1.166, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.2765382014874915, | |
| "grad_norm": 5.6363361298308545, | |
| "learning_rate": 3.2697071821604986e-06, | |
| "loss": 1.0634, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.2819472616632859, | |
| "grad_norm": 5.96879436061098, | |
| "learning_rate": 3.255514332172979e-06, | |
| "loss": 1.0877, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.2873563218390804, | |
| "grad_norm": 5.298279579130303, | |
| "learning_rate": 3.2412946336910778e-06, | |
| "loss": 1.1096, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.292765382014875, | |
| "grad_norm": 5.24661665042464, | |
| "learning_rate": 3.2270485920361093e-06, | |
| "loss": 1.0553, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.2981744421906694, | |
| "grad_norm": 5.2866147649117305, | |
| "learning_rate": 3.2127767134655374e-06, | |
| "loss": 1.0457, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.3035835023664637, | |
| "grad_norm": 4.959907157136451, | |
| "learning_rate": 3.198479505154984e-06, | |
| "loss": 1.1399, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.3089925625422583, | |
| "grad_norm": 5.238554311773635, | |
| "learning_rate": 3.184157475180208e-06, | |
| "loss": 1.1271, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.3144016227180528, | |
| "grad_norm": 6.098173197957714, | |
| "learning_rate": 3.1698111324990454e-06, | |
| "loss": 1.0782, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.3198106828938472, | |
| "grad_norm": 11.039229160599406, | |
| "learning_rate": 3.15544098693333e-06, | |
| "loss": 1.05, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.3252197430696415, | |
| "grad_norm": 5.499488969453295, | |
| "learning_rate": 3.14104754915077e-06, | |
| "loss": 1.0368, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.330628803245436, | |
| "grad_norm": 5.138414489753493, | |
| "learning_rate": 3.1266313306468018e-06, | |
| "loss": 1.0835, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.3360378634212307, | |
| "grad_norm": 5.106872795478693, | |
| "learning_rate": 3.1121928437264138e-06, | |
| "loss": 0.9708, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.341446923597025, | |
| "grad_norm": 5.002564538923287, | |
| "learning_rate": 3.0977326014859415e-06, | |
| "loss": 1.0687, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.3468559837728193, | |
| "grad_norm": 5.387403166077266, | |
| "learning_rate": 3.0832511177948326e-06, | |
| "loss": 1.1496, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.352265043948614, | |
| "grad_norm": 8.271273394288187, | |
| "learning_rate": 3.0687489072773864e-06, | |
| "loss": 1.0601, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.3576741041244085, | |
| "grad_norm": 5.909473723884275, | |
| "learning_rate": 3.0542264852944635e-06, | |
| "loss": 1.1829, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.3630831643002028, | |
| "grad_norm": 5.390843669805894, | |
| "learning_rate": 3.0396843679251777e-06, | |
| "loss": 1.1672, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.3684922244759972, | |
| "grad_norm": 4.9990577854234095, | |
| "learning_rate": 3.0251230719485465e-06, | |
| "loss": 1.0904, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.3739012846517917, | |
| "grad_norm": 5.2939324887679495, | |
| "learning_rate": 3.0105431148251364e-06, | |
| "loss": 1.0716, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 5.423330025150549, | |
| "learning_rate": 2.9959450146786674e-06, | |
| "loss": 1.1325, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.3847194050033806, | |
| "grad_norm": 5.023381297619562, | |
| "learning_rate": 2.981329290277605e-06, | |
| "loss": 0.987, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.390128465179175, | |
| "grad_norm": 5.282148760283777, | |
| "learning_rate": 2.966696461016721e-06, | |
| "loss": 1.0894, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.3955375253549696, | |
| "grad_norm": 5.15061299365383, | |
| "learning_rate": 2.952047046898637e-06, | |
| "loss": 1.1275, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.4009465855307641, | |
| "grad_norm": 5.6780949048605285, | |
| "learning_rate": 2.9373815685153485e-06, | |
| "loss": 0.9549, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.4063556457065585, | |
| "grad_norm": 5.251474491048633, | |
| "learning_rate": 2.9227005470297194e-06, | |
| "loss": 1.0549, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 5.418474582825882, | |
| "learning_rate": 2.9080045041569647e-06, | |
| "loss": 1.1761, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.4171737660581474, | |
| "grad_norm": 5.363207500979156, | |
| "learning_rate": 2.893293962146114e-06, | |
| "loss": 1.132, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.422582826233942, | |
| "grad_norm": 6.178387649415905, | |
| "learning_rate": 2.878569443761442e-06, | |
| "loss": 1.0214, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.4279918864097363, | |
| "grad_norm": 5.718107503992589, | |
| "learning_rate": 2.863831472263904e-06, | |
| "loss": 1.0028, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.4334009465855306, | |
| "grad_norm": 5.279016485794886, | |
| "learning_rate": 2.8490805713925298e-06, | |
| "loss": 1.0827, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.4388100067613252, | |
| "grad_norm": 5.087811977719158, | |
| "learning_rate": 2.8343172653458194e-06, | |
| "loss": 1.0937, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.4442190669371198, | |
| "grad_norm": 5.4070064172712184, | |
| "learning_rate": 2.8195420787631113e-06, | |
| "loss": 1.123, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.4496281271129141, | |
| "grad_norm": 5.587801015021142, | |
| "learning_rate": 2.8047555367059404e-06, | |
| "loss": 1.0621, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.4550371872887085, | |
| "grad_norm": 5.552901382895493, | |
| "learning_rate": 2.7899581646393746e-06, | |
| "loss": 0.9631, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.460446247464503, | |
| "grad_norm": 5.4717166148353185, | |
| "learning_rate": 2.7751504884133484e-06, | |
| "loss": 1.0253, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.4658553076402976, | |
| "grad_norm": 5.3779298143381356, | |
| "learning_rate": 2.7603330342439686e-06, | |
| "loss": 0.9938, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.471264367816092, | |
| "grad_norm": 6.033013211604236, | |
| "learning_rate": 2.745506328694822e-06, | |
| "loss": 1.0509, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.4766734279918863, | |
| "grad_norm": 5.832457552507952, | |
| "learning_rate": 2.730670898658255e-06, | |
| "loss": 1.1067, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.4820824881676808, | |
| "grad_norm": 5.7329943331650295, | |
| "learning_rate": 2.7158272713366573e-06, | |
| "loss": 1.0657, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.4874915483434754, | |
| "grad_norm": 5.506227442307808, | |
| "learning_rate": 2.700975974223719e-06, | |
| "loss": 1.1391, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.4929006085192698, | |
| "grad_norm": 4.978032312700087, | |
| "learning_rate": 2.6861175350856937e-06, | |
| "loss": 0.9931, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.4929006085192698, | |
| "eval_loss": 1.3746687173843384, | |
| "eval_runtime": 80.4525, | |
| "eval_samples_per_second": 15.487, | |
| "eval_steps_per_second": 1.939, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.498309668695064, | |
| "grad_norm": 4.962349655742751, | |
| "learning_rate": 2.6712524819426355e-06, | |
| "loss": 1.0006, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.5037187288708587, | |
| "grad_norm": 4.925015437741922, | |
| "learning_rate": 2.656381343049641e-06, | |
| "loss": 1.1016, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.5091277890466532, | |
| "grad_norm": 4.738818602054919, | |
| "learning_rate": 2.6415046468780726e-06, | |
| "loss": 1.0465, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.5145368492224476, | |
| "grad_norm": 4.961032669335967, | |
| "learning_rate": 2.626622922096782e-06, | |
| "loss": 0.9809, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.519945909398242, | |
| "grad_norm": 5.043527496081483, | |
| "learning_rate": 2.6117366975533187e-06, | |
| "loss": 1.0272, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.5253549695740365, | |
| "grad_norm": 5.913998311820069, | |
| "learning_rate": 2.596846502255142e-06, | |
| "loss": 1.0146, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.530764029749831, | |
| "grad_norm": 5.295260207722928, | |
| "learning_rate": 2.581952865350815e-06, | |
| "loss": 1.0956, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.5361730899256254, | |
| "grad_norm": 4.466395085273886, | |
| "learning_rate": 2.5670563161112073e-06, | |
| "loss": 1.0354, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.5415821501014197, | |
| "grad_norm": 5.399245625271399, | |
| "learning_rate": 2.5521573839106815e-06, | |
| "loss": 0.9433, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.5469912102772143, | |
| "grad_norm": 5.270037129845648, | |
| "learning_rate": 2.5372565982082843e-06, | |
| "loss": 0.9744, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.5524002704530089, | |
| "grad_norm": 5.161110168680388, | |
| "learning_rate": 2.5223544885289287e-06, | |
| "loss": 1.0077, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.5578093306288032, | |
| "grad_norm": 5.6674277481523045, | |
| "learning_rate": 2.5074515844445774e-06, | |
| "loss": 1.0805, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.5632183908045976, | |
| "grad_norm": 5.089380302532208, | |
| "learning_rate": 2.4925484155554235e-06, | |
| "loss": 0.9904, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.5686274509803921, | |
| "grad_norm": 4.6826224514211, | |
| "learning_rate": 2.477645511471073e-06, | |
| "loss": 1.0843, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.5740365111561867, | |
| "grad_norm": 5.231440812973741, | |
| "learning_rate": 2.462743401791716e-06, | |
| "loss": 1.0343, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.579445571331981, | |
| "grad_norm": 6.063260236415264, | |
| "learning_rate": 2.4478426160893197e-06, | |
| "loss": 1.0377, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.5848546315077754, | |
| "grad_norm": 4.850759349852377, | |
| "learning_rate": 2.4329436838887936e-06, | |
| "loss": 0.984, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.59026369168357, | |
| "grad_norm": 5.356535713038874, | |
| "learning_rate": 2.4180471346491864e-06, | |
| "loss": 1.0006, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.5956727518593645, | |
| "grad_norm": 5.843575763516113, | |
| "learning_rate": 2.403153497744859e-06, | |
| "loss": 1.0172, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.6010818120351589, | |
| "grad_norm": 5.383170068467428, | |
| "learning_rate": 2.3882633024466813e-06, | |
| "loss": 0.9683, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.6064908722109532, | |
| "grad_norm": 5.754163901430788, | |
| "learning_rate": 2.3733770779032185e-06, | |
| "loss": 0.9569, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.6118999323867478, | |
| "grad_norm": 5.20527028666756, | |
| "learning_rate": 2.3584953531219278e-06, | |
| "loss": 1.0422, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.6173089925625423, | |
| "grad_norm": 7.36652597560724, | |
| "learning_rate": 2.3436186569503598e-06, | |
| "loss": 1.0485, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.6227180527383367, | |
| "grad_norm": 5.265722625187695, | |
| "learning_rate": 2.3287475180573653e-06, | |
| "loss": 1.0536, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.628127112914131, | |
| "grad_norm": 5.352738983760188, | |
| "learning_rate": 2.3138824649143076e-06, | |
| "loss": 1.0608, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.6335361730899256, | |
| "grad_norm": 5.069558233311071, | |
| "learning_rate": 2.2990240257762817e-06, | |
| "loss": 0.9955, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.6389452332657202, | |
| "grad_norm": 5.3724147320331435, | |
| "learning_rate": 2.2841727286633444e-06, | |
| "loss": 1.0888, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.6443542934415145, | |
| "grad_norm": 4.963379896141614, | |
| "learning_rate": 2.269329101341745e-06, | |
| "loss": 0.98, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.6497633536173089, | |
| "grad_norm": 5.132481781947378, | |
| "learning_rate": 2.254493671305179e-06, | |
| "loss": 0.9771, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.6551724137931034, | |
| "grad_norm": 5.347290600611583, | |
| "learning_rate": 2.239666965756032e-06, | |
| "loss": 0.9705, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.660581473968898, | |
| "grad_norm": 5.722698577141815, | |
| "learning_rate": 2.224849511586652e-06, | |
| "loss": 0.9266, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.6659905341446923, | |
| "grad_norm": 5.733293031193421, | |
| "learning_rate": 2.2100418353606262e-06, | |
| "loss": 1.0429, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.6713995943204867, | |
| "grad_norm": 6.274637328200243, | |
| "learning_rate": 2.19524446329406e-06, | |
| "loss": 1.0208, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.6768086544962812, | |
| "grad_norm": 5.5296462714582955, | |
| "learning_rate": 2.180457921236889e-06, | |
| "loss": 0.9945, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.6822177146720758, | |
| "grad_norm": 5.336860694103749, | |
| "learning_rate": 2.165682734654181e-06, | |
| "loss": 0.9847, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.6876267748478702, | |
| "grad_norm": 5.016859607786664, | |
| "learning_rate": 2.150919428607472e-06, | |
| "loss": 1.0305, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.6930358350236645, | |
| "grad_norm": 4.918636259586344, | |
| "learning_rate": 2.1361685277360973e-06, | |
| "loss": 1.0611, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.698444895199459, | |
| "grad_norm": 5.520829989468747, | |
| "learning_rate": 2.1214305562385592e-06, | |
| "loss": 0.8441, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.7038539553752536, | |
| "grad_norm": 5.327879216317322, | |
| "learning_rate": 2.106706037853887e-06, | |
| "loss": 1.0225, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.709263015551048, | |
| "grad_norm": 5.3843566459704135, | |
| "learning_rate": 2.0919954958430357e-06, | |
| "loss": 1.0082, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.7146720757268423, | |
| "grad_norm": 5.395175713065742, | |
| "learning_rate": 2.077299452970282e-06, | |
| "loss": 1.023, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.720081135902637, | |
| "grad_norm": 5.887472074041824, | |
| "learning_rate": 2.062618431484652e-06, | |
| "loss": 0.9278, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.7254901960784315, | |
| "grad_norm": 6.537709956987992, | |
| "learning_rate": 2.047952953101363e-06, | |
| "loss": 1.0788, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.7308992562542258, | |
| "grad_norm": 6.118159510578694, | |
| "learning_rate": 2.0333035389832795e-06, | |
| "loss": 1.1197, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.7363083164300201, | |
| "grad_norm": 5.2165328231189765, | |
| "learning_rate": 2.0186707097223952e-06, | |
| "loss": 1.0348, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.7417173766058147, | |
| "grad_norm": 5.79401876729717, | |
| "learning_rate": 2.0040549853213326e-06, | |
| "loss": 1.0581, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.7417173766058147, | |
| "eval_loss": 1.306036353111267, | |
| "eval_runtime": 80.4204, | |
| "eval_samples_per_second": 15.494, | |
| "eval_steps_per_second": 1.94, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.7471264367816093, | |
| "grad_norm": 5.1382450667316455, | |
| "learning_rate": 1.989456885174865e-06, | |
| "loss": 0.9931, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.7525354969574036, | |
| "grad_norm": 5.923288803726085, | |
| "learning_rate": 1.9748769280514544e-06, | |
| "loss": 1.048, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.757944557133198, | |
| "grad_norm": 5.27150872289333, | |
| "learning_rate": 1.960315632074824e-06, | |
| "loss": 0.9658, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.7633536173089925, | |
| "grad_norm": 5.594406875794453, | |
| "learning_rate": 1.945773514705537e-06, | |
| "loss": 0.9441, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.768762677484787, | |
| "grad_norm": 5.205336223607601, | |
| "learning_rate": 1.931251092722615e-06, | |
| "loss": 1.0016, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.7741717376605814, | |
| "grad_norm": 5.104727647984587, | |
| "learning_rate": 1.916748882205168e-06, | |
| "loss": 0.917, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.7795807978363758, | |
| "grad_norm": 4.929606382507894, | |
| "learning_rate": 1.9022673985140585e-06, | |
| "loss": 1.0204, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.7849898580121704, | |
| "grad_norm": 4.872805835810651, | |
| "learning_rate": 1.8878071562735873e-06, | |
| "loss": 0.9421, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.790398918187965, | |
| "grad_norm": 5.328166060135325, | |
| "learning_rate": 1.8733686693531986e-06, | |
| "loss": 1.0683, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.7958079783637593, | |
| "grad_norm": 5.20759899782578, | |
| "learning_rate": 1.8589524508492308e-06, | |
| "loss": 0.9228, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.8012170385395536, | |
| "grad_norm": 5.362139404873552, | |
| "learning_rate": 1.84455901306667e-06, | |
| "loss": 1.0189, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.8066260987153482, | |
| "grad_norm": 5.165887526933946, | |
| "learning_rate": 1.8301888675009554e-06, | |
| "loss": 1.0733, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.8120351588911427, | |
| "grad_norm": 5.69463106350885, | |
| "learning_rate": 1.8158425248197931e-06, | |
| "loss": 1.0076, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.817444219066937, | |
| "grad_norm": 5.001839035249502, | |
| "learning_rate": 1.8015204948450166e-06, | |
| "loss": 1.009, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.8228532792427314, | |
| "grad_norm": 5.335741327820695, | |
| "learning_rate": 1.787223286534463e-06, | |
| "loss": 0.9548, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.828262339418526, | |
| "grad_norm": 6.290040565183058, | |
| "learning_rate": 1.7729514079638915e-06, | |
| "loss": 1.091, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.8336713995943206, | |
| "grad_norm": 4.84945280688225, | |
| "learning_rate": 1.7587053663089233e-06, | |
| "loss": 0.9476, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.839080459770115, | |
| "grad_norm": 4.954078085115975, | |
| "learning_rate": 1.7444856678270218e-06, | |
| "loss": 0.8989, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.8444895199459093, | |
| "grad_norm": 5.322267912027798, | |
| "learning_rate": 1.7302928178395018e-06, | |
| "loss": 1.0021, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.8498985801217038, | |
| "grad_norm": 4.608936147804684, | |
| "learning_rate": 1.716127320713568e-06, | |
| "loss": 0.9097, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.8553076402974984, | |
| "grad_norm": 5.1258935953966605, | |
| "learning_rate": 1.7019896798443984e-06, | |
| "loss": 0.9417, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.8607167004732927, | |
| "grad_norm": 5.5152859470375475, | |
| "learning_rate": 1.6878803976372465e-06, | |
| "loss": 1.0601, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.866125760649087, | |
| "grad_norm": 8.364778472582735, | |
| "learning_rate": 1.6737999754895965e-06, | |
| "loss": 0.9963, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.8715348208248817, | |
| "grad_norm": 4.9035977247931575, | |
| "learning_rate": 1.6597489137733377e-06, | |
| "loss": 0.8301, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.8769438810006762, | |
| "grad_norm": 5.3057895439765845, | |
| "learning_rate": 1.6457277118169893e-06, | |
| "loss": 0.9862, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 5.656903858451251, | |
| "learning_rate": 1.6317368678879497e-06, | |
| "loss": 0.9841, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.887762001352265, | |
| "grad_norm": 5.054670138966832, | |
| "learning_rate": 1.6177768791747957e-06, | |
| "loss": 0.9873, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.8931710615280595, | |
| "grad_norm": 4.872926655885509, | |
| "learning_rate": 1.6038482417696095e-06, | |
| "loss": 1.0795, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.898580121703854, | |
| "grad_norm": 5.313971672771769, | |
| "learning_rate": 1.5899514506503499e-06, | |
| "loss": 0.9203, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.9039891818796484, | |
| "grad_norm": 4.964610105036094, | |
| "learning_rate": 1.5760869996632685e-06, | |
| "loss": 1.0121, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.9093982420554427, | |
| "grad_norm": 4.990950102894019, | |
| "learning_rate": 1.5622553815053476e-06, | |
| "loss": 0.9234, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.9148073022312373, | |
| "grad_norm": 5.392912217914297, | |
| "learning_rate": 1.5484570877068055e-06, | |
| "loss": 0.9205, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.9202163624070319, | |
| "grad_norm": 5.4989192390271935, | |
| "learning_rate": 1.5346926086136171e-06, | |
| "loss": 1.0099, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.9256254225828262, | |
| "grad_norm": 4.982181015464049, | |
| "learning_rate": 1.5209624333700985e-06, | |
| "loss": 0.937, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.9310344827586206, | |
| "grad_norm": 5.49893209655002, | |
| "learning_rate": 1.5072670499015151e-06, | |
| "loss": 0.9491, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.9364435429344151, | |
| "grad_norm": 5.018154393664902, | |
| "learning_rate": 1.493606944896751e-06, | |
| "loss": 1.0217, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.9418526031102097, | |
| "grad_norm": 5.78990835618344, | |
| "learning_rate": 1.4799826037910082e-06, | |
| "loss": 0.9641, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.947261663286004, | |
| "grad_norm": 4.991442418184385, | |
| "learning_rate": 1.4663945107485567e-06, | |
| "loss": 0.9263, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.9526707234617984, | |
| "grad_norm": 6.0772021735310675, | |
| "learning_rate": 1.4528431486455311e-06, | |
| "loss": 1.0178, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.958079783637593, | |
| "grad_norm": 5.265699493164527, | |
| "learning_rate": 1.4393289990527665e-06, | |
| "loss": 0.9058, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.9634888438133875, | |
| "grad_norm": 5.332909399121638, | |
| "learning_rate": 1.425852542218692e-06, | |
| "loss": 0.8859, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.9688979039891819, | |
| "grad_norm": 5.145306432714561, | |
| "learning_rate": 1.412414257052256e-06, | |
| "loss": 0.9829, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.9743069641649762, | |
| "grad_norm": 5.392247944243994, | |
| "learning_rate": 1.3990146211059141e-06, | |
| "loss": 0.9439, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.9797160243407708, | |
| "grad_norm": 5.346185673554031, | |
| "learning_rate": 1.3856541105586545e-06, | |
| "loss": 0.9911, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.9851250845165653, | |
| "grad_norm": 5.1690668100026285, | |
| "learning_rate": 1.3723332001990774e-06, | |
| "loss": 0.8723, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.9905341446923597, | |
| "grad_norm": 5.00656915809359, | |
| "learning_rate": 1.3590523634085218e-06, | |
| "loss": 0.915, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.9905341446923597, | |
| "eval_loss": 1.247275710105896, | |
| "eval_runtime": 80.2409, | |
| "eval_samples_per_second": 15.528, | |
| "eval_steps_per_second": 1.944, | |
| "step": 368 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 552, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 184, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 526384229253120.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |