| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.6199095022624435, | |
| "eval_steps": 1000, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01809954751131222, | |
| "grad_norm": 4.442032814025879, | |
| "learning_rate": 0.0, | |
| "loss": 1.497, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03619909502262444, | |
| "grad_norm": 4.569097995758057, | |
| "learning_rate": 1.5051499783199055e-07, | |
| "loss": 1.4506, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05429864253393665, | |
| "grad_norm": 3.9136922359466553, | |
| "learning_rate": 2.385606273598312e-07, | |
| "loss": 1.2728, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07239819004524888, | |
| "grad_norm": 3.9136922359466553, | |
| "learning_rate": 2.385606273598312e-07, | |
| "loss": 1.587, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.09049773755656108, | |
| "grad_norm": 3.9136922359466553, | |
| "learning_rate": 2.385606273598312e-07, | |
| "loss": 1.408, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1085972850678733, | |
| "grad_norm": 4.174969673156738, | |
| "learning_rate": 3.010299956639811e-07, | |
| "loss": 1.3948, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12669683257918551, | |
| "grad_norm": 4.174969673156738, | |
| "learning_rate": 3.010299956639811e-07, | |
| "loss": 1.515, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.14479638009049775, | |
| "grad_norm": 4.5925774574279785, | |
| "learning_rate": 3.494850021680093e-07, | |
| "loss": 1.5149, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.16289592760180996, | |
| "grad_norm": 4.349692344665527, | |
| "learning_rate": 3.8907562519182173e-07, | |
| "loss": 1.5055, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.18099547511312217, | |
| "grad_norm": 5.4546027183532715, | |
| "learning_rate": 4.2254902000712834e-07, | |
| "loss": 1.4242, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19909502262443438, | |
| "grad_norm": 4.096962928771973, | |
| "learning_rate": 4.5154499349597166e-07, | |
| "loss": 1.4094, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.2171945701357466, | |
| "grad_norm": 5.804754734039307, | |
| "learning_rate": 4.771212547196623e-07, | |
| "loss": 1.6699, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 4.950932502746582, | |
| "learning_rate": 4.999999999999999e-07, | |
| "loss": 1.4682, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.25339366515837103, | |
| "grad_norm": 4.728433609008789, | |
| "learning_rate": 5.206963425791124e-07, | |
| "loss": 1.3831, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.27149321266968324, | |
| "grad_norm": 4.042943000793457, | |
| "learning_rate": 5.395906230238123e-07, | |
| "loss": 1.4438, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2895927601809955, | |
| "grad_norm": 3.768977642059326, | |
| "learning_rate": 5.569716761534182e-07, | |
| "loss": 1.3996, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 4.375296115875244, | |
| "learning_rate": 5.730640178391189e-07, | |
| "loss": 1.3065, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3257918552036199, | |
| "grad_norm": 3.959031105041504, | |
| "learning_rate": 5.880456295278405e-07, | |
| "loss": 1.4533, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3438914027149321, | |
| "grad_norm": 4.150556564331055, | |
| "learning_rate": 6.020599913279622e-07, | |
| "loss": 1.4256, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.36199095022624433, | |
| "grad_norm": 5.503087520599365, | |
| "learning_rate": 6.15224460689137e-07, | |
| "loss": 1.3212, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.38009049773755654, | |
| "grad_norm": 5.915998935699463, | |
| "learning_rate": 6.276362525516529e-07, | |
| "loss": 1.4347, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.39819004524886875, | |
| "grad_norm": 3.7581424713134766, | |
| "learning_rate": 6.393768004764143e-07, | |
| "loss": 1.3012, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.416289592760181, | |
| "grad_norm": 3.728571653366089, | |
| "learning_rate": 6.505149978319905e-07, | |
| "loss": 1.3426, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4343891402714932, | |
| "grad_norm": 4.000448226928711, | |
| "learning_rate": 6.611096473669595e-07, | |
| "loss": 1.3064, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.45248868778280543, | |
| "grad_norm": 3.8997511863708496, | |
| "learning_rate": 6.712113404111031e-07, | |
| "loss": 1.4217, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 3.4601259231567383, | |
| "learning_rate": 6.808639180087963e-07, | |
| "loss": 1.278, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.48868778280542985, | |
| "grad_norm": 3.3370354175567627, | |
| "learning_rate": 6.901056208558029e-07, | |
| "loss": 1.402, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5067873303167421, | |
| "grad_norm": 4.348806381225586, | |
| "learning_rate": 6.989700043360186e-07, | |
| "loss": 1.3881, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5248868778280543, | |
| "grad_norm": 3.3785560131073, | |
| "learning_rate": 7.074866739854088e-07, | |
| "loss": 1.2949, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5429864253393665, | |
| "grad_norm": 3.269310474395752, | |
| "learning_rate": 7.156818820794935e-07, | |
| "loss": 1.3877, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5610859728506787, | |
| "grad_norm": 3.2553048133850098, | |
| "learning_rate": 7.235790156711094e-07, | |
| "loss": 1.3625, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.579185520361991, | |
| "grad_norm": 4.357567310333252, | |
| "learning_rate": 7.311989989494779e-07, | |
| "loss": 1.2856, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5972850678733032, | |
| "grad_norm": 3.816030263900757, | |
| "learning_rate": 7.38560627359831e-07, | |
| "loss": 1.2666, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 3.816030263900757, | |
| "learning_rate": 7.38560627359831e-07, | |
| "loss": 1.288, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6334841628959276, | |
| "grad_norm": 3.095264434814453, | |
| "learning_rate": 7.456808469171361e-07, | |
| "loss": 1.3014, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6515837104072398, | |
| "grad_norm": 5.308295726776123, | |
| "learning_rate": 7.525749891599529e-07, | |
| "loss": 1.3896, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.669683257918552, | |
| "grad_norm": 3.2766010761260986, | |
| "learning_rate": 7.592569699389436e-07, | |
| "loss": 1.3191, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6877828054298643, | |
| "grad_norm": 3.6091532707214355, | |
| "learning_rate": 7.657394585211274e-07, | |
| "loss": 1.3227, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 4.882140159606934, | |
| "learning_rate": 7.720340221751376e-07, | |
| "loss": 1.2763, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7239819004524887, | |
| "grad_norm": 3.5777502059936523, | |
| "learning_rate": 7.781512503836435e-07, | |
| "loss": 1.2811, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7420814479638009, | |
| "grad_norm": 3.520531415939331, | |
| "learning_rate": 7.841008620334974e-07, | |
| "loss": 1.3677, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7601809954751131, | |
| "grad_norm": 4.111598014831543, | |
| "learning_rate": 7.89891798308405e-07, | |
| "loss": 1.3009, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7782805429864253, | |
| "grad_norm": 3.3306169509887695, | |
| "learning_rate": 7.955323035132494e-07, | |
| "loss": 1.1605, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7963800904977375, | |
| "grad_norm": 3.112687349319458, | |
| "learning_rate": 8.01029995663981e-07, | |
| "loss": 1.2589, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.8144796380090498, | |
| "grad_norm": 3.7117624282836914, | |
| "learning_rate": 8.063919283598676e-07, | |
| "loss": 1.2737, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.832579185520362, | |
| "grad_norm": 4.166692733764648, | |
| "learning_rate": 8.116246451989502e-07, | |
| "loss": 1.2353, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8506787330316742, | |
| "grad_norm": 6.097745895385742, | |
| "learning_rate": 8.16734227789793e-07, | |
| "loss": 1.221, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8687782805429864, | |
| "grad_norm": 3.3959226608276367, | |
| "learning_rate": 8.217263382430935e-07, | |
| "loss": 1.2574, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8868778280542986, | |
| "grad_norm": 4.395689964294434, | |
| "learning_rate": 8.266062568876716e-07, | |
| "loss": 1.1438, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.9049773755656109, | |
| "grad_norm": 3.479315757751465, | |
| "learning_rate": 8.313789158407869e-07, | |
| "loss": 1.2905, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 3.4640543460845947, | |
| "learning_rate": 8.360489289678585e-07, | |
| "loss": 1.2397, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 5.382558822631836, | |
| "learning_rate": 8.406206186877934e-07, | |
| "loss": 1.2131, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9592760180995475, | |
| "grad_norm": 5.364609718322754, | |
| "learning_rate": 8.450980400142567e-07, | |
| "loss": 1.2098, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.9773755656108597, | |
| "grad_norm": 3.814605951309204, | |
| "learning_rate": 8.494850021680092e-07, | |
| "loss": 1.2415, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.995475113122172, | |
| "grad_norm": 2.9692606925964355, | |
| "learning_rate": 8.53785088048968e-07, | |
| "loss": 1.1955, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.0135746606334841, | |
| "grad_norm": 3.130876302719116, | |
| "learning_rate": 8.580016718173995e-07, | |
| "loss": 1.158, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0316742081447963, | |
| "grad_norm": 3.4549965858459473, | |
| "learning_rate": 8.621379348003944e-07, | |
| "loss": 1.2838, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.0497737556561086, | |
| "grad_norm": 3.4735677242279053, | |
| "learning_rate": 8.661968799114842e-07, | |
| "loss": 1.0405, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0678733031674208, | |
| "grad_norm": 4.529227256774902, | |
| "learning_rate": 8.701813447471218e-07, | |
| "loss": 1.1188, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.085972850678733, | |
| "grad_norm": 3.53712797164917, | |
| "learning_rate": 8.740940135031001e-07, | |
| "loss": 1.2113, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.1040723981900453, | |
| "grad_norm": 3.5053398609161377, | |
| "learning_rate": 8.779374278362456e-07, | |
| "loss": 1.197, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.1221719457013575, | |
| "grad_norm": 2.57523512840271, | |
| "learning_rate": 8.817139967814684e-07, | |
| "loss": 1.0894, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.1402714932126696, | |
| "grad_norm": 3.3335084915161133, | |
| "learning_rate": 8.854260058210719e-07, | |
| "loss": 1.026, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1583710407239818, | |
| "grad_norm": 3.1159112453460693, | |
| "learning_rate": 8.890756251918216e-07, | |
| "loss": 1.1336, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 4.9009294509887695, | |
| "learning_rate": 8.926649175053833e-07, | |
| "loss": 1.1167, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1945701357466063, | |
| "grad_norm": 3.9954421520233154, | |
| "learning_rate": 8.961958447491268e-07, | |
| "loss": 1.0955, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.2126696832579185, | |
| "grad_norm": 4.976899147033691, | |
| "learning_rate": 8.996702747267907e-07, | |
| "loss": 1.1722, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 4.318509101867676, | |
| "learning_rate": 9.030899869919433e-07, | |
| "loss": 1.0919, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.248868778280543, | |
| "grad_norm": 3.509249210357666, | |
| "learning_rate": 9.064566783214276e-07, | |
| "loss": 1.1013, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.2669683257918551, | |
| "grad_norm": 2.86033034324646, | |
| "learning_rate": 9.097719677709341e-07, | |
| "loss": 1.0721, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2850678733031673, | |
| "grad_norm": 2.95745587348938, | |
| "learning_rate": 9.13037401350413e-07, | |
| "loss": 1.1235, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.3031674208144797, | |
| "grad_norm": 3.4295952320098877, | |
| "learning_rate": 9.162544563531181e-07, | |
| "loss": 1.1075, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.3212669683257918, | |
| "grad_norm": 4.3247575759887695, | |
| "learning_rate": 9.194245453686276e-07, | |
| "loss": 1.0976, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.3393665158371042, | |
| "grad_norm": 3.853940010070801, | |
| "learning_rate": 9.225490200071283e-07, | |
| "loss": 0.997, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.3574660633484164, | |
| "grad_norm": 3.1195785999298096, | |
| "learning_rate": 9.256291743595375e-07, | |
| "loss": 1.1499, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.3755656108597285, | |
| "grad_norm": 3.3039655685424805, | |
| "learning_rate": 9.28666248215634e-07, | |
| "loss": 1.0822, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3936651583710407, | |
| "grad_norm": 2.9753613471984863, | |
| "learning_rate": 9.316614300602277e-07, | |
| "loss": 1.0524, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 2.7271316051483154, | |
| "learning_rate": 9.346158598654879e-07, | |
| "loss": 1.0233, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.4298642533936652, | |
| "grad_norm": 5.75250244140625, | |
| "learning_rate": 9.375306316958498e-07, | |
| "loss": 1.0652, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.4479638009049773, | |
| "grad_norm": 3.271881341934204, | |
| "learning_rate": 9.404067961403955e-07, | |
| "loss": 0.9659, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4660633484162897, | |
| "grad_norm": 3.8478872776031494, | |
| "learning_rate": 9.432453625862408e-07, | |
| "loss": 1.1446, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.4841628959276019, | |
| "grad_norm": 2.925736904144287, | |
| "learning_rate": 9.4604730134524e-07, | |
| "loss": 0.9846, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.502262443438914, | |
| "grad_norm": 2.8262815475463867, | |
| "learning_rate": 9.488135456452205e-07, | |
| "loss": 0.9395, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.5203619909502262, | |
| "grad_norm": 4.063136100769043, | |
| "learning_rate": 9.515449934959715e-07, | |
| "loss": 1.0066, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 2.5695972442626953, | |
| "learning_rate": 9.542425094393247e-07, | |
| "loss": 0.9592, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.5565610859728507, | |
| "grad_norm": 3.1798267364501953, | |
| "learning_rate": 9.569069261918583e-07, | |
| "loss": 0.9213, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.5746606334841629, | |
| "grad_norm": 6.351828098297119, | |
| "learning_rate": 9.59539046188037e-07, | |
| "loss": 0.9416, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.5927601809954752, | |
| "grad_norm": 2.9044878482818604, | |
| "learning_rate": 9.621396430309406e-07, | |
| "loss": 0.9626, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.6108597285067874, | |
| "grad_norm": 3.130195140838623, | |
| "learning_rate": 9.647094628571462e-07, | |
| "loss": 0.9767, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.6289592760180995, | |
| "grad_norm": 6.14766263961792, | |
| "learning_rate": 9.672492256217836e-07, | |
| "loss": 1.029, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 3.2492759227752686, | |
| "learning_rate": 9.69759626309309e-07, | |
| "loss": 0.9621, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.6651583710407238, | |
| "grad_norm": 2.8760759830474854, | |
| "learning_rate": 9.722413360750842e-07, | |
| "loss": 1.0525, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.6832579185520362, | |
| "grad_norm": 2.910680055618286, | |
| "learning_rate": 9.74695003322456e-07, | |
| "loss": 0.8777, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.7013574660633484, | |
| "grad_norm": 3.2539334297180176, | |
| "learning_rate": 9.771212547196622e-07, | |
| "loss": 0.9407, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.7194570135746607, | |
| "grad_norm": 2.7406673431396484, | |
| "learning_rate": 9.795206961605466e-07, | |
| "loss": 0.9077, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.737556561085973, | |
| "grad_norm": 3.0724148750305176, | |
| "learning_rate": 9.818939136727774e-07, | |
| "loss": 0.8289, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.755656108597285, | |
| "grad_norm": 3.2282042503356934, | |
| "learning_rate": 9.842414742769674e-07, | |
| "loss": 0.9638, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.7737556561085972, | |
| "grad_norm": 6.185736656188965, | |
| "learning_rate": 9.865639267998492e-07, | |
| "loss": 0.9356, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.7918552036199094, | |
| "grad_norm": 3.5250043869018555, | |
| "learning_rate": 9.888618026444236e-07, | |
| "loss": 0.8963, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.8099547511312217, | |
| "grad_norm": 3.418933629989624, | |
| "learning_rate": 9.91135616519784e-07, | |
| "loss": 0.9555, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.8280542986425339, | |
| "grad_norm": 2.943357467651367, | |
| "learning_rate": 9.933858671331222e-07, | |
| "loss": 0.8866, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 3.657294511795044, | |
| "learning_rate": 9.956130378462473e-07, | |
| "loss": 0.9626, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.8642533936651584, | |
| "grad_norm": 3.3619449138641357, | |
| "learning_rate": 9.978175972987748e-07, | |
| "loss": 0.8601, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 3.6125662326812744, | |
| "learning_rate": 9.999999999999997e-07, | |
| "loss": 0.8527, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.9004524886877827, | |
| "grad_norm": 3.5924038887023926, | |
| "learning_rate": 1e-06, | |
| "loss": 0.879, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.9185520361990949, | |
| "grad_norm": 2.770998001098633, | |
| "learning_rate": 1e-06, | |
| "loss": 0.8364, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.9366515837104072, | |
| "grad_norm": 3.2162160873413086, | |
| "learning_rate": 1e-06, | |
| "loss": 0.8591, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.9547511312217196, | |
| "grad_norm": 5.3574910163879395, | |
| "learning_rate": 1e-06, | |
| "loss": 0.9122, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.9728506787330318, | |
| "grad_norm": 3.564077138900757, | |
| "learning_rate": 1e-06, | |
| "loss": 0.9706, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.990950226244344, | |
| "grad_norm": 3.347325086593628, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7614, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.009049773755656, | |
| "grad_norm": 2.967604637145996, | |
| "learning_rate": 1e-06, | |
| "loss": 0.8534, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.0271493212669682, | |
| "grad_norm": 2.7884786128997803, | |
| "learning_rate": 1e-06, | |
| "loss": 0.8786, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.0452488687782804, | |
| "grad_norm": 3.3380022048950195, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7821, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.0633484162895925, | |
| "grad_norm": 2.7749805450439453, | |
| "learning_rate": 1e-06, | |
| "loss": 0.749, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.081447963800905, | |
| "grad_norm": 2.6747641563415527, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7501, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.0995475113122173, | |
| "grad_norm": 3.078357458114624, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7979, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 5.836832523345947, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7917, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.1357466063348416, | |
| "grad_norm": 6.190191745758057, | |
| "learning_rate": 1e-06, | |
| "loss": 0.8743, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 3.3209612369537354, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7989, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.171945701357466, | |
| "grad_norm": 3.5375821590423584, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7356, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.1900452488687785, | |
| "grad_norm": 2.635437488555908, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7602, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.2081447963800906, | |
| "grad_norm": 4.8566460609436035, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7611, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.226244343891403, | |
| "grad_norm": 3.1864778995513916, | |
| "learning_rate": 1e-06, | |
| "loss": 0.767, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.244343891402715, | |
| "grad_norm": 4.00085973739624, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7164, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.262443438914027, | |
| "grad_norm": 3.608243465423584, | |
| "learning_rate": 1e-06, | |
| "loss": 0.8259, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.2805429864253393, | |
| "grad_norm": 2.6522486209869385, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6852, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.2986425339366514, | |
| "grad_norm": 3.137711524963379, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7703, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.3167420814479636, | |
| "grad_norm": 3.867400884628296, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7472, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.334841628959276, | |
| "grad_norm": 3.147169351577759, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6673, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 5.177737236022949, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7478, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.3710407239819005, | |
| "grad_norm": 4.302467346191406, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7025, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.3891402714932126, | |
| "grad_norm": 3.5397591590881348, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6615, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.4072398190045248, | |
| "grad_norm": 5.074939250946045, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7043, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.425339366515837, | |
| "grad_norm": 7.0850090980529785, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6587, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.4434389140271495, | |
| "grad_norm": 3.6032073497772217, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6726, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 2.707559108734131, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6529, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.479638009049774, | |
| "grad_norm": 3.5912561416625977, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6655, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.497737556561086, | |
| "grad_norm": 3.537712335586548, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6617, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.515837104072398, | |
| "grad_norm": 8.936192512512207, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6757, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.5339366515837103, | |
| "grad_norm": 2.4215872287750244, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6143, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.5520361990950224, | |
| "grad_norm": 3.1526424884796143, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6308, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.5701357466063346, | |
| "grad_norm": 4.358147144317627, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7556, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 3.423694610595703, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5853, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.6063348416289593, | |
| "grad_norm": 3.792518138885498, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6154, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.6244343891402715, | |
| "grad_norm": 7.347434997558594, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5983, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.6425339366515836, | |
| "grad_norm": 3.5371973514556885, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6173, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.660633484162896, | |
| "grad_norm": 3.084972620010376, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5556, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.6787330316742084, | |
| "grad_norm": 5.569381237030029, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6146, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.6968325791855206, | |
| "grad_norm": 3.270057201385498, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5938, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.7149321266968327, | |
| "grad_norm": 3.9021122455596924, | |
| "learning_rate": 1e-06, | |
| "loss": 0.532, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.733031674208145, | |
| "grad_norm": 3.702974319458008, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5419, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.751131221719457, | |
| "grad_norm": 3.640477418899536, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5927, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 4.119628429412842, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6556, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.7873303167420813, | |
| "grad_norm": 4.196484565734863, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5473, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.8054298642533935, | |
| "grad_norm": 3.049004316329956, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5382, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 3.8105475902557373, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5667, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.841628959276018, | |
| "grad_norm": 7.120466232299805, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5815, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.8597285067873304, | |
| "grad_norm": 2.96946120262146, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5488, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.8778280542986425, | |
| "grad_norm": 3.927828073501587, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5051, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.8959276018099547, | |
| "grad_norm": 3.6861846446990967, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5683, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.914027149321267, | |
| "grad_norm": 3.383025646209717, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4956, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.9321266968325794, | |
| "grad_norm": 3.9769487380981445, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5101, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.9502262443438916, | |
| "grad_norm": 3.262488842010498, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5589, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.9683257918552037, | |
| "grad_norm": 3.582789182662964, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5064, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.986425339366516, | |
| "grad_norm": 3.441208839416504, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4696, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.004524886877828, | |
| "grad_norm": 4.338072299957275, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5083, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 3.02262443438914, | |
| "grad_norm": 3.537062644958496, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5328, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 3.0407239819004523, | |
| "grad_norm": 3.4206771850585938, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4615, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.0588235294117645, | |
| "grad_norm": 3.4206771850585938, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4958, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 3.015188694000244, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5143, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.0950226244343892, | |
| "grad_norm": 3.314436435699463, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4972, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 3.1131221719457014, | |
| "grad_norm": 4.298553466796875, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4794, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 3.1312217194570136, | |
| "grad_norm": 4.354773044586182, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4287, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 3.1493212669683257, | |
| "grad_norm": 3.558988332748413, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4092, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.167420814479638, | |
| "grad_norm": 5.127811431884766, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4465, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.1855203619909505, | |
| "grad_norm": 2.5903167724609375, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4119, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 3.2036199095022626, | |
| "grad_norm": 3.8117196559906006, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4529, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 3.2217194570135748, | |
| "grad_norm": 3.164416551589966, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4856, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 3.239819004524887, | |
| "grad_norm": 4.106442451477051, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4312, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 3.257918552036199, | |
| "grad_norm": 3.762195110321045, | |
| "learning_rate": 1e-06, | |
| "loss": 0.411, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.276018099547511, | |
| "grad_norm": 3.0889034271240234, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3977, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 3.6932199001312256, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4528, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 3.3122171945701355, | |
| "grad_norm": 3.405756711959839, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4042, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 3.330316742081448, | |
| "grad_norm": 5.229561805725098, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4171, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 3.3484162895927603, | |
| "grad_norm": 5.6756439208984375, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3891, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.3665158371040724, | |
| "grad_norm": 3.706697940826416, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3673, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 2.9856503009796143, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3989, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 3.4027149321266967, | |
| "grad_norm": 3.5144922733306885, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3885, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 3.420814479638009, | |
| "grad_norm": 5.046453475952148, | |
| "learning_rate": 1e-06, | |
| "loss": 0.402, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 3.4389140271493215, | |
| "grad_norm": 4.306224822998047, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3568, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.4570135746606336, | |
| "grad_norm": 2.7284157276153564, | |
| "learning_rate": 1e-06, | |
| "loss": 0.4109, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.475113122171946, | |
| "grad_norm": 5.036966323852539, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3958, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.493212669683258, | |
| "grad_norm": 2.3863677978515625, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3674, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 3.51131221719457, | |
| "grad_norm": 3.3838043212890625, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3627, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 6.964299201965332, | |
| "learning_rate": 1e-06, | |
| "loss": 0.405, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.5475113122171944, | |
| "grad_norm": 2.7131292819976807, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3599, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 3.5656108597285066, | |
| "grad_norm": 2.6383941173553467, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3335, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 3.583710407239819, | |
| "grad_norm": 2.710670232772827, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3979, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 3.6018099547511313, | |
| "grad_norm": 2.6803207397460938, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3695, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 3.6199095022624435, | |
| "grad_norm": 3.22158145904541, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3832, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 182, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1675256582811156e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |