| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 4.990948709353001, |
| "eval_steps": 500, |
| "global_step": 930, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0053637277908146165, |
| "grad_norm": 6.338560672252144, |
| "learning_rate": 8.60215053763441e-07, |
| "loss": 1.0751, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.010727455581629233, |
| "grad_norm": 6.34459310995922, |
| "learning_rate": 1.720430107526882e-06, |
| "loss": 1.0724, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.01609118337244385, |
| "grad_norm": 6.24204412097311, |
| "learning_rate": 2.580645161290323e-06, |
| "loss": 1.0656, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.021454911163258466, |
| "grad_norm": 5.836679159124573, |
| "learning_rate": 3.440860215053764e-06, |
| "loss": 1.0643, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.02681863895407308, |
| "grad_norm": 4.516135892252434, |
| "learning_rate": 4.3010752688172045e-06, |
| "loss": 1.0254, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0321823667448877, |
| "grad_norm": 2.6133078802773384, |
| "learning_rate": 5.161290322580646e-06, |
| "loss": 0.9857, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.03754609453570231, |
| "grad_norm": 2.356431435775483, |
| "learning_rate": 6.021505376344087e-06, |
| "loss": 0.9684, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.04290982232651693, |
| "grad_norm": 3.8129269226867466, |
| "learning_rate": 6.881720430107528e-06, |
| "loss": 0.9487, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.048273550117331546, |
| "grad_norm": 3.837805004980831, |
| "learning_rate": 7.741935483870968e-06, |
| "loss": 0.958, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.05363727790814616, |
| "grad_norm": 3.226513127256882, |
| "learning_rate": 8.602150537634409e-06, |
| "loss": 0.9352, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.05900100569896078, |
| "grad_norm": 3.1817577843556006, |
| "learning_rate": 9.46236559139785e-06, |
| "loss": 0.8991, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.0643647334897754, |
| "grad_norm": 2.734625287133208, |
| "learning_rate": 1.0322580645161291e-05, |
| "loss": 0.8769, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.06972846128059002, |
| "grad_norm": 1.9702603165063668, |
| "learning_rate": 1.118279569892473e-05, |
| "loss": 0.8565, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.07509218907140462, |
| "grad_norm": 1.4326323487408281, |
| "learning_rate": 1.2043010752688173e-05, |
| "loss": 0.8304, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.08045591686221924, |
| "grad_norm": 1.5771781448342905, |
| "learning_rate": 1.2903225806451613e-05, |
| "loss": 0.8223, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.08581964465303386, |
| "grad_norm": 1.2675059911437987, |
| "learning_rate": 1.3763440860215056e-05, |
| "loss": 0.8036, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.09118337244384847, |
| "grad_norm": 0.998013780742343, |
| "learning_rate": 1.4623655913978497e-05, |
| "loss": 0.7903, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.09654710023466309, |
| "grad_norm": 1.0123573361429459, |
| "learning_rate": 1.5483870967741936e-05, |
| "loss": 0.7909, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.10191082802547771, |
| "grad_norm": 0.867381543747644, |
| "learning_rate": 1.6344086021505377e-05, |
| "loss": 0.7807, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.10727455581629232, |
| "grad_norm": 0.8142913834018638, |
| "learning_rate": 1.7204301075268818e-05, |
| "loss": 0.7675, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.11263828360710694, |
| "grad_norm": 0.848005958837766, |
| "learning_rate": 1.806451612903226e-05, |
| "loss": 0.7644, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.11800201139792156, |
| "grad_norm": 0.7118287539298433, |
| "learning_rate": 1.89247311827957e-05, |
| "loss": 0.7569, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.12336573918873617, |
| "grad_norm": 0.5855004670255447, |
| "learning_rate": 1.978494623655914e-05, |
| "loss": 0.7487, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.1287294669795508, |
| "grad_norm": 0.6728204797382786, |
| "learning_rate": 2.0645161290322582e-05, |
| "loss": 0.7416, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.1340931947703654, |
| "grad_norm": 0.5967666869348331, |
| "learning_rate": 2.1505376344086024e-05, |
| "loss": 0.7363, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.13945692256118003, |
| "grad_norm": 0.6160085099956011, |
| "learning_rate": 2.236559139784946e-05, |
| "loss": 0.734, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.14482065035199462, |
| "grad_norm": 0.6363894242487178, |
| "learning_rate": 2.3225806451612906e-05, |
| "loss": 0.7262, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.15018437814280924, |
| "grad_norm": 0.6626486374605112, |
| "learning_rate": 2.4086021505376347e-05, |
| "loss": 0.7188, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.15554810593362386, |
| "grad_norm": 0.7590807753214363, |
| "learning_rate": 2.4946236559139788e-05, |
| "loss": 0.7223, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.16091183372443849, |
| "grad_norm": 0.8043401688063679, |
| "learning_rate": 2.5806451612903226e-05, |
| "loss": 0.7165, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.1662755615152531, |
| "grad_norm": 0.8116558016551599, |
| "learning_rate": 2.6666666666666667e-05, |
| "loss": 0.7128, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.17163928930606773, |
| "grad_norm": 0.7050286092608573, |
| "learning_rate": 2.752688172043011e-05, |
| "loss": 0.709, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.17700301709688232, |
| "grad_norm": 0.605916542924061, |
| "learning_rate": 2.8387096774193552e-05, |
| "loss": 0.7035, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.18236674488769694, |
| "grad_norm": 0.48538643613649285, |
| "learning_rate": 2.9247311827956993e-05, |
| "loss": 0.7025, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.18773047267851156, |
| "grad_norm": 0.4038151178490983, |
| "learning_rate": 3.010752688172043e-05, |
| "loss": 0.7043, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.19309420046932618, |
| "grad_norm": 0.5220993966401706, |
| "learning_rate": 3.096774193548387e-05, |
| "loss": 0.6972, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.1984579282601408, |
| "grad_norm": 1.1471863644219147, |
| "learning_rate": 3.182795698924731e-05, |
| "loss": 0.6992, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.20382165605095542, |
| "grad_norm": 11.637743206594976, |
| "learning_rate": 3.2688172043010754e-05, |
| "loss": 0.7018, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.20918538384177002, |
| "grad_norm": 2.51589445225404, |
| "learning_rate": 3.3548387096774195e-05, |
| "loss": 0.7004, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.21454911163258464, |
| "grad_norm": 0.8841029393673814, |
| "learning_rate": 3.4408602150537636e-05, |
| "loss": 0.6959, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.21991283942339926, |
| "grad_norm": 2.0189356445033613, |
| "learning_rate": 3.526881720430108e-05, |
| "loss": 0.6962, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.22527656721421388, |
| "grad_norm": 0.8438418075337472, |
| "learning_rate": 3.612903225806452e-05, |
| "loss": 0.6948, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.2306402950050285, |
| "grad_norm": 2.280049552210539, |
| "learning_rate": 3.698924731182796e-05, |
| "loss": 0.691, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.23600402279584312, |
| "grad_norm": 1.5063081335495634, |
| "learning_rate": 3.78494623655914e-05, |
| "loss": 0.6911, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.24136775058665771, |
| "grad_norm": 1.8084395582302646, |
| "learning_rate": 3.870967741935484e-05, |
| "loss": 0.6906, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.24673147837747234, |
| "grad_norm": 1.6588924960337472, |
| "learning_rate": 3.956989247311828e-05, |
| "loss": 0.6929, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.25209520616828696, |
| "grad_norm": 1.014045775957985, |
| "learning_rate": 4.0430107526881724e-05, |
| "loss": 0.6908, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.2574589339591016, |
| "grad_norm": 1.9198362282460184, |
| "learning_rate": 4.1290322580645165e-05, |
| "loss": 0.687, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.2628226617499162, |
| "grad_norm": 1.744156574216378, |
| "learning_rate": 4.2150537634408606e-05, |
| "loss": 0.6804, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.2681863895407308, |
| "grad_norm": 1.1668744438759826, |
| "learning_rate": 4.301075268817205e-05, |
| "loss": 0.6784, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.27355011733154544, |
| "grad_norm": 2.3338212759533854, |
| "learning_rate": 4.387096774193548e-05, |
| "loss": 0.6808, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.27891384512236006, |
| "grad_norm": 1.7244150150466557, |
| "learning_rate": 4.473118279569892e-05, |
| "loss": 0.6838, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.2842775729131747, |
| "grad_norm": 2.419331402256656, |
| "learning_rate": 4.559139784946237e-05, |
| "loss": 0.68, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.28964130070398925, |
| "grad_norm": 4.265232664271197, |
| "learning_rate": 4.645161290322581e-05, |
| "loss": 0.6877, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.29500502849480387, |
| "grad_norm": 1.6428312936958347, |
| "learning_rate": 4.731182795698925e-05, |
| "loss": 0.6866, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.3003687562856185, |
| "grad_norm": 1.0611731524926487, |
| "learning_rate": 4.8172043010752693e-05, |
| "loss": 0.6722, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.3057324840764331, |
| "grad_norm": 1.7504655396715425, |
| "learning_rate": 4.9032258064516135e-05, |
| "loss": 0.6823, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.31109621186724773, |
| "grad_norm": 1.4266527434683252, |
| "learning_rate": 4.9892473118279576e-05, |
| "loss": 0.678, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.31645993965806235, |
| "grad_norm": 1.3715035848857164, |
| "learning_rate": 5.075268817204302e-05, |
| "loss": 0.6669, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.32182366744887697, |
| "grad_norm": 1.4341866180855634, |
| "learning_rate": 5.161290322580645e-05, |
| "loss": 0.6809, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.3271873952396916, |
| "grad_norm": 1.5209173251883614, |
| "learning_rate": 5.247311827956989e-05, |
| "loss": 0.6694, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.3325511230305062, |
| "grad_norm": 1.3054315394599638, |
| "learning_rate": 5.333333333333333e-05, |
| "loss": 0.6637, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.33791485082132083, |
| "grad_norm": 1.5264153975136474, |
| "learning_rate": 5.4193548387096774e-05, |
| "loss": 0.6726, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.34327857861213545, |
| "grad_norm": 1.2283408808572964, |
| "learning_rate": 5.505376344086022e-05, |
| "loss": 0.6746, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.3486423064029501, |
| "grad_norm": 1.4653304361520723, |
| "learning_rate": 5.591397849462366e-05, |
| "loss": 0.6734, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.35400603419376464, |
| "grad_norm": 1.1322311838365562, |
| "learning_rate": 5.6774193548387104e-05, |
| "loss": 0.6665, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.35936976198457926, |
| "grad_norm": 1.4758546005558055, |
| "learning_rate": 5.7634408602150545e-05, |
| "loss": 0.6696, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.3647334897753939, |
| "grad_norm": 1.1524101733290173, |
| "learning_rate": 5.8494623655913986e-05, |
| "loss": 0.6698, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.3700972175662085, |
| "grad_norm": 1.7469295396906894, |
| "learning_rate": 5.935483870967743e-05, |
| "loss": 0.6709, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.3754609453570231, |
| "grad_norm": 1.1846563978993125, |
| "learning_rate": 6.021505376344086e-05, |
| "loss": 0.6673, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.38082467314783774, |
| "grad_norm": 1.1459305938713673, |
| "learning_rate": 6.10752688172043e-05, |
| "loss": 0.6575, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.38618840093865237, |
| "grad_norm": 1.8209165087943502, |
| "learning_rate": 6.193548387096774e-05, |
| "loss": 0.6742, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.391552128729467, |
| "grad_norm": 1.111422678506876, |
| "learning_rate": 6.279569892473119e-05, |
| "loss": 0.669, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.3969158565202816, |
| "grad_norm": 1.8170583952074655, |
| "learning_rate": 6.365591397849463e-05, |
| "loss": 0.6677, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.4022795843110962, |
| "grad_norm": 1.1599786495072175, |
| "learning_rate": 6.451612903225807e-05, |
| "loss": 0.6666, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.40764331210191085, |
| "grad_norm": 1.4841350723375708, |
| "learning_rate": 6.537634408602151e-05, |
| "loss": 0.6707, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.41300703989272547, |
| "grad_norm": 1.0958715407789212, |
| "learning_rate": 6.623655913978495e-05, |
| "loss": 0.6593, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.41837076768354003, |
| "grad_norm": 1.4849874395428133, |
| "learning_rate": 6.709677419354839e-05, |
| "loss": 0.6658, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.42373449547435466, |
| "grad_norm": 1.3922476398945494, |
| "learning_rate": 6.795698924731183e-05, |
| "loss": 0.6627, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.4290982232651693, |
| "grad_norm": 1.349253537020458, |
| "learning_rate": 6.881720430107527e-05, |
| "loss": 0.6651, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.4344619510559839, |
| "grad_norm": 1.214908872484667, |
| "learning_rate": 6.967741935483871e-05, |
| "loss": 0.667, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.4398256788467985, |
| "grad_norm": 1.345794151853099, |
| "learning_rate": 7.053763440860215e-05, |
| "loss": 0.663, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.44518940663761314, |
| "grad_norm": 1.8118348713875871, |
| "learning_rate": 7.13978494623656e-05, |
| "loss": 0.6627, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.45055313442842776, |
| "grad_norm": 1.106435642838999, |
| "learning_rate": 7.225806451612904e-05, |
| "loss": 0.6583, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.4559168622192424, |
| "grad_norm": 1.4776781772073297, |
| "learning_rate": 7.311827956989248e-05, |
| "loss": 0.6647, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.461280590010057, |
| "grad_norm": 1.0848264815933542, |
| "learning_rate": 7.397849462365592e-05, |
| "loss": 0.6557, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.4666443178008716, |
| "grad_norm": 1.8994900378306065, |
| "learning_rate": 7.483870967741936e-05, |
| "loss": 0.6561, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.47200804559168624, |
| "grad_norm": 1.1008594371177927, |
| "learning_rate": 7.56989247311828e-05, |
| "loss": 0.6541, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.47737177338250086, |
| "grad_norm": 1.7741666055604592, |
| "learning_rate": 7.655913978494624e-05, |
| "loss": 0.6532, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.48273550117331543, |
| "grad_norm": 1.1479464223196045, |
| "learning_rate": 7.741935483870968e-05, |
| "loss": 0.6639, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.48809922896413005, |
| "grad_norm": 1.3988914997706596, |
| "learning_rate": 7.827956989247312e-05, |
| "loss": 0.6511, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.49346295675494467, |
| "grad_norm": 1.2477219744293453, |
| "learning_rate": 7.913978494623657e-05, |
| "loss": 0.6605, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.4988266845457593, |
| "grad_norm": 0.9097553948609914, |
| "learning_rate": 8e-05, |
| "loss": 0.65, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.5041904123365739, |
| "grad_norm": 1.5836301376704878, |
| "learning_rate": 7.999971824066397e-05, |
| "loss": 0.6542, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.5095541401273885, |
| "grad_norm": 1.4923647833076734, |
| "learning_rate": 7.99988729666253e-05, |
| "loss": 0.6531, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.5149178679182032, |
| "grad_norm": 0.9849105190117686, |
| "learning_rate": 7.999746418979217e-05, |
| "loss": 0.643, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.5202815957090178, |
| "grad_norm": 1.012040139532396, |
| "learning_rate": 7.99954919300114e-05, |
| "loss": 0.6485, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.5256453234998324, |
| "grad_norm": 1.3986606888790023, |
| "learning_rate": 7.999295621506808e-05, |
| "loss": 0.6549, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.531009051290647, |
| "grad_norm": 1.1230453443468684, |
| "learning_rate": 7.998985708068532e-05, |
| "loss": 0.6459, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.5363727790814616, |
| "grad_norm": 1.3637771512073018, |
| "learning_rate": 7.998619457052362e-05, |
| "loss": 0.6398, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.5417365068722763, |
| "grad_norm": 1.11658331412697, |
| "learning_rate": 7.998196873618028e-05, |
| "loss": 0.6486, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.5471002346630909, |
| "grad_norm": 1.2713507835255546, |
| "learning_rate": 7.997717963718872e-05, |
| "loss": 0.6429, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.5524639624539055, |
| "grad_norm": 1.2063346906341978, |
| "learning_rate": 7.997182734101763e-05, |
| "loss": 0.6453, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.5578276902447201, |
| "grad_norm": 1.5786954989331352, |
| "learning_rate": 7.996591192306995e-05, |
| "loss": 0.6481, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.5631914180355347, |
| "grad_norm": 2.037493924714199, |
| "learning_rate": 7.99594334666819e-05, |
| "loss": 0.6452, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.5685551458263494, |
| "grad_norm": 1.076036499954452, |
| "learning_rate": 7.995239206312176e-05, |
| "loss": 0.6366, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.573918873617164, |
| "grad_norm": 2.8730059630882607, |
| "learning_rate": 7.994478781158861e-05, |
| "loss": 0.6585, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.5792826014079785, |
| "grad_norm": 2.526449754135261, |
| "learning_rate": 7.993662081921086e-05, |
| "loss": 0.6701, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.5846463291987931, |
| "grad_norm": 1.4700621099416509, |
| "learning_rate": 7.992789120104486e-05, |
| "loss": 0.6466, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.5900100569896077, |
| "grad_norm": 1.3276483909541417, |
| "learning_rate": 7.991859908007314e-05, |
| "loss": 0.6535, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.5953737847804224, |
| "grad_norm": 0.9621323964779972, |
| "learning_rate": 7.990874458720283e-05, |
| "loss": 0.6381, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.600737512571237, |
| "grad_norm": 1.7905082277492617, |
| "learning_rate": 7.989832786126369e-05, |
| "loss": 0.6477, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.6061012403620516, |
| "grad_norm": 0.7687516563695098, |
| "learning_rate": 7.98873490490062e-05, |
| "loss": 0.6512, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.6114649681528662, |
| "grad_norm": 1.9114078648875095, |
| "learning_rate": 7.987580830509949e-05, |
| "loss": 0.6532, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.6168286959436808, |
| "grad_norm": 1.1288630311317658, |
| "learning_rate": 7.986370579212921e-05, |
| "loss": 0.6521, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.6221924237344955, |
| "grad_norm": 1.7375799757138244, |
| "learning_rate": 7.985104168059514e-05, |
| "loss": 0.6482, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.6275561515253101, |
| "grad_norm": 1.4087756071428208, |
| "learning_rate": 7.983781614890886e-05, |
| "loss": 0.6542, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.6329198793161247, |
| "grad_norm": 1.4278272442020752, |
| "learning_rate": 7.982402938339123e-05, |
| "loss": 0.6467, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.6382836071069393, |
| "grad_norm": 0.9613841064478965, |
| "learning_rate": 7.980968157826976e-05, |
| "loss": 0.6392, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.6436473348977539, |
| "grad_norm": 1.4516378583492695, |
| "learning_rate": 7.97947729356758e-05, |
| "loss": 0.6405, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.6490110626885686, |
| "grad_norm": 0.8053425444626354, |
| "learning_rate": 7.977930366564188e-05, |
| "loss": 0.6413, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.6543747904793832, |
| "grad_norm": 1.0531976112785613, |
| "learning_rate": 7.976327398609851e-05, |
| "loss": 0.6371, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.6597385182701978, |
| "grad_norm": 0.7167882513685747, |
| "learning_rate": 7.974668412287133e-05, |
| "loss": 0.6363, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.6651022460610124, |
| "grad_norm": 0.9378177990169265, |
| "learning_rate": 7.972953430967773e-05, |
| "loss": 0.6341, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.670465973851827, |
| "grad_norm": 0.8552565042049741, |
| "learning_rate": 7.971182478812374e-05, |
| "loss": 0.6324, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.6758297016426417, |
| "grad_norm": 0.7453958012562631, |
| "learning_rate": 7.96935558077005e-05, |
| "loss": 0.63, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.6811934294334563, |
| "grad_norm": 0.814810784183303, |
| "learning_rate": 7.967472762578082e-05, |
| "loss": 0.6363, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.6865571572242709, |
| "grad_norm": 1.048879669741169, |
| "learning_rate": 7.965534050761548e-05, |
| "loss": 0.6316, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.6919208850150855, |
| "grad_norm": 1.5279961244780853, |
| "learning_rate": 7.963539472632956e-05, |
| "loss": 0.6455, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.6972846128059002, |
| "grad_norm": 0.7981994579669985, |
| "learning_rate": 7.961489056291858e-05, |
| "loss": 0.633, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.7026483405967148, |
| "grad_norm": 1.0473944191586428, |
| "learning_rate": 7.95938283062445e-05, |
| "loss": 0.631, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.7080120683875293, |
| "grad_norm": 1.2737207552833516, |
| "learning_rate": 7.957220825303168e-05, |
| "loss": 0.6318, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.7133757961783439, |
| "grad_norm": 0.8894810359025763, |
| "learning_rate": 7.955003070786275e-05, |
| "loss": 0.6269, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.7187395239691585, |
| "grad_norm": 1.122051061656217, |
| "learning_rate": 7.95272959831742e-05, |
| "loss": 0.6337, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.7241032517599731, |
| "grad_norm": 0.7525451166224594, |
| "learning_rate": 7.950400439925207e-05, |
| "loss": 0.6335, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.7294669795507878, |
| "grad_norm": 0.7016690417729449, |
| "learning_rate": 7.948015628422745e-05, |
| "loss": 0.6286, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.7348307073416024, |
| "grad_norm": 0.6789098247390296, |
| "learning_rate": 7.945575197407177e-05, |
| "loss": 0.6329, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.740194435132417, |
| "grad_norm": 0.6404674159758782, |
| "learning_rate": 7.943079181259215e-05, |
| "loss": 0.6287, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.7455581629232316, |
| "grad_norm": 0.9093960160044203, |
| "learning_rate": 7.940527615142653e-05, |
| "loss": 0.6289, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.7509218907140462, |
| "grad_norm": 1.1758965813037474, |
| "learning_rate": 7.937920535003866e-05, |
| "loss": 0.6351, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.7562856185048609, |
| "grad_norm": 0.742370532155386, |
| "learning_rate": 7.935257977571317e-05, |
| "loss": 0.6261, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.7616493462956755, |
| "grad_norm": 0.542139953471737, |
| "learning_rate": 7.932539980355023e-05, |
| "loss": 0.6268, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.7670130740864901, |
| "grad_norm": 0.6206950110046128, |
| "learning_rate": 7.92976658164604e-05, |
| "loss": 0.6202, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.7723768018773047, |
| "grad_norm": 0.729311584778343, |
| "learning_rate": 7.926937820515918e-05, |
| "loss": 0.6212, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.7777405296681194, |
| "grad_norm": 0.654400750098185, |
| "learning_rate": 7.924053736816148e-05, |
| "loss": 0.6182, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.783104257458934, |
| "grad_norm": 0.6195531403475248, |
| "learning_rate": 7.921114371177607e-05, |
| "loss": 0.6245, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.7884679852497486, |
| "grad_norm": 0.7976556334248974, |
| "learning_rate": 7.918119765009979e-05, |
| "loss": 0.6249, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.7938317130405632, |
| "grad_norm": 0.8396887583061823, |
| "learning_rate": 7.915069960501177e-05, |
| "loss": 0.6248, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.7991954408313778, |
| "grad_norm": 0.8502364179472985, |
| "learning_rate": 7.911965000616746e-05, |
| "loss": 0.6194, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.8045591686221925, |
| "grad_norm": 0.9898063147614041, |
| "learning_rate": 7.908804929099256e-05, |
| "loss": 0.62, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.8099228964130071, |
| "grad_norm": 1.2455473972915074, |
| "learning_rate": 7.905589790467694e-05, |
| "loss": 0.6198, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.8152866242038217, |
| "grad_norm": 0.6806078960215027, |
| "learning_rate": 7.902319630016822e-05, |
| "loss": 0.6162, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.8206503519946363, |
| "grad_norm": 0.4476574519688243, |
| "learning_rate": 7.898994493816553e-05, |
| "loss": 0.6214, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.8260140797854509, |
| "grad_norm": 0.8696643453540888, |
| "learning_rate": 7.895614428711296e-05, |
| "loss": 0.6148, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.8313778075762656, |
| "grad_norm": 1.246010486407204, |
| "learning_rate": 7.892179482319297e-05, |
| "loss": 0.6231, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.8367415353670801, |
| "grad_norm": 0.5040497252873622, |
| "learning_rate": 7.888689703031963e-05, |
| "loss": 0.6192, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.8421052631578947, |
| "grad_norm": 0.5624852670118361, |
| "learning_rate": 7.885145140013192e-05, |
| "loss": 0.6173, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.8474689909487093, |
| "grad_norm": 0.9368458860548771, |
| "learning_rate": 7.88154584319867e-05, |
| "loss": 0.6137, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.8528327187395239, |
| "grad_norm": 0.8845383469044678, |
| "learning_rate": 7.87789186329517e-05, |
| "loss": 0.6182, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.8581964465303386, |
| "grad_norm": 0.5894895395765093, |
| "learning_rate": 7.87418325177984e-05, |
| "loss": 0.6091, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.8635601743211532, |
| "grad_norm": 0.5731381526470704, |
| "learning_rate": 7.870420060899476e-05, |
| "loss": 0.6187, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.8689239021119678, |
| "grad_norm": 0.7549017840945705, |
| "learning_rate": 7.866602343669785e-05, |
| "loss": 0.6203, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.8742876299027824, |
| "grad_norm": 0.962856257611003, |
| "learning_rate": 7.862730153874642e-05, |
| "loss": 0.6176, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.879651357693597, |
| "grad_norm": 1.3065614325608164, |
| "learning_rate": 7.858803546065328e-05, |
| "loss": 0.6247, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.8850150854844117, |
| "grad_norm": 0.7514931987981934, |
| "learning_rate": 7.854822575559764e-05, |
| "loss": 0.6167, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.8903788132752263, |
| "grad_norm": 0.6171596328035264, |
| "learning_rate": 7.85078729844173e-05, |
| "loss": 0.6191, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.8957425410660409, |
| "grad_norm": 0.7129708826504616, |
| "learning_rate": 7.846697771560075e-05, |
| "loss": 0.6176, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.9011062688568555, |
| "grad_norm": 1.0448469281639305, |
| "learning_rate": 7.842554052527918e-05, |
| "loss": 0.6186, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.9064699966476701, |
| "grad_norm": 0.8929655231467754, |
| "learning_rate": 7.838356199721836e-05, |
| "loss": 0.6123, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.9118337244384848, |
| "grad_norm": 0.6547241466981105, |
| "learning_rate": 7.834104272281041e-05, |
| "loss": 0.6157, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.9171974522292994, |
| "grad_norm": 0.820018666142171, |
| "learning_rate": 7.829798330106544e-05, |
| "loss": 0.6151, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.922561180020114, |
| "grad_norm": 0.8525569097143616, |
| "learning_rate": 7.825438433860314e-05, |
| "loss": 0.6227, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.9279249078109286, |
| "grad_norm": 0.7094520065383435, |
| "learning_rate": 7.821024644964429e-05, |
| "loss": 0.6123, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.9332886356017432, |
| "grad_norm": 0.5885877988974711, |
| "learning_rate": 7.816557025600196e-05, |
| "loss": 0.6042, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.9386523633925579, |
| "grad_norm": 0.610585149507006, |
| "learning_rate": 7.81203563870729e-05, |
| "loss": 0.6115, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.9440160911833725, |
| "grad_norm": 0.4909861004310254, |
| "learning_rate": 7.807460547982861e-05, |
| "loss": 0.6136, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.9493798189741871, |
| "grad_norm": 0.552485503833097, |
| "learning_rate": 7.802831817880633e-05, |
| "loss": 0.607, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.9547435467650017, |
| "grad_norm": 0.7187876946145139, |
| "learning_rate": 7.798149513610003e-05, |
| "loss": 0.6081, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.9601072745558162, |
| "grad_norm": 0.6564549638726072, |
| "learning_rate": 7.793413701135119e-05, |
| "loss": 0.6092, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.9654710023466309, |
| "grad_norm": 0.7342350726235318, |
| "learning_rate": 7.788624447173948e-05, |
| "loss": 0.6082, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.9708347301374455, |
| "grad_norm": 0.6410201096556201, |
| "learning_rate": 7.783781819197341e-05, |
| "loss": 0.6098, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.9761984579282601, |
| "grad_norm": 0.7068555927099052, |
| "learning_rate": 7.778885885428082e-05, |
| "loss": 0.6159, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.9815621857190747, |
| "grad_norm": 0.972774444436176, |
| "learning_rate": 7.773936714839922e-05, |
| "loss": 0.6077, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.9869259135098893, |
| "grad_norm": 0.982305769799067, |
| "learning_rate": 7.768934377156613e-05, |
| "loss": 0.6074, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.992289641300704, |
| "grad_norm": 0.792349928717434, |
| "learning_rate": 7.76387894285092e-05, |
| "loss": 0.6071, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.9976533690915186, |
| "grad_norm": 0.6150582773472084, |
| "learning_rate": 7.758770483143634e-05, |
| "loss": 0.6055, |
| "step": 186 |
| }, |
| { |
| "epoch": 1.003687562856185, |
| "grad_norm": 1.0092690602500856, |
| "learning_rate": 7.75360907000257e-05, |
| "loss": 1.0004, |
| "step": 187 |
| }, |
| { |
| "epoch": 1.0090512906469997, |
| "grad_norm": 1.4743324999618854, |
| "learning_rate": 7.748394776141539e-05, |
| "loss": 0.5946, |
| "step": 188 |
| }, |
| { |
| "epoch": 1.0144150184378142, |
| "grad_norm": 0.6277281735974461, |
| "learning_rate": 7.743127675019344e-05, |
| "loss": 0.5838, |
| "step": 189 |
| }, |
| { |
| "epoch": 1.019778746228629, |
| "grad_norm": 1.448463639411485, |
| "learning_rate": 7.737807840838728e-05, |
| "loss": 0.5989, |
| "step": 190 |
| }, |
| { |
| "epoch": 1.0251424740194435, |
| "grad_norm": 0.853892200703028, |
| "learning_rate": 7.732435348545341e-05, |
| "loss": 0.5831, |
| "step": 191 |
| }, |
| { |
| "epoch": 1.0305062018102582, |
| "grad_norm": 0.9983332294058066, |
| "learning_rate": 7.727010273826674e-05, |
| "loss": 0.5929, |
| "step": 192 |
| }, |
| { |
| "epoch": 1.0358699296010727, |
| "grad_norm": 0.828704229898241, |
| "learning_rate": 7.721532693111002e-05, |
| "loss": 0.5829, |
| "step": 193 |
| }, |
| { |
| "epoch": 1.0412336573918874, |
| "grad_norm": 0.6753173564927042, |
| "learning_rate": 7.716002683566298e-05, |
| "loss": 0.5876, |
| "step": 194 |
| }, |
| { |
| "epoch": 1.046597385182702, |
| "grad_norm": 0.7363007051587145, |
| "learning_rate": 7.710420323099151e-05, |
| "loss": 0.5817, |
| "step": 195 |
| }, |
| { |
| "epoch": 1.0519611129735167, |
| "grad_norm": 0.6189597764944422, |
| "learning_rate": 7.704785690353674e-05, |
| "loss": 0.5813, |
| "step": 196 |
| }, |
| { |
| "epoch": 1.0573248407643312, |
| "grad_norm": 0.548504465316847, |
| "learning_rate": 7.699098864710385e-05, |
| "loss": 0.5833, |
| "step": 197 |
| }, |
| { |
| "epoch": 1.062688568555146, |
| "grad_norm": 0.5803042076698839, |
| "learning_rate": 7.693359926285095e-05, |
| "loss": 0.5862, |
| "step": 198 |
| }, |
| { |
| "epoch": 1.0680522963459604, |
| "grad_norm": 0.43453807837921604, |
| "learning_rate": 7.687568955927776e-05, |
| "loss": 0.5786, |
| "step": 199 |
| }, |
| { |
| "epoch": 1.0734160241367752, |
| "grad_norm": 0.5542433696373542, |
| "learning_rate": 7.681726035221428e-05, |
| "loss": 0.5712, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.0787797519275897, |
| "grad_norm": 0.616342629969921, |
| "learning_rate": 7.675831246480923e-05, |
| "loss": 0.5765, |
| "step": 201 |
| }, |
| { |
| "epoch": 1.0841434797184042, |
| "grad_norm": 0.5795433672277299, |
| "learning_rate": 7.66988467275185e-05, |
| "loss": 0.5833, |
| "step": 202 |
| }, |
| { |
| "epoch": 1.089507207509219, |
| "grad_norm": 0.4501970786991801, |
| "learning_rate": 7.663886397809341e-05, |
| "loss": 0.5791, |
| "step": 203 |
| }, |
| { |
| "epoch": 1.0948709353000334, |
| "grad_norm": 0.4353910474176125, |
| "learning_rate": 7.657836506156896e-05, |
| "loss": 0.5741, |
| "step": 204 |
| }, |
| { |
| "epoch": 1.1002346630908482, |
| "grad_norm": 0.3477330860396474, |
| "learning_rate": 7.651735083025187e-05, |
| "loss": 0.5753, |
| "step": 205 |
| }, |
| { |
| "epoch": 1.1055983908816627, |
| "grad_norm": 0.3477663464279302, |
| "learning_rate": 7.64558221437086e-05, |
| "loss": 0.5771, |
| "step": 206 |
| }, |
| { |
| "epoch": 1.1109621186724774, |
| "grad_norm": 0.4507396818143408, |
| "learning_rate": 7.639377986875323e-05, |
| "loss": 0.5735, |
| "step": 207 |
| }, |
| { |
| "epoch": 1.116325846463292, |
| "grad_norm": 0.3794218987883004, |
| "learning_rate": 7.63312248794353e-05, |
| "loss": 0.5776, |
| "step": 208 |
| }, |
| { |
| "epoch": 1.1216895742541066, |
| "grad_norm": 0.3574145388894302, |
| "learning_rate": 7.626815805702741e-05, |
| "loss": 0.5776, |
| "step": 209 |
| }, |
| { |
| "epoch": 1.1270533020449212, |
| "grad_norm": 0.473623038030073, |
| "learning_rate": 7.620458029001286e-05, |
| "loss": 0.5687, |
| "step": 210 |
| }, |
| { |
| "epoch": 1.1324170298357359, |
| "grad_norm": 0.513881497788833, |
| "learning_rate": 7.61404924740731e-05, |
| "loss": 0.571, |
| "step": 211 |
| }, |
| { |
| "epoch": 1.1377807576265504, |
| "grad_norm": 0.5631346609022523, |
| "learning_rate": 7.60758955120752e-05, |
| "loss": 0.5784, |
| "step": 212 |
| }, |
| { |
| "epoch": 1.1431444854173651, |
| "grad_norm": 0.6816476567358593, |
| "learning_rate": 7.601079031405899e-05, |
| "loss": 0.575, |
| "step": 213 |
| }, |
| { |
| "epoch": 1.1485082132081796, |
| "grad_norm": 0.852628392408615, |
| "learning_rate": 7.594517779722432e-05, |
| "loss": 0.5799, |
| "step": 214 |
| }, |
| { |
| "epoch": 1.1538719409989944, |
| "grad_norm": 0.8589566472396788, |
| "learning_rate": 7.587905888591818e-05, |
| "loss": 0.5774, |
| "step": 215 |
| }, |
| { |
| "epoch": 1.1592356687898089, |
| "grad_norm": 0.7183732341560192, |
| "learning_rate": 7.581243451162157e-05, |
| "loss": 0.5725, |
| "step": 216 |
| }, |
| { |
| "epoch": 1.1645993965806236, |
| "grad_norm": 0.5764553522343966, |
| "learning_rate": 7.57453056129365e-05, |
| "loss": 0.571, |
| "step": 217 |
| }, |
| { |
| "epoch": 1.1699631243714381, |
| "grad_norm": 0.4929901875121555, |
| "learning_rate": 7.567767313557262e-05, |
| "loss": 0.5757, |
| "step": 218 |
| }, |
| { |
| "epoch": 1.1753268521622529, |
| "grad_norm": 0.4723018655052622, |
| "learning_rate": 7.560953803233407e-05, |
| "loss": 0.5707, |
| "step": 219 |
| }, |
| { |
| "epoch": 1.1806905799530674, |
| "grad_norm": 0.4230581385616684, |
| "learning_rate": 7.554090126310589e-05, |
| "loss": 0.5741, |
| "step": 220 |
| }, |
| { |
| "epoch": 1.1860543077438819, |
| "grad_norm": 0.38511465865648165, |
| "learning_rate": 7.547176379484063e-05, |
| "loss": 0.5703, |
| "step": 221 |
| }, |
| { |
| "epoch": 1.1914180355346966, |
| "grad_norm": 0.3152141271515254, |
| "learning_rate": 7.540212660154462e-05, |
| "loss": 0.5746, |
| "step": 222 |
| }, |
| { |
| "epoch": 1.1967817633255113, |
| "grad_norm": 0.31639051914639366, |
| "learning_rate": 7.533199066426435e-05, |
| "loss": 0.5746, |
| "step": 223 |
| }, |
| { |
| "epoch": 1.2021454911163258, |
| "grad_norm": 0.377554126720378, |
| "learning_rate": 7.526135697107258e-05, |
| "loss": 0.5679, |
| "step": 224 |
| }, |
| { |
| "epoch": 1.2075092189071404, |
| "grad_norm": 0.3964575276420449, |
| "learning_rate": 7.519022651705441e-05, |
| "loss": 0.5703, |
| "step": 225 |
| }, |
| { |
| "epoch": 1.212872946697955, |
| "grad_norm": 0.3112698936775079, |
| "learning_rate": 7.511860030429333e-05, |
| "loss": 0.5679, |
| "step": 226 |
| }, |
| { |
| "epoch": 1.2182366744887696, |
| "grad_norm": 0.32379714073410504, |
| "learning_rate": 7.504647934185706e-05, |
| "loss": 0.5693, |
| "step": 227 |
| }, |
| { |
| "epoch": 1.2236004022795843, |
| "grad_norm": 0.4040411094185397, |
| "learning_rate": 7.497386464578329e-05, |
| "loss": 0.5731, |
| "step": 228 |
| }, |
| { |
| "epoch": 1.2289641300703988, |
| "grad_norm": 0.4908973335455194, |
| "learning_rate": 7.490075723906548e-05, |
| "loss": 0.5698, |
| "step": 229 |
| }, |
| { |
| "epoch": 1.2343278578612136, |
| "grad_norm": 0.5659873818172788, |
| "learning_rate": 7.482715815163833e-05, |
| "loss": 0.5677, |
| "step": 230 |
| }, |
| { |
| "epoch": 1.239691585652028, |
| "grad_norm": 0.6149372973578214, |
| "learning_rate": 7.475306842036336e-05, |
| "loss": 0.5686, |
| "step": 231 |
| }, |
| { |
| "epoch": 1.2450553134428428, |
| "grad_norm": 0.6411124815404463, |
| "learning_rate": 7.467848908901422e-05, |
| "loss": 0.5731, |
| "step": 232 |
| }, |
| { |
| "epoch": 1.2504190412336573, |
| "grad_norm": 0.7525181257988096, |
| "learning_rate": 7.460342120826207e-05, |
| "loss": 0.5634, |
| "step": 233 |
| }, |
| { |
| "epoch": 1.255782769024472, |
| "grad_norm": 0.8271566546622127, |
| "learning_rate": 7.452786583566072e-05, |
| "loss": 0.5741, |
| "step": 234 |
| }, |
| { |
| "epoch": 1.2611464968152866, |
| "grad_norm": 0.886416613492408, |
| "learning_rate": 7.445182403563176e-05, |
| "loss": 0.5728, |
| "step": 235 |
| }, |
| { |
| "epoch": 1.2665102246061013, |
| "grad_norm": 0.8924881545698343, |
| "learning_rate": 7.437529687944951e-05, |
| "loss": 0.5717, |
| "step": 236 |
| }, |
| { |
| "epoch": 1.2718739523969158, |
| "grad_norm": 0.7745460347324375, |
| "learning_rate": 7.429828544522604e-05, |
| "loss": 0.5722, |
| "step": 237 |
| }, |
| { |
| "epoch": 1.2772376801877305, |
| "grad_norm": 0.7338807909601145, |
| "learning_rate": 7.422079081789587e-05, |
| "loss": 0.5695, |
| "step": 238 |
| }, |
| { |
| "epoch": 1.282601407978545, |
| "grad_norm": 0.7633048091889993, |
| "learning_rate": 7.414281408920074e-05, |
| "loss": 0.5671, |
| "step": 239 |
| }, |
| { |
| "epoch": 1.2879651357693598, |
| "grad_norm": 0.7923736165710045, |
| "learning_rate": 7.40643563576742e-05, |
| "loss": 0.5678, |
| "step": 240 |
| }, |
| { |
| "epoch": 1.2933288635601743, |
| "grad_norm": 0.7486989851097655, |
| "learning_rate": 7.398541872862619e-05, |
| "loss": 0.5667, |
| "step": 241 |
| }, |
| { |
| "epoch": 1.298692591350989, |
| "grad_norm": 0.6208239838109744, |
| "learning_rate": 7.390600231412737e-05, |
| "loss": 0.5736, |
| "step": 242 |
| }, |
| { |
| "epoch": 1.3040563191418035, |
| "grad_norm": 0.7404604089133735, |
| "learning_rate": 7.382610823299359e-05, |
| "loss": 0.5727, |
| "step": 243 |
| }, |
| { |
| "epoch": 1.309420046932618, |
| "grad_norm": 0.8945885885911449, |
| "learning_rate": 7.374573761077001e-05, |
| "loss": 0.5711, |
| "step": 244 |
| }, |
| { |
| "epoch": 1.3147837747234328, |
| "grad_norm": 0.7417028846645015, |
| "learning_rate": 7.366489157971525e-05, |
| "loss": 0.5632, |
| "step": 245 |
| }, |
| { |
| "epoch": 1.3201475025142475, |
| "grad_norm": 0.40901103836642044, |
| "learning_rate": 7.358357127878557e-05, |
| "loss": 0.5643, |
| "step": 246 |
| }, |
| { |
| "epoch": 1.325511230305062, |
| "grad_norm": 0.4165629890181651, |
| "learning_rate": 7.35017778536186e-05, |
| "loss": 0.5655, |
| "step": 247 |
| }, |
| { |
| "epoch": 1.3308749580958765, |
| "grad_norm": 0.5243219329421154, |
| "learning_rate": 7.341951245651747e-05, |
| "loss": 0.5633, |
| "step": 248 |
| }, |
| { |
| "epoch": 1.3362386858866913, |
| "grad_norm": 0.4504046228772055, |
| "learning_rate": 7.333677624643431e-05, |
| "loss": 0.566, |
| "step": 249 |
| }, |
| { |
| "epoch": 1.341602413677506, |
| "grad_norm": 0.35827695239534885, |
| "learning_rate": 7.325357038895413e-05, |
| "loss": 0.5743, |
| "step": 250 |
| }, |
| { |
| "epoch": 1.3469661414683205, |
| "grad_norm": 0.32152110100309084, |
| "learning_rate": 7.316989605627825e-05, |
| "loss": 0.5629, |
| "step": 251 |
| }, |
| { |
| "epoch": 1.352329869259135, |
| "grad_norm": 0.30834448489072347, |
| "learning_rate": 7.308575442720796e-05, |
| "loss": 0.5657, |
| "step": 252 |
| }, |
| { |
| "epoch": 1.3576935970499497, |
| "grad_norm": 0.27834088197947626, |
| "learning_rate": 7.300114668712767e-05, |
| "loss": 0.5644, |
| "step": 253 |
| }, |
| { |
| "epoch": 1.3630573248407643, |
| "grad_norm": 0.2294870511600659, |
| "learning_rate": 7.291607402798843e-05, |
| "loss": 0.5663, |
| "step": 254 |
| }, |
| { |
| "epoch": 1.368421052631579, |
| "grad_norm": 0.31049794029517497, |
| "learning_rate": 7.283053764829106e-05, |
| "loss": 0.5662, |
| "step": 255 |
| }, |
| { |
| "epoch": 1.3737847804223935, |
| "grad_norm": 0.3317352484927729, |
| "learning_rate": 7.274453875306922e-05, |
| "loss": 0.5624, |
| "step": 256 |
| }, |
| { |
| "epoch": 1.3791485082132082, |
| "grad_norm": 0.2991395872386746, |
| "learning_rate": 7.265807855387251e-05, |
| "loss": 0.5671, |
| "step": 257 |
| }, |
| { |
| "epoch": 1.3845122360040227, |
| "grad_norm": 0.36719705545840825, |
| "learning_rate": 7.257115826874931e-05, |
| "loss": 0.5696, |
| "step": 258 |
| }, |
| { |
| "epoch": 1.3898759637948375, |
| "grad_norm": 0.42253987097227313, |
| "learning_rate": 7.248377912222974e-05, |
| "loss": 0.5619, |
| "step": 259 |
| }, |
| { |
| "epoch": 1.395239691585652, |
| "grad_norm": 0.45185019659051934, |
| "learning_rate": 7.239594234530831e-05, |
| "loss": 0.5635, |
| "step": 260 |
| }, |
| { |
| "epoch": 1.4006034193764667, |
| "grad_norm": 0.49545413351898454, |
| "learning_rate": 7.23076491754266e-05, |
| "loss": 0.5634, |
| "step": 261 |
| }, |
| { |
| "epoch": 1.4059671471672812, |
| "grad_norm": 0.5802415947227182, |
| "learning_rate": 7.221890085645588e-05, |
| "loss": 0.5686, |
| "step": 262 |
| }, |
| { |
| "epoch": 1.411330874958096, |
| "grad_norm": 0.6878874855186285, |
| "learning_rate": 7.212969863867953e-05, |
| "loss": 0.5648, |
| "step": 263 |
| }, |
| { |
| "epoch": 1.4166946027489105, |
| "grad_norm": 0.815358307587891, |
| "learning_rate": 7.204004377877539e-05, |
| "loss": 0.5676, |
| "step": 264 |
| }, |
| { |
| "epoch": 1.4220583305397252, |
| "grad_norm": 0.7718060228087203, |
| "learning_rate": 7.194993753979818e-05, |
| "loss": 0.567, |
| "step": 265 |
| }, |
| { |
| "epoch": 1.4274220583305397, |
| "grad_norm": 0.5699309161254399, |
| "learning_rate": 7.185938119116161e-05, |
| "loss": 0.5613, |
| "step": 266 |
| }, |
| { |
| "epoch": 1.4327857861213542, |
| "grad_norm": 0.5107074162600784, |
| "learning_rate": 7.176837600862049e-05, |
| "loss": 0.5658, |
| "step": 267 |
| }, |
| { |
| "epoch": 1.438149513912169, |
| "grad_norm": 0.6522642870784296, |
| "learning_rate": 7.167692327425282e-05, |
| "loss": 0.5717, |
| "step": 268 |
| }, |
| { |
| "epoch": 1.4435132417029837, |
| "grad_norm": 0.7567797167312894, |
| "learning_rate": 7.15850242764417e-05, |
| "loss": 0.5678, |
| "step": 269 |
| }, |
| { |
| "epoch": 1.4488769694937982, |
| "grad_norm": 0.5791144265821572, |
| "learning_rate": 7.149268030985714e-05, |
| "loss": 0.5689, |
| "step": 270 |
| }, |
| { |
| "epoch": 1.4542406972846127, |
| "grad_norm": 0.3877687625030928, |
| "learning_rate": 7.139989267543787e-05, |
| "loss": 0.5674, |
| "step": 271 |
| }, |
| { |
| "epoch": 1.4596044250754274, |
| "grad_norm": 0.5209778834900987, |
| "learning_rate": 7.130666268037303e-05, |
| "loss": 0.5599, |
| "step": 272 |
| }, |
| { |
| "epoch": 1.4649681528662422, |
| "grad_norm": 0.467485068394913, |
| "learning_rate": 7.121299163808368e-05, |
| "loss": 0.5666, |
| "step": 273 |
| }, |
| { |
| "epoch": 1.4703318806570567, |
| "grad_norm": 0.3765790034913379, |
| "learning_rate": 7.111888086820435e-05, |
| "loss": 0.5687, |
| "step": 274 |
| }, |
| { |
| "epoch": 1.4756956084478712, |
| "grad_norm": 0.4702163124649037, |
| "learning_rate": 7.102433169656445e-05, |
| "loss": 0.5653, |
| "step": 275 |
| }, |
| { |
| "epoch": 1.481059336238686, |
| "grad_norm": 0.4595483441798163, |
| "learning_rate": 7.092934545516958e-05, |
| "loss": 0.5668, |
| "step": 276 |
| }, |
| { |
| "epoch": 1.4864230640295006, |
| "grad_norm": 0.3806817840099515, |
| "learning_rate": 7.083392348218274e-05, |
| "loss": 0.5575, |
| "step": 277 |
| }, |
| { |
| "epoch": 1.4917867918203152, |
| "grad_norm": 0.2933836832087611, |
| "learning_rate": 7.073806712190551e-05, |
| "loss": 0.566, |
| "step": 278 |
| }, |
| { |
| "epoch": 1.4971505196111297, |
| "grad_norm": 0.29502601995192856, |
| "learning_rate": 7.064177772475912e-05, |
| "loss": 0.5579, |
| "step": 279 |
| }, |
| { |
| "epoch": 1.5025142474019444, |
| "grad_norm": 0.2446527849116459, |
| "learning_rate": 7.054505664726542e-05, |
| "loss": 0.5588, |
| "step": 280 |
| }, |
| { |
| "epoch": 1.5078779751927591, |
| "grad_norm": 0.23256693103187187, |
| "learning_rate": 7.044790525202772e-05, |
| "loss": 0.5575, |
| "step": 281 |
| }, |
| { |
| "epoch": 1.5132417029835736, |
| "grad_norm": 0.31883595054809466, |
| "learning_rate": 7.035032490771165e-05, |
| "loss": 0.5608, |
| "step": 282 |
| }, |
| { |
| "epoch": 1.5186054307743881, |
| "grad_norm": 0.35371535591743564, |
| "learning_rate": 7.025231698902585e-05, |
| "loss": 0.5622, |
| "step": 283 |
| }, |
| { |
| "epoch": 1.5239691585652029, |
| "grad_norm": 0.30500733849822, |
| "learning_rate": 7.015388287670264e-05, |
| "loss": 0.561, |
| "step": 284 |
| }, |
| { |
| "epoch": 1.5293328863560174, |
| "grad_norm": 0.27539890689680196, |
| "learning_rate": 7.005502395747854e-05, |
| "loss": 0.5632, |
| "step": 285 |
| }, |
| { |
| "epoch": 1.534696614146832, |
| "grad_norm": 0.3163401305287628, |
| "learning_rate": 6.995574162407471e-05, |
| "loss": 0.5647, |
| "step": 286 |
| }, |
| { |
| "epoch": 1.5400603419376466, |
| "grad_norm": 0.3537933994086675, |
| "learning_rate": 6.985603727517736e-05, |
| "loss": 0.5559, |
| "step": 287 |
| }, |
| { |
| "epoch": 1.5454240697284614, |
| "grad_norm": 0.40878100621394714, |
| "learning_rate": 6.975591231541805e-05, |
| "loss": 0.5602, |
| "step": 288 |
| }, |
| { |
| "epoch": 1.5507877975192759, |
| "grad_norm": 0.44323300216184397, |
| "learning_rate": 6.96553681553539e-05, |
| "loss": 0.5628, |
| "step": 289 |
| }, |
| { |
| "epoch": 1.5561515253100904, |
| "grad_norm": 0.4339681253670531, |
| "learning_rate": 6.955440621144766e-05, |
| "loss": 0.5622, |
| "step": 290 |
| }, |
| { |
| "epoch": 1.561515253100905, |
| "grad_norm": 0.42574808327461705, |
| "learning_rate": 6.945302790604789e-05, |
| "loss": 0.5656, |
| "step": 291 |
| }, |
| { |
| "epoch": 1.5668789808917198, |
| "grad_norm": 0.49984403935729443, |
| "learning_rate": 6.935123466736878e-05, |
| "loss": 0.5615, |
| "step": 292 |
| }, |
| { |
| "epoch": 1.5722427086825344, |
| "grad_norm": 0.6728978442736605, |
| "learning_rate": 6.92490279294701e-05, |
| "loss": 0.5612, |
| "step": 293 |
| }, |
| { |
| "epoch": 1.5776064364733489, |
| "grad_norm": 0.7429424392949961, |
| "learning_rate": 6.914640913223695e-05, |
| "loss": 0.5666, |
| "step": 294 |
| }, |
| { |
| "epoch": 1.5829701642641636, |
| "grad_norm": 0.6836320262283112, |
| "learning_rate": 6.904337972135958e-05, |
| "loss": 0.5607, |
| "step": 295 |
| }, |
| { |
| "epoch": 1.5883338920549783, |
| "grad_norm": 0.5867869403119099, |
| "learning_rate": 6.893994114831287e-05, |
| "loss": 0.5608, |
| "step": 296 |
| }, |
| { |
| "epoch": 1.5936976198457928, |
| "grad_norm": 0.5953453881284777, |
| "learning_rate": 6.883609487033605e-05, |
| "loss": 0.56, |
| "step": 297 |
| }, |
| { |
| "epoch": 1.5990613476366073, |
| "grad_norm": 0.6338055543705274, |
| "learning_rate": 6.8731842350412e-05, |
| "loss": 0.5641, |
| "step": 298 |
| }, |
| { |
| "epoch": 1.604425075427422, |
| "grad_norm": 0.6648867991249984, |
| "learning_rate": 6.862718505724677e-05, |
| "loss": 0.566, |
| "step": 299 |
| }, |
| { |
| "epoch": 1.6097888032182368, |
| "grad_norm": 0.7077186281640041, |
| "learning_rate": 6.852212446524881e-05, |
| "loss": 0.5603, |
| "step": 300 |
| }, |
| { |
| "epoch": 1.6151525310090513, |
| "grad_norm": 0.7046651305076743, |
| "learning_rate": 6.84166620545083e-05, |
| "loss": 0.5619, |
| "step": 301 |
| }, |
| { |
| "epoch": 1.6205162587998658, |
| "grad_norm": 0.5221466006404026, |
| "learning_rate": 6.831079931077615e-05, |
| "loss": 0.5561, |
| "step": 302 |
| }, |
| { |
| "epoch": 1.6258799865906806, |
| "grad_norm": 0.37944143421384124, |
| "learning_rate": 6.820453772544318e-05, |
| "loss": 0.5631, |
| "step": 303 |
| }, |
| { |
| "epoch": 1.6312437143814953, |
| "grad_norm": 0.5844863656673001, |
| "learning_rate": 6.80978787955191e-05, |
| "loss": 0.5545, |
| "step": 304 |
| }, |
| { |
| "epoch": 1.6366074421723098, |
| "grad_norm": 0.5727271056764134, |
| "learning_rate": 6.799082402361131e-05, |
| "loss": 0.5602, |
| "step": 305 |
| }, |
| { |
| "epoch": 1.6419711699631243, |
| "grad_norm": 0.3690953808626758, |
| "learning_rate": 6.788337491790397e-05, |
| "loss": 0.5613, |
| "step": 306 |
| }, |
| { |
| "epoch": 1.647334897753939, |
| "grad_norm": 0.31408506346634396, |
| "learning_rate": 6.777553299213646e-05, |
| "loss": 0.5587, |
| "step": 307 |
| }, |
| { |
| "epoch": 1.6526986255447536, |
| "grad_norm": 0.36584423383126785, |
| "learning_rate": 6.766729976558226e-05, |
| "loss": 0.5579, |
| "step": 308 |
| }, |
| { |
| "epoch": 1.658062353335568, |
| "grad_norm": 0.34138280754486466, |
| "learning_rate": 6.755867676302747e-05, |
| "loss": 0.5597, |
| "step": 309 |
| }, |
| { |
| "epoch": 1.6634260811263828, |
| "grad_norm": 0.2631015494780705, |
| "learning_rate": 6.744966551474936e-05, |
| "loss": 0.558, |
| "step": 310 |
| }, |
| { |
| "epoch": 1.6687898089171975, |
| "grad_norm": 0.34093472748141784, |
| "learning_rate": 6.734026755649474e-05, |
| "loss": 0.5619, |
| "step": 311 |
| }, |
| { |
| "epoch": 1.674153536708012, |
| "grad_norm": 0.43169917288860177, |
| "learning_rate": 6.723048442945845e-05, |
| "loss": 0.5617, |
| "step": 312 |
| }, |
| { |
| "epoch": 1.6795172644988265, |
| "grad_norm": 0.4064955294037096, |
| "learning_rate": 6.712031768026154e-05, |
| "loss": 0.562, |
| "step": 313 |
| }, |
| { |
| "epoch": 1.6848809922896413, |
| "grad_norm": 0.4126204222058468, |
| "learning_rate": 6.70097688609295e-05, |
| "loss": 0.5592, |
| "step": 314 |
| }, |
| { |
| "epoch": 1.690244720080456, |
| "grad_norm": 0.4791759638439358, |
| "learning_rate": 6.689883952887042e-05, |
| "loss": 0.5612, |
| "step": 315 |
| }, |
| { |
| "epoch": 1.6956084478712705, |
| "grad_norm": 0.5431515205787689, |
| "learning_rate": 6.678753124685306e-05, |
| "loss": 0.5653, |
| "step": 316 |
| }, |
| { |
| "epoch": 1.700972175662085, |
| "grad_norm": 0.6320848556741568, |
| "learning_rate": 6.667584558298481e-05, |
| "loss": 0.5576, |
| "step": 317 |
| }, |
| { |
| "epoch": 1.7063359034528998, |
| "grad_norm": 0.7004122845515078, |
| "learning_rate": 6.656378411068958e-05, |
| "loss": 0.5609, |
| "step": 318 |
| }, |
| { |
| "epoch": 1.7116996312437145, |
| "grad_norm": 0.6089676867039382, |
| "learning_rate": 6.645134840868566e-05, |
| "loss": 0.5647, |
| "step": 319 |
| }, |
| { |
| "epoch": 1.717063359034529, |
| "grad_norm": 0.4272876666051787, |
| "learning_rate": 6.633854006096351e-05, |
| "loss": 0.5579, |
| "step": 320 |
| }, |
| { |
| "epoch": 1.7224270868253435, |
| "grad_norm": 0.4606786107127789, |
| "learning_rate": 6.622536065676338e-05, |
| "loss": 0.5598, |
| "step": 321 |
| }, |
| { |
| "epoch": 1.7277908146161582, |
| "grad_norm": 0.5009778503199392, |
| "learning_rate": 6.611181179055296e-05, |
| "loss": 0.5581, |
| "step": 322 |
| }, |
| { |
| "epoch": 1.733154542406973, |
| "grad_norm": 0.4252208216862443, |
| "learning_rate": 6.599789506200491e-05, |
| "loss": 0.5598, |
| "step": 323 |
| }, |
| { |
| "epoch": 1.7385182701977875, |
| "grad_norm": 0.3441171701556595, |
| "learning_rate": 6.588361207597432e-05, |
| "loss": 0.5592, |
| "step": 324 |
| }, |
| { |
| "epoch": 1.743881997988602, |
| "grad_norm": 0.3879024267741338, |
| "learning_rate": 6.576896444247609e-05, |
| "loss": 0.5579, |
| "step": 325 |
| }, |
| { |
| "epoch": 1.7492457257794167, |
| "grad_norm": 0.30111558862635124, |
| "learning_rate": 6.565395377666228e-05, |
| "loss": 0.5594, |
| "step": 326 |
| }, |
| { |
| "epoch": 1.7546094535702315, |
| "grad_norm": 0.300732536138267, |
| "learning_rate": 6.553858169879934e-05, |
| "loss": 0.56, |
| "step": 327 |
| }, |
| { |
| "epoch": 1.759973181361046, |
| "grad_norm": 0.340331678535412, |
| "learning_rate": 6.542284983424528e-05, |
| "loss": 0.5558, |
| "step": 328 |
| }, |
| { |
| "epoch": 1.7653369091518605, |
| "grad_norm": 0.3902331189981612, |
| "learning_rate": 6.530675981342674e-05, |
| "loss": 0.5586, |
| "step": 329 |
| }, |
| { |
| "epoch": 1.7707006369426752, |
| "grad_norm": 0.43726753368459903, |
| "learning_rate": 6.51903132718161e-05, |
| "loss": 0.5525, |
| "step": 330 |
| }, |
| { |
| "epoch": 1.7760643647334897, |
| "grad_norm": 0.39280624763313293, |
| "learning_rate": 6.507351184990837e-05, |
| "loss": 0.555, |
| "step": 331 |
| }, |
| { |
| "epoch": 1.7814280925243042, |
| "grad_norm": 0.4556458085809101, |
| "learning_rate": 6.49563571931981e-05, |
| "loss": 0.5605, |
| "step": 332 |
| }, |
| { |
| "epoch": 1.786791820315119, |
| "grad_norm": 0.47834893977036586, |
| "learning_rate": 6.48388509521562e-05, |
| "loss": 0.5619, |
| "step": 333 |
| }, |
| { |
| "epoch": 1.7921555481059337, |
| "grad_norm": 0.37688804258392605, |
| "learning_rate": 6.47209947822067e-05, |
| "loss": 0.5591, |
| "step": 334 |
| }, |
| { |
| "epoch": 1.7975192758967482, |
| "grad_norm": 0.31185260137172, |
| "learning_rate": 6.46027903437034e-05, |
| "loss": 0.5601, |
| "step": 335 |
| }, |
| { |
| "epoch": 1.8028830036875627, |
| "grad_norm": 0.38273388647684164, |
| "learning_rate": 6.448423930190653e-05, |
| "loss": 0.5586, |
| "step": 336 |
| }, |
| { |
| "epoch": 1.8082467314783774, |
| "grad_norm": 0.3659890752118683, |
| "learning_rate": 6.43653433269592e-05, |
| "loss": 0.5593, |
| "step": 337 |
| }, |
| { |
| "epoch": 1.8136104592691922, |
| "grad_norm": 0.28697986611796433, |
| "learning_rate": 6.424610409386396e-05, |
| "loss": 0.5531, |
| "step": 338 |
| }, |
| { |
| "epoch": 1.8189741870600067, |
| "grad_norm": 0.3578108779196417, |
| "learning_rate": 6.41265232824592e-05, |
| "loss": 0.5576, |
| "step": 339 |
| }, |
| { |
| "epoch": 1.8243379148508212, |
| "grad_norm": 0.3595477545374786, |
| "learning_rate": 6.40066025773954e-05, |
| "loss": 0.5536, |
| "step": 340 |
| }, |
| { |
| "epoch": 1.829701642641636, |
| "grad_norm": 0.24495176220429252, |
| "learning_rate": 6.388634366811146e-05, |
| "loss": 0.5561, |
| "step": 341 |
| }, |
| { |
| "epoch": 1.8350653704324507, |
| "grad_norm": 0.34910710828586194, |
| "learning_rate": 6.376574824881092e-05, |
| "loss": 0.5554, |
| "step": 342 |
| }, |
| { |
| "epoch": 1.8404290982232652, |
| "grad_norm": 0.4115446924161809, |
| "learning_rate": 6.364481801843802e-05, |
| "loss": 0.556, |
| "step": 343 |
| }, |
| { |
| "epoch": 1.8457928260140797, |
| "grad_norm": 0.45492770981427666, |
| "learning_rate": 6.352355468065386e-05, |
| "loss": 0.5574, |
| "step": 344 |
| }, |
| { |
| "epoch": 1.8511565538048944, |
| "grad_norm": 0.4502627142016844, |
| "learning_rate": 6.34019599438123e-05, |
| "loss": 0.5578, |
| "step": 345 |
| }, |
| { |
| "epoch": 1.8565202815957091, |
| "grad_norm": 0.3764500404139869, |
| "learning_rate": 6.328003552093597e-05, |
| "loss": 0.5554, |
| "step": 346 |
| }, |
| { |
| "epoch": 1.8618840093865237, |
| "grad_norm": 0.3785656825858741, |
| "learning_rate": 6.315778312969208e-05, |
| "loss": 0.5551, |
| "step": 347 |
| }, |
| { |
| "epoch": 1.8672477371773382, |
| "grad_norm": 0.43002728271919993, |
| "learning_rate": 6.303520449236827e-05, |
| "loss": 0.5502, |
| "step": 348 |
| }, |
| { |
| "epoch": 1.872611464968153, |
| "grad_norm": 0.3968933759349479, |
| "learning_rate": 6.291230133584829e-05, |
| "loss": 0.5563, |
| "step": 349 |
| }, |
| { |
| "epoch": 1.8779751927589676, |
| "grad_norm": 0.3445454221288414, |
| "learning_rate": 6.278907539158775e-05, |
| "loss": 0.5593, |
| "step": 350 |
| }, |
| { |
| "epoch": 1.8833389205497821, |
| "grad_norm": 0.26807986583588095, |
| "learning_rate": 6.266552839558965e-05, |
| "loss": 0.5495, |
| "step": 351 |
| }, |
| { |
| "epoch": 1.8887026483405966, |
| "grad_norm": 0.25008221285974236, |
| "learning_rate": 6.254166208837998e-05, |
| "loss": 0.5517, |
| "step": 352 |
| }, |
| { |
| "epoch": 1.8940663761314114, |
| "grad_norm": 0.24820258665637415, |
| "learning_rate": 6.241747821498315e-05, |
| "loss": 0.5522, |
| "step": 353 |
| }, |
| { |
| "epoch": 1.899430103922226, |
| "grad_norm": 0.27580712951039277, |
| "learning_rate": 6.229297852489746e-05, |
| "loss": 0.5583, |
| "step": 354 |
| }, |
| { |
| "epoch": 1.9047938317130404, |
| "grad_norm": 0.25679911776420106, |
| "learning_rate": 6.21681647720704e-05, |
| "loss": 0.5499, |
| "step": 355 |
| }, |
| { |
| "epoch": 1.9101575595038551, |
| "grad_norm": 0.2197401652483747, |
| "learning_rate": 6.204303871487399e-05, |
| "loss": 0.5549, |
| "step": 356 |
| }, |
| { |
| "epoch": 1.9155212872946699, |
| "grad_norm": 0.2598523527517371, |
| "learning_rate": 6.191760211607995e-05, |
| "loss": 0.5581, |
| "step": 357 |
| }, |
| { |
| "epoch": 1.9208850150854844, |
| "grad_norm": 0.25865343855556444, |
| "learning_rate": 6.179185674283493e-05, |
| "loss": 0.5561, |
| "step": 358 |
| }, |
| { |
| "epoch": 1.9262487428762989, |
| "grad_norm": 0.23265157023397873, |
| "learning_rate": 6.166580436663557e-05, |
| "loss": 0.5538, |
| "step": 359 |
| }, |
| { |
| "epoch": 1.9316124706671136, |
| "grad_norm": 0.21694798938619186, |
| "learning_rate": 6.153944676330357e-05, |
| "loss": 0.5488, |
| "step": 360 |
| }, |
| { |
| "epoch": 1.9369761984579283, |
| "grad_norm": 0.2266361665583917, |
| "learning_rate": 6.141278571296064e-05, |
| "loss": 0.5479, |
| "step": 361 |
| }, |
| { |
| "epoch": 1.9423399262487429, |
| "grad_norm": 0.20869559940686064, |
| "learning_rate": 6.128582300000345e-05, |
| "loss": 0.5507, |
| "step": 362 |
| }, |
| { |
| "epoch": 1.9477036540395574, |
| "grad_norm": 0.24712363397279544, |
| "learning_rate": 6.11585604130785e-05, |
| "loss": 0.5521, |
| "step": 363 |
| }, |
| { |
| "epoch": 1.953067381830372, |
| "grad_norm": 0.2625049015072699, |
| "learning_rate": 6.103099974505689e-05, |
| "loss": 0.5538, |
| "step": 364 |
| }, |
| { |
| "epoch": 1.9584311096211868, |
| "grad_norm": 0.26045303486127896, |
| "learning_rate": 6.0903142793009055e-05, |
| "loss": 0.5521, |
| "step": 365 |
| }, |
| { |
| "epoch": 1.9637948374120013, |
| "grad_norm": 0.2729210974767953, |
| "learning_rate": 6.0774991358179505e-05, |
| "loss": 0.5545, |
| "step": 366 |
| }, |
| { |
| "epoch": 1.9691585652028158, |
| "grad_norm": 0.2661858868407906, |
| "learning_rate": 6.064654724596141e-05, |
| "loss": 0.5565, |
| "step": 367 |
| }, |
| { |
| "epoch": 1.9745222929936306, |
| "grad_norm": 0.3161992287956601, |
| "learning_rate": 6.051781226587114e-05, |
| "loss": 0.5566, |
| "step": 368 |
| }, |
| { |
| "epoch": 1.9798860207844453, |
| "grad_norm": 0.3247068065937088, |
| "learning_rate": 6.038878823152283e-05, |
| "loss": 0.5485, |
| "step": 369 |
| }, |
| { |
| "epoch": 1.9852497485752598, |
| "grad_norm": 0.31042344714670567, |
| "learning_rate": 6.0259476960602795e-05, |
| "loss": 0.5529, |
| "step": 370 |
| }, |
| { |
| "epoch": 1.9906134763660743, |
| "grad_norm": 0.28663489099284756, |
| "learning_rate": 6.012988027484392e-05, |
| "loss": 0.553, |
| "step": 371 |
| }, |
| { |
| "epoch": 1.995977204156889, |
| "grad_norm": 0.32048941108555556, |
| "learning_rate": 6.000000000000001e-05, |
| "loss": 0.5491, |
| "step": 372 |
| }, |
| { |
| "epoch": 2.0020113979215557, |
| "grad_norm": 0.5619929857332273, |
| "learning_rate": 5.9869837965820064e-05, |
| "loss": 0.9077, |
| "step": 373 |
| }, |
| { |
| "epoch": 2.00737512571237, |
| "grad_norm": 0.9276754960897458, |
| "learning_rate": 5.973939600602251e-05, |
| "loss": 0.5122, |
| "step": 374 |
| }, |
| { |
| "epoch": 2.0127388535031847, |
| "grad_norm": 1.1885282416108998, |
| "learning_rate": 5.960867595826934e-05, |
| "loss": 0.5236, |
| "step": 375 |
| }, |
| { |
| "epoch": 2.0181025812939994, |
| "grad_norm": 0.6750751948664354, |
| "learning_rate": 5.9477679664140256e-05, |
| "loss": 0.5117, |
| "step": 376 |
| }, |
| { |
| "epoch": 2.0234663090848137, |
| "grad_norm": 0.7850363466297992, |
| "learning_rate": 5.934640896910668e-05, |
| "loss": 0.5088, |
| "step": 377 |
| }, |
| { |
| "epoch": 2.0288300368756285, |
| "grad_norm": 0.9014636634663518, |
| "learning_rate": 5.921486572250582e-05, |
| "loss": 0.5097, |
| "step": 378 |
| }, |
| { |
| "epoch": 2.034193764666443, |
| "grad_norm": 0.8754469407524593, |
| "learning_rate": 5.908305177751457e-05, |
| "loss": 0.5147, |
| "step": 379 |
| }, |
| { |
| "epoch": 2.039557492457258, |
| "grad_norm": 0.4769303794801558, |
| "learning_rate": 5.89509689911234e-05, |
| "loss": 0.5087, |
| "step": 380 |
| }, |
| { |
| "epoch": 2.044921220248072, |
| "grad_norm": 0.7004101867500354, |
| "learning_rate": 5.881861922411023e-05, |
| "loss": 0.5088, |
| "step": 381 |
| }, |
| { |
| "epoch": 2.050284948038887, |
| "grad_norm": 0.518970595004128, |
| "learning_rate": 5.8686004341014175e-05, |
| "loss": 0.5106, |
| "step": 382 |
| }, |
| { |
| "epoch": 2.0556486758297017, |
| "grad_norm": 0.562641653776111, |
| "learning_rate": 5.855312621010932e-05, |
| "loss": 0.5076, |
| "step": 383 |
| }, |
| { |
| "epoch": 2.0610124036205164, |
| "grad_norm": 0.6089557991308883, |
| "learning_rate": 5.841998670337834e-05, |
| "loss": 0.5098, |
| "step": 384 |
| }, |
| { |
| "epoch": 2.0663761314113307, |
| "grad_norm": 0.3745335212494372, |
| "learning_rate": 5.828658769648621e-05, |
| "loss": 0.5107, |
| "step": 385 |
| }, |
| { |
| "epoch": 2.0717398592021454, |
| "grad_norm": 0.5038019010238222, |
| "learning_rate": 5.815293106875369e-05, |
| "loss": 0.5086, |
| "step": 386 |
| }, |
| { |
| "epoch": 2.07710358699296, |
| "grad_norm": 0.4417411581025683, |
| "learning_rate": 5.8019018703130924e-05, |
| "loss": 0.5082, |
| "step": 387 |
| }, |
| { |
| "epoch": 2.082467314783775, |
| "grad_norm": 0.4645390192932457, |
| "learning_rate": 5.788485248617088e-05, |
| "loss": 0.5088, |
| "step": 388 |
| }, |
| { |
| "epoch": 2.087831042574589, |
| "grad_norm": 0.38534363222977375, |
| "learning_rate": 5.775043430800274e-05, |
| "loss": 0.5044, |
| "step": 389 |
| }, |
| { |
| "epoch": 2.093194770365404, |
| "grad_norm": 0.3177873591397125, |
| "learning_rate": 5.761576606230538e-05, |
| "loss": 0.5037, |
| "step": 390 |
| }, |
| { |
| "epoch": 2.0985584981562186, |
| "grad_norm": 0.31327769202934297, |
| "learning_rate": 5.7480849646280536e-05, |
| "loss": 0.5065, |
| "step": 391 |
| }, |
| { |
| "epoch": 2.1039222259470334, |
| "grad_norm": 0.3096007175172773, |
| "learning_rate": 5.7345686960626216e-05, |
| "loss": 0.5089, |
| "step": 392 |
| }, |
| { |
| "epoch": 2.1092859537378477, |
| "grad_norm": 0.3250396986727379, |
| "learning_rate": 5.7210279909509846e-05, |
| "loss": 0.5075, |
| "step": 393 |
| }, |
| { |
| "epoch": 2.1146496815286624, |
| "grad_norm": 0.3335529585038257, |
| "learning_rate": 5.707463040054147e-05, |
| "loss": 0.5034, |
| "step": 394 |
| }, |
| { |
| "epoch": 2.120013409319477, |
| "grad_norm": 0.27510903683348653, |
| "learning_rate": 5.693874034474686e-05, |
| "loss": 0.5058, |
| "step": 395 |
| }, |
| { |
| "epoch": 2.125377137110292, |
| "grad_norm": 0.30416211894239636, |
| "learning_rate": 5.6802611656540605e-05, |
| "loss": 0.5052, |
| "step": 396 |
| }, |
| { |
| "epoch": 2.130740864901106, |
| "grad_norm": 0.3309159515345732, |
| "learning_rate": 5.666624625369915e-05, |
| "loss": 0.5011, |
| "step": 397 |
| }, |
| { |
| "epoch": 2.136104592691921, |
| "grad_norm": 0.2633474754073962, |
| "learning_rate": 5.652964605733378e-05, |
| "loss": 0.5031, |
| "step": 398 |
| }, |
| { |
| "epoch": 2.1414683204827356, |
| "grad_norm": 0.23995436691844568, |
| "learning_rate": 5.6392812991863505e-05, |
| "loss": 0.5013, |
| "step": 399 |
| }, |
| { |
| "epoch": 2.1468320482735503, |
| "grad_norm": 0.21205441071144687, |
| "learning_rate": 5.6255748984988026e-05, |
| "loss": 0.5026, |
| "step": 400 |
| }, |
| { |
| "epoch": 2.1521957760643646, |
| "grad_norm": 0.1978028326665556, |
| "learning_rate": 5.61184559676605e-05, |
| "loss": 0.5029, |
| "step": 401 |
| }, |
| { |
| "epoch": 2.1575595038551794, |
| "grad_norm": 0.23744700406570274, |
| "learning_rate": 5.598093587406042e-05, |
| "loss": 0.5036, |
| "step": 402 |
| }, |
| { |
| "epoch": 2.162923231645994, |
| "grad_norm": 0.22982879915535342, |
| "learning_rate": 5.584319064156628e-05, |
| "loss": 0.5007, |
| "step": 403 |
| }, |
| { |
| "epoch": 2.1682869594368084, |
| "grad_norm": 0.2332943467736946, |
| "learning_rate": 5.570522221072835e-05, |
| "loss": 0.5023, |
| "step": 404 |
| }, |
| { |
| "epoch": 2.173650687227623, |
| "grad_norm": 0.21085441964760665, |
| "learning_rate": 5.5567032525241315e-05, |
| "loss": 0.5009, |
| "step": 405 |
| }, |
| { |
| "epoch": 2.179014415018438, |
| "grad_norm": 0.22875835498415426, |
| "learning_rate": 5.542862353191686e-05, |
| "loss": 0.5004, |
| "step": 406 |
| }, |
| { |
| "epoch": 2.1843781428092526, |
| "grad_norm": 0.2508518468627133, |
| "learning_rate": 5.528999718065629e-05, |
| "loss": 0.5027, |
| "step": 407 |
| }, |
| { |
| "epoch": 2.189741870600067, |
| "grad_norm": 0.2133758915716313, |
| "learning_rate": 5.515115542442305e-05, |
| "loss": 0.5039, |
| "step": 408 |
| }, |
| { |
| "epoch": 2.1951055983908816, |
| "grad_norm": 0.21013033352445007, |
| "learning_rate": 5.501210021921518e-05, |
| "loss": 0.509, |
| "step": 409 |
| }, |
| { |
| "epoch": 2.2004693261816963, |
| "grad_norm": 0.24089413117157768, |
| "learning_rate": 5.48728335240378e-05, |
| "loss": 0.5034, |
| "step": 410 |
| }, |
| { |
| "epoch": 2.205833053972511, |
| "grad_norm": 0.2673713830759583, |
| "learning_rate": 5.47333573008755e-05, |
| "loss": 0.5003, |
| "step": 411 |
| }, |
| { |
| "epoch": 2.2111967817633253, |
| "grad_norm": 0.26775112246936006, |
| "learning_rate": 5.459367351466466e-05, |
| "loss": 0.5023, |
| "step": 412 |
| }, |
| { |
| "epoch": 2.21656050955414, |
| "grad_norm": 0.29448398829721556, |
| "learning_rate": 5.445378413326582e-05, |
| "loss": 0.5026, |
| "step": 413 |
| }, |
| { |
| "epoch": 2.221924237344955, |
| "grad_norm": 0.26959315515895454, |
| "learning_rate": 5.4313691127435975e-05, |
| "loss": 0.5037, |
| "step": 414 |
| }, |
| { |
| "epoch": 2.2272879651357695, |
| "grad_norm": 0.21092593980509475, |
| "learning_rate": 5.417339647080071e-05, |
| "loss": 0.5018, |
| "step": 415 |
| }, |
| { |
| "epoch": 2.232651692926584, |
| "grad_norm": 0.27287941846913594, |
| "learning_rate": 5.40329021398265e-05, |
| "loss": 0.504, |
| "step": 416 |
| }, |
| { |
| "epoch": 2.2380154207173986, |
| "grad_norm": 0.2243145959214538, |
| "learning_rate": 5.389221011379281e-05, |
| "loss": 0.5016, |
| "step": 417 |
| }, |
| { |
| "epoch": 2.2433791485082133, |
| "grad_norm": 0.23263330992814404, |
| "learning_rate": 5.3751322374764254e-05, |
| "loss": 0.496, |
| "step": 418 |
| }, |
| { |
| "epoch": 2.248742876299028, |
| "grad_norm": 0.2512271712538734, |
| "learning_rate": 5.361024090756259e-05, |
| "loss": 0.5035, |
| "step": 419 |
| }, |
| { |
| "epoch": 2.2541066040898423, |
| "grad_norm": 0.2551912835527355, |
| "learning_rate": 5.346896769973886e-05, |
| "loss": 0.5007, |
| "step": 420 |
| }, |
| { |
| "epoch": 2.259470331880657, |
| "grad_norm": 0.22385870672760472, |
| "learning_rate": 5.3327504741545326e-05, |
| "loss": 0.5029, |
| "step": 421 |
| }, |
| { |
| "epoch": 2.2648340596714718, |
| "grad_norm": 0.19743299945786424, |
| "learning_rate": 5.318585402590745e-05, |
| "loss": 0.5028, |
| "step": 422 |
| }, |
| { |
| "epoch": 2.270197787462286, |
| "grad_norm": 0.20277764076614224, |
| "learning_rate": 5.3044017548395804e-05, |
| "loss": 0.5005, |
| "step": 423 |
| }, |
| { |
| "epoch": 2.275561515253101, |
| "grad_norm": 0.19076747830149005, |
| "learning_rate": 5.290199730719798e-05, |
| "loss": 0.4979, |
| "step": 424 |
| }, |
| { |
| "epoch": 2.2809252430439155, |
| "grad_norm": 0.20942827686948592, |
| "learning_rate": 5.275979530309042e-05, |
| "loss": 0.5038, |
| "step": 425 |
| }, |
| { |
| "epoch": 2.2862889708347303, |
| "grad_norm": 0.21537452339336768, |
| "learning_rate": 5.2617413539410236e-05, |
| "loss": 0.5053, |
| "step": 426 |
| }, |
| { |
| "epoch": 2.291652698625545, |
| "grad_norm": 0.25627491704457783, |
| "learning_rate": 5.247485402202697e-05, |
| "loss": 0.4995, |
| "step": 427 |
| }, |
| { |
| "epoch": 2.2970164264163593, |
| "grad_norm": 0.2829522803374003, |
| "learning_rate": 5.2332118759314394e-05, |
| "loss": 0.4969, |
| "step": 428 |
| }, |
| { |
| "epoch": 2.302380154207174, |
| "grad_norm": 0.22098956424737304, |
| "learning_rate": 5.218920976212215e-05, |
| "loss": 0.504, |
| "step": 429 |
| }, |
| { |
| "epoch": 2.3077438819979887, |
| "grad_norm": 0.2239527428611037, |
| "learning_rate": 5.204612904374745e-05, |
| "loss": 0.5001, |
| "step": 430 |
| }, |
| { |
| "epoch": 2.313107609788803, |
| "grad_norm": 0.20447198338227954, |
| "learning_rate": 5.1902878619906694e-05, |
| "loss": 0.5008, |
| "step": 431 |
| }, |
| { |
| "epoch": 2.3184713375796178, |
| "grad_norm": 0.1994784390562372, |
| "learning_rate": 5.1759460508707085e-05, |
| "loss": 0.4954, |
| "step": 432 |
| }, |
| { |
| "epoch": 2.3238350653704325, |
| "grad_norm": 0.20376249784991973, |
| "learning_rate": 5.1615876730618226e-05, |
| "loss": 0.5051, |
| "step": 433 |
| }, |
| { |
| "epoch": 2.3291987931612472, |
| "grad_norm": 0.1777004640090449, |
| "learning_rate": 5.1472129308443616e-05, |
| "loss": 0.5059, |
| "step": 434 |
| }, |
| { |
| "epoch": 2.3345625209520615, |
| "grad_norm": 0.17919423701924383, |
| "learning_rate": 5.132822026729216e-05, |
| "loss": 0.4997, |
| "step": 435 |
| }, |
| { |
| "epoch": 2.3399262487428762, |
| "grad_norm": 0.1952145263768232, |
| "learning_rate": 5.118415163454968e-05, |
| "loss": 0.4964, |
| "step": 436 |
| }, |
| { |
| "epoch": 2.345289976533691, |
| "grad_norm": 0.23136811372687383, |
| "learning_rate": 5.1039925439850244e-05, |
| "loss": 0.4853, |
| "step": 437 |
| }, |
| { |
| "epoch": 2.3506537043245057, |
| "grad_norm": 0.2527914130838122, |
| "learning_rate": 5.0895543715047737e-05, |
| "loss": 0.5063, |
| "step": 438 |
| }, |
| { |
| "epoch": 2.35601743211532, |
| "grad_norm": 0.23487193932679698, |
| "learning_rate": 5.075100849418708e-05, |
| "loss": 0.5067, |
| "step": 439 |
| }, |
| { |
| "epoch": 2.3613811599061347, |
| "grad_norm": 0.17865390789281577, |
| "learning_rate": 5.060632181347568e-05, |
| "loss": 0.5051, |
| "step": 440 |
| }, |
| { |
| "epoch": 2.3667448876969495, |
| "grad_norm": 0.19380856376412295, |
| "learning_rate": 5.046148571125468e-05, |
| "loss": 0.5012, |
| "step": 441 |
| }, |
| { |
| "epoch": 2.3721086154877637, |
| "grad_norm": 0.2136191065502088, |
| "learning_rate": 5.031650222797028e-05, |
| "loss": 0.4994, |
| "step": 442 |
| }, |
| { |
| "epoch": 2.3774723432785785, |
| "grad_norm": 0.26432526505519316, |
| "learning_rate": 5.0171373406144985e-05, |
| "loss": 0.4977, |
| "step": 443 |
| }, |
| { |
| "epoch": 2.382836071069393, |
| "grad_norm": 0.27436611300421115, |
| "learning_rate": 5.002610129034883e-05, |
| "loss": 0.5019, |
| "step": 444 |
| }, |
| { |
| "epoch": 2.388199798860208, |
| "grad_norm": 0.20385491901178016, |
| "learning_rate": 4.9880687927170534e-05, |
| "loss": 0.4977, |
| "step": 445 |
| }, |
| { |
| "epoch": 2.3935635266510227, |
| "grad_norm": 0.16876671632134482, |
| "learning_rate": 4.973513536518875e-05, |
| "loss": 0.4975, |
| "step": 446 |
| }, |
| { |
| "epoch": 2.398927254441837, |
| "grad_norm": 0.1925770688555282, |
| "learning_rate": 4.958944565494314e-05, |
| "loss": 0.5, |
| "step": 447 |
| }, |
| { |
| "epoch": 2.4042909822326517, |
| "grad_norm": 0.20179672034929158, |
| "learning_rate": 4.944362084890548e-05, |
| "loss": 0.4984, |
| "step": 448 |
| }, |
| { |
| "epoch": 2.4096547100234664, |
| "grad_norm": 0.21265428318185095, |
| "learning_rate": 4.929766300145083e-05, |
| "loss": 0.4974, |
| "step": 449 |
| }, |
| { |
| "epoch": 2.4150184378142807, |
| "grad_norm": 0.25978527351861586, |
| "learning_rate": 4.915157416882849e-05, |
| "loss": 0.5055, |
| "step": 450 |
| }, |
| { |
| "epoch": 2.4203821656050954, |
| "grad_norm": 0.21775715781130206, |
| "learning_rate": 4.9005356409133044e-05, |
| "loss": 0.5016, |
| "step": 451 |
| }, |
| { |
| "epoch": 2.42574589339591, |
| "grad_norm": 0.19027697490800918, |
| "learning_rate": 4.88590117822755e-05, |
| "loss": 0.5022, |
| "step": 452 |
| }, |
| { |
| "epoch": 2.431109621186725, |
| "grad_norm": 0.22937605026076494, |
| "learning_rate": 4.871254234995406e-05, |
| "loss": 0.5041, |
| "step": 453 |
| }, |
| { |
| "epoch": 2.436473348977539, |
| "grad_norm": 0.24637066107726857, |
| "learning_rate": 4.856595017562525e-05, |
| "loss": 0.5024, |
| "step": 454 |
| }, |
| { |
| "epoch": 2.441837076768354, |
| "grad_norm": 0.21654301339170334, |
| "learning_rate": 4.8419237324474734e-05, |
| "loss": 0.499, |
| "step": 455 |
| }, |
| { |
| "epoch": 2.4472008045591687, |
| "grad_norm": 0.23644059339454185, |
| "learning_rate": 4.82724058633883e-05, |
| "loss": 0.5013, |
| "step": 456 |
| }, |
| { |
| "epoch": 2.4525645323499834, |
| "grad_norm": 0.22992449590363817, |
| "learning_rate": 4.812545786092269e-05, |
| "loss": 0.507, |
| "step": 457 |
| }, |
| { |
| "epoch": 2.4579282601407977, |
| "grad_norm": 0.23917593718094943, |
| "learning_rate": 4.7978395387276475e-05, |
| "loss": 0.5017, |
| "step": 458 |
| }, |
| { |
| "epoch": 2.4632919879316124, |
| "grad_norm": 0.20812657582063035, |
| "learning_rate": 4.783122051426093e-05, |
| "loss": 0.5019, |
| "step": 459 |
| }, |
| { |
| "epoch": 2.468655715722427, |
| "grad_norm": 0.16762169107118158, |
| "learning_rate": 4.768393531527077e-05, |
| "loss": 0.5096, |
| "step": 460 |
| }, |
| { |
| "epoch": 2.474019443513242, |
| "grad_norm": 0.21282987152563532, |
| "learning_rate": 4.753654186525497e-05, |
| "loss": 0.4999, |
| "step": 461 |
| }, |
| { |
| "epoch": 2.479383171304056, |
| "grad_norm": 0.21742448911503381, |
| "learning_rate": 4.738904224068758e-05, |
| "loss": 0.5036, |
| "step": 462 |
| }, |
| { |
| "epoch": 2.484746899094871, |
| "grad_norm": 0.1925300663006384, |
| "learning_rate": 4.724143851953841e-05, |
| "loss": 0.4977, |
| "step": 463 |
| }, |
| { |
| "epoch": 2.4901106268856856, |
| "grad_norm": 0.18087700742417484, |
| "learning_rate": 4.7093732781243777e-05, |
| "loss": 0.5079, |
| "step": 464 |
| }, |
| { |
| "epoch": 2.4954743546765004, |
| "grad_norm": 0.15639702761113433, |
| "learning_rate": 4.694592710667723e-05, |
| "loss": 0.5072, |
| "step": 465 |
| }, |
| { |
| "epoch": 2.5008380824673146, |
| "grad_norm": 0.19597235332511273, |
| "learning_rate": 4.6798023578120184e-05, |
| "loss": 0.5018, |
| "step": 466 |
| }, |
| { |
| "epoch": 2.5062018102581294, |
| "grad_norm": 0.1684316769198244, |
| "learning_rate": 4.6650024279232666e-05, |
| "loss": 0.4929, |
| "step": 467 |
| }, |
| { |
| "epoch": 2.511565538048944, |
| "grad_norm": 0.21055212869799955, |
| "learning_rate": 4.650193129502386e-05, |
| "loss": 0.5006, |
| "step": 468 |
| }, |
| { |
| "epoch": 2.5169292658397584, |
| "grad_norm": 0.24733144692822004, |
| "learning_rate": 4.635374671182283e-05, |
| "loss": 0.4994, |
| "step": 469 |
| }, |
| { |
| "epoch": 2.522292993630573, |
| "grad_norm": 0.24533597541006635, |
| "learning_rate": 4.620547261724906e-05, |
| "loss": 0.4958, |
| "step": 470 |
| }, |
| { |
| "epoch": 2.527656721421388, |
| "grad_norm": 0.2066880499616437, |
| "learning_rate": 4.605711110018307e-05, |
| "loss": 0.5005, |
| "step": 471 |
| }, |
| { |
| "epoch": 2.5330204492122026, |
| "grad_norm": 0.16991247454378994, |
| "learning_rate": 4.590866425073698e-05, |
| "loss": 0.4937, |
| "step": 472 |
| }, |
| { |
| "epoch": 2.5383841770030173, |
| "grad_norm": 0.1921709883210605, |
| "learning_rate": 4.576013416022511e-05, |
| "loss": 0.4986, |
| "step": 473 |
| }, |
| { |
| "epoch": 2.5437479047938316, |
| "grad_norm": 0.21100456607229467, |
| "learning_rate": 4.5611522921134394e-05, |
| "loss": 0.5096, |
| "step": 474 |
| }, |
| { |
| "epoch": 2.5491116325846463, |
| "grad_norm": 0.21922187130563403, |
| "learning_rate": 4.546283262709506e-05, |
| "loss": 0.4961, |
| "step": 475 |
| }, |
| { |
| "epoch": 2.554475360375461, |
| "grad_norm": 0.2138728259158541, |
| "learning_rate": 4.531406537285103e-05, |
| "loss": 0.4978, |
| "step": 476 |
| }, |
| { |
| "epoch": 2.5598390881662754, |
| "grad_norm": 0.17501395717864893, |
| "learning_rate": 4.516522325423046e-05, |
| "loss": 0.4978, |
| "step": 477 |
| }, |
| { |
| "epoch": 2.56520281595709, |
| "grad_norm": 0.2043784564828151, |
| "learning_rate": 4.5016308368116155e-05, |
| "loss": 0.5071, |
| "step": 478 |
| }, |
| { |
| "epoch": 2.570566543747905, |
| "grad_norm": 0.23306913353410968, |
| "learning_rate": 4.486732281241611e-05, |
| "loss": 0.4983, |
| "step": 479 |
| }, |
| { |
| "epoch": 2.5759302715387196, |
| "grad_norm": 0.2446931512816206, |
| "learning_rate": 4.471826868603385e-05, |
| "loss": 0.4967, |
| "step": 480 |
| }, |
| { |
| "epoch": 2.5812939993295343, |
| "grad_norm": 0.19818051787599317, |
| "learning_rate": 4.456914808883898e-05, |
| "loss": 0.5007, |
| "step": 481 |
| }, |
| { |
| "epoch": 2.5866577271203486, |
| "grad_norm": 0.16087879372692104, |
| "learning_rate": 4.4419963121637526e-05, |
| "loss": 0.4944, |
| "step": 482 |
| }, |
| { |
| "epoch": 2.5920214549111633, |
| "grad_norm": 0.20092825461048774, |
| "learning_rate": 4.427071588614236e-05, |
| "loss": 0.5002, |
| "step": 483 |
| }, |
| { |
| "epoch": 2.597385182701978, |
| "grad_norm": 0.2281785256323823, |
| "learning_rate": 4.412140848494356e-05, |
| "loss": 0.5031, |
| "step": 484 |
| }, |
| { |
| "epoch": 2.6027489104927923, |
| "grad_norm": 0.20204680377867434, |
| "learning_rate": 4.397204302147886e-05, |
| "loss": 0.496, |
| "step": 485 |
| }, |
| { |
| "epoch": 2.608112638283607, |
| "grad_norm": 0.1977652017995082, |
| "learning_rate": 4.3822621600003934e-05, |
| "loss": 0.4959, |
| "step": 486 |
| }, |
| { |
| "epoch": 2.613476366074422, |
| "grad_norm": 0.19244577039841376, |
| "learning_rate": 4.367314632556281e-05, |
| "loss": 0.4949, |
| "step": 487 |
| }, |
| { |
| "epoch": 2.618840093865236, |
| "grad_norm": 0.24680562797983854, |
| "learning_rate": 4.3523619303958196e-05, |
| "loss": 0.5001, |
| "step": 488 |
| }, |
| { |
| "epoch": 2.624203821656051, |
| "grad_norm": 0.2608000397905511, |
| "learning_rate": 4.3374042641721787e-05, |
| "loss": 0.5006, |
| "step": 489 |
| }, |
| { |
| "epoch": 2.6295675494468655, |
| "grad_norm": 0.16561810253679096, |
| "learning_rate": 4.322441844608469e-05, |
| "loss": 0.4972, |
| "step": 490 |
| }, |
| { |
| "epoch": 2.6349312772376803, |
| "grad_norm": 0.22607914890990072, |
| "learning_rate": 4.3074748824947546e-05, |
| "loss": 0.5002, |
| "step": 491 |
| }, |
| { |
| "epoch": 2.640295005028495, |
| "grad_norm": 0.2918667204791213, |
| "learning_rate": 4.292503588685105e-05, |
| "loss": 0.5, |
| "step": 492 |
| }, |
| { |
| "epoch": 2.6456587328193093, |
| "grad_norm": 0.2047093208105323, |
| "learning_rate": 4.277528174094607e-05, |
| "loss": 0.4999, |
| "step": 493 |
| }, |
| { |
| "epoch": 2.651022460610124, |
| "grad_norm": 0.1740999252433492, |
| "learning_rate": 4.262548849696407e-05, |
| "loss": 0.495, |
| "step": 494 |
| }, |
| { |
| "epoch": 2.6563861884009388, |
| "grad_norm": 0.1585431462983745, |
| "learning_rate": 4.247565826518728e-05, |
| "loss": 0.4913, |
| "step": 495 |
| }, |
| { |
| "epoch": 2.661749916191753, |
| "grad_norm": 0.18550331479957272, |
| "learning_rate": 4.2325793156419035e-05, |
| "loss": 0.4991, |
| "step": 496 |
| }, |
| { |
| "epoch": 2.667113643982568, |
| "grad_norm": 0.18017249852635744, |
| "learning_rate": 4.217589528195403e-05, |
| "loss": 0.4952, |
| "step": 497 |
| }, |
| { |
| "epoch": 2.6724773717733825, |
| "grad_norm": 0.2170267245439509, |
| "learning_rate": 4.202596675354851e-05, |
| "loss": 0.4979, |
| "step": 498 |
| }, |
| { |
| "epoch": 2.6778410995641972, |
| "grad_norm": 0.20466183171637414, |
| "learning_rate": 4.187600968339064e-05, |
| "loss": 0.4977, |
| "step": 499 |
| }, |
| { |
| "epoch": 2.683204827355012, |
| "grad_norm": 0.1859091442920982, |
| "learning_rate": 4.1726026184070625e-05, |
| "loss": 0.5037, |
| "step": 500 |
| }, |
| { |
| "epoch": 2.6885685551458263, |
| "grad_norm": 0.16577581911976427, |
| "learning_rate": 4.157601836855103e-05, |
| "loss": 0.4996, |
| "step": 501 |
| }, |
| { |
| "epoch": 2.693932282936641, |
| "grad_norm": 0.20982822011955427, |
| "learning_rate": 4.142598835013698e-05, |
| "loss": 0.5003, |
| "step": 502 |
| }, |
| { |
| "epoch": 2.6992960107274557, |
| "grad_norm": 0.1760144931085274, |
| "learning_rate": 4.12759382424464e-05, |
| "loss": 0.5048, |
| "step": 503 |
| }, |
| { |
| "epoch": 2.70465973851827, |
| "grad_norm": 0.16308891153060442, |
| "learning_rate": 4.11258701593802e-05, |
| "loss": 0.4995, |
| "step": 504 |
| }, |
| { |
| "epoch": 2.7100234663090847, |
| "grad_norm": 0.18485290922951852, |
| "learning_rate": 4.0975786215092596e-05, |
| "loss": 0.4974, |
| "step": 505 |
| }, |
| { |
| "epoch": 2.7153871940998995, |
| "grad_norm": 0.16453966274325355, |
| "learning_rate": 4.0825688523961176e-05, |
| "loss": 0.4961, |
| "step": 506 |
| }, |
| { |
| "epoch": 2.7207509218907138, |
| "grad_norm": 0.17078151543088202, |
| "learning_rate": 4.0675579200557246e-05, |
| "loss": 0.4996, |
| "step": 507 |
| }, |
| { |
| "epoch": 2.7261146496815285, |
| "grad_norm": 0.18413562304090325, |
| "learning_rate": 4.052546035961596e-05, |
| "loss": 0.5047, |
| "step": 508 |
| }, |
| { |
| "epoch": 2.7314783774723432, |
| "grad_norm": 0.1807528229446923, |
| "learning_rate": 4.0375334116006596e-05, |
| "loss": 0.4999, |
| "step": 509 |
| }, |
| { |
| "epoch": 2.736842105263158, |
| "grad_norm": 0.1708195041200041, |
| "learning_rate": 4.0225202584702643e-05, |
| "loss": 0.5039, |
| "step": 510 |
| }, |
| { |
| "epoch": 2.7422058330539727, |
| "grad_norm": 0.1772112061283108, |
| "learning_rate": 4.0075067880752165e-05, |
| "loss": 0.5006, |
| "step": 511 |
| }, |
| { |
| "epoch": 2.747569560844787, |
| "grad_norm": 0.16771636982644764, |
| "learning_rate": 3.9924932119247855e-05, |
| "loss": 0.5008, |
| "step": 512 |
| }, |
| { |
| "epoch": 2.7529332886356017, |
| "grad_norm": 0.16916780256561517, |
| "learning_rate": 3.977479741529738e-05, |
| "loss": 0.4939, |
| "step": 513 |
| }, |
| { |
| "epoch": 2.7582970164264164, |
| "grad_norm": 0.17680260996970482, |
| "learning_rate": 3.962466588399342e-05, |
| "loss": 0.4997, |
| "step": 514 |
| }, |
| { |
| "epoch": 2.7636607442172307, |
| "grad_norm": 0.1780764938281788, |
| "learning_rate": 3.947453964038404e-05, |
| "loss": 0.4939, |
| "step": 515 |
| }, |
| { |
| "epoch": 2.7690244720080455, |
| "grad_norm": 0.20020918095732, |
| "learning_rate": 3.932442079944276e-05, |
| "loss": 0.4932, |
| "step": 516 |
| }, |
| { |
| "epoch": 2.77438819979886, |
| "grad_norm": 0.19304746634023665, |
| "learning_rate": 3.9174311476038824e-05, |
| "loss": 0.5026, |
| "step": 517 |
| }, |
| { |
| "epoch": 2.779751927589675, |
| "grad_norm": 0.19887309023558672, |
| "learning_rate": 3.902421378490742e-05, |
| "loss": 0.4953, |
| "step": 518 |
| }, |
| { |
| "epoch": 2.7851156553804897, |
| "grad_norm": 0.204450641639583, |
| "learning_rate": 3.887412984061979e-05, |
| "loss": 0.5001, |
| "step": 519 |
| }, |
| { |
| "epoch": 2.790479383171304, |
| "grad_norm": 0.1651139970176456, |
| "learning_rate": 3.872406175755362e-05, |
| "loss": 0.4985, |
| "step": 520 |
| }, |
| { |
| "epoch": 2.7958431109621187, |
| "grad_norm": 0.2563097667637881, |
| "learning_rate": 3.857401164986303e-05, |
| "loss": 0.4948, |
| "step": 521 |
| }, |
| { |
| "epoch": 2.8012068387529334, |
| "grad_norm": 0.234054045546983, |
| "learning_rate": 3.842398163144899e-05, |
| "loss": 0.4947, |
| "step": 522 |
| }, |
| { |
| "epoch": 2.8065705665437477, |
| "grad_norm": 0.15811375667315455, |
| "learning_rate": 3.827397381592939e-05, |
| "loss": 0.5048, |
| "step": 523 |
| }, |
| { |
| "epoch": 2.8119342943345624, |
| "grad_norm": 0.1987863698237863, |
| "learning_rate": 3.8123990316609364e-05, |
| "loss": 0.501, |
| "step": 524 |
| }, |
| { |
| "epoch": 2.817298022125377, |
| "grad_norm": 0.17172619805487405, |
| "learning_rate": 3.7974033246451496e-05, |
| "loss": 0.4991, |
| "step": 525 |
| }, |
| { |
| "epoch": 2.822661749916192, |
| "grad_norm": 0.16007263014351186, |
| "learning_rate": 3.782410471804599e-05, |
| "loss": 0.4972, |
| "step": 526 |
| }, |
| { |
| "epoch": 2.8280254777070066, |
| "grad_norm": 0.16534850292679054, |
| "learning_rate": 3.767420684358097e-05, |
| "loss": 0.4984, |
| "step": 527 |
| }, |
| { |
| "epoch": 2.833389205497821, |
| "grad_norm": 0.14773257937726625, |
| "learning_rate": 3.752434173481273e-05, |
| "loss": 0.504, |
| "step": 528 |
| }, |
| { |
| "epoch": 2.8387529332886356, |
| "grad_norm": 0.17349714217162007, |
| "learning_rate": 3.737451150303595e-05, |
| "loss": 0.4984, |
| "step": 529 |
| }, |
| { |
| "epoch": 2.8441166610794504, |
| "grad_norm": 0.15861554785434998, |
| "learning_rate": 3.722471825905394e-05, |
| "loss": 0.5046, |
| "step": 530 |
| }, |
| { |
| "epoch": 2.8494803888702647, |
| "grad_norm": 0.17356548128864052, |
| "learning_rate": 3.707496411314896e-05, |
| "loss": 0.4965, |
| "step": 531 |
| }, |
| { |
| "epoch": 2.8548441166610794, |
| "grad_norm": 0.16137459004699495, |
| "learning_rate": 3.692525117505246e-05, |
| "loss": 0.4966, |
| "step": 532 |
| }, |
| { |
| "epoch": 2.860207844451894, |
| "grad_norm": 0.1543679059675596, |
| "learning_rate": 3.677558155391532e-05, |
| "loss": 0.4969, |
| "step": 533 |
| }, |
| { |
| "epoch": 2.8655715722427084, |
| "grad_norm": 0.17840356309672536, |
| "learning_rate": 3.662595735827822e-05, |
| "loss": 0.4928, |
| "step": 534 |
| }, |
| { |
| "epoch": 2.870935300033523, |
| "grad_norm": 0.15915427976526053, |
| "learning_rate": 3.647638069604182e-05, |
| "loss": 0.491, |
| "step": 535 |
| }, |
| { |
| "epoch": 2.876299027824338, |
| "grad_norm": 0.1960858603291955, |
| "learning_rate": 3.632685367443721e-05, |
| "loss": 0.4981, |
| "step": 536 |
| }, |
| { |
| "epoch": 2.8816627556151526, |
| "grad_norm": 0.20296992977396197, |
| "learning_rate": 3.617737839999608e-05, |
| "loss": 0.4984, |
| "step": 537 |
| }, |
| { |
| "epoch": 2.8870264834059673, |
| "grad_norm": 0.2694415740935397, |
| "learning_rate": 3.602795697852116e-05, |
| "loss": 0.4907, |
| "step": 538 |
| }, |
| { |
| "epoch": 2.8923902111967816, |
| "grad_norm": 0.19651514395841568, |
| "learning_rate": 3.587859151505645e-05, |
| "loss": 0.4907, |
| "step": 539 |
| }, |
| { |
| "epoch": 2.8977539389875964, |
| "grad_norm": 0.17781434805395327, |
| "learning_rate": 3.572928411385765e-05, |
| "loss": 0.5042, |
| "step": 540 |
| }, |
| { |
| "epoch": 2.903117666778411, |
| "grad_norm": 0.20415382211791988, |
| "learning_rate": 3.558003687836249e-05, |
| "loss": 0.4973, |
| "step": 541 |
| }, |
| { |
| "epoch": 2.9084813945692254, |
| "grad_norm": 0.25361323116982154, |
| "learning_rate": 3.5430851911161025e-05, |
| "loss": 0.4957, |
| "step": 542 |
| }, |
| { |
| "epoch": 2.91384512236004, |
| "grad_norm": 0.20713936406930122, |
| "learning_rate": 3.528173131396617e-05, |
| "loss": 0.5036, |
| "step": 543 |
| }, |
| { |
| "epoch": 2.919208850150855, |
| "grad_norm": 0.17026848748787526, |
| "learning_rate": 3.513267718758391e-05, |
| "loss": 0.4975, |
| "step": 544 |
| }, |
| { |
| "epoch": 2.9245725779416696, |
| "grad_norm": 0.21622589822789534, |
| "learning_rate": 3.498369163188385e-05, |
| "loss": 0.4976, |
| "step": 545 |
| }, |
| { |
| "epoch": 2.9299363057324843, |
| "grad_norm": 0.19646465258305665, |
| "learning_rate": 3.483477674576955e-05, |
| "loss": 0.4987, |
| "step": 546 |
| }, |
| { |
| "epoch": 2.9353000335232986, |
| "grad_norm": 0.17789915328753722, |
| "learning_rate": 3.468593462714896e-05, |
| "loss": 0.4952, |
| "step": 547 |
| }, |
| { |
| "epoch": 2.9406637613141133, |
| "grad_norm": 0.16632647338954235, |
| "learning_rate": 3.453716737290495e-05, |
| "loss": 0.498, |
| "step": 548 |
| }, |
| { |
| "epoch": 2.946027489104928, |
| "grad_norm": 0.18269935667563644, |
| "learning_rate": 3.438847707886561e-05, |
| "loss": 0.4993, |
| "step": 549 |
| }, |
| { |
| "epoch": 2.9513912168957424, |
| "grad_norm": 0.1574198748294102, |
| "learning_rate": 3.4239865839774906e-05, |
| "loss": 0.4938, |
| "step": 550 |
| }, |
| { |
| "epoch": 2.956754944686557, |
| "grad_norm": 0.16957001146132789, |
| "learning_rate": 3.409133574926302e-05, |
| "loss": 0.4969, |
| "step": 551 |
| }, |
| { |
| "epoch": 2.962118672477372, |
| "grad_norm": 0.15435240809940032, |
| "learning_rate": 3.394288889981695e-05, |
| "loss": 0.4946, |
| "step": 552 |
| }, |
| { |
| "epoch": 2.9674824002681865, |
| "grad_norm": 0.154260292592301, |
| "learning_rate": 3.379452738275095e-05, |
| "loss": 0.4929, |
| "step": 553 |
| }, |
| { |
| "epoch": 2.9728461280590013, |
| "grad_norm": 0.15155738642592756, |
| "learning_rate": 3.364625328817717e-05, |
| "loss": 0.4972, |
| "step": 554 |
| }, |
| { |
| "epoch": 2.9782098558498156, |
| "grad_norm": 0.17467421089382126, |
| "learning_rate": 3.349806870497615e-05, |
| "loss": 0.5008, |
| "step": 555 |
| }, |
| { |
| "epoch": 2.9835735836406303, |
| "grad_norm": 0.17056105214141762, |
| "learning_rate": 3.334997572076734e-05, |
| "loss": 0.4905, |
| "step": 556 |
| }, |
| { |
| "epoch": 2.988937311431445, |
| "grad_norm": 0.1673670147899835, |
| "learning_rate": 3.320197642187983e-05, |
| "loss": 0.494, |
| "step": 557 |
| }, |
| { |
| "epoch": 2.9943010392222593, |
| "grad_norm": 0.1850966895118181, |
| "learning_rate": 3.305407289332279e-05, |
| "loss": 0.4992, |
| "step": 558 |
| }, |
| { |
| "epoch": 3.000335232986926, |
| "grad_norm": 0.2722101703808084, |
| "learning_rate": 3.2906267218756244e-05, |
| "loss": 0.836, |
| "step": 559 |
| }, |
| { |
| "epoch": 3.0056989607777407, |
| "grad_norm": 0.3632821992997869, |
| "learning_rate": 3.2758561480461606e-05, |
| "loss": 0.4507, |
| "step": 560 |
| }, |
| { |
| "epoch": 3.011062688568555, |
| "grad_norm": 0.3428188312032687, |
| "learning_rate": 3.261095775931244e-05, |
| "loss": 0.46, |
| "step": 561 |
| }, |
| { |
| "epoch": 3.0164264163593697, |
| "grad_norm": 0.2992573935904219, |
| "learning_rate": 3.2463458134745036e-05, |
| "loss": 0.4511, |
| "step": 562 |
| }, |
| { |
| "epoch": 3.0217901441501844, |
| "grad_norm": 0.31765288073417064, |
| "learning_rate": 3.2316064684729246e-05, |
| "loss": 0.4476, |
| "step": 563 |
| }, |
| { |
| "epoch": 3.027153871940999, |
| "grad_norm": 0.26053192252342355, |
| "learning_rate": 3.2168779485739086e-05, |
| "loss": 0.4473, |
| "step": 564 |
| }, |
| { |
| "epoch": 3.0325175997318135, |
| "grad_norm": 0.298616923861466, |
| "learning_rate": 3.2021604612723525e-05, |
| "loss": 0.4521, |
| "step": 565 |
| }, |
| { |
| "epoch": 3.037881327522628, |
| "grad_norm": 0.28423459110220495, |
| "learning_rate": 3.187454213907733e-05, |
| "loss": 0.4507, |
| "step": 566 |
| }, |
| { |
| "epoch": 3.043245055313443, |
| "grad_norm": 0.23009487403692486, |
| "learning_rate": 3.172759413661172e-05, |
| "loss": 0.448, |
| "step": 567 |
| }, |
| { |
| "epoch": 3.0486087831042576, |
| "grad_norm": 0.28376655658935424, |
| "learning_rate": 3.1580762675525286e-05, |
| "loss": 0.4509, |
| "step": 568 |
| }, |
| { |
| "epoch": 3.053972510895072, |
| "grad_norm": 0.22381560292455216, |
| "learning_rate": 3.143404982437476e-05, |
| "loss": 0.4521, |
| "step": 569 |
| }, |
| { |
| "epoch": 3.0593362386858867, |
| "grad_norm": 0.24070713732730523, |
| "learning_rate": 3.128745765004594e-05, |
| "loss": 0.4439, |
| "step": 570 |
| }, |
| { |
| "epoch": 3.0646999664767014, |
| "grad_norm": 0.2632500650736694, |
| "learning_rate": 3.114098821772451e-05, |
| "loss": 0.4489, |
| "step": 571 |
| }, |
| { |
| "epoch": 3.070063694267516, |
| "grad_norm": 0.19380561177767147, |
| "learning_rate": 3.099464359086695e-05, |
| "loss": 0.4473, |
| "step": 572 |
| }, |
| { |
| "epoch": 3.0754274220583304, |
| "grad_norm": 0.28710006884974615, |
| "learning_rate": 3.084842583117153e-05, |
| "loss": 0.4527, |
| "step": 573 |
| }, |
| { |
| "epoch": 3.080791149849145, |
| "grad_norm": 0.18312560702847464, |
| "learning_rate": 3.0702336998549175e-05, |
| "loss": 0.451, |
| "step": 574 |
| }, |
| { |
| "epoch": 3.08615487763996, |
| "grad_norm": 0.23079325997690583, |
| "learning_rate": 3.055637915109453e-05, |
| "loss": 0.4427, |
| "step": 575 |
| }, |
| { |
| "epoch": 3.0915186054307746, |
| "grad_norm": 0.19698651608881707, |
| "learning_rate": 3.0410554345056876e-05, |
| "loss": 0.4534, |
| "step": 576 |
| }, |
| { |
| "epoch": 3.096882333221589, |
| "grad_norm": 0.18838697788327152, |
| "learning_rate": 3.026486463481125e-05, |
| "loss": 0.451, |
| "step": 577 |
| }, |
| { |
| "epoch": 3.1022460610124036, |
| "grad_norm": 0.17680792058130787, |
| "learning_rate": 3.0119312072829476e-05, |
| "loss": 0.4435, |
| "step": 578 |
| }, |
| { |
| "epoch": 3.1076097888032184, |
| "grad_norm": 0.1601761416116848, |
| "learning_rate": 2.997389870965118e-05, |
| "loss": 0.4511, |
| "step": 579 |
| }, |
| { |
| "epoch": 3.1129735165940327, |
| "grad_norm": 0.17773164119539953, |
| "learning_rate": 2.982862659385502e-05, |
| "loss": 0.4513, |
| "step": 580 |
| }, |
| { |
| "epoch": 3.1183372443848474, |
| "grad_norm": 0.17193804443547908, |
| "learning_rate": 2.968349777202973e-05, |
| "loss": 0.4484, |
| "step": 581 |
| }, |
| { |
| "epoch": 3.123700972175662, |
| "grad_norm": 0.16914799438808237, |
| "learning_rate": 2.9538514288745336e-05, |
| "loss": 0.4463, |
| "step": 582 |
| }, |
| { |
| "epoch": 3.129064699966477, |
| "grad_norm": 0.16873259416647984, |
| "learning_rate": 2.939367818652434e-05, |
| "loss": 0.4491, |
| "step": 583 |
| }, |
| { |
| "epoch": 3.134428427757291, |
| "grad_norm": 0.16119241415751986, |
| "learning_rate": 2.9248991505812944e-05, |
| "loss": 0.4473, |
| "step": 584 |
| }, |
| { |
| "epoch": 3.139792155548106, |
| "grad_norm": 0.15930346566056242, |
| "learning_rate": 2.9104456284952277e-05, |
| "loss": 0.4464, |
| "step": 585 |
| }, |
| { |
| "epoch": 3.1451558833389206, |
| "grad_norm": 0.15103719252838246, |
| "learning_rate": 2.8960074560149752e-05, |
| "loss": 0.4458, |
| "step": 586 |
| }, |
| { |
| "epoch": 3.1505196111297353, |
| "grad_norm": 0.14739728413169334, |
| "learning_rate": 2.8815848365450336e-05, |
| "loss": 0.4461, |
| "step": 587 |
| }, |
| { |
| "epoch": 3.1558833389205496, |
| "grad_norm": 0.1636187894417558, |
| "learning_rate": 2.867177973270784e-05, |
| "loss": 0.4444, |
| "step": 588 |
| }, |
| { |
| "epoch": 3.1612470667113644, |
| "grad_norm": 0.15742478053043885, |
| "learning_rate": 2.8527870691556404e-05, |
| "loss": 0.4476, |
| "step": 589 |
| }, |
| { |
| "epoch": 3.166610794502179, |
| "grad_norm": 0.611468578577517, |
| "learning_rate": 2.8384123269381784e-05, |
| "loss": 0.4607, |
| "step": 590 |
| }, |
| { |
| "epoch": 3.171974522292994, |
| "grad_norm": 0.17061785943367402, |
| "learning_rate": 2.8240539491292938e-05, |
| "loss": 0.4518, |
| "step": 591 |
| }, |
| { |
| "epoch": 3.177338250083808, |
| "grad_norm": 0.14588623033185896, |
| "learning_rate": 2.8097121380093323e-05, |
| "loss": 0.4404, |
| "step": 592 |
| }, |
| { |
| "epoch": 3.182701977874623, |
| "grad_norm": 0.18047802738317803, |
| "learning_rate": 2.7953870956252562e-05, |
| "loss": 0.4472, |
| "step": 593 |
| }, |
| { |
| "epoch": 3.1880657056654376, |
| "grad_norm": 0.15324942876302464, |
| "learning_rate": 2.7810790237877857e-05, |
| "loss": 0.4471, |
| "step": 594 |
| }, |
| { |
| "epoch": 3.1934294334562523, |
| "grad_norm": 0.19424542098598574, |
| "learning_rate": 2.7667881240685606e-05, |
| "loss": 0.4441, |
| "step": 595 |
| }, |
| { |
| "epoch": 3.1987931612470666, |
| "grad_norm": 0.16898760852013042, |
| "learning_rate": 2.7525145977973045e-05, |
| "loss": 0.4435, |
| "step": 596 |
| }, |
| { |
| "epoch": 3.2041568890378813, |
| "grad_norm": 0.19404967889677557, |
| "learning_rate": 2.738258646058978e-05, |
| "loss": 0.4427, |
| "step": 597 |
| }, |
| { |
| "epoch": 3.209520616828696, |
| "grad_norm": 0.16937723295023943, |
| "learning_rate": 2.7240204696909603e-05, |
| "loss": 0.4493, |
| "step": 598 |
| }, |
| { |
| "epoch": 3.2148843446195103, |
| "grad_norm": 0.14987599858224868, |
| "learning_rate": 2.7098002692802033e-05, |
| "loss": 0.4482, |
| "step": 599 |
| }, |
| { |
| "epoch": 3.220248072410325, |
| "grad_norm": 0.14816239745817442, |
| "learning_rate": 2.6955982451604206e-05, |
| "loss": 0.4512, |
| "step": 600 |
| }, |
| { |
| "epoch": 3.22561180020114, |
| "grad_norm": 0.13895435016957602, |
| "learning_rate": 2.6814145974092566e-05, |
| "loss": 0.4479, |
| "step": 601 |
| }, |
| { |
| "epoch": 3.2309755279919545, |
| "grad_norm": 0.13200544957430468, |
| "learning_rate": 2.6672495258454678e-05, |
| "loss": 0.4543, |
| "step": 602 |
| }, |
| { |
| "epoch": 3.236339255782769, |
| "grad_norm": 0.1456637489059305, |
| "learning_rate": 2.6531032300261153e-05, |
| "loss": 0.4475, |
| "step": 603 |
| }, |
| { |
| "epoch": 3.2417029835735836, |
| "grad_norm": 0.13818311152408982, |
| "learning_rate": 2.6389759092437418e-05, |
| "loss": 0.4446, |
| "step": 604 |
| }, |
| { |
| "epoch": 3.2470667113643983, |
| "grad_norm": 0.1330695845217543, |
| "learning_rate": 2.6248677625235763e-05, |
| "loss": 0.4513, |
| "step": 605 |
| }, |
| { |
| "epoch": 3.252430439155213, |
| "grad_norm": 0.12671554312337807, |
| "learning_rate": 2.6107789886207195e-05, |
| "loss": 0.443, |
| "step": 606 |
| }, |
| { |
| "epoch": 3.2577941669460273, |
| "grad_norm": 0.148616489659857, |
| "learning_rate": 2.5967097860173514e-05, |
| "loss": 0.4494, |
| "step": 607 |
| }, |
| { |
| "epoch": 3.263157894736842, |
| "grad_norm": 0.13556510213657127, |
| "learning_rate": 2.58266035291993e-05, |
| "loss": 0.4535, |
| "step": 608 |
| }, |
| { |
| "epoch": 3.2685216225276568, |
| "grad_norm": 0.15241618420915304, |
| "learning_rate": 2.5686308872564028e-05, |
| "loss": 0.4496, |
| "step": 609 |
| }, |
| { |
| "epoch": 3.2738853503184715, |
| "grad_norm": 0.144757212701804, |
| "learning_rate": 2.5546215866734185e-05, |
| "loss": 0.4465, |
| "step": 610 |
| }, |
| { |
| "epoch": 3.279249078109286, |
| "grad_norm": 0.14396415920602024, |
| "learning_rate": 2.540632648533536e-05, |
| "loss": 0.448, |
| "step": 611 |
| }, |
| { |
| "epoch": 3.2846128059001005, |
| "grad_norm": 0.15210625698722938, |
| "learning_rate": 2.526664269912452e-05, |
| "loss": 0.4453, |
| "step": 612 |
| }, |
| { |
| "epoch": 3.2899765336909153, |
| "grad_norm": 0.1536053230061205, |
| "learning_rate": 2.5127166475962205e-05, |
| "loss": 0.4487, |
| "step": 613 |
| }, |
| { |
| "epoch": 3.29534026148173, |
| "grad_norm": 0.15509713077990947, |
| "learning_rate": 2.4987899780784836e-05, |
| "loss": 0.4457, |
| "step": 614 |
| }, |
| { |
| "epoch": 3.3007039892725443, |
| "grad_norm": 0.13085919728595163, |
| "learning_rate": 2.484884457557696e-05, |
| "loss": 0.4443, |
| "step": 615 |
| }, |
| { |
| "epoch": 3.306067717063359, |
| "grad_norm": 0.16886670984395816, |
| "learning_rate": 2.4710002819343712e-05, |
| "loss": 0.4538, |
| "step": 616 |
| }, |
| { |
| "epoch": 3.3114314448541737, |
| "grad_norm": 0.12562181708750828, |
| "learning_rate": 2.457137646808315e-05, |
| "loss": 0.4477, |
| "step": 617 |
| }, |
| { |
| "epoch": 3.316795172644988, |
| "grad_norm": 0.1577346712555077, |
| "learning_rate": 2.44329674747587e-05, |
| "loss": 0.4542, |
| "step": 618 |
| }, |
| { |
| "epoch": 3.3221589004358028, |
| "grad_norm": 0.13554447992255508, |
| "learning_rate": 2.4294777789271663e-05, |
| "loss": 0.4495, |
| "step": 619 |
| }, |
| { |
| "epoch": 3.3275226282266175, |
| "grad_norm": 0.14611739712605668, |
| "learning_rate": 2.4156809358433728e-05, |
| "loss": 0.4472, |
| "step": 620 |
| }, |
| { |
| "epoch": 3.332886356017432, |
| "grad_norm": 0.1400607484697624, |
| "learning_rate": 2.4019064125939603e-05, |
| "loss": 0.4521, |
| "step": 621 |
| }, |
| { |
| "epoch": 3.338250083808247, |
| "grad_norm": 0.14259331457489, |
| "learning_rate": 2.3881544032339506e-05, |
| "loss": 0.448, |
| "step": 622 |
| }, |
| { |
| "epoch": 3.3436138115990612, |
| "grad_norm": 0.13315687312240615, |
| "learning_rate": 2.3744251015011987e-05, |
| "loss": 0.4463, |
| "step": 623 |
| }, |
| { |
| "epoch": 3.348977539389876, |
| "grad_norm": 0.12598643415214605, |
| "learning_rate": 2.360718700813651e-05, |
| "loss": 0.4447, |
| "step": 624 |
| }, |
| { |
| "epoch": 3.3543412671806907, |
| "grad_norm": 0.14066053069067935, |
| "learning_rate": 2.347035394266623e-05, |
| "loss": 0.446, |
| "step": 625 |
| }, |
| { |
| "epoch": 3.359704994971505, |
| "grad_norm": 0.12802549460203494, |
| "learning_rate": 2.333375374630086e-05, |
| "loss": 0.4419, |
| "step": 626 |
| }, |
| { |
| "epoch": 3.3650687227623197, |
| "grad_norm": 0.12761654852469786, |
| "learning_rate": 2.3197388343459405e-05, |
| "loss": 0.4501, |
| "step": 627 |
| }, |
| { |
| "epoch": 3.3704324505531345, |
| "grad_norm": 0.14182894960939563, |
| "learning_rate": 2.3061259655253165e-05, |
| "loss": 0.4463, |
| "step": 628 |
| }, |
| { |
| "epoch": 3.375796178343949, |
| "grad_norm": 0.41133093836911544, |
| "learning_rate": 2.2925369599458543e-05, |
| "loss": 0.4551, |
| "step": 629 |
| }, |
| { |
| "epoch": 3.3811599061347635, |
| "grad_norm": 0.13182899206147553, |
| "learning_rate": 2.2789720090490167e-05, |
| "loss": 0.4446, |
| "step": 630 |
| }, |
| { |
| "epoch": 3.386523633925578, |
| "grad_norm": 0.1170971831574601, |
| "learning_rate": 2.265431303937379e-05, |
| "loss": 0.4417, |
| "step": 631 |
| }, |
| { |
| "epoch": 3.391887361716393, |
| "grad_norm": 0.12294985745042769, |
| "learning_rate": 2.2519150353719478e-05, |
| "loss": 0.4494, |
| "step": 632 |
| }, |
| { |
| "epoch": 3.3972510895072077, |
| "grad_norm": 0.13039963065660728, |
| "learning_rate": 2.2384233937694626e-05, |
| "loss": 0.447, |
| "step": 633 |
| }, |
| { |
| "epoch": 3.402614817298022, |
| "grad_norm": 0.11831506164421687, |
| "learning_rate": 2.2249565691997263e-05, |
| "loss": 0.4499, |
| "step": 634 |
| }, |
| { |
| "epoch": 3.4079785450888367, |
| "grad_norm": 0.12143409559233521, |
| "learning_rate": 2.2115147513829145e-05, |
| "loss": 0.4476, |
| "step": 635 |
| }, |
| { |
| "epoch": 3.4133422728796514, |
| "grad_norm": 0.1291445962430486, |
| "learning_rate": 2.1980981296869083e-05, |
| "loss": 0.4486, |
| "step": 636 |
| }, |
| { |
| "epoch": 3.418706000670466, |
| "grad_norm": 0.13989097404575776, |
| "learning_rate": 2.184706893124633e-05, |
| "loss": 0.4387, |
| "step": 637 |
| }, |
| { |
| "epoch": 3.4240697284612804, |
| "grad_norm": 0.3715077464881586, |
| "learning_rate": 2.1713412303513803e-05, |
| "loss": 0.4472, |
| "step": 638 |
| }, |
| { |
| "epoch": 3.429433456252095, |
| "grad_norm": 0.1251774557356555, |
| "learning_rate": 2.1580013296621657e-05, |
| "loss": 0.4539, |
| "step": 639 |
| }, |
| { |
| "epoch": 3.43479718404291, |
| "grad_norm": 0.13346719081780015, |
| "learning_rate": 2.144687378989069e-05, |
| "loss": 0.4494, |
| "step": 640 |
| }, |
| { |
| "epoch": 3.4401609118337246, |
| "grad_norm": 0.13105974532551945, |
| "learning_rate": 2.1313995658985825e-05, |
| "loss": 0.4457, |
| "step": 641 |
| }, |
| { |
| "epoch": 3.445524639624539, |
| "grad_norm": 0.12874889844793144, |
| "learning_rate": 2.118138077588978e-05, |
| "loss": 0.4471, |
| "step": 642 |
| }, |
| { |
| "epoch": 3.4508883674153537, |
| "grad_norm": 0.13187822242861885, |
| "learning_rate": 2.1049031008876603e-05, |
| "loss": 0.4486, |
| "step": 643 |
| }, |
| { |
| "epoch": 3.4562520952061684, |
| "grad_norm": 0.11696538375561746, |
| "learning_rate": 2.0916948222485446e-05, |
| "loss": 0.4509, |
| "step": 644 |
| }, |
| { |
| "epoch": 3.4616158229969827, |
| "grad_norm": 0.121989247700871, |
| "learning_rate": 2.0785134277494202e-05, |
| "loss": 0.4462, |
| "step": 645 |
| }, |
| { |
| "epoch": 3.4669795507877974, |
| "grad_norm": 0.893427164640271, |
| "learning_rate": 2.065359103089333e-05, |
| "loss": 0.4619, |
| "step": 646 |
| }, |
| { |
| "epoch": 3.472343278578612, |
| "grad_norm": 0.45677961197929307, |
| "learning_rate": 2.0522320335859768e-05, |
| "loss": 0.4513, |
| "step": 647 |
| }, |
| { |
| "epoch": 3.477707006369427, |
| "grad_norm": 0.13666000173799434, |
| "learning_rate": 2.0391324041730665e-05, |
| "loss": 0.4538, |
| "step": 648 |
| }, |
| { |
| "epoch": 3.4830707341602416, |
| "grad_norm": 0.2608095297691742, |
| "learning_rate": 2.026060399397751e-05, |
| "loss": 0.4489, |
| "step": 649 |
| }, |
| { |
| "epoch": 3.488434461951056, |
| "grad_norm": 0.18107927299339568, |
| "learning_rate": 2.013016203417994e-05, |
| "loss": 0.4526, |
| "step": 650 |
| }, |
| { |
| "epoch": 3.4937981897418706, |
| "grad_norm": 0.14754327484468957, |
| "learning_rate": 2.0000000000000012e-05, |
| "loss": 0.4467, |
| "step": 651 |
| }, |
| { |
| "epoch": 3.4991619175326854, |
| "grad_norm": 0.13398053465109833, |
| "learning_rate": 1.9870119725156094e-05, |
| "loss": 0.4526, |
| "step": 652 |
| }, |
| { |
| "epoch": 3.5045256453234996, |
| "grad_norm": 0.13952089135435625, |
| "learning_rate": 1.9740523039397225e-05, |
| "loss": 0.4407, |
| "step": 653 |
| }, |
| { |
| "epoch": 3.5098893731143144, |
| "grad_norm": 0.14080074484432106, |
| "learning_rate": 1.9611211768477173e-05, |
| "loss": 0.4454, |
| "step": 654 |
| }, |
| { |
| "epoch": 3.515253100905129, |
| "grad_norm": 0.1198388915000314, |
| "learning_rate": 1.948218773412886e-05, |
| "loss": 0.4492, |
| "step": 655 |
| }, |
| { |
| "epoch": 3.520616828695944, |
| "grad_norm": 0.13993217838224647, |
| "learning_rate": 1.935345275403859e-05, |
| "loss": 0.44, |
| "step": 656 |
| }, |
| { |
| "epoch": 3.5259805564867586, |
| "grad_norm": 0.12889045963571127, |
| "learning_rate": 1.9225008641820498e-05, |
| "loss": 0.4443, |
| "step": 657 |
| }, |
| { |
| "epoch": 3.531344284277573, |
| "grad_norm": 0.12378896201897356, |
| "learning_rate": 1.909685720699096e-05, |
| "loss": 0.4456, |
| "step": 658 |
| }, |
| { |
| "epoch": 3.5367080120683876, |
| "grad_norm": 0.1276674265766728, |
| "learning_rate": 1.8969000254943125e-05, |
| "loss": 0.4481, |
| "step": 659 |
| }, |
| { |
| "epoch": 3.5420717398592023, |
| "grad_norm": 0.12390916606673247, |
| "learning_rate": 1.8841439586921515e-05, |
| "loss": 0.4451, |
| "step": 660 |
| }, |
| { |
| "epoch": 3.5474354676500166, |
| "grad_norm": 0.12273677799318142, |
| "learning_rate": 1.871417699999656e-05, |
| "loss": 0.4502, |
| "step": 661 |
| }, |
| { |
| "epoch": 3.5527991954408313, |
| "grad_norm": 0.11996313423286693, |
| "learning_rate": 1.858721428703937e-05, |
| "loss": 0.4561, |
| "step": 662 |
| }, |
| { |
| "epoch": 3.558162923231646, |
| "grad_norm": 0.12358256215052507, |
| "learning_rate": 1.8460553236696448e-05, |
| "loss": 0.4475, |
| "step": 663 |
| }, |
| { |
| "epoch": 3.5635266510224604, |
| "grad_norm": 1.460296103471973, |
| "learning_rate": 1.8334195633364435e-05, |
| "loss": 0.4528, |
| "step": 664 |
| }, |
| { |
| "epoch": 3.568890378813275, |
| "grad_norm": 0.12331491175077743, |
| "learning_rate": 1.8208143257165085e-05, |
| "loss": 0.4509, |
| "step": 665 |
| }, |
| { |
| "epoch": 3.57425410660409, |
| "grad_norm": 0.11750213968895096, |
| "learning_rate": 1.808239788392006e-05, |
| "loss": 0.4489, |
| "step": 666 |
| }, |
| { |
| "epoch": 3.5796178343949046, |
| "grad_norm": 0.11714707650553612, |
| "learning_rate": 1.795696128512603e-05, |
| "loss": 0.4466, |
| "step": 667 |
| }, |
| { |
| "epoch": 3.5849815621857193, |
| "grad_norm": 0.12196764778181841, |
| "learning_rate": 1.78318352279296e-05, |
| "loss": 0.4447, |
| "step": 668 |
| }, |
| { |
| "epoch": 3.5903452899765336, |
| "grad_norm": 0.12562029298509103, |
| "learning_rate": 1.7707021475102548e-05, |
| "loss": 0.4533, |
| "step": 669 |
| }, |
| { |
| "epoch": 3.5957090177673483, |
| "grad_norm": 0.14349002013949913, |
| "learning_rate": 1.758252178501686e-05, |
| "loss": 0.4436, |
| "step": 670 |
| }, |
| { |
| "epoch": 3.601072745558163, |
| "grad_norm": 0.11537479526408528, |
| "learning_rate": 1.745833791162003e-05, |
| "loss": 0.4482, |
| "step": 671 |
| }, |
| { |
| "epoch": 3.6064364733489773, |
| "grad_norm": 0.3056266641244182, |
| "learning_rate": 1.7334471604410367e-05, |
| "loss": 0.4419, |
| "step": 672 |
| }, |
| { |
| "epoch": 3.611800201139792, |
| "grad_norm": 0.11213414375074307, |
| "learning_rate": 1.7210924608412257e-05, |
| "loss": 0.4517, |
| "step": 673 |
| }, |
| { |
| "epoch": 3.617163928930607, |
| "grad_norm": 0.13465101181054254, |
| "learning_rate": 1.7087698664151724e-05, |
| "loss": 0.4386, |
| "step": 674 |
| }, |
| { |
| "epoch": 3.6225276567214215, |
| "grad_norm": 0.1265640161713725, |
| "learning_rate": 1.6964795507631745e-05, |
| "loss": 0.4495, |
| "step": 675 |
| }, |
| { |
| "epoch": 3.6278913845122363, |
| "grad_norm": 0.13249341487141286, |
| "learning_rate": 1.684221687030793e-05, |
| "loss": 0.4505, |
| "step": 676 |
| }, |
| { |
| "epoch": 3.6332551123030505, |
| "grad_norm": 0.10926988434260812, |
| "learning_rate": 1.671996447906403e-05, |
| "loss": 0.4507, |
| "step": 677 |
| }, |
| { |
| "epoch": 3.6386188400938653, |
| "grad_norm": 0.1160947830417163, |
| "learning_rate": 1.659804005618769e-05, |
| "loss": 0.4548, |
| "step": 678 |
| }, |
| { |
| "epoch": 3.64398256788468, |
| "grad_norm": 0.14474625249969114, |
| "learning_rate": 1.6476445319346143e-05, |
| "loss": 0.4415, |
| "step": 679 |
| }, |
| { |
| "epoch": 3.6493462956754943, |
| "grad_norm": 0.11464961989670115, |
| "learning_rate": 1.6355181981561976e-05, |
| "loss": 0.4493, |
| "step": 680 |
| }, |
| { |
| "epoch": 3.654710023466309, |
| "grad_norm": 0.11514676338990276, |
| "learning_rate": 1.62342517511891e-05, |
| "loss": 0.4464, |
| "step": 681 |
| }, |
| { |
| "epoch": 3.6600737512571238, |
| "grad_norm": 0.10388737077706772, |
| "learning_rate": 1.6113656331888563e-05, |
| "loss": 0.4431, |
| "step": 682 |
| }, |
| { |
| "epoch": 3.665437479047938, |
| "grad_norm": 0.21254990175343727, |
| "learning_rate": 1.599339742260463e-05, |
| "loss": 0.4494, |
| "step": 683 |
| }, |
| { |
| "epoch": 3.6708012068387528, |
| "grad_norm": 0.11018791870383089, |
| "learning_rate": 1.5873476717540818e-05, |
| "loss": 0.4453, |
| "step": 684 |
| }, |
| { |
| "epoch": 3.6761649346295675, |
| "grad_norm": 0.11202984963659915, |
| "learning_rate": 1.575389590613604e-05, |
| "loss": 0.4421, |
| "step": 685 |
| }, |
| { |
| "epoch": 3.6815286624203822, |
| "grad_norm": 0.5609794652204741, |
| "learning_rate": 1.5634656673040824e-05, |
| "loss": 0.4453, |
| "step": 686 |
| }, |
| { |
| "epoch": 3.686892390211197, |
| "grad_norm": 0.12056021537980104, |
| "learning_rate": 1.5515760698093485e-05, |
| "loss": 0.4422, |
| "step": 687 |
| }, |
| { |
| "epoch": 3.6922561180020113, |
| "grad_norm": 0.1143368342750797, |
| "learning_rate": 1.539720965629661e-05, |
| "loss": 0.4459, |
| "step": 688 |
| }, |
| { |
| "epoch": 3.697619845792826, |
| "grad_norm": 0.12221336375445264, |
| "learning_rate": 1.5279005217793307e-05, |
| "loss": 0.4454, |
| "step": 689 |
| }, |
| { |
| "epoch": 3.7029835735836407, |
| "grad_norm": 0.12874007735808712, |
| "learning_rate": 1.5161149047843813e-05, |
| "loss": 0.4501, |
| "step": 690 |
| }, |
| { |
| "epoch": 3.708347301374455, |
| "grad_norm": 0.23200442640856928, |
| "learning_rate": 1.504364280680191e-05, |
| "loss": 0.4421, |
| "step": 691 |
| }, |
| { |
| "epoch": 3.7137110291652697, |
| "grad_norm": 0.118738423956062, |
| "learning_rate": 1.492648815009163e-05, |
| "loss": 0.4473, |
| "step": 692 |
| }, |
| { |
| "epoch": 3.7190747569560845, |
| "grad_norm": 0.11440430785036737, |
| "learning_rate": 1.4809686728183903e-05, |
| "loss": 0.4441, |
| "step": 693 |
| }, |
| { |
| "epoch": 3.724438484746899, |
| "grad_norm": 0.11276169565585069, |
| "learning_rate": 1.4693240186573267e-05, |
| "loss": 0.4441, |
| "step": 694 |
| }, |
| { |
| "epoch": 3.729802212537714, |
| "grad_norm": 0.10878975487213734, |
| "learning_rate": 1.4577150165754739e-05, |
| "loss": 0.4455, |
| "step": 695 |
| }, |
| { |
| "epoch": 3.7351659403285282, |
| "grad_norm": 0.11093388920860042, |
| "learning_rate": 1.4461418301200665e-05, |
| "loss": 0.4422, |
| "step": 696 |
| }, |
| { |
| "epoch": 3.740529668119343, |
| "grad_norm": 0.10930519937275324, |
| "learning_rate": 1.4346046223337737e-05, |
| "loss": 0.4458, |
| "step": 697 |
| }, |
| { |
| "epoch": 3.7458933959101577, |
| "grad_norm": 0.11767456655079721, |
| "learning_rate": 1.4231035557523925e-05, |
| "loss": 0.4455, |
| "step": 698 |
| }, |
| { |
| "epoch": 3.751257123700972, |
| "grad_norm": 0.11061581216489039, |
| "learning_rate": 1.4116387924025703e-05, |
| "loss": 0.4451, |
| "step": 699 |
| }, |
| { |
| "epoch": 3.7566208514917867, |
| "grad_norm": 0.12532437758806012, |
| "learning_rate": 1.4002104937995103e-05, |
| "loss": 0.4435, |
| "step": 700 |
| }, |
| { |
| "epoch": 3.7619845792826014, |
| "grad_norm": 0.1101987534346296, |
| "learning_rate": 1.388818820944704e-05, |
| "loss": 0.4445, |
| "step": 701 |
| }, |
| { |
| "epoch": 3.767348307073416, |
| "grad_norm": 0.11241200124363177, |
| "learning_rate": 1.377463934323663e-05, |
| "loss": 0.4424, |
| "step": 702 |
| }, |
| { |
| "epoch": 3.772712034864231, |
| "grad_norm": 0.11769736086113589, |
| "learning_rate": 1.3661459939036493e-05, |
| "loss": 0.4441, |
| "step": 703 |
| }, |
| { |
| "epoch": 3.778075762655045, |
| "grad_norm": 0.10530984853730485, |
| "learning_rate": 1.354865159131435e-05, |
| "loss": 0.4454, |
| "step": 704 |
| }, |
| { |
| "epoch": 3.78343949044586, |
| "grad_norm": 0.128491542213919, |
| "learning_rate": 1.3436215889310433e-05, |
| "loss": 0.4483, |
| "step": 705 |
| }, |
| { |
| "epoch": 3.7888032182366747, |
| "grad_norm": 0.10663013232583238, |
| "learning_rate": 1.3324154417015205e-05, |
| "loss": 0.4477, |
| "step": 706 |
| }, |
| { |
| "epoch": 3.794166946027489, |
| "grad_norm": 0.11468548568373477, |
| "learning_rate": 1.3212468753146955e-05, |
| "loss": 0.4399, |
| "step": 707 |
| }, |
| { |
| "epoch": 3.7995306738183037, |
| "grad_norm": 0.11277620565805507, |
| "learning_rate": 1.3101160471129588e-05, |
| "loss": 0.4412, |
| "step": 708 |
| }, |
| { |
| "epoch": 3.8048944016091184, |
| "grad_norm": 0.10429631660245424, |
| "learning_rate": 1.2990231139070519e-05, |
| "loss": 0.439, |
| "step": 709 |
| }, |
| { |
| "epoch": 3.8102581293999327, |
| "grad_norm": 0.11117524802508165, |
| "learning_rate": 1.2879682319738467e-05, |
| "loss": 0.448, |
| "step": 710 |
| }, |
| { |
| "epoch": 3.8156218571907474, |
| "grad_norm": 0.10565379203871825, |
| "learning_rate": 1.2769515570541554e-05, |
| "loss": 0.4506, |
| "step": 711 |
| }, |
| { |
| "epoch": 3.820985584981562, |
| "grad_norm": 0.10995579742720774, |
| "learning_rate": 1.2659732443505263e-05, |
| "loss": 0.4378, |
| "step": 712 |
| }, |
| { |
| "epoch": 3.826349312772377, |
| "grad_norm": 0.1094854319106533, |
| "learning_rate": 1.2550334485250661e-05, |
| "loss": 0.4388, |
| "step": 713 |
| }, |
| { |
| "epoch": 3.8317130405631916, |
| "grad_norm": 0.10426452233943717, |
| "learning_rate": 1.2441323236972536e-05, |
| "loss": 0.449, |
| "step": 714 |
| }, |
| { |
| "epoch": 3.837076768354006, |
| "grad_norm": 0.11501738794377854, |
| "learning_rate": 1.2332700234417745e-05, |
| "loss": 0.4502, |
| "step": 715 |
| }, |
| { |
| "epoch": 3.8424404961448206, |
| "grad_norm": 0.10783773069849821, |
| "learning_rate": 1.222446700786355e-05, |
| "loss": 0.443, |
| "step": 716 |
| }, |
| { |
| "epoch": 3.8478042239356354, |
| "grad_norm": 0.1158852559257513, |
| "learning_rate": 1.2116625082096043e-05, |
| "loss": 0.4439, |
| "step": 717 |
| }, |
| { |
| "epoch": 3.8531679517264497, |
| "grad_norm": 0.10640912577878574, |
| "learning_rate": 1.2009175976388683e-05, |
| "loss": 0.4416, |
| "step": 718 |
| }, |
| { |
| "epoch": 3.8585316795172644, |
| "grad_norm": 0.11306196171729263, |
| "learning_rate": 1.1902121204480928e-05, |
| "loss": 0.4475, |
| "step": 719 |
| }, |
| { |
| "epoch": 3.863895407308079, |
| "grad_norm": 0.10374780127454973, |
| "learning_rate": 1.1795462274556835e-05, |
| "loss": 0.4413, |
| "step": 720 |
| }, |
| { |
| "epoch": 3.869259135098894, |
| "grad_norm": 0.10797835729481477, |
| "learning_rate": 1.1689200689223862e-05, |
| "loss": 0.4467, |
| "step": 721 |
| }, |
| { |
| "epoch": 3.8746228628897086, |
| "grad_norm": 0.10466854161183786, |
| "learning_rate": 1.1583337945491717e-05, |
| "loss": 0.4435, |
| "step": 722 |
| }, |
| { |
| "epoch": 3.879986590680523, |
| "grad_norm": 0.1037575420149514, |
| "learning_rate": 1.1477875534751192e-05, |
| "loss": 0.437, |
| "step": 723 |
| }, |
| { |
| "epoch": 3.8853503184713376, |
| "grad_norm": 0.11088366593365082, |
| "learning_rate": 1.1372814942753246e-05, |
| "loss": 0.4431, |
| "step": 724 |
| }, |
| { |
| "epoch": 3.8907140462621523, |
| "grad_norm": 0.09889840234162205, |
| "learning_rate": 1.1268157649588018e-05, |
| "loss": 0.4406, |
| "step": 725 |
| }, |
| { |
| "epoch": 3.8960777740529666, |
| "grad_norm": 0.10215017922131649, |
| "learning_rate": 1.1163905129663956e-05, |
| "loss": 0.4483, |
| "step": 726 |
| }, |
| { |
| "epoch": 3.9014415018437814, |
| "grad_norm": 0.1077840732925271, |
| "learning_rate": 1.1060058851687128e-05, |
| "loss": 0.4417, |
| "step": 727 |
| }, |
| { |
| "epoch": 3.906805229634596, |
| "grad_norm": 0.10317372106874026, |
| "learning_rate": 1.0956620278640427e-05, |
| "loss": 0.4446, |
| "step": 728 |
| }, |
| { |
| "epoch": 3.912168957425411, |
| "grad_norm": 0.1328432948301595, |
| "learning_rate": 1.0853590867763054e-05, |
| "loss": 0.431, |
| "step": 729 |
| }, |
| { |
| "epoch": 3.9175326852162256, |
| "grad_norm": 0.10066539948258632, |
| "learning_rate": 1.0750972070529922e-05, |
| "loss": 0.4474, |
| "step": 730 |
| }, |
| { |
| "epoch": 3.92289641300704, |
| "grad_norm": 0.10308611398222804, |
| "learning_rate": 1.064876533263122e-05, |
| "loss": 0.4466, |
| "step": 731 |
| }, |
| { |
| "epoch": 3.9282601407978546, |
| "grad_norm": 0.10575171336887328, |
| "learning_rate": 1.0546972093952114e-05, |
| "loss": 0.4448, |
| "step": 732 |
| }, |
| { |
| "epoch": 3.9336238685886693, |
| "grad_norm": 0.1090751248770618, |
| "learning_rate": 1.0445593788552344e-05, |
| "loss": 0.4405, |
| "step": 733 |
| }, |
| { |
| "epoch": 3.9389875963794836, |
| "grad_norm": 0.10355561012627491, |
| "learning_rate": 1.0344631844646128e-05, |
| "loss": 0.4432, |
| "step": 734 |
| }, |
| { |
| "epoch": 3.9443513241702983, |
| "grad_norm": 0.10595362039676634, |
| "learning_rate": 1.024408768458196e-05, |
| "loss": 0.4379, |
| "step": 735 |
| }, |
| { |
| "epoch": 3.949715051961113, |
| "grad_norm": 0.11198347644741112, |
| "learning_rate": 1.0143962724822653e-05, |
| "loss": 0.4491, |
| "step": 736 |
| }, |
| { |
| "epoch": 3.9550787797519273, |
| "grad_norm": 0.11560971244461928, |
| "learning_rate": 1.0044258375925295e-05, |
| "loss": 0.4482, |
| "step": 737 |
| }, |
| { |
| "epoch": 3.960442507542742, |
| "grad_norm": 0.10482045247799242, |
| "learning_rate": 9.944976042521465e-06, |
| "loss": 0.4427, |
| "step": 738 |
| }, |
| { |
| "epoch": 3.965806235333557, |
| "grad_norm": 0.10640773573235275, |
| "learning_rate": 9.846117123297353e-06, |
| "loss": 0.4455, |
| "step": 739 |
| }, |
| { |
| "epoch": 3.9711699631243715, |
| "grad_norm": 0.1038041198852965, |
| "learning_rate": 9.747683010974147e-06, |
| "loss": 0.4427, |
| "step": 740 |
| }, |
| { |
| "epoch": 3.9765336909151863, |
| "grad_norm": 0.10117239899529366, |
| "learning_rate": 9.649675092288366e-06, |
| "loss": 0.4413, |
| "step": 741 |
| }, |
| { |
| "epoch": 3.9818974187060006, |
| "grad_norm": 0.10620766516638773, |
| "learning_rate": 9.552094747972297e-06, |
| "loss": 0.4444, |
| "step": 742 |
| }, |
| { |
| "epoch": 3.9872611464968153, |
| "grad_norm": 0.10086629667003526, |
| "learning_rate": 9.454943352734598e-06, |
| "loss": 0.4417, |
| "step": 743 |
| }, |
| { |
| "epoch": 3.99262487428763, |
| "grad_norm": 0.09919272930882302, |
| "learning_rate": 9.358222275240884e-06, |
| "loss": 0.4487, |
| "step": 744 |
| }, |
| { |
| "epoch": 3.9979886020784443, |
| "grad_norm": 0.11767811064838585, |
| "learning_rate": 9.26193287809451e-06, |
| "loss": 0.5044, |
| "step": 745 |
| }, |
| { |
| "epoch": 4.004022795843111, |
| "grad_norm": 0.23793487869261376, |
| "learning_rate": 9.166076517817281e-06, |
| "loss": 0.661, |
| "step": 746 |
| }, |
| { |
| "epoch": 4.009386523633926, |
| "grad_norm": 0.13621220417679167, |
| "learning_rate": 9.07065454483043e-06, |
| "loss": 0.4149, |
| "step": 747 |
| }, |
| { |
| "epoch": 4.01475025142474, |
| "grad_norm": 0.17673227803196773, |
| "learning_rate": 8.975668303435556e-06, |
| "loss": 0.4189, |
| "step": 748 |
| }, |
| { |
| "epoch": 4.020113979215555, |
| "grad_norm": 0.18439349920499457, |
| "learning_rate": 8.881119131795652e-06, |
| "loss": 0.4107, |
| "step": 749 |
| }, |
| { |
| "epoch": 4.025477707006369, |
| "grad_norm": 0.12260746721261573, |
| "learning_rate": 8.787008361916332e-06, |
| "loss": 0.4125, |
| "step": 750 |
| }, |
| { |
| "epoch": 4.030841434797184, |
| "grad_norm": 0.15019363340139308, |
| "learning_rate": 8.693337319626978e-06, |
| "loss": 0.4128, |
| "step": 751 |
| }, |
| { |
| "epoch": 4.036205162587999, |
| "grad_norm": 0.1375901687390116, |
| "learning_rate": 8.60010732456214e-06, |
| "loss": 0.4055, |
| "step": 752 |
| }, |
| { |
| "epoch": 4.041568890378813, |
| "grad_norm": 0.12724391705577723, |
| "learning_rate": 8.507319690142871e-06, |
| "loss": 0.4146, |
| "step": 753 |
| }, |
| { |
| "epoch": 4.0469326181696275, |
| "grad_norm": 0.13895885677671274, |
| "learning_rate": 8.414975723558317e-06, |
| "loss": 0.4142, |
| "step": 754 |
| }, |
| { |
| "epoch": 4.052296345960443, |
| "grad_norm": 0.14044158258051825, |
| "learning_rate": 8.323076725747192e-06, |
| "loss": 0.4106, |
| "step": 755 |
| }, |
| { |
| "epoch": 4.057660073751257, |
| "grad_norm": 0.11701974422585514, |
| "learning_rate": 8.23162399137952e-06, |
| "loss": 0.4192, |
| "step": 756 |
| }, |
| { |
| "epoch": 4.063023801542072, |
| "grad_norm": 0.12971366592995717, |
| "learning_rate": 8.140618808838408e-06, |
| "loss": 0.4031, |
| "step": 757 |
| }, |
| { |
| "epoch": 4.068387529332886, |
| "grad_norm": 0.12697445696160525, |
| "learning_rate": 8.050062460201827e-06, |
| "loss": 0.4131, |
| "step": 758 |
| }, |
| { |
| "epoch": 4.073751257123701, |
| "grad_norm": 0.11378592889355851, |
| "learning_rate": 7.959956221224626e-06, |
| "loss": 0.4111, |
| "step": 759 |
| }, |
| { |
| "epoch": 4.079114984914516, |
| "grad_norm": 0.12117141039045348, |
| "learning_rate": 7.870301361320485e-06, |
| "loss": 0.4165, |
| "step": 760 |
| }, |
| { |
| "epoch": 4.08447871270533, |
| "grad_norm": 0.1197961111627781, |
| "learning_rate": 7.781099143544124e-06, |
| "loss": 0.4129, |
| "step": 761 |
| }, |
| { |
| "epoch": 4.089842440496144, |
| "grad_norm": 0.10807056855124261, |
| "learning_rate": 7.692350824573402e-06, |
| "loss": 0.4134, |
| "step": 762 |
| }, |
| { |
| "epoch": 4.09520616828696, |
| "grad_norm": 0.11324627112259107, |
| "learning_rate": 7.604057654691699e-06, |
| "loss": 0.4079, |
| "step": 763 |
| }, |
| { |
| "epoch": 4.100569896077774, |
| "grad_norm": 0.11724445449675143, |
| "learning_rate": 7.516220877770273e-06, |
| "loss": 0.411, |
| "step": 764 |
| }, |
| { |
| "epoch": 4.105933623868589, |
| "grad_norm": 0.10958796773895332, |
| "learning_rate": 7.428841731250695e-06, |
| "loss": 0.415, |
| "step": 765 |
| }, |
| { |
| "epoch": 4.111297351659403, |
| "grad_norm": 0.11066322738663384, |
| "learning_rate": 7.341921446127509e-06, |
| "loss": 0.4106, |
| "step": 766 |
| }, |
| { |
| "epoch": 4.116661079450218, |
| "grad_norm": 0.11185922785392959, |
| "learning_rate": 7.255461246930791e-06, |
| "loss": 0.4101, |
| "step": 767 |
| }, |
| { |
| "epoch": 4.122024807241033, |
| "grad_norm": 0.11501328533411367, |
| "learning_rate": 7.169462351708958e-06, |
| "loss": 0.4154, |
| "step": 768 |
| }, |
| { |
| "epoch": 4.127388535031847, |
| "grad_norm": 0.10891964674735306, |
| "learning_rate": 7.083925972011583e-06, |
| "loss": 0.4109, |
| "step": 769 |
| }, |
| { |
| "epoch": 4.132752262822661, |
| "grad_norm": 0.1084357061608093, |
| "learning_rate": 6.998853312872347e-06, |
| "loss": 0.4176, |
| "step": 770 |
| }, |
| { |
| "epoch": 4.138115990613477, |
| "grad_norm": 0.11666376677405135, |
| "learning_rate": 6.914245572792064e-06, |
| "loss": 0.4063, |
| "step": 771 |
| }, |
| { |
| "epoch": 4.143479718404291, |
| "grad_norm": 0.10587157920780786, |
| "learning_rate": 6.830103943721749e-06, |
| "loss": 0.4149, |
| "step": 772 |
| }, |
| { |
| "epoch": 4.148843446195105, |
| "grad_norm": 1.5402775013214236, |
| "learning_rate": 6.7464296110458925e-06, |
| "loss": 0.4212, |
| "step": 773 |
| }, |
| { |
| "epoch": 4.15420717398592, |
| "grad_norm": 0.1127675778731525, |
| "learning_rate": 6.6632237535656995e-06, |
| "loss": 0.4123, |
| "step": 774 |
| }, |
| { |
| "epoch": 4.159570901776735, |
| "grad_norm": 0.10636517110929453, |
| "learning_rate": 6.58048754348255e-06, |
| "loss": 0.4031, |
| "step": 775 |
| }, |
| { |
| "epoch": 4.16493462956755, |
| "grad_norm": 0.10408511766608668, |
| "learning_rate": 6.4982221463813965e-06, |
| "loss": 0.4092, |
| "step": 776 |
| }, |
| { |
| "epoch": 4.170298357358364, |
| "grad_norm": 0.10980308429743034, |
| "learning_rate": 6.41642872121444e-06, |
| "loss": 0.4133, |
| "step": 777 |
| }, |
| { |
| "epoch": 4.175662085149178, |
| "grad_norm": 0.1811518845819661, |
| "learning_rate": 6.335108420284748e-06, |
| "loss": 0.4114, |
| "step": 778 |
| }, |
| { |
| "epoch": 4.1810258129399935, |
| "grad_norm": 0.10929364439171839, |
| "learning_rate": 6.254262389230006e-06, |
| "loss": 0.4175, |
| "step": 779 |
| }, |
| { |
| "epoch": 4.186389540730808, |
| "grad_norm": 0.10394422643680504, |
| "learning_rate": 6.1738917670064194e-06, |
| "loss": 0.418, |
| "step": 780 |
| }, |
| { |
| "epoch": 4.191753268521622, |
| "grad_norm": 0.1012577329095609, |
| "learning_rate": 6.09399768587263e-06, |
| "loss": 0.4118, |
| "step": 781 |
| }, |
| { |
| "epoch": 4.197116996312437, |
| "grad_norm": 0.10687807539475037, |
| "learning_rate": 6.014581271373829e-06, |
| "loss": 0.4062, |
| "step": 782 |
| }, |
| { |
| "epoch": 4.202480724103252, |
| "grad_norm": 0.10262903089768421, |
| "learning_rate": 5.935643642325808e-06, |
| "loss": 0.4067, |
| "step": 783 |
| }, |
| { |
| "epoch": 4.207844451894067, |
| "grad_norm": 0.09927029282657747, |
| "learning_rate": 5.857185910799277e-06, |
| "loss": 0.4141, |
| "step": 784 |
| }, |
| { |
| "epoch": 4.213208179684881, |
| "grad_norm": 0.10324140566962033, |
| "learning_rate": 5.779209182104133e-06, |
| "loss": 0.4087, |
| "step": 785 |
| }, |
| { |
| "epoch": 4.218571907475695, |
| "grad_norm": 0.09857973369656504, |
| "learning_rate": 5.701714554773956e-06, |
| "loss": 0.4122, |
| "step": 786 |
| }, |
| { |
| "epoch": 4.2239356352665105, |
| "grad_norm": 0.10704699849654356, |
| "learning_rate": 5.624703120550492e-06, |
| "loss": 0.4132, |
| "step": 787 |
| }, |
| { |
| "epoch": 4.229299363057325, |
| "grad_norm": 0.09932400738261876, |
| "learning_rate": 5.548175964368248e-06, |
| "loss": 0.4136, |
| "step": 788 |
| }, |
| { |
| "epoch": 4.234663090848139, |
| "grad_norm": 0.10182388734864921, |
| "learning_rate": 5.4721341643392845e-06, |
| "loss": 0.4165, |
| "step": 789 |
| }, |
| { |
| "epoch": 4.240026818638954, |
| "grad_norm": 0.10458339193532643, |
| "learning_rate": 5.39657879173793e-06, |
| "loss": 0.4136, |
| "step": 790 |
| }, |
| { |
| "epoch": 4.2453905464297685, |
| "grad_norm": 0.10492190568796772, |
| "learning_rate": 5.3215109109857835e-06, |
| "loss": 0.4156, |
| "step": 791 |
| }, |
| { |
| "epoch": 4.250754274220584, |
| "grad_norm": 0.09833089016292879, |
| "learning_rate": 5.246931579636654e-06, |
| "loss": 0.4213, |
| "step": 792 |
| }, |
| { |
| "epoch": 4.256118002011398, |
| "grad_norm": 0.09631929734391337, |
| "learning_rate": 5.172841848361674e-06, |
| "loss": 0.4126, |
| "step": 793 |
| }, |
| { |
| "epoch": 4.261481729802212, |
| "grad_norm": 0.09916119673198169, |
| "learning_rate": 5.099242760934533e-06, |
| "loss": 0.4093, |
| "step": 794 |
| }, |
| { |
| "epoch": 4.2668454575930275, |
| "grad_norm": 0.09868198578261592, |
| "learning_rate": 5.026135354216717e-06, |
| "loss": 0.4136, |
| "step": 795 |
| }, |
| { |
| "epoch": 4.272209185383842, |
| "grad_norm": 0.09504747159177464, |
| "learning_rate": 4.953520658142958e-06, |
| "loss": 0.4156, |
| "step": 796 |
| }, |
| { |
| "epoch": 4.277572913174656, |
| "grad_norm": 0.09411067382234167, |
| "learning_rate": 4.881399695706677e-06, |
| "loss": 0.4136, |
| "step": 797 |
| }, |
| { |
| "epoch": 4.282936640965471, |
| "grad_norm": 0.10081365538952461, |
| "learning_rate": 4.809773482945601e-06, |
| "loss": 0.4122, |
| "step": 798 |
| }, |
| { |
| "epoch": 4.2883003687562855, |
| "grad_norm": 1.2440830128881666, |
| "learning_rate": 4.738643028927432e-06, |
| "loss": 0.4265, |
| "step": 799 |
| }, |
| { |
| "epoch": 4.293664096547101, |
| "grad_norm": 0.09904601559918061, |
| "learning_rate": 4.668009335735648e-06, |
| "loss": 0.4119, |
| "step": 800 |
| }, |
| { |
| "epoch": 4.299027824337915, |
| "grad_norm": 0.10410871952221983, |
| "learning_rate": 4.5978733984553835e-06, |
| "loss": 0.4116, |
| "step": 801 |
| }, |
| { |
| "epoch": 4.304391552128729, |
| "grad_norm": 0.11936741762617652, |
| "learning_rate": 4.528236205159386e-06, |
| "loss": 0.4163, |
| "step": 802 |
| }, |
| { |
| "epoch": 4.309755279919544, |
| "grad_norm": 0.7752585101606566, |
| "learning_rate": 4.459098736894114e-06, |
| "loss": 0.4168, |
| "step": 803 |
| }, |
| { |
| "epoch": 4.315119007710359, |
| "grad_norm": 0.10147283835455163, |
| "learning_rate": 4.39046196766594e-06, |
| "loss": 0.409, |
| "step": 804 |
| }, |
| { |
| "epoch": 4.320482735501173, |
| "grad_norm": 0.10494531512367294, |
| "learning_rate": 4.322326864427387e-06, |
| "loss": 0.4108, |
| "step": 805 |
| }, |
| { |
| "epoch": 4.325846463291988, |
| "grad_norm": 0.09812759496228979, |
| "learning_rate": 4.254694387063514e-06, |
| "loss": 0.4139, |
| "step": 806 |
| }, |
| { |
| "epoch": 4.3312101910828025, |
| "grad_norm": 0.10039748389385268, |
| "learning_rate": 4.187565488378434e-06, |
| "loss": 0.4099, |
| "step": 807 |
| }, |
| { |
| "epoch": 4.336573918873617, |
| "grad_norm": 0.1000059293045232, |
| "learning_rate": 4.120941114081833e-06, |
| "loss": 0.4112, |
| "step": 808 |
| }, |
| { |
| "epoch": 4.341937646664432, |
| "grad_norm": 0.09526496629638147, |
| "learning_rate": 4.0548222027756835e-06, |
| "loss": 0.4115, |
| "step": 809 |
| }, |
| { |
| "epoch": 4.347301374455246, |
| "grad_norm": 0.09813119886581101, |
| "learning_rate": 3.989209685941027e-06, |
| "loss": 0.412, |
| "step": 810 |
| }, |
| { |
| "epoch": 4.352665102246061, |
| "grad_norm": 0.09732532571840974, |
| "learning_rate": 3.924104487924805e-06, |
| "loss": 0.4061, |
| "step": 811 |
| }, |
| { |
| "epoch": 4.358028830036876, |
| "grad_norm": 0.09543190553673306, |
| "learning_rate": 3.859507525926897e-06, |
| "loss": 0.4141, |
| "step": 812 |
| }, |
| { |
| "epoch": 4.36339255782769, |
| "grad_norm": 0.09873317668913352, |
| "learning_rate": 3.795419709987149e-06, |
| "loss": 0.4153, |
| "step": 813 |
| }, |
| { |
| "epoch": 4.368756285618505, |
| "grad_norm": 0.23096877402734253, |
| "learning_rate": 3.7318419429726025e-06, |
| "loss": 0.4102, |
| "step": 814 |
| }, |
| { |
| "epoch": 4.374120013409319, |
| "grad_norm": 0.10185639946736366, |
| "learning_rate": 3.6687751205647117e-06, |
| "loss": 0.4145, |
| "step": 815 |
| }, |
| { |
| "epoch": 4.379483741200134, |
| "grad_norm": 0.38509111321061773, |
| "learning_rate": 3.606220131246776e-06, |
| "loss": 0.4162, |
| "step": 816 |
| }, |
| { |
| "epoch": 4.384847468990949, |
| "grad_norm": 0.09806919042690798, |
| "learning_rate": 3.5441778562914242e-06, |
| "loss": 0.4141, |
| "step": 817 |
| }, |
| { |
| "epoch": 4.390211196781763, |
| "grad_norm": 0.0969544886038649, |
| "learning_rate": 3.482649169748147e-06, |
| "loss": 0.4212, |
| "step": 818 |
| }, |
| { |
| "epoch": 4.395574924572578, |
| "grad_norm": 0.11496521800565809, |
| "learning_rate": 3.4216349384310533e-06, |
| "loss": 0.4115, |
| "step": 819 |
| }, |
| { |
| "epoch": 4.400938652363393, |
| "grad_norm": 0.09584588428602607, |
| "learning_rate": 3.3611360219065925e-06, |
| "loss": 0.4113, |
| "step": 820 |
| }, |
| { |
| "epoch": 4.406302380154207, |
| "grad_norm": 0.0925654699857564, |
| "learning_rate": 3.3011532724815142e-06, |
| "loss": 0.4102, |
| "step": 821 |
| }, |
| { |
| "epoch": 4.411666107945022, |
| "grad_norm": 0.09392896960421455, |
| "learning_rate": 3.241687535190776e-06, |
| "loss": 0.4105, |
| "step": 822 |
| }, |
| { |
| "epoch": 4.417029835735836, |
| "grad_norm": 0.09291428552953933, |
| "learning_rate": 3.1827396477857264e-06, |
| "loss": 0.4135, |
| "step": 823 |
| }, |
| { |
| "epoch": 4.422393563526651, |
| "grad_norm": 0.09323644383847197, |
| "learning_rate": 3.124310440722247e-06, |
| "loss": 0.4159, |
| "step": 824 |
| }, |
| { |
| "epoch": 4.427757291317466, |
| "grad_norm": 0.23059082156453775, |
| "learning_rate": 3.0664007371490558e-06, |
| "loss": 0.4185, |
| "step": 825 |
| }, |
| { |
| "epoch": 4.43312101910828, |
| "grad_norm": 0.09839755683407715, |
| "learning_rate": 3.009011352896152e-06, |
| "loss": 0.4144, |
| "step": 826 |
| }, |
| { |
| "epoch": 4.4384847468990944, |
| "grad_norm": 0.09990156860630273, |
| "learning_rate": 2.9521430964632602e-06, |
| "loss": 0.4056, |
| "step": 827 |
| }, |
| { |
| "epoch": 4.44384847468991, |
| "grad_norm": 0.093523667605744, |
| "learning_rate": 2.8957967690084986e-06, |
| "loss": 0.4165, |
| "step": 828 |
| }, |
| { |
| "epoch": 4.449212202480724, |
| "grad_norm": 0.09192510131863983, |
| "learning_rate": 2.839973164337044e-06, |
| "loss": 0.4115, |
| "step": 829 |
| }, |
| { |
| "epoch": 4.454575930271539, |
| "grad_norm": 0.09410786987122588, |
| "learning_rate": 2.7846730688900003e-06, |
| "loss": 0.4138, |
| "step": 830 |
| }, |
| { |
| "epoch": 4.459939658062353, |
| "grad_norm": 0.09707065586772506, |
| "learning_rate": 2.729897261733263e-06, |
| "loss": 0.4109, |
| "step": 831 |
| }, |
| { |
| "epoch": 4.465303385853168, |
| "grad_norm": 0.09262040362541676, |
| "learning_rate": 2.675646514546597e-06, |
| "loss": 0.4139, |
| "step": 832 |
| }, |
| { |
| "epoch": 4.470667113643983, |
| "grad_norm": 0.09786358492563979, |
| "learning_rate": 2.6219215916127283e-06, |
| "loss": 0.4112, |
| "step": 833 |
| }, |
| { |
| "epoch": 4.476030841434797, |
| "grad_norm": 0.09168116998441043, |
| "learning_rate": 2.568723249806575e-06, |
| "loss": 0.4106, |
| "step": 834 |
| }, |
| { |
| "epoch": 4.481394569225611, |
| "grad_norm": 0.09262516025294427, |
| "learning_rate": 2.516052238584625e-06, |
| "loss": 0.4149, |
| "step": 835 |
| }, |
| { |
| "epoch": 4.486758297016427, |
| "grad_norm": 0.09205026270318697, |
| "learning_rate": 2.463909299974323e-06, |
| "loss": 0.4062, |
| "step": 836 |
| }, |
| { |
| "epoch": 4.492122024807241, |
| "grad_norm": 0.09122547354614034, |
| "learning_rate": 2.4122951685636674e-06, |
| "loss": 0.4177, |
| "step": 837 |
| }, |
| { |
| "epoch": 4.497485752598056, |
| "grad_norm": 0.09211390919384611, |
| "learning_rate": 2.3612105714908173e-06, |
| "loss": 0.4046, |
| "step": 838 |
| }, |
| { |
| "epoch": 4.50284948038887, |
| "grad_norm": 0.091478554430513, |
| "learning_rate": 2.310656228433894e-06, |
| "loss": 0.412, |
| "step": 839 |
| }, |
| { |
| "epoch": 4.508213208179685, |
| "grad_norm": 0.09399408591730637, |
| "learning_rate": 2.260632851600795e-06, |
| "loss": 0.4149, |
| "step": 840 |
| }, |
| { |
| "epoch": 4.5135769359705, |
| "grad_norm": 0.09195607008303311, |
| "learning_rate": 2.211141145719191e-06, |
| "loss": 0.413, |
| "step": 841 |
| }, |
| { |
| "epoch": 4.518940663761314, |
| "grad_norm": 0.47959804465608913, |
| "learning_rate": 2.1621818080265955e-06, |
| "loss": 0.4182, |
| "step": 842 |
| }, |
| { |
| "epoch": 4.524304391552128, |
| "grad_norm": 0.09051553997088063, |
| "learning_rate": 2.1137555282605325e-06, |
| "loss": 0.4114, |
| "step": 843 |
| }, |
| { |
| "epoch": 4.5296681193429436, |
| "grad_norm": 0.08829292877572836, |
| "learning_rate": 2.0658629886488234e-06, |
| "loss": 0.4165, |
| "step": 844 |
| }, |
| { |
| "epoch": 4.535031847133758, |
| "grad_norm": 0.09080343947058375, |
| "learning_rate": 2.0185048638999706e-06, |
| "loss": 0.4142, |
| "step": 845 |
| }, |
| { |
| "epoch": 4.540395574924572, |
| "grad_norm": 0.09243369068431759, |
| "learning_rate": 1.9716818211936674e-06, |
| "loss": 0.4005, |
| "step": 846 |
| }, |
| { |
| "epoch": 4.545759302715387, |
| "grad_norm": 0.09353671619124167, |
| "learning_rate": 1.925394520171393e-06, |
| "loss": 0.4143, |
| "step": 847 |
| }, |
| { |
| "epoch": 4.551123030506202, |
| "grad_norm": 0.09072830507170011, |
| "learning_rate": 1.8796436129270955e-06, |
| "loss": 0.4116, |
| "step": 848 |
| }, |
| { |
| "epoch": 4.556486758297017, |
| "grad_norm": 0.09201634812096428, |
| "learning_rate": 1.8344297439980475e-06, |
| "loss": 0.4093, |
| "step": 849 |
| }, |
| { |
| "epoch": 4.561850486087831, |
| "grad_norm": 0.20856475534127397, |
| "learning_rate": 1.7897535503557196e-06, |
| "loss": 0.4161, |
| "step": 850 |
| }, |
| { |
| "epoch": 4.567214213878645, |
| "grad_norm": 0.0894737416874815, |
| "learning_rate": 1.7456156613968644e-06, |
| "loss": 0.4128, |
| "step": 851 |
| }, |
| { |
| "epoch": 4.5725779416694605, |
| "grad_norm": 0.09249497786193772, |
| "learning_rate": 1.702016698934581e-06, |
| "loss": 0.4104, |
| "step": 852 |
| }, |
| { |
| "epoch": 4.577941669460275, |
| "grad_norm": 0.10099307480578555, |
| "learning_rate": 1.65895727718961e-06, |
| "loss": 0.4134, |
| "step": 853 |
| }, |
| { |
| "epoch": 4.58330539725109, |
| "grad_norm": 0.0894656321450511, |
| "learning_rate": 1.6164380027816485e-06, |
| "loss": 0.4153, |
| "step": 854 |
| }, |
| { |
| "epoch": 4.588669125041904, |
| "grad_norm": 0.08952970498041048, |
| "learning_rate": 1.5744594747208308e-06, |
| "loss": 0.4118, |
| "step": 855 |
| }, |
| { |
| "epoch": 4.594032852832719, |
| "grad_norm": 0.08843654007502101, |
| "learning_rate": 1.5330222843992658e-06, |
| "loss": 0.4156, |
| "step": 856 |
| }, |
| { |
| "epoch": 4.599396580623534, |
| "grad_norm": 0.18591112023947728, |
| "learning_rate": 1.492127015582714e-06, |
| "loss": 0.4154, |
| "step": 857 |
| }, |
| { |
| "epoch": 4.604760308414348, |
| "grad_norm": 0.0944881909519439, |
| "learning_rate": 1.4517742444023665e-06, |
| "loss": 0.4129, |
| "step": 858 |
| }, |
| { |
| "epoch": 4.610124036205162, |
| "grad_norm": 0.09301181960862621, |
| "learning_rate": 1.4119645393467196e-06, |
| "loss": 0.4193, |
| "step": 859 |
| }, |
| { |
| "epoch": 4.6154877639959775, |
| "grad_norm": 0.08874447140825721, |
| "learning_rate": 1.3726984612535854e-06, |
| "loss": 0.4126, |
| "step": 860 |
| }, |
| { |
| "epoch": 4.620851491786792, |
| "grad_norm": 0.09037539019783776, |
| "learning_rate": 1.3339765633021551e-06, |
| "loss": 0.4164, |
| "step": 861 |
| }, |
| { |
| "epoch": 4.626215219577606, |
| "grad_norm": 0.09159608133309274, |
| "learning_rate": 1.2957993910052503e-06, |
| "loss": 0.4156, |
| "step": 862 |
| }, |
| { |
| "epoch": 4.631578947368421, |
| "grad_norm": 0.0876847986538798, |
| "learning_rate": 1.2581674822016087e-06, |
| "loss": 0.4133, |
| "step": 863 |
| }, |
| { |
| "epoch": 4.6369426751592355, |
| "grad_norm": 0.08889717326718397, |
| "learning_rate": 1.221081367048309e-06, |
| "loss": 0.4173, |
| "step": 864 |
| }, |
| { |
| "epoch": 4.64230640295005, |
| "grad_norm": 0.08911910146230485, |
| "learning_rate": 1.1845415680133089e-06, |
| "loss": 0.4041, |
| "step": 865 |
| }, |
| { |
| "epoch": 4.647670130740865, |
| "grad_norm": 0.0912001297197578, |
| "learning_rate": 1.1485485998680822e-06, |
| "loss": 0.4057, |
| "step": 866 |
| }, |
| { |
| "epoch": 4.653033858531679, |
| "grad_norm": 0.08889958044076257, |
| "learning_rate": 1.1131029696803774e-06, |
| "loss": 0.4078, |
| "step": 867 |
| }, |
| { |
| "epoch": 4.6583975863224945, |
| "grad_norm": 0.09028811157516371, |
| "learning_rate": 1.0782051768070477e-06, |
| "loss": 0.417, |
| "step": 868 |
| }, |
| { |
| "epoch": 4.663761314113309, |
| "grad_norm": 0.09213399459832071, |
| "learning_rate": 1.0438557128870408e-06, |
| "loss": 0.4148, |
| "step": 869 |
| }, |
| { |
| "epoch": 4.669125041904123, |
| "grad_norm": 0.08974457665914247, |
| "learning_rate": 1.010055061834474e-06, |
| "loss": 0.4075, |
| "step": 870 |
| }, |
| { |
| "epoch": 4.674488769694938, |
| "grad_norm": 0.08738862355780086, |
| "learning_rate": 9.768036998317875e-07, |
| "loss": 0.4167, |
| "step": 871 |
| }, |
| { |
| "epoch": 4.6798524974857525, |
| "grad_norm": 0.08744381941505991, |
| "learning_rate": 9.441020953230696e-07, |
| "loss": 0.4044, |
| "step": 872 |
| }, |
| { |
| "epoch": 4.685216225276568, |
| "grad_norm": 0.08768023798398505, |
| "learning_rate": 9.119507090074342e-07, |
| "loss": 0.4094, |
| "step": 873 |
| }, |
| { |
| "epoch": 4.690579953067382, |
| "grad_norm": 0.08529166544500341, |
| "learning_rate": 8.803499938325477e-07, |
| "loss": 0.4129, |
| "step": 874 |
| }, |
| { |
| "epoch": 4.695943680858196, |
| "grad_norm": 0.08637519839116772, |
| "learning_rate": 8.493003949882373e-07, |
| "loss": 0.4133, |
| "step": 875 |
| }, |
| { |
| "epoch": 4.701307408649011, |
| "grad_norm": 0.0887897705514168, |
| "learning_rate": 8.188023499002206e-07, |
| "loss": 0.4051, |
| "step": 876 |
| }, |
| { |
| "epoch": 4.706671136439826, |
| "grad_norm": 0.09055462354291803, |
| "learning_rate": 7.888562882239425e-07, |
| "loss": 0.4161, |
| "step": 877 |
| }, |
| { |
| "epoch": 4.71203486423064, |
| "grad_norm": 0.08675152093159026, |
| "learning_rate": 7.594626318385256e-07, |
| "loss": 0.4039, |
| "step": 878 |
| }, |
| { |
| "epoch": 4.717398592021455, |
| "grad_norm": 0.08841042189110157, |
| "learning_rate": 7.30621794840829e-07, |
| "loss": 0.411, |
| "step": 879 |
| }, |
| { |
| "epoch": 4.7227623198122695, |
| "grad_norm": 0.09048394753620571, |
| "learning_rate": 7.023341835396036e-07, |
| "loss": 0.4158, |
| "step": 880 |
| }, |
| { |
| "epoch": 4.728126047603084, |
| "grad_norm": 0.0937542868591101, |
| "learning_rate": 6.746001964497773e-07, |
| "loss": 0.4108, |
| "step": 881 |
| }, |
| { |
| "epoch": 4.733489775393899, |
| "grad_norm": 0.09183387737661795, |
| "learning_rate": 6.474202242868411e-07, |
| "loss": 0.4132, |
| "step": 882 |
| }, |
| { |
| "epoch": 4.738853503184713, |
| "grad_norm": 0.08673828591623912, |
| "learning_rate": 6.207946499613382e-07, |
| "loss": 0.4138, |
| "step": 883 |
| }, |
| { |
| "epoch": 4.7442172309755275, |
| "grad_norm": 0.1221653497024461, |
| "learning_rate": 5.947238485734819e-07, |
| "loss": 0.411, |
| "step": 884 |
| }, |
| { |
| "epoch": 4.749580958766343, |
| "grad_norm": 0.08800343336015064, |
| "learning_rate": 5.692081874078481e-07, |
| "loss": 0.4044, |
| "step": 885 |
| }, |
| { |
| "epoch": 4.754944686557157, |
| "grad_norm": 0.08826254435249375, |
| "learning_rate": 5.442480259282335e-07, |
| "loss": 0.414, |
| "step": 886 |
| }, |
| { |
| "epoch": 4.760308414347972, |
| "grad_norm": 0.0886779509344544, |
| "learning_rate": 5.198437157725567e-07, |
| "loss": 0.4116, |
| "step": 887 |
| }, |
| { |
| "epoch": 4.765672142138786, |
| "grad_norm": 0.09045462951093008, |
| "learning_rate": 4.959956007479338e-07, |
| "loss": 0.4035, |
| "step": 888 |
| }, |
| { |
| "epoch": 4.771035869929601, |
| "grad_norm": 0.08680316882639165, |
| "learning_rate": 4.7270401682581567e-07, |
| "loss": 0.415, |
| "step": 889 |
| }, |
| { |
| "epoch": 4.776399597720416, |
| "grad_norm": 0.0867270897509586, |
| "learning_rate": 4.499692921372667e-07, |
| "loss": 0.4143, |
| "step": 890 |
| }, |
| { |
| "epoch": 4.78176332551123, |
| "grad_norm": 0.08467163312345347, |
| "learning_rate": 4.277917469683246e-07, |
| "loss": 0.4081, |
| "step": 891 |
| }, |
| { |
| "epoch": 4.787127053302045, |
| "grad_norm": 0.14351442138219858, |
| "learning_rate": 4.061716937555149e-07, |
| "loss": 0.4149, |
| "step": 892 |
| }, |
| { |
| "epoch": 4.79249078109286, |
| "grad_norm": 0.08694159920502774, |
| "learning_rate": 3.851094370814323e-07, |
| "loss": 0.4173, |
| "step": 893 |
| }, |
| { |
| "epoch": 4.797854508883674, |
| "grad_norm": 0.08529615404396881, |
| "learning_rate": 3.646052736704464e-07, |
| "loss": 0.4062, |
| "step": 894 |
| }, |
| { |
| "epoch": 4.803218236674489, |
| "grad_norm": 0.08879884361022528, |
| "learning_rate": 3.4465949238453144e-07, |
| "loss": 0.4197, |
| "step": 895 |
| }, |
| { |
| "epoch": 4.808581964465303, |
| "grad_norm": 0.0862432359092941, |
| "learning_rate": 3.252723742191899e-07, |
| "loss": 0.4139, |
| "step": 896 |
| }, |
| { |
| "epoch": 4.813945692256118, |
| "grad_norm": 0.08670212847739438, |
| "learning_rate": 3.064441922995043e-07, |
| "loss": 0.421, |
| "step": 897 |
| }, |
| { |
| "epoch": 4.819309420046933, |
| "grad_norm": 0.0861260158634291, |
| "learning_rate": 2.8817521187626926e-07, |
| "loss": 0.4114, |
| "step": 898 |
| }, |
| { |
| "epoch": 4.824673147837747, |
| "grad_norm": 0.13041000271914696, |
| "learning_rate": 2.704656903222791e-07, |
| "loss": 0.4089, |
| "step": 899 |
| }, |
| { |
| "epoch": 4.830036875628561, |
| "grad_norm": 0.0853917982287929, |
| "learning_rate": 2.533158771286903e-07, |
| "loss": 0.4189, |
| "step": 900 |
| }, |
| { |
| "epoch": 4.835400603419377, |
| "grad_norm": 0.08585480202942483, |
| "learning_rate": 2.3672601390148707e-07, |
| "loss": 0.4102, |
| "step": 901 |
| }, |
| { |
| "epoch": 4.840764331210191, |
| "grad_norm": 0.08514537079223002, |
| "learning_rate": 2.206963343581281e-07, |
| "loss": 0.4186, |
| "step": 902 |
| }, |
| { |
| "epoch": 4.846128059001006, |
| "grad_norm": 0.0854519613880308, |
| "learning_rate": 2.0522706432419382e-07, |
| "loss": 0.4166, |
| "step": 903 |
| }, |
| { |
| "epoch": 4.85149178679182, |
| "grad_norm": 0.08595768719036412, |
| "learning_rate": 1.903184217302556e-07, |
| "loss": 0.4138, |
| "step": 904 |
| }, |
| { |
| "epoch": 4.856855514582635, |
| "grad_norm": 0.08593142424333411, |
| "learning_rate": 1.7597061660877157e-07, |
| "loss": 0.4187, |
| "step": 905 |
| }, |
| { |
| "epoch": 4.86221924237345, |
| "grad_norm": 0.08605519303696639, |
| "learning_rate": 1.6218385109114665e-07, |
| "loss": 0.4152, |
| "step": 906 |
| }, |
| { |
| "epoch": 4.867582970164264, |
| "grad_norm": 0.08778080722891127, |
| "learning_rate": 1.4895831940486827e-07, |
| "loss": 0.408, |
| "step": 907 |
| }, |
| { |
| "epoch": 4.872946697955078, |
| "grad_norm": 0.5236720260419594, |
| "learning_rate": 1.3629420787079738e-07, |
| "loss": 0.4086, |
| "step": 908 |
| }, |
| { |
| "epoch": 4.878310425745894, |
| "grad_norm": 0.08691961311755944, |
| "learning_rate": 1.2419169490051287e-07, |
| "loss": 0.4162, |
| "step": 909 |
| }, |
| { |
| "epoch": 4.883674153536708, |
| "grad_norm": 0.08604130165924209, |
| "learning_rate": 1.1265095099381118e-07, |
| "loss": 0.4161, |
| "step": 910 |
| }, |
| { |
| "epoch": 4.889037881327523, |
| "grad_norm": 0.08562809530743228, |
| "learning_rate": 1.0167213873631732e-07, |
| "loss": 0.4068, |
| "step": 911 |
| }, |
| { |
| "epoch": 4.894401609118337, |
| "grad_norm": 0.08788772763541514, |
| "learning_rate": 9.125541279717098e-08, |
| "loss": 0.4117, |
| "step": 912 |
| }, |
| { |
| "epoch": 4.899765336909152, |
| "grad_norm": 0.08465883837484985, |
| "learning_rate": 8.14009199268595e-08, |
| "loss": 0.412, |
| "step": 913 |
| }, |
| { |
| "epoch": 4.905129064699967, |
| "grad_norm": 0.08635396133541602, |
| "learning_rate": 7.210879895515277e-08, |
| "loss": 0.4114, |
| "step": 914 |
| }, |
| { |
| "epoch": 4.910492792490781, |
| "grad_norm": 0.08583085092727974, |
| "learning_rate": 6.337918078914041e-08, |
| "loss": 0.4129, |
| "step": 915 |
| }, |
| { |
| "epoch": 4.915856520281595, |
| "grad_norm": 0.08664926249547512, |
| "learning_rate": 5.521218841139764e-08, |
| "loss": 0.4154, |
| "step": 916 |
| }, |
| { |
| "epoch": 4.9212202480724105, |
| "grad_norm": 0.08448787096930517, |
| "learning_rate": 4.7607936878235614e-08, |
| "loss": 0.406, |
| "step": 917 |
| }, |
| { |
| "epoch": 4.926583975863225, |
| "grad_norm": 0.0850836523204834, |
| "learning_rate": 4.0566533318102676e-08, |
| "loss": 0.4122, |
| "step": 918 |
| }, |
| { |
| "epoch": 4.931947703654039, |
| "grad_norm": 0.08844457157599679, |
| "learning_rate": 3.408807693006111e-08, |
| "loss": 0.4186, |
| "step": 919 |
| }, |
| { |
| "epoch": 4.937311431444854, |
| "grad_norm": 0.08507338195494901, |
| "learning_rate": 2.817265898237942e-08, |
| "loss": 0.4044, |
| "step": 920 |
| }, |
| { |
| "epoch": 4.942675159235669, |
| "grad_norm": 0.08409639688790976, |
| "learning_rate": 2.2820362811279973e-08, |
| "loss": 0.4084, |
| "step": 921 |
| }, |
| { |
| "epoch": 4.948038887026484, |
| "grad_norm": 0.08816595457478153, |
| "learning_rate": 1.8031263819726642e-08, |
| "loss": 0.4145, |
| "step": 922 |
| }, |
| { |
| "epoch": 4.953402614817298, |
| "grad_norm": 0.08680564134665435, |
| "learning_rate": 1.3805429476385634e-08, |
| "loss": 0.4068, |
| "step": 923 |
| }, |
| { |
| "epoch": 4.958766342608112, |
| "grad_norm": 0.09052432530510993, |
| "learning_rate": 1.0142919314679588e-08, |
| "loss": 0.4153, |
| "step": 924 |
| }, |
| { |
| "epoch": 4.9641300703989275, |
| "grad_norm": 0.09320878153934302, |
| "learning_rate": 7.043784931921593e-09, |
| "loss": 0.4103, |
| "step": 925 |
| }, |
| { |
| "epoch": 4.969493798189742, |
| "grad_norm": 0.08708424563628227, |
| "learning_rate": 4.508069988617969e-09, |
| "loss": 0.4131, |
| "step": 926 |
| }, |
| { |
| "epoch": 4.974857525980557, |
| "grad_norm": 0.08389684177916036, |
| "learning_rate": 2.5358102078376635e-09, |
| "loss": 0.4079, |
| "step": 927 |
| }, |
| { |
| "epoch": 4.980221253771371, |
| "grad_norm": 0.08673227922167183, |
| "learning_rate": 1.127033374705988e-09, |
| "loss": 0.4102, |
| "step": 928 |
| }, |
| { |
| "epoch": 4.9855849815621855, |
| "grad_norm": 0.08435533244348804, |
| "learning_rate": 2.8175933603158403e-10, |
| "loss": 0.412, |
| "step": 929 |
| }, |
| { |
| "epoch": 4.990948709353001, |
| "grad_norm": 0.0856513074559934, |
| "learning_rate": 0.0, |
| "loss": 0.4095, |
| "step": 930 |
| }, |
| { |
| "epoch": 4.990948709353001, |
| "step": 930, |
| "total_flos": 2.393166916865753e+19, |
| "train_loss": 0.5230806940986265, |
| "train_runtime": 141229.2058, |
| "train_samples_per_second": 3.379, |
| "train_steps_per_second": 0.007 |
| } |
| ], |
| "logging_steps": 1.0, |
| "max_steps": 930, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "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": 2.393166916865753e+19, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
