| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9640213461869512, | |
| "eval_steps": 200, | |
| "global_step": 1400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006885866758478223, | |
| "grad_norm": 1.5874253511428833, | |
| "learning_rate": 0.0, | |
| "loss": 0.1378, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0013771733516956446, | |
| "grad_norm": 2.887026309967041, | |
| "learning_rate": 1.0273972602739726e-08, | |
| "loss": 0.1301, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002065760027543467, | |
| "grad_norm": 1.5245386362075806, | |
| "learning_rate": 2.0547945205479452e-08, | |
| "loss": 0.1035, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.002754346703391289, | |
| "grad_norm": 1.233545184135437, | |
| "learning_rate": 3.082191780821918e-08, | |
| "loss": 0.1226, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.003442933379239112, | |
| "grad_norm": 2.1326398849487305, | |
| "learning_rate": 4.1095890410958903e-08, | |
| "loss": 0.1215, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.004131520055086934, | |
| "grad_norm": 2.2501494884490967, | |
| "learning_rate": 5.136986301369863e-08, | |
| "loss": 0.1366, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0048201067309347565, | |
| "grad_norm": 2.3572044372558594, | |
| "learning_rate": 6.164383561643836e-08, | |
| "loss": 0.1283, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.005508693406782578, | |
| "grad_norm": 1.3646363019943237, | |
| "learning_rate": 7.191780821917807e-08, | |
| "loss": 0.1099, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.006197280082630401, | |
| "grad_norm": 3.3486526012420654, | |
| "learning_rate": 8.219178082191781e-08, | |
| "loss": 0.1232, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.006885866758478224, | |
| "grad_norm": 2.454549789428711, | |
| "learning_rate": 9.246575342465753e-08, | |
| "loss": 0.1485, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.007574453434326046, | |
| "grad_norm": 1.8196218013763428, | |
| "learning_rate": 1.0273972602739726e-07, | |
| "loss": 0.1124, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.008263040110173868, | |
| "grad_norm": 2.2766010761260986, | |
| "learning_rate": 1.1301369863013698e-07, | |
| "loss": 0.1295, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.008951626786021691, | |
| "grad_norm": 1.2352460622787476, | |
| "learning_rate": 1.232876712328767e-07, | |
| "loss": 0.13, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.009640213461869513, | |
| "grad_norm": 5.661056995391846, | |
| "learning_rate": 1.3356164383561644e-07, | |
| "loss": 0.1215, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.010328800137717335, | |
| "grad_norm": 1.2685704231262207, | |
| "learning_rate": 1.4383561643835615e-07, | |
| "loss": 0.1033, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.011017386813565157, | |
| "grad_norm": 3.1796512603759766, | |
| "learning_rate": 1.5410958904109588e-07, | |
| "loss": 0.1133, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.01170597348941298, | |
| "grad_norm": 4.14359188079834, | |
| "learning_rate": 1.6438356164383561e-07, | |
| "loss": 0.1147, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.012394560165260802, | |
| "grad_norm": 1.7137943506240845, | |
| "learning_rate": 1.7465753424657535e-07, | |
| "loss": 0.1102, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.013083146841108624, | |
| "grad_norm": 1.869036078453064, | |
| "learning_rate": 1.8493150684931505e-07, | |
| "loss": 0.1395, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.013771733516956448, | |
| "grad_norm": 1.4325156211853027, | |
| "learning_rate": 1.9520547945205478e-07, | |
| "loss": 0.1051, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01446032019280427, | |
| "grad_norm": 1.183656096458435, | |
| "learning_rate": 2.0547945205479452e-07, | |
| "loss": 0.0956, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.015148906868652091, | |
| "grad_norm": 1.6014477014541626, | |
| "learning_rate": 2.1575342465753425e-07, | |
| "loss": 0.0966, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.015837493544499915, | |
| "grad_norm": 1.5043327808380127, | |
| "learning_rate": 2.2602739726027396e-07, | |
| "loss": 0.1223, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.016526080220347735, | |
| "grad_norm": 5.508129119873047, | |
| "learning_rate": 2.363013698630137e-07, | |
| "loss": 0.135, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.01721466689619556, | |
| "grad_norm": 2.1750261783599854, | |
| "learning_rate": 2.465753424657534e-07, | |
| "loss": 0.1441, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.017903253572043382, | |
| "grad_norm": 4.23982572555542, | |
| "learning_rate": 2.568493150684931e-07, | |
| "loss": 0.1431, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.018591840247891202, | |
| "grad_norm": 1.3620706796646118, | |
| "learning_rate": 2.671232876712329e-07, | |
| "loss": 0.1046, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.019280426923739026, | |
| "grad_norm": 1.568567156791687, | |
| "learning_rate": 2.773972602739726e-07, | |
| "loss": 0.1234, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.01996901359958685, | |
| "grad_norm": 1.8222662210464478, | |
| "learning_rate": 2.876712328767123e-07, | |
| "loss": 0.1215, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.02065760027543467, | |
| "grad_norm": 2.091226816177368, | |
| "learning_rate": 2.9794520547945206e-07, | |
| "loss": 0.1224, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.021346186951282493, | |
| "grad_norm": 1.533294439315796, | |
| "learning_rate": 3.0821917808219176e-07, | |
| "loss": 0.1036, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.022034773627130314, | |
| "grad_norm": 3.3673996925354004, | |
| "learning_rate": 3.184931506849315e-07, | |
| "loss": 0.1002, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.022723360302978137, | |
| "grad_norm": 2.213367223739624, | |
| "learning_rate": 3.2876712328767123e-07, | |
| "loss": 0.1086, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.02341194697882596, | |
| "grad_norm": 1.5319344997406006, | |
| "learning_rate": 3.3904109589041093e-07, | |
| "loss": 0.1209, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.02410053365467378, | |
| "grad_norm": 2.030628204345703, | |
| "learning_rate": 3.493150684931507e-07, | |
| "loss": 0.1081, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.024789120330521604, | |
| "grad_norm": 4.031522750854492, | |
| "learning_rate": 3.595890410958904e-07, | |
| "loss": 0.1263, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.025477707006369428, | |
| "grad_norm": 2.492892026901245, | |
| "learning_rate": 3.698630136986301e-07, | |
| "loss": 0.1018, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.026166293682217248, | |
| "grad_norm": 1.691985845565796, | |
| "learning_rate": 3.801369863013699e-07, | |
| "loss": 0.1146, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.026854880358065072, | |
| "grad_norm": 3.1290664672851562, | |
| "learning_rate": 3.9041095890410957e-07, | |
| "loss": 0.1281, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.027543467033912895, | |
| "grad_norm": 1.4528034925460815, | |
| "learning_rate": 4.0068493150684933e-07, | |
| "loss": 0.1094, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.028232053709760715, | |
| "grad_norm": 2.95461106300354, | |
| "learning_rate": 4.1095890410958903e-07, | |
| "loss": 0.1161, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.02892064038560854, | |
| "grad_norm": 1.3321444988250732, | |
| "learning_rate": 4.212328767123288e-07, | |
| "loss": 0.1146, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.02960922706145636, | |
| "grad_norm": 2.259939432144165, | |
| "learning_rate": 4.315068493150685e-07, | |
| "loss": 0.1176, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.030297813737304183, | |
| "grad_norm": 4.673003196716309, | |
| "learning_rate": 4.4178082191780826e-07, | |
| "loss": 0.1196, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.030986400413152006, | |
| "grad_norm": 2.461071014404297, | |
| "learning_rate": 4.520547945205479e-07, | |
| "loss": 0.1163, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03167498708899983, | |
| "grad_norm": 2.1340038776397705, | |
| "learning_rate": 4.623287671232877e-07, | |
| "loss": 0.1189, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.03236357376484765, | |
| "grad_norm": 3.9452526569366455, | |
| "learning_rate": 4.726027397260274e-07, | |
| "loss": 0.1394, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03305216044069547, | |
| "grad_norm": 3.155883550643921, | |
| "learning_rate": 4.828767123287671e-07, | |
| "loss": 0.1083, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0337407471165433, | |
| "grad_norm": 3.6826388835906982, | |
| "learning_rate": 4.931506849315068e-07, | |
| "loss": 0.1077, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03442933379239112, | |
| "grad_norm": 3.4725987911224365, | |
| "learning_rate": 5.034246575342467e-07, | |
| "loss": 0.1194, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03511792046823894, | |
| "grad_norm": 2.661391496658325, | |
| "learning_rate": 5.136986301369863e-07, | |
| "loss": 0.1108, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.035806507144086765, | |
| "grad_norm": 1.5056617259979248, | |
| "learning_rate": 5.239726027397261e-07, | |
| "loss": 0.1195, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.036495093819934585, | |
| "grad_norm": 1.8250921964645386, | |
| "learning_rate": 5.342465753424658e-07, | |
| "loss": 0.1049, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.037183680495782405, | |
| "grad_norm": 1.7526438236236572, | |
| "learning_rate": 5.445205479452055e-07, | |
| "loss": 0.1198, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.03787226717163023, | |
| "grad_norm": 1.9744874238967896, | |
| "learning_rate": 5.547945205479452e-07, | |
| "loss": 0.1232, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03856085384747805, | |
| "grad_norm": 2.8387508392333984, | |
| "learning_rate": 5.65068493150685e-07, | |
| "loss": 0.116, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.03924944052332587, | |
| "grad_norm": 2.2203245162963867, | |
| "learning_rate": 5.753424657534246e-07, | |
| "loss": 0.1064, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0399380271991737, | |
| "grad_norm": 1.6614450216293335, | |
| "learning_rate": 5.856164383561644e-07, | |
| "loss": 0.109, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.04062661387502152, | |
| "grad_norm": 3.8924262523651123, | |
| "learning_rate": 5.958904109589041e-07, | |
| "loss": 0.0901, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.04131520055086934, | |
| "grad_norm": 2.0824718475341797, | |
| "learning_rate": 6.061643835616438e-07, | |
| "loss": 0.1262, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04200378722671716, | |
| "grad_norm": 2.9020180702209473, | |
| "learning_rate": 6.164383561643835e-07, | |
| "loss": 0.1036, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.04269237390256499, | |
| "grad_norm": 1.4103049039840698, | |
| "learning_rate": 6.267123287671233e-07, | |
| "loss": 0.0987, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.04338096057841281, | |
| "grad_norm": 1.1437251567840576, | |
| "learning_rate": 6.36986301369863e-07, | |
| "loss": 0.0982, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.04406954725426063, | |
| "grad_norm": 1.1304177045822144, | |
| "learning_rate": 6.472602739726027e-07, | |
| "loss": 0.1046, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.044758133930108454, | |
| "grad_norm": 1.483987808227539, | |
| "learning_rate": 6.575342465753425e-07, | |
| "loss": 0.1394, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.045446720605956274, | |
| "grad_norm": 0.9523391127586365, | |
| "learning_rate": 6.678082191780823e-07, | |
| "loss": 0.0872, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.046135307281804094, | |
| "grad_norm": 1.3211452960968018, | |
| "learning_rate": 6.780821917808219e-07, | |
| "loss": 0.1217, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.04682389395765192, | |
| "grad_norm": 1.481589436531067, | |
| "learning_rate": 6.883561643835617e-07, | |
| "loss": 0.1074, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.04751248063349974, | |
| "grad_norm": 1.7939581871032715, | |
| "learning_rate": 6.986301369863014e-07, | |
| "loss": 0.1063, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.04820106730934756, | |
| "grad_norm": 2.629153251647949, | |
| "learning_rate": 7.089041095890411e-07, | |
| "loss": 0.1216, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04888965398519539, | |
| "grad_norm": 4.473081111907959, | |
| "learning_rate": 7.191780821917808e-07, | |
| "loss": 0.1217, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.04957824066104321, | |
| "grad_norm": 2.7944514751434326, | |
| "learning_rate": 7.294520547945206e-07, | |
| "loss": 0.096, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.05026682733689103, | |
| "grad_norm": 1.3520607948303223, | |
| "learning_rate": 7.397260273972602e-07, | |
| "loss": 0.0894, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.050955414012738856, | |
| "grad_norm": 1.3510197401046753, | |
| "learning_rate": 7.5e-07, | |
| "loss": 0.086, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.051644000688586676, | |
| "grad_norm": 2.2130050659179688, | |
| "learning_rate": 7.602739726027398e-07, | |
| "loss": 0.1053, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.052332587364434496, | |
| "grad_norm": 1.8730335235595703, | |
| "learning_rate": 7.705479452054795e-07, | |
| "loss": 0.118, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.05302117404028232, | |
| "grad_norm": 4.524833679199219, | |
| "learning_rate": 7.808219178082191e-07, | |
| "loss": 0.0962, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.053709760716130144, | |
| "grad_norm": 2.070676326751709, | |
| "learning_rate": 7.91095890410959e-07, | |
| "loss": 0.1206, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.054398347391977964, | |
| "grad_norm": 4.2869648933410645, | |
| "learning_rate": 8.013698630136987e-07, | |
| "loss": 0.0986, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.05508693406782579, | |
| "grad_norm": 2.3519368171691895, | |
| "learning_rate": 8.116438356164385e-07, | |
| "loss": 0.103, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05577552074367361, | |
| "grad_norm": 2.1209590435028076, | |
| "learning_rate": 8.219178082191781e-07, | |
| "loss": 0.1354, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.05646410741952143, | |
| "grad_norm": 1.8561289310455322, | |
| "learning_rate": 8.321917808219178e-07, | |
| "loss": 0.1163, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.05715269409536926, | |
| "grad_norm": 3.3910579681396484, | |
| "learning_rate": 8.424657534246576e-07, | |
| "loss": 0.0895, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.05784128077121708, | |
| "grad_norm": 1.069259762763977, | |
| "learning_rate": 8.527397260273974e-07, | |
| "loss": 0.093, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.0585298674470649, | |
| "grad_norm": 1.2470117807388306, | |
| "learning_rate": 8.63013698630137e-07, | |
| "loss": 0.08, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.05921845412291272, | |
| "grad_norm": 5.087569713592529, | |
| "learning_rate": 8.732876712328767e-07, | |
| "loss": 0.1091, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.059907040798760545, | |
| "grad_norm": 1.0401315689086914, | |
| "learning_rate": 8.835616438356165e-07, | |
| "loss": 0.1311, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.060595627474608366, | |
| "grad_norm": 1.485530972480774, | |
| "learning_rate": 8.938356164383562e-07, | |
| "loss": 0.1062, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.061284214150456186, | |
| "grad_norm": 1.2075278759002686, | |
| "learning_rate": 9.041095890410958e-07, | |
| "loss": 0.1027, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.06197280082630401, | |
| "grad_norm": 1.7765151262283325, | |
| "learning_rate": 9.143835616438356e-07, | |
| "loss": 0.1408, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06266138750215183, | |
| "grad_norm": 2.0588860511779785, | |
| "learning_rate": 9.246575342465754e-07, | |
| "loss": 0.0929, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.06334997417799966, | |
| "grad_norm": 1.2527514696121216, | |
| "learning_rate": 9.349315068493152e-07, | |
| "loss": 0.104, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.06403856085384747, | |
| "grad_norm": 1.4009658098220825, | |
| "learning_rate": 9.452054794520548e-07, | |
| "loss": 0.1272, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.0647271475296953, | |
| "grad_norm": 1.8319647312164307, | |
| "learning_rate": 9.554794520547946e-07, | |
| "loss": 0.0844, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.06541573420554313, | |
| "grad_norm": 3.1829638481140137, | |
| "learning_rate": 9.657534246575343e-07, | |
| "loss": 0.1366, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06610432088139094, | |
| "grad_norm": 1.3211784362792969, | |
| "learning_rate": 9.76027397260274e-07, | |
| "loss": 0.1012, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.06679290755723877, | |
| "grad_norm": 0.9261050820350647, | |
| "learning_rate": 9.863013698630137e-07, | |
| "loss": 0.0849, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.0674814942330866, | |
| "grad_norm": 1.131234884262085, | |
| "learning_rate": 9.965753424657534e-07, | |
| "loss": 0.0824, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.06817008090893441, | |
| "grad_norm": 4.053731918334961, | |
| "learning_rate": 1.0068493150684933e-06, | |
| "loss": 0.083, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.06885866758478223, | |
| "grad_norm": 0.857279896736145, | |
| "learning_rate": 1.017123287671233e-06, | |
| "loss": 0.0787, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06954725426063006, | |
| "grad_norm": 2.5226166248321533, | |
| "learning_rate": 1.0273972602739725e-06, | |
| "loss": 0.1149, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.07023584093647788, | |
| "grad_norm": 1.1416969299316406, | |
| "learning_rate": 1.0376712328767124e-06, | |
| "loss": 0.1192, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.0709244276123257, | |
| "grad_norm": 2.4472241401672363, | |
| "learning_rate": 1.0479452054794521e-06, | |
| "loss": 0.0866, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.07161301428817353, | |
| "grad_norm": 1.6066569089889526, | |
| "learning_rate": 1.0582191780821918e-06, | |
| "loss": 0.0828, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.07230160096402134, | |
| "grad_norm": 1.4382472038269043, | |
| "learning_rate": 1.0684931506849315e-06, | |
| "loss": 0.0769, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.07299018763986917, | |
| "grad_norm": 0.9668927192687988, | |
| "learning_rate": 1.0787671232876712e-06, | |
| "loss": 0.0851, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.073678774315717, | |
| "grad_norm": 1.8680490255355835, | |
| "learning_rate": 1.089041095890411e-06, | |
| "loss": 0.0789, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.07436736099156481, | |
| "grad_norm": 1.3016374111175537, | |
| "learning_rate": 1.0993150684931509e-06, | |
| "loss": 0.0866, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.07505594766741264, | |
| "grad_norm": 1.1077319383621216, | |
| "learning_rate": 1.1095890410958904e-06, | |
| "loss": 0.0805, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.07574453434326046, | |
| "grad_norm": 1.5466182231903076, | |
| "learning_rate": 1.11986301369863e-06, | |
| "loss": 0.0947, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07643312101910828, | |
| "grad_norm": 3.4772799015045166, | |
| "learning_rate": 1.13013698630137e-06, | |
| "loss": 0.1183, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.0771217076949561, | |
| "grad_norm": 1.2641544342041016, | |
| "learning_rate": 1.1404109589041097e-06, | |
| "loss": 0.0591, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.07781029437080393, | |
| "grad_norm": 1.66828453540802, | |
| "learning_rate": 1.1506849315068492e-06, | |
| "loss": 0.0984, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.07849888104665174, | |
| "grad_norm": 2.605940341949463, | |
| "learning_rate": 1.160958904109589e-06, | |
| "loss": 0.0859, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.07918746772249957, | |
| "grad_norm": 2.146186351776123, | |
| "learning_rate": 1.1712328767123288e-06, | |
| "loss": 0.1184, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0798760543983474, | |
| "grad_norm": 1.3627345561981201, | |
| "learning_rate": 1.1815068493150685e-06, | |
| "loss": 0.0947, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.08056464107419521, | |
| "grad_norm": 1.2085281610488892, | |
| "learning_rate": 1.1917808219178082e-06, | |
| "loss": 0.1049, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.08125322775004304, | |
| "grad_norm": 1.843820333480835, | |
| "learning_rate": 1.202054794520548e-06, | |
| "loss": 0.1, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.08194181442589087, | |
| "grad_norm": 1.540426254272461, | |
| "learning_rate": 1.2123287671232876e-06, | |
| "loss": 0.084, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.08263040110173868, | |
| "grad_norm": 1.6418871879577637, | |
| "learning_rate": 1.2226027397260276e-06, | |
| "loss": 0.0781, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.0833189877775865, | |
| "grad_norm": 1.3550796508789062, | |
| "learning_rate": 1.232876712328767e-06, | |
| "loss": 0.0957, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.08400757445343432, | |
| "grad_norm": 1.7492413520812988, | |
| "learning_rate": 1.2431506849315068e-06, | |
| "loss": 0.0825, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.08469616112928215, | |
| "grad_norm": 3.9397377967834473, | |
| "learning_rate": 1.2534246575342467e-06, | |
| "loss": 0.1042, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.08538474780512997, | |
| "grad_norm": 0.572508692741394, | |
| "learning_rate": 1.2636986301369864e-06, | |
| "loss": 0.0803, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.08607333448097779, | |
| "grad_norm": 0.9744763374328613, | |
| "learning_rate": 1.273972602739726e-06, | |
| "loss": 0.0743, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08676192115682561, | |
| "grad_norm": 1.3532994985580444, | |
| "learning_rate": 1.2842465753424658e-06, | |
| "loss": 0.0957, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.08745050783267344, | |
| "grad_norm": 1.25874924659729, | |
| "learning_rate": 1.2945205479452055e-06, | |
| "loss": 0.0831, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.08813909450852125, | |
| "grad_norm": 0.786328136920929, | |
| "learning_rate": 1.3047945205479452e-06, | |
| "loss": 0.0897, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.08882768118436908, | |
| "grad_norm": 0.8441939353942871, | |
| "learning_rate": 1.315068493150685e-06, | |
| "loss": 0.0752, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.08951626786021691, | |
| "grad_norm": 1.7906488180160522, | |
| "learning_rate": 1.3253424657534246e-06, | |
| "loss": 0.0966, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.09020485453606472, | |
| "grad_norm": 1.4520841836929321, | |
| "learning_rate": 1.3356164383561645e-06, | |
| "loss": 0.1341, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.09089344121191255, | |
| "grad_norm": 1.3951728343963623, | |
| "learning_rate": 1.3458904109589042e-06, | |
| "loss": 0.0898, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.09158202788776038, | |
| "grad_norm": 1.0645533800125122, | |
| "learning_rate": 1.3561643835616437e-06, | |
| "loss": 0.0754, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.09227061456360819, | |
| "grad_norm": 2.151301145553589, | |
| "learning_rate": 1.3664383561643836e-06, | |
| "loss": 0.0711, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.09295920123945602, | |
| "grad_norm": 1.4348841905593872, | |
| "learning_rate": 1.3767123287671234e-06, | |
| "loss": 0.0672, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.09364778791530384, | |
| "grad_norm": 1.8859331607818604, | |
| "learning_rate": 1.386986301369863e-06, | |
| "loss": 0.0938, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.09433637459115166, | |
| "grad_norm": 4.374384880065918, | |
| "learning_rate": 1.3972602739726028e-06, | |
| "loss": 0.0772, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.09502496126699948, | |
| "grad_norm": 1.7930854558944702, | |
| "learning_rate": 1.4075342465753425e-06, | |
| "loss": 0.0854, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.09571354794284731, | |
| "grad_norm": 2.969715118408203, | |
| "learning_rate": 1.4178082191780822e-06, | |
| "loss": 0.1152, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.09640213461869512, | |
| "grad_norm": 3.479867935180664, | |
| "learning_rate": 1.428082191780822e-06, | |
| "loss": 0.1028, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09709072129454295, | |
| "grad_norm": 2.0213847160339355, | |
| "learning_rate": 1.4383561643835616e-06, | |
| "loss": 0.1096, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.09777930797039078, | |
| "grad_norm": 0.92966228723526, | |
| "learning_rate": 1.4486301369863013e-06, | |
| "loss": 0.0792, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.09846789464623859, | |
| "grad_norm": 2.740649938583374, | |
| "learning_rate": 1.4589041095890412e-06, | |
| "loss": 0.0571, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.09915648132208642, | |
| "grad_norm": 1.513682246208191, | |
| "learning_rate": 1.469178082191781e-06, | |
| "loss": 0.0997, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.09984506799793424, | |
| "grad_norm": 4.071322441101074, | |
| "learning_rate": 1.4794520547945204e-06, | |
| "loss": 0.1025, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.10053365467378206, | |
| "grad_norm": 2.535306453704834, | |
| "learning_rate": 1.4897260273972603e-06, | |
| "loss": 0.0992, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.10122224134962989, | |
| "grad_norm": 1.0436655282974243, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.0725, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.10191082802547771, | |
| "grad_norm": 3.913482427597046, | |
| "learning_rate": 1.4999978333955598e-06, | |
| "loss": 0.0648, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.10259941470132553, | |
| "grad_norm": 2.273953676223755, | |
| "learning_rate": 1.499991333594757e-06, | |
| "loss": 0.0944, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.10328800137717335, | |
| "grad_norm": 0.742541491985321, | |
| "learning_rate": 1.4999805006351445e-06, | |
| "loss": 0.0847, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.10397658805302118, | |
| "grad_norm": 0.7346779108047485, | |
| "learning_rate": 1.4999653345793113e-06, | |
| "loss": 0.0556, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.10466517472886899, | |
| "grad_norm": 3.0982794761657715, | |
| "learning_rate": 1.4999458355148812e-06, | |
| "loss": 0.0737, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.10535376140471682, | |
| "grad_norm": 2.795851945877075, | |
| "learning_rate": 1.499922003554512e-06, | |
| "loss": 0.0752, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.10604234808056465, | |
| "grad_norm": 0.8724387288093567, | |
| "learning_rate": 1.4998938388358954e-06, | |
| "loss": 0.0797, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.10673093475641246, | |
| "grad_norm": 0.9423794150352478, | |
| "learning_rate": 1.499861341521756e-06, | |
| "loss": 0.0909, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.10741952143226029, | |
| "grad_norm": 1.4669538736343384, | |
| "learning_rate": 1.4998245117998516e-06, | |
| "loss": 0.1168, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.10810810810810811, | |
| "grad_norm": 3.28376841545105, | |
| "learning_rate": 1.499783349882969e-06, | |
| "loss": 0.0852, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.10879669478395593, | |
| "grad_norm": 0.9011422395706177, | |
| "learning_rate": 1.4997378560089264e-06, | |
| "loss": 0.1331, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.10948528145980375, | |
| "grad_norm": 1.4032050371170044, | |
| "learning_rate": 1.4996880304405692e-06, | |
| "loss": 0.0797, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.11017386813565158, | |
| "grad_norm": 2.6546716690063477, | |
| "learning_rate": 1.4996338734657707e-06, | |
| "loss": 0.0806, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1108624548114994, | |
| "grad_norm": 10.069770812988281, | |
| "learning_rate": 1.4995753853974287e-06, | |
| "loss": 0.0645, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.11155104148734722, | |
| "grad_norm": 1.4410676956176758, | |
| "learning_rate": 1.4995125665734647e-06, | |
| "loss": 0.098, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.11223962816319505, | |
| "grad_norm": 2.0679640769958496, | |
| "learning_rate": 1.4994454173568212e-06, | |
| "loss": 0.0914, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.11292821483904286, | |
| "grad_norm": 0.6671273112297058, | |
| "learning_rate": 1.4993739381354604e-06, | |
| "loss": 0.0732, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.11361680151489069, | |
| "grad_norm": 0.8500239849090576, | |
| "learning_rate": 1.4992981293223616e-06, | |
| "loss": 0.074, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.11430538819073852, | |
| "grad_norm": 1.320983648300171, | |
| "learning_rate": 1.499217991355519e-06, | |
| "loss": 0.0916, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.11499397486658633, | |
| "grad_norm": 1.8575754165649414, | |
| "learning_rate": 1.4991335246979376e-06, | |
| "loss": 0.0752, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.11568256154243416, | |
| "grad_norm": 1.9247512817382812, | |
| "learning_rate": 1.4990447298376342e-06, | |
| "loss": 0.075, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.11637114821828197, | |
| "grad_norm": 2.7277333736419678, | |
| "learning_rate": 1.4989516072876302e-06, | |
| "loss": 0.1134, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.1170597348941298, | |
| "grad_norm": 1.0003513097763062, | |
| "learning_rate": 1.498854157585952e-06, | |
| "loss": 0.089, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.11774832156997762, | |
| "grad_norm": 0.9851368069648743, | |
| "learning_rate": 1.4987523812956263e-06, | |
| "loss": 0.0756, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.11843690824582544, | |
| "grad_norm": 0.9414574503898621, | |
| "learning_rate": 1.4986462790046762e-06, | |
| "loss": 0.0919, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.11912549492167326, | |
| "grad_norm": 1.6840728521347046, | |
| "learning_rate": 1.4985358513261205e-06, | |
| "loss": 0.0866, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.11981408159752109, | |
| "grad_norm": 1.581800103187561, | |
| "learning_rate": 1.4984210988979667e-06, | |
| "loss": 0.0741, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.1205026682733689, | |
| "grad_norm": 5.528469085693359, | |
| "learning_rate": 1.49830202238321e-06, | |
| "loss": 0.0904, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.12119125494921673, | |
| "grad_norm": 2.592730760574341, | |
| "learning_rate": 1.498178622469829e-06, | |
| "loss": 0.0728, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.12187984162506456, | |
| "grad_norm": 1.9295653104782104, | |
| "learning_rate": 1.498050899870779e-06, | |
| "loss": 0.0674, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.12256842830091237, | |
| "grad_norm": 1.214874029159546, | |
| "learning_rate": 1.4979188553239933e-06, | |
| "loss": 0.0839, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.1232570149767602, | |
| "grad_norm": 1.151961088180542, | |
| "learning_rate": 1.4977824895923727e-06, | |
| "loss": 0.0928, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.12394560165260803, | |
| "grad_norm": 1.966047763824463, | |
| "learning_rate": 1.4976418034637863e-06, | |
| "loss": 0.0691, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.12463418832845584, | |
| "grad_norm": 1.163633942604065, | |
| "learning_rate": 1.4974967977510637e-06, | |
| "loss": 0.0725, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.12532277500430367, | |
| "grad_norm": 1.1855480670928955, | |
| "learning_rate": 1.497347473291991e-06, | |
| "loss": 0.0618, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.12601136168015148, | |
| "grad_norm": 1.2485895156860352, | |
| "learning_rate": 1.4971938309493083e-06, | |
| "loss": 0.0815, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.12669994835599932, | |
| "grad_norm": 1.1105977296829224, | |
| "learning_rate": 1.4970358716107002e-06, | |
| "loss": 0.0837, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.12738853503184713, | |
| "grad_norm": 0.9790412187576294, | |
| "learning_rate": 1.496873596188795e-06, | |
| "loss": 0.0791, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.12807712170769495, | |
| "grad_norm": 1.4629616737365723, | |
| "learning_rate": 1.496707005621157e-06, | |
| "loss": 0.0912, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.1287657083835428, | |
| "grad_norm": 1.36525559425354, | |
| "learning_rate": 1.4965361008702815e-06, | |
| "loss": 0.0885, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.1294542950593906, | |
| "grad_norm": 2.7682502269744873, | |
| "learning_rate": 1.4963608829235905e-06, | |
| "loss": 0.0801, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.1301428817352384, | |
| "grad_norm": 0.8608899116516113, | |
| "learning_rate": 1.4961813527934247e-06, | |
| "loss": 0.0694, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.13083146841108625, | |
| "grad_norm": 0.822589635848999, | |
| "learning_rate": 1.49599751151704e-06, | |
| "loss": 0.081, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.13152005508693407, | |
| "grad_norm": 0.818540632724762, | |
| "learning_rate": 1.4958093601565992e-06, | |
| "loss": 0.093, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.13220864176278188, | |
| "grad_norm": 1.299375295639038, | |
| "learning_rate": 1.4956168997991687e-06, | |
| "loss": 0.1023, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.13289722843862972, | |
| "grad_norm": 1.2853596210479736, | |
| "learning_rate": 1.4954201315567092e-06, | |
| "loss": 0.0734, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.13358581511447754, | |
| "grad_norm": 3.053440570831299, | |
| "learning_rate": 1.4952190565660715e-06, | |
| "loss": 0.0708, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.13427440179032535, | |
| "grad_norm": 1.0819454193115234, | |
| "learning_rate": 1.4950136759889885e-06, | |
| "loss": 0.068, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.1349629884661732, | |
| "grad_norm": 1.9725898504257202, | |
| "learning_rate": 1.4948039910120698e-06, | |
| "loss": 0.0809, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.135651575142021, | |
| "grad_norm": 1.2086131572723389, | |
| "learning_rate": 1.4945900028467934e-06, | |
| "loss": 0.0854, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.13634016181786882, | |
| "grad_norm": 1.4241052865982056, | |
| "learning_rate": 1.4943717127295003e-06, | |
| "loss": 0.0694, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.13702874849371666, | |
| "grad_norm": 1.4852848052978516, | |
| "learning_rate": 1.494149121921386e-06, | |
| "loss": 0.0809, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.13771733516956447, | |
| "grad_norm": 0.8755311965942383, | |
| "learning_rate": 1.4939222317084934e-06, | |
| "loss": 0.0726, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13771733516956447, | |
| "eval_loss": NaN, | |
| "eval_runtime": 84.2314, | |
| "eval_samples_per_second": 11.16, | |
| "eval_steps_per_second": 0.356, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13840592184541228, | |
| "grad_norm": 1.5212584733963013, | |
| "learning_rate": 1.4936910434017064e-06, | |
| "loss": 0.0645, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.13909450852126012, | |
| "grad_norm": 2.667065143585205, | |
| "learning_rate": 1.4934555583367412e-06, | |
| "loss": 0.0924, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.13978309519710794, | |
| "grad_norm": 1.479569911956787, | |
| "learning_rate": 1.4932157778741394e-06, | |
| "loss": 0.0694, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.14047168187295575, | |
| "grad_norm": 0.9887720346450806, | |
| "learning_rate": 1.4929717033992592e-06, | |
| "loss": 0.0752, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.1411602685488036, | |
| "grad_norm": 1.2442289590835571, | |
| "learning_rate": 1.4927233363222682e-06, | |
| "loss": 0.0788, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.1418488552246514, | |
| "grad_norm": 1.651498794555664, | |
| "learning_rate": 1.4924706780781346e-06, | |
| "loss": 0.0716, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.14253744190049922, | |
| "grad_norm": 1.3058912754058838, | |
| "learning_rate": 1.4922137301266201e-06, | |
| "loss": 0.0736, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.14322602857634706, | |
| "grad_norm": 2.032660484313965, | |
| "learning_rate": 1.4919524939522702e-06, | |
| "loss": 0.0948, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.14391461525219487, | |
| "grad_norm": 2.2852706909179688, | |
| "learning_rate": 1.491686971064406e-06, | |
| "loss": 0.091, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.14460320192804269, | |
| "grad_norm": 1.2744024991989136, | |
| "learning_rate": 1.4914171629971158e-06, | |
| "loss": 0.0641, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.14529178860389053, | |
| "grad_norm": 1.1681936979293823, | |
| "learning_rate": 1.4911430713092457e-06, | |
| "loss": 0.0872, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.14598037527973834, | |
| "grad_norm": 1.0655605792999268, | |
| "learning_rate": 1.4908646975843913e-06, | |
| "loss": 0.0714, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.14666896195558615, | |
| "grad_norm": 2.672621726989746, | |
| "learning_rate": 1.4905820434308877e-06, | |
| "loss": 0.072, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.147357548631434, | |
| "grad_norm": 1.1016126871109009, | |
| "learning_rate": 1.490295110481801e-06, | |
| "loss": 0.0834, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.1480461353072818, | |
| "grad_norm": 1.2881873846054077, | |
| "learning_rate": 1.4900039003949186e-06, | |
| "loss": 0.0705, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.14873472198312962, | |
| "grad_norm": 1.059262990951538, | |
| "learning_rate": 1.4897084148527387e-06, | |
| "loss": 0.0774, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.14942330865897746, | |
| "grad_norm": 1.5575706958770752, | |
| "learning_rate": 1.489408655562463e-06, | |
| "loss": 0.0744, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.15011189533482527, | |
| "grad_norm": 2.3055214881896973, | |
| "learning_rate": 1.4891046242559834e-06, | |
| "loss": 0.0835, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.1508004820106731, | |
| "grad_norm": 2.3411362171173096, | |
| "learning_rate": 1.4887963226898756e-06, | |
| "loss": 0.0669, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.15148906868652093, | |
| "grad_norm": 0.925907552242279, | |
| "learning_rate": 1.4884837526453857e-06, | |
| "loss": 0.0755, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.15217765536236874, | |
| "grad_norm": 1.427729606628418, | |
| "learning_rate": 1.4881669159284226e-06, | |
| "loss": 0.0873, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.15286624203821655, | |
| "grad_norm": 0.9466798305511475, | |
| "learning_rate": 1.4878458143695456e-06, | |
| "loss": 0.0665, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.1535548287140644, | |
| "grad_norm": 2.1973841190338135, | |
| "learning_rate": 1.4875204498239548e-06, | |
| "loss": 0.1028, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.1542434153899122, | |
| "grad_norm": 0.9114904403686523, | |
| "learning_rate": 1.4871908241714804e-06, | |
| "loss": 0.0695, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.15493200206576002, | |
| "grad_norm": 1.8247462511062622, | |
| "learning_rate": 1.4868569393165714e-06, | |
| "loss": 0.0812, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.15562058874160786, | |
| "grad_norm": 1.0007606744766235, | |
| "learning_rate": 1.486518797188285e-06, | |
| "loss": 0.0638, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.15630917541745568, | |
| "grad_norm": 0.9683484435081482, | |
| "learning_rate": 1.4861763997402753e-06, | |
| "loss": 0.0798, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.1569977620933035, | |
| "grad_norm": 0.763083279132843, | |
| "learning_rate": 1.4858297489507813e-06, | |
| "loss": 0.0674, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.15768634876915133, | |
| "grad_norm": 0.9944437742233276, | |
| "learning_rate": 1.4854788468226172e-06, | |
| "loss": 0.0853, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.15837493544499914, | |
| "grad_norm": 0.9368122816085815, | |
| "learning_rate": 1.4851236953831592e-06, | |
| "loss": 0.0688, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.15906352212084696, | |
| "grad_norm": 0.6713323593139648, | |
| "learning_rate": 1.4847642966843345e-06, | |
| "loss": 0.072, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.1597521087966948, | |
| "grad_norm": 1.0318669080734253, | |
| "learning_rate": 1.484400652802609e-06, | |
| "loss": 0.0735, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.1604406954725426, | |
| "grad_norm": 1.212418556213379, | |
| "learning_rate": 1.4840327658389762e-06, | |
| "loss": 0.0694, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.16112928214839042, | |
| "grad_norm": 1.4916894435882568, | |
| "learning_rate": 1.4836606379189438e-06, | |
| "loss": 0.1084, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.16181786882423826, | |
| "grad_norm": 1.3681950569152832, | |
| "learning_rate": 1.483284271192523e-06, | |
| "loss": 0.0698, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.16250645550008608, | |
| "grad_norm": 1.0745643377304077, | |
| "learning_rate": 1.4829036678342143e-06, | |
| "loss": 0.0672, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.1631950421759339, | |
| "grad_norm": 0.7977475523948669, | |
| "learning_rate": 1.4825188300429964e-06, | |
| "loss": 0.0644, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.16388362885178173, | |
| "grad_norm": 2.3173305988311768, | |
| "learning_rate": 1.4821297600423124e-06, | |
| "loss": 0.0884, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.16457221552762955, | |
| "grad_norm": 1.0547959804534912, | |
| "learning_rate": 1.4817364600800578e-06, | |
| "loss": 0.0734, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.16526080220347736, | |
| "grad_norm": 1.466292142868042, | |
| "learning_rate": 1.4813389324285674e-06, | |
| "loss": 0.0764, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.16594938887932517, | |
| "grad_norm": 1.8221032619476318, | |
| "learning_rate": 1.4809371793846014e-06, | |
| "loss": 0.0731, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.166637975555173, | |
| "grad_norm": 1.3423607349395752, | |
| "learning_rate": 1.4805312032693328e-06, | |
| "loss": 0.0796, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.16732656223102083, | |
| "grad_norm": 1.1996288299560547, | |
| "learning_rate": 1.4801210064283344e-06, | |
| "loss": 0.063, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.16801514890686864, | |
| "grad_norm": 2.8401691913604736, | |
| "learning_rate": 1.479706591231564e-06, | |
| "loss": 0.1115, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.16870373558271648, | |
| "grad_norm": 2.180941581726074, | |
| "learning_rate": 1.4792879600733523e-06, | |
| "loss": 0.0782, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.1693923222585643, | |
| "grad_norm": 0.8152029514312744, | |
| "learning_rate": 1.4788651153723867e-06, | |
| "loss": 0.0717, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.1700809089344121, | |
| "grad_norm": 1.4131059646606445, | |
| "learning_rate": 1.4784380595717006e-06, | |
| "loss": 0.0982, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.17076949561025995, | |
| "grad_norm": 0.6604526042938232, | |
| "learning_rate": 1.4780067951386564e-06, | |
| "loss": 0.0666, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.17145808228610776, | |
| "grad_norm": 1.7243674993515015, | |
| "learning_rate": 1.4775713245649324e-06, | |
| "loss": 0.0559, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.17214666896195557, | |
| "grad_norm": 1.2893307209014893, | |
| "learning_rate": 1.4771316503665087e-06, | |
| "loss": 0.0726, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.17283525563780341, | |
| "grad_norm": 0.6680165529251099, | |
| "learning_rate": 1.4766877750836521e-06, | |
| "loss": 0.0581, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.17352384231365123, | |
| "grad_norm": 1.9170761108398438, | |
| "learning_rate": 1.4762397012809018e-06, | |
| "loss": 0.075, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.17421242898949904, | |
| "grad_norm": 1.1394575834274292, | |
| "learning_rate": 1.475787431547054e-06, | |
| "loss": 0.0819, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.17490101566534688, | |
| "grad_norm": 1.3237025737762451, | |
| "learning_rate": 1.4753309684951481e-06, | |
| "loss": 0.0683, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.1755896023411947, | |
| "grad_norm": 1.7598793506622314, | |
| "learning_rate": 1.47487031476245e-06, | |
| "loss": 0.0967, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.1762781890170425, | |
| "grad_norm": 1.3406288623809814, | |
| "learning_rate": 1.4744054730104387e-06, | |
| "loss": 0.0812, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.17696677569289035, | |
| "grad_norm": 0.6121050715446472, | |
| "learning_rate": 1.4739364459247888e-06, | |
| "loss": 0.0581, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.17765536236873816, | |
| "grad_norm": 1.8967972993850708, | |
| "learning_rate": 1.4734632362153573e-06, | |
| "loss": 0.0751, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.17834394904458598, | |
| "grad_norm": 2.95050311088562, | |
| "learning_rate": 1.4729858466161657e-06, | |
| "loss": 0.0589, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.17903253572043382, | |
| "grad_norm": 2.333331823348999, | |
| "learning_rate": 1.4725042798853862e-06, | |
| "loss": 0.0958, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.17972112239628163, | |
| "grad_norm": 2.1714529991149902, | |
| "learning_rate": 1.4720185388053244e-06, | |
| "loss": 0.0777, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.18040970907212944, | |
| "grad_norm": 1.1433035135269165, | |
| "learning_rate": 1.4715286261824036e-06, | |
| "loss": 0.0883, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.18109829574797728, | |
| "grad_norm": 1.5877351760864258, | |
| "learning_rate": 1.4710345448471487e-06, | |
| "loss": 0.0724, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.1817868824238251, | |
| "grad_norm": 0.9370065927505493, | |
| "learning_rate": 1.47053629765417e-06, | |
| "loss": 0.0758, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.1824754690996729, | |
| "grad_norm": 1.0883688926696777, | |
| "learning_rate": 1.4700338874821463e-06, | |
| "loss": 0.0775, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.18316405577552075, | |
| "grad_norm": 1.8363934755325317, | |
| "learning_rate": 1.4695273172338088e-06, | |
| "loss": 0.1059, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.18385264245136856, | |
| "grad_norm": 1.6309915781021118, | |
| "learning_rate": 1.4690165898359232e-06, | |
| "loss": 0.071, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.18454122912721638, | |
| "grad_norm": 1.3420311212539673, | |
| "learning_rate": 1.4685017082392748e-06, | |
| "loss": 0.0633, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.18522981580306422, | |
| "grad_norm": 0.8939677476882935, | |
| "learning_rate": 1.4679826754186493e-06, | |
| "loss": 0.0509, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.18591840247891203, | |
| "grad_norm": 1.6941150426864624, | |
| "learning_rate": 1.4674594943728168e-06, | |
| "loss": 0.0718, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.18660698915475984, | |
| "grad_norm": 0.8134205341339111, | |
| "learning_rate": 1.4669321681245148e-06, | |
| "loss": 0.0615, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.18729557583060769, | |
| "grad_norm": 2.2507824897766113, | |
| "learning_rate": 1.466400699720429e-06, | |
| "loss": 0.0884, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.1879841625064555, | |
| "grad_norm": 1.3280833959579468, | |
| "learning_rate": 1.465865092231178e-06, | |
| "loss": 0.0772, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.1886727491823033, | |
| "grad_norm": 1.360690951347351, | |
| "learning_rate": 1.4653253487512935e-06, | |
| "loss": 0.07, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.18936133585815115, | |
| "grad_norm": 1.1449198722839355, | |
| "learning_rate": 1.4647814723992044e-06, | |
| "loss": 0.0818, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.19004992253399897, | |
| "grad_norm": 1.3556039333343506, | |
| "learning_rate": 1.4642334663172168e-06, | |
| "loss": 0.0853, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.19073850920984678, | |
| "grad_norm": 0.913157045841217, | |
| "learning_rate": 1.4636813336714972e-06, | |
| "loss": 0.0622, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.19142709588569462, | |
| "grad_norm": 0.7217982411384583, | |
| "learning_rate": 1.4631250776520536e-06, | |
| "loss": 0.0692, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.19211568256154243, | |
| "grad_norm": 0.8475648760795593, | |
| "learning_rate": 1.4625647014727178e-06, | |
| "loss": 0.0804, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.19280426923739025, | |
| "grad_norm": 1.1824219226837158, | |
| "learning_rate": 1.4620002083711252e-06, | |
| "loss": 0.0787, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1934928559132381, | |
| "grad_norm": 1.7809088230133057, | |
| "learning_rate": 1.4614316016086983e-06, | |
| "loss": 0.0588, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.1941814425890859, | |
| "grad_norm": 1.6327368021011353, | |
| "learning_rate": 1.4608588844706261e-06, | |
| "loss": 0.064, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.19487002926493371, | |
| "grad_norm": 0.8754256367683411, | |
| "learning_rate": 1.4602820602658461e-06, | |
| "loss": 0.0792, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.19555861594078155, | |
| "grad_norm": 0.7843604683876038, | |
| "learning_rate": 1.4597011323270242e-06, | |
| "loss": 0.0602, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.19624720261662937, | |
| "grad_norm": 1.9893841743469238, | |
| "learning_rate": 1.4591161040105372e-06, | |
| "loss": 0.073, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.19693578929247718, | |
| "grad_norm": 0.7418936491012573, | |
| "learning_rate": 1.4585269786964512e-06, | |
| "loss": 0.0757, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.19762437596832502, | |
| "grad_norm": 0.8246138095855713, | |
| "learning_rate": 1.4579337597885034e-06, | |
| "loss": 0.0612, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.19831296264417284, | |
| "grad_norm": 1.3396852016448975, | |
| "learning_rate": 1.4573364507140827e-06, | |
| "loss": 0.0849, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.19900154932002065, | |
| "grad_norm": 1.2449584007263184, | |
| "learning_rate": 1.4567350549242091e-06, | |
| "loss": 0.087, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.1996901359958685, | |
| "grad_norm": 1.2178400754928589, | |
| "learning_rate": 1.4561295758935142e-06, | |
| "loss": 0.0849, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2003787226717163, | |
| "grad_norm": 0.8280997276306152, | |
| "learning_rate": 1.4555200171202204e-06, | |
| "loss": 0.0632, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.20106730934756412, | |
| "grad_norm": 1.4488314390182495, | |
| "learning_rate": 1.4549063821261218e-06, | |
| "loss": 0.066, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.20175589602341196, | |
| "grad_norm": 2.008378267288208, | |
| "learning_rate": 1.4542886744565632e-06, | |
| "loss": 0.0626, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.20244448269925977, | |
| "grad_norm": 1.3806315660476685, | |
| "learning_rate": 1.4536668976804201e-06, | |
| "loss": 0.0808, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.20313306937510758, | |
| "grad_norm": 1.518605351448059, | |
| "learning_rate": 1.4530410553900767e-06, | |
| "loss": 0.0663, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.20382165605095542, | |
| "grad_norm": 0.6927845478057861, | |
| "learning_rate": 1.4524111512014075e-06, | |
| "loss": 0.0574, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.20451024272680324, | |
| "grad_norm": 2.699159860610962, | |
| "learning_rate": 1.4517771887537541e-06, | |
| "loss": 0.0719, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.20519882940265105, | |
| "grad_norm": 0.9919278025627136, | |
| "learning_rate": 1.4511391717099052e-06, | |
| "loss": 0.069, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.2058874160784989, | |
| "grad_norm": 1.0398292541503906, | |
| "learning_rate": 1.4504971037560761e-06, | |
| "loss": 0.0673, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.2065760027543467, | |
| "grad_norm": 2.8945765495300293, | |
| "learning_rate": 1.449850988601886e-06, | |
| "loss": 0.1054, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.20726458943019452, | |
| "grad_norm": 0.999775767326355, | |
| "learning_rate": 1.4492008299803375e-06, | |
| "loss": 0.0788, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.20795317610604236, | |
| "grad_norm": 1.3050086498260498, | |
| "learning_rate": 1.4485466316477945e-06, | |
| "loss": 0.064, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.20864176278189017, | |
| "grad_norm": 1.0358120203018188, | |
| "learning_rate": 1.4478883973839614e-06, | |
| "loss": 0.0688, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.20933034945773799, | |
| "grad_norm": 1.0159348249435425, | |
| "learning_rate": 1.44722613099186e-06, | |
| "loss": 0.0686, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.21001893613358583, | |
| "grad_norm": 1.2458326816558838, | |
| "learning_rate": 1.4465598362978086e-06, | |
| "loss": 0.0665, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.21070752280943364, | |
| "grad_norm": 1.3187693357467651, | |
| "learning_rate": 1.4458895171513993e-06, | |
| "loss": 0.0774, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.21139610948528145, | |
| "grad_norm": 1.6980345249176025, | |
| "learning_rate": 1.445215177425476e-06, | |
| "loss": 0.0782, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.2120846961611293, | |
| "grad_norm": 0.9206858277320862, | |
| "learning_rate": 1.4445368210161116e-06, | |
| "loss": 0.0459, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.2127732828369771, | |
| "grad_norm": 0.7377665638923645, | |
| "learning_rate": 1.4438544518425866e-06, | |
| "loss": 0.0523, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.21346186951282492, | |
| "grad_norm": 1.3793962001800537, | |
| "learning_rate": 1.4431680738473647e-06, | |
| "loss": 0.08, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.21415045618867276, | |
| "grad_norm": 0.9102684855461121, | |
| "learning_rate": 1.442477690996072e-06, | |
| "loss": 0.0651, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.21483904286452057, | |
| "grad_norm": 1.606674313545227, | |
| "learning_rate": 1.4417833072774721e-06, | |
| "loss": 0.077, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.2155276295403684, | |
| "grad_norm": 0.9576410055160522, | |
| "learning_rate": 1.4410849267034452e-06, | |
| "loss": 0.0671, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.21621621621621623, | |
| "grad_norm": 0.6433281302452087, | |
| "learning_rate": 1.4403825533089628e-06, | |
| "loss": 0.056, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.21690480289206404, | |
| "grad_norm": 1.459549069404602, | |
| "learning_rate": 1.4396761911520656e-06, | |
| "loss": 0.0774, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.21759338956791185, | |
| "grad_norm": 0.6870599389076233, | |
| "learning_rate": 1.4389658443138404e-06, | |
| "loss": 0.0643, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.2182819762437597, | |
| "grad_norm": 0.7249013781547546, | |
| "learning_rate": 1.4382515168983953e-06, | |
| "loss": 0.0639, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.2189705629196075, | |
| "grad_norm": 2.184231996536255, | |
| "learning_rate": 1.4375332130328366e-06, | |
| "loss": 0.0739, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.21965914959545532, | |
| "grad_norm": 0.781061053276062, | |
| "learning_rate": 1.4368109368672457e-06, | |
| "loss": 0.0592, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.22034773627130316, | |
| "grad_norm": 0.595102071762085, | |
| "learning_rate": 1.4360846925746535e-06, | |
| "loss": 0.0552, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.22103632294715098, | |
| "grad_norm": 0.770702064037323, | |
| "learning_rate": 1.4353544843510178e-06, | |
| "loss": 0.0727, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.2217249096229988, | |
| "grad_norm": 1.023535966873169, | |
| "learning_rate": 1.4346203164151984e-06, | |
| "loss": 0.0772, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.22241349629884663, | |
| "grad_norm": 1.0462602376937866, | |
| "learning_rate": 1.4338821930089322e-06, | |
| "loss": 0.0725, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.22310208297469444, | |
| "grad_norm": 0.9319475293159485, | |
| "learning_rate": 1.4331401183968103e-06, | |
| "loss": 0.0685, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.22379066965054226, | |
| "grad_norm": 1.269640564918518, | |
| "learning_rate": 1.4323940968662517e-06, | |
| "loss": 0.0643, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.2244792563263901, | |
| "grad_norm": 1.1411255598068237, | |
| "learning_rate": 1.4316441327274787e-06, | |
| "loss": 0.0908, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.2251678430022379, | |
| "grad_norm": 1.5144968032836914, | |
| "learning_rate": 1.4308902303134935e-06, | |
| "loss": 0.0825, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.22585642967808572, | |
| "grad_norm": 0.9155645966529846, | |
| "learning_rate": 1.4301323939800514e-06, | |
| "loss": 0.0673, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.22654501635393356, | |
| "grad_norm": 1.5559560060501099, | |
| "learning_rate": 1.4293706281056369e-06, | |
| "loss": 0.0794, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.22723360302978138, | |
| "grad_norm": 1.014963984489441, | |
| "learning_rate": 1.428604937091437e-06, | |
| "loss": 0.0765, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.2279221897056292, | |
| "grad_norm": 1.5839638710021973, | |
| "learning_rate": 1.4278353253613174e-06, | |
| "loss": 0.0704, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.22861077638147703, | |
| "grad_norm": 2.0470452308654785, | |
| "learning_rate": 1.4270617973617962e-06, | |
| "loss": 0.0844, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.22929936305732485, | |
| "grad_norm": 0.9143749475479126, | |
| "learning_rate": 1.4262843575620175e-06, | |
| "loss": 0.0655, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.22998794973317266, | |
| "grad_norm": 1.262157678604126, | |
| "learning_rate": 1.4255030104537268e-06, | |
| "loss": 0.0598, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.23067653640902047, | |
| "grad_norm": 1.0421985387802124, | |
| "learning_rate": 1.4247177605512447e-06, | |
| "loss": 0.0875, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.2313651230848683, | |
| "grad_norm": 1.1137545108795166, | |
| "learning_rate": 1.42392861239144e-06, | |
| "loss": 0.0668, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.23205370976071613, | |
| "grad_norm": 0.9526805877685547, | |
| "learning_rate": 1.4231355705337047e-06, | |
| "loss": 0.077, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.23274229643656394, | |
| "grad_norm": 0.8724916577339172, | |
| "learning_rate": 1.4223386395599267e-06, | |
| "loss": 0.0691, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.23343088311241178, | |
| "grad_norm": 0.6975535154342651, | |
| "learning_rate": 1.4215378240744638e-06, | |
| "loss": 0.0539, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.2341194697882596, | |
| "grad_norm": 1.0890775918960571, | |
| "learning_rate": 1.420733128704117e-06, | |
| "loss": 0.0668, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.2348080564641074, | |
| "grad_norm": 1.278756856918335, | |
| "learning_rate": 1.4199245580981043e-06, | |
| "loss": 0.0716, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.23549664313995525, | |
| "grad_norm": 3.335198163986206, | |
| "learning_rate": 1.419112116928032e-06, | |
| "loss": 0.1018, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.23618522981580306, | |
| "grad_norm": 2.4635980129241943, | |
| "learning_rate": 1.41829580988787e-06, | |
| "loss": 0.0704, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.23687381649165087, | |
| "grad_norm": 1.4701200723648071, | |
| "learning_rate": 1.4174756416939244e-06, | |
| "loss": 0.0762, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.23756240316749871, | |
| "grad_norm": 1.6528270244598389, | |
| "learning_rate": 1.4166516170848075e-06, | |
| "loss": 0.0783, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.23825098984334653, | |
| "grad_norm": 1.0045890808105469, | |
| "learning_rate": 1.415823740821414e-06, | |
| "loss": 0.0648, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.23893957651919434, | |
| "grad_norm": 0.7579544186592102, | |
| "learning_rate": 1.4149920176868922e-06, | |
| "loss": 0.0736, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.23962816319504218, | |
| "grad_norm": 0.9279513359069824, | |
| "learning_rate": 1.4141564524866149e-06, | |
| "loss": 0.0695, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.24031674987089, | |
| "grad_norm": 1.477232575416565, | |
| "learning_rate": 1.4133170500481535e-06, | |
| "loss": 0.0633, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.2410053365467378, | |
| "grad_norm": 1.8822599649429321, | |
| "learning_rate": 1.4124738152212497e-06, | |
| "loss": 0.0738, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.24169392322258565, | |
| "grad_norm": 1.2123138904571533, | |
| "learning_rate": 1.4116267528777872e-06, | |
| "loss": 0.0701, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.24238250989843346, | |
| "grad_norm": 1.932442545890808, | |
| "learning_rate": 1.410775867911763e-06, | |
| "loss": 0.1131, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.24307109657428128, | |
| "grad_norm": 3.9569449424743652, | |
| "learning_rate": 1.4099211652392602e-06, | |
| "loss": 0.0692, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.24375968325012912, | |
| "grad_norm": 1.976380705833435, | |
| "learning_rate": 1.4090626497984193e-06, | |
| "loss": 0.0742, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.24444826992597693, | |
| "grad_norm": 1.6463158130645752, | |
| "learning_rate": 1.4082003265494093e-06, | |
| "loss": 0.0725, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.24513685660182474, | |
| "grad_norm": 1.5660310983657837, | |
| "learning_rate": 1.4073342004743988e-06, | |
| "loss": 0.0668, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.24582544327767258, | |
| "grad_norm": 1.7993230819702148, | |
| "learning_rate": 1.4064642765775287e-06, | |
| "loss": 0.0881, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.2465140299535204, | |
| "grad_norm": 1.2761452198028564, | |
| "learning_rate": 1.4055905598848812e-06, | |
| "loss": 0.0763, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.2472026166293682, | |
| "grad_norm": 0.9085745811462402, | |
| "learning_rate": 1.4047130554444523e-06, | |
| "loss": 0.0737, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.24789120330521605, | |
| "grad_norm": 1.363052487373352, | |
| "learning_rate": 1.403831768326122e-06, | |
| "loss": 0.085, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.24857978998106386, | |
| "grad_norm": 0.7843564748764038, | |
| "learning_rate": 1.402946703621625e-06, | |
| "loss": 0.0716, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.24926837665691168, | |
| "grad_norm": 0.7529016733169556, | |
| "learning_rate": 1.4020578664445222e-06, | |
| "loss": 0.0605, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.24995696333275952, | |
| "grad_norm": 0.7520154118537903, | |
| "learning_rate": 1.4011652619301693e-06, | |
| "loss": 0.0564, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.25064555000860733, | |
| "grad_norm": 3.100435733795166, | |
| "learning_rate": 1.4002688952356887e-06, | |
| "loss": 0.0609, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.25133413668445514, | |
| "grad_norm": 1.572684645652771, | |
| "learning_rate": 1.3993687715399397e-06, | |
| "loss": 0.0741, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.25202272336030296, | |
| "grad_norm": 0.8189212083816528, | |
| "learning_rate": 1.3984648960434868e-06, | |
| "loss": 0.0747, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.2527113100361508, | |
| "grad_norm": 1.6113169193267822, | |
| "learning_rate": 1.3975572739685727e-06, | |
| "loss": 0.0751, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.25339989671199864, | |
| "grad_norm": 0.921216607093811, | |
| "learning_rate": 1.396645910559085e-06, | |
| "loss": 0.0835, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.25408848338784645, | |
| "grad_norm": 0.9955066442489624, | |
| "learning_rate": 1.3957308110805275e-06, | |
| "loss": 0.074, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.25477707006369427, | |
| "grad_norm": 1.3020073175430298, | |
| "learning_rate": 1.3948119808199902e-06, | |
| "loss": 0.0732, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2554656567395421, | |
| "grad_norm": 1.2504678964614868, | |
| "learning_rate": 1.3938894250861172e-06, | |
| "loss": 0.0692, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.2561542434153899, | |
| "grad_norm": 1.1123673915863037, | |
| "learning_rate": 1.3929631492090781e-06, | |
| "loss": 0.0687, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.25684283009123776, | |
| "grad_norm": 0.8279098272323608, | |
| "learning_rate": 1.392033158540535e-06, | |
| "loss": 0.0645, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.2575314167670856, | |
| "grad_norm": 0.8439050912857056, | |
| "learning_rate": 1.3910994584536129e-06, | |
| "loss": 0.075, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.2582200034429334, | |
| "grad_norm": 1.8508895635604858, | |
| "learning_rate": 1.3901620543428686e-06, | |
| "loss": 0.0681, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.2589085901187812, | |
| "grad_norm": 1.7394272089004517, | |
| "learning_rate": 1.389220951624259e-06, | |
| "loss": 0.0675, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.259597176794629, | |
| "grad_norm": 0.7236210703849792, | |
| "learning_rate": 1.3882761557351108e-06, | |
| "loss": 0.0764, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.2602857634704768, | |
| "grad_norm": 1.813814640045166, | |
| "learning_rate": 1.3873276721340877e-06, | |
| "loss": 0.0851, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.2609743501463247, | |
| "grad_norm": 0.5362693667411804, | |
| "learning_rate": 1.3863755063011594e-06, | |
| "loss": 0.0472, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.2616629368221725, | |
| "grad_norm": 1.2949661016464233, | |
| "learning_rate": 1.385419663737571e-06, | |
| "loss": 0.066, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.2623515234980203, | |
| "grad_norm": 2.056469440460205, | |
| "learning_rate": 1.3844601499658096e-06, | |
| "loss": 0.0822, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.26304011017386814, | |
| "grad_norm": 1.49033522605896, | |
| "learning_rate": 1.3834969705295732e-06, | |
| "loss": 0.0709, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.26372869684971595, | |
| "grad_norm": 1.3653734922409058, | |
| "learning_rate": 1.382530130993739e-06, | |
| "loss": 0.0744, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.26441728352556376, | |
| "grad_norm": 0.7207341194152832, | |
| "learning_rate": 1.3815596369443303e-06, | |
| "loss": 0.0747, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.2651058702014116, | |
| "grad_norm": 0.7893703579902649, | |
| "learning_rate": 1.380585493988485e-06, | |
| "loss": 0.065, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.26579445687725944, | |
| "grad_norm": 1.1721290349960327, | |
| "learning_rate": 1.3796077077544234e-06, | |
| "loss": 0.0761, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.26648304355310726, | |
| "grad_norm": 0.7020180225372314, | |
| "learning_rate": 1.378626283891414e-06, | |
| "loss": 0.0536, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.26717163022895507, | |
| "grad_norm": 4.96091365814209, | |
| "learning_rate": 1.377641228069744e-06, | |
| "loss": 0.0705, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.2678602169048029, | |
| "grad_norm": 1.319156527519226, | |
| "learning_rate": 1.3766525459806827e-06, | |
| "loss": 0.0584, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.2685488035806507, | |
| "grad_norm": 0.8134905695915222, | |
| "learning_rate": 1.3756602433364518e-06, | |
| "loss": 0.0583, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.2692373902564985, | |
| "grad_norm": 0.4383983016014099, | |
| "learning_rate": 1.374664325870191e-06, | |
| "loss": 0.0582, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.2699259769323464, | |
| "grad_norm": 1.2263388633728027, | |
| "learning_rate": 1.3736647993359243e-06, | |
| "loss": 0.0686, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.2706145636081942, | |
| "grad_norm": 0.8428287506103516, | |
| "learning_rate": 1.3726616695085285e-06, | |
| "loss": 0.0726, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.271303150284042, | |
| "grad_norm": 1.329553246498108, | |
| "learning_rate": 1.3716549421836981e-06, | |
| "loss": 0.075, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.2719917369598898, | |
| "grad_norm": 0.9827485680580139, | |
| "learning_rate": 1.370644623177913e-06, | |
| "loss": 0.0545, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.27268032363573763, | |
| "grad_norm": 2.0422122478485107, | |
| "learning_rate": 1.3696307183284044e-06, | |
| "loss": 0.0709, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.27336891031158544, | |
| "grad_norm": 0.6993587017059326, | |
| "learning_rate": 1.3686132334931205e-06, | |
| "loss": 0.0611, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.2740574969874333, | |
| "grad_norm": 1.3707586526870728, | |
| "learning_rate": 1.367592174550694e-06, | |
| "loss": 0.0866, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.2747460836632811, | |
| "grad_norm": 0.8292105197906494, | |
| "learning_rate": 1.3665675474004072e-06, | |
| "loss": 0.0684, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.27543467033912894, | |
| "grad_norm": 2.963430404663086, | |
| "learning_rate": 1.3655393579621579e-06, | |
| "loss": 0.097, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.27543467033912894, | |
| "eval_loss": NaN, | |
| "eval_runtime": 65.2621, | |
| "eval_samples_per_second": 14.403, | |
| "eval_steps_per_second": 0.46, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.27612325701497675, | |
| "grad_norm": 2.321105480194092, | |
| "learning_rate": 1.3645076121764256e-06, | |
| "loss": 0.0704, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.27681184369082457, | |
| "grad_norm": 1.274804949760437, | |
| "learning_rate": 1.3634723160042372e-06, | |
| "loss": 0.0586, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.2775004303666724, | |
| "grad_norm": 0.7802310585975647, | |
| "learning_rate": 1.3624334754271319e-06, | |
| "loss": 0.0498, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.27818901704252025, | |
| "grad_norm": 2.0682902336120605, | |
| "learning_rate": 1.3613910964471274e-06, | |
| "loss": 0.0762, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.27887760371836806, | |
| "grad_norm": 1.4176132678985596, | |
| "learning_rate": 1.3603451850866845e-06, | |
| "loss": 0.0893, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.2795661903942159, | |
| "grad_norm": 1.4302617311477661, | |
| "learning_rate": 1.3592957473886735e-06, | |
| "loss": 0.0804, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.2802547770700637, | |
| "grad_norm": 1.5312399864196777, | |
| "learning_rate": 1.3582427894163382e-06, | |
| "loss": 0.0838, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.2809433637459115, | |
| "grad_norm": 0.5655006766319275, | |
| "learning_rate": 1.3571863172532604e-06, | |
| "loss": 0.0618, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.2816319504217593, | |
| "grad_norm": 1.0927455425262451, | |
| "learning_rate": 1.3561263370033267e-06, | |
| "loss": 0.0647, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.2823205370976072, | |
| "grad_norm": 1.2012804746627808, | |
| "learning_rate": 1.3550628547906912e-06, | |
| "loss": 0.0745, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.283009123773455, | |
| "grad_norm": 2.0218346118927, | |
| "learning_rate": 1.353995876759742e-06, | |
| "loss": 0.0837, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.2836977104493028, | |
| "grad_norm": 1.6230448484420776, | |
| "learning_rate": 1.3529254090750632e-06, | |
| "loss": 0.0674, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.2843862971251506, | |
| "grad_norm": 1.415094017982483, | |
| "learning_rate": 1.351851457921402e-06, | |
| "loss": 0.0718, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.28507488380099844, | |
| "grad_norm": 2.0649023056030273, | |
| "learning_rate": 1.3507740295036311e-06, | |
| "loss": 0.0854, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.28576347047684625, | |
| "grad_norm": 1.873093605041504, | |
| "learning_rate": 1.3496931300467137e-06, | |
| "loss": 0.0787, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.2864520571526941, | |
| "grad_norm": 1.908851981163025, | |
| "learning_rate": 1.3486087657956674e-06, | |
| "loss": 0.0767, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.28714064382854193, | |
| "grad_norm": 0.9365402460098267, | |
| "learning_rate": 1.3475209430155279e-06, | |
| "loss": 0.0552, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.28782923050438974, | |
| "grad_norm": 2.2832300662994385, | |
| "learning_rate": 1.3464296679913132e-06, | |
| "loss": 0.0813, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.28851781718023756, | |
| "grad_norm": 1.1070810556411743, | |
| "learning_rate": 1.3453349470279865e-06, | |
| "loss": 0.0713, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.28920640385608537, | |
| "grad_norm": 0.6457180380821228, | |
| "learning_rate": 1.3442367864504208e-06, | |
| "loss": 0.0575, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2898949905319332, | |
| "grad_norm": 1.8618677854537964, | |
| "learning_rate": 1.3431351926033613e-06, | |
| "loss": 0.0701, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.29058357720778105, | |
| "grad_norm": 0.9280813336372375, | |
| "learning_rate": 1.3420301718513902e-06, | |
| "loss": 0.0642, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.29127216388362887, | |
| "grad_norm": 1.831061601638794, | |
| "learning_rate": 1.340921730578888e-06, | |
| "loss": 0.0818, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.2919607505594767, | |
| "grad_norm": 0.8554335236549377, | |
| "learning_rate": 1.3398098751899982e-06, | |
| "loss": 0.0741, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.2926493372353245, | |
| "grad_norm": 1.663798451423645, | |
| "learning_rate": 1.33869461210859e-06, | |
| "loss": 0.0485, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.2933379239111723, | |
| "grad_norm": 1.5381135940551758, | |
| "learning_rate": 1.3375759477782202e-06, | |
| "loss": 0.0725, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.2940265105870201, | |
| "grad_norm": 0.8437361121177673, | |
| "learning_rate": 1.3364538886620971e-06, | |
| "loss": 0.0641, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.294715097262868, | |
| "grad_norm": 0.704498291015625, | |
| "learning_rate": 1.3353284412430433e-06, | |
| "loss": 0.0564, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.2954036839387158, | |
| "grad_norm": 1.3558584451675415, | |
| "learning_rate": 1.3341996120234562e-06, | |
| "loss": 0.0733, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.2960922706145636, | |
| "grad_norm": 1.3674967288970947, | |
| "learning_rate": 1.3330674075252737e-06, | |
| "loss": 0.0696, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.2967808572904114, | |
| "grad_norm": 1.8234331607818604, | |
| "learning_rate": 1.3319318342899334e-06, | |
| "loss": 0.0678, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.29746944396625924, | |
| "grad_norm": 0.9533807039260864, | |
| "learning_rate": 1.330792898878337e-06, | |
| "loss": 0.0743, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.29815803064210705, | |
| "grad_norm": 0.9827238321304321, | |
| "learning_rate": 1.329650607870811e-06, | |
| "loss": 0.0568, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.2988466173179549, | |
| "grad_norm": 1.220139503479004, | |
| "learning_rate": 1.3285049678670695e-06, | |
| "loss": 0.0709, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.29953520399380273, | |
| "grad_norm": 2.355515480041504, | |
| "learning_rate": 1.3273559854861758e-06, | |
| "loss": 0.0801, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.30022379066965055, | |
| "grad_norm": 0.8433701992034912, | |
| "learning_rate": 1.326203667366504e-06, | |
| "loss": 0.063, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.30091237734549836, | |
| "grad_norm": 1.749002456665039, | |
| "learning_rate": 1.3250480201657012e-06, | |
| "loss": 0.0827, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.3016009640213462, | |
| "grad_norm": 1.7491220235824585, | |
| "learning_rate": 1.323889050560648e-06, | |
| "loss": 0.0816, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.302289550697194, | |
| "grad_norm": 0.9220266342163086, | |
| "learning_rate": 1.322726765247421e-06, | |
| "loss": 0.0914, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.30297813737304186, | |
| "grad_norm": 2.6664085388183594, | |
| "learning_rate": 1.3215611709412537e-06, | |
| "loss": 0.0762, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.30366672404888967, | |
| "grad_norm": 0.7559336423873901, | |
| "learning_rate": 1.3203922743764975e-06, | |
| "loss": 0.0496, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.3043553107247375, | |
| "grad_norm": 0.9362991452217102, | |
| "learning_rate": 1.3192200823065832e-06, | |
| "loss": 0.08, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.3050438974005853, | |
| "grad_norm": 0.9287563562393188, | |
| "learning_rate": 1.3180446015039812e-06, | |
| "loss": 0.0553, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.3057324840764331, | |
| "grad_norm": 0.662253737449646, | |
| "learning_rate": 1.3168658387601636e-06, | |
| "loss": 0.0545, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.3064210707522809, | |
| "grad_norm": 1.268722414970398, | |
| "learning_rate": 1.315683800885564e-06, | |
| "loss": 0.0644, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.3071096574281288, | |
| "grad_norm": 0.9683335423469543, | |
| "learning_rate": 1.3144984947095382e-06, | |
| "loss": 0.0704, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.3077982441039766, | |
| "grad_norm": 1.3514573574066162, | |
| "learning_rate": 1.3133099270803253e-06, | |
| "loss": 0.0644, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.3084868307798244, | |
| "grad_norm": 1.2376309633255005, | |
| "learning_rate": 1.312118104865008e-06, | |
| "loss": 0.0891, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.30917541745567223, | |
| "grad_norm": 0.9724456667900085, | |
| "learning_rate": 1.310923034949472e-06, | |
| "loss": 0.0695, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.30986400413152004, | |
| "grad_norm": 1.492810606956482, | |
| "learning_rate": 1.3097247242383676e-06, | |
| "loss": 0.0786, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.31055259080736786, | |
| "grad_norm": 0.6909795999526978, | |
| "learning_rate": 1.3085231796550688e-06, | |
| "loss": 0.0663, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.3112411774832157, | |
| "grad_norm": 0.9493128061294556, | |
| "learning_rate": 1.3073184081416343e-06, | |
| "loss": 0.0679, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.31192976415906354, | |
| "grad_norm": 1.309522032737732, | |
| "learning_rate": 1.3061104166587656e-06, | |
| "loss": 0.0655, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.31261835083491135, | |
| "grad_norm": 1.089707612991333, | |
| "learning_rate": 1.304899212185769e-06, | |
| "loss": 0.0845, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.31330693751075916, | |
| "grad_norm": 1.0757405757904053, | |
| "learning_rate": 1.3036848017205135e-06, | |
| "loss": 0.0692, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.313995524186607, | |
| "grad_norm": 0.884185791015625, | |
| "learning_rate": 1.302467192279392e-06, | |
| "loss": 0.0711, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.3146841108624548, | |
| "grad_norm": 1.2772009372711182, | |
| "learning_rate": 1.3012463908972784e-06, | |
| "loss": 0.0659, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.31537269753830266, | |
| "grad_norm": 1.3834528923034668, | |
| "learning_rate": 1.3000224046274898e-06, | |
| "loss": 0.0681, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.3160612842141505, | |
| "grad_norm": 0.7710753083229065, | |
| "learning_rate": 1.2987952405417437e-06, | |
| "loss": 0.0564, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.3167498708899983, | |
| "grad_norm": 0.7007287740707397, | |
| "learning_rate": 1.2975649057301176e-06, | |
| "loss": 0.0683, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.3174384575658461, | |
| "grad_norm": 1.799493432044983, | |
| "learning_rate": 1.2963314073010087e-06, | |
| "loss": 0.0725, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.3181270442416939, | |
| "grad_norm": 2.191889524459839, | |
| "learning_rate": 1.295094752381092e-06, | |
| "loss": 0.0761, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.3188156309175417, | |
| "grad_norm": 0.775509238243103, | |
| "learning_rate": 1.2938549481152796e-06, | |
| "loss": 0.0577, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.3195042175933896, | |
| "grad_norm": 1.184441089630127, | |
| "learning_rate": 1.2926120016666797e-06, | |
| "loss": 0.082, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.3201928042692374, | |
| "grad_norm": 0.9733319282531738, | |
| "learning_rate": 1.2913659202165538e-06, | |
| "loss": 0.0614, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.3208813909450852, | |
| "grad_norm": 1.4761791229248047, | |
| "learning_rate": 1.2901167109642771e-06, | |
| "loss": 0.0564, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.32156997762093303, | |
| "grad_norm": 1.2912704944610596, | |
| "learning_rate": 1.288864381127296e-06, | |
| "loss": 0.0908, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.32225856429678085, | |
| "grad_norm": 0.9779075384140015, | |
| "learning_rate": 1.2876089379410857e-06, | |
| "loss": 0.0628, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.32294715097262866, | |
| "grad_norm": 1.5734461545944214, | |
| "learning_rate": 1.2863503886591101e-06, | |
| "loss": 0.0938, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.32363573764847653, | |
| "grad_norm": 1.2200082540512085, | |
| "learning_rate": 1.2850887405527784e-06, | |
| "loss": 0.0823, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.32432432432432434, | |
| "grad_norm": 1.3125114440917969, | |
| "learning_rate": 1.2838240009114031e-06, | |
| "loss": 0.0681, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.32501291100017216, | |
| "grad_norm": 0.6019772887229919, | |
| "learning_rate": 1.2825561770421595e-06, | |
| "loss": 0.0736, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.32570149767601997, | |
| "grad_norm": 0.5353420376777649, | |
| "learning_rate": 1.2812852762700416e-06, | |
| "loss": 0.0483, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.3263900843518678, | |
| "grad_norm": 0.9439918398857117, | |
| "learning_rate": 1.280011305937821e-06, | |
| "loss": 0.0505, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.3270786710277156, | |
| "grad_norm": 1.1888117790222168, | |
| "learning_rate": 1.278734273406003e-06, | |
| "loss": 0.0717, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.32776725770356346, | |
| "grad_norm": 1.5723661184310913, | |
| "learning_rate": 1.2774541860527868e-06, | |
| "loss": 0.0747, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.3284558443794113, | |
| "grad_norm": 0.9167183041572571, | |
| "learning_rate": 1.2761710512740197e-06, | |
| "loss": 0.0693, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.3291444310552591, | |
| "grad_norm": 1.7774821519851685, | |
| "learning_rate": 1.2748848764831567e-06, | |
| "loss": 0.0819, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.3298330177311069, | |
| "grad_norm": 0.7332593202590942, | |
| "learning_rate": 1.2735956691112163e-06, | |
| "loss": 0.0636, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.3305216044069547, | |
| "grad_norm": 2.0606303215026855, | |
| "learning_rate": 1.2723034366067382e-06, | |
| "loss": 0.0836, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.33121019108280253, | |
| "grad_norm": 0.8616734147071838, | |
| "learning_rate": 1.2710081864357405e-06, | |
| "loss": 0.0615, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.33189877775865034, | |
| "grad_norm": 1.0574530363082886, | |
| "learning_rate": 1.2697099260816757e-06, | |
| "loss": 0.0824, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.3325873644344982, | |
| "grad_norm": 0.8800958395004272, | |
| "learning_rate": 1.2684086630453882e-06, | |
| "loss": 0.0721, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.333275951110346, | |
| "grad_norm": 0.660064160823822, | |
| "learning_rate": 1.2671044048450708e-06, | |
| "loss": 0.068, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.33396453778619384, | |
| "grad_norm": 0.7648721933364868, | |
| "learning_rate": 1.265797159016221e-06, | |
| "loss": 0.0824, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.33465312446204165, | |
| "grad_norm": 0.845352053642273, | |
| "learning_rate": 1.2644869331115981e-06, | |
| "loss": 0.0705, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.33534171113788946, | |
| "grad_norm": 1.0151258707046509, | |
| "learning_rate": 1.2631737347011786e-06, | |
| "loss": 0.0757, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.3360302978137373, | |
| "grad_norm": 0.7538004517555237, | |
| "learning_rate": 1.2618575713721133e-06, | |
| "loss": 0.0625, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.33671888448958515, | |
| "grad_norm": 1.7425694465637207, | |
| "learning_rate": 1.2605384507286825e-06, | |
| "loss": 0.0983, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.33740747116543296, | |
| "grad_norm": 0.5900651812553406, | |
| "learning_rate": 1.2592163803922538e-06, | |
| "loss": 0.0633, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.3380960578412808, | |
| "grad_norm": 1.029417634010315, | |
| "learning_rate": 1.2578913680012365e-06, | |
| "loss": 0.056, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.3387846445171286, | |
| "grad_norm": 0.6467821002006531, | |
| "learning_rate": 1.2565634212110376e-06, | |
| "loss": 0.0664, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.3394732311929764, | |
| "grad_norm": 1.9054161310195923, | |
| "learning_rate": 1.2552325476940182e-06, | |
| "loss": 0.0734, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.3401618178688242, | |
| "grad_norm": 1.4536879062652588, | |
| "learning_rate": 1.253898755139449e-06, | |
| "loss": 0.0764, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.3408504045446721, | |
| "grad_norm": 0.7609797120094299, | |
| "learning_rate": 1.2525620512534655e-06, | |
| "loss": 0.0708, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.3415389912205199, | |
| "grad_norm": 1.7982048988342285, | |
| "learning_rate": 1.251222443759024e-06, | |
| "loss": 0.0614, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.3422275778963677, | |
| "grad_norm": 1.6682957410812378, | |
| "learning_rate": 1.2498799403958568e-06, | |
| "loss": 0.0714, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.3429161645722155, | |
| "grad_norm": 0.6978663206100464, | |
| "learning_rate": 1.248534548920427e-06, | |
| "loss": 0.0617, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.34360475124806333, | |
| "grad_norm": 0.776599109172821, | |
| "learning_rate": 1.2471862771058844e-06, | |
| "loss": 0.0612, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.34429333792391115, | |
| "grad_norm": 0.7338681221008301, | |
| "learning_rate": 1.24583513274202e-06, | |
| "loss": 0.0735, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.344981924599759, | |
| "grad_norm": 1.1987628936767578, | |
| "learning_rate": 1.244481123635222e-06, | |
| "loss": 0.0609, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.34567051127560683, | |
| "grad_norm": 1.0987433195114136, | |
| "learning_rate": 1.2431242576084287e-06, | |
| "loss": 0.0653, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.34635909795145464, | |
| "grad_norm": 2.0642261505126953, | |
| "learning_rate": 1.241764542501086e-06, | |
| "loss": 0.0889, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.34704768462730246, | |
| "grad_norm": 0.829462468624115, | |
| "learning_rate": 1.2404019861690996e-06, | |
| "loss": 0.081, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.34773627130315027, | |
| "grad_norm": 1.9418965578079224, | |
| "learning_rate": 1.2390365964847914e-06, | |
| "loss": 0.0696, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.3484248579789981, | |
| "grad_norm": 1.235224962234497, | |
| "learning_rate": 1.2376683813368525e-06, | |
| "loss": 0.0548, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.34911344465484595, | |
| "grad_norm": 1.5201059579849243, | |
| "learning_rate": 1.2362973486302997e-06, | |
| "loss": 0.0656, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.34980203133069376, | |
| "grad_norm": 0.5023059844970703, | |
| "learning_rate": 1.2349235062864271e-06, | |
| "loss": 0.0628, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.3504906180065416, | |
| "grad_norm": 0.9694226980209351, | |
| "learning_rate": 1.233546862242763e-06, | |
| "loss": 0.0776, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.3511792046823894, | |
| "grad_norm": 0.9070531129837036, | |
| "learning_rate": 1.2321674244530216e-06, | |
| "loss": 0.0571, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.3518677913582372, | |
| "grad_norm": 0.7685413360595703, | |
| "learning_rate": 1.23078520088706e-06, | |
| "loss": 0.0692, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.352556378034085, | |
| "grad_norm": 0.6996450424194336, | |
| "learning_rate": 1.2294001995308288e-06, | |
| "loss": 0.0695, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.3532449647099329, | |
| "grad_norm": 0.9781703948974609, | |
| "learning_rate": 1.2280124283863285e-06, | |
| "loss": 0.065, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.3539335513857807, | |
| "grad_norm": 0.7418661117553711, | |
| "learning_rate": 1.2266218954715614e-06, | |
| "loss": 0.0554, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.3546221380616285, | |
| "grad_norm": 1.1466076374053955, | |
| "learning_rate": 1.2252286088204882e-06, | |
| "loss": 0.0495, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.3553107247374763, | |
| "grad_norm": 0.8437464833259583, | |
| "learning_rate": 1.2238325764829774e-06, | |
| "loss": 0.0632, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.35599931141332414, | |
| "grad_norm": 1.469531774520874, | |
| "learning_rate": 1.2224338065247623e-06, | |
| "loss": 0.0664, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.35668789808917195, | |
| "grad_norm": 1.5107131004333496, | |
| "learning_rate": 1.2210323070273924e-06, | |
| "loss": 0.0741, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.3573764847650198, | |
| "grad_norm": 1.238404631614685, | |
| "learning_rate": 1.2196280860881882e-06, | |
| "loss": 0.0561, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.35806507144086763, | |
| "grad_norm": 1.3876500129699707, | |
| "learning_rate": 1.2182211518201932e-06, | |
| "loss": 0.0566, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.35875365811671545, | |
| "grad_norm": 1.5162941217422485, | |
| "learning_rate": 1.2168115123521275e-06, | |
| "loss": 0.0882, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.35944224479256326, | |
| "grad_norm": 0.6646033525466919, | |
| "learning_rate": 1.2153991758283405e-06, | |
| "loss": 0.062, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.3601308314684111, | |
| "grad_norm": 1.850542664527893, | |
| "learning_rate": 1.2139841504087648e-06, | |
| "loss": 0.06, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.3608194181442589, | |
| "grad_norm": 1.7398889064788818, | |
| "learning_rate": 1.2125664442688676e-06, | |
| "loss": 0.0666, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.36150800482010675, | |
| "grad_norm": 0.640017569065094, | |
| "learning_rate": 1.2111460655996052e-06, | |
| "loss": 0.054, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.36219659149595457, | |
| "grad_norm": 1.4830689430236816, | |
| "learning_rate": 1.2097230226073736e-06, | |
| "loss": 0.0703, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.3628851781718024, | |
| "grad_norm": 0.9098523855209351, | |
| "learning_rate": 1.2082973235139631e-06, | |
| "loss": 0.0564, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.3635737648476502, | |
| "grad_norm": 0.8333926200866699, | |
| "learning_rate": 1.20686897655651e-06, | |
| "loss": 0.0583, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.364262351523498, | |
| "grad_norm": 0.6616759300231934, | |
| "learning_rate": 1.2054379899874481e-06, | |
| "loss": 0.0561, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.3649509381993458, | |
| "grad_norm": 1.463626503944397, | |
| "learning_rate": 1.2040043720744629e-06, | |
| "loss": 0.0673, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.3656395248751937, | |
| "grad_norm": 1.154791235923767, | |
| "learning_rate": 1.2025681311004414e-06, | |
| "loss": 0.072, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.3663281115510415, | |
| "grad_norm": 1.2037138938903809, | |
| "learning_rate": 1.2011292753634273e-06, | |
| "loss": 0.0793, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.3670166982268893, | |
| "grad_norm": 1.1648156642913818, | |
| "learning_rate": 1.19968781317657e-06, | |
| "loss": 0.0866, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.36770528490273713, | |
| "grad_norm": 0.7248892188072205, | |
| "learning_rate": 1.198243752868079e-06, | |
| "loss": 0.0684, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.36839387157858494, | |
| "grad_norm": 1.5188297033309937, | |
| "learning_rate": 1.1967971027811736e-06, | |
| "loss": 0.0676, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.36908245825443275, | |
| "grad_norm": 0.8715887665748596, | |
| "learning_rate": 1.1953478712740369e-06, | |
| "loss": 0.0735, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.3697710449302806, | |
| "grad_norm": 5.552928447723389, | |
| "learning_rate": 1.1938960667197659e-06, | |
| "loss": 0.0746, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.37045963160612844, | |
| "grad_norm": 0.7330494523048401, | |
| "learning_rate": 1.192441697506324e-06, | |
| "loss": 0.0527, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.37114821828197625, | |
| "grad_norm": 0.8480969667434692, | |
| "learning_rate": 1.1909847720364916e-06, | |
| "loss": 0.0401, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.37183680495782406, | |
| "grad_norm": 1.340770959854126, | |
| "learning_rate": 1.1895252987278186e-06, | |
| "loss": 0.0663, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.3725253916336719, | |
| "grad_norm": 0.6979343891143799, | |
| "learning_rate": 1.1880632860125758e-06, | |
| "loss": 0.0575, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.3732139783095197, | |
| "grad_norm": 1.3765246868133545, | |
| "learning_rate": 1.1865987423377045e-06, | |
| "loss": 0.0536, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.37390256498536756, | |
| "grad_norm": 1.0142332315444946, | |
| "learning_rate": 1.18513167616477e-06, | |
| "loss": 0.0584, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.37459115166121537, | |
| "grad_norm": 1.2334593534469604, | |
| "learning_rate": 1.1836620959699108e-06, | |
| "loss": 0.0715, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.3752797383370632, | |
| "grad_norm": 0.9497774839401245, | |
| "learning_rate": 1.1821900102437912e-06, | |
| "loss": 0.0782, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.375968325012911, | |
| "grad_norm": 0.9996221661567688, | |
| "learning_rate": 1.1807154274915511e-06, | |
| "loss": 0.083, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.3766569116887588, | |
| "grad_norm": 1.4065628051757812, | |
| "learning_rate": 1.179238356232757e-06, | |
| "loss": 0.0629, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.3773454983646066, | |
| "grad_norm": 0.8860158324241638, | |
| "learning_rate": 1.1777588050013535e-06, | |
| "loss": 0.0584, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.3780340850404545, | |
| "grad_norm": 1.6017625331878662, | |
| "learning_rate": 1.1762767823456133e-06, | |
| "loss": 0.0827, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.3787226717163023, | |
| "grad_norm": 0.788905918598175, | |
| "learning_rate": 1.1747922968280882e-06, | |
| "loss": 0.0598, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3794112583921501, | |
| "grad_norm": 0.8911811113357544, | |
| "learning_rate": 1.1733053570255591e-06, | |
| "loss": 0.0564, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.38009984506799793, | |
| "grad_norm": 3.0224835872650146, | |
| "learning_rate": 1.1718159715289867e-06, | |
| "loss": 0.0688, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.38078843174384575, | |
| "grad_norm": 0.5552683472633362, | |
| "learning_rate": 1.170324148943463e-06, | |
| "loss": 0.0564, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.38147701841969356, | |
| "grad_norm": 0.5574796795845032, | |
| "learning_rate": 1.168829897888159e-06, | |
| "loss": 0.0417, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.3821656050955414, | |
| "grad_norm": 0.82485032081604, | |
| "learning_rate": 1.167333226996278e-06, | |
| "loss": 0.0697, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.38285419177138924, | |
| "grad_norm": 0.5813287496566772, | |
| "learning_rate": 1.1658341449150032e-06, | |
| "loss": 0.0627, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.38354277844723705, | |
| "grad_norm": 1.1604082584381104, | |
| "learning_rate": 1.164332660305449e-06, | |
| "loss": 0.0748, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.38423136512308487, | |
| "grad_norm": 0.5904826521873474, | |
| "learning_rate": 1.162828781842611e-06, | |
| "loss": 0.0597, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.3849199517989327, | |
| "grad_norm": 0.5884749293327332, | |
| "learning_rate": 1.1613225182153145e-06, | |
| "loss": 0.058, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.3856085384747805, | |
| "grad_norm": 1.0348842144012451, | |
| "learning_rate": 1.1598138781261665e-06, | |
| "loss": 0.0964, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.38629712515062836, | |
| "grad_norm": 3.1019513607025146, | |
| "learning_rate": 1.1583028702915034e-06, | |
| "loss": 0.0874, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.3869857118264762, | |
| "grad_norm": 0.8248711228370667, | |
| "learning_rate": 1.156789503441343e-06, | |
| "loss": 0.0694, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.387674298502324, | |
| "grad_norm": 0.7153844237327576, | |
| "learning_rate": 1.1552737863193303e-06, | |
| "loss": 0.0602, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.3883628851781718, | |
| "grad_norm": 0.9267993569374084, | |
| "learning_rate": 1.1537557276826913e-06, | |
| "loss": 0.0758, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.3890514718540196, | |
| "grad_norm": 0.640562117099762, | |
| "learning_rate": 1.1522353363021794e-06, | |
| "loss": 0.0616, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.38974005852986743, | |
| "grad_norm": 1.1064201593399048, | |
| "learning_rate": 1.1507126209620256e-06, | |
| "loss": 0.0741, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.3904286452057153, | |
| "grad_norm": 2.026252269744873, | |
| "learning_rate": 1.149187590459888e-06, | |
| "loss": 0.0772, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.3911172318815631, | |
| "grad_norm": 0.6178696155548096, | |
| "learning_rate": 1.1476602536068014e-06, | |
| "loss": 0.0811, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.3918058185574109, | |
| "grad_norm": 0.9859865307807922, | |
| "learning_rate": 1.1461306192271252e-06, | |
| "loss": 0.0656, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.39249440523325874, | |
| "grad_norm": 1.1877422332763672, | |
| "learning_rate": 1.1445986961584925e-06, | |
| "loss": 0.0576, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.39318299190910655, | |
| "grad_norm": 0.9550566673278809, | |
| "learning_rate": 1.1430644932517608e-06, | |
| "loss": 0.0694, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.39387157858495436, | |
| "grad_norm": 0.9525591135025024, | |
| "learning_rate": 1.1415280193709588e-06, | |
| "loss": 0.0852, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.3945601652608022, | |
| "grad_norm": 0.8233121037483215, | |
| "learning_rate": 1.139989283393236e-06, | |
| "loss": 0.0635, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.39524875193665004, | |
| "grad_norm": 1.1796543598175049, | |
| "learning_rate": 1.1384482942088117e-06, | |
| "loss": 0.0659, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.39593733861249786, | |
| "grad_norm": 1.0039130449295044, | |
| "learning_rate": 1.1369050607209233e-06, | |
| "loss": 0.0714, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.39662592528834567, | |
| "grad_norm": 0.9246156215667725, | |
| "learning_rate": 1.1353595918457751e-06, | |
| "loss": 0.0654, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.3973145119641935, | |
| "grad_norm": 6.412201881408691, | |
| "learning_rate": 1.1338118965124859e-06, | |
| "loss": 0.0623, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.3980030986400413, | |
| "grad_norm": 2.1875622272491455, | |
| "learning_rate": 1.1322619836630388e-06, | |
| "loss": 0.0705, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.3986916853158891, | |
| "grad_norm": 1.5282702445983887, | |
| "learning_rate": 1.1307098622522289e-06, | |
| "loss": 0.0671, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.399380271991737, | |
| "grad_norm": 0.506668210029602, | |
| "learning_rate": 1.1291555412476107e-06, | |
| "loss": 0.0527, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.4000688586675848, | |
| "grad_norm": 1.0984927415847778, | |
| "learning_rate": 1.127599029629448e-06, | |
| "loss": 0.1052, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.4007574453434326, | |
| "grad_norm": 0.995433509349823, | |
| "learning_rate": 1.1260403363906608e-06, | |
| "loss": 0.0695, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.4014460320192804, | |
| "grad_norm": 2.201193332672119, | |
| "learning_rate": 1.1244794705367735e-06, | |
| "loss": 0.0766, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.40213461869512823, | |
| "grad_norm": 1.5904078483581543, | |
| "learning_rate": 1.1229164410858632e-06, | |
| "loss": 0.0611, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.40282320537097605, | |
| "grad_norm": 1.0457239151000977, | |
| "learning_rate": 1.121351257068507e-06, | |
| "loss": 0.0672, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.4035117920468239, | |
| "grad_norm": 0.9336317181587219, | |
| "learning_rate": 1.1197839275277313e-06, | |
| "loss": 0.083, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.4042003787226717, | |
| "grad_norm": 0.9990373849868774, | |
| "learning_rate": 1.1182144615189574e-06, | |
| "loss": 0.0567, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.40488896539851954, | |
| "grad_norm": 1.1609365940093994, | |
| "learning_rate": 1.1166428681099505e-06, | |
| "loss": 0.0664, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.40557755207436735, | |
| "grad_norm": 0.8399966955184937, | |
| "learning_rate": 1.1150691563807676e-06, | |
| "loss": 0.0858, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.40626613875021517, | |
| "grad_norm": 0.7864644527435303, | |
| "learning_rate": 1.1134933354237042e-06, | |
| "loss": 0.0602, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.406954725426063, | |
| "grad_norm": 1.6738113164901733, | |
| "learning_rate": 1.111915414343242e-06, | |
| "loss": 0.0591, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.40764331210191085, | |
| "grad_norm": 1.3056893348693848, | |
| "learning_rate": 1.1103354022559963e-06, | |
| "loss": 0.0556, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.40833189877775866, | |
| "grad_norm": 0.8307862877845764, | |
| "learning_rate": 1.1087533082906643e-06, | |
| "loss": 0.0783, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.4090204854536065, | |
| "grad_norm": 1.2549020051956177, | |
| "learning_rate": 1.1071691415879704e-06, | |
| "loss": 0.0781, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.4097090721294543, | |
| "grad_norm": 0.7206241488456726, | |
| "learning_rate": 1.105582911300615e-06, | |
| "loss": 0.0638, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.4103976588053021, | |
| "grad_norm": 0.9231994152069092, | |
| "learning_rate": 1.1039946265932207e-06, | |
| "loss": 0.0751, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.4110862454811499, | |
| "grad_norm": 0.8608548045158386, | |
| "learning_rate": 1.1024042966422805e-06, | |
| "loss": 0.0662, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.4117748321569978, | |
| "grad_norm": 0.8011870980262756, | |
| "learning_rate": 1.100811930636103e-06, | |
| "loss": 0.0499, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.4124634188328456, | |
| "grad_norm": 0.6414494514465332, | |
| "learning_rate": 1.0992175377747615e-06, | |
| "loss": 0.0679, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.4131520055086934, | |
| "grad_norm": 1.301715612411499, | |
| "learning_rate": 1.0976211272700387e-06, | |
| "loss": 0.0764, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4131520055086934, | |
| "eval_loss": NaN, | |
| "eval_runtime": 70.502, | |
| "eval_samples_per_second": 13.333, | |
| "eval_steps_per_second": 0.426, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4138405921845412, | |
| "grad_norm": 0.7710323333740234, | |
| "learning_rate": 1.096022708345375e-06, | |
| "loss": 0.0773, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.41452917886038904, | |
| "grad_norm": 0.6463258266448975, | |
| "learning_rate": 1.0944222902358142e-06, | |
| "loss": 0.0518, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.41521776553623685, | |
| "grad_norm": 0.764061689376831, | |
| "learning_rate": 1.0928198821879514e-06, | |
| "loss": 0.06, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.4159063522120847, | |
| "grad_norm": 1.3276740312576294, | |
| "learning_rate": 1.0912154934598781e-06, | |
| "loss": 0.0647, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.41659493888793253, | |
| "grad_norm": 1.1279891729354858, | |
| "learning_rate": 1.0896091333211294e-06, | |
| "loss": 0.0799, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.41728352556378034, | |
| "grad_norm": 1.2612184286117554, | |
| "learning_rate": 1.0880008110526307e-06, | |
| "loss": 0.0848, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.41797211223962816, | |
| "grad_norm": 0.5697112679481506, | |
| "learning_rate": 1.086390535946644e-06, | |
| "loss": 0.0583, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.41866069891547597, | |
| "grad_norm": 0.8816370964050293, | |
| "learning_rate": 1.084778317306714e-06, | |
| "loss": 0.0593, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.4193492855913238, | |
| "grad_norm": 1.256788730621338, | |
| "learning_rate": 1.0831641644476136e-06, | |
| "loss": 0.068, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.42003787226717165, | |
| "grad_norm": 1.2928075790405273, | |
| "learning_rate": 1.0815480866952917e-06, | |
| "loss": 0.0708, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.42072645894301947, | |
| "grad_norm": 1.6780614852905273, | |
| "learning_rate": 1.0799300933868188e-06, | |
| "loss": 0.0655, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.4214150456188673, | |
| "grad_norm": 1.049795150756836, | |
| "learning_rate": 1.0783101938703316e-06, | |
| "loss": 0.092, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.4221036322947151, | |
| "grad_norm": 1.0051531791687012, | |
| "learning_rate": 1.0766883975049809e-06, | |
| "loss": 0.0661, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.4227922189705629, | |
| "grad_norm": 0.9689170122146606, | |
| "learning_rate": 1.0750647136608764e-06, | |
| "loss": 0.0771, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.4234808056464107, | |
| "grad_norm": 1.548276662826538, | |
| "learning_rate": 1.0734391517190334e-06, | |
| "loss": 0.0897, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.4241693923222586, | |
| "grad_norm": 1.1861604452133179, | |
| "learning_rate": 1.0718117210713177e-06, | |
| "loss": 0.0751, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.4248579789981064, | |
| "grad_norm": 1.4423487186431885, | |
| "learning_rate": 1.070182431120392e-06, | |
| "loss": 0.0897, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.4255465656739542, | |
| "grad_norm": 0.8759933114051819, | |
| "learning_rate": 1.0685512912796607e-06, | |
| "loss": 0.0665, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.426235152349802, | |
| "grad_norm": 0.7939784526824951, | |
| "learning_rate": 1.0669183109732174e-06, | |
| "loss": 0.0699, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.42692373902564984, | |
| "grad_norm": 0.6089954972267151, | |
| "learning_rate": 1.065283499635788e-06, | |
| "loss": 0.0559, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.42761232570149765, | |
| "grad_norm": 1.0511106252670288, | |
| "learning_rate": 1.0636468667126779e-06, | |
| "loss": 0.0524, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.4283009123773455, | |
| "grad_norm": 1.1751030683517456, | |
| "learning_rate": 1.0620084216597168e-06, | |
| "loss": 0.0624, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.42898949905319333, | |
| "grad_norm": 1.3557554483413696, | |
| "learning_rate": 1.0603681739432045e-06, | |
| "loss": 0.0723, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.42967808572904115, | |
| "grad_norm": 1.5584670305252075, | |
| "learning_rate": 1.0587261330398554e-06, | |
| "loss": 0.1011, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.43036667240488896, | |
| "grad_norm": 0.8870041966438293, | |
| "learning_rate": 1.0570823084367446e-06, | |
| "loss": 0.0494, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.4310552590807368, | |
| "grad_norm": 1.565441370010376, | |
| "learning_rate": 1.0554367096312525e-06, | |
| "loss": 0.0822, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.4317438457565846, | |
| "grad_norm": 0.6839600801467896, | |
| "learning_rate": 1.0537893461310102e-06, | |
| "loss": 0.057, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.43243243243243246, | |
| "grad_norm": 0.534784734249115, | |
| "learning_rate": 1.0521402274538447e-06, | |
| "loss": 0.0607, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.43312101910828027, | |
| "grad_norm": 0.9460827708244324, | |
| "learning_rate": 1.0504893631277235e-06, | |
| "loss": 0.0721, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.4338096057841281, | |
| "grad_norm": 0.9717746376991272, | |
| "learning_rate": 1.0488367626906999e-06, | |
| "loss": 0.0765, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.4344981924599759, | |
| "grad_norm": 0.7943449020385742, | |
| "learning_rate": 1.0471824356908577e-06, | |
| "loss": 0.0629, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.4351867791358237, | |
| "grad_norm": 0.9659263491630554, | |
| "learning_rate": 1.0455263916862563e-06, | |
| "loss": 0.0654, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.4358753658116715, | |
| "grad_norm": 1.4025851488113403, | |
| "learning_rate": 1.043868640244875e-06, | |
| "loss": 0.0732, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.4365639524875194, | |
| "grad_norm": 1.0346896648406982, | |
| "learning_rate": 1.042209190944558e-06, | |
| "loss": 0.0661, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.4372525391633672, | |
| "grad_norm": 1.959477424621582, | |
| "learning_rate": 1.04054805337296e-06, | |
| "loss": 0.064, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.437941125839215, | |
| "grad_norm": 0.6854409575462341, | |
| "learning_rate": 1.0388852371274878e-06, | |
| "loss": 0.0632, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.43862971251506283, | |
| "grad_norm": 0.9316849112510681, | |
| "learning_rate": 1.0372207518152497e-06, | |
| "loss": 0.0896, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.43931829919091064, | |
| "grad_norm": 1.10444974899292, | |
| "learning_rate": 1.0355546070529947e-06, | |
| "loss": 0.0507, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.44000688586675846, | |
| "grad_norm": 1.0166298151016235, | |
| "learning_rate": 1.033886812467061e-06, | |
| "loss": 0.0745, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.4406954725426063, | |
| "grad_norm": 1.988134741783142, | |
| "learning_rate": 1.032217377693318e-06, | |
| "loss": 0.0907, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.44138405921845414, | |
| "grad_norm": 1.0151528120040894, | |
| "learning_rate": 1.030546312377112e-06, | |
| "loss": 0.051, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.44207264589430195, | |
| "grad_norm": 0.4793682098388672, | |
| "learning_rate": 1.02887362617321e-06, | |
| "loss": 0.0503, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.44276123257014977, | |
| "grad_norm": 1.3316080570220947, | |
| "learning_rate": 1.027199328745743e-06, | |
| "loss": 0.0858, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.4434498192459976, | |
| "grad_norm": 1.0065698623657227, | |
| "learning_rate": 1.0255234297681523e-06, | |
| "loss": 0.0767, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.4441384059218454, | |
| "grad_norm": 1.0908037424087524, | |
| "learning_rate": 1.0238459389231315e-06, | |
| "loss": 0.0583, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.44482699259769326, | |
| "grad_norm": 1.080434799194336, | |
| "learning_rate": 1.0221668659025711e-06, | |
| "loss": 0.0621, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.4455155792735411, | |
| "grad_norm": 0.57427978515625, | |
| "learning_rate": 1.0204862204075037e-06, | |
| "loss": 0.0483, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.4462041659493889, | |
| "grad_norm": 1.8378747701644897, | |
| "learning_rate": 1.0188040121480466e-06, | |
| "loss": 0.0809, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.4468927526252367, | |
| "grad_norm": 0.7844899892807007, | |
| "learning_rate": 1.0171202508433462e-06, | |
| "loss": 0.0575, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.4475813393010845, | |
| "grad_norm": 0.9027393460273743, | |
| "learning_rate": 1.0154349462215217e-06, | |
| "loss": 0.0793, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.4482699259769323, | |
| "grad_norm": 2.9562268257141113, | |
| "learning_rate": 1.0137481080196088e-06, | |
| "loss": 0.0724, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.4489585126527802, | |
| "grad_norm": 2.5500500202178955, | |
| "learning_rate": 1.0120597459835041e-06, | |
| "loss": 0.054, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.449647099328628, | |
| "grad_norm": 1.079923152923584, | |
| "learning_rate": 1.010369869867908e-06, | |
| "loss": 0.057, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.4503356860044758, | |
| "grad_norm": 1.3031681776046753, | |
| "learning_rate": 1.008678489436269e-06, | |
| "loss": 0.0626, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.45102427268032363, | |
| "grad_norm": 0.8103233575820923, | |
| "learning_rate": 1.0069856144607262e-06, | |
| "loss": 0.059, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.45171285935617145, | |
| "grad_norm": 1.1583657264709473, | |
| "learning_rate": 1.0052912547220545e-06, | |
| "loss": 0.0519, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.45240144603201926, | |
| "grad_norm": 1.240957260131836, | |
| "learning_rate": 1.003595420009607e-06, | |
| "loss": 0.0921, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.45309003270786713, | |
| "grad_norm": 1.5437582731246948, | |
| "learning_rate": 1.001898120121258e-06, | |
| "loss": 0.0828, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.45377861938371494, | |
| "grad_norm": 0.7383193373680115, | |
| "learning_rate": 1.0001993648633474e-06, | |
| "loss": 0.0703, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.45446720605956276, | |
| "grad_norm": 0.9811745285987854, | |
| "learning_rate": 9.98499164050624e-07, | |
| "loss": 0.0456, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.45515579273541057, | |
| "grad_norm": 2.0876340866088867, | |
| "learning_rate": 9.967975275061877e-07, | |
| "loss": 0.0635, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.4558443794112584, | |
| "grad_norm": 0.8804057240486145, | |
| "learning_rate": 9.950944650614346e-07, | |
| "loss": 0.0487, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.4565329660871062, | |
| "grad_norm": 1.2256144285202026, | |
| "learning_rate": 9.933899865559977e-07, | |
| "loss": 0.0693, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.45722155276295406, | |
| "grad_norm": 0.8635557293891907, | |
| "learning_rate": 9.91684101837693e-07, | |
| "loss": 0.0515, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.4579101394388019, | |
| "grad_norm": 0.6876556873321533, | |
| "learning_rate": 9.899768207624593e-07, | |
| "loss": 0.0675, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.4585987261146497, | |
| "grad_norm": 1.759175419807434, | |
| "learning_rate": 9.882681531943048e-07, | |
| "loss": 0.075, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.4592873127904975, | |
| "grad_norm": 0.6456977128982544, | |
| "learning_rate": 9.865581090052471e-07, | |
| "loss": 0.0409, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.4599758994663453, | |
| "grad_norm": 1.0619314908981323, | |
| "learning_rate": 9.84846698075258e-07, | |
| "loss": 0.0637, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.46066448614219313, | |
| "grad_norm": 0.6641900539398193, | |
| "learning_rate": 9.831339302922052e-07, | |
| "loss": 0.0507, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.46135307281804094, | |
| "grad_norm": 1.0717488527297974, | |
| "learning_rate": 9.814198155517963e-07, | |
| "loss": 0.1022, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.4620416594938888, | |
| "grad_norm": 1.3715516328811646, | |
| "learning_rate": 9.797043637575208e-07, | |
| "loss": 0.0666, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.4627302461697366, | |
| "grad_norm": 0.8388981819152832, | |
| "learning_rate": 9.779875848205935e-07, | |
| "loss": 0.0643, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.46341883284558444, | |
| "grad_norm": 1.0023797750473022, | |
| "learning_rate": 9.762694886598966e-07, | |
| "loss": 0.0676, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.46410741952143225, | |
| "grad_norm": 1.0432517528533936, | |
| "learning_rate": 9.74550085201923e-07, | |
| "loss": 0.0818, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.46479600619728006, | |
| "grad_norm": 6.716395854949951, | |
| "learning_rate": 9.728293843807181e-07, | |
| "loss": 0.072, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.4654845928731279, | |
| "grad_norm": 0.6503903865814209, | |
| "learning_rate": 9.711073961378237e-07, | |
| "loss": 0.0522, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.46617317954897575, | |
| "grad_norm": 0.9570492506027222, | |
| "learning_rate": 9.69384130422219e-07, | |
| "loss": 0.0791, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.46686176622482356, | |
| "grad_norm": 1.2909079790115356, | |
| "learning_rate": 9.676595971902651e-07, | |
| "loss": 0.0724, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.4675503529006714, | |
| "grad_norm": 0.8854145407676697, | |
| "learning_rate": 9.65933806405645e-07, | |
| "loss": 0.0772, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.4682389395765192, | |
| "grad_norm": 1.7351726293563843, | |
| "learning_rate": 9.642067680393085e-07, | |
| "loss": 0.0692, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.468927526252367, | |
| "grad_norm": 1.7346372604370117, | |
| "learning_rate": 9.624784920694125e-07, | |
| "loss": 0.074, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.4696161129282148, | |
| "grad_norm": 0.8814237117767334, | |
| "learning_rate": 9.607489884812644e-07, | |
| "loss": 0.0546, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.4703046996040627, | |
| "grad_norm": 0.7480579018592834, | |
| "learning_rate": 9.590182672672657e-07, | |
| "loss": 0.0631, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.4709932862799105, | |
| "grad_norm": 0.7951589226722717, | |
| "learning_rate": 9.572863384268507e-07, | |
| "loss": 0.0566, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.4716818729557583, | |
| "grad_norm": 1.2843166589736938, | |
| "learning_rate": 9.555532119664326e-07, | |
| "loss": 0.0687, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.4723704596316061, | |
| "grad_norm": 0.6581158638000488, | |
| "learning_rate": 9.53818897899343e-07, | |
| "loss": 0.0567, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.47305904630745393, | |
| "grad_norm": 0.9683729410171509, | |
| "learning_rate": 9.520834062457757e-07, | |
| "loss": 0.0671, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.47374763298330175, | |
| "grad_norm": 0.5884563326835632, | |
| "learning_rate": 9.503467470327278e-07, | |
| "loss": 0.0701, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.4744362196591496, | |
| "grad_norm": 1.0241869688034058, | |
| "learning_rate": 9.486089302939416e-07, | |
| "loss": 0.0685, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.47512480633499743, | |
| "grad_norm": 0.9679346680641174, | |
| "learning_rate": 9.468699660698482e-07, | |
| "loss": 0.0709, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.47581339301084524, | |
| "grad_norm": 1.4202946424484253, | |
| "learning_rate": 9.451298644075082e-07, | |
| "loss": 0.0485, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.47650197968669306, | |
| "grad_norm": 0.7277752757072449, | |
| "learning_rate": 9.433886353605525e-07, | |
| "loss": 0.0655, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.47719056636254087, | |
| "grad_norm": 1.3876338005065918, | |
| "learning_rate": 9.416462889891278e-07, | |
| "loss": 0.0618, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.4778791530383887, | |
| "grad_norm": 0.7729571461677551, | |
| "learning_rate": 9.399028353598343e-07, | |
| "loss": 0.0542, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.47856773971423655, | |
| "grad_norm": 0.7464489936828613, | |
| "learning_rate": 9.381582845456707e-07, | |
| "loss": 0.0547, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.47925632639008436, | |
| "grad_norm": 1.2545253038406372, | |
| "learning_rate": 9.364126466259746e-07, | |
| "loss": 0.0575, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.4799449130659322, | |
| "grad_norm": 0.7285553216934204, | |
| "learning_rate": 9.346659316863638e-07, | |
| "loss": 0.0581, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.48063349974178, | |
| "grad_norm": 1.683297872543335, | |
| "learning_rate": 9.329181498186797e-07, | |
| "loss": 0.0729, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.4813220864176278, | |
| "grad_norm": 1.1888490915298462, | |
| "learning_rate": 9.311693111209272e-07, | |
| "loss": 0.058, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.4820106730934756, | |
| "grad_norm": 1.1569621562957764, | |
| "learning_rate": 9.294194256972179e-07, | |
| "loss": 0.0625, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.4826992597693235, | |
| "grad_norm": 1.0742779970169067, | |
| "learning_rate": 9.2766850365771e-07, | |
| "loss": 0.0654, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.4833878464451713, | |
| "grad_norm": 0.7003283500671387, | |
| "learning_rate": 9.259165551185518e-07, | |
| "loss": 0.0521, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.4840764331210191, | |
| "grad_norm": 1.9456571340560913, | |
| "learning_rate": 9.241635902018215e-07, | |
| "loss": 0.057, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.4847650197968669, | |
| "grad_norm": 1.381656527519226, | |
| "learning_rate": 9.224096190354705e-07, | |
| "loss": 0.0667, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.48545360647271474, | |
| "grad_norm": 0.6088836789131165, | |
| "learning_rate": 9.206546517532632e-07, | |
| "loss": 0.0493, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.48614219314856255, | |
| "grad_norm": 2.731722354888916, | |
| "learning_rate": 9.188986984947191e-07, | |
| "loss": 0.0701, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.4868307798244104, | |
| "grad_norm": 0.9402956366539001, | |
| "learning_rate": 9.171417694050546e-07, | |
| "loss": 0.0518, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.48751936650025823, | |
| "grad_norm": 0.6874627470970154, | |
| "learning_rate": 9.153838746351244e-07, | |
| "loss": 0.0559, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.48820795317610605, | |
| "grad_norm": 0.6484014987945557, | |
| "learning_rate": 9.136250243413615e-07, | |
| "loss": 0.0557, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.48889653985195386, | |
| "grad_norm": 1.3081269264221191, | |
| "learning_rate": 9.11865228685721e-07, | |
| "loss": 0.0602, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.4895851265278017, | |
| "grad_norm": 1.0417407751083374, | |
| "learning_rate": 9.10104497835618e-07, | |
| "loss": 0.0653, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.4902737132036495, | |
| "grad_norm": 0.9068683981895447, | |
| "learning_rate": 9.083428419638731e-07, | |
| "loss": 0.0631, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.49096229987949735, | |
| "grad_norm": 0.6790648698806763, | |
| "learning_rate": 9.065802712486495e-07, | |
| "loss": 0.0493, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.49165088655534517, | |
| "grad_norm": 2.2056519985198975, | |
| "learning_rate": 9.048167958733967e-07, | |
| "loss": 0.0687, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.492339473231193, | |
| "grad_norm": 0.6747722625732422, | |
| "learning_rate": 9.030524260267908e-07, | |
| "loss": 0.0742, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.4930280599070408, | |
| "grad_norm": 1.328088402748108, | |
| "learning_rate": 9.012871719026761e-07, | |
| "loss": 0.0901, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.4937166465828886, | |
| "grad_norm": 0.6692374348640442, | |
| "learning_rate": 8.995210437000057e-07, | |
| "loss": 0.0945, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.4944052332587364, | |
| "grad_norm": 0.7426961064338684, | |
| "learning_rate": 8.977540516227829e-07, | |
| "loss": 0.0513, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.4950938199345843, | |
| "grad_norm": 1.1755691766738892, | |
| "learning_rate": 8.959862058800017e-07, | |
| "loss": 0.0706, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.4957824066104321, | |
| "grad_norm": 0.8059654235839844, | |
| "learning_rate": 8.942175166855892e-07, | |
| "loss": 0.0571, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.4964709932862799, | |
| "grad_norm": 1.2894006967544556, | |
| "learning_rate": 8.924479942583443e-07, | |
| "loss": 0.0573, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.49715957996212773, | |
| "grad_norm": 0.7330460548400879, | |
| "learning_rate": 8.906776488218814e-07, | |
| "loss": 0.0847, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.49784816663797554, | |
| "grad_norm": 0.9233019351959229, | |
| "learning_rate": 8.889064906045685e-07, | |
| "loss": 0.072, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.49853675331382336, | |
| "grad_norm": 0.7230072617530823, | |
| "learning_rate": 8.871345298394711e-07, | |
| "loss": 0.0632, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.4992253399896712, | |
| "grad_norm": 0.9280722141265869, | |
| "learning_rate": 8.853617767642904e-07, | |
| "loss": 0.0559, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.49991392666551904, | |
| "grad_norm": 1.118358850479126, | |
| "learning_rate": 8.835882416213052e-07, | |
| "loss": 0.0653, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.5006025133413669, | |
| "grad_norm": 0.9093000292778015, | |
| "learning_rate": 8.818139346573137e-07, | |
| "loss": 0.0536, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.5012911000172147, | |
| "grad_norm": 1.1668764352798462, | |
| "learning_rate": 8.800388661235722e-07, | |
| "loss": 0.0647, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.5019796866930625, | |
| "grad_norm": 1.3647384643554688, | |
| "learning_rate": 8.782630462757385e-07, | |
| "loss": 0.0652, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.5026682733689103, | |
| "grad_norm": 1.5207314491271973, | |
| "learning_rate": 8.764864853738098e-07, | |
| "loss": 0.0663, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.5033568600447581, | |
| "grad_norm": 1.2120305299758911, | |
| "learning_rate": 8.747091936820653e-07, | |
| "loss": 0.0592, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.5040454467206059, | |
| "grad_norm": 0.7326256632804871, | |
| "learning_rate": 8.729311814690068e-07, | |
| "loss": 0.0532, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.5047340333964537, | |
| "grad_norm": 0.6355156302452087, | |
| "learning_rate": 8.711524590072988e-07, | |
| "loss": 0.0623, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.5054226200723017, | |
| "grad_norm": 1.2010247707366943, | |
| "learning_rate": 8.693730365737092e-07, | |
| "loss": 0.0644, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.5061112067481495, | |
| "grad_norm": 1.1812398433685303, | |
| "learning_rate": 8.675929244490498e-07, | |
| "loss": 0.0729, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.5067997934239973, | |
| "grad_norm": 1.1439708471298218, | |
| "learning_rate": 8.658121329181176e-07, | |
| "loss": 0.0739, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.5074883800998451, | |
| "grad_norm": 0.9230808019638062, | |
| "learning_rate": 8.640306722696352e-07, | |
| "loss": 0.0649, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.5081769667756929, | |
| "grad_norm": 0.8821923732757568, | |
| "learning_rate": 8.622485527961906e-07, | |
| "loss": 0.0635, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.5088655534515407, | |
| "grad_norm": 1.6253490447998047, | |
| "learning_rate": 8.604657847941781e-07, | |
| "loss": 0.0533, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.5095541401273885, | |
| "grad_norm": 0.8836159706115723, | |
| "learning_rate": 8.586823785637393e-07, | |
| "loss": 0.0699, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5102427268032363, | |
| "grad_norm": 0.7093166708946228, | |
| "learning_rate": 8.568983444087034e-07, | |
| "loss": 0.064, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.5109313134790842, | |
| "grad_norm": 1.1154043674468994, | |
| "learning_rate": 8.551136926365272e-07, | |
| "loss": 0.0541, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.511619900154932, | |
| "grad_norm": 0.6078494191169739, | |
| "learning_rate": 8.533284335582355e-07, | |
| "loss": 0.0577, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.5123084868307798, | |
| "grad_norm": 0.7072514295578003, | |
| "learning_rate": 8.515425774883629e-07, | |
| "loss": 0.0654, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.5129970735066276, | |
| "grad_norm": 1.0565458536148071, | |
| "learning_rate": 8.497561347448919e-07, | |
| "loss": 0.0693, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.5136856601824755, | |
| "grad_norm": 0.6897779107093811, | |
| "learning_rate": 8.47969115649196e-07, | |
| "loss": 0.0538, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.5143742468583233, | |
| "grad_norm": 2.510122299194336, | |
| "learning_rate": 8.461815305259775e-07, | |
| "loss": 0.071, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.5150628335341711, | |
| "grad_norm": 2.774024248123169, | |
| "learning_rate": 8.443933897032092e-07, | |
| "loss": 0.0866, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.515751420210019, | |
| "grad_norm": 0.7493607401847839, | |
| "learning_rate": 8.426047035120748e-07, | |
| "loss": 0.0564, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.5164400068858668, | |
| "grad_norm": 0.9617034792900085, | |
| "learning_rate": 8.408154822869088e-07, | |
| "loss": 0.0472, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5171285935617146, | |
| "grad_norm": 0.8855732083320618, | |
| "learning_rate": 8.390257363651374e-07, | |
| "loss": 0.0651, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.5178171802375624, | |
| "grad_norm": 0.7951659560203552, | |
| "learning_rate": 8.372354760872173e-07, | |
| "loss": 0.0556, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.5185057669134102, | |
| "grad_norm": 0.7143845558166504, | |
| "learning_rate": 8.354447117965775e-07, | |
| "loss": 0.0607, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.519194353589258, | |
| "grad_norm": 0.6807696223258972, | |
| "learning_rate": 8.336534538395593e-07, | |
| "loss": 0.0553, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.5198829402651058, | |
| "grad_norm": 1.0912590026855469, | |
| "learning_rate": 8.318617125653556e-07, | |
| "loss": 0.0755, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.5205715269409537, | |
| "grad_norm": 2.645685911178589, | |
| "learning_rate": 8.300694983259521e-07, | |
| "loss": 0.0702, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.5212601136168015, | |
| "grad_norm": 1.1050947904586792, | |
| "learning_rate": 8.282768214760669e-07, | |
| "loss": 0.0619, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.5219487002926494, | |
| "grad_norm": 1.451743721961975, | |
| "learning_rate": 8.264836923730912e-07, | |
| "loss": 0.0553, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.5226372869684972, | |
| "grad_norm": 1.4950122833251953, | |
| "learning_rate": 8.246901213770288e-07, | |
| "loss": 0.0648, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.523325873644345, | |
| "grad_norm": 0.9115877747535706, | |
| "learning_rate": 8.228961188504367e-07, | |
| "loss": 0.0751, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.5240144603201928, | |
| "grad_norm": 0.923009991645813, | |
| "learning_rate": 8.211016951583652e-07, | |
| "loss": 0.0703, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.5247030469960406, | |
| "grad_norm": 0.9123929738998413, | |
| "learning_rate": 8.193068606682978e-07, | |
| "loss": 0.0673, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.5253916336718885, | |
| "grad_norm": 0.8821322321891785, | |
| "learning_rate": 8.175116257500919e-07, | |
| "loss": 0.0666, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.5260802203477363, | |
| "grad_norm": 1.231985330581665, | |
| "learning_rate": 8.157160007759173e-07, | |
| "loss": 0.0625, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.5267688070235841, | |
| "grad_norm": 0.6930813789367676, | |
| "learning_rate": 8.139199961201986e-07, | |
| "loss": 0.0706, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.5274573936994319, | |
| "grad_norm": 1.2772164344787598, | |
| "learning_rate": 8.121236221595536e-07, | |
| "loss": 0.0559, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.5281459803752797, | |
| "grad_norm": 0.8855903744697571, | |
| "learning_rate": 8.103268892727335e-07, | |
| "loss": 0.0635, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.5288345670511275, | |
| "grad_norm": 1.218009352684021, | |
| "learning_rate": 8.085298078405637e-07, | |
| "loss": 0.0699, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.5295231537269753, | |
| "grad_norm": 1.2904692888259888, | |
| "learning_rate": 8.06732388245883e-07, | |
| "loss": 0.0699, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.5302117404028232, | |
| "grad_norm": 1.2834588289260864, | |
| "learning_rate": 8.049346408734843e-07, | |
| "loss": 0.0483, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.5309003270786711, | |
| "grad_norm": 0.6944026947021484, | |
| "learning_rate": 8.03136576110054e-07, | |
| "loss": 0.0705, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.5315889137545189, | |
| "grad_norm": 0.9683507084846497, | |
| "learning_rate": 8.013382043441125e-07, | |
| "loss": 0.0653, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.5322775004303667, | |
| "grad_norm": 0.5813625454902649, | |
| "learning_rate": 7.995395359659536e-07, | |
| "loss": 0.0597, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.5329660871062145, | |
| "grad_norm": 0.6153557300567627, | |
| "learning_rate": 7.977405813675849e-07, | |
| "loss": 0.0571, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.5336546737820623, | |
| "grad_norm": 1.0479635000228882, | |
| "learning_rate": 7.959413509426683e-07, | |
| "loss": 0.0594, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.5343432604579101, | |
| "grad_norm": 1.587385892868042, | |
| "learning_rate": 7.941418550864585e-07, | |
| "loss": 0.0604, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.535031847133758, | |
| "grad_norm": 1.0947626829147339, | |
| "learning_rate": 7.923421041957441e-07, | |
| "loss": 0.0525, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.5357204338096058, | |
| "grad_norm": 1.3159915208816528, | |
| "learning_rate": 7.905421086687869e-07, | |
| "loss": 0.0647, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.5364090204854536, | |
| "grad_norm": 1.9973822832107544, | |
| "learning_rate": 7.887418789052627e-07, | |
| "loss": 0.0626, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.5370976071613014, | |
| "grad_norm": 1.3887907266616821, | |
| "learning_rate": 7.869414253062005e-07, | |
| "loss": 0.0613, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.5377861938371492, | |
| "grad_norm": 1.3360315561294556, | |
| "learning_rate": 7.851407582739217e-07, | |
| "loss": 0.0518, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.538474780512997, | |
| "grad_norm": 1.0960276126861572, | |
| "learning_rate": 7.833398882119819e-07, | |
| "loss": 0.0451, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.5391633671888449, | |
| "grad_norm": 0.7975735664367676, | |
| "learning_rate": 7.815388255251093e-07, | |
| "loss": 0.0608, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.5398519538646928, | |
| "grad_norm": 0.635464608669281, | |
| "learning_rate": 7.797375806191448e-07, | |
| "loss": 0.0674, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.5405405405405406, | |
| "grad_norm": 2.0832087993621826, | |
| "learning_rate": 7.779361639009826e-07, | |
| "loss": 0.0611, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.5412291272163884, | |
| "grad_norm": 0.7622383832931519, | |
| "learning_rate": 7.761345857785089e-07, | |
| "loss": 0.0548, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.5419177138922362, | |
| "grad_norm": 0.9756510257720947, | |
| "learning_rate": 7.743328566605428e-07, | |
| "loss": 0.0668, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.542606300568084, | |
| "grad_norm": 0.8937543034553528, | |
| "learning_rate": 7.725309869567762e-07, | |
| "loss": 0.0815, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.5432948872439318, | |
| "grad_norm": 1.4156814813613892, | |
| "learning_rate": 7.707289870777125e-07, | |
| "loss": 0.0742, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.5439834739197796, | |
| "grad_norm": 1.117104411125183, | |
| "learning_rate": 7.689268674346077e-07, | |
| "loss": 0.064, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.5446720605956274, | |
| "grad_norm": 1.0066673755645752, | |
| "learning_rate": 7.671246384394092e-07, | |
| "loss": 0.0563, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.5453606472714753, | |
| "grad_norm": 1.0756511688232422, | |
| "learning_rate": 7.653223105046972e-07, | |
| "loss": 0.0808, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.5460492339473231, | |
| "grad_norm": 1.359854817390442, | |
| "learning_rate": 7.635198940436224e-07, | |
| "loss": 0.0649, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.5467378206231709, | |
| "grad_norm": 1.2628271579742432, | |
| "learning_rate": 7.617173994698479e-07, | |
| "loss": 0.0668, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.5474264072990188, | |
| "grad_norm": 0.8253432512283325, | |
| "learning_rate": 7.599148371974873e-07, | |
| "loss": 0.0718, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.5481149939748666, | |
| "grad_norm": 0.6616992354393005, | |
| "learning_rate": 7.581122176410458e-07, | |
| "loss": 0.0539, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.5488035806507144, | |
| "grad_norm": 1.1883723735809326, | |
| "learning_rate": 7.563095512153599e-07, | |
| "loss": 0.0659, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.5494921673265623, | |
| "grad_norm": 0.6985335946083069, | |
| "learning_rate": 7.54506848335536e-07, | |
| "loss": 0.0535, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.5501807540024101, | |
| "grad_norm": 0.9789535999298096, | |
| "learning_rate": 7.527041194168919e-07, | |
| "loss": 0.0612, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.5508693406782579, | |
| "grad_norm": 1.8599963188171387, | |
| "learning_rate": 7.509013748748953e-07, | |
| "loss": 0.1093, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5508693406782579, | |
| "eval_loss": NaN, | |
| "eval_runtime": 65.384, | |
| "eval_samples_per_second": 14.377, | |
| "eval_steps_per_second": 0.459, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5515579273541057, | |
| "grad_norm": 0.669114887714386, | |
| "learning_rate": 7.490986251251047e-07, | |
| "loss": 0.0583, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.5522465140299535, | |
| "grad_norm": 1.12394118309021, | |
| "learning_rate": 7.472958805831083e-07, | |
| "loss": 0.0678, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.5529351007058013, | |
| "grad_norm": 1.0287014245986938, | |
| "learning_rate": 7.454931516644641e-07, | |
| "loss": 0.0812, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.5536236873816491, | |
| "grad_norm": 1.0715841054916382, | |
| "learning_rate": 7.436904487846403e-07, | |
| "loss": 0.0908, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.554312274057497, | |
| "grad_norm": 0.9400249719619751, | |
| "learning_rate": 7.418877823589541e-07, | |
| "loss": 0.078, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.5550008607333448, | |
| "grad_norm": 0.9705797433853149, | |
| "learning_rate": 7.400851628025128e-07, | |
| "loss": 0.0483, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.5556894474091927, | |
| "grad_norm": 1.586593747138977, | |
| "learning_rate": 7.382826005301522e-07, | |
| "loss": 0.0555, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.5563780340850405, | |
| "grad_norm": 0.8826665282249451, | |
| "learning_rate": 7.364801059563776e-07, | |
| "loss": 0.0504, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.5570666207608883, | |
| "grad_norm": 1.2483572959899902, | |
| "learning_rate": 7.34677689495303e-07, | |
| "loss": 0.0631, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.5577552074367361, | |
| "grad_norm": 0.9542162418365479, | |
| "learning_rate": 7.32875361560591e-07, | |
| "loss": 0.0561, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.5584437941125839, | |
| "grad_norm": 0.7871468663215637, | |
| "learning_rate": 7.310731325653926e-07, | |
| "loss": 0.0585, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.5591323807884317, | |
| "grad_norm": 1.8042188882827759, | |
| "learning_rate": 7.292710129222878e-07, | |
| "loss": 0.0833, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.5598209674642796, | |
| "grad_norm": 0.7162600159645081, | |
| "learning_rate": 7.27469013043224e-07, | |
| "loss": 0.0633, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.5605095541401274, | |
| "grad_norm": 0.8553561568260193, | |
| "learning_rate": 7.256671433394572e-07, | |
| "loss": 0.0609, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.5611981408159752, | |
| "grad_norm": 1.5041810274124146, | |
| "learning_rate": 7.238654142214913e-07, | |
| "loss": 0.0667, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.561886727491823, | |
| "grad_norm": 0.855491578578949, | |
| "learning_rate": 7.220638360990176e-07, | |
| "loss": 0.0567, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.5625753141676708, | |
| "grad_norm": 0.6591017842292786, | |
| "learning_rate": 7.202624193808553e-07, | |
| "loss": 0.0585, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.5632639008435186, | |
| "grad_norm": 1.0353714227676392, | |
| "learning_rate": 7.184611744748908e-07, | |
| "loss": 0.0614, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.5639524875193666, | |
| "grad_norm": 0.8775607943534851, | |
| "learning_rate": 7.166601117880181e-07, | |
| "loss": 0.0731, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.5646410741952144, | |
| "grad_norm": 1.0326716899871826, | |
| "learning_rate": 7.148592417260784e-07, | |
| "loss": 0.0569, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.5653296608710622, | |
| "grad_norm": 2.2870261669158936, | |
| "learning_rate": 7.130585746937999e-07, | |
| "loss": 0.0705, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.56601824754691, | |
| "grad_norm": 0.7294363975524902, | |
| "learning_rate": 7.112581210947371e-07, | |
| "loss": 0.0691, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.5667068342227578, | |
| "grad_norm": 1.2912131547927856, | |
| "learning_rate": 7.094578913312129e-07, | |
| "loss": 0.0815, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.5673954208986056, | |
| "grad_norm": 0.8258212804794312, | |
| "learning_rate": 7.076578958042562e-07, | |
| "loss": 0.0593, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.5680840075744534, | |
| "grad_norm": 1.1386826038360596, | |
| "learning_rate": 7.058581449135415e-07, | |
| "loss": 0.0638, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.5687725942503012, | |
| "grad_norm": 1.535383939743042, | |
| "learning_rate": 7.040586490573317e-07, | |
| "loss": 0.0872, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.5694611809261491, | |
| "grad_norm": 1.2025022506713867, | |
| "learning_rate": 7.022594186324152e-07, | |
| "loss": 0.0792, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.5701497676019969, | |
| "grad_norm": 0.8510024547576904, | |
| "learning_rate": 7.004604640340467e-07, | |
| "loss": 0.0607, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.5708383542778447, | |
| "grad_norm": 1.6770105361938477, | |
| "learning_rate": 6.986617956558879e-07, | |
| "loss": 0.0864, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.5715269409536925, | |
| "grad_norm": 0.8720819354057312, | |
| "learning_rate": 6.968634238899461e-07, | |
| "loss": 0.0627, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.5722155276295404, | |
| "grad_norm": 0.9065396785736084, | |
| "learning_rate": 6.950653591265157e-07, | |
| "loss": 0.0517, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.5729041143053882, | |
| "grad_norm": 1.2403937578201294, | |
| "learning_rate": 6.932676117541171e-07, | |
| "loss": 0.0711, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.573592700981236, | |
| "grad_norm": 0.6375687718391418, | |
| "learning_rate": 6.914701921594364e-07, | |
| "loss": 0.0726, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.5742812876570839, | |
| "grad_norm": 3.1151282787323, | |
| "learning_rate": 6.896731107272666e-07, | |
| "loss": 0.0731, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.5749698743329317, | |
| "grad_norm": 1.1992899179458618, | |
| "learning_rate": 6.878763778404465e-07, | |
| "loss": 0.0628, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.5756584610087795, | |
| "grad_norm": 0.9321326613426208, | |
| "learning_rate": 6.860800038798014e-07, | |
| "loss": 0.074, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.5763470476846273, | |
| "grad_norm": 0.6797460317611694, | |
| "learning_rate": 6.842839992240828e-07, | |
| "loss": 0.0609, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.5770356343604751, | |
| "grad_norm": 1.9307366609573364, | |
| "learning_rate": 6.824883742499084e-07, | |
| "loss": 0.0642, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.5777242210363229, | |
| "grad_norm": 0.5758196711540222, | |
| "learning_rate": 6.806931393317021e-07, | |
| "loss": 0.06, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.5784128077121707, | |
| "grad_norm": 0.7546324133872986, | |
| "learning_rate": 6.788983048416348e-07, | |
| "loss": 0.0797, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.5791013943880186, | |
| "grad_norm": 0.6275709271430969, | |
| "learning_rate": 6.771038811495634e-07, | |
| "loss": 0.0651, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.5797899810638664, | |
| "grad_norm": 0.8625208735466003, | |
| "learning_rate": 6.753098786229712e-07, | |
| "loss": 0.0747, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.5804785677397143, | |
| "grad_norm": 3.9727537631988525, | |
| "learning_rate": 6.735163076269089e-07, | |
| "loss": 0.0726, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.5811671544155621, | |
| "grad_norm": 1.0236303806304932, | |
| "learning_rate": 6.717231785239332e-07, | |
| "loss": 0.0703, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.5818557410914099, | |
| "grad_norm": 1.2694447040557861, | |
| "learning_rate": 6.699305016740482e-07, | |
| "loss": 0.0593, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.5825443277672577, | |
| "grad_norm": 1.1452184915542603, | |
| "learning_rate": 6.681382874346448e-07, | |
| "loss": 0.0507, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.5832329144431055, | |
| "grad_norm": 1.218812346458435, | |
| "learning_rate": 6.663465461604408e-07, | |
| "loss": 0.0795, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.5839215011189534, | |
| "grad_norm": 1.0507051944732666, | |
| "learning_rate": 6.645552882034224e-07, | |
| "loss": 0.0521, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.5846100877948012, | |
| "grad_norm": 0.8243395686149597, | |
| "learning_rate": 6.627645239127829e-07, | |
| "loss": 0.0744, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.585298674470649, | |
| "grad_norm": 0.9169445633888245, | |
| "learning_rate": 6.609742636348628e-07, | |
| "loss": 0.0595, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.5859872611464968, | |
| "grad_norm": 0.7174025177955627, | |
| "learning_rate": 6.591845177130912e-07, | |
| "loss": 0.06, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.5866758478223446, | |
| "grad_norm": 0.6325875520706177, | |
| "learning_rate": 6.573952964879253e-07, | |
| "loss": 0.0538, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.5873644344981924, | |
| "grad_norm": 0.8738754391670227, | |
| "learning_rate": 6.556066102967912e-07, | |
| "loss": 0.0673, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.5880530211740402, | |
| "grad_norm": 0.7679761052131653, | |
| "learning_rate": 6.538184694740228e-07, | |
| "loss": 0.0467, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.5887416078498882, | |
| "grad_norm": 2.8244290351867676, | |
| "learning_rate": 6.52030884350804e-07, | |
| "loss": 0.0764, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.589430194525736, | |
| "grad_norm": 1.1413732767105103, | |
| "learning_rate": 6.50243865255108e-07, | |
| "loss": 0.0549, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.5901187812015838, | |
| "grad_norm": 0.7082204818725586, | |
| "learning_rate": 6.484574225116372e-07, | |
| "loss": 0.0484, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.5908073678774316, | |
| "grad_norm": 1.241410732269287, | |
| "learning_rate": 6.466715664417646e-07, | |
| "loss": 0.0544, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.5914959545532794, | |
| "grad_norm": 0.8751795291900635, | |
| "learning_rate": 6.44886307363473e-07, | |
| "loss": 0.0615, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.5921845412291272, | |
| "grad_norm": 1.266106367111206, | |
| "learning_rate": 6.431016555912967e-07, | |
| "loss": 0.0646, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.592873127904975, | |
| "grad_norm": 1.332010269165039, | |
| "learning_rate": 6.413176214362607e-07, | |
| "loss": 0.052, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.5935617145808229, | |
| "grad_norm": 0.8078105449676514, | |
| "learning_rate": 6.395342152058221e-07, | |
| "loss": 0.0576, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.5942503012566707, | |
| "grad_norm": 0.7434984445571899, | |
| "learning_rate": 6.377514472038098e-07, | |
| "loss": 0.0637, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.5949388879325185, | |
| "grad_norm": 0.8975241780281067, | |
| "learning_rate": 6.359693277303647e-07, | |
| "loss": 0.0583, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.5956274746083663, | |
| "grad_norm": 1.3634800910949707, | |
| "learning_rate": 6.341878670818823e-07, | |
| "loss": 0.075, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.5963160612842141, | |
| "grad_norm": 0.6128592491149902, | |
| "learning_rate": 6.324070755509505e-07, | |
| "loss": 0.0547, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.5970046479600619, | |
| "grad_norm": 1.5584526062011719, | |
| "learning_rate": 6.306269634262911e-07, | |
| "loss": 0.0681, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.5976932346359098, | |
| "grad_norm": 0.6906041502952576, | |
| "learning_rate": 6.288475409927014e-07, | |
| "loss": 0.049, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.5983818213117577, | |
| "grad_norm": 0.5763527750968933, | |
| "learning_rate": 6.270688185309933e-07, | |
| "loss": 0.0496, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.5990704079876055, | |
| "grad_norm": 0.8747649788856506, | |
| "learning_rate": 6.25290806317935e-07, | |
| "loss": 0.0562, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.5997589946634533, | |
| "grad_norm": 1.1523576974868774, | |
| "learning_rate": 6.235135146261905e-07, | |
| "loss": 0.0463, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.6004475813393011, | |
| "grad_norm": 1.9650055170059204, | |
| "learning_rate": 6.217369537242615e-07, | |
| "loss": 0.0552, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.6011361680151489, | |
| "grad_norm": 0.8052722811698914, | |
| "learning_rate": 6.199611338764277e-07, | |
| "loss": 0.0551, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.6018247546909967, | |
| "grad_norm": 3.1632587909698486, | |
| "learning_rate": 6.181860653426865e-07, | |
| "loss": 0.0627, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.6025133413668445, | |
| "grad_norm": 1.2010645866394043, | |
| "learning_rate": 6.164117583786949e-07, | |
| "loss": 0.0764, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.6032019280426923, | |
| "grad_norm": 0.8220053315162659, | |
| "learning_rate": 6.146382232357099e-07, | |
| "loss": 0.055, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.6038905147185402, | |
| "grad_norm": 1.2072614431381226, | |
| "learning_rate": 6.12865470160529e-07, | |
| "loss": 0.0707, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.604579101394388, | |
| "grad_norm": 0.6051107048988342, | |
| "learning_rate": 6.110935093954315e-07, | |
| "loss": 0.0549, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.6052676880702358, | |
| "grad_norm": 0.9836848378181458, | |
| "learning_rate": 6.09322351178119e-07, | |
| "loss": 0.0509, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.6059562747460837, | |
| "grad_norm": 1.1881619691848755, | |
| "learning_rate": 6.075520057416561e-07, | |
| "loss": 0.0667, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.6066448614219315, | |
| "grad_norm": 1.4986263513565063, | |
| "learning_rate": 6.057824833144109e-07, | |
| "loss": 0.0787, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.6073334480977793, | |
| "grad_norm": 1.1127889156341553, | |
| "learning_rate": 6.040137941199982e-07, | |
| "loss": 0.062, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.6080220347736272, | |
| "grad_norm": 1.376637578010559, | |
| "learning_rate": 6.022459483772174e-07, | |
| "loss": 0.0591, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.608710621449475, | |
| "grad_norm": 0.9518951773643494, | |
| "learning_rate": 6.004789562999943e-07, | |
| "loss": 0.0679, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.6093992081253228, | |
| "grad_norm": 1.337842345237732, | |
| "learning_rate": 5.987128280973241e-07, | |
| "loss": 0.0468, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.6100877948011706, | |
| "grad_norm": 0.774268627166748, | |
| "learning_rate": 5.969475739732093e-07, | |
| "loss": 0.0498, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.6107763814770184, | |
| "grad_norm": 1.0546382665634155, | |
| "learning_rate": 5.951832041266037e-07, | |
| "loss": 0.0714, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.6114649681528662, | |
| "grad_norm": 1.2130711078643799, | |
| "learning_rate": 5.934197287513507e-07, | |
| "loss": 0.0861, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.612153554828714, | |
| "grad_norm": 1.6215195655822754, | |
| "learning_rate": 5.916571580361268e-07, | |
| "loss": 0.0804, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.6128421415045618, | |
| "grad_norm": 1.2330925464630127, | |
| "learning_rate": 5.898955021643818e-07, | |
| "loss": 0.0695, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.6135307281804097, | |
| "grad_norm": 2.0326528549194336, | |
| "learning_rate": 5.881347713142792e-07, | |
| "loss": 0.0674, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.6142193148562576, | |
| "grad_norm": 1.2848743200302124, | |
| "learning_rate": 5.863749756586385e-07, | |
| "loss": 0.0777, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.6149079015321054, | |
| "grad_norm": 0.8805621862411499, | |
| "learning_rate": 5.846161253648758e-07, | |
| "loss": 0.0539, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.6155964882079532, | |
| "grad_norm": 1.2727314233779907, | |
| "learning_rate": 5.828582305949455e-07, | |
| "loss": 0.0725, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.616285074883801, | |
| "grad_norm": 0.9873810410499573, | |
| "learning_rate": 5.811013015052811e-07, | |
| "loss": 0.0684, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.6169736615596488, | |
| "grad_norm": 1.8348257541656494, | |
| "learning_rate": 5.793453482467372e-07, | |
| "loss": 0.0807, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.6176622482354966, | |
| "grad_norm": 0.45594796538352966, | |
| "learning_rate": 5.775903809645295e-07, | |
| "loss": 0.0551, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.6183508349113445, | |
| "grad_norm": 0.9970102906227112, | |
| "learning_rate": 5.758364097981783e-07, | |
| "loss": 0.0523, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.6190394215871923, | |
| "grad_norm": 1.9257394075393677, | |
| "learning_rate": 5.740834448814483e-07, | |
| "loss": 0.0701, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.6197280082630401, | |
| "grad_norm": 0.5468912124633789, | |
| "learning_rate": 5.723314963422902e-07, | |
| "loss": 0.0561, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6204165949388879, | |
| "grad_norm": 1.1196551322937012, | |
| "learning_rate": 5.705805743027822e-07, | |
| "loss": 0.0725, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.6211051816147357, | |
| "grad_norm": 0.7054963111877441, | |
| "learning_rate": 5.688306888790729e-07, | |
| "loss": 0.0636, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.6217937682905835, | |
| "grad_norm": 0.7172769904136658, | |
| "learning_rate": 5.670818501813205e-07, | |
| "loss": 0.0637, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.6224823549664315, | |
| "grad_norm": 0.7681342959403992, | |
| "learning_rate": 5.653340683136364e-07, | |
| "loss": 0.0587, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.6231709416422793, | |
| "grad_norm": 0.7725322842597961, | |
| "learning_rate": 5.635873533740257e-07, | |
| "loss": 0.055, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.6238595283181271, | |
| "grad_norm": 0.627192497253418, | |
| "learning_rate": 5.618417154543293e-07, | |
| "loss": 0.0627, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.6245481149939749, | |
| "grad_norm": 1.0197110176086426, | |
| "learning_rate": 5.600971646401658e-07, | |
| "loss": 0.0666, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.6252367016698227, | |
| "grad_norm": 0.5483493804931641, | |
| "learning_rate": 5.583537110108724e-07, | |
| "loss": 0.0486, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.6259252883456705, | |
| "grad_norm": 1.1244122982025146, | |
| "learning_rate": 5.566113646394475e-07, | |
| "loss": 0.0626, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.6266138750215183, | |
| "grad_norm": 2.424347162246704, | |
| "learning_rate": 5.548701355924921e-07, | |
| "loss": 0.0721, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.6273024616973661, | |
| "grad_norm": 1.1833293437957764, | |
| "learning_rate": 5.531300339301518e-07, | |
| "loss": 0.0723, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.627991048373214, | |
| "grad_norm": 0.6992703080177307, | |
| "learning_rate": 5.513910697060585e-07, | |
| "loss": 0.0554, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.6286796350490618, | |
| "grad_norm": 1.3111495971679688, | |
| "learning_rate": 5.496532529672726e-07, | |
| "loss": 0.0886, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.6293682217249096, | |
| "grad_norm": 2.576478958129883, | |
| "learning_rate": 5.479165937542243e-07, | |
| "loss": 0.0933, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.6300568084007574, | |
| "grad_norm": 1.6602191925048828, | |
| "learning_rate": 5.46181102100657e-07, | |
| "loss": 0.0574, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.6307453950766053, | |
| "grad_norm": 0.7469310164451599, | |
| "learning_rate": 5.444467880335675e-07, | |
| "loss": 0.0627, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.6314339817524531, | |
| "grad_norm": 0.6428240537643433, | |
| "learning_rate": 5.427136615731493e-07, | |
| "loss": 0.046, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.632122568428301, | |
| "grad_norm": 0.5253523588180542, | |
| "learning_rate": 5.409817327327345e-07, | |
| "loss": 0.0462, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.6328111551041488, | |
| "grad_norm": 0.6976008415222168, | |
| "learning_rate": 5.392510115187357e-07, | |
| "loss": 0.0567, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.6334997417799966, | |
| "grad_norm": 0.6858130097389221, | |
| "learning_rate": 5.375215079305878e-07, | |
| "loss": 0.0498, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.6341883284558444, | |
| "grad_norm": 0.9298868179321289, | |
| "learning_rate": 5.357932319606919e-07, | |
| "loss": 0.0666, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.6348769151316922, | |
| "grad_norm": 0.6549275517463684, | |
| "learning_rate": 5.340661935943549e-07, | |
| "loss": 0.0482, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.63556550180754, | |
| "grad_norm": 1.0457004308700562, | |
| "learning_rate": 5.323404028097349e-07, | |
| "loss": 0.0599, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.6362540884833878, | |
| "grad_norm": 0.8807022571563721, | |
| "learning_rate": 5.30615869577781e-07, | |
| "loss": 0.067, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 1.3472622632980347, | |
| "learning_rate": 5.288926038621765e-07, | |
| "loss": 0.0549, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.6376312618350835, | |
| "grad_norm": 1.059233546257019, | |
| "learning_rate": 5.271706156192821e-07, | |
| "loss": 0.0548, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.6383198485109313, | |
| "grad_norm": 1.3020786046981812, | |
| "learning_rate": 5.254499147980773e-07, | |
| "loss": 0.0813, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.6390084351867792, | |
| "grad_norm": 1.4036099910736084, | |
| "learning_rate": 5.237305113401036e-07, | |
| "loss": 0.0672, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.639697021862627, | |
| "grad_norm": 0.7069018483161926, | |
| "learning_rate": 5.220124151794066e-07, | |
| "loss": 0.0576, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.6403856085384748, | |
| "grad_norm": 0.6201725602149963, | |
| "learning_rate": 5.202956362424795e-07, | |
| "loss": 0.0519, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.6410741952143226, | |
| "grad_norm": 0.9227588176727295, | |
| "learning_rate": 5.185801844482038e-07, | |
| "loss": 0.0686, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.6417627818901704, | |
| "grad_norm": 1.3585100173950195, | |
| "learning_rate": 5.168660697077948e-07, | |
| "loss": 0.0501, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.6424513685660183, | |
| "grad_norm": 0.9505581259727478, | |
| "learning_rate": 5.151533019247421e-07, | |
| "loss": 0.0682, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.6431399552418661, | |
| "grad_norm": 1.9560891389846802, | |
| "learning_rate": 5.134418909947528e-07, | |
| "loss": 0.0595, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.6438285419177139, | |
| "grad_norm": 0.598139226436615, | |
| "learning_rate": 5.117318468056952e-07, | |
| "loss": 0.0755, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.6445171285935617, | |
| "grad_norm": 0.944092333316803, | |
| "learning_rate": 5.100231792375409e-07, | |
| "loss": 0.0581, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.6452057152694095, | |
| "grad_norm": 1.233586072921753, | |
| "learning_rate": 5.083158981623073e-07, | |
| "loss": 0.0759, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.6458943019452573, | |
| "grad_norm": 1.4512312412261963, | |
| "learning_rate": 5.066100134440024e-07, | |
| "loss": 0.0649, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.6465828886211051, | |
| "grad_norm": 1.2120344638824463, | |
| "learning_rate": 5.049055349385656e-07, | |
| "loss": 0.0723, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.6472714752969531, | |
| "grad_norm": 1.418200969696045, | |
| "learning_rate": 5.032024724938122e-07, | |
| "loss": 0.0723, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.6479600619728009, | |
| "grad_norm": 0.6435151100158691, | |
| "learning_rate": 5.015008359493763e-07, | |
| "loss": 0.0596, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.6486486486486487, | |
| "grad_norm": 0.9242047071456909, | |
| "learning_rate": 4.998006351366527e-07, | |
| "loss": 0.0671, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.6493372353244965, | |
| "grad_norm": 0.827181875705719, | |
| "learning_rate": 4.981018798787422e-07, | |
| "loss": 0.0573, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.6500258220003443, | |
| "grad_norm": 0.9310402274131775, | |
| "learning_rate": 4.964045799903931e-07, | |
| "loss": 0.0602, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.6507144086761921, | |
| "grad_norm": 0.6727992296218872, | |
| "learning_rate": 4.947087452779455e-07, | |
| "loss": 0.0361, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.6514029953520399, | |
| "grad_norm": 1.1100339889526367, | |
| "learning_rate": 4.930143855392738e-07, | |
| "loss": 0.0584, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.6520915820278878, | |
| "grad_norm": 0.8588836193084717, | |
| "learning_rate": 4.913215105637313e-07, | |
| "loss": 0.061, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.6527801687037356, | |
| "grad_norm": 0.7795128226280212, | |
| "learning_rate": 4.89630130132092e-07, | |
| "loss": 0.0507, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.6534687553795834, | |
| "grad_norm": 1.1720349788665771, | |
| "learning_rate": 4.879402540164959e-07, | |
| "loss": 0.0674, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.6541573420554312, | |
| "grad_norm": 0.669318675994873, | |
| "learning_rate": 4.862518919803913e-07, | |
| "loss": 0.0766, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.654845928731279, | |
| "grad_norm": 1.0669409036636353, | |
| "learning_rate": 4.845650537784785e-07, | |
| "loss": 0.0766, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.6555345154071269, | |
| "grad_norm": 4.53080940246582, | |
| "learning_rate": 4.828797491566539e-07, | |
| "loss": 0.0507, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.6562231020829747, | |
| "grad_norm": 0.9759594202041626, | |
| "learning_rate": 4.811959878519535e-07, | |
| "loss": 0.0632, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.6569116887588226, | |
| "grad_norm": 0.7316691875457764, | |
| "learning_rate": 4.795137795924964e-07, | |
| "loss": 0.0722, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.6576002754346704, | |
| "grad_norm": 0.912428617477417, | |
| "learning_rate": 4.778331340974292e-07, | |
| "loss": 0.0705, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.6582888621105182, | |
| "grad_norm": 0.7177801728248596, | |
| "learning_rate": 4.761540610768688e-07, | |
| "loss": 0.0547, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.658977448786366, | |
| "grad_norm": 0.47414833307266235, | |
| "learning_rate": 4.7447657023184774e-07, | |
| "loss": 0.0559, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.6596660354622138, | |
| "grad_norm": 0.9815657734870911, | |
| "learning_rate": 4.7280067125425705e-07, | |
| "loss": 0.068, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.6603546221380616, | |
| "grad_norm": 0.636580228805542, | |
| "learning_rate": 4.7112637382679027e-07, | |
| "loss": 0.052, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.6610432088139094, | |
| "grad_norm": 0.8123611211776733, | |
| "learning_rate": 4.6945368762288817e-07, | |
| "loss": 0.0618, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.6617317954897572, | |
| "grad_norm": 0.5711089372634888, | |
| "learning_rate": 4.677826223066821e-07, | |
| "loss": 0.0546, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.6624203821656051, | |
| "grad_norm": 1.0411585569381714, | |
| "learning_rate": 4.661131875329393e-07, | |
| "loss": 0.0557, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.6631089688414529, | |
| "grad_norm": 0.8787512183189392, | |
| "learning_rate": 4.6444539294700546e-07, | |
| "loss": 0.0686, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.6637975555173007, | |
| "grad_norm": 1.9264168739318848, | |
| "learning_rate": 4.627792481847503e-07, | |
| "loss": 0.0767, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.6644861421931486, | |
| "grad_norm": 1.11745023727417, | |
| "learning_rate": 4.6111476287251196e-07, | |
| "loss": 0.0727, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.6651747288689964, | |
| "grad_norm": 0.7174832820892334, | |
| "learning_rate": 4.5945194662704025e-07, | |
| "loss": 0.0683, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.6658633155448442, | |
| "grad_norm": 1.3074320554733276, | |
| "learning_rate": 4.5779080905544195e-07, | |
| "loss": 0.0561, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.666551902220692, | |
| "grad_norm": 0.48328328132629395, | |
| "learning_rate": 4.5613135975512516e-07, | |
| "loss": 0.0639, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.6672404888965399, | |
| "grad_norm": 1.2463585138320923, | |
| "learning_rate": 4.54473608313744e-07, | |
| "loss": 0.0664, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.6679290755723877, | |
| "grad_norm": 0.6928512454032898, | |
| "learning_rate": 4.5281756430914247e-07, | |
| "loss": 0.0572, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.6686176622482355, | |
| "grad_norm": 0.8524710536003113, | |
| "learning_rate": 4.511632373093004e-07, | |
| "loss": 0.0533, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.6693062489240833, | |
| "grad_norm": 1.7520873546600342, | |
| "learning_rate": 4.4951063687227695e-07, | |
| "loss": 0.0663, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.6699948355999311, | |
| "grad_norm": 1.2154256105422974, | |
| "learning_rate": 4.4785977254615547e-07, | |
| "loss": 0.044, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.6706834222757789, | |
| "grad_norm": 0.9355267286300659, | |
| "learning_rate": 4.4621065386898985e-07, | |
| "loss": 0.0568, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.6713720089516267, | |
| "grad_norm": 0.6941357851028442, | |
| "learning_rate": 4.445632903687477e-07, | |
| "loss": 0.0689, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.6720605956274746, | |
| "grad_norm": 0.6184002757072449, | |
| "learning_rate": 4.429176915632555e-07, | |
| "loss": 0.0603, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.6727491823033225, | |
| "grad_norm": 1.1245633363723755, | |
| "learning_rate": 4.4127386696014475e-07, | |
| "loss": 0.0716, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.6734377689791703, | |
| "grad_norm": 0.5176686644554138, | |
| "learning_rate": 4.396318260567956e-07, | |
| "loss": 0.052, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.6741263556550181, | |
| "grad_norm": 1.0823880434036255, | |
| "learning_rate": 4.379915783402833e-07, | |
| "loss": 0.0638, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.6748149423308659, | |
| "grad_norm": 1.414318561553955, | |
| "learning_rate": 4.363531332873225e-07, | |
| "loss": 0.0816, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.6755035290067137, | |
| "grad_norm": 0.947318971157074, | |
| "learning_rate": 4.3471650036421207e-07, | |
| "loss": 0.0518, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.6761921156825615, | |
| "grad_norm": 1.2991081476211548, | |
| "learning_rate": 4.330816890267826e-07, | |
| "loss": 0.0539, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.6768807023584094, | |
| "grad_norm": 0.6339100003242493, | |
| "learning_rate": 4.314487087203394e-07, | |
| "loss": 0.0534, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.6775692890342572, | |
| "grad_norm": 1.4862066507339478, | |
| "learning_rate": 4.298175688796083e-07, | |
| "loss": 0.0536, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.678257875710105, | |
| "grad_norm": 1.5165352821350098, | |
| "learning_rate": 4.281882789286824e-07, | |
| "loss": 0.0637, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.6789464623859528, | |
| "grad_norm": 0.7241688370704651, | |
| "learning_rate": 4.2656084828096663e-07, | |
| "loss": 0.055, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.6796350490618006, | |
| "grad_norm": 8.256474494934082, | |
| "learning_rate": 4.249352863391238e-07, | |
| "loss": 0.0933, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.6803236357376484, | |
| "grad_norm": 1.5877653360366821, | |
| "learning_rate": 4.2331160249501943e-07, | |
| "loss": 0.0543, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.6810122224134963, | |
| "grad_norm": 1.0932315587997437, | |
| "learning_rate": 4.216898061296685e-07, | |
| "loss": 0.0618, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.6817008090893442, | |
| "grad_norm": 1.8836272954940796, | |
| "learning_rate": 4.2006990661318124e-07, | |
| "loss": 0.0607, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.682389395765192, | |
| "grad_norm": 1.326351523399353, | |
| "learning_rate": 4.184519133047081e-07, | |
| "loss": 0.0702, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.6830779824410398, | |
| "grad_norm": 1.0623064041137695, | |
| "learning_rate": 4.168358355523864e-07, | |
| "loss": 0.0554, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.6837665691168876, | |
| "grad_norm": 3.3747942447662354, | |
| "learning_rate": 4.1522168269328635e-07, | |
| "loss": 0.0499, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.6844551557927354, | |
| "grad_norm": 0.9363127946853638, | |
| "learning_rate": 4.136094640533561e-07, | |
| "loss": 0.0641, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.6851437424685832, | |
| "grad_norm": 0.8857420682907104, | |
| "learning_rate": 4.119991889473694e-07, | |
| "loss": 0.0505, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.685832329144431, | |
| "grad_norm": 0.546501636505127, | |
| "learning_rate": 4.103908666788708e-07, | |
| "loss": 0.0566, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.6865209158202789, | |
| "grad_norm": 0.6877146363258362, | |
| "learning_rate": 4.087845065401224e-07, | |
| "loss": 0.0505, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.6872095024961267, | |
| "grad_norm": 0.7885240912437439, | |
| "learning_rate": 4.071801178120486e-07, | |
| "loss": 0.0681, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.6878980891719745, | |
| "grad_norm": 0.8773134350776672, | |
| "learning_rate": 4.0557770976418567e-07, | |
| "loss": 0.0516, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.6885866758478223, | |
| "grad_norm": 0.5082318186759949, | |
| "learning_rate": 4.039772916546252e-07, | |
| "loss": 0.0543, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6885866758478223, | |
| "eval_loss": NaN, | |
| "eval_runtime": 65.6181, | |
| "eval_samples_per_second": 14.325, | |
| "eval_steps_per_second": 0.457, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6892752625236702, | |
| "grad_norm": 0.7060466408729553, | |
| "learning_rate": 4.0237887272996137e-07, | |
| "loss": 0.0552, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.689963849199518, | |
| "grad_norm": 1.0729354619979858, | |
| "learning_rate": 4.0078246222523856e-07, | |
| "loss": 0.0679, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.6906524358753658, | |
| "grad_norm": 0.745944619178772, | |
| "learning_rate": 3.9918806936389695e-07, | |
| "loss": 0.0559, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.6913410225512137, | |
| "grad_norm": 1.1027761697769165, | |
| "learning_rate": 3.975957033577199e-07, | |
| "loss": 0.0549, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.6920296092270615, | |
| "grad_norm": 1.2901561260223389, | |
| "learning_rate": 3.960053734067795e-07, | |
| "loss": 0.0674, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.6927181959029093, | |
| "grad_norm": 0.7766295671463013, | |
| "learning_rate": 3.944170886993849e-07, | |
| "loss": 0.0606, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.6934067825787571, | |
| "grad_norm": 0.8777790665626526, | |
| "learning_rate": 3.928308584120297e-07, | |
| "loss": 0.0643, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.6940953692546049, | |
| "grad_norm": 1.5662516355514526, | |
| "learning_rate": 3.9124669170933565e-07, | |
| "loss": 0.0687, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.6947839559304527, | |
| "grad_norm": 1.2681442499160767, | |
| "learning_rate": 3.8966459774400347e-07, | |
| "loss": 0.0515, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.6954725426063005, | |
| "grad_norm": 0.7493829727172852, | |
| "learning_rate": 3.8808458565675835e-07, | |
| "loss": 0.0521, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.6961611292821483, | |
| "grad_norm": 0.8877359628677368, | |
| "learning_rate": 3.865066645762961e-07, | |
| "loss": 0.067, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.6968497159579962, | |
| "grad_norm": 0.9092392325401306, | |
| "learning_rate": 3.849308436192326e-07, | |
| "loss": 0.0547, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.6975383026338441, | |
| "grad_norm": 1.1998776197433472, | |
| "learning_rate": 3.833571318900496e-07, | |
| "loss": 0.059, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.6982268893096919, | |
| "grad_norm": 1.0754103660583496, | |
| "learning_rate": 3.817855384810431e-07, | |
| "loss": 0.0688, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.6989154759855397, | |
| "grad_norm": 1.3370611667633057, | |
| "learning_rate": 3.802160724722687e-07, | |
| "loss": 0.0614, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.6996040626613875, | |
| "grad_norm": 0.9170624017715454, | |
| "learning_rate": 3.7864874293149283e-07, | |
| "loss": 0.0491, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.7002926493372353, | |
| "grad_norm": 0.6947853565216064, | |
| "learning_rate": 3.7708355891413704e-07, | |
| "loss": 0.0592, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.7009812360130832, | |
| "grad_norm": 0.8483502864837646, | |
| "learning_rate": 3.7552052946322657e-07, | |
| "loss": 0.0722, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.701669822688931, | |
| "grad_norm": 0.6551016569137573, | |
| "learning_rate": 3.739596636093392e-07, | |
| "loss": 0.0738, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.7023584093647788, | |
| "grad_norm": 0.7901301383972168, | |
| "learning_rate": 3.72400970370552e-07, | |
| "loss": 0.045, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.7030469960406266, | |
| "grad_norm": 0.6407663822174072, | |
| "learning_rate": 3.7084445875238954e-07, | |
| "loss": 0.0542, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.7037355827164744, | |
| "grad_norm": 0.620256245136261, | |
| "learning_rate": 3.6929013774777145e-07, | |
| "loss": 0.0565, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.7044241693923222, | |
| "grad_norm": 1.1071631908416748, | |
| "learning_rate": 3.677380163369611e-07, | |
| "loss": 0.0694, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.70511275606817, | |
| "grad_norm": 0.9421183466911316, | |
| "learning_rate": 3.661881034875143e-07, | |
| "loss": 0.0725, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.705801342744018, | |
| "grad_norm": 0.7301518321037292, | |
| "learning_rate": 3.6464040815422507e-07, | |
| "loss": 0.0583, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.7064899294198658, | |
| "grad_norm": 0.8051359057426453, | |
| "learning_rate": 3.630949392790767e-07, | |
| "loss": 0.0529, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.7071785160957136, | |
| "grad_norm": 0.49650633335113525, | |
| "learning_rate": 3.6155170579118854e-07, | |
| "loss": 0.0409, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.7078671027715614, | |
| "grad_norm": 1.2859952449798584, | |
| "learning_rate": 3.600107166067642e-07, | |
| "loss": 0.0775, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.7085556894474092, | |
| "grad_norm": 0.6026453971862793, | |
| "learning_rate": 3.584719806290414e-07, | |
| "loss": 0.0571, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.709244276123257, | |
| "grad_norm": 2.1153674125671387, | |
| "learning_rate": 3.569355067482393e-07, | |
| "loss": 0.0657, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.7099328627991048, | |
| "grad_norm": 0.5973625183105469, | |
| "learning_rate": 3.554013038415074e-07, | |
| "loss": 0.0553, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.7106214494749526, | |
| "grad_norm": 1.2447024583816528, | |
| "learning_rate": 3.538693807728749e-07, | |
| "loss": 0.0902, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.7113100361508005, | |
| "grad_norm": 0.9078012108802795, | |
| "learning_rate": 3.523397463931984e-07, | |
| "loss": 0.059, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.7119986228266483, | |
| "grad_norm": 1.5184681415557861, | |
| "learning_rate": 3.50812409540112e-07, | |
| "loss": 0.085, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.7126872095024961, | |
| "grad_norm": 1.7193793058395386, | |
| "learning_rate": 3.492873790379746e-07, | |
| "loss": 0.0752, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.7133757961783439, | |
| "grad_norm": 0.6111558079719543, | |
| "learning_rate": 3.4776466369782085e-07, | |
| "loss": 0.0466, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.7140643828541918, | |
| "grad_norm": 0.4839138984680176, | |
| "learning_rate": 3.462442723173087e-07, | |
| "loss": 0.0622, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.7147529695300396, | |
| "grad_norm": 1.4164533615112305, | |
| "learning_rate": 3.4472621368066987e-07, | |
| "loss": 0.0842, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.7154415562058875, | |
| "grad_norm": 0.66438889503479, | |
| "learning_rate": 3.4321049655865736e-07, | |
| "loss": 0.0473, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.7161301428817353, | |
| "grad_norm": 1.491725206375122, | |
| "learning_rate": 3.4169712970849627e-07, | |
| "loss": 0.0621, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.7168187295575831, | |
| "grad_norm": 0.7163612246513367, | |
| "learning_rate": 3.4018612187383375e-07, | |
| "loss": 0.0532, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.7175073162334309, | |
| "grad_norm": 1.5005770921707153, | |
| "learning_rate": 3.386774817846857e-07, | |
| "loss": 0.0754, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.7181959029092787, | |
| "grad_norm": 0.8143193125724792, | |
| "learning_rate": 3.3717121815738917e-07, | |
| "loss": 0.048, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.7188844895851265, | |
| "grad_norm": 1.1266696453094482, | |
| "learning_rate": 3.3566733969455105e-07, | |
| "loss": 0.0663, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.7195730762609743, | |
| "grad_norm": 1.5463004112243652, | |
| "learning_rate": 3.341658550849968e-07, | |
| "loss": 0.0934, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.7202616629368221, | |
| "grad_norm": 1.2996982336044312, | |
| "learning_rate": 3.3266677300372194e-07, | |
| "loss": 0.0472, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.72095024961267, | |
| "grad_norm": 1.383406639099121, | |
| "learning_rate": 3.31170102111841e-07, | |
| "loss": 0.0589, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.7216388362885178, | |
| "grad_norm": 0.43949881196022034, | |
| "learning_rate": 3.296758510565372e-07, | |
| "loss": 0.0635, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.7223274229643656, | |
| "grad_norm": 0.8941599130630493, | |
| "learning_rate": 3.2818402847101323e-07, | |
| "loss": 0.0728, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 0.7230160096402135, | |
| "grad_norm": 0.8952903151512146, | |
| "learning_rate": 3.26694642974441e-07, | |
| "loss": 0.0651, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.7237045963160613, | |
| "grad_norm": 1.3641283512115479, | |
| "learning_rate": 3.25207703171912e-07, | |
| "loss": 0.0622, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 0.7243931829919091, | |
| "grad_norm": 0.7892813682556152, | |
| "learning_rate": 3.2372321765438675e-07, | |
| "loss": 0.0565, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.725081769667757, | |
| "grad_norm": 1.0692788362503052, | |
| "learning_rate": 3.2224119499864655e-07, | |
| "loss": 0.0772, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.7257703563436048, | |
| "grad_norm": 0.8681766390800476, | |
| "learning_rate": 3.20761643767243e-07, | |
| "loss": 0.0448, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.7264589430194526, | |
| "grad_norm": 0.8469806909561157, | |
| "learning_rate": 3.192845725084492e-07, | |
| "loss": 0.0648, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.7271475296953004, | |
| "grad_norm": 0.8979076743125916, | |
| "learning_rate": 3.1780998975620896e-07, | |
| "loss": 0.0605, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.7278361163711482, | |
| "grad_norm": 1.2326704263687134, | |
| "learning_rate": 3.163379040300891e-07, | |
| "loss": 0.057, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 0.728524703046996, | |
| "grad_norm": 1.183748483657837, | |
| "learning_rate": 3.148683238352303e-07, | |
| "loss": 0.0584, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.7292132897228438, | |
| "grad_norm": 0.7711232304573059, | |
| "learning_rate": 3.134012576622956e-07, | |
| "loss": 0.0563, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.7299018763986916, | |
| "grad_norm": 1.0204769372940063, | |
| "learning_rate": 3.119367139874243e-07, | |
| "loss": 0.0572, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.7305904630745395, | |
| "grad_norm": 1.6725730895996094, | |
| "learning_rate": 3.104747012721815e-07, | |
| "loss": 0.0731, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 0.7312790497503874, | |
| "grad_norm": 0.7798736691474915, | |
| "learning_rate": 3.090152279635086e-07, | |
| "loss": 0.0713, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.7319676364262352, | |
| "grad_norm": 0.7971237897872925, | |
| "learning_rate": 3.075583024936762e-07, | |
| "loss": 0.0638, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 0.732656223102083, | |
| "grad_norm": 1.1695165634155273, | |
| "learning_rate": 3.061039332802341e-07, | |
| "loss": 0.0653, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.7333448097779308, | |
| "grad_norm": 0.6647826433181763, | |
| "learning_rate": 3.046521287259631e-07, | |
| "loss": 0.0458, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.7340333964537786, | |
| "grad_norm": 0.7141667604446411, | |
| "learning_rate": 3.0320289721882635e-07, | |
| "loss": 0.0646, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.7347219831296264, | |
| "grad_norm": 1.2314116954803467, | |
| "learning_rate": 3.01756247131921e-07, | |
| "loss": 0.0535, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 0.7354105698054743, | |
| "grad_norm": 0.859336256980896, | |
| "learning_rate": 3.0031218682343007e-07, | |
| "loss": 0.0845, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.7360991564813221, | |
| "grad_norm": 0.9779216051101685, | |
| "learning_rate": 2.9887072463657285e-07, | |
| "loss": 0.0605, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 0.7367877431571699, | |
| "grad_norm": 1.0691299438476562, | |
| "learning_rate": 2.974318688995586e-07, | |
| "loss": 0.0641, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.7374763298330177, | |
| "grad_norm": 0.7054864764213562, | |
| "learning_rate": 2.959956279255373e-07, | |
| "loss": 0.0632, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.7381649165088655, | |
| "grad_norm": 0.8933075666427612, | |
| "learning_rate": 2.9456201001255207e-07, | |
| "loss": 0.0616, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.7388535031847133, | |
| "grad_norm": 1.6148409843444824, | |
| "learning_rate": 2.9313102344349e-07, | |
| "loss": 0.0825, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 0.7395420898605612, | |
| "grad_norm": 0.7739663124084473, | |
| "learning_rate": 2.9170267648603673e-07, | |
| "loss": 0.0609, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.7402306765364091, | |
| "grad_norm": 1.013547658920288, | |
| "learning_rate": 2.9027697739262663e-07, | |
| "loss": 0.0732, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.7409192632122569, | |
| "grad_norm": 0.6638150215148926, | |
| "learning_rate": 2.8885393440039515e-07, | |
| "loss": 0.074, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.7416078498881047, | |
| "grad_norm": 0.8091201186180115, | |
| "learning_rate": 2.874335557311325e-07, | |
| "loss": 0.0473, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.7422964365639525, | |
| "grad_norm": 0.6808359622955322, | |
| "learning_rate": 2.860158495912356e-07, | |
| "loss": 0.0663, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.7429850232398003, | |
| "grad_norm": 0.7049098014831543, | |
| "learning_rate": 2.846008241716597e-07, | |
| "loss": 0.0631, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 0.7436736099156481, | |
| "grad_norm": 1.1154870986938477, | |
| "learning_rate": 2.831884876478727e-07, | |
| "loss": 0.0647, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.7443621965914959, | |
| "grad_norm": 0.5480945706367493, | |
| "learning_rate": 2.8177884817980686e-07, | |
| "loss": 0.0643, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 0.7450507832673438, | |
| "grad_norm": 1.0349327325820923, | |
| "learning_rate": 2.803719139118119e-07, | |
| "loss": 0.0658, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.7457393699431916, | |
| "grad_norm": 0.5763936042785645, | |
| "learning_rate": 2.789676929726077e-07, | |
| "loss": 0.0533, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.7464279566190394, | |
| "grad_norm": 1.1502076387405396, | |
| "learning_rate": 2.775661934752379e-07, | |
| "loss": 0.0402, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.7471165432948872, | |
| "grad_norm": 0.9040001034736633, | |
| "learning_rate": 2.761674235170229e-07, | |
| "loss": 0.0703, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.7478051299707351, | |
| "grad_norm": 1.0969592332839966, | |
| "learning_rate": 2.7477139117951204e-07, | |
| "loss": 0.0712, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.7484937166465829, | |
| "grad_norm": 1.1855138540267944, | |
| "learning_rate": 2.7337810452843853e-07, | |
| "loss": 0.066, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 0.7491823033224307, | |
| "grad_norm": 0.7665496468544006, | |
| "learning_rate": 2.719875716136718e-07, | |
| "loss": 0.0608, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.7498708899982786, | |
| "grad_norm": 0.9379006028175354, | |
| "learning_rate": 2.7059980046917157e-07, | |
| "loss": 0.0515, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.7505594766741264, | |
| "grad_norm": 1.091181993484497, | |
| "learning_rate": 2.6921479911294004e-07, | |
| "loss": 0.0721, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.7512480633499742, | |
| "grad_norm": 1.3554106950759888, | |
| "learning_rate": 2.678325755469782e-07, | |
| "loss": 0.0529, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 0.751936650025822, | |
| "grad_norm": 2.0800187587738037, | |
| "learning_rate": 2.6645313775723735e-07, | |
| "loss": 0.0751, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.7526252367016698, | |
| "grad_norm": 1.056557297706604, | |
| "learning_rate": 2.6507649371357303e-07, | |
| "loss": 0.07, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 0.7533138233775176, | |
| "grad_norm": 0.9215115308761597, | |
| "learning_rate": 2.637026513697004e-07, | |
| "loss": 0.0576, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.7540024100533654, | |
| "grad_norm": 0.9977719187736511, | |
| "learning_rate": 2.623316186631474e-07, | |
| "loss": 0.0635, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.7546909967292132, | |
| "grad_norm": 0.5847204923629761, | |
| "learning_rate": 2.609634035152089e-07, | |
| "loss": 0.0572, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.7553795834050611, | |
| "grad_norm": 2.722033977508545, | |
| "learning_rate": 2.5959801383090053e-07, | |
| "loss": 0.0703, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 0.756068170080909, | |
| "grad_norm": 1.430222988128662, | |
| "learning_rate": 2.582354574989141e-07, | |
| "loss": 0.066, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 0.7567567567567568, | |
| "grad_norm": 0.8249462842941284, | |
| "learning_rate": 2.5687574239157133e-07, | |
| "loss": 0.0501, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 0.7574453434326046, | |
| "grad_norm": 0.6355162262916565, | |
| "learning_rate": 2.555188763647782e-07, | |
| "loss": 0.0596, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7581339301084524, | |
| "grad_norm": 0.6785207390785217, | |
| "learning_rate": 2.541648672579799e-07, | |
| "loss": 0.0486, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 0.7588225167843002, | |
| "grad_norm": 2.902787685394287, | |
| "learning_rate": 2.5281372289411586e-07, | |
| "loss": 0.0629, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 0.759511103460148, | |
| "grad_norm": 1.7893404960632324, | |
| "learning_rate": 2.5146545107957316e-07, | |
| "loss": 0.0669, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 0.7601996901359959, | |
| "grad_norm": 0.5831757187843323, | |
| "learning_rate": 2.5012005960414336e-07, | |
| "loss": 0.0523, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.7608882768118437, | |
| "grad_norm": 1.552027702331543, | |
| "learning_rate": 2.4877755624097596e-07, | |
| "loss": 0.0714, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.7615768634876915, | |
| "grad_norm": 1.1877832412719727, | |
| "learning_rate": 2.4743794874653473e-07, | |
| "loss": 0.0652, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 0.7622654501635393, | |
| "grad_norm": 0.6878598928451538, | |
| "learning_rate": 2.461012448605511e-07, | |
| "loss": 0.0624, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 0.7629540368393871, | |
| "grad_norm": 0.6071801781654358, | |
| "learning_rate": 2.447674523059818e-07, | |
| "loss": 0.0828, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 0.7636426235152349, | |
| "grad_norm": 1.2923734188079834, | |
| "learning_rate": 2.434365787889626e-07, | |
| "loss": 0.0633, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 0.7643312101910829, | |
| "grad_norm": 0.46784406900405884, | |
| "learning_rate": 2.421086319987636e-07, | |
| "loss": 0.0487, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.7650197968669307, | |
| "grad_norm": 0.752988874912262, | |
| "learning_rate": 2.407836196077461e-07, | |
| "loss": 0.0489, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 0.7657083835427785, | |
| "grad_norm": 0.5726896524429321, | |
| "learning_rate": 2.394615492713174e-07, | |
| "loss": 0.0519, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 0.7663969702186263, | |
| "grad_norm": 1.234387755393982, | |
| "learning_rate": 2.3814242862788708e-07, | |
| "loss": 0.0607, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 0.7670855568944741, | |
| "grad_norm": 0.7004895806312561, | |
| "learning_rate": 2.3682626529882154e-07, | |
| "loss": 0.0581, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 0.7677741435703219, | |
| "grad_norm": 1.1989398002624512, | |
| "learning_rate": 2.3551306688840193e-07, | |
| "loss": 0.0579, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.7684627302461697, | |
| "grad_norm": 1.3043203353881836, | |
| "learning_rate": 2.3420284098377892e-07, | |
| "loss": 0.0677, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.7691513169220175, | |
| "grad_norm": 1.6823842525482178, | |
| "learning_rate": 2.3289559515492922e-07, | |
| "loss": 0.0593, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 0.7698399035978654, | |
| "grad_norm": 1.1932395696640015, | |
| "learning_rate": 2.3159133695461185e-07, | |
| "loss": 0.0612, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 0.7705284902737132, | |
| "grad_norm": 0.7435174584388733, | |
| "learning_rate": 2.302900739183246e-07, | |
| "loss": 0.0445, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 0.771217076949561, | |
| "grad_norm": 1.0695222616195679, | |
| "learning_rate": 2.289918135642597e-07, | |
| "loss": 0.0762, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.7719056636254088, | |
| "grad_norm": 1.2553892135620117, | |
| "learning_rate": 2.276965633932619e-07, | |
| "loss": 0.0713, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 0.7725942503012567, | |
| "grad_norm": 1.7449697256088257, | |
| "learning_rate": 2.264043308887838e-07, | |
| "loss": 0.078, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 0.7732828369771045, | |
| "grad_norm": 1.0760855674743652, | |
| "learning_rate": 2.2511512351684363e-07, | |
| "loss": 0.0604, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 0.7739714236529524, | |
| "grad_norm": 1.5066118240356445, | |
| "learning_rate": 2.238289487259802e-07, | |
| "loss": 0.0619, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 0.7746600103288002, | |
| "grad_norm": 0.9582607746124268, | |
| "learning_rate": 2.2254581394721314e-07, | |
| "loss": 0.0732, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.775348597004648, | |
| "grad_norm": 0.8618611097335815, | |
| "learning_rate": 2.2126572659399702e-07, | |
| "loss": 0.0639, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 0.7760371836804958, | |
| "grad_norm": 0.8515923619270325, | |
| "learning_rate": 2.199886940621793e-07, | |
| "loss": 0.058, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 0.7767257703563436, | |
| "grad_norm": 1.9144634008407593, | |
| "learning_rate": 2.187147237299584e-07, | |
| "loss": 0.0687, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 0.7774143570321914, | |
| "grad_norm": 1.0546441078186035, | |
| "learning_rate": 2.1744382295784051e-07, | |
| "loss": 0.0578, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 0.7781029437080392, | |
| "grad_norm": 0.9172005653381348, | |
| "learning_rate": 2.1617599908859704e-07, | |
| "loss": 0.0703, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.778791530383887, | |
| "grad_norm": 0.8583981394767761, | |
| "learning_rate": 2.149112594472219e-07, | |
| "loss": 0.0489, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 0.7794801170597349, | |
| "grad_norm": 0.8677762746810913, | |
| "learning_rate": 2.1364961134088974e-07, | |
| "loss": 0.0608, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 0.7801687037355827, | |
| "grad_norm": 0.9285558462142944, | |
| "learning_rate": 2.1239106205891434e-07, | |
| "loss": 0.0645, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 0.7808572904114306, | |
| "grad_norm": 1.6503733396530151, | |
| "learning_rate": 2.1113561887270414e-07, | |
| "loss": 0.0916, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 0.7815458770872784, | |
| "grad_norm": 1.0598102807998657, | |
| "learning_rate": 2.0988328903572294e-07, | |
| "loss": 0.0668, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.7822344637631262, | |
| "grad_norm": 0.985866129398346, | |
| "learning_rate": 2.0863407978344646e-07, | |
| "loss": 0.0678, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 0.782923050438974, | |
| "grad_norm": 1.042116403579712, | |
| "learning_rate": 2.073879983333206e-07, | |
| "loss": 0.0666, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 0.7836116371148218, | |
| "grad_norm": 1.3594166040420532, | |
| "learning_rate": 2.0614505188472043e-07, | |
| "loss": 0.0623, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 0.7843002237906697, | |
| "grad_norm": 0.9141592383384705, | |
| "learning_rate": 2.049052476189081e-07, | |
| "loss": 0.0708, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 0.7849888104665175, | |
| "grad_norm": 0.7261631488800049, | |
| "learning_rate": 2.0366859269899139e-07, | |
| "loss": 0.0665, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.7856773971423653, | |
| "grad_norm": 0.7423864603042603, | |
| "learning_rate": 2.024350942698824e-07, | |
| "loss": 0.0591, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 0.7863659838182131, | |
| "grad_norm": 0.8459689617156982, | |
| "learning_rate": 2.0120475945825635e-07, | |
| "loss": 0.0592, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 0.7870545704940609, | |
| "grad_norm": 0.6239163279533386, | |
| "learning_rate": 1.9997759537251027e-07, | |
| "loss": 0.0409, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 0.7877431571699087, | |
| "grad_norm": 0.9537420272827148, | |
| "learning_rate": 1.987536091027217e-07, | |
| "loss": 0.0681, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 0.7884317438457565, | |
| "grad_norm": 0.5007339715957642, | |
| "learning_rate": 1.9753280772060819e-07, | |
| "loss": 0.0517, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.7891203305216044, | |
| "grad_norm": 1.2504444122314453, | |
| "learning_rate": 1.9631519827948646e-07, | |
| "loss": 0.0663, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 0.7898089171974523, | |
| "grad_norm": 1.1952650547027588, | |
| "learning_rate": 1.9510078781423127e-07, | |
| "loss": 0.0546, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 0.7904975038733001, | |
| "grad_norm": 0.7635645866394043, | |
| "learning_rate": 1.9388958334123465e-07, | |
| "loss": 0.0533, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 0.7911860905491479, | |
| "grad_norm": 0.7804190516471863, | |
| "learning_rate": 1.9268159185836575e-07, | |
| "loss": 0.0594, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 0.7918746772249957, | |
| "grad_norm": 0.6885479688644409, | |
| "learning_rate": 1.914768203449312e-07, | |
| "loss": 0.063, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.7925632639008435, | |
| "grad_norm": 0.667660653591156, | |
| "learning_rate": 1.9027527576163248e-07, | |
| "loss": 0.07, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 0.7932518505766913, | |
| "grad_norm": 1.9460809230804443, | |
| "learning_rate": 1.8907696505052805e-07, | |
| "loss": 0.0777, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 0.7939404372525392, | |
| "grad_norm": 1.4023791551589966, | |
| "learning_rate": 1.8788189513499225e-07, | |
| "loss": 0.0607, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 0.794629023928387, | |
| "grad_norm": 1.2282487154006958, | |
| "learning_rate": 1.8669007291967474e-07, | |
| "loss": 0.058, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 0.7953176106042348, | |
| "grad_norm": 2.0234920978546143, | |
| "learning_rate": 1.855015052904619e-07, | |
| "loss": 0.0718, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.7960061972800826, | |
| "grad_norm": 0.5690957903862, | |
| "learning_rate": 1.843161991144361e-07, | |
| "loss": 0.0482, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 0.7966947839559304, | |
| "grad_norm": 0.6931769847869873, | |
| "learning_rate": 1.8313416123983648e-07, | |
| "loss": 0.0675, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 0.7973833706317782, | |
| "grad_norm": 1.6891932487487793, | |
| "learning_rate": 1.8195539849601882e-07, | |
| "loss": 0.066, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 0.7980719573076261, | |
| "grad_norm": 0.6087167263031006, | |
| "learning_rate": 1.807799176934169e-07, | |
| "loss": 0.0716, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 0.798760543983474, | |
| "grad_norm": 0.9160038828849792, | |
| "learning_rate": 1.7960772562350263e-07, | |
| "loss": 0.0565, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.7994491306593218, | |
| "grad_norm": 0.6283268332481384, | |
| "learning_rate": 1.7843882905874638e-07, | |
| "loss": 0.0628, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 0.8001377173351696, | |
| "grad_norm": 1.298898696899414, | |
| "learning_rate": 1.7727323475257906e-07, | |
| "loss": 0.0732, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 0.8008263040110174, | |
| "grad_norm": 0.9403923153877258, | |
| "learning_rate": 1.7611094943935205e-07, | |
| "loss": 0.0623, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 0.8015148906868652, | |
| "grad_norm": 0.912321925163269, | |
| "learning_rate": 1.7495197983429907e-07, | |
| "loss": 0.0477, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 0.802203477362713, | |
| "grad_norm": 0.926247239112854, | |
| "learning_rate": 1.7379633263349588e-07, | |
| "loss": 0.0648, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.8028920640385608, | |
| "grad_norm": 0.9041908979415894, | |
| "learning_rate": 1.726440145138241e-07, | |
| "loss": 0.0583, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 0.8035806507144087, | |
| "grad_norm": 0.9324719905853271, | |
| "learning_rate": 1.7149503213293054e-07, | |
| "loss": 0.0636, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 0.8042692373902565, | |
| "grad_norm": 1.0776565074920654, | |
| "learning_rate": 1.7034939212918905e-07, | |
| "loss": 0.0544, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 0.8049578240661043, | |
| "grad_norm": 0.5352507829666138, | |
| "learning_rate": 1.6920710112166303e-07, | |
| "loss": 0.0519, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 0.8056464107419521, | |
| "grad_norm": 0.7596953511238098, | |
| "learning_rate": 1.6806816571006675e-07, | |
| "loss": 0.0517, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.8063349974178, | |
| "grad_norm": 0.8265296816825867, | |
| "learning_rate": 1.6693259247472645e-07, | |
| "loss": 0.0579, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 0.8070235840936478, | |
| "grad_norm": 1.5133427381515503, | |
| "learning_rate": 1.6580038797654388e-07, | |
| "loss": 0.0628, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 0.8077121707694956, | |
| "grad_norm": 0.9565699696540833, | |
| "learning_rate": 1.6467155875695691e-07, | |
| "loss": 0.0545, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 0.8084007574453435, | |
| "grad_norm": 1.4020029306411743, | |
| "learning_rate": 1.635461113379028e-07, | |
| "loss": 0.0784, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 0.8090893441211913, | |
| "grad_norm": 1.5086497068405151, | |
| "learning_rate": 1.6242405222177986e-07, | |
| "loss": 0.0739, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.8097779307970391, | |
| "grad_norm": 0.9495487809181213, | |
| "learning_rate": 1.613053878914101e-07, | |
| "loss": 0.0646, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 0.8104665174728869, | |
| "grad_norm": 0.8247946500778198, | |
| "learning_rate": 1.6019012481000194e-07, | |
| "loss": 0.059, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 0.8111551041487347, | |
| "grad_norm": 1.0125800371170044, | |
| "learning_rate": 1.5907826942111224e-07, | |
| "loss": 0.0424, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 0.8118436908245825, | |
| "grad_norm": 1.3568590879440308, | |
| "learning_rate": 1.5796982814861001e-07, | |
| "loss": 0.0635, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 0.8125322775004303, | |
| "grad_norm": 0.7913120985031128, | |
| "learning_rate": 1.5686480739663873e-07, | |
| "loss": 0.0481, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.8132208641762781, | |
| "grad_norm": 1.2215467691421509, | |
| "learning_rate": 1.557632135495795e-07, | |
| "loss": 0.0854, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 0.813909450852126, | |
| "grad_norm": 0.7948154807090759, | |
| "learning_rate": 1.5466505297201342e-07, | |
| "loss": 0.0718, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 0.8145980375279739, | |
| "grad_norm": 1.113890528678894, | |
| "learning_rate": 1.5357033200868667e-07, | |
| "loss": 0.0598, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 0.8152866242038217, | |
| "grad_norm": 1.6052463054656982, | |
| "learning_rate": 1.5247905698447205e-07, | |
| "loss": 0.0707, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 0.8159752108796695, | |
| "grad_norm": 0.8825914263725281, | |
| "learning_rate": 1.5139123420433265e-07, | |
| "loss": 0.0652, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.8166637975555173, | |
| "grad_norm": 0.9417052268981934, | |
| "learning_rate": 1.5030686995328635e-07, | |
| "loss": 0.0621, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 0.8173523842313651, | |
| "grad_norm": 2.241471529006958, | |
| "learning_rate": 1.4922597049636923e-07, | |
| "loss": 0.058, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 0.818040970907213, | |
| "grad_norm": 0.6243368983268738, | |
| "learning_rate": 1.4814854207859832e-07, | |
| "loss": 0.0556, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 0.8187295575830608, | |
| "grad_norm": 3.0077273845672607, | |
| "learning_rate": 1.4707459092493706e-07, | |
| "loss": 0.0553, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 0.8194181442589086, | |
| "grad_norm": 0.6388275027275085, | |
| "learning_rate": 1.460041232402582e-07, | |
| "loss": 0.0678, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.8201067309347564, | |
| "grad_norm": 1.3966974020004272, | |
| "learning_rate": 1.449371452093087e-07, | |
| "loss": 0.0604, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 0.8207953176106042, | |
| "grad_norm": 0.6352896690368652, | |
| "learning_rate": 1.4387366299667333e-07, | |
| "loss": 0.0543, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 0.821483904286452, | |
| "grad_norm": 0.7193229794502258, | |
| "learning_rate": 1.4281368274673965e-07, | |
| "loss": 0.0505, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 0.8221724909622998, | |
| "grad_norm": 1.2771050930023193, | |
| "learning_rate": 1.4175721058366212e-07, | |
| "loss": 0.0717, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 0.8228610776381478, | |
| "grad_norm": 1.4145382642745972, | |
| "learning_rate": 1.4070425261132658e-07, | |
| "loss": 0.0759, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.8235496643139956, | |
| "grad_norm": 0.8068073987960815, | |
| "learning_rate": 1.3965481491331558e-07, | |
| "loss": 0.0597, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 0.8242382509898434, | |
| "grad_norm": 1.2401487827301025, | |
| "learning_rate": 1.3860890355287287e-07, | |
| "loss": 0.0836, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 0.8249268376656912, | |
| "grad_norm": 0.6385102868080139, | |
| "learning_rate": 1.375665245728684e-07, | |
| "loss": 0.0518, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 0.825615424341539, | |
| "grad_norm": 0.7058199048042297, | |
| "learning_rate": 1.3652768399576275e-07, | |
| "loss": 0.0758, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 0.8263040110173868, | |
| "grad_norm": 1.0258511304855347, | |
| "learning_rate": 1.354923878235743e-07, | |
| "loss": 0.0475, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.8263040110173868, | |
| "eval_loss": NaN, | |
| "eval_runtime": 65.7672, | |
| "eval_samples_per_second": 14.293, | |
| "eval_steps_per_second": 0.456, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.8269925976932346, | |
| "grad_norm": 1.4682613611221313, | |
| "learning_rate": 1.3446064203784223e-07, | |
| "loss": 0.072, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 0.8276811843690824, | |
| "grad_norm": 1.379334807395935, | |
| "learning_rate": 1.3343245259959292e-07, | |
| "loss": 0.0602, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 0.8283697710449303, | |
| "grad_norm": 0.716903567314148, | |
| "learning_rate": 1.3240782544930604e-07, | |
| "loss": 0.064, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 0.8290583577207781, | |
| "grad_norm": 0.719710648059845, | |
| "learning_rate": 1.3138676650687977e-07, | |
| "loss": 0.0499, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 0.8297469443966259, | |
| "grad_norm": 1.8048508167266846, | |
| "learning_rate": 1.3036928167159591e-07, | |
| "loss": 0.0619, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.8304355310724737, | |
| "grad_norm": 1.060123085975647, | |
| "learning_rate": 1.2935537682208706e-07, | |
| "loss": 0.0625, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 0.8311241177483216, | |
| "grad_norm": 0.9757376909255981, | |
| "learning_rate": 1.2834505781630198e-07, | |
| "loss": 0.0594, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 0.8318127044241694, | |
| "grad_norm": 4.1545610427856445, | |
| "learning_rate": 1.2733833049147165e-07, | |
| "loss": 0.0578, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 0.8325012911000172, | |
| "grad_norm": 0.7732447981834412, | |
| "learning_rate": 1.2633520066407583e-07, | |
| "loss": 0.0604, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 0.8331898777758651, | |
| "grad_norm": 1.139441967010498, | |
| "learning_rate": 1.2533567412980914e-07, | |
| "loss": 0.0665, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.8338784644517129, | |
| "grad_norm": 0.9618107676506042, | |
| "learning_rate": 1.2433975666354833e-07, | |
| "loss": 0.0614, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 0.8345670511275607, | |
| "grad_norm": 1.0712355375289917, | |
| "learning_rate": 1.2334745401931746e-07, | |
| "loss": 0.0605, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 0.8352556378034085, | |
| "grad_norm": 2.2155399322509766, | |
| "learning_rate": 1.223587719302562e-07, | |
| "loss": 0.0834, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 0.8359442244792563, | |
| "grad_norm": 1.6424671411514282, | |
| "learning_rate": 1.2137371610858585e-07, | |
| "loss": 0.0664, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 0.8366328111551041, | |
| "grad_norm": 0.849981427192688, | |
| "learning_rate": 1.2039229224557694e-07, | |
| "loss": 0.0587, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.8373213978309519, | |
| "grad_norm": 0.9269185662269592, | |
| "learning_rate": 1.194145060115149e-07, | |
| "loss": 0.0531, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 0.8380099845067998, | |
| "grad_norm": 1.2190477848052979, | |
| "learning_rate": 1.1844036305566975e-07, | |
| "loss": 0.0766, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 0.8386985711826476, | |
| "grad_norm": 1.1311004161834717, | |
| "learning_rate": 1.1746986900626117e-07, | |
| "loss": 0.0551, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 0.8393871578584955, | |
| "grad_norm": 1.0408507585525513, | |
| "learning_rate": 1.1650302947042685e-07, | |
| "loss": 0.0551, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 0.8400757445343433, | |
| "grad_norm": 1.3155499696731567, | |
| "learning_rate": 1.1553985003419051e-07, | |
| "loss": 0.057, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.8407643312101911, | |
| "grad_norm": 0.6170788407325745, | |
| "learning_rate": 1.1458033626242914e-07, | |
| "loss": 0.0643, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 0.8414529178860389, | |
| "grad_norm": 1.4344539642333984, | |
| "learning_rate": 1.1362449369884064e-07, | |
| "loss": 0.0782, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 0.8421415045618867, | |
| "grad_norm": 1.439078688621521, | |
| "learning_rate": 1.126723278659125e-07, | |
| "loss": 0.0545, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 0.8428300912377346, | |
| "grad_norm": 0.7359664440155029, | |
| "learning_rate": 1.1172384426488916e-07, | |
| "loss": 0.062, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 0.8435186779135824, | |
| "grad_norm": 0.4754618704319, | |
| "learning_rate": 1.1077904837574086e-07, | |
| "loss": 0.0554, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.8442072645894302, | |
| "grad_norm": 0.9606807827949524, | |
| "learning_rate": 1.0983794565713152e-07, | |
| "loss": 0.0576, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 0.844895851265278, | |
| "grad_norm": 0.5315238237380981, | |
| "learning_rate": 1.0890054154638715e-07, | |
| "loss": 0.0492, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 0.8455844379411258, | |
| "grad_norm": 0.6970733404159546, | |
| "learning_rate": 1.0796684145946528e-07, | |
| "loss": 0.0641, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 0.8462730246169736, | |
| "grad_norm": 1.3414822816848755, | |
| "learning_rate": 1.0703685079092204e-07, | |
| "loss": 0.0562, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 0.8469616112928214, | |
| "grad_norm": 1.322902798652649, | |
| "learning_rate": 1.0611057491388279e-07, | |
| "loss": 0.0631, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.8476501979686694, | |
| "grad_norm": 0.8792802691459656, | |
| "learning_rate": 1.0518801918001006e-07, | |
| "loss": 0.052, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 0.8483387846445172, | |
| "grad_norm": 0.8927648663520813, | |
| "learning_rate": 1.0426918891947265e-07, | |
| "loss": 0.055, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 0.849027371320365, | |
| "grad_norm": 0.6912924647331238, | |
| "learning_rate": 1.0335408944091526e-07, | |
| "loss": 0.0679, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 0.8497159579962128, | |
| "grad_norm": 0.9239325523376465, | |
| "learning_rate": 1.0244272603142741e-07, | |
| "loss": 0.0578, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 0.8504045446720606, | |
| "grad_norm": 0.6665210127830505, | |
| "learning_rate": 1.0153510395651327e-07, | |
| "loss": 0.0604, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.8510931313479084, | |
| "grad_norm": 0.656692385673523, | |
| "learning_rate": 1.0063122846006061e-07, | |
| "loss": 0.0581, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 0.8517817180237562, | |
| "grad_norm": 0.9401907920837402, | |
| "learning_rate": 9.973110476431141e-08, | |
| "loss": 0.0487, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 0.852470304699604, | |
| "grad_norm": 0.7068642973899841, | |
| "learning_rate": 9.883473806983087e-08, | |
| "loss": 0.0594, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 0.8531588913754519, | |
| "grad_norm": 1.5799263715744019, | |
| "learning_rate": 9.79421335554781e-08, | |
| "loss": 0.072, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 0.8538474780512997, | |
| "grad_norm": 0.9455007910728455, | |
| "learning_rate": 9.705329637837506e-08, | |
| "loss": 0.0677, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.8545360647271475, | |
| "grad_norm": 1.209568977355957, | |
| "learning_rate": 9.616823167387826e-08, | |
| "loss": 0.1128, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 0.8552246514029953, | |
| "grad_norm": 1.2844939231872559, | |
| "learning_rate": 9.52869445555479e-08, | |
| "loss": 0.0685, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 0.8559132380788431, | |
| "grad_norm": 0.6665671467781067, | |
| "learning_rate": 9.440944011511887e-08, | |
| "loss": 0.0932, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 0.856601824754691, | |
| "grad_norm": 1.0164397954940796, | |
| "learning_rate": 9.353572342247135e-08, | |
| "loss": 0.0535, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 0.8572904114305389, | |
| "grad_norm": 1.0427037477493286, | |
| "learning_rate": 9.266579952560119e-08, | |
| "loss": 0.051, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.8579789981063867, | |
| "grad_norm": 0.5913142561912537, | |
| "learning_rate": 9.179967345059095e-08, | |
| "loss": 0.0519, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 0.8586675847822345, | |
| "grad_norm": 0.4813862144947052, | |
| "learning_rate": 9.09373502015808e-08, | |
| "loss": 0.0429, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 0.8593561714580823, | |
| "grad_norm": 1.7151315212249756, | |
| "learning_rate": 9.007883476073985e-08, | |
| "loss": 0.067, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 0.8600447581339301, | |
| "grad_norm": 2.2725136280059814, | |
| "learning_rate": 8.92241320882372e-08, | |
| "loss": 0.0644, | |
| "step": 1249 | |
| }, | |
| { | |
| "epoch": 0.8607333448097779, | |
| "grad_norm": 0.8944635987281799, | |
| "learning_rate": 8.837324712221292e-08, | |
| "loss": 0.0627, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.8614219314856257, | |
| "grad_norm": 0.9078942537307739, | |
| "learning_rate": 8.752618477875021e-08, | |
| "loss": 0.06, | |
| "step": 1251 | |
| }, | |
| { | |
| "epoch": 0.8621105181614735, | |
| "grad_norm": 0.8876720070838928, | |
| "learning_rate": 8.668294995184653e-08, | |
| "loss": 0.0669, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 0.8627991048373214, | |
| "grad_norm": 0.6233602166175842, | |
| "learning_rate": 8.584354751338528e-08, | |
| "loss": 0.061, | |
| "step": 1253 | |
| }, | |
| { | |
| "epoch": 0.8634876915131692, | |
| "grad_norm": 0.5211963057518005, | |
| "learning_rate": 8.500798231310789e-08, | |
| "loss": 0.0484, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 0.864176278189017, | |
| "grad_norm": 0.8618441224098206, | |
| "learning_rate": 8.417625917858587e-08, | |
| "loss": 0.0647, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.8648648648648649, | |
| "grad_norm": 1.8776743412017822, | |
| "learning_rate": 8.334838291519268e-08, | |
| "loss": 0.0674, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 0.8655534515407127, | |
| "grad_norm": 0.6075620651245117, | |
| "learning_rate": 8.25243583060759e-08, | |
| "loss": 0.0549, | |
| "step": 1257 | |
| }, | |
| { | |
| "epoch": 0.8662420382165605, | |
| "grad_norm": 0.7557206153869629, | |
| "learning_rate": 8.170419011212976e-08, | |
| "loss": 0.0556, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 0.8669306248924084, | |
| "grad_norm": 1.145349144935608, | |
| "learning_rate": 8.088788307196814e-08, | |
| "loss": 0.0525, | |
| "step": 1259 | |
| }, | |
| { | |
| "epoch": 0.8676192115682562, | |
| "grad_norm": 0.7890598773956299, | |
| "learning_rate": 8.007544190189591e-08, | |
| "loss": 0.0593, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.868307798244104, | |
| "grad_norm": 2.0478196144104004, | |
| "learning_rate": 7.926687129588286e-08, | |
| "loss": 0.0626, | |
| "step": 1261 | |
| }, | |
| { | |
| "epoch": 0.8689963849199518, | |
| "grad_norm": 1.6047842502593994, | |
| "learning_rate": 7.846217592553626e-08, | |
| "loss": 0.0567, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 0.8696849715957996, | |
| "grad_norm": 0.5736270546913147, | |
| "learning_rate": 7.766136044007335e-08, | |
| "loss": 0.0539, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 0.8703735582716474, | |
| "grad_norm": 1.2750167846679688, | |
| "learning_rate": 7.686442946629534e-08, | |
| "loss": 0.0782, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 0.8710621449474952, | |
| "grad_norm": 2.8505361080169678, | |
| "learning_rate": 7.607138760855997e-08, | |
| "loss": 0.0947, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.871750731623343, | |
| "grad_norm": 2.215938091278076, | |
| "learning_rate": 7.528223944875535e-08, | |
| "loss": 0.0713, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 0.8724393182991909, | |
| "grad_norm": 0.8500781059265137, | |
| "learning_rate": 7.449698954627316e-08, | |
| "loss": 0.0595, | |
| "step": 1267 | |
| }, | |
| { | |
| "epoch": 0.8731279049750388, | |
| "grad_norm": 1.5429556369781494, | |
| "learning_rate": 7.371564243798253e-08, | |
| "loss": 0.0838, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 0.8738164916508866, | |
| "grad_norm": 1.1179783344268799, | |
| "learning_rate": 7.2938202638204e-08, | |
| "loss": 0.0443, | |
| "step": 1269 | |
| }, | |
| { | |
| "epoch": 0.8745050783267344, | |
| "grad_norm": 0.7166243195533752, | |
| "learning_rate": 7.216467463868267e-08, | |
| "loss": 0.0502, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.8751936650025822, | |
| "grad_norm": 0.5423519015312195, | |
| "learning_rate": 7.139506290856313e-08, | |
| "loss": 0.0577, | |
| "step": 1271 | |
| }, | |
| { | |
| "epoch": 0.87588225167843, | |
| "grad_norm": 1.6516828536987305, | |
| "learning_rate": 7.062937189436316e-08, | |
| "loss": 0.0641, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 0.8765708383542778, | |
| "grad_norm": 0.5345444083213806, | |
| "learning_rate": 6.986760601994857e-08, | |
| "loss": 0.059, | |
| "step": 1273 | |
| }, | |
| { | |
| "epoch": 0.8772594250301257, | |
| "grad_norm": 0.5780264735221863, | |
| "learning_rate": 6.910976968650634e-08, | |
| "loss": 0.0716, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 0.8779480117059735, | |
| "grad_norm": 0.6236739158630371, | |
| "learning_rate": 6.835586727252114e-08, | |
| "loss": 0.0586, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.8786365983818213, | |
| "grad_norm": 0.6632158756256104, | |
| "learning_rate": 6.760590313374836e-08, | |
| "loss": 0.0609, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 0.8793251850576691, | |
| "grad_norm": 0.906181812286377, | |
| "learning_rate": 6.685988160318959e-08, | |
| "loss": 0.0598, | |
| "step": 1277 | |
| }, | |
| { | |
| "epoch": 0.8800137717335169, | |
| "grad_norm": 0.6677340865135193, | |
| "learning_rate": 6.611780699106765e-08, | |
| "loss": 0.0482, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 0.8807023584093647, | |
| "grad_norm": 1.232237696647644, | |
| "learning_rate": 6.537968358480179e-08, | |
| "loss": 0.0646, | |
| "step": 1279 | |
| }, | |
| { | |
| "epoch": 0.8813909450852127, | |
| "grad_norm": 0.5725114941596985, | |
| "learning_rate": 6.464551564898227e-08, | |
| "loss": 0.0542, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.8820795317610605, | |
| "grad_norm": 1.2437443733215332, | |
| "learning_rate": 6.391530742534657e-08, | |
| "loss": 0.0562, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 0.8827681184369083, | |
| "grad_norm": 1.6437773704528809, | |
| "learning_rate": 6.318906313275435e-08, | |
| "loss": 0.0746, | |
| "step": 1282 | |
| }, | |
| { | |
| "epoch": 0.8834567051127561, | |
| "grad_norm": 1.1963529586791992, | |
| "learning_rate": 6.246678696716337e-08, | |
| "loss": 0.0703, | |
| "step": 1283 | |
| }, | |
| { | |
| "epoch": 0.8841452917886039, | |
| "grad_norm": 0.7588176131248474, | |
| "learning_rate": 6.17484831016048e-08, | |
| "loss": 0.0758, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 0.8848338784644517, | |
| "grad_norm": 0.8116660118103027, | |
| "learning_rate": 6.103415568615959e-08, | |
| "loss": 0.0509, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.8855224651402995, | |
| "grad_norm": 0.6952071189880371, | |
| "learning_rate": 6.032380884793445e-08, | |
| "loss": 0.0441, | |
| "step": 1286 | |
| }, | |
| { | |
| "epoch": 0.8862110518161473, | |
| "grad_norm": 0.81709885597229, | |
| "learning_rate": 5.961744669103739e-08, | |
| "loss": 0.0623, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 0.8868996384919952, | |
| "grad_norm": 0.9331743717193604, | |
| "learning_rate": 5.891507329655493e-08, | |
| "loss": 0.0477, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 0.887588225167843, | |
| "grad_norm": 0.852997899055481, | |
| "learning_rate": 5.821669272252792e-08, | |
| "loss": 0.0728, | |
| "step": 1289 | |
| }, | |
| { | |
| "epoch": 0.8882768118436908, | |
| "grad_norm": 0.9943188428878784, | |
| "learning_rate": 5.752230900392827e-08, | |
| "loss": 0.0654, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.8889653985195386, | |
| "grad_norm": 0.9971889853477478, | |
| "learning_rate": 5.683192615263533e-08, | |
| "loss": 0.0452, | |
| "step": 1291 | |
| }, | |
| { | |
| "epoch": 0.8896539851953865, | |
| "grad_norm": 1.0278383493423462, | |
| "learning_rate": 5.614554815741349e-08, | |
| "loss": 0.0663, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 0.8903425718712343, | |
| "grad_norm": 0.996644914150238, | |
| "learning_rate": 5.546317898388842e-08, | |
| "loss": 0.0573, | |
| "step": 1293 | |
| }, | |
| { | |
| "epoch": 0.8910311585470821, | |
| "grad_norm": 0.6578953266143799, | |
| "learning_rate": 5.478482257452414e-08, | |
| "loss": 0.0519, | |
| "step": 1294 | |
| }, | |
| { | |
| "epoch": 0.89171974522293, | |
| "grad_norm": 1.1061625480651855, | |
| "learning_rate": 5.411048284860071e-08, | |
| "loss": 0.0679, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.8924083318987778, | |
| "grad_norm": 1.2438205480575562, | |
| "learning_rate": 5.3440163702191514e-08, | |
| "loss": 0.0535, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 0.8930969185746256, | |
| "grad_norm": 1.1850031614303589, | |
| "learning_rate": 5.27738690081401e-08, | |
| "loss": 0.0579, | |
| "step": 1297 | |
| }, | |
| { | |
| "epoch": 0.8937855052504734, | |
| "grad_norm": 1.737141489982605, | |
| "learning_rate": 5.2111602616038787e-08, | |
| "loss": 0.0623, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 0.8944740919263212, | |
| "grad_norm": 1.0074365139007568, | |
| "learning_rate": 5.145336835220557e-08, | |
| "loss": 0.0715, | |
| "step": 1299 | |
| }, | |
| { | |
| "epoch": 0.895162678602169, | |
| "grad_norm": 0.7543210983276367, | |
| "learning_rate": 5.07991700196627e-08, | |
| "loss": 0.0564, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.8958512652780168, | |
| "grad_norm": 1.1668027639389038, | |
| "learning_rate": 5.0149011398114036e-08, | |
| "loss": 0.059, | |
| "step": 1301 | |
| }, | |
| { | |
| "epoch": 0.8965398519538647, | |
| "grad_norm": 1.0464317798614502, | |
| "learning_rate": 4.95028962439239e-08, | |
| "loss": 0.0682, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 0.8972284386297125, | |
| "grad_norm": 1.7445652484893799, | |
| "learning_rate": 4.886082829009486e-08, | |
| "loss": 0.0597, | |
| "step": 1303 | |
| }, | |
| { | |
| "epoch": 0.8979170253055604, | |
| "grad_norm": 0.8748214840888977, | |
| "learning_rate": 4.8222811246246113e-08, | |
| "loss": 0.0587, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 0.8986056119814082, | |
| "grad_norm": 1.7114619016647339, | |
| "learning_rate": 4.758884879859257e-08, | |
| "loss": 0.0814, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.899294198657256, | |
| "grad_norm": 0.62885582447052, | |
| "learning_rate": 4.695894460992331e-08, | |
| "loss": 0.0463, | |
| "step": 1306 | |
| }, | |
| { | |
| "epoch": 0.8999827853331038, | |
| "grad_norm": 1.0485039949417114, | |
| "learning_rate": 4.633310231958016e-08, | |
| "loss": 0.0537, | |
| "step": 1307 | |
| }, | |
| { | |
| "epoch": 0.9006713720089516, | |
| "grad_norm": 0.8217902183532715, | |
| "learning_rate": 4.57113255434368e-08, | |
| "loss": 0.0702, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 0.9013599586847995, | |
| "grad_norm": 1.468031406402588, | |
| "learning_rate": 4.509361787387828e-08, | |
| "loss": 0.0606, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 0.9020485453606473, | |
| "grad_norm": 0.8831107020378113, | |
| "learning_rate": 4.447998287977981e-08, | |
| "loss": 0.0656, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.9027371320364951, | |
| "grad_norm": 1.1209571361541748, | |
| "learning_rate": 4.3870424106485896e-08, | |
| "loss": 0.0553, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 0.9034257187123429, | |
| "grad_norm": 0.8163548111915588, | |
| "learning_rate": 4.326494507579074e-08, | |
| "loss": 0.06, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 0.9041143053881907, | |
| "grad_norm": 0.6159918308258057, | |
| "learning_rate": 4.2663549285917284e-08, | |
| "loss": 0.0488, | |
| "step": 1313 | |
| }, | |
| { | |
| "epoch": 0.9048028920640385, | |
| "grad_norm": 0.9278672337532043, | |
| "learning_rate": 4.2066240211496634e-08, | |
| "loss": 0.0582, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 0.9054914787398863, | |
| "grad_norm": 0.6402984261512756, | |
| "learning_rate": 4.1473021303548994e-08, | |
| "loss": 0.0411, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.9061800654157343, | |
| "grad_norm": 1.0502867698669434, | |
| "learning_rate": 4.088389598946291e-08, | |
| "loss": 0.0794, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 0.9068686520915821, | |
| "grad_norm": 0.5800631642341614, | |
| "learning_rate": 4.029886767297567e-08, | |
| "loss": 0.0515, | |
| "step": 1317 | |
| }, | |
| { | |
| "epoch": 0.9075572387674299, | |
| "grad_norm": 1.1622531414031982, | |
| "learning_rate": 3.9717939734154e-08, | |
| "loss": 0.0556, | |
| "step": 1318 | |
| }, | |
| { | |
| "epoch": 0.9082458254432777, | |
| "grad_norm": 0.5939826369285583, | |
| "learning_rate": 3.914111552937383e-08, | |
| "loss": 0.0526, | |
| "step": 1319 | |
| }, | |
| { | |
| "epoch": 0.9089344121191255, | |
| "grad_norm": 0.6807751059532166, | |
| "learning_rate": 3.85683983913018e-08, | |
| "loss": 0.053, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.9096229987949733, | |
| "grad_norm": 1.1925545930862427, | |
| "learning_rate": 3.799979162887487e-08, | |
| "loss": 0.0756, | |
| "step": 1321 | |
| }, | |
| { | |
| "epoch": 0.9103115854708211, | |
| "grad_norm": 0.7052732110023499, | |
| "learning_rate": 3.7435298527282417e-08, | |
| "loss": 0.0475, | |
| "step": 1322 | |
| }, | |
| { | |
| "epoch": 0.911000172146669, | |
| "grad_norm": 0.7789724469184875, | |
| "learning_rate": 3.68749223479464e-08, | |
| "loss": 0.0672, | |
| "step": 1323 | |
| }, | |
| { | |
| "epoch": 0.9116887588225168, | |
| "grad_norm": 1.242823600769043, | |
| "learning_rate": 3.6318666328503056e-08, | |
| "loss": 0.0623, | |
| "step": 1324 | |
| }, | |
| { | |
| "epoch": 0.9123773454983646, | |
| "grad_norm": 1.072239875793457, | |
| "learning_rate": 3.5766533682783265e-08, | |
| "loss": 0.0516, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.9130659321742124, | |
| "grad_norm": 0.794519305229187, | |
| "learning_rate": 3.521852760079569e-08, | |
| "loss": 0.0809, | |
| "step": 1326 | |
| }, | |
| { | |
| "epoch": 0.9137545188500602, | |
| "grad_norm": 0.6489912271499634, | |
| "learning_rate": 3.4674651248706584e-08, | |
| "loss": 0.0572, | |
| "step": 1327 | |
| }, | |
| { | |
| "epoch": 0.9144431055259081, | |
| "grad_norm": 0.526973307132721, | |
| "learning_rate": 3.4134907768822274e-08, | |
| "loss": 0.0526, | |
| "step": 1328 | |
| }, | |
| { | |
| "epoch": 0.9151316922017559, | |
| "grad_norm": 0.7113784551620483, | |
| "learning_rate": 3.359930027957112e-08, | |
| "loss": 0.0561, | |
| "step": 1329 | |
| }, | |
| { | |
| "epoch": 0.9158202788776038, | |
| "grad_norm": 0.7585306167602539, | |
| "learning_rate": 3.30678318754854e-08, | |
| "loss": 0.0612, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.9165088655534516, | |
| "grad_norm": 0.8375997543334961, | |
| "learning_rate": 3.2540505627183135e-08, | |
| "loss": 0.0606, | |
| "step": 1331 | |
| }, | |
| { | |
| "epoch": 0.9171974522292994, | |
| "grad_norm": 0.7716051936149597, | |
| "learning_rate": 3.201732458135073e-08, | |
| "loss": 0.0551, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 0.9178860389051472, | |
| "grad_norm": 0.9730448126792908, | |
| "learning_rate": 3.1498291760725245e-08, | |
| "loss": 0.0574, | |
| "step": 1333 | |
| }, | |
| { | |
| "epoch": 0.918574625580995, | |
| "grad_norm": 1.3575196266174316, | |
| "learning_rate": 3.098341016407674e-08, | |
| "loss": 0.0648, | |
| "step": 1334 | |
| }, | |
| { | |
| "epoch": 0.9192632122568428, | |
| "grad_norm": 0.8249568939208984, | |
| "learning_rate": 3.047268276619139e-08, | |
| "loss": 0.0684, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.9199517989326906, | |
| "grad_norm": 0.7877727746963501, | |
| "learning_rate": 2.9966112517853724e-08, | |
| "loss": 0.0679, | |
| "step": 1336 | |
| }, | |
| { | |
| "epoch": 0.9206403856085384, | |
| "grad_norm": 1.2159273624420166, | |
| "learning_rate": 2.946370234583015e-08, | |
| "loss": 0.0802, | |
| "step": 1337 | |
| }, | |
| { | |
| "epoch": 0.9213289722843863, | |
| "grad_norm": 1.197150707244873, | |
| "learning_rate": 2.89654551528514e-08, | |
| "loss": 0.0798, | |
| "step": 1338 | |
| }, | |
| { | |
| "epoch": 0.9220175589602341, | |
| "grad_norm": 0.7052870988845825, | |
| "learning_rate": 2.8471373817596515e-08, | |
| "loss": 0.0612, | |
| "step": 1339 | |
| }, | |
| { | |
| "epoch": 0.9227061456360819, | |
| "grad_norm": 0.5017627477645874, | |
| "learning_rate": 2.7981461194675623e-08, | |
| "loss": 0.0566, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.9233947323119298, | |
| "grad_norm": 0.8716162443161011, | |
| "learning_rate": 2.749572011461371e-08, | |
| "loss": 0.0698, | |
| "step": 1341 | |
| }, | |
| { | |
| "epoch": 0.9240833189877776, | |
| "grad_norm": 1.0268443822860718, | |
| "learning_rate": 2.70141533838342e-08, | |
| "loss": 0.0795, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 0.9247719056636254, | |
| "grad_norm": 1.5873016119003296, | |
| "learning_rate": 2.6536763784642732e-08, | |
| "loss": 0.0732, | |
| "step": 1343 | |
| }, | |
| { | |
| "epoch": 0.9254604923394733, | |
| "grad_norm": 1.1690579652786255, | |
| "learning_rate": 2.606355407521116e-08, | |
| "loss": 0.0715, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 0.9261490790153211, | |
| "grad_norm": 0.8380487561225891, | |
| "learning_rate": 2.559452698956141e-08, | |
| "loss": 0.0555, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.9268376656911689, | |
| "grad_norm": 0.8684842586517334, | |
| "learning_rate": 2.512968523754991e-08, | |
| "loss": 0.0589, | |
| "step": 1346 | |
| }, | |
| { | |
| "epoch": 0.9275262523670167, | |
| "grad_norm": 0.48923298716545105, | |
| "learning_rate": 2.4669031504852085e-08, | |
| "loss": 0.0536, | |
| "step": 1347 | |
| }, | |
| { | |
| "epoch": 0.9282148390428645, | |
| "grad_norm": 0.6461222171783447, | |
| "learning_rate": 2.421256845294606e-08, | |
| "loss": 0.0595, | |
| "step": 1348 | |
| }, | |
| { | |
| "epoch": 0.9289034257187123, | |
| "grad_norm": 0.8027684688568115, | |
| "learning_rate": 2.3760298719098406e-08, | |
| "loss": 0.0714, | |
| "step": 1349 | |
| }, | |
| { | |
| "epoch": 0.9295920123945601, | |
| "grad_norm": 1.3216205835342407, | |
| "learning_rate": 2.3312224916347992e-08, | |
| "loss": 0.057, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.9302805990704079, | |
| "grad_norm": 1.0285162925720215, | |
| "learning_rate": 2.286834963349141e-08, | |
| "loss": 0.0716, | |
| "step": 1351 | |
| }, | |
| { | |
| "epoch": 0.9309691857462558, | |
| "grad_norm": 0.8122292757034302, | |
| "learning_rate": 2.2428675435067658e-08, | |
| "loss": 0.0576, | |
| "step": 1352 | |
| }, | |
| { | |
| "epoch": 0.9316577724221037, | |
| "grad_norm": 1.3762600421905518, | |
| "learning_rate": 2.199320486134365e-08, | |
| "loss": 0.0633, | |
| "step": 1353 | |
| }, | |
| { | |
| "epoch": 0.9323463590979515, | |
| "grad_norm": 0.9969989657402039, | |
| "learning_rate": 2.15619404282994e-08, | |
| "loss": 0.065, | |
| "step": 1354 | |
| }, | |
| { | |
| "epoch": 0.9330349457737993, | |
| "grad_norm": 0.8273980021476746, | |
| "learning_rate": 2.1134884627613178e-08, | |
| "loss": 0.0706, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.9337235324496471, | |
| "grad_norm": 1.2278693914413452, | |
| "learning_rate": 2.0712039926647886e-08, | |
| "loss": 0.0662, | |
| "step": 1356 | |
| }, | |
| { | |
| "epoch": 0.9344121191254949, | |
| "grad_norm": 0.490213006734848, | |
| "learning_rate": 2.0293408768435883e-08, | |
| "loss": 0.0516, | |
| "step": 1357 | |
| }, | |
| { | |
| "epoch": 0.9351007058013427, | |
| "grad_norm": 0.6395962238311768, | |
| "learning_rate": 1.9878993571665587e-08, | |
| "loss": 0.0606, | |
| "step": 1358 | |
| }, | |
| { | |
| "epoch": 0.9357892924771906, | |
| "grad_norm": 1.7359540462493896, | |
| "learning_rate": 1.946879673066715e-08, | |
| "loss": 0.0546, | |
| "step": 1359 | |
| }, | |
| { | |
| "epoch": 0.9364778791530384, | |
| "grad_norm": 1.2886805534362793, | |
| "learning_rate": 1.9062820615398714e-08, | |
| "loss": 0.0639, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.9371664658288862, | |
| "grad_norm": 0.7155603170394897, | |
| "learning_rate": 1.8661067571432767e-08, | |
| "loss": 0.0466, | |
| "step": 1361 | |
| }, | |
| { | |
| "epoch": 0.937855052504734, | |
| "grad_norm": 0.7551135420799255, | |
| "learning_rate": 1.826353991994223e-08, | |
| "loss": 0.0712, | |
| "step": 1362 | |
| }, | |
| { | |
| "epoch": 0.9385436391805818, | |
| "grad_norm": 0.9385201334953308, | |
| "learning_rate": 1.7870239957687717e-08, | |
| "loss": 0.0738, | |
| "step": 1363 | |
| }, | |
| { | |
| "epoch": 0.9392322258564296, | |
| "grad_norm": 1.1299532651901245, | |
| "learning_rate": 1.74811699570038e-08, | |
| "loss": 0.0474, | |
| "step": 1364 | |
| }, | |
| { | |
| "epoch": 0.9399208125322775, | |
| "grad_norm": 1.9095109701156616, | |
| "learning_rate": 1.7096332165785765e-08, | |
| "loss": 0.059, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.9406093992081254, | |
| "grad_norm": 1.5280030965805054, | |
| "learning_rate": 1.6715728807477043e-08, | |
| "loss": 0.081, | |
| "step": 1366 | |
| }, | |
| { | |
| "epoch": 0.9412979858839732, | |
| "grad_norm": 0.7835893034934998, | |
| "learning_rate": 1.633936208105613e-08, | |
| "loss": 0.0597, | |
| "step": 1367 | |
| }, | |
| { | |
| "epoch": 0.941986572559821, | |
| "grad_norm": 0.6840996146202087, | |
| "learning_rate": 1.5967234161023943e-08, | |
| "loss": 0.0722, | |
| "step": 1368 | |
| }, | |
| { | |
| "epoch": 0.9426751592356688, | |
| "grad_norm": 1.3263771533966064, | |
| "learning_rate": 1.559934719739106e-08, | |
| "loss": 0.0878, | |
| "step": 1369 | |
| }, | |
| { | |
| "epoch": 0.9433637459115166, | |
| "grad_norm": 1.2267135381698608, | |
| "learning_rate": 1.5235703315665594e-08, | |
| "loss": 0.0606, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.9440523325873644, | |
| "grad_norm": 1.2609857320785522, | |
| "learning_rate": 1.4876304616840836e-08, | |
| "loss": 0.0597, | |
| "step": 1371 | |
| }, | |
| { | |
| "epoch": 0.9447409192632122, | |
| "grad_norm": 0.7519130706787109, | |
| "learning_rate": 1.4521153177382785e-08, | |
| "loss": 0.0536, | |
| "step": 1372 | |
| }, | |
| { | |
| "epoch": 0.94542950593906, | |
| "grad_norm": 1.0901200771331787, | |
| "learning_rate": 1.4170251049218652e-08, | |
| "loss": 0.0503, | |
| "step": 1373 | |
| }, | |
| { | |
| "epoch": 0.9461180926149079, | |
| "grad_norm": 1.362013339996338, | |
| "learning_rate": 1.382360025972479e-08, | |
| "loss": 0.0642, | |
| "step": 1374 | |
| }, | |
| { | |
| "epoch": 0.9468066792907557, | |
| "grad_norm": 1.2554388046264648, | |
| "learning_rate": 1.3481202811714944e-08, | |
| "loss": 0.0683, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.9474952659666035, | |
| "grad_norm": 1.3946216106414795, | |
| "learning_rate": 1.314306068342852e-08, | |
| "loss": 0.0653, | |
| "step": 1376 | |
| }, | |
| { | |
| "epoch": 0.9481838526424514, | |
| "grad_norm": 1.222314715385437, | |
| "learning_rate": 1.2809175828519593e-08, | |
| "loss": 0.0765, | |
| "step": 1377 | |
| }, | |
| { | |
| "epoch": 0.9488724393182992, | |
| "grad_norm": 0.9005465507507324, | |
| "learning_rate": 1.2479550176045323e-08, | |
| "loss": 0.0597, | |
| "step": 1378 | |
| }, | |
| { | |
| "epoch": 0.949561025994147, | |
| "grad_norm": 0.9284931421279907, | |
| "learning_rate": 1.2154185630454562e-08, | |
| "loss": 0.063, | |
| "step": 1379 | |
| }, | |
| { | |
| "epoch": 0.9502496126699949, | |
| "grad_norm": 0.9003196358680725, | |
| "learning_rate": 1.1833084071577516e-08, | |
| "loss": 0.0587, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.9509381993458427, | |
| "grad_norm": 1.0747370719909668, | |
| "learning_rate": 1.1516247354614345e-08, | |
| "loss": 0.0615, | |
| "step": 1381 | |
| }, | |
| { | |
| "epoch": 0.9516267860216905, | |
| "grad_norm": 1.0467376708984375, | |
| "learning_rate": 1.1203677310124588e-08, | |
| "loss": 0.0634, | |
| "step": 1382 | |
| }, | |
| { | |
| "epoch": 0.9523153726975383, | |
| "grad_norm": 1.0511927604675293, | |
| "learning_rate": 1.0895375744016583e-08, | |
| "loss": 0.0754, | |
| "step": 1383 | |
| }, | |
| { | |
| "epoch": 0.9530039593733861, | |
| "grad_norm": 0.6931172013282776, | |
| "learning_rate": 1.0591344437537232e-08, | |
| "loss": 0.061, | |
| "step": 1384 | |
| }, | |
| { | |
| "epoch": 0.9536925460492339, | |
| "grad_norm": 0.5200431942939758, | |
| "learning_rate": 1.0291585147261251e-08, | |
| "loss": 0.0465, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.9543811327250817, | |
| "grad_norm": 0.8292903304100037, | |
| "learning_rate": 9.996099605081605e-09, | |
| "loss": 0.0575, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 0.9550697194009296, | |
| "grad_norm": 0.9346963763237, | |
| "learning_rate": 9.70488951819909e-09, | |
| "loss": 0.064, | |
| "step": 1387 | |
| }, | |
| { | |
| "epoch": 0.9557583060767774, | |
| "grad_norm": 0.8979430794715881, | |
| "learning_rate": 9.417956569112429e-09, | |
| "loss": 0.0517, | |
| "step": 1388 | |
| }, | |
| { | |
| "epoch": 0.9564468927526253, | |
| "grad_norm": 1.2510077953338623, | |
| "learning_rate": 9.135302415608864e-09, | |
| "loss": 0.059, | |
| "step": 1389 | |
| }, | |
| { | |
| "epoch": 0.9571354794284731, | |
| "grad_norm": 1.1528666019439697, | |
| "learning_rate": 8.85692869075433e-09, | |
| "loss": 0.0703, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.9578240661043209, | |
| "grad_norm": 1.6137210130691528, | |
| "learning_rate": 8.582837002884287e-09, | |
| "loss": 0.0497, | |
| "step": 1391 | |
| }, | |
| { | |
| "epoch": 0.9585126527801687, | |
| "grad_norm": 1.5655218362808228, | |
| "learning_rate": 8.313028935593991e-09, | |
| "loss": 0.0741, | |
| "step": 1392 | |
| }, | |
| { | |
| "epoch": 0.9592012394560165, | |
| "grad_norm": 0.9392473101615906, | |
| "learning_rate": 8.04750604772983e-09, | |
| "loss": 0.0521, | |
| "step": 1393 | |
| }, | |
| { | |
| "epoch": 0.9598898261318644, | |
| "grad_norm": 1.004064679145813, | |
| "learning_rate": 7.78626987337991e-09, | |
| "loss": 0.0535, | |
| "step": 1394 | |
| }, | |
| { | |
| "epoch": 0.9605784128077122, | |
| "grad_norm": 0.8733145594596863, | |
| "learning_rate": 7.529321921865478e-09, | |
| "loss": 0.0718, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.96126699948356, | |
| "grad_norm": 1.284411907196045, | |
| "learning_rate": 7.276663677732109e-09, | |
| "loss": 0.064, | |
| "step": 1396 | |
| }, | |
| { | |
| "epoch": 0.9619555861594078, | |
| "grad_norm": 0.6915552020072937, | |
| "learning_rate": 7.028296600740947e-09, | |
| "loss": 0.0523, | |
| "step": 1397 | |
| }, | |
| { | |
| "epoch": 0.9626441728352556, | |
| "grad_norm": 0.9719657897949219, | |
| "learning_rate": 6.7842221258605506e-09, | |
| "loss": 0.0502, | |
| "step": 1398 | |
| }, | |
| { | |
| "epoch": 0.9633327595111034, | |
| "grad_norm": 1.8920620679855347, | |
| "learning_rate": 6.544441663258655e-09, | |
| "loss": 0.0513, | |
| "step": 1399 | |
| }, | |
| { | |
| "epoch": 0.9640213461869512, | |
| "grad_norm": 2.993326187133789, | |
| "learning_rate": 6.308956598293586e-09, | |
| "loss": 0.0845, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.9640213461869512, | |
| "eval_loss": NaN, | |
| "eval_runtime": 65.1384, | |
| "eval_samples_per_second": 14.431, | |
| "eval_steps_per_second": 0.461, | |
| "step": 1400 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1453, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.9728944545805107e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |