| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 184, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010869565217391304, | |
| "grad_norm": 4.7562408971323125, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 1.4045, | |
| "num_tokens": 90966.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.021739130434782608, | |
| "grad_norm": 4.856641936347526, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 1.4286, | |
| "num_tokens": 176808.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03260869565217391, | |
| "grad_norm": 4.632629643475528, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 1.3767, | |
| "num_tokens": 268011.0, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.043478260869565216, | |
| "grad_norm": 4.433390852741152, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 1.3293, | |
| "num_tokens": 369286.0, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05434782608695652, | |
| "grad_norm": 4.267197673905712, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 1.3839, | |
| "num_tokens": 465540.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06521739130434782, | |
| "grad_norm": 3.717050440728708, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 1.2858, | |
| "num_tokens": 567659.0, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07608695652173914, | |
| "grad_norm": 3.180745867673301, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 1.188, | |
| "num_tokens": 665757.0, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08695652173913043, | |
| "grad_norm": 2.756203802259368, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 1.2036, | |
| "num_tokens": 764351.0, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09782608695652174, | |
| "grad_norm": 2.722957346536631, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 1.1352, | |
| "num_tokens": 859181.0, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.10869565217391304, | |
| "grad_norm": 3.149236707148637, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 1.0399, | |
| "num_tokens": 958891.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11956521739130435, | |
| "grad_norm": 2.5830391272833104, | |
| "learning_rate": 1.1578947368421053e-05, | |
| "loss": 1.0534, | |
| "num_tokens": 1057964.0, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.13043478260869565, | |
| "grad_norm": 2.1386887474211687, | |
| "learning_rate": 1.263157894736842e-05, | |
| "loss": 0.9946, | |
| "num_tokens": 1150907.0, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.14130434782608695, | |
| "grad_norm": 2.211121746514346, | |
| "learning_rate": 1.3684210526315791e-05, | |
| "loss": 1.0225, | |
| "num_tokens": 1238881.0, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15217391304347827, | |
| "grad_norm": 2.0317494078733, | |
| "learning_rate": 1.4736842105263159e-05, | |
| "loss": 0.9779, | |
| "num_tokens": 1336961.0, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16304347826086957, | |
| "grad_norm": 1.6535140053442847, | |
| "learning_rate": 1.578947368421053e-05, | |
| "loss": 0.9645, | |
| "num_tokens": 1427193.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 3.7204342856811095, | |
| "learning_rate": 1.6842105263157896e-05, | |
| "loss": 0.9152, | |
| "num_tokens": 1519751.0, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.18478260869565216, | |
| "grad_norm": 1.5902509836923946, | |
| "learning_rate": 1.7894736842105264e-05, | |
| "loss": 0.9191, | |
| "num_tokens": 1616777.0, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1956521739130435, | |
| "grad_norm": 1.4483857314150967, | |
| "learning_rate": 1.894736842105263e-05, | |
| "loss": 0.8791, | |
| "num_tokens": 1714839.0, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.20652173913043478, | |
| "grad_norm": 1.3039225121757552, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9332, | |
| "num_tokens": 1815763.0, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.21739130434782608, | |
| "grad_norm": 1.3126973712776373, | |
| "learning_rate": 1.9878787878787878e-05, | |
| "loss": 0.8782, | |
| "num_tokens": 1908693.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.22826086956521738, | |
| "grad_norm": 1.4053049528153356, | |
| "learning_rate": 1.975757575757576e-05, | |
| "loss": 0.9232, | |
| "num_tokens": 2004172.0, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2391304347826087, | |
| "grad_norm": 1.1849138803900574, | |
| "learning_rate": 1.963636363636364e-05, | |
| "loss": 0.873, | |
| "num_tokens": 2119371.0, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.01085287355782, | |
| "learning_rate": 1.9515151515151515e-05, | |
| "loss": 0.9279, | |
| "num_tokens": 2212596.0, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 1.2458008758973527, | |
| "learning_rate": 1.9393939393939395e-05, | |
| "loss": 0.8654, | |
| "num_tokens": 2294656.0, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2717391304347826, | |
| "grad_norm": 1.338715786372464, | |
| "learning_rate": 1.9272727272727275e-05, | |
| "loss": 0.8992, | |
| "num_tokens": 2391442.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2826086956521739, | |
| "grad_norm": 1.0044343113961611, | |
| "learning_rate": 1.9151515151515152e-05, | |
| "loss": 0.8313, | |
| "num_tokens": 2494467.0, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.29347826086956524, | |
| "grad_norm": 1.1561251486922257, | |
| "learning_rate": 1.9030303030303032e-05, | |
| "loss": 0.8516, | |
| "num_tokens": 2584063.0, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.30434782608695654, | |
| "grad_norm": 1.132844292406038, | |
| "learning_rate": 1.8909090909090912e-05, | |
| "loss": 0.7765, | |
| "num_tokens": 2670487.0, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.31521739130434784, | |
| "grad_norm": 1.0830282801682392, | |
| "learning_rate": 1.8787878787878792e-05, | |
| "loss": 0.855, | |
| "num_tokens": 2762813.0, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.32608695652173914, | |
| "grad_norm": 1.0350686000005194, | |
| "learning_rate": 1.866666666666667e-05, | |
| "loss": 0.8527, | |
| "num_tokens": 2850331.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33695652173913043, | |
| "grad_norm": 1.0369499924518195, | |
| "learning_rate": 1.8545454545454545e-05, | |
| "loss": 0.7974, | |
| "num_tokens": 2942596.0, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 1.1454834580125053, | |
| "learning_rate": 1.8424242424242425e-05, | |
| "loss": 0.8509, | |
| "num_tokens": 3039852.0, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.358695652173913, | |
| "grad_norm": 1.0966923132147153, | |
| "learning_rate": 1.8303030303030305e-05, | |
| "loss": 0.8426, | |
| "num_tokens": 3136550.0, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3695652173913043, | |
| "grad_norm": 0.9615713216664558, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 0.8209, | |
| "num_tokens": 3226573.0, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3804347826086957, | |
| "grad_norm": 1.0463868635910665, | |
| "learning_rate": 1.8060606060606062e-05, | |
| "loss": 0.8322, | |
| "num_tokens": 3319993.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.391304347826087, | |
| "grad_norm": 0.9016694267436932, | |
| "learning_rate": 1.7939393939393942e-05, | |
| "loss": 0.8494, | |
| "num_tokens": 3409783.0, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.40217391304347827, | |
| "grad_norm": 0.8120321544484976, | |
| "learning_rate": 1.781818181818182e-05, | |
| "loss": 0.8246, | |
| "num_tokens": 3508366.0, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.41304347826086957, | |
| "grad_norm": 0.7992720431772786, | |
| "learning_rate": 1.76969696969697e-05, | |
| "loss": 0.7703, | |
| "num_tokens": 3602876.0, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.42391304347826086, | |
| "grad_norm": 0.7908788366960535, | |
| "learning_rate": 1.7575757575757576e-05, | |
| "loss": 0.8249, | |
| "num_tokens": 3697241.0, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 0.8154728597756308, | |
| "learning_rate": 1.7454545454545456e-05, | |
| "loss": 0.825, | |
| "num_tokens": 3783861.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.44565217391304346, | |
| "grad_norm": 0.8244532763608124, | |
| "learning_rate": 1.7333333333333336e-05, | |
| "loss": 0.845, | |
| "num_tokens": 3884088.0, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.45652173913043476, | |
| "grad_norm": 0.7694222272972445, | |
| "learning_rate": 1.7212121212121212e-05, | |
| "loss": 0.7804, | |
| "num_tokens": 3972883.0, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4673913043478261, | |
| "grad_norm": 0.7461120769896388, | |
| "learning_rate": 1.7090909090909092e-05, | |
| "loss": 0.7936, | |
| "num_tokens": 4065958.0, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4782608695652174, | |
| "grad_norm": 0.7808686279876842, | |
| "learning_rate": 1.6969696969696972e-05, | |
| "loss": 0.8452, | |
| "num_tokens": 4164566.0, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4891304347826087, | |
| "grad_norm": 0.7343804102094451, | |
| "learning_rate": 1.684848484848485e-05, | |
| "loss": 0.8031, | |
| "num_tokens": 4254036.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.7439703254330587, | |
| "learning_rate": 1.672727272727273e-05, | |
| "loss": 0.8043, | |
| "num_tokens": 4340424.0, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5108695652173914, | |
| "grad_norm": 0.7396028883918819, | |
| "learning_rate": 1.660606060606061e-05, | |
| "loss": 0.7837, | |
| "num_tokens": 4436870.0, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 0.7608846409786219, | |
| "learning_rate": 1.6484848484848486e-05, | |
| "loss": 0.8515, | |
| "num_tokens": 4530847.0, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.532608695652174, | |
| "grad_norm": 0.7181352448640346, | |
| "learning_rate": 1.6363636363636366e-05, | |
| "loss": 0.8198, | |
| "num_tokens": 4624892.0, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5434782608695652, | |
| "grad_norm": 0.7087375538457228, | |
| "learning_rate": 1.6242424242424243e-05, | |
| "loss": 0.8, | |
| "num_tokens": 4726543.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5543478260869565, | |
| "grad_norm": 0.7463024608938168, | |
| "learning_rate": 1.6121212121212123e-05, | |
| "loss": 0.7789, | |
| "num_tokens": 4819956.0, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5652173913043478, | |
| "grad_norm": 0.7442588718390604, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.8054, | |
| "num_tokens": 4917035.0, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5760869565217391, | |
| "grad_norm": 0.7183312987900197, | |
| "learning_rate": 1.587878787878788e-05, | |
| "loss": 0.7923, | |
| "num_tokens": 5017078.0, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5869565217391305, | |
| "grad_norm": 0.7212235895921912, | |
| "learning_rate": 1.575757575757576e-05, | |
| "loss": 0.7816, | |
| "num_tokens": 5110982.0, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5978260869565217, | |
| "grad_norm": 0.6984711344390122, | |
| "learning_rate": 1.563636363636364e-05, | |
| "loss": 0.802, | |
| "num_tokens": 5200799.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6086956521739131, | |
| "grad_norm": 0.7262030804489741, | |
| "learning_rate": 1.5515151515151516e-05, | |
| "loss": 0.7887, | |
| "num_tokens": 5296443.0, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6195652173913043, | |
| "grad_norm": 0.7665109508827059, | |
| "learning_rate": 1.5393939393939393e-05, | |
| "loss": 0.8081, | |
| "num_tokens": 5392587.0, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6304347826086957, | |
| "grad_norm": 0.7222105761333159, | |
| "learning_rate": 1.5272727272727276e-05, | |
| "loss": 0.7779, | |
| "num_tokens": 5487722.0, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.6413043478260869, | |
| "grad_norm": 0.7297075918942776, | |
| "learning_rate": 1.5151515151515153e-05, | |
| "loss": 0.7518, | |
| "num_tokens": 5585642.0, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 0.7571906625686418, | |
| "learning_rate": 1.5030303030303031e-05, | |
| "loss": 0.7858, | |
| "num_tokens": 5675685.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6630434782608695, | |
| "grad_norm": 0.7361173105602014, | |
| "learning_rate": 1.4909090909090911e-05, | |
| "loss": 0.8163, | |
| "num_tokens": 5766441.0, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6739130434782609, | |
| "grad_norm": 0.7105038659620458, | |
| "learning_rate": 1.478787878787879e-05, | |
| "loss": 0.7505, | |
| "num_tokens": 5856978.0, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6847826086956522, | |
| "grad_norm": 0.6825484337570947, | |
| "learning_rate": 1.4666666666666666e-05, | |
| "loss": 0.7751, | |
| "num_tokens": 5941585.0, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.7314955103489276, | |
| "learning_rate": 1.4545454545454546e-05, | |
| "loss": 0.8163, | |
| "num_tokens": 6043058.0, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7065217391304348, | |
| "grad_norm": 0.7103929889225186, | |
| "learning_rate": 1.4424242424242425e-05, | |
| "loss": 0.7832, | |
| "num_tokens": 6147716.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.717391304347826, | |
| "grad_norm": 0.6793775565981599, | |
| "learning_rate": 1.4303030303030305e-05, | |
| "loss": 0.7264, | |
| "num_tokens": 6238620.0, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7282608695652174, | |
| "grad_norm": 0.7124999007971334, | |
| "learning_rate": 1.4181818181818183e-05, | |
| "loss": 0.7876, | |
| "num_tokens": 6339589.0, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7391304347826086, | |
| "grad_norm": 0.6758760592365682, | |
| "learning_rate": 1.4060606060606061e-05, | |
| "loss": 0.8137, | |
| "num_tokens": 6452148.0, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.7083257729536743, | |
| "learning_rate": 1.3939393939393942e-05, | |
| "loss": 0.7611, | |
| "num_tokens": 6546822.0, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7608695652173914, | |
| "grad_norm": 0.6894843161471015, | |
| "learning_rate": 1.381818181818182e-05, | |
| "loss": 0.8409, | |
| "num_tokens": 6652603.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7717391304347826, | |
| "grad_norm": 0.7320922145311975, | |
| "learning_rate": 1.3696969696969698e-05, | |
| "loss": 0.7523, | |
| "num_tokens": 6745015.0, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.782608695652174, | |
| "grad_norm": 0.7040695482776516, | |
| "learning_rate": 1.3575757575757578e-05, | |
| "loss": 0.7612, | |
| "num_tokens": 6838204.0, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7934782608695652, | |
| "grad_norm": 0.6847801952071528, | |
| "learning_rate": 1.3454545454545455e-05, | |
| "loss": 0.8138, | |
| "num_tokens": 6933388.0, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8043478260869565, | |
| "grad_norm": 0.7550074032673957, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.7649, | |
| "num_tokens": 7029939.0, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8152173913043478, | |
| "grad_norm": 0.6925774304998896, | |
| "learning_rate": 1.3212121212121213e-05, | |
| "loss": 0.7576, | |
| "num_tokens": 7128091.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8260869565217391, | |
| "grad_norm": 0.7562225545168595, | |
| "learning_rate": 1.3090909090909092e-05, | |
| "loss": 0.7916, | |
| "num_tokens": 7214845.0, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8369565217391305, | |
| "grad_norm": 0.6510325534353791, | |
| "learning_rate": 1.296969696969697e-05, | |
| "loss": 0.7825, | |
| "num_tokens": 7323115.0, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8478260869565217, | |
| "grad_norm": 0.7026385885074842, | |
| "learning_rate": 1.284848484848485e-05, | |
| "loss": 0.7866, | |
| "num_tokens": 7420683.0, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8586956521739131, | |
| "grad_norm": 0.7213706799641426, | |
| "learning_rate": 1.2727272727272728e-05, | |
| "loss": 0.7789, | |
| "num_tokens": 7504805.0, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 0.7056975956417906, | |
| "learning_rate": 1.2606060606060607e-05, | |
| "loss": 0.779, | |
| "num_tokens": 7594818.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8804347826086957, | |
| "grad_norm": 0.6742743459752469, | |
| "learning_rate": 1.2484848484848487e-05, | |
| "loss": 0.7664, | |
| "num_tokens": 7687306.0, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8913043478260869, | |
| "grad_norm": 0.6788480168543541, | |
| "learning_rate": 1.2363636363636364e-05, | |
| "loss": 0.7481, | |
| "num_tokens": 7787926.0, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.9021739130434783, | |
| "grad_norm": 0.7159837226779882, | |
| "learning_rate": 1.2242424242424242e-05, | |
| "loss": 0.7658, | |
| "num_tokens": 7886472.0, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9130434782608695, | |
| "grad_norm": 0.6843998675172349, | |
| "learning_rate": 1.2121212121212122e-05, | |
| "loss": 0.8125, | |
| "num_tokens": 7979894.0, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9239130434782609, | |
| "grad_norm": 0.7299584101150774, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.7764, | |
| "num_tokens": 8072266.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9347826086956522, | |
| "grad_norm": 0.6972344562839721, | |
| "learning_rate": 1.187878787878788e-05, | |
| "loss": 0.7486, | |
| "num_tokens": 8169019.0, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9456521739130435, | |
| "grad_norm": 0.695628203135494, | |
| "learning_rate": 1.1757575757575759e-05, | |
| "loss": 0.7216, | |
| "num_tokens": 8263517.0, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9565217391304348, | |
| "grad_norm": 0.7199580320946095, | |
| "learning_rate": 1.1636363636363637e-05, | |
| "loss": 0.753, | |
| "num_tokens": 8354397.0, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.967391304347826, | |
| "grad_norm": 0.688593706715687, | |
| "learning_rate": 1.1515151515151517e-05, | |
| "loss": 0.7736, | |
| "num_tokens": 8457076.0, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9782608695652174, | |
| "grad_norm": 0.7039533145498842, | |
| "learning_rate": 1.1393939393939395e-05, | |
| "loss": 0.747, | |
| "num_tokens": 8543392.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9891304347826086, | |
| "grad_norm": 0.7417219845924928, | |
| "learning_rate": 1.1272727272727272e-05, | |
| "loss": 0.8096, | |
| "num_tokens": 8634673.0, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.716787642824306, | |
| "learning_rate": 1.1151515151515154e-05, | |
| "loss": 0.7479, | |
| "num_tokens": 8737000.0, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.0108695652173914, | |
| "grad_norm": 0.9557939792897692, | |
| "learning_rate": 1.103030303030303e-05, | |
| "loss": 0.5859, | |
| "num_tokens": 8841242.0, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0217391304347827, | |
| "grad_norm": 0.9640008403759153, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 0.5499, | |
| "num_tokens": 8939461.0, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0326086956521738, | |
| "grad_norm": 0.8358505468758168, | |
| "learning_rate": 1.0787878787878789e-05, | |
| "loss": 0.5827, | |
| "num_tokens": 9041498.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 0.8333101189839462, | |
| "learning_rate": 1.0666666666666667e-05, | |
| "loss": 0.5649, | |
| "num_tokens": 9135657.0, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0543478260869565, | |
| "grad_norm": 0.8267054342321523, | |
| "learning_rate": 1.0545454545454546e-05, | |
| "loss": 0.4877, | |
| "num_tokens": 9223861.0, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.065217391304348, | |
| "grad_norm": 0.9083108087266245, | |
| "learning_rate": 1.0424242424242426e-05, | |
| "loss": 0.5178, | |
| "num_tokens": 9306683.0, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0760869565217392, | |
| "grad_norm": 1.0509887548129675, | |
| "learning_rate": 1.0303030303030304e-05, | |
| "loss": 0.5147, | |
| "num_tokens": 9403981.0, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0869565217391304, | |
| "grad_norm": 0.9617465165511758, | |
| "learning_rate": 1.0181818181818182e-05, | |
| "loss": 0.5772, | |
| "num_tokens": 9510065.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0978260869565217, | |
| "grad_norm": 0.8289148974326247, | |
| "learning_rate": 1.0060606060606062e-05, | |
| "loss": 0.4808, | |
| "num_tokens": 9594692.0, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.108695652173913, | |
| "grad_norm": 0.7858587143009836, | |
| "learning_rate": 9.939393939393939e-06, | |
| "loss": 0.4966, | |
| "num_tokens": 9686151.0, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.1195652173913044, | |
| "grad_norm": 0.7571756541563724, | |
| "learning_rate": 9.81818181818182e-06, | |
| "loss": 0.5566, | |
| "num_tokens": 9785432.0, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1304347826086956, | |
| "grad_norm": 0.7871226692453429, | |
| "learning_rate": 9.696969696969698e-06, | |
| "loss": 0.5247, | |
| "num_tokens": 9872855.0, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.141304347826087, | |
| "grad_norm": 0.7536149334148949, | |
| "learning_rate": 9.575757575757576e-06, | |
| "loss": 0.5696, | |
| "num_tokens": 9966227.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1521739130434783, | |
| "grad_norm": 0.7674793225688019, | |
| "learning_rate": 9.454545454545456e-06, | |
| "loss": 0.5085, | |
| "num_tokens": 10066104.0, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1630434782608696, | |
| "grad_norm": 0.6536542113768675, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.4929, | |
| "num_tokens": 10163479.0, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.1739130434782608, | |
| "grad_norm": 0.705223538854848, | |
| "learning_rate": 9.212121212121213e-06, | |
| "loss": 0.5041, | |
| "num_tokens": 10256103.0, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.184782608695652, | |
| "grad_norm": 0.7303409820035304, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 0.5068, | |
| "num_tokens": 10346350.0, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1956521739130435, | |
| "grad_norm": 0.669830511929235, | |
| "learning_rate": 8.969696969696971e-06, | |
| "loss": 0.5058, | |
| "num_tokens": 10446545.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2065217391304348, | |
| "grad_norm": 0.6616180902943235, | |
| "learning_rate": 8.84848484848485e-06, | |
| "loss": 0.5212, | |
| "num_tokens": 10543987.0, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.2173913043478262, | |
| "grad_norm": 0.7140780588825608, | |
| "learning_rate": 8.727272727272728e-06, | |
| "loss": 0.5542, | |
| "num_tokens": 10637982.0, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2282608695652173, | |
| "grad_norm": 0.6797454711812516, | |
| "learning_rate": 8.606060606060606e-06, | |
| "loss": 0.5146, | |
| "num_tokens": 10736295.0, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2391304347826086, | |
| "grad_norm": 0.6970790226374188, | |
| "learning_rate": 8.484848484848486e-06, | |
| "loss": 0.4858, | |
| "num_tokens": 10838089.0, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.7455739079700027, | |
| "learning_rate": 8.363636363636365e-06, | |
| "loss": 0.4863, | |
| "num_tokens": 10924837.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2608695652173914, | |
| "grad_norm": 0.6801436530929313, | |
| "learning_rate": 8.242424242424243e-06, | |
| "loss": 0.5271, | |
| "num_tokens": 11021046.0, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.2717391304347827, | |
| "grad_norm": 0.6727217360986417, | |
| "learning_rate": 8.121212121212121e-06, | |
| "loss": 0.5162, | |
| "num_tokens": 11117240.0, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.2826086956521738, | |
| "grad_norm": 0.7023642689024054, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4953, | |
| "num_tokens": 11210211.0, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2934782608695652, | |
| "grad_norm": 0.6447357115502464, | |
| "learning_rate": 7.87878787878788e-06, | |
| "loss": 0.5144, | |
| "num_tokens": 11303329.0, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.3043478260869565, | |
| "grad_norm": 0.6492607932024856, | |
| "learning_rate": 7.757575757575758e-06, | |
| "loss": 0.5034, | |
| "num_tokens": 11394106.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.315217391304348, | |
| "grad_norm": 0.6515181308514346, | |
| "learning_rate": 7.636363636363638e-06, | |
| "loss": 0.535, | |
| "num_tokens": 11490446.0, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.3260869565217392, | |
| "grad_norm": 0.672198656598516, | |
| "learning_rate": 7.515151515151516e-06, | |
| "loss": 0.5065, | |
| "num_tokens": 11581978.0, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3369565217391304, | |
| "grad_norm": 0.6741993377594862, | |
| "learning_rate": 7.393939393939395e-06, | |
| "loss": 0.4874, | |
| "num_tokens": 11665181.0, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.3478260869565217, | |
| "grad_norm": 0.6125558254135477, | |
| "learning_rate": 7.272727272727273e-06, | |
| "loss": 0.5178, | |
| "num_tokens": 11764989.0, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.358695652173913, | |
| "grad_norm": 0.6452975872864739, | |
| "learning_rate": 7.151515151515152e-06, | |
| "loss": 0.5048, | |
| "num_tokens": 11860837.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.3695652173913042, | |
| "grad_norm": 0.6942100514879066, | |
| "learning_rate": 7.030303030303031e-06, | |
| "loss": 0.5337, | |
| "num_tokens": 11949817.0, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.3804347826086958, | |
| "grad_norm": 0.6803177844769924, | |
| "learning_rate": 6.90909090909091e-06, | |
| "loss": 0.5114, | |
| "num_tokens": 12046147.0, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.6485232956728539, | |
| "learning_rate": 6.787878787878789e-06, | |
| "loss": 0.5136, | |
| "num_tokens": 12155046.0, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.4021739130434783, | |
| "grad_norm": 0.6399901029660079, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.4899, | |
| "num_tokens": 12245026.0, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4130434782608696, | |
| "grad_norm": 0.6778428524940225, | |
| "learning_rate": 6.545454545454546e-06, | |
| "loss": 0.5007, | |
| "num_tokens": 12339918.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4239130434782608, | |
| "grad_norm": 0.6506595090929513, | |
| "learning_rate": 6.424242424242425e-06, | |
| "loss": 0.5123, | |
| "num_tokens": 12431057.0, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.434782608695652, | |
| "grad_norm": 0.6042493738676704, | |
| "learning_rate": 6.303030303030303e-06, | |
| "loss": 0.5028, | |
| "num_tokens": 12528957.0, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4456521739130435, | |
| "grad_norm": 0.6655662091714003, | |
| "learning_rate": 6.181818181818182e-06, | |
| "loss": 0.5159, | |
| "num_tokens": 12630846.0, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4565217391304348, | |
| "grad_norm": 0.6581649207568613, | |
| "learning_rate": 6.060606060606061e-06, | |
| "loss": 0.5468, | |
| "num_tokens": 12734120.0, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4673913043478262, | |
| "grad_norm": 0.6792172891456207, | |
| "learning_rate": 5.93939393939394e-06, | |
| "loss": 0.515, | |
| "num_tokens": 12825758.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.4782608695652173, | |
| "grad_norm": 0.6486207741620927, | |
| "learning_rate": 5.8181818181818185e-06, | |
| "loss": 0.4768, | |
| "num_tokens": 12915706.0, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.4891304347826086, | |
| "grad_norm": 0.6229517910332318, | |
| "learning_rate": 5.696969696969698e-06, | |
| "loss": 0.5253, | |
| "num_tokens": 13006745.0, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.6931899962995233, | |
| "learning_rate": 5.575757575757577e-06, | |
| "loss": 0.518, | |
| "num_tokens": 13094837.0, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5108695652173914, | |
| "grad_norm": 0.6790435623873506, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 0.5225, | |
| "num_tokens": 13186663.0, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5217391304347827, | |
| "grad_norm": 0.6370361345960944, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.5281, | |
| "num_tokens": 13275324.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5326086956521738, | |
| "grad_norm": 0.6988457221148147, | |
| "learning_rate": 5.212121212121213e-06, | |
| "loss": 0.4471, | |
| "num_tokens": 13354804.0, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.5434782608695652, | |
| "grad_norm": 0.6267926090215254, | |
| "learning_rate": 5.090909090909091e-06, | |
| "loss": 0.5081, | |
| "num_tokens": 13459278.0, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.5543478260869565, | |
| "grad_norm": 0.6034410178549162, | |
| "learning_rate": 4.9696969696969696e-06, | |
| "loss": 0.4722, | |
| "num_tokens": 13549338.0, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.5652173913043477, | |
| "grad_norm": 0.7074387736221774, | |
| "learning_rate": 4.848484848484849e-06, | |
| "loss": 0.5115, | |
| "num_tokens": 13642023.0, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5760869565217392, | |
| "grad_norm": 0.625505288129055, | |
| "learning_rate": 4.727272727272728e-06, | |
| "loss": 0.4854, | |
| "num_tokens": 13734577.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.5869565217391304, | |
| "grad_norm": 0.6337610272860343, | |
| "learning_rate": 4.606060606060606e-06, | |
| "loss": 0.5218, | |
| "num_tokens": 13827853.0, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5978260869565217, | |
| "grad_norm": 0.5960020890354443, | |
| "learning_rate": 4.4848484848484855e-06, | |
| "loss": 0.4842, | |
| "num_tokens": 13926421.0, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.608695652173913, | |
| "grad_norm": 0.6099226198445001, | |
| "learning_rate": 4.363636363636364e-06, | |
| "loss": 0.5045, | |
| "num_tokens": 14027160.0, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6195652173913042, | |
| "grad_norm": 0.6403047778481467, | |
| "learning_rate": 4.242424242424243e-06, | |
| "loss": 0.5112, | |
| "num_tokens": 14123445.0, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.6304347826086958, | |
| "grad_norm": 0.6428678963084046, | |
| "learning_rate": 4.1212121212121215e-06, | |
| "loss": 0.4792, | |
| "num_tokens": 14213841.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.641304347826087, | |
| "grad_norm": 0.682404747047711, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.4825, | |
| "num_tokens": 14299391.0, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.6521739130434783, | |
| "grad_norm": 0.6392379137193949, | |
| "learning_rate": 3.878787878787879e-06, | |
| "loss": 0.5237, | |
| "num_tokens": 14406803.0, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.6630434782608696, | |
| "grad_norm": 0.6336927405705557, | |
| "learning_rate": 3.757575757575758e-06, | |
| "loss": 0.4933, | |
| "num_tokens": 14495074.0, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6739130434782608, | |
| "grad_norm": 0.613723947203496, | |
| "learning_rate": 3.6363636363636366e-06, | |
| "loss": 0.4987, | |
| "num_tokens": 14583035.0, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6847826086956523, | |
| "grad_norm": 0.6454282865195516, | |
| "learning_rate": 3.5151515151515154e-06, | |
| "loss": 0.4844, | |
| "num_tokens": 14675739.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.6956521739130435, | |
| "grad_norm": 0.5744835923909174, | |
| "learning_rate": 3.3939393939393946e-06, | |
| "loss": 0.5221, | |
| "num_tokens": 14771295.0, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.7065217391304348, | |
| "grad_norm": 0.6513316019714479, | |
| "learning_rate": 3.272727272727273e-06, | |
| "loss": 0.5196, | |
| "num_tokens": 14868828.0, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7173913043478262, | |
| "grad_norm": 0.6051344138316549, | |
| "learning_rate": 3.1515151515151517e-06, | |
| "loss": 0.5399, | |
| "num_tokens": 14962321.0, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.7282608695652173, | |
| "grad_norm": 0.5997374654265664, | |
| "learning_rate": 3.0303030303030305e-06, | |
| "loss": 0.566, | |
| "num_tokens": 15071474.0, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.6172378822692283, | |
| "learning_rate": 2.9090909090909093e-06, | |
| "loss": 0.5205, | |
| "num_tokens": 15172326.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.6790771644844449, | |
| "learning_rate": 2.7878787878787885e-06, | |
| "loss": 0.5405, | |
| "num_tokens": 15263118.0, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.7608695652173914, | |
| "grad_norm": 0.6250126730731829, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.5358, | |
| "num_tokens": 15366482.0, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.7717391304347827, | |
| "grad_norm": 0.616414057408985, | |
| "learning_rate": 2.5454545454545456e-06, | |
| "loss": 0.5124, | |
| "num_tokens": 15466519.0, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.7826086956521738, | |
| "grad_norm": 0.6196261922311431, | |
| "learning_rate": 2.4242424242424244e-06, | |
| "loss": 0.5011, | |
| "num_tokens": 15568813.0, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.7934782608695652, | |
| "grad_norm": 0.6012492351790968, | |
| "learning_rate": 2.303030303030303e-06, | |
| "loss": 0.4866, | |
| "num_tokens": 15663551.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.8043478260869565, | |
| "grad_norm": 0.6093660563249755, | |
| "learning_rate": 2.181818181818182e-06, | |
| "loss": 0.4806, | |
| "num_tokens": 15751970.0, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.8152173913043477, | |
| "grad_norm": 0.5733134815640453, | |
| "learning_rate": 2.0606060606060607e-06, | |
| "loss": 0.5342, | |
| "num_tokens": 15851852.0, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.8260869565217392, | |
| "grad_norm": 0.5629607438246448, | |
| "learning_rate": 1.9393939393939395e-06, | |
| "loss": 0.4899, | |
| "num_tokens": 15953339.0, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.8369565217391304, | |
| "grad_norm": 0.5705735862059056, | |
| "learning_rate": 1.8181818181818183e-06, | |
| "loss": 0.5279, | |
| "num_tokens": 16049218.0, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.8478260869565217, | |
| "grad_norm": 0.6346619776519064, | |
| "learning_rate": 1.6969696969696973e-06, | |
| "loss": 0.5129, | |
| "num_tokens": 16148821.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.858695652173913, | |
| "grad_norm": 0.6442961172463525, | |
| "learning_rate": 1.5757575757575759e-06, | |
| "loss": 0.4936, | |
| "num_tokens": 16236603.0, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.8695652173913042, | |
| "grad_norm": 0.6198075225575228, | |
| "learning_rate": 1.4545454545454546e-06, | |
| "loss": 0.5051, | |
| "num_tokens": 16331317.0, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.8804347826086958, | |
| "grad_norm": 0.5952174959924715, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 0.5081, | |
| "num_tokens": 16435421.0, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.891304347826087, | |
| "grad_norm": 0.608740415655138, | |
| "learning_rate": 1.2121212121212122e-06, | |
| "loss": 0.4985, | |
| "num_tokens": 16525942.0, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.9021739130434783, | |
| "grad_norm": 0.6186344271280536, | |
| "learning_rate": 1.090909090909091e-06, | |
| "loss": 0.4966, | |
| "num_tokens": 16612253.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "grad_norm": 0.6128028556805487, | |
| "learning_rate": 9.696969696969698e-07, | |
| "loss": 0.4677, | |
| "num_tokens": 16695763.0, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9239130434782608, | |
| "grad_norm": 0.6186139521919142, | |
| "learning_rate": 8.484848484848486e-07, | |
| "loss": 0.5063, | |
| "num_tokens": 16792394.0, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.9347826086956523, | |
| "grad_norm": 0.621118150443857, | |
| "learning_rate": 7.272727272727273e-07, | |
| "loss": 0.4859, | |
| "num_tokens": 16887893.0, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.9456521739130435, | |
| "grad_norm": 0.63455296339652, | |
| "learning_rate": 6.060606060606061e-07, | |
| "loss": 0.5323, | |
| "num_tokens": 16978736.0, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.9565217391304348, | |
| "grad_norm": 0.6338215890082622, | |
| "learning_rate": 4.848484848484849e-07, | |
| "loss": 0.5063, | |
| "num_tokens": 17069163.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.9673913043478262, | |
| "grad_norm": 0.6122769745533587, | |
| "learning_rate": 3.6363636363636366e-07, | |
| "loss": 0.5523, | |
| "num_tokens": 17170696.0, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.9782608695652173, | |
| "grad_norm": 0.5640059310251839, | |
| "learning_rate": 2.4242424242424244e-07, | |
| "loss": 0.5126, | |
| "num_tokens": 17276399.0, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.9891304347826086, | |
| "grad_norm": 0.67719515967601, | |
| "learning_rate": 1.2121212121212122e-07, | |
| "loss": 0.549, | |
| "num_tokens": 17377036.0, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5303896052463619, | |
| "learning_rate": 0.0, | |
| "loss": 0.476, | |
| "num_tokens": 17486096.0, | |
| "step": 184 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 184, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 198078707531776.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |