| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.5420054200542005, | |
| "eval_steps": 500, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0018066847335140017, | |
| "grad_norm": 2.833585783869219, | |
| "learning_rate": 0.0, | |
| "loss": 0.8251, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0036133694670280035, | |
| "grad_norm": 2.8814908232125855, | |
| "learning_rate": 1.0714285714285714e-06, | |
| "loss": 0.8284, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005420054200542005, | |
| "grad_norm": 2.8770896332736204, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.8426, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007226738934056007, | |
| "grad_norm": 2.7937082319961015, | |
| "learning_rate": 3.2142857142857143e-06, | |
| "loss": 0.8378, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009033423667570008, | |
| "grad_norm": 2.6064730264267317, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.8226, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01084010840108401, | |
| "grad_norm": 2.08053811254878, | |
| "learning_rate": 5.357142857142857e-06, | |
| "loss": 0.8094, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.012646793134598013, | |
| "grad_norm": 1.4474261934350556, | |
| "learning_rate": 6.428571428571429e-06, | |
| "loss": 0.7737, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014453477868112014, | |
| "grad_norm": 1.3829973165220981, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.7732, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.016260162601626018, | |
| "grad_norm": 1.286324724661706, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.7653, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.018066847335140017, | |
| "grad_norm": 2.3802925033552484, | |
| "learning_rate": 9.642857142857144e-06, | |
| "loss": 0.7502, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01987353206865402, | |
| "grad_norm": 2.332362112237301, | |
| "learning_rate": 1.0714285714285714e-05, | |
| "loss": 0.7352, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02168021680216802, | |
| "grad_norm": 2.255275511725211, | |
| "learning_rate": 1.1785714285714286e-05, | |
| "loss": 0.7485, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.023486901535682024, | |
| "grad_norm": 1.862321465187453, | |
| "learning_rate": 1.2857142857142857e-05, | |
| "loss": 0.7252, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.025293586269196026, | |
| "grad_norm": 1.7503296986837806, | |
| "learning_rate": 1.3928571428571429e-05, | |
| "loss": 0.7286, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02710027100271003, | |
| "grad_norm": 1.3910078564328057, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.7179, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.028906955736224028, | |
| "grad_norm": 1.1447088267333418, | |
| "learning_rate": 1.6071428571428572e-05, | |
| "loss": 0.7019, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03071364046973803, | |
| "grad_norm": 0.8309495264632762, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 0.6998, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.032520325203252036, | |
| "grad_norm": 0.7727537242977915, | |
| "learning_rate": 1.8214285714285712e-05, | |
| "loss": 0.6887, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03432700993676603, | |
| "grad_norm": 0.8344364990521232, | |
| "learning_rate": 1.928571428571429e-05, | |
| "loss": 0.6834, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.036133694670280034, | |
| "grad_norm": 0.6823059422974896, | |
| "learning_rate": 2.0357142857142858e-05, | |
| "loss": 0.6597, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.037940379403794036, | |
| "grad_norm": 0.6081366218483659, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.666, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03974706413730804, | |
| "grad_norm": 0.5681526786044694, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.6549, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04155374887082204, | |
| "grad_norm": 0.5770791293540624, | |
| "learning_rate": 2.357142857142857e-05, | |
| "loss": 0.6552, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04336043360433604, | |
| "grad_norm": 0.5162292144131758, | |
| "learning_rate": 2.464285714285714e-05, | |
| "loss": 0.6604, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.045167118337850046, | |
| "grad_norm": 0.5490380070807557, | |
| "learning_rate": 2.5714285714285714e-05, | |
| "loss": 0.6421, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04697380307136405, | |
| "grad_norm": 0.4844741148097692, | |
| "learning_rate": 2.6785714285714288e-05, | |
| "loss": 0.6566, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04878048780487805, | |
| "grad_norm": 0.4632816755153851, | |
| "learning_rate": 2.7857142857142858e-05, | |
| "loss": 0.6346, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.05058717253839205, | |
| "grad_norm": 0.47413499783511553, | |
| "learning_rate": 2.892857142857143e-05, | |
| "loss": 0.6372, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.052393857271906055, | |
| "grad_norm": 0.46848340364244656, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6485, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05420054200542006, | |
| "grad_norm": 0.4249862814563853, | |
| "learning_rate": 2.9999731440137413e-05, | |
| "loss": 0.649, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05600722673893405, | |
| "grad_norm": 0.3741955197649565, | |
| "learning_rate": 2.9998925770166232e-05, | |
| "loss": 0.6379, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.057813911472448055, | |
| "grad_norm": 0.3881925801129848, | |
| "learning_rate": 2.9997583018935875e-05, | |
| "loss": 0.6262, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05962059620596206, | |
| "grad_norm": 0.3922238715881848, | |
| "learning_rate": 2.9995703234527553e-05, | |
| "loss": 0.6364, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06142728093947606, | |
| "grad_norm": 0.33920092930264795, | |
| "learning_rate": 2.999328648425255e-05, | |
| "loss": 0.6233, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06323396567299007, | |
| "grad_norm": 0.38360312337448227, | |
| "learning_rate": 2.999033285464982e-05, | |
| "loss": 0.6291, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06504065040650407, | |
| "grad_norm": 0.37438403307661106, | |
| "learning_rate": 2.9986842451482876e-05, | |
| "loss": 0.6248, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06684733514001806, | |
| "grad_norm": 0.3660585694043342, | |
| "learning_rate": 2.9982815399736008e-05, | |
| "loss": 0.6297, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06865401987353206, | |
| "grad_norm": 0.3536480766826361, | |
| "learning_rate": 2.9978251843609816e-05, | |
| "loss": 0.6221, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07046070460704607, | |
| "grad_norm": 0.4002479177979865, | |
| "learning_rate": 2.9973151946516027e-05, | |
| "loss": 0.6245, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07226738934056007, | |
| "grad_norm": 0.3063169356732776, | |
| "learning_rate": 2.996751589107167e-05, | |
| "loss": 0.6123, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 0.35680447479317984, | |
| "learning_rate": 2.9961343879092512e-05, | |
| "loss": 0.6186, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07588075880758807, | |
| "grad_norm": 0.29028093811805317, | |
| "learning_rate": 2.9954636131585845e-05, | |
| "loss": 0.6117, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07768744354110207, | |
| "grad_norm": 0.327087930172019, | |
| "learning_rate": 2.9947392888742566e-05, | |
| "loss": 0.6116, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.07949412827461608, | |
| "grad_norm": 0.30051645313739944, | |
| "learning_rate": 2.993961440992859e-05, | |
| "loss": 0.6057, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08130081300813008, | |
| "grad_norm": 0.32276337072694117, | |
| "learning_rate": 2.993130097367553e-05, | |
| "loss": 0.6129, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08310749774164408, | |
| "grad_norm": 0.29149163779977444, | |
| "learning_rate": 2.9922452877670775e-05, | |
| "loss": 0.6145, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08491418247515808, | |
| "grad_norm": 0.2848254537728177, | |
| "learning_rate": 2.991307043874677e-05, | |
| "loss": 0.6171, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08672086720867209, | |
| "grad_norm": 0.2736548047269907, | |
| "learning_rate": 2.9903153992869734e-05, | |
| "loss": 0.6033, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.08852755194218609, | |
| "grad_norm": 0.3041549583089305, | |
| "learning_rate": 2.989270389512756e-05, | |
| "loss": 0.6108, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09033423667570009, | |
| "grad_norm": 0.28826353495304036, | |
| "learning_rate": 2.988172051971717e-05, | |
| "loss": 0.6101, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0921409214092141, | |
| "grad_norm": 0.28025059682410447, | |
| "learning_rate": 2.9870204259931062e-05, | |
| "loss": 0.6, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.0939476061427281, | |
| "grad_norm": 0.27249462262783575, | |
| "learning_rate": 2.9858155528143256e-05, | |
| "loss": 0.617, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0957542908762421, | |
| "grad_norm": 0.3172471705535385, | |
| "learning_rate": 2.9845574755794522e-05, | |
| "loss": 0.6026, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0975609756097561, | |
| "grad_norm": 0.24824194798422894, | |
| "learning_rate": 2.9832462393376926e-05, | |
| "loss": 0.5921, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0993676603432701, | |
| "grad_norm": 0.3434388403796833, | |
| "learning_rate": 2.9818818910417706e-05, | |
| "loss": 0.6079, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1011743450767841, | |
| "grad_norm": 0.30129494072287116, | |
| "learning_rate": 2.9804644795462437e-05, | |
| "loss": 0.6121, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10298102981029811, | |
| "grad_norm": 0.29313587299072186, | |
| "learning_rate": 2.9789940556057574e-05, | |
| "loss": 0.5974, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10478771454381211, | |
| "grad_norm": 0.32928460639595075, | |
| "learning_rate": 2.9774706718732255e-05, | |
| "loss": 0.6061, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10659439927732611, | |
| "grad_norm": 0.2680249584703676, | |
| "learning_rate": 2.9758943828979444e-05, | |
| "loss": 0.592, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.10840108401084012, | |
| "grad_norm": 0.34348625850914644, | |
| "learning_rate": 2.9742652451236414e-05, | |
| "loss": 0.5934, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1102077687443541, | |
| "grad_norm": 0.28452776028596916, | |
| "learning_rate": 2.972583316886451e-05, | |
| "loss": 0.597, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1120144534778681, | |
| "grad_norm": 0.3322182395606521, | |
| "learning_rate": 2.9708486584128303e-05, | |
| "loss": 0.6018, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11382113821138211, | |
| "grad_norm": 0.30021383036006005, | |
| "learning_rate": 2.9690613318173966e-05, | |
| "loss": 0.6017, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.11562782294489611, | |
| "grad_norm": 0.276650237644649, | |
| "learning_rate": 2.9672214011007087e-05, | |
| "loss": 0.6001, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.11743450767841011, | |
| "grad_norm": 0.5862319982197399, | |
| "learning_rate": 2.9653289321469715e-05, | |
| "loss": 0.5941, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.11924119241192412, | |
| "grad_norm": 0.26853876464603094, | |
| "learning_rate": 2.9633839927216793e-05, | |
| "loss": 0.5971, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.12104787714543812, | |
| "grad_norm": 0.2829683425685341, | |
| "learning_rate": 2.9613866524691867e-05, | |
| "loss": 0.5874, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12285456187895212, | |
| "grad_norm": 0.2819305232851346, | |
| "learning_rate": 2.9593369829102173e-05, | |
| "loss": 0.6031, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12466124661246612, | |
| "grad_norm": 0.27702734164669873, | |
| "learning_rate": 2.957235057439301e-05, | |
| "loss": 0.6047, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.12646793134598014, | |
| "grad_norm": 0.29063816468584863, | |
| "learning_rate": 2.955080951322147e-05, | |
| "loss": 0.6089, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12827461607949414, | |
| "grad_norm": 0.3063448907538783, | |
| "learning_rate": 2.9528747416929467e-05, | |
| "loss": 0.582, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.13008130081300814, | |
| "grad_norm": 0.2996205495531954, | |
| "learning_rate": 2.9506165075516148e-05, | |
| "loss": 0.5989, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13188798554652212, | |
| "grad_norm": 0.3741556351883678, | |
| "learning_rate": 2.9483063297609577e-05, | |
| "loss": 0.5978, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.13369467028003612, | |
| "grad_norm": 0.2987012098136024, | |
| "learning_rate": 2.9459442910437798e-05, | |
| "loss": 0.5894, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.13550135501355012, | |
| "grad_norm": 0.33060763601185644, | |
| "learning_rate": 2.94353047597992e-05, | |
| "loss": 0.5888, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.13730803974706413, | |
| "grad_norm": 0.2858723573255857, | |
| "learning_rate": 2.941064971003224e-05, | |
| "loss": 0.5843, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.13911472448057813, | |
| "grad_norm": 0.30513180176369226, | |
| "learning_rate": 2.9385478643984484e-05, | |
| "loss": 0.58, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.14092140921409213, | |
| "grad_norm": 0.3480721530294818, | |
| "learning_rate": 2.9359792462981007e-05, | |
| "loss": 0.5909, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14272809394760613, | |
| "grad_norm": 0.3098785903465462, | |
| "learning_rate": 2.9333592086792113e-05, | |
| "loss": 0.5987, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.14453477868112014, | |
| "grad_norm": 0.32360139759712253, | |
| "learning_rate": 2.9306878453600382e-05, | |
| "loss": 0.6078, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14634146341463414, | |
| "grad_norm": 0.31681100648297805, | |
| "learning_rate": 2.9279652519967105e-05, | |
| "loss": 0.5928, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 0.33637175342033093, | |
| "learning_rate": 2.9251915260798024e-05, | |
| "loss": 0.5851, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.14995483288166214, | |
| "grad_norm": 0.32126065455334785, | |
| "learning_rate": 2.9223667669308395e-05, | |
| "loss": 0.595, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.15176151761517614, | |
| "grad_norm": 0.28162115413110317, | |
| "learning_rate": 2.9194910756987464e-05, | |
| "loss": 0.5886, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.15356820234869015, | |
| "grad_norm": 0.2927832038468001, | |
| "learning_rate": 2.9165645553562215e-05, | |
| "loss": 0.5919, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.15537488708220415, | |
| "grad_norm": 0.2842368528997179, | |
| "learning_rate": 2.9135873106960525e-05, | |
| "loss": 0.5907, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.15718157181571815, | |
| "grad_norm": 0.3044892933421028, | |
| "learning_rate": 2.9105594483273603e-05, | |
| "loss": 0.5932, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.15898825654923215, | |
| "grad_norm": 0.2787616554815814, | |
| "learning_rate": 2.9074810766717865e-05, | |
| "loss": 0.5972, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.16079494128274616, | |
| "grad_norm": 0.31536931569551113, | |
| "learning_rate": 2.904352305959606e-05, | |
| "loss": 0.5941, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.16260162601626016, | |
| "grad_norm": 0.2812317162219802, | |
| "learning_rate": 2.9011732482257835e-05, | |
| "loss": 0.5858, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16440831074977416, | |
| "grad_norm": 0.3243626162570077, | |
| "learning_rate": 2.89794401730596e-05, | |
| "loss": 0.5839, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.16621499548328816, | |
| "grad_norm": 0.31149750778219104, | |
| "learning_rate": 2.894664728832377e-05, | |
| "loss": 0.5879, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.16802168021680217, | |
| "grad_norm": 0.2799988189260621, | |
| "learning_rate": 2.8913355002297367e-05, | |
| "loss": 0.5939, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.16982836495031617, | |
| "grad_norm": 0.3659704073921071, | |
| "learning_rate": 2.887956450710995e-05, | |
| "loss": 0.594, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.17163504968383017, | |
| "grad_norm": 0.3155038381999107, | |
| "learning_rate": 2.8845277012730963e-05, | |
| "loss": 0.5857, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17344173441734417, | |
| "grad_norm": 0.28670310893186607, | |
| "learning_rate": 2.8810493746926364e-05, | |
| "loss": 0.5854, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.17524841915085818, | |
| "grad_norm": 0.3027782704454321, | |
| "learning_rate": 2.87752159552147e-05, | |
| "loss": 0.5913, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.17705510388437218, | |
| "grad_norm": 0.29488509286302705, | |
| "learning_rate": 2.87394449008225e-05, | |
| "loss": 0.5842, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.17886178861788618, | |
| "grad_norm": 0.3023043372178035, | |
| "learning_rate": 2.8703181864639013e-05, | |
| "loss": 0.5971, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.18066847335140018, | |
| "grad_norm": 0.2920343922477137, | |
| "learning_rate": 2.8666428145170385e-05, | |
| "loss": 0.5703, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18247515808491419, | |
| "grad_norm": 0.26724172739119173, | |
| "learning_rate": 2.8629185058493116e-05, | |
| "loss": 0.5864, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.1842818428184282, | |
| "grad_norm": 0.34770771323943805, | |
| "learning_rate": 2.8591453938206985e-05, | |
| "loss": 0.6044, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.1860885275519422, | |
| "grad_norm": 0.273213628338846, | |
| "learning_rate": 2.8553236135387247e-05, | |
| "loss": 0.5831, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.1878952122854562, | |
| "grad_norm": 0.34065241651987704, | |
| "learning_rate": 2.8514533018536286e-05, | |
| "loss": 0.5812, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.1897018970189702, | |
| "grad_norm": 0.28586917405518736, | |
| "learning_rate": 2.8475345973534605e-05, | |
| "loss": 0.5734, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1915085817524842, | |
| "grad_norm": 0.2702413811100727, | |
| "learning_rate": 2.8435676403591193e-05, | |
| "loss": 0.5764, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1933152664859982, | |
| "grad_norm": 0.28403756910607586, | |
| "learning_rate": 2.8395525729193284e-05, | |
| "loss": 0.5826, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1951219512195122, | |
| "grad_norm": 0.2578481483409949, | |
| "learning_rate": 2.835489538805548e-05, | |
| "loss": 0.5803, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.1969286359530262, | |
| "grad_norm": 0.2757886473056093, | |
| "learning_rate": 2.8313786835068314e-05, | |
| "loss": 0.5893, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.1987353206865402, | |
| "grad_norm": 0.28782114342933773, | |
| "learning_rate": 2.8272201542246077e-05, | |
| "loss": 0.571, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2005420054200542, | |
| "grad_norm": 0.28817871732654127, | |
| "learning_rate": 2.8230140998674185e-05, | |
| "loss": 0.5907, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2023486901535682, | |
| "grad_norm": 0.27487749659574895, | |
| "learning_rate": 2.8187606710455807e-05, | |
| "loss": 0.584, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2041553748870822, | |
| "grad_norm": 0.29745293014063967, | |
| "learning_rate": 2.8144600200657953e-05, | |
| "loss": 0.5881, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.20596205962059622, | |
| "grad_norm": 0.30986472376082147, | |
| "learning_rate": 2.8101123009256946e-05, | |
| "loss": 0.5713, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.20776874435411022, | |
| "grad_norm": 0.258458540249251, | |
| "learning_rate": 2.8057176693083253e-05, | |
| "loss": 0.5869, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.20957542908762422, | |
| "grad_norm": 0.31218243071865426, | |
| "learning_rate": 2.8012762825765763e-05, | |
| "loss": 0.5808, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.21138211382113822, | |
| "grad_norm": 0.3015978973078328, | |
| "learning_rate": 2.7967882997675424e-05, | |
| "loss": 0.5921, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.21318879855465223, | |
| "grad_norm": 0.2906725555402426, | |
| "learning_rate": 2.7922538815868287e-05, | |
| "loss": 0.5793, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.21499548328816623, | |
| "grad_norm": 0.2686708606832387, | |
| "learning_rate": 2.7876731904027994e-05, | |
| "loss": 0.5696, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.21680216802168023, | |
| "grad_norm": 0.306920182031447, | |
| "learning_rate": 2.78304639024076e-05, | |
| "loss": 0.5886, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2186088527551942, | |
| "grad_norm": 0.2675555748745438, | |
| "learning_rate": 2.7783736467770863e-05, | |
| "loss": 0.5841, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2204155374887082, | |
| "grad_norm": 0.2961259999734875, | |
| "learning_rate": 2.7736551273332908e-05, | |
| "loss": 0.5888, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.2979054845453111, | |
| "learning_rate": 2.7688910008700305e-05, | |
| "loss": 0.5792, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2240289069557362, | |
| "grad_norm": 0.2643324547090545, | |
| "learning_rate": 2.764081437981059e-05, | |
| "loss": 0.565, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.22583559168925021, | |
| "grad_norm": 0.27546367306007746, | |
| "learning_rate": 2.7592266108871158e-05, | |
| "loss": 0.5723, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.22764227642276422, | |
| "grad_norm": 0.3046480301612398, | |
| "learning_rate": 2.754326693429761e-05, | |
| "loss": 0.5847, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.22944896115627822, | |
| "grad_norm": 0.2706720967609819, | |
| "learning_rate": 2.7493818610651493e-05, | |
| "loss": 0.5774, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.23125564588979222, | |
| "grad_norm": 0.25419185973706093, | |
| "learning_rate": 2.744392290857747e-05, | |
| "loss": 0.5763, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.23306233062330622, | |
| "grad_norm": 0.27770964690614774, | |
| "learning_rate": 2.7393581614739924e-05, | |
| "loss": 0.5717, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.23486901535682023, | |
| "grad_norm": 0.28692816251688974, | |
| "learning_rate": 2.7342796531758984e-05, | |
| "loss": 0.5901, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.23667570009033423, | |
| "grad_norm": 0.291082485601733, | |
| "learning_rate": 2.729156947814598e-05, | |
| "loss": 0.5849, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.23848238482384823, | |
| "grad_norm": 0.2655019529750113, | |
| "learning_rate": 2.7239902288238297e-05, | |
| "loss": 0.5633, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.24028906955736223, | |
| "grad_norm": 0.2703759360941128, | |
| "learning_rate": 2.7187796812133733e-05, | |
| "loss": 0.5788, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.24209575429087624, | |
| "grad_norm": 0.2975052958330194, | |
| "learning_rate": 2.7135254915624213e-05, | |
| "loss": 0.5795, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.24390243902439024, | |
| "grad_norm": 0.27240132194115707, | |
| "learning_rate": 2.708227848012901e-05, | |
| "loss": 0.5771, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.24570912375790424, | |
| "grad_norm": 0.27811650822381756, | |
| "learning_rate": 2.7028869402627357e-05, | |
| "loss": 0.5803, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.24751580849141824, | |
| "grad_norm": 0.2830474782550724, | |
| "learning_rate": 2.6975029595590523e-05, | |
| "loss": 0.5645, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.24932249322493225, | |
| "grad_norm": 0.2938710437376192, | |
| "learning_rate": 2.6920760986913332e-05, | |
| "loss": 0.5814, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.25112917795844625, | |
| "grad_norm": 0.27314197829276166, | |
| "learning_rate": 2.6866065519845124e-05, | |
| "loss": 0.584, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.2529358626919603, | |
| "grad_norm": 0.2828223996342519, | |
| "learning_rate": 2.681094515292018e-05, | |
| "loss": 0.5834, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.25474254742547425, | |
| "grad_norm": 0.24839151099849252, | |
| "learning_rate": 2.6755401859887598e-05, | |
| "loss": 0.5644, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2565492321589883, | |
| "grad_norm": 0.32766177067955377, | |
| "learning_rate": 2.6699437629640595e-05, | |
| "loss": 0.5843, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.25835591689250226, | |
| "grad_norm": 0.2953450112150113, | |
| "learning_rate": 2.6643054466145297e-05, | |
| "loss": 0.5763, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2601626016260163, | |
| "grad_norm": 0.2867043623826392, | |
| "learning_rate": 2.6586254388368995e-05, | |
| "loss": 0.5647, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.26196928635953026, | |
| "grad_norm": 0.28484485619651345, | |
| "learning_rate": 2.652903943020783e-05, | |
| "loss": 0.5772, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.26377597109304424, | |
| "grad_norm": 0.2783119448944388, | |
| "learning_rate": 2.647141164041398e-05, | |
| "loss": 0.582, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.26558265582655827, | |
| "grad_norm": 0.286490896354593, | |
| "learning_rate": 2.641337308252228e-05, | |
| "loss": 0.5715, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.26738934056007224, | |
| "grad_norm": 0.2986882378912008, | |
| "learning_rate": 2.6354925834776346e-05, | |
| "loss": 0.573, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.26919602529358627, | |
| "grad_norm": 0.308835984272729, | |
| "learning_rate": 2.6296071990054167e-05, | |
| "loss": 0.5703, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.27100271002710025, | |
| "grad_norm": 0.2665260387800768, | |
| "learning_rate": 2.6236813655793123e-05, | |
| "loss": 0.58, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2728093947606143, | |
| "grad_norm": 0.28323735489775853, | |
| "learning_rate": 2.617715295391457e-05, | |
| "loss": 0.562, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.27461607949412825, | |
| "grad_norm": 0.27363701184963624, | |
| "learning_rate": 2.6117092020747824e-05, | |
| "loss": 0.5697, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2764227642276423, | |
| "grad_norm": 0.2791323579500085, | |
| "learning_rate": 2.6056633006953677e-05, | |
| "loss": 0.5697, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.27822944896115626, | |
| "grad_norm": 0.2742265379635391, | |
| "learning_rate": 2.5995778077447393e-05, | |
| "loss": 0.5736, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2800361336946703, | |
| "grad_norm": 0.3098153479837776, | |
| "learning_rate": 2.5934529411321174e-05, | |
| "loss": 0.5665, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.28184281842818426, | |
| "grad_norm": 0.25253625108017397, | |
| "learning_rate": 2.587288920176613e-05, | |
| "loss": 0.5661, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2836495031616983, | |
| "grad_norm": 0.3596999576828901, | |
| "learning_rate": 2.581085965599375e-05, | |
| "loss": 0.5794, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.28545618789521227, | |
| "grad_norm": 0.26441404761712806, | |
| "learning_rate": 2.5748442995156882e-05, | |
| "loss": 0.5767, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.2872628726287263, | |
| "grad_norm": 0.28376911904978996, | |
| "learning_rate": 2.5685641454270172e-05, | |
| "loss": 0.5576, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.28906955736224027, | |
| "grad_norm": 0.31680061742974575, | |
| "learning_rate": 2.5622457282130046e-05, | |
| "loss": 0.5664, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2908762420957543, | |
| "grad_norm": 0.25774838728867683, | |
| "learning_rate": 2.5558892741234173e-05, | |
| "loss": 0.5746, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2926829268292683, | |
| "grad_norm": 0.32688394361315487, | |
| "learning_rate": 2.5494950107700482e-05, | |
| "loss": 0.5605, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2944896115627823, | |
| "grad_norm": 0.25439549403737804, | |
| "learning_rate": 2.5430631671185616e-05, | |
| "loss": 0.5689, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 0.3068192818856112, | |
| "learning_rate": 2.5365939734802973e-05, | |
| "loss": 0.5784, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.2981029810298103, | |
| "grad_norm": 0.2628542180584984, | |
| "learning_rate": 2.5300876615040223e-05, | |
| "loss": 0.5694, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2999096657633243, | |
| "grad_norm": 0.3126576916530388, | |
| "learning_rate": 2.523544464167637e-05, | |
| "loss": 0.5756, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3017163504968383, | |
| "grad_norm": 0.2661513382628665, | |
| "learning_rate": 2.5169646157698313e-05, | |
| "loss": 0.5743, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3035230352303523, | |
| "grad_norm": 0.2794359843877779, | |
| "learning_rate": 2.5103483519216964e-05, | |
| "loss": 0.5731, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3053297199638663, | |
| "grad_norm": 0.267337015732144, | |
| "learning_rate": 2.5036959095382875e-05, | |
| "loss": 0.5652, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3071364046973803, | |
| "grad_norm": 0.283872456782531, | |
| "learning_rate": 2.4970075268301388e-05, | |
| "loss": 0.5787, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3089430894308943, | |
| "grad_norm": 0.2615383089315913, | |
| "learning_rate": 2.4902834432947353e-05, | |
| "loss": 0.5616, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.3107497741644083, | |
| "grad_norm": 0.27454236304750723, | |
| "learning_rate": 2.4835238997079382e-05, | |
| "loss": 0.569, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.31255645889792233, | |
| "grad_norm": 0.26790606034696746, | |
| "learning_rate": 2.4767291381153603e-05, | |
| "loss": 0.5721, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3143631436314363, | |
| "grad_norm": 0.24338452694103885, | |
| "learning_rate": 2.4698994018236994e-05, | |
| "loss": 0.5635, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.31616982836495033, | |
| "grad_norm": 0.27640360012160314, | |
| "learning_rate": 2.4630349353920284e-05, | |
| "loss": 0.5676, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3179765130984643, | |
| "grad_norm": 0.2701056771293538, | |
| "learning_rate": 2.4561359846230346e-05, | |
| "loss": 0.582, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.31978319783197834, | |
| "grad_norm": 0.24753738293952274, | |
| "learning_rate": 2.4492027965542217e-05, | |
| "loss": 0.5571, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.3215898825654923, | |
| "grad_norm": 0.26161253512639976, | |
| "learning_rate": 2.44223561944906e-05, | |
| "loss": 0.5641, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.32339656729900634, | |
| "grad_norm": 0.2570144122258103, | |
| "learning_rate": 2.4352347027881003e-05, | |
| "loss": 0.5738, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.3252032520325203, | |
| "grad_norm": 0.27930694625424896, | |
| "learning_rate": 2.4282002972600382e-05, | |
| "loss": 0.5817, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.32700993676603435, | |
| "grad_norm": 0.23436344914835724, | |
| "learning_rate": 2.4211326547527377e-05, | |
| "loss": 0.5602, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.3288166214995483, | |
| "grad_norm": 0.2729605037353879, | |
| "learning_rate": 2.4140320283442125e-05, | |
| "loss": 0.5704, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.33062330623306235, | |
| "grad_norm": 0.2798695539410275, | |
| "learning_rate": 2.4068986722935625e-05, | |
| "loss": 0.5808, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.3324299909665763, | |
| "grad_norm": 0.2789180229196701, | |
| "learning_rate": 2.3997328420318705e-05, | |
| "loss": 0.5769, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.33423667570009036, | |
| "grad_norm": 0.27131707792538245, | |
| "learning_rate": 2.3925347941530556e-05, | |
| "loss": 0.5733, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.33604336043360433, | |
| "grad_norm": 0.27412770965167466, | |
| "learning_rate": 2.3853047864046843e-05, | |
| "loss": 0.5684, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.33785004516711836, | |
| "grad_norm": 0.2782894711462652, | |
| "learning_rate": 2.3780430776787413e-05, | |
| "loss": 0.5659, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.33965672990063234, | |
| "grad_norm": 0.2430242526823377, | |
| "learning_rate": 2.3707499280023604e-05, | |
| "loss": 0.5658, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.34146341463414637, | |
| "grad_norm": 0.28013753256340473, | |
| "learning_rate": 2.3634255985285104e-05, | |
| "loss": 0.5729, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.34327009936766034, | |
| "grad_norm": 0.2595735350692398, | |
| "learning_rate": 2.356070351526648e-05, | |
| "loss": 0.5621, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.34507678410117437, | |
| "grad_norm": 0.2675696222506633, | |
| "learning_rate": 2.348684450373322e-05, | |
| "loss": 0.5598, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.34688346883468835, | |
| "grad_norm": 0.25005185921177986, | |
| "learning_rate": 2.3412681595427467e-05, | |
| "loss": 0.5674, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.3486901535682023, | |
| "grad_norm": 0.2830143632836114, | |
| "learning_rate": 2.3338217445973268e-05, | |
| "loss": 0.5688, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.35049683830171635, | |
| "grad_norm": 0.2512108464193193, | |
| "learning_rate": 2.3263454721781537e-05, | |
| "loss": 0.5706, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.3523035230352303, | |
| "grad_norm": 0.2768218752940674, | |
| "learning_rate": 2.318839609995453e-05, | |
| "loss": 0.5536, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.35411020776874436, | |
| "grad_norm": 0.25016072504753256, | |
| "learning_rate": 2.3113044268189995e-05, | |
| "loss": 0.5669, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.35591689250225833, | |
| "grad_norm": 0.25210388765899894, | |
| "learning_rate": 2.303740192468495e-05, | |
| "loss": 0.5668, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.35772357723577236, | |
| "grad_norm": 0.28513394233627476, | |
| "learning_rate": 2.2961471778039045e-05, | |
| "loss": 0.5438, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.35953026196928634, | |
| "grad_norm": 0.2548176732782281, | |
| "learning_rate": 2.288525654715757e-05, | |
| "loss": 0.5671, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.36133694670280037, | |
| "grad_norm": 0.27079689306361104, | |
| "learning_rate": 2.280875896115413e-05, | |
| "loss": 0.556, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.36314363143631434, | |
| "grad_norm": 0.24997335486124786, | |
| "learning_rate": 2.2731981759252876e-05, | |
| "loss": 0.5507, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.36495031616982837, | |
| "grad_norm": 0.24175290069909067, | |
| "learning_rate": 2.2654927690690445e-05, | |
| "loss": 0.5774, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.36675700090334235, | |
| "grad_norm": 0.2532933339218424, | |
| "learning_rate": 2.257759951461752e-05, | |
| "loss": 0.5641, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.3685636856368564, | |
| "grad_norm": 0.2621868163907268, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.577, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 0.24085524911800554, | |
| "learning_rate": 2.24221319255199e-05, | |
| "loss": 0.5677, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3721770551038844, | |
| "grad_norm": 0.2632917461430485, | |
| "learning_rate": 2.234399807947579e-05, | |
| "loss": 0.5546, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.37398373983739835, | |
| "grad_norm": 0.23460851905806784, | |
| "learning_rate": 2.2265601259683e-05, | |
| "loss": 0.5626, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3757904245709124, | |
| "grad_norm": 0.24036262844750986, | |
| "learning_rate": 2.2186944273373426e-05, | |
| "loss": 0.55, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.37759710930442636, | |
| "grad_norm": 0.25653419125069626, | |
| "learning_rate": 2.210802993709498e-05, | |
| "loss": 0.5714, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.3794037940379404, | |
| "grad_norm": 0.25795821587482737, | |
| "learning_rate": 2.202886107661078e-05, | |
| "loss": 0.5674, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.38121047877145436, | |
| "grad_norm": 0.2682406799824342, | |
| "learning_rate": 2.1949440526797928e-05, | |
| "loss": 0.5671, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.3830171635049684, | |
| "grad_norm": 0.25405140008902716, | |
| "learning_rate": 2.1869771131546015e-05, | |
| "loss": 0.5748, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.38482384823848237, | |
| "grad_norm": 0.277498893414283, | |
| "learning_rate": 2.178985574365529e-05, | |
| "loss": 0.5751, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.3866305329719964, | |
| "grad_norm": 0.2621687610495017, | |
| "learning_rate": 2.170969722473449e-05, | |
| "loss": 0.561, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.3884372177055104, | |
| "grad_norm": 0.33519408751709534, | |
| "learning_rate": 2.1629298445098403e-05, | |
| "loss": 0.588, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.3902439024390244, | |
| "grad_norm": 0.25011729377513336, | |
| "learning_rate": 2.154866228366505e-05, | |
| "loss": 0.5765, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.3920505871725384, | |
| "grad_norm": 0.24393104363426, | |
| "learning_rate": 2.146779162785263e-05, | |
| "loss": 0.5767, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3938572719060524, | |
| "grad_norm": 0.2712254007014316, | |
| "learning_rate": 2.138668937347609e-05, | |
| "loss": 0.5684, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.3956639566395664, | |
| "grad_norm": 0.23805865346251903, | |
| "learning_rate": 2.1305358424643484e-05, | |
| "loss": 0.5634, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.3974706413730804, | |
| "grad_norm": 0.23597914856820135, | |
| "learning_rate": 2.1223801693651927e-05, | |
| "loss": 0.5656, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3992773261065944, | |
| "grad_norm": 0.2606858225008041, | |
| "learning_rate": 2.114202210088336e-05, | |
| "loss": 0.5547, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.4010840108401084, | |
| "grad_norm": 0.5424755323054944, | |
| "learning_rate": 2.106002257469993e-05, | |
| "loss": 0.572, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.4028906955736224, | |
| "grad_norm": 0.2507723535221933, | |
| "learning_rate": 2.0977806051339172e-05, | |
| "loss": 0.555, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.4046973803071364, | |
| "grad_norm": 0.258072944340719, | |
| "learning_rate": 2.0895375474808857e-05, | |
| "loss": 0.5593, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.4065040650406504, | |
| "grad_norm": 0.2872582066072117, | |
| "learning_rate": 2.0812733796781544e-05, | |
| "loss": 0.565, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4083107497741644, | |
| "grad_norm": 0.2364423377916723, | |
| "learning_rate": 2.0729883976488936e-05, | |
| "loss": 0.5703, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.4101174345076784, | |
| "grad_norm": 0.26270639218601316, | |
| "learning_rate": 2.064682898061588e-05, | |
| "loss": 0.5747, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.41192411924119243, | |
| "grad_norm": 0.2631951060473206, | |
| "learning_rate": 2.0563571783194146e-05, | |
| "loss": 0.5615, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.4137308039747064, | |
| "grad_norm": 0.2317741311806957, | |
| "learning_rate": 2.0480115365495928e-05, | |
| "loss": 0.5639, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.41553748870822044, | |
| "grad_norm": 0.30997167552763305, | |
| "learning_rate": 2.0396462715927107e-05, | |
| "loss": 0.5661, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4173441734417344, | |
| "grad_norm": 0.2480886169428658, | |
| "learning_rate": 2.0312616829920222e-05, | |
| "loss": 0.5652, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.41915085817524844, | |
| "grad_norm": 0.23699658602720144, | |
| "learning_rate": 2.022858070982723e-05, | |
| "loss": 0.568, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.4209575429087624, | |
| "grad_norm": 0.24827523861823048, | |
| "learning_rate": 2.0144357364811973e-05, | |
| "loss": 0.5504, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.42276422764227645, | |
| "grad_norm": 0.25171738807398075, | |
| "learning_rate": 2.0059949810742452e-05, | |
| "loss": 0.5656, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.4245709123757904, | |
| "grad_norm": 0.2205290140571607, | |
| "learning_rate": 1.997536107008281e-05, | |
| "loss": 0.5603, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.42637759710930445, | |
| "grad_norm": 0.23922227707786625, | |
| "learning_rate": 1.9890594171785128e-05, | |
| "loss": 0.5722, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.4281842818428184, | |
| "grad_norm": 0.24378531634524622, | |
| "learning_rate": 1.9805652151180945e-05, | |
| "loss": 0.5719, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.42999096657633246, | |
| "grad_norm": 0.22020061853610345, | |
| "learning_rate": 1.972053804987258e-05, | |
| "loss": 0.58, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.43179765130984643, | |
| "grad_norm": 0.24680402960694306, | |
| "learning_rate": 1.963525491562421e-05, | |
| "loss": 0.561, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.43360433604336046, | |
| "grad_norm": 0.22284712718565894, | |
| "learning_rate": 1.954980580225275e-05, | |
| "loss": 0.5617, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.43541102077687444, | |
| "grad_norm": 0.2484592670769788, | |
| "learning_rate": 1.946419376951848e-05, | |
| "loss": 0.5603, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.4372177055103884, | |
| "grad_norm": 0.2388978520409849, | |
| "learning_rate": 1.9378421883015505e-05, | |
| "loss": 0.5608, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.43902439024390244, | |
| "grad_norm": 0.2523677374867834, | |
| "learning_rate": 1.9292493214061953e-05, | |
| "loss": 0.5654, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.4408310749774164, | |
| "grad_norm": 0.2383883235220744, | |
| "learning_rate": 1.9206410839590042e-05, | |
| "loss": 0.5719, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.44263775971093045, | |
| "grad_norm": 0.29119882075217696, | |
| "learning_rate": 1.9120177842035853e-05, | |
| "loss": 0.5636, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.24457061677600248, | |
| "learning_rate": 1.9033797309228984e-05, | |
| "loss": 0.5703, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.44625112917795845, | |
| "grad_norm": 0.245042331857765, | |
| "learning_rate": 1.8947272334281977e-05, | |
| "loss": 0.5523, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.4480578139114724, | |
| "grad_norm": 0.28714938771611187, | |
| "learning_rate": 1.8860606015479537e-05, | |
| "loss": 0.5707, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.44986449864498645, | |
| "grad_norm": 0.24121091985031393, | |
| "learning_rate": 1.877380145616763e-05, | |
| "loss": 0.5548, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.45167118337850043, | |
| "grad_norm": 0.2606569077339658, | |
| "learning_rate": 1.868686176464232e-05, | |
| "loss": 0.5697, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.45347786811201446, | |
| "grad_norm": 0.23580420111011996, | |
| "learning_rate": 1.8599790054038487e-05, | |
| "loss": 0.5565, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.45528455284552843, | |
| "grad_norm": 0.23127251461544818, | |
| "learning_rate": 1.8512589442218358e-05, | |
| "loss": 0.5387, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.45709123757904246, | |
| "grad_norm": 0.24560555350844132, | |
| "learning_rate": 1.8425263051659838e-05, | |
| "loss": 0.5618, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.45889792231255644, | |
| "grad_norm": 0.2563321621344126, | |
| "learning_rate": 1.8337814009344716e-05, | |
| "loss": 0.5606, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.46070460704607047, | |
| "grad_norm": 0.2605334576865085, | |
| "learning_rate": 1.8250245446646707e-05, | |
| "loss": 0.5541, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.46251129177958444, | |
| "grad_norm": 0.25888298184378367, | |
| "learning_rate": 1.8162560499219286e-05, | |
| "loss": 0.5539, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.4643179765130985, | |
| "grad_norm": 0.25609617770435344, | |
| "learning_rate": 1.807476230688346e-05, | |
| "loss": 0.5647, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.46612466124661245, | |
| "grad_norm": 0.25864894166343505, | |
| "learning_rate": 1.7986854013515274e-05, | |
| "loss": 0.5591, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.4679313459801265, | |
| "grad_norm": 0.24989575985544257, | |
| "learning_rate": 1.78988387669333e-05, | |
| "loss": 0.5629, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.46973803071364045, | |
| "grad_norm": 0.25850049217533544, | |
| "learning_rate": 1.781071971878587e-05, | |
| "loss": 0.5765, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4715447154471545, | |
| "grad_norm": 0.26195883131588654, | |
| "learning_rate": 1.7722500024438244e-05, | |
| "loss": 0.5669, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.47335140018066846, | |
| "grad_norm": 0.2643441022284166, | |
| "learning_rate": 1.7634182842859628e-05, | |
| "loss": 0.5559, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.4751580849141825, | |
| "grad_norm": 0.25888645697984375, | |
| "learning_rate": 1.7545771336510033e-05, | |
| "loss": 0.5608, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.47696476964769646, | |
| "grad_norm": 0.25422111952455206, | |
| "learning_rate": 1.7457268671227067e-05, | |
| "loss": 0.566, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.4787714543812105, | |
| "grad_norm": 0.27122044053732625, | |
| "learning_rate": 1.736867801611254e-05, | |
| "loss": 0.5484, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.48057813911472447, | |
| "grad_norm": 0.23792779766126487, | |
| "learning_rate": 1.728000254341901e-05, | |
| "loss": 0.5718, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.4823848238482385, | |
| "grad_norm": 0.22612645512697185, | |
| "learning_rate": 1.7191245428436175e-05, | |
| "loss": 0.5567, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.48419150858175247, | |
| "grad_norm": 0.242856441232987, | |
| "learning_rate": 1.7102409849377188e-05, | |
| "loss": 0.5565, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.4859981933152665, | |
| "grad_norm": 0.22598008818744872, | |
| "learning_rate": 1.7013498987264832e-05, | |
| "loss": 0.5634, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.4878048780487805, | |
| "grad_norm": 0.28177732615064893, | |
| "learning_rate": 1.6924516025817636e-05, | |
| "loss": 0.545, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4896115627822945, | |
| "grad_norm": 0.2440174064317069, | |
| "learning_rate": 1.683546415133584e-05, | |
| "loss": 0.5667, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.4914182475158085, | |
| "grad_norm": 0.2617297152457466, | |
| "learning_rate": 1.6746346552587342e-05, | |
| "loss": 0.5612, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.4932249322493225, | |
| "grad_norm": 0.25621942967426764, | |
| "learning_rate": 1.665716642069349e-05, | |
| "loss": 0.5692, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.4950316169828365, | |
| "grad_norm": 0.25791288879390334, | |
| "learning_rate": 1.6567926949014805e-05, | |
| "loss": 0.5568, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.4968383017163505, | |
| "grad_norm": 0.2418430510543647, | |
| "learning_rate": 1.6478631333036655e-05, | |
| "loss": 0.5509, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.4986449864498645, | |
| "grad_norm": 0.22083223278151687, | |
| "learning_rate": 1.638928277025482e-05, | |
| "loss": 0.5435, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.5004516711833785, | |
| "grad_norm": 0.2291376855754137, | |
| "learning_rate": 1.6299884460061005e-05, | |
| "loss": 0.5551, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.5022583559168925, | |
| "grad_norm": 0.2302128127786599, | |
| "learning_rate": 1.621043960362826e-05, | |
| "loss": 0.5419, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.5040650406504065, | |
| "grad_norm": 0.228883386803342, | |
| "learning_rate": 1.6120951403796367e-05, | |
| "loss": 0.5571, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.5058717253839206, | |
| "grad_norm": 0.2401265638231575, | |
| "learning_rate": 1.603142306495714e-05, | |
| "loss": 0.5532, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5076784101174345, | |
| "grad_norm": 0.2114239387262738, | |
| "learning_rate": 1.5941857792939702e-05, | |
| "loss": 0.5541, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.5094850948509485, | |
| "grad_norm": 0.22325184258966826, | |
| "learning_rate": 1.585225879489567e-05, | |
| "loss": 0.5487, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.5112917795844625, | |
| "grad_norm": 0.22702669120714522, | |
| "learning_rate": 1.5762629279184326e-05, | |
| "loss": 0.5635, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.5130984643179766, | |
| "grad_norm": 0.2165975196960249, | |
| "learning_rate": 1.5672972455257726e-05, | |
| "loss": 0.552, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5149051490514905, | |
| "grad_norm": 0.22658048500397085, | |
| "learning_rate": 1.5583291533545775e-05, | |
| "loss": 0.5465, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5167118337850045, | |
| "grad_norm": 0.23175050113812948, | |
| "learning_rate": 1.549358972534128e-05, | |
| "loss": 0.5441, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 0.21294545077390667, | |
| "learning_rate": 1.5403870242684942e-05, | |
| "loss": 0.5654, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.5203252032520326, | |
| "grad_norm": 0.2504268423528941, | |
| "learning_rate": 1.5314136298250355e-05, | |
| "loss": 0.5628, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.5221318879855466, | |
| "grad_norm": 0.2099986397700923, | |
| "learning_rate": 1.5224391105228956e-05, | |
| "loss": 0.5474, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5239385727190605, | |
| "grad_norm": 0.20317375678953348, | |
| "learning_rate": 1.5134637877214968e-05, | |
| "loss": 0.5588, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5257452574525745, | |
| "grad_norm": 0.23346942768814213, | |
| "learning_rate": 1.5044879828090346e-05, | |
| "loss": 0.5631, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.5275519421860885, | |
| "grad_norm": 0.23620525056730876, | |
| "learning_rate": 1.4955120171909658e-05, | |
| "loss": 0.5671, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5293586269196026, | |
| "grad_norm": 0.20358818421382174, | |
| "learning_rate": 1.4865362122785031e-05, | |
| "loss": 0.5598, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.5311653116531165, | |
| "grad_norm": 0.24259950926712837, | |
| "learning_rate": 1.4775608894771048e-05, | |
| "loss": 0.5521, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5329719963866305, | |
| "grad_norm": 0.2643019881724919, | |
| "learning_rate": 1.4685863701749648e-05, | |
| "loss": 0.5497, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5347786811201445, | |
| "grad_norm": 0.24143753848243574, | |
| "learning_rate": 1.4596129757315062e-05, | |
| "loss": 0.554, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.5365853658536586, | |
| "grad_norm": 0.23280547368412433, | |
| "learning_rate": 1.4506410274658718e-05, | |
| "loss": 0.5589, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5383920505871725, | |
| "grad_norm": 0.23707855241324371, | |
| "learning_rate": 1.441670846645423e-05, | |
| "loss": 0.5536, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.5401987353206865, | |
| "grad_norm": 0.23874631751676395, | |
| "learning_rate": 1.4327027544742281e-05, | |
| "loss": 0.569, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5420054200542005, | |
| "grad_norm": 0.22915529315710134, | |
| "learning_rate": 1.4237370720815675e-05, | |
| "loss": 0.5569, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 553, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 347755278106624.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |