| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9790419161676647, | |
| "eval_steps": 167, | |
| "global_step": 1336, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0014970059880239522, | |
| "grad_norm": 0.09437490254640579, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.5976, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0014970059880239522, | |
| "eval_loss": 0.4935530424118042, | |
| "eval_runtime": 96.1134, | |
| "eval_samples_per_second": 7.658, | |
| "eval_steps_per_second": 0.957, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0029940119760479044, | |
| "grad_norm": 0.07851643115282059, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.5631, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.004491017964071856, | |
| "grad_norm": 0.07771598547697067, | |
| "learning_rate": 3e-06, | |
| "loss": 0.4231, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.005988023952095809, | |
| "grad_norm": 0.07919719070196152, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.3963, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0074850299401197605, | |
| "grad_norm": 0.08890288323163986, | |
| "learning_rate": 5e-06, | |
| "loss": 0.57, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.008982035928143712, | |
| "grad_norm": 0.0997990146279335, | |
| "learning_rate": 6e-06, | |
| "loss": 0.3526, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.010479041916167664, | |
| "grad_norm": 0.09869150817394257, | |
| "learning_rate": 7e-06, | |
| "loss": 0.5726, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.011976047904191617, | |
| "grad_norm": 0.08410096168518066, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4443, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01347305389221557, | |
| "grad_norm": 0.10099457949399948, | |
| "learning_rate": 9e-06, | |
| "loss": 0.67, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.014970059880239521, | |
| "grad_norm": 0.08075608313083649, | |
| "learning_rate": 1e-05, | |
| "loss": 0.512, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.016467065868263474, | |
| "grad_norm": 0.0809493139386177, | |
| "learning_rate": 9.9999859669361e-06, | |
| "loss": 0.4853, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.017964071856287425, | |
| "grad_norm": 0.0719803050160408, | |
| "learning_rate": 9.999943867823174e-06, | |
| "loss": 0.3089, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.019461077844311378, | |
| "grad_norm": 0.09597136080265045, | |
| "learning_rate": 9.999873702897528e-06, | |
| "loss": 0.5944, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.020958083832335328, | |
| "grad_norm": 0.05794112756848335, | |
| "learning_rate": 9.999775472553019e-06, | |
| "loss": 0.2852, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02245508982035928, | |
| "grad_norm": 0.08182471245527267, | |
| "learning_rate": 9.999649177341036e-06, | |
| "loss": 0.4125, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.023952095808383235, | |
| "grad_norm": 0.0892665907740593, | |
| "learning_rate": 9.999494817970498e-06, | |
| "loss": 0.5022, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.025449101796407185, | |
| "grad_norm": 0.11126008629798889, | |
| "learning_rate": 9.999312395307861e-06, | |
| "loss": 0.6561, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.02694610778443114, | |
| "grad_norm": 0.08598649501800537, | |
| "learning_rate": 9.999101910377107e-06, | |
| "loss": 0.423, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.02844311377245509, | |
| "grad_norm": 0.0690973624587059, | |
| "learning_rate": 9.998863364359734e-06, | |
| "loss": 0.2771, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.029940119760479042, | |
| "grad_norm": 0.10876584053039551, | |
| "learning_rate": 9.998596758594752e-06, | |
| "loss": 0.4564, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03143712574850299, | |
| "grad_norm": 0.08262284100055695, | |
| "learning_rate": 9.998302094578685e-06, | |
| "loss": 0.4124, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03293413173652695, | |
| "grad_norm": 0.1151677817106247, | |
| "learning_rate": 9.997979373965542e-06, | |
| "loss": 0.594, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0344311377245509, | |
| "grad_norm": 0.06878229230642319, | |
| "learning_rate": 9.99762859856683e-06, | |
| "loss": 0.367, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03592814371257485, | |
| "grad_norm": 0.09845534712076187, | |
| "learning_rate": 9.997249770351531e-06, | |
| "loss": 0.4928, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0374251497005988, | |
| "grad_norm": 0.11114847660064697, | |
| "learning_rate": 9.996842891446092e-06, | |
| "loss": 0.513, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.038922155688622756, | |
| "grad_norm": 0.0876867026090622, | |
| "learning_rate": 9.996407964134416e-06, | |
| "loss": 0.4446, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.040419161676646706, | |
| "grad_norm": 0.08587277680635452, | |
| "learning_rate": 9.995944990857848e-06, | |
| "loss": 0.4212, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.041916167664670656, | |
| "grad_norm": 0.1094743087887764, | |
| "learning_rate": 9.995453974215164e-06, | |
| "loss": 0.5395, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.04341317365269461, | |
| "grad_norm": 0.09168315678834915, | |
| "learning_rate": 9.994934916962547e-06, | |
| "loss": 0.5382, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.04491017964071856, | |
| "grad_norm": 0.0933656170964241, | |
| "learning_rate": 9.994387822013586e-06, | |
| "loss": 0.4935, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04640718562874251, | |
| "grad_norm": 0.09951834380626678, | |
| "learning_rate": 9.993812692439247e-06, | |
| "loss": 0.4884, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.04790419161676647, | |
| "grad_norm": 0.08234664052724838, | |
| "learning_rate": 9.99320953146786e-06, | |
| "loss": 0.3578, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.04940119760479042, | |
| "grad_norm": 0.0895133763551712, | |
| "learning_rate": 9.992578342485107e-06, | |
| "loss": 0.5871, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.05089820359281437, | |
| "grad_norm": 0.12492594122886658, | |
| "learning_rate": 9.991919129033994e-06, | |
| "loss": 0.5202, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.05239520958083832, | |
| "grad_norm": 0.09857629239559174, | |
| "learning_rate": 9.99123189481483e-06, | |
| "loss": 0.5833, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05389221556886228, | |
| "grad_norm": 0.13012069463729858, | |
| "learning_rate": 9.990516643685222e-06, | |
| "loss": 0.6599, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.05538922155688623, | |
| "grad_norm": 0.08107103407382965, | |
| "learning_rate": 9.98977337966003e-06, | |
| "loss": 0.4434, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.05688622754491018, | |
| "grad_norm": 0.08469574898481369, | |
| "learning_rate": 9.989002106911368e-06, | |
| "loss": 0.4689, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.058383233532934134, | |
| "grad_norm": 0.06690481305122375, | |
| "learning_rate": 9.988202829768562e-06, | |
| "loss": 0.4292, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.059880239520958084, | |
| "grad_norm": 0.07395423203706741, | |
| "learning_rate": 9.987375552718133e-06, | |
| "loss": 0.3295, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.061377245508982034, | |
| "grad_norm": 0.10279028117656708, | |
| "learning_rate": 9.986520280403775e-06, | |
| "loss": 0.5975, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.06287425149700598, | |
| "grad_norm": 0.08501698821783066, | |
| "learning_rate": 9.985637017626326e-06, | |
| "loss": 0.4974, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.06437125748502993, | |
| "grad_norm": 0.08820093423128128, | |
| "learning_rate": 9.984725769343737e-06, | |
| "loss": 0.5446, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0658682634730539, | |
| "grad_norm": 0.0912579894065857, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.4359, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.06736526946107785, | |
| "grad_norm": 0.06515525281429291, | |
| "learning_rate": 9.982819336880369e-06, | |
| "loss": 0.3298, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0688622754491018, | |
| "grad_norm": 0.09884097427129745, | |
| "learning_rate": 9.981824163400827e-06, | |
| "loss": 0.4986, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.07035928143712575, | |
| "grad_norm": 0.09501009434461594, | |
| "learning_rate": 9.980801025818556e-06, | |
| "loss": 0.5314, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.0718562874251497, | |
| "grad_norm": 0.09527381509542465, | |
| "learning_rate": 9.979749929876658e-06, | |
| "loss": 0.4653, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.07335329341317365, | |
| "grad_norm": 0.07869047671556473, | |
| "learning_rate": 9.978670881475173e-06, | |
| "loss": 0.3973, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.0748502994011976, | |
| "grad_norm": 0.0786195695400238, | |
| "learning_rate": 9.977563886671043e-06, | |
| "loss": 0.4364, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07634730538922156, | |
| "grad_norm": 0.08818861842155457, | |
| "learning_rate": 9.976428951678077e-06, | |
| "loss": 0.5149, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.07784431137724551, | |
| "grad_norm": 0.0732090175151825, | |
| "learning_rate": 9.975266082866923e-06, | |
| "loss": 0.3853, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.07934131736526946, | |
| "grad_norm": 0.07830777764320374, | |
| "learning_rate": 9.974075286765027e-06, | |
| "loss": 0.5106, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.08083832335329341, | |
| "grad_norm": 0.08423200249671936, | |
| "learning_rate": 9.972856570056594e-06, | |
| "loss": 0.4341, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.08233532934131736, | |
| "grad_norm": 0.12318050861358643, | |
| "learning_rate": 9.971609939582556e-06, | |
| "loss": 0.3335, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08383233532934131, | |
| "grad_norm": 0.10663409531116486, | |
| "learning_rate": 9.970335402340534e-06, | |
| "loss": 0.6584, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.08532934131736528, | |
| "grad_norm": 0.09056587517261505, | |
| "learning_rate": 9.969032965484789e-06, | |
| "loss": 0.4559, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.08682634730538923, | |
| "grad_norm": 0.08034185320138931, | |
| "learning_rate": 9.967702636326195e-06, | |
| "loss": 0.3497, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.08832335329341318, | |
| "grad_norm": 0.10354162752628326, | |
| "learning_rate": 9.96634442233219e-06, | |
| "loss": 0.6226, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.08982035928143713, | |
| "grad_norm": 0.0716947466135025, | |
| "learning_rate": 9.964958331126735e-06, | |
| "loss": 0.3743, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09131736526946108, | |
| "grad_norm": 0.06358286738395691, | |
| "learning_rate": 9.96354437049027e-06, | |
| "loss": 0.3273, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.09281437125748503, | |
| "grad_norm": 0.07961272448301315, | |
| "learning_rate": 9.96210254835968e-06, | |
| "loss": 0.4669, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.09431137724550898, | |
| "grad_norm": 0.07491420954465866, | |
| "learning_rate": 9.960632872828233e-06, | |
| "loss": 0.4228, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.09580838323353294, | |
| "grad_norm": 0.07608773559331894, | |
| "learning_rate": 9.959135352145552e-06, | |
| "loss": 0.447, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.09730538922155689, | |
| "grad_norm": 0.0806804671883583, | |
| "learning_rate": 9.957609994717559e-06, | |
| "loss": 0.5003, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.09880239520958084, | |
| "grad_norm": 0.08095613121986389, | |
| "learning_rate": 9.956056809106426e-06, | |
| "loss": 0.3929, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.10029940119760479, | |
| "grad_norm": 0.09194113314151764, | |
| "learning_rate": 9.954475804030539e-06, | |
| "loss": 0.5485, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.10179640718562874, | |
| "grad_norm": 0.08301645517349243, | |
| "learning_rate": 9.952866988364431e-06, | |
| "loss": 0.3857, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.10329341317365269, | |
| "grad_norm": 0.08787499368190765, | |
| "learning_rate": 9.95123037113875e-06, | |
| "loss": 0.5438, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.10479041916167664, | |
| "grad_norm": 0.08403324335813522, | |
| "learning_rate": 9.9495659615402e-06, | |
| "loss": 0.5825, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1062874251497006, | |
| "grad_norm": 0.07904759049415588, | |
| "learning_rate": 9.947873768911483e-06, | |
| "loss": 0.3771, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.10778443113772455, | |
| "grad_norm": 0.07713301479816437, | |
| "learning_rate": 9.946153802751257e-06, | |
| "loss": 0.4679, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.1092814371257485, | |
| "grad_norm": 0.07544517517089844, | |
| "learning_rate": 9.944406072714086e-06, | |
| "loss": 0.3567, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.11077844311377245, | |
| "grad_norm": 0.0836024358868599, | |
| "learning_rate": 9.942630588610368e-06, | |
| "loss": 0.4644, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1122754491017964, | |
| "grad_norm": 0.08232344686985016, | |
| "learning_rate": 9.940827360406297e-06, | |
| "loss": 0.4551, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11377245508982035, | |
| "grad_norm": 0.0864156037569046, | |
| "learning_rate": 9.938996398223802e-06, | |
| "loss": 0.4393, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.11526946107784432, | |
| "grad_norm": 0.06776531785726547, | |
| "learning_rate": 9.937137712340483e-06, | |
| "loss": 0.2865, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.11676646706586827, | |
| "grad_norm": 0.092351995408535, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.4635, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.11826347305389222, | |
| "grad_norm": 0.10527883470058441, | |
| "learning_rate": 9.933337211359833e-06, | |
| "loss": 0.4317, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.11976047904191617, | |
| "grad_norm": 0.06179358810186386, | |
| "learning_rate": 9.931395417595568e-06, | |
| "loss": 0.3493, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12125748502994012, | |
| "grad_norm": 0.07964818924665451, | |
| "learning_rate": 9.929425942796502e-06, | |
| "loss": 0.4229, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.12275449101796407, | |
| "grad_norm": 0.07189369946718216, | |
| "learning_rate": 9.927428798017738e-06, | |
| "loss": 0.2973, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.12425149700598802, | |
| "grad_norm": 0.07853538542985916, | |
| "learning_rate": 9.925403994469702e-06, | |
| "loss": 0.4828, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.12574850299401197, | |
| "grad_norm": 0.08836135268211365, | |
| "learning_rate": 9.92335154351807e-06, | |
| "loss": 0.544, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.12724550898203593, | |
| "grad_norm": 0.08305364102125168, | |
| "learning_rate": 9.921271456683716e-06, | |
| "loss": 0.3718, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.12874251497005987, | |
| "grad_norm": 0.1029515191912651, | |
| "learning_rate": 9.919163745642633e-06, | |
| "loss": 0.5272, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.13023952095808383, | |
| "grad_norm": 0.09859377145767212, | |
| "learning_rate": 9.91702842222588e-06, | |
| "loss": 0.6271, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1317365269461078, | |
| "grad_norm": 0.08649040758609772, | |
| "learning_rate": 9.91486549841951e-06, | |
| "loss": 0.3192, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.13323353293413173, | |
| "grad_norm": 0.07640533149242401, | |
| "learning_rate": 9.912674986364502e-06, | |
| "loss": 0.315, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.1347305389221557, | |
| "grad_norm": 0.0803142637014389, | |
| "learning_rate": 9.91045689835669e-06, | |
| "loss": 0.3875, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.13622754491017963, | |
| "grad_norm": 0.08381187170743942, | |
| "learning_rate": 9.908211246846708e-06, | |
| "loss": 0.2947, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.1377245508982036, | |
| "grad_norm": 0.06045156344771385, | |
| "learning_rate": 9.905938044439904e-06, | |
| "loss": 0.3409, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.13922155688622753, | |
| "grad_norm": 0.0911700651049614, | |
| "learning_rate": 9.903637303896272e-06, | |
| "loss": 0.5051, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.1407185628742515, | |
| "grad_norm": 0.08980625867843628, | |
| "learning_rate": 9.901309038130392e-06, | |
| "loss": 0.512, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.14221556886227546, | |
| "grad_norm": 0.1140458956360817, | |
| "learning_rate": 9.89895326021134e-06, | |
| "loss": 0.5091, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.1437125748502994, | |
| "grad_norm": 0.06530901044607162, | |
| "learning_rate": 9.896569983362632e-06, | |
| "loss": 0.1685, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.14520958083832336, | |
| "grad_norm": 0.09205546230077744, | |
| "learning_rate": 9.894159220962138e-06, | |
| "loss": 0.4877, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.1467065868263473, | |
| "grad_norm": 0.074747733771801, | |
| "learning_rate": 9.891720986542011e-06, | |
| "loss": 0.4084, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.14820359281437126, | |
| "grad_norm": 0.09208723902702332, | |
| "learning_rate": 9.889255293788613e-06, | |
| "loss": 0.4498, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.1497005988023952, | |
| "grad_norm": 0.06876664608716965, | |
| "learning_rate": 9.886762156542428e-06, | |
| "loss": 0.3346, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15119760479041916, | |
| "grad_norm": 0.09977904707193375, | |
| "learning_rate": 9.884241588798004e-06, | |
| "loss": 0.4219, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.15269461077844312, | |
| "grad_norm": 0.0814763605594635, | |
| "learning_rate": 9.881693604703853e-06, | |
| "loss": 0.4438, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.15419161676646706, | |
| "grad_norm": 0.07697126269340515, | |
| "learning_rate": 9.879118218562384e-06, | |
| "loss": 0.4969, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.15568862275449102, | |
| "grad_norm": 0.10940787941217422, | |
| "learning_rate": 9.876515444829822e-06, | |
| "loss": 0.5485, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.15718562874251496, | |
| "grad_norm": 0.09441516548395157, | |
| "learning_rate": 9.873885298116123e-06, | |
| "loss": 0.5654, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.15868263473053892, | |
| "grad_norm": 0.07836193591356277, | |
| "learning_rate": 9.871227793184893e-06, | |
| "loss": 0.4448, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1601796407185629, | |
| "grad_norm": 0.08903878927230835, | |
| "learning_rate": 9.868542944953304e-06, | |
| "loss": 0.4793, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.16167664670658682, | |
| "grad_norm": 0.0880504921078682, | |
| "learning_rate": 9.865830768492019e-06, | |
| "loss": 0.4365, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.1631736526946108, | |
| "grad_norm": 0.06745749711990356, | |
| "learning_rate": 9.863091279025095e-06, | |
| "loss": 0.2266, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.16467065868263472, | |
| "grad_norm": 0.137594074010849, | |
| "learning_rate": 9.860324491929905e-06, | |
| "loss": 0.3052, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1661676646706587, | |
| "grad_norm": 0.07614412158727646, | |
| "learning_rate": 9.857530422737045e-06, | |
| "loss": 0.3063, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.16766467065868262, | |
| "grad_norm": 0.07592976093292236, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.3864, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.1691616766467066, | |
| "grad_norm": 0.07913090288639069, | |
| "learning_rate": 9.851860500946342e-06, | |
| "loss": 0.4046, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.17065868263473055, | |
| "grad_norm": 0.08463918417692184, | |
| "learning_rate": 9.848984680175049e-06, | |
| "loss": 0.3976, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.1721556886227545, | |
| "grad_norm": 0.08824311941862106, | |
| "learning_rate": 9.846081640959008e-06, | |
| "loss": 0.3154, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.17365269461077845, | |
| "grad_norm": 0.07617625594139099, | |
| "learning_rate": 9.843151399593636e-06, | |
| "loss": 0.4338, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.1751497005988024, | |
| "grad_norm": 0.0766025111079216, | |
| "learning_rate": 9.840193972527037e-06, | |
| "loss": 0.336, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.17664670658682635, | |
| "grad_norm": 0.07912255823612213, | |
| "learning_rate": 9.837209376359918e-06, | |
| "loss": 0.3919, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1781437125748503, | |
| "grad_norm": 0.10425494611263275, | |
| "learning_rate": 9.834197627845488e-06, | |
| "loss": 0.3993, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.17964071856287425, | |
| "grad_norm": 0.09615318477153778, | |
| "learning_rate": 9.831158743889373e-06, | |
| "loss": 0.434, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.18113772455089822, | |
| "grad_norm": 0.0898306667804718, | |
| "learning_rate": 9.828092741549513e-06, | |
| "loss": 0.5143, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.18263473053892215, | |
| "grad_norm": 0.09334494918584824, | |
| "learning_rate": 9.82499963803607e-06, | |
| "loss": 0.4423, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.18413173652694612, | |
| "grad_norm": 0.09541749954223633, | |
| "learning_rate": 9.821879450711336e-06, | |
| "loss": 0.4826, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.18562874251497005, | |
| "grad_norm": 0.09210306406021118, | |
| "learning_rate": 9.81873219708962e-06, | |
| "loss": 0.512, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.18712574850299402, | |
| "grad_norm": 0.08098453283309937, | |
| "learning_rate": 9.815557894837171e-06, | |
| "loss": 0.2722, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.18862275449101795, | |
| "grad_norm": 0.10569058358669281, | |
| "learning_rate": 9.81235656177206e-06, | |
| "loss": 0.4803, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.19011976047904192, | |
| "grad_norm": 0.07749304175376892, | |
| "learning_rate": 9.809128215864096e-06, | |
| "loss": 0.2676, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.19161676646706588, | |
| "grad_norm": 0.08790621906518936, | |
| "learning_rate": 9.80587287523471e-06, | |
| "loss": 0.3458, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.19311377245508982, | |
| "grad_norm": 0.07159163057804108, | |
| "learning_rate": 9.802590558156863e-06, | |
| "loss": 0.2287, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.19461077844311378, | |
| "grad_norm": 0.08052971214056015, | |
| "learning_rate": 9.79928128305494e-06, | |
| "loss": 0.3894, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.19610778443113772, | |
| "grad_norm": 0.08126161992549896, | |
| "learning_rate": 9.795945068504654e-06, | |
| "loss": 0.3881, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.19760479041916168, | |
| "grad_norm": 0.1161087155342102, | |
| "learning_rate": 9.792581933232924e-06, | |
| "loss": 0.574, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.19910179640718562, | |
| "grad_norm": 0.13015466928482056, | |
| "learning_rate": 9.789191896117786e-06, | |
| "loss": 0.4662, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.20059880239520958, | |
| "grad_norm": 0.07646424323320389, | |
| "learning_rate": 9.78577497618829e-06, | |
| "loss": 0.3165, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.20209580838323354, | |
| "grad_norm": 0.07416396588087082, | |
| "learning_rate": 9.782331192624372e-06, | |
| "loss": 0.2962, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.20359281437125748, | |
| "grad_norm": 0.10298296809196472, | |
| "learning_rate": 9.778860564756769e-06, | |
| "loss": 0.4891, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.20508982035928144, | |
| "grad_norm": 0.10729490220546722, | |
| "learning_rate": 9.775363112066897e-06, | |
| "loss": 0.4275, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.20658682634730538, | |
| "grad_norm": 0.08321131765842438, | |
| "learning_rate": 9.771838854186748e-06, | |
| "loss": 0.3733, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.20808383233532934, | |
| "grad_norm": 0.0714513510465622, | |
| "learning_rate": 9.768287810898773e-06, | |
| "loss": 0.2682, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.20958083832335328, | |
| "grad_norm": 0.10221531242132187, | |
| "learning_rate": 9.764710002135784e-06, | |
| "loss": 0.434, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21107784431137724, | |
| "grad_norm": 0.09798867255449295, | |
| "learning_rate": 9.761105447980824e-06, | |
| "loss": 0.4254, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2125748502994012, | |
| "grad_norm": 0.0933586061000824, | |
| "learning_rate": 9.757474168667072e-06, | |
| "loss": 0.4123, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.21407185628742514, | |
| "grad_norm": 0.10116757452487946, | |
| "learning_rate": 9.753816184577715e-06, | |
| "loss": 0.4565, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2155688622754491, | |
| "grad_norm": 0.09178843349218369, | |
| "learning_rate": 9.750131516245844e-06, | |
| "loss": 0.3424, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.21706586826347304, | |
| "grad_norm": 0.12792278826236725, | |
| "learning_rate": 9.746420184354334e-06, | |
| "loss": 0.3526, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.218562874251497, | |
| "grad_norm": 0.10231195390224457, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.432, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.22005988023952097, | |
| "grad_norm": 0.10359717905521393, | |
| "learning_rate": 9.738917613372121e-06, | |
| "loss": 0.4257, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2215568862275449, | |
| "grad_norm": 0.10195307433605194, | |
| "learning_rate": 9.73512641639504e-06, | |
| "loss": 0.413, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.22305389221556887, | |
| "grad_norm": 0.09156057238578796, | |
| "learning_rate": 9.731308640085329e-06, | |
| "loss": 0.377, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.2245508982035928, | |
| "grad_norm": 0.10193546861410141, | |
| "learning_rate": 9.72746430587303e-06, | |
| "loss": 0.3818, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.22604790419161677, | |
| "grad_norm": 0.09497050940990448, | |
| "learning_rate": 9.723593435337252e-06, | |
| "loss": 0.4343, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2275449101796407, | |
| "grad_norm": 0.08387334644794464, | |
| "learning_rate": 9.719696050206072e-06, | |
| "loss": 0.3136, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.22904191616766467, | |
| "grad_norm": 0.1258871853351593, | |
| "learning_rate": 9.715772172356388e-06, | |
| "loss": 0.4261, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.23053892215568864, | |
| "grad_norm": 0.11211127787828445, | |
| "learning_rate": 9.711821823813812e-06, | |
| "loss": 0.5238, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.23203592814371257, | |
| "grad_norm": 0.10336726158857346, | |
| "learning_rate": 9.70784502675254e-06, | |
| "loss": 0.4502, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.23353293413173654, | |
| "grad_norm": 0.06710167229175568, | |
| "learning_rate": 9.703841803495234e-06, | |
| "loss": 0.1438, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.23502994011976047, | |
| "grad_norm": 0.07125142216682434, | |
| "learning_rate": 9.699812176512887e-06, | |
| "loss": 0.2077, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.23652694610778444, | |
| "grad_norm": 0.1243571862578392, | |
| "learning_rate": 9.695756168424703e-06, | |
| "loss": 0.4075, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.23802395209580837, | |
| "grad_norm": 0.09641040116548538, | |
| "learning_rate": 9.691673801997974e-06, | |
| "loss": 0.5389, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.23952095808383234, | |
| "grad_norm": 0.10448335111141205, | |
| "learning_rate": 9.68756510014794e-06, | |
| "loss": 0.2794, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2410179640718563, | |
| "grad_norm": 0.0995652824640274, | |
| "learning_rate": 9.683430085937672e-06, | |
| "loss": 0.4319, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.24251497005988024, | |
| "grad_norm": 0.09269218891859055, | |
| "learning_rate": 9.67926878257794e-06, | |
| "loss": 0.3553, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2440119760479042, | |
| "grad_norm": 0.11025439202785492, | |
| "learning_rate": 9.675081213427076e-06, | |
| "loss": 0.445, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.24550898203592814, | |
| "grad_norm": 0.0948467031121254, | |
| "learning_rate": 9.67086740199085e-06, | |
| "loss": 0.4348, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.2470059880239521, | |
| "grad_norm": 0.09545961767435074, | |
| "learning_rate": 9.666627371922335e-06, | |
| "loss": 0.364, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.24850299401197604, | |
| "grad_norm": 0.11892738938331604, | |
| "learning_rate": 9.66236114702178e-06, | |
| "loss": 0.4303, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.09273416548967361, | |
| "learning_rate": 9.658068751236464e-06, | |
| "loss": 0.3282, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.385356605052948, | |
| "eval_runtime": 96.1718, | |
| "eval_samples_per_second": 7.653, | |
| "eval_steps_per_second": 0.957, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.25149700598802394, | |
| "grad_norm": 0.13071858882904053, | |
| "learning_rate": 9.653750208660577e-06, | |
| "loss": 0.5143, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.25299401197604793, | |
| "grad_norm": 0.13147501647472382, | |
| "learning_rate": 9.649405543535067e-06, | |
| "loss": 0.4322, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.25449101796407186, | |
| "grad_norm": 0.12748856842517853, | |
| "learning_rate": 9.645034780247521e-06, | |
| "loss": 0.3911, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2559880239520958, | |
| "grad_norm": 0.11920400708913803, | |
| "learning_rate": 9.640637943332025e-06, | |
| "loss": 0.3082, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.25748502994011974, | |
| "grad_norm": 0.1189412921667099, | |
| "learning_rate": 9.636215057469009e-06, | |
| "loss": 0.3137, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.25898203592814373, | |
| "grad_norm": 0.11915755271911621, | |
| "learning_rate": 9.631766147485131e-06, | |
| "loss": 0.3861, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.26047904191616766, | |
| "grad_norm": 0.10477300733327866, | |
| "learning_rate": 9.627291238353127e-06, | |
| "loss": 0.4244, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2619760479041916, | |
| "grad_norm": 0.11633849143981934, | |
| "learning_rate": 9.622790355191672e-06, | |
| "loss": 0.4226, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2634730538922156, | |
| "grad_norm": 0.1252908855676651, | |
| "learning_rate": 9.618263523265238e-06, | |
| "loss": 0.4436, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.26497005988023953, | |
| "grad_norm": 0.10194990783929825, | |
| "learning_rate": 9.613710767983953e-06, | |
| "loss": 0.4307, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.26646706586826346, | |
| "grad_norm": 0.11449652910232544, | |
| "learning_rate": 9.609132114903458e-06, | |
| "loss": 0.4771, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.2679640718562874, | |
| "grad_norm": 0.08113080263137817, | |
| "learning_rate": 9.60452758972477e-06, | |
| "loss": 0.2791, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.2694610778443114, | |
| "grad_norm": 0.11075828969478607, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.389, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.27095808383233533, | |
| "grad_norm": 0.09724461287260056, | |
| "learning_rate": 9.595241026602836e-06, | |
| "loss": 0.2496, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.27245508982035926, | |
| "grad_norm": 0.09227737784385681, | |
| "learning_rate": 9.590559040787168e-06, | |
| "loss": 0.2457, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.27395209580838326, | |
| "grad_norm": 0.10152530670166016, | |
| "learning_rate": 9.585851287128157e-06, | |
| "loss": 0.4295, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.2754491017964072, | |
| "grad_norm": 0.12347866594791412, | |
| "learning_rate": 9.581117792051487e-06, | |
| "loss": 0.5124, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.27694610778443113, | |
| "grad_norm": 0.09101920574903488, | |
| "learning_rate": 9.576358582127334e-06, | |
| "loss": 0.3037, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.27844311377245506, | |
| "grad_norm": 0.11212054640054703, | |
| "learning_rate": 9.57157368407022e-06, | |
| "loss": 0.4074, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.27994011976047906, | |
| "grad_norm": 0.12494377791881561, | |
| "learning_rate": 9.56676312473885e-06, | |
| "loss": 0.3777, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.281437125748503, | |
| "grad_norm": 0.09690834581851959, | |
| "learning_rate": 9.561926931135985e-06, | |
| "loss": 0.3416, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.28293413173652693, | |
| "grad_norm": 0.0795978531241417, | |
| "learning_rate": 9.557065130408267e-06, | |
| "loss": 0.214, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.2844311377245509, | |
| "grad_norm": 0.11251098662614822, | |
| "learning_rate": 9.552177749846083e-06, | |
| "loss": 0.3523, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.28592814371257486, | |
| "grad_norm": 0.11772778630256653, | |
| "learning_rate": 9.5472648168834e-06, | |
| "loss": 0.3147, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.2874251497005988, | |
| "grad_norm": 0.1310431957244873, | |
| "learning_rate": 9.542326359097619e-06, | |
| "loss": 0.3656, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.28892215568862273, | |
| "grad_norm": 0.10823767632246017, | |
| "learning_rate": 9.537362404209419e-06, | |
| "loss": 0.4717, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2904191616766467, | |
| "grad_norm": 0.1049995943903923, | |
| "learning_rate": 9.532372980082598e-06, | |
| "loss": 0.3117, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.29191616766467066, | |
| "grad_norm": 0.10875418782234192, | |
| "learning_rate": 9.527358114723917e-06, | |
| "loss": 0.4098, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2934131736526946, | |
| "grad_norm": 0.10218145698308945, | |
| "learning_rate": 9.522317836282949e-06, | |
| "loss": 0.3279, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.2949101796407186, | |
| "grad_norm": 0.13674795627593994, | |
| "learning_rate": 9.517252173051912e-06, | |
| "loss": 0.4229, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.2964071856287425, | |
| "grad_norm": 0.18106026947498322, | |
| "learning_rate": 9.512161153465518e-06, | |
| "loss": 0.4479, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.29790419161676646, | |
| "grad_norm": 0.1282949596643448, | |
| "learning_rate": 9.507044806100806e-06, | |
| "loss": 0.5003, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.2994011976047904, | |
| "grad_norm": 0.10633736848831177, | |
| "learning_rate": 9.501903159676993e-06, | |
| "loss": 0.3367, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3008982035928144, | |
| "grad_norm": 0.14237818121910095, | |
| "learning_rate": 9.496736243055293e-06, | |
| "loss": 0.5059, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.3023952095808383, | |
| "grad_norm": 0.14547429978847504, | |
| "learning_rate": 9.491544085238778e-06, | |
| "loss": 0.4906, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.30389221556886226, | |
| "grad_norm": 0.12434766441583633, | |
| "learning_rate": 9.486326715372201e-06, | |
| "loss": 0.4527, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.30538922155688625, | |
| "grad_norm": 0.11812443286180496, | |
| "learning_rate": 9.481084162741835e-06, | |
| "loss": 0.4048, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.3068862275449102, | |
| "grad_norm": 0.1250404566526413, | |
| "learning_rate": 9.475816456775313e-06, | |
| "loss": 0.48, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3083832335329341, | |
| "grad_norm": 0.11976836621761322, | |
| "learning_rate": 9.470523627041452e-06, | |
| "loss": 0.3601, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.30988023952095806, | |
| "grad_norm": 0.10545119643211365, | |
| "learning_rate": 9.465205703250105e-06, | |
| "loss": 0.332, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.31137724550898205, | |
| "grad_norm": 0.12188291549682617, | |
| "learning_rate": 9.459862715251973e-06, | |
| "loss": 0.4266, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.312874251497006, | |
| "grad_norm": 0.17091204226016998, | |
| "learning_rate": 9.454494693038455e-06, | |
| "loss": 0.3737, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.3143712574850299, | |
| "grad_norm": 0.10543544590473175, | |
| "learning_rate": 9.44910166674147e-06, | |
| "loss": 0.291, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3158682634730539, | |
| "grad_norm": 0.12531983852386475, | |
| "learning_rate": 9.44368366663329e-06, | |
| "loss": 0.3832, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.31736526946107785, | |
| "grad_norm": 0.12134228646755219, | |
| "learning_rate": 9.438240723126376e-06, | |
| "loss": 0.2531, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.3188622754491018, | |
| "grad_norm": 0.12980221211910248, | |
| "learning_rate": 9.43277286677319e-06, | |
| "loss": 0.3047, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.3203592814371258, | |
| "grad_norm": 0.13721132278442383, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.3008, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.3218562874251497, | |
| "grad_norm": 0.12022245675325394, | |
| "learning_rate": 9.421762538436933e-06, | |
| "loss": 0.4336, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.32335329341317365, | |
| "grad_norm": 0.13329699635505676, | |
| "learning_rate": 9.416220128257317e-06, | |
| "loss": 0.4004, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.3248502994011976, | |
| "grad_norm": 0.10903549939393997, | |
| "learning_rate": 9.410652928837998e-06, | |
| "loss": 0.2668, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3263473053892216, | |
| "grad_norm": 0.11044671386480331, | |
| "learning_rate": 9.405060971428924e-06, | |
| "loss": 0.2933, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.3278443113772455, | |
| "grad_norm": 0.13679952919483185, | |
| "learning_rate": 9.399444287419012e-06, | |
| "loss": 0.4049, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.32934131736526945, | |
| "grad_norm": 0.1408192664384842, | |
| "learning_rate": 9.393802908335978e-06, | |
| "loss": 0.4256, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.33083832335329344, | |
| "grad_norm": 0.14757537841796875, | |
| "learning_rate": 9.388136865846153e-06, | |
| "loss": 0.4321, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.3323353293413174, | |
| "grad_norm": 0.09091170132160187, | |
| "learning_rate": 9.382446191754313e-06, | |
| "loss": 0.2193, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.3338323353293413, | |
| "grad_norm": 0.10384230315685272, | |
| "learning_rate": 9.376730918003495e-06, | |
| "loss": 0.2178, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.33532934131736525, | |
| "grad_norm": 0.163813054561615, | |
| "learning_rate": 9.370991076674821e-06, | |
| "loss": 0.3887, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.33682634730538924, | |
| "grad_norm": 0.129890576004982, | |
| "learning_rate": 9.36522669998731e-06, | |
| "loss": 0.3996, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.3383233532934132, | |
| "grad_norm": 0.12134243547916412, | |
| "learning_rate": 9.359437820297716e-06, | |
| "loss": 0.3961, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.3398203592814371, | |
| "grad_norm": 0.142124205827713, | |
| "learning_rate": 9.353624470100321e-06, | |
| "loss": 0.3307, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.3413173652694611, | |
| "grad_norm": 0.11390755325555801, | |
| "learning_rate": 9.347786682026774e-06, | |
| "loss": 0.2885, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.34281437125748504, | |
| "grad_norm": 0.1425858587026596, | |
| "learning_rate": 9.341924488845892e-06, | |
| "loss": 0.3786, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.344311377245509, | |
| "grad_norm": 0.12148135900497437, | |
| "learning_rate": 9.336037923463494e-06, | |
| "loss": 0.4657, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3458083832335329, | |
| "grad_norm": 0.13119304180145264, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.4115, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.3473053892215569, | |
| "grad_norm": 0.12425543367862701, | |
| "learning_rate": 9.324191808401235e-06, | |
| "loss": 0.3414, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.34880239520958084, | |
| "grad_norm": 0.0954410657286644, | |
| "learning_rate": 9.31823232521629e-06, | |
| "loss": 0.1629, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.3502994011976048, | |
| "grad_norm": 0.14815054833889008, | |
| "learning_rate": 9.312248602819284e-06, | |
| "loss": 0.4772, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.35179640718562877, | |
| "grad_norm": 0.13234567642211914, | |
| "learning_rate": 9.306240674798203e-06, | |
| "loss": 0.3067, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.3532934131736527, | |
| "grad_norm": 0.10531976073980331, | |
| "learning_rate": 9.300208574876897e-06, | |
| "loss": 0.2415, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.35479041916167664, | |
| "grad_norm": 0.11296534538269043, | |
| "learning_rate": 9.294152336914907e-06, | |
| "loss": 0.2706, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3562874251497006, | |
| "grad_norm": 0.16313917934894562, | |
| "learning_rate": 9.288071994907262e-06, | |
| "loss": 0.4391, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.35778443113772457, | |
| "grad_norm": 0.13125663995742798, | |
| "learning_rate": 9.281967582984292e-06, | |
| "loss": 0.4955, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.3592814371257485, | |
| "grad_norm": 0.1586073786020279, | |
| "learning_rate": 9.275839135411439e-06, | |
| "loss": 0.4565, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.36077844311377244, | |
| "grad_norm": 0.1046639159321785, | |
| "learning_rate": 9.269686686589063e-06, | |
| "loss": 0.2976, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.36227544910179643, | |
| "grad_norm": 0.15066610276699066, | |
| "learning_rate": 9.263510271052243e-06, | |
| "loss": 0.4158, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.36377245508982037, | |
| "grad_norm": 0.16172873973846436, | |
| "learning_rate": 9.257309923470596e-06, | |
| "loss": 0.3956, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.3652694610778443, | |
| "grad_norm": 0.18391874432563782, | |
| "learning_rate": 9.251085678648072e-06, | |
| "loss": 0.5078, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.36676646706586824, | |
| "grad_norm": 0.1277581751346588, | |
| "learning_rate": 9.244837571522758e-06, | |
| "loss": 0.2947, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.36826347305389223, | |
| "grad_norm": 0.16508063673973083, | |
| "learning_rate": 9.238565637166692e-06, | |
| "loss": 0.4845, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.36976047904191617, | |
| "grad_norm": 0.13770635426044464, | |
| "learning_rate": 9.232269910785651e-06, | |
| "loss": 0.413, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.3712574850299401, | |
| "grad_norm": 0.16841325163841248, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.4875, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.3727544910179641, | |
| "grad_norm": 0.16400140523910522, | |
| "learning_rate": 9.219607223439343e-06, | |
| "loss": 0.436, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.37425149700598803, | |
| "grad_norm": 0.18191541731357574, | |
| "learning_rate": 9.213240333552589e-06, | |
| "loss": 0.4318, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.37574850299401197, | |
| "grad_norm": 0.1227971538901329, | |
| "learning_rate": 9.206849793797508e-06, | |
| "loss": 0.3745, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.3772455089820359, | |
| "grad_norm": 0.13938282430171967, | |
| "learning_rate": 9.200435640045637e-06, | |
| "loss": 0.2679, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.3787425149700599, | |
| "grad_norm": 0.190107524394989, | |
| "learning_rate": 9.193997908301069e-06, | |
| "loss": 0.4332, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.38023952095808383, | |
| "grad_norm": 0.14340659976005554, | |
| "learning_rate": 9.187536634700244e-06, | |
| "loss": 0.3839, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.38173652694610777, | |
| "grad_norm": 0.14708517491817474, | |
| "learning_rate": 9.181051855511749e-06, | |
| "loss": 0.4324, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.38323353293413176, | |
| "grad_norm": 0.1463197022676468, | |
| "learning_rate": 9.174543607136111e-06, | |
| "loss": 0.4632, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.3847305389221557, | |
| "grad_norm": 0.1626986563205719, | |
| "learning_rate": 9.168011926105598e-06, | |
| "loss": 0.4655, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.38622754491017963, | |
| "grad_norm": 0.13527408242225647, | |
| "learning_rate": 9.161456849084007e-06, | |
| "loss": 0.3422, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.38772455089820357, | |
| "grad_norm": 0.1195373684167862, | |
| "learning_rate": 9.154878412866465e-06, | |
| "loss": 0.2734, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.38922155688622756, | |
| "grad_norm": 0.16617508232593536, | |
| "learning_rate": 9.14827665437922e-06, | |
| "loss": 0.3768, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3907185628742515, | |
| "grad_norm": 0.09267516434192657, | |
| "learning_rate": 9.141651610679427e-06, | |
| "loss": 0.1452, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.39221556886227543, | |
| "grad_norm": 0.13747261464595795, | |
| "learning_rate": 9.135003318954954e-06, | |
| "loss": 0.3527, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.3937125748502994, | |
| "grad_norm": 0.1347353607416153, | |
| "learning_rate": 9.12833181652416e-06, | |
| "loss": 0.3078, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.39520958083832336, | |
| "grad_norm": 0.10927855968475342, | |
| "learning_rate": 9.121637140835696e-06, | |
| "loss": 0.2403, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.3967065868263473, | |
| "grad_norm": 0.14766469597816467, | |
| "learning_rate": 9.114919329468283e-06, | |
| "loss": 0.4768, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.39820359281437123, | |
| "grad_norm": 0.17013344168663025, | |
| "learning_rate": 9.108178420130514e-06, | |
| "loss": 0.4459, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.3997005988023952, | |
| "grad_norm": 0.16462120413780212, | |
| "learning_rate": 9.101414450660633e-06, | |
| "loss": 0.3027, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.40119760479041916, | |
| "grad_norm": 0.15908250212669373, | |
| "learning_rate": 9.094627459026326e-06, | |
| "loss": 0.4534, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.4026946107784431, | |
| "grad_norm": 0.11128409951925278, | |
| "learning_rate": 9.087817483324507e-06, | |
| "loss": 0.2694, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.4041916167664671, | |
| "grad_norm": 0.15646331012248993, | |
| "learning_rate": 9.08098456178111e-06, | |
| "loss": 0.3605, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.405688622754491, | |
| "grad_norm": 0.17366009950637817, | |
| "learning_rate": 9.074128732750859e-06, | |
| "loss": 0.3855, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.40718562874251496, | |
| "grad_norm": 0.204384908080101, | |
| "learning_rate": 9.067250034717072e-06, | |
| "loss": 0.4084, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.4086826347305389, | |
| "grad_norm": 0.11996804177761078, | |
| "learning_rate": 9.060348506291432e-06, | |
| "loss": 0.2457, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.4101796407185629, | |
| "grad_norm": 0.18958471715450287, | |
| "learning_rate": 9.053424186213776e-06, | |
| "loss": 0.3692, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.4116766467065868, | |
| "grad_norm": 0.16402485966682434, | |
| "learning_rate": 9.046477113351871e-06, | |
| "loss": 0.3062, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.41317365269461076, | |
| "grad_norm": 0.1784193068742752, | |
| "learning_rate": 9.039507326701207e-06, | |
| "loss": 0.4527, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.41467065868263475, | |
| "grad_norm": 0.12558554112911224, | |
| "learning_rate": 9.032514865384767e-06, | |
| "loss": 0.2494, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.4161676646706587, | |
| "grad_norm": 0.1545354127883911, | |
| "learning_rate": 9.025499768652817e-06, | |
| "loss": 0.4017, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.4176646706586826, | |
| "grad_norm": 0.18937596678733826, | |
| "learning_rate": 9.018462075882673e-06, | |
| "loss": 0.4244, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.41916167664670656, | |
| "grad_norm": 0.1287037581205368, | |
| "learning_rate": 9.011401826578492e-06, | |
| "loss": 0.2686, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.42065868263473055, | |
| "grad_norm": 0.11539726704359055, | |
| "learning_rate": 9.00431906037105e-06, | |
| "loss": 0.3171, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.4221556886227545, | |
| "grad_norm": 0.15977144241333008, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.3253, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.4236526946107784, | |
| "grad_norm": 0.1216585785150528, | |
| "learning_rate": 8.990086136401199e-06, | |
| "loss": 0.2122, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.4251497005988024, | |
| "grad_norm": 0.13447438180446625, | |
| "learning_rate": 8.982936058531403e-06, | |
| "loss": 0.3615, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.42664670658682635, | |
| "grad_norm": 0.12826186418533325, | |
| "learning_rate": 8.975763623543121e-06, | |
| "loss": 0.368, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.4281437125748503, | |
| "grad_norm": 0.12770900130271912, | |
| "learning_rate": 8.968568871696847e-06, | |
| "loss": 0.2612, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.4296407185628742, | |
| "grad_norm": 0.23060199618339539, | |
| "learning_rate": 8.961351843378349e-06, | |
| "loss": 0.3835, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.4311377245508982, | |
| "grad_norm": 0.11921186745166779, | |
| "learning_rate": 8.95411257909843e-06, | |
| "loss": 0.2427, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.43263473053892215, | |
| "grad_norm": 0.13262028992176056, | |
| "learning_rate": 8.946851119492717e-06, | |
| "loss": 0.25, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.4341317365269461, | |
| "grad_norm": 0.14844539761543274, | |
| "learning_rate": 8.939567505321418e-06, | |
| "loss": 0.3788, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4356287425149701, | |
| "grad_norm": 0.23767727613449097, | |
| "learning_rate": 8.932261777469105e-06, | |
| "loss": 0.3833, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.437125748502994, | |
| "grad_norm": 0.18635216355323792, | |
| "learning_rate": 8.924933976944474e-06, | |
| "loss": 0.3438, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.43862275449101795, | |
| "grad_norm": 0.2509608566761017, | |
| "learning_rate": 8.917584144880124e-06, | |
| "loss": 0.3578, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.44011976047904194, | |
| "grad_norm": 0.2058945745229721, | |
| "learning_rate": 8.910212322532317e-06, | |
| "loss": 0.3542, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.4416167664670659, | |
| "grad_norm": 0.17262375354766846, | |
| "learning_rate": 8.902818551280758e-06, | |
| "loss": 0.4504, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4431137724550898, | |
| "grad_norm": 0.1604480743408203, | |
| "learning_rate": 8.895402872628352e-06, | |
| "loss": 0.2744, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.44461077844311375, | |
| "grad_norm": 0.14820334315299988, | |
| "learning_rate": 8.887965328200975e-06, | |
| "loss": 0.3349, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.44610778443113774, | |
| "grad_norm": 0.15106992423534393, | |
| "learning_rate": 8.880505959747245e-06, | |
| "loss": 0.2446, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.4476047904191617, | |
| "grad_norm": 0.10235021263360977, | |
| "learning_rate": 8.873024809138272e-06, | |
| "loss": 0.2411, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.4491017964071856, | |
| "grad_norm": 0.1582055240869522, | |
| "learning_rate": 8.86552191836745e-06, | |
| "loss": 0.4247, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4505988023952096, | |
| "grad_norm": 0.1799710988998413, | |
| "learning_rate": 8.857997329550195e-06, | |
| "loss": 0.4035, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.45209580838323354, | |
| "grad_norm": 0.17223840951919556, | |
| "learning_rate": 8.850451084923717e-06, | |
| "loss": 0.2541, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.4535928143712575, | |
| "grad_norm": 0.21316657960414886, | |
| "learning_rate": 8.842883226846792e-06, | |
| "loss": 0.5618, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.4550898203592814, | |
| "grad_norm": 0.19103510677814484, | |
| "learning_rate": 8.835293797799517e-06, | |
| "loss": 0.3129, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.4565868263473054, | |
| "grad_norm": 0.14965221285820007, | |
| "learning_rate": 8.827682840383065e-06, | |
| "loss": 0.4006, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.45808383233532934, | |
| "grad_norm": 0.2670743465423584, | |
| "learning_rate": 8.82005039731946e-06, | |
| "loss": 0.4408, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.4595808383233533, | |
| "grad_norm": 0.1590014398097992, | |
| "learning_rate": 8.812396511451324e-06, | |
| "loss": 0.3762, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.46107784431137727, | |
| "grad_norm": 0.27246150374412537, | |
| "learning_rate": 8.804721225741646e-06, | |
| "loss": 0.4455, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.4625748502994012, | |
| "grad_norm": 0.1429322361946106, | |
| "learning_rate": 8.797024583273536e-06, | |
| "loss": 0.1568, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.46407185628742514, | |
| "grad_norm": 0.14144589006900787, | |
| "learning_rate": 8.789306627249985e-06, | |
| "loss": 0.2598, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4655688622754491, | |
| "grad_norm": 0.1922285109758377, | |
| "learning_rate": 8.781567400993617e-06, | |
| "loss": 0.3722, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.46706586826347307, | |
| "grad_norm": 0.15292677283287048, | |
| "learning_rate": 8.77380694794646e-06, | |
| "loss": 0.3738, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.468562874251497, | |
| "grad_norm": 0.16995134949684143, | |
| "learning_rate": 8.766025311669685e-06, | |
| "loss": 0.4324, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.47005988023952094, | |
| "grad_norm": 0.2128669172525406, | |
| "learning_rate": 8.75822253584337e-06, | |
| "loss": 0.4218, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.47155688622754494, | |
| "grad_norm": 0.13214029371738434, | |
| "learning_rate": 8.75039866426626e-06, | |
| "loss": 0.3243, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.47305389221556887, | |
| "grad_norm": 0.17079909145832062, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.3964, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.4745508982035928, | |
| "grad_norm": 0.12262991815805435, | |
| "learning_rate": 8.734687809646437e-06, | |
| "loss": 0.2449, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.47604790419161674, | |
| "grad_norm": 0.14602302014827728, | |
| "learning_rate": 8.726800914792296e-06, | |
| "loss": 0.2784, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.47754491017964074, | |
| "grad_norm": 0.22383739054203033, | |
| "learning_rate": 8.718893100564002e-06, | |
| "loss": 0.337, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.47904191616766467, | |
| "grad_norm": 0.102297842502594, | |
| "learning_rate": 8.710964411349902e-06, | |
| "loss": 0.2467, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4805389221556886, | |
| "grad_norm": 0.1584213376045227, | |
| "learning_rate": 8.703014891655518e-06, | |
| "loss": 0.4198, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.4820359281437126, | |
| "grad_norm": 0.1733960062265396, | |
| "learning_rate": 8.695044586103297e-06, | |
| "loss": 0.3446, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.48353293413173654, | |
| "grad_norm": 0.18755660951137543, | |
| "learning_rate": 8.687053539432358e-06, | |
| "loss": 0.3846, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.48502994011976047, | |
| "grad_norm": 0.13098154962062836, | |
| "learning_rate": 8.679041796498253e-06, | |
| "loss": 0.2938, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.4865269461077844, | |
| "grad_norm": 0.14376209676265717, | |
| "learning_rate": 8.6710094022727e-06, | |
| "loss": 0.2437, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4880239520958084, | |
| "grad_norm": 0.17319519817829132, | |
| "learning_rate": 8.66295640184334e-06, | |
| "loss": 0.3055, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.48952095808383234, | |
| "grad_norm": 0.168416827917099, | |
| "learning_rate": 8.65488284041348e-06, | |
| "loss": 0.3925, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.49101796407185627, | |
| "grad_norm": 0.16502583026885986, | |
| "learning_rate": 8.646788763301842e-06, | |
| "loss": 0.3697, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.49251497005988026, | |
| "grad_norm": 0.13545477390289307, | |
| "learning_rate": 8.638674215942307e-06, | |
| "loss": 0.3017, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.4940119760479042, | |
| "grad_norm": 0.14142753183841705, | |
| "learning_rate": 8.630539243883659e-06, | |
| "loss": 0.3287, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.49550898203592814, | |
| "grad_norm": 0.17961451411247253, | |
| "learning_rate": 8.62238389278933e-06, | |
| "loss": 0.3309, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.49700598802395207, | |
| "grad_norm": 0.14708763360977173, | |
| "learning_rate": 8.61420820843715e-06, | |
| "loss": 0.3669, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.49850299401197606, | |
| "grad_norm": 0.21809980273246765, | |
| "learning_rate": 8.606012236719073e-06, | |
| "loss": 0.31, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.22139891982078552, | |
| "learning_rate": 8.59779602364094e-06, | |
| "loss": 0.4451, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.3353196680545807, | |
| "eval_runtime": 96.09, | |
| "eval_samples_per_second": 7.659, | |
| "eval_steps_per_second": 0.957, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.5014970059880239, | |
| "grad_norm": 0.13361559808254242, | |
| "learning_rate": 8.58955961532221e-06, | |
| "loss": 0.198, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.5029940119760479, | |
| "grad_norm": 0.17649687826633453, | |
| "learning_rate": 8.581303057995697e-06, | |
| "loss": 0.3268, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.5044910179640718, | |
| "grad_norm": 0.1601666957139969, | |
| "learning_rate": 8.573026398007323e-06, | |
| "loss": 0.4069, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.5059880239520959, | |
| "grad_norm": 0.189590185880661, | |
| "learning_rate": 8.564729681815846e-06, | |
| "loss": 0.3302, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.5074850299401198, | |
| "grad_norm": 0.18572555482387543, | |
| "learning_rate": 8.556412955992604e-06, | |
| "loss": 0.4025, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.5089820359281437, | |
| "grad_norm": 0.1558508276939392, | |
| "learning_rate": 8.548076267221258e-06, | |
| "loss": 0.396, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5104790419161677, | |
| "grad_norm": 0.19267967343330383, | |
| "learning_rate": 8.539719662297519e-06, | |
| "loss": 0.418, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.5119760479041916, | |
| "grad_norm": 0.13435742259025574, | |
| "learning_rate": 8.531343188128896e-06, | |
| "loss": 0.266, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.5134730538922155, | |
| "grad_norm": 0.2617822289466858, | |
| "learning_rate": 8.52294689173443e-06, | |
| "loss": 0.2821, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.5149700598802395, | |
| "grad_norm": 0.17945440113544464, | |
| "learning_rate": 8.514530820244427e-06, | |
| "loss": 0.3387, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.5164670658682635, | |
| "grad_norm": 0.13977111876010895, | |
| "learning_rate": 8.506095020900192e-06, | |
| "loss": 0.3336, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5179640718562875, | |
| "grad_norm": 0.17441023886203766, | |
| "learning_rate": 8.497639541053769e-06, | |
| "loss": 0.3207, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.5194610778443114, | |
| "grad_norm": 0.16159437596797943, | |
| "learning_rate": 8.489164428167677e-06, | |
| "loss": 0.395, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.5209580838323353, | |
| "grad_norm": 0.14497527480125427, | |
| "learning_rate": 8.480669729814635e-06, | |
| "loss": 0.3077, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.5224550898203593, | |
| "grad_norm": 0.17960812151432037, | |
| "learning_rate": 8.472155493677299e-06, | |
| "loss": 0.3512, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.5239520958083832, | |
| "grad_norm": 0.11662229150533676, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.1755, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5254491017964071, | |
| "grad_norm": 0.17493830621242523, | |
| "learning_rate": 8.455068599328462e-06, | |
| "loss": 0.4597, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.5269461077844312, | |
| "grad_norm": 0.15957583487033844, | |
| "learning_rate": 8.446496037029555e-06, | |
| "loss": 0.3225, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.5284431137724551, | |
| "grad_norm": 0.18921981751918793, | |
| "learning_rate": 8.437904128770999e-06, | |
| "loss": 0.3952, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.5299401197604791, | |
| "grad_norm": 0.15334290266036987, | |
| "learning_rate": 8.429292922781115e-06, | |
| "loss": 0.4056, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.531437125748503, | |
| "grad_norm": 0.16442593932151794, | |
| "learning_rate": 8.420662467396548e-06, | |
| "loss": 0.3826, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5329341317365269, | |
| "grad_norm": 0.17233920097351074, | |
| "learning_rate": 8.412012811061985e-06, | |
| "loss": 0.2516, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.5344311377245509, | |
| "grad_norm": 0.16326873004436493, | |
| "learning_rate": 8.403344002329901e-06, | |
| "loss": 0.3362, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.5359281437125748, | |
| "grad_norm": 0.20600523054599762, | |
| "learning_rate": 8.394656089860274e-06, | |
| "loss": 0.3266, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.5374251497005988, | |
| "grad_norm": 0.19378286600112915, | |
| "learning_rate": 8.385949122420318e-06, | |
| "loss": 0.3118, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.5389221556886228, | |
| "grad_norm": 0.17837578058242798, | |
| "learning_rate": 8.377223148884202e-06, | |
| "loss": 0.3373, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5404191616766467, | |
| "grad_norm": 0.17606143653392792, | |
| "learning_rate": 8.368478218232787e-06, | |
| "loss": 0.2534, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.5419161676646707, | |
| "grad_norm": 0.15157122910022736, | |
| "learning_rate": 8.359714379553338e-06, | |
| "loss": 0.1722, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.5434131736526946, | |
| "grad_norm": 0.23432698845863342, | |
| "learning_rate": 8.350931682039262e-06, | |
| "loss": 0.4558, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.5449101796407185, | |
| "grad_norm": 0.1647920310497284, | |
| "learning_rate": 8.342130174989819e-06, | |
| "loss": 0.3593, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5464071856287425, | |
| "grad_norm": 0.17406417429447174, | |
| "learning_rate": 8.333309907809852e-06, | |
| "loss": 0.2793, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5479041916167665, | |
| "grad_norm": 0.1906813532114029, | |
| "learning_rate": 8.324470930009514e-06, | |
| "loss": 0.4096, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.5494011976047904, | |
| "grad_norm": 0.1391858011484146, | |
| "learning_rate": 8.315613291203977e-06, | |
| "loss": 0.3213, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.5508982035928144, | |
| "grad_norm": 0.20214834809303284, | |
| "learning_rate": 8.306737041113169e-06, | |
| "loss": 0.3283, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.5523952095808383, | |
| "grad_norm": 0.2006218433380127, | |
| "learning_rate": 8.29784222956148e-06, | |
| "loss": 0.3451, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.5538922155688623, | |
| "grad_norm": 0.21113942563533783, | |
| "learning_rate": 8.288928906477497e-06, | |
| "loss": 0.3681, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5553892215568862, | |
| "grad_norm": 0.1564386785030365, | |
| "learning_rate": 8.279997121893713e-06, | |
| "loss": 0.2562, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.5568862275449101, | |
| "grad_norm": 0.18803362548351288, | |
| "learning_rate": 8.271046925946247e-06, | |
| "loss": 0.2923, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.5583832335329342, | |
| "grad_norm": 0.15986989438533783, | |
| "learning_rate": 8.262078368874566e-06, | |
| "loss": 0.3348, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.5598802395209581, | |
| "grad_norm": 0.18634189665317535, | |
| "learning_rate": 8.25309150102121e-06, | |
| "loss": 0.3538, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.561377245508982, | |
| "grad_norm": 0.18781672418117523, | |
| "learning_rate": 8.244086372831492e-06, | |
| "loss": 0.3305, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.562874251497006, | |
| "grad_norm": 0.22481568157672882, | |
| "learning_rate": 8.235063034853228e-06, | |
| "loss": 0.3972, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.5643712574850299, | |
| "grad_norm": 0.1664562225341797, | |
| "learning_rate": 8.226021537736449e-06, | |
| "loss": 0.5461, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.5658682634730539, | |
| "grad_norm": 0.133816659450531, | |
| "learning_rate": 8.216961932233118e-06, | |
| "loss": 0.1557, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.5673652694610778, | |
| "grad_norm": 0.23958347737789154, | |
| "learning_rate": 8.207884269196845e-06, | |
| "loss": 0.4871, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.5688622754491018, | |
| "grad_norm": 0.1876394897699356, | |
| "learning_rate": 8.198788599582596e-06, | |
| "loss": 0.2644, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5703592814371258, | |
| "grad_norm": 0.18707850575447083, | |
| "learning_rate": 8.189674974446423e-06, | |
| "loss": 0.3008, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.5718562874251497, | |
| "grad_norm": 0.21212701499462128, | |
| "learning_rate": 8.180543444945154e-06, | |
| "loss": 0.4367, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.5733532934131736, | |
| "grad_norm": 0.22406694293022156, | |
| "learning_rate": 8.171394062336127e-06, | |
| "loss": 0.3987, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.5748502994011976, | |
| "grad_norm": 0.1596011221408844, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.3971, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.5763473053892215, | |
| "grad_norm": 0.19374164938926697, | |
| "learning_rate": 8.153041943324912e-06, | |
| "loss": 0.2919, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5778443113772455, | |
| "grad_norm": 0.17940644919872284, | |
| "learning_rate": 8.143839309937307e-06, | |
| "loss": 0.268, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.5793413173652695, | |
| "grad_norm": 0.2111639380455017, | |
| "learning_rate": 8.134619029470535e-06, | |
| "loss": 0.3569, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.5808383233532934, | |
| "grad_norm": 0.19277545809745789, | |
| "learning_rate": 8.125381153680103e-06, | |
| "loss": 0.5122, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.5823353293413174, | |
| "grad_norm": 0.1731099933385849, | |
| "learning_rate": 8.116125734420297e-06, | |
| "loss": 0.3467, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.5838323353293413, | |
| "grad_norm": 0.1366906315088272, | |
| "learning_rate": 8.10685282364387e-06, | |
| "loss": 0.2594, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5853293413173652, | |
| "grad_norm": 0.22104769945144653, | |
| "learning_rate": 8.097562473401764e-06, | |
| "loss": 0.3096, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.5868263473053892, | |
| "grad_norm": 0.17460548877716064, | |
| "learning_rate": 8.088254735842808e-06, | |
| "loss": 0.2145, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.5883233532934131, | |
| "grad_norm": 0.20823459327220917, | |
| "learning_rate": 8.078929663213432e-06, | |
| "loss": 0.2993, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.5898203592814372, | |
| "grad_norm": 0.15346956253051758, | |
| "learning_rate": 8.069587307857377e-06, | |
| "loss": 0.2628, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.5913173652694611, | |
| "grad_norm": 0.24852439761161804, | |
| "learning_rate": 8.060227722215385e-06, | |
| "loss": 0.3749, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.592814371257485, | |
| "grad_norm": 0.17771868407726288, | |
| "learning_rate": 8.050850958824926e-06, | |
| "loss": 0.4067, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.594311377245509, | |
| "grad_norm": 0.2503078281879425, | |
| "learning_rate": 8.041457070319884e-06, | |
| "loss": 0.4665, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.5958083832335329, | |
| "grad_norm": 0.26984649896621704, | |
| "learning_rate": 8.032046109430276e-06, | |
| "loss": 0.3278, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.5973053892215568, | |
| "grad_norm": 0.25289425253868103, | |
| "learning_rate": 8.02261812898195e-06, | |
| "loss": 0.2592, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.5988023952095808, | |
| "grad_norm": 0.1754600703716278, | |
| "learning_rate": 8.013173181896283e-06, | |
| "loss": 0.2665, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6002994011976048, | |
| "grad_norm": 0.22235795855522156, | |
| "learning_rate": 8.003711321189895e-06, | |
| "loss": 0.2513, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.6017964071856288, | |
| "grad_norm": 0.36419808864593506, | |
| "learning_rate": 7.994232599974346e-06, | |
| "loss": 0.4715, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.6032934131736527, | |
| "grad_norm": 0.166230246424675, | |
| "learning_rate": 7.984737071455834e-06, | |
| "loss": 0.2649, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.6047904191616766, | |
| "grad_norm": 0.27781379222869873, | |
| "learning_rate": 7.975224788934903e-06, | |
| "loss": 0.3881, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.6062874251497006, | |
| "grad_norm": 0.18180640041828156, | |
| "learning_rate": 7.965695805806141e-06, | |
| "loss": 0.1962, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6077844311377245, | |
| "grad_norm": 0.1591605544090271, | |
| "learning_rate": 7.95615017555788e-06, | |
| "loss": 0.2133, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.6092814371257484, | |
| "grad_norm": 0.16064637899398804, | |
| "learning_rate": 7.946587951771894e-06, | |
| "loss": 0.3003, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.6107784431137725, | |
| "grad_norm": 0.22653892636299133, | |
| "learning_rate": 7.937009188123102e-06, | |
| "loss": 0.3814, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.6122754491017964, | |
| "grad_norm": 0.19928650557994843, | |
| "learning_rate": 7.927413938379268e-06, | |
| "loss": 0.3206, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.6137724550898204, | |
| "grad_norm": 0.2117091864347458, | |
| "learning_rate": 7.917802256400688e-06, | |
| "loss": 0.4773, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6152694610778443, | |
| "grad_norm": 0.1339869201183319, | |
| "learning_rate": 7.908174196139907e-06, | |
| "loss": 0.1727, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.6167664670658682, | |
| "grad_norm": 0.1618313491344452, | |
| "learning_rate": 7.898529811641393e-06, | |
| "loss": 0.3444, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.6182634730538922, | |
| "grad_norm": 0.1486457735300064, | |
| "learning_rate": 7.888869157041257e-06, | |
| "loss": 0.2223, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.6197604790419161, | |
| "grad_norm": 0.19519484043121338, | |
| "learning_rate": 7.879192286566929e-06, | |
| "loss": 0.426, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.6212574850299402, | |
| "grad_norm": 0.20848150551319122, | |
| "learning_rate": 7.869499254536865e-06, | |
| "loss": 0.3986, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6227544910179641, | |
| "grad_norm": 0.21209968626499176, | |
| "learning_rate": 7.859790115360243e-06, | |
| "loss": 0.3407, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.624251497005988, | |
| "grad_norm": 0.23103849589824677, | |
| "learning_rate": 7.850064923536649e-06, | |
| "loss": 0.4584, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.625748502994012, | |
| "grad_norm": 0.16797013580799103, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.3119, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.6272455089820359, | |
| "grad_norm": 0.16341273486614227, | |
| "learning_rate": 7.83056660039713e-06, | |
| "loss": 0.2981, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.6287425149700598, | |
| "grad_norm": 0.19721773266792297, | |
| "learning_rate": 7.82079357852969e-06, | |
| "loss": 0.4564, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6302395209580839, | |
| "grad_norm": 0.18293291330337524, | |
| "learning_rate": 7.811004722911637e-06, | |
| "loss": 0.437, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.6317365269461078, | |
| "grad_norm": 0.18640290200710297, | |
| "learning_rate": 7.801200088490026e-06, | |
| "loss": 0.3639, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.6332335329341318, | |
| "grad_norm": 0.22067849338054657, | |
| "learning_rate": 7.791379730300476e-06, | |
| "loss": 0.4027, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.6347305389221557, | |
| "grad_norm": 0.13944509625434875, | |
| "learning_rate": 7.781543703466881e-06, | |
| "loss": 0.2165, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.6362275449101796, | |
| "grad_norm": 0.187015101313591, | |
| "learning_rate": 7.771692063201072e-06, | |
| "loss": 0.2628, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6377245508982036, | |
| "grad_norm": 0.23688088357448578, | |
| "learning_rate": 7.76182486480253e-06, | |
| "loss": 0.3946, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.6392215568862275, | |
| "grad_norm": 0.22390404343605042, | |
| "learning_rate": 7.751942163658066e-06, | |
| "loss": 0.3308, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.6407185628742516, | |
| "grad_norm": 0.26804405450820923, | |
| "learning_rate": 7.742044015241508e-06, | |
| "loss": 0.3903, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.6422155688622755, | |
| "grad_norm": 0.14200535416603088, | |
| "learning_rate": 7.7321304751134e-06, | |
| "loss": 0.3099, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.6437125748502994, | |
| "grad_norm": 0.17333939671516418, | |
| "learning_rate": 7.722201598920673e-06, | |
| "loss": 0.3332, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6452095808383234, | |
| "grad_norm": 0.1965278834104538, | |
| "learning_rate": 7.712257442396355e-06, | |
| "loss": 0.344, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.6467065868263473, | |
| "grad_norm": 0.18956130743026733, | |
| "learning_rate": 7.702298061359236e-06, | |
| "loss": 0.3551, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.6482035928143712, | |
| "grad_norm": 0.2348889261484146, | |
| "learning_rate": 7.692323511713568e-06, | |
| "loss": 0.2989, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.6497005988023952, | |
| "grad_norm": 0.23371171951293945, | |
| "learning_rate": 7.682333849448749e-06, | |
| "loss": 0.4422, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.6511976047904192, | |
| "grad_norm": 0.18746024370193481, | |
| "learning_rate": 7.672329130639007e-06, | |
| "loss": 0.2378, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6526946107784432, | |
| "grad_norm": 0.21659070253372192, | |
| "learning_rate": 7.662309411443084e-06, | |
| "loss": 0.3655, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.6541916167664671, | |
| "grad_norm": 0.24405577778816223, | |
| "learning_rate": 7.652274748103924e-06, | |
| "loss": 0.4455, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.655688622754491, | |
| "grad_norm": 0.2173536866903305, | |
| "learning_rate": 7.642225196948357e-06, | |
| "loss": 0.3585, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.657185628742515, | |
| "grad_norm": 0.21628788113594055, | |
| "learning_rate": 7.63216081438678e-06, | |
| "loss": 0.2962, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.6586826347305389, | |
| "grad_norm": 0.1556631177663803, | |
| "learning_rate": 7.622081656912842e-06, | |
| "loss": 0.1789, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6601796407185628, | |
| "grad_norm": 0.21247422695159912, | |
| "learning_rate": 7.611987781103128e-06, | |
| "loss": 0.243, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.6616766467065869, | |
| "grad_norm": 0.16574670374393463, | |
| "learning_rate": 7.601879243616838e-06, | |
| "loss": 0.2957, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.6631736526946108, | |
| "grad_norm": 0.19471777975559235, | |
| "learning_rate": 7.5917561011954755e-06, | |
| "loss": 0.2709, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.6646706586826348, | |
| "grad_norm": 0.2093854397535324, | |
| "learning_rate": 7.581618410662519e-06, | |
| "loss": 0.3786, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.6661676646706587, | |
| "grad_norm": 0.1995527446269989, | |
| "learning_rate": 7.571466228923115e-06, | |
| "loss": 0.282, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6676646706586826, | |
| "grad_norm": 0.20918631553649902, | |
| "learning_rate": 7.56129961296375e-06, | |
| "loss": 0.3328, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.6691616766467066, | |
| "grad_norm": 0.3352525532245636, | |
| "learning_rate": 7.551118619851929e-06, | |
| "loss": 0.5474, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.6706586826347305, | |
| "grad_norm": 0.16128239035606384, | |
| "learning_rate": 7.540923306735868e-06, | |
| "loss": 0.2257, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.6721556886227545, | |
| "grad_norm": 0.19775696098804474, | |
| "learning_rate": 7.530713730844153e-06, | |
| "loss": 0.399, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.6736526946107785, | |
| "grad_norm": 0.16437461972236633, | |
| "learning_rate": 7.5204899494854415e-06, | |
| "loss": 0.3229, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6751497005988024, | |
| "grad_norm": 0.1884375959634781, | |
| "learning_rate": 7.510252020048121e-06, | |
| "loss": 0.2652, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.6766467065868264, | |
| "grad_norm": 0.22506847977638245, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.293, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.6781437125748503, | |
| "grad_norm": 0.24144884943962097, | |
| "learning_rate": 7.489733946887982e-06, | |
| "loss": 0.2653, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.6796407185628742, | |
| "grad_norm": 0.2532469928264618, | |
| "learning_rate": 7.479453918337733e-06, | |
| "loss": 0.3346, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.6811377245508982, | |
| "grad_norm": 0.20879314839839935, | |
| "learning_rate": 7.469159972053377e-06, | |
| "loss": 0.4094, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6826347305389222, | |
| "grad_norm": 0.3141922354698181, | |
| "learning_rate": 7.458852165817153e-06, | |
| "loss": 0.5201, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.6841317365269461, | |
| "grad_norm": 0.22677458822727203, | |
| "learning_rate": 7.448530557489105e-06, | |
| "loss": 0.2836, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.6856287425149701, | |
| "grad_norm": 0.20220258831977844, | |
| "learning_rate": 7.438195205006749e-06, | |
| "loss": 0.331, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.687125748502994, | |
| "grad_norm": 0.22227731347084045, | |
| "learning_rate": 7.427846166384747e-06, | |
| "loss": 0.3344, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.688622754491018, | |
| "grad_norm": 0.2184399664402008, | |
| "learning_rate": 7.417483499714589e-06, | |
| "loss": 0.2932, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6901197604790419, | |
| "grad_norm": 0.16358090937137604, | |
| "learning_rate": 7.40710726316426e-06, | |
| "loss": 0.2363, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.6916167664670658, | |
| "grad_norm": 0.25432059168815613, | |
| "learning_rate": 7.396717514977916e-06, | |
| "loss": 0.3201, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.6931137724550899, | |
| "grad_norm": 0.30511903762817383, | |
| "learning_rate": 7.386314313475557e-06, | |
| "loss": 0.4459, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.6946107784431138, | |
| "grad_norm": 0.26864519715309143, | |
| "learning_rate": 7.3758977170527e-06, | |
| "loss": 0.2934, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.6961077844311377, | |
| "grad_norm": 0.21390244364738464, | |
| "learning_rate": 7.365467784180051e-06, | |
| "loss": 0.1999, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.6976047904191617, | |
| "grad_norm": 0.17398962378501892, | |
| "learning_rate": 7.355024573403174e-06, | |
| "loss": 0.3066, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.6991017964071856, | |
| "grad_norm": 0.3039548695087433, | |
| "learning_rate": 7.3445681433421675e-06, | |
| "loss": 0.2559, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.7005988023952096, | |
| "grad_norm": 0.24252189695835114, | |
| "learning_rate": 7.3340985526913335e-06, | |
| "loss": 0.2586, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.7020958083832335, | |
| "grad_norm": 0.25120341777801514, | |
| "learning_rate": 7.323615860218844e-06, | |
| "loss": 0.3042, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.7035928143712575, | |
| "grad_norm": 0.38001149892807007, | |
| "learning_rate": 7.313120124766417e-06, | |
| "loss": 0.3586, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7050898203592815, | |
| "grad_norm": 0.20646889507770538, | |
| "learning_rate": 7.30261140524898e-06, | |
| "loss": 0.2903, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.7065868263473054, | |
| "grad_norm": 0.26575809717178345, | |
| "learning_rate": 7.292089760654352e-06, | |
| "loss": 0.2822, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.7080838323353293, | |
| "grad_norm": 0.18457849323749542, | |
| "learning_rate": 7.281555250042893e-06, | |
| "loss": 0.3372, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.7095808383233533, | |
| "grad_norm": 0.16875137388706207, | |
| "learning_rate": 7.271007932547188e-06, | |
| "loss": 0.2172, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.7110778443113772, | |
| "grad_norm": 0.22017961740493774, | |
| "learning_rate": 7.2604478673717095e-06, | |
| "loss": 0.3113, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7125748502994012, | |
| "grad_norm": 0.31057754158973694, | |
| "learning_rate": 7.249875113792485e-06, | |
| "loss": 0.3059, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.7140718562874252, | |
| "grad_norm": 0.22742144763469696, | |
| "learning_rate": 7.239289731156767e-06, | |
| "loss": 0.4326, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.7155688622754491, | |
| "grad_norm": 0.19704562425613403, | |
| "learning_rate": 7.2286917788826926e-06, | |
| "loss": 0.3139, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.7170658682634731, | |
| "grad_norm": 0.18587222695350647, | |
| "learning_rate": 7.218081316458959e-06, | |
| "loss": 0.3197, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.718562874251497, | |
| "grad_norm": 0.2095022052526474, | |
| "learning_rate": 7.207458403444488e-06, | |
| "loss": 0.3275, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7200598802395209, | |
| "grad_norm": 0.2593580186367035, | |
| "learning_rate": 7.196823099468084e-06, | |
| "loss": 0.2981, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.7215568862275449, | |
| "grad_norm": 0.2533735930919647, | |
| "learning_rate": 7.186175464228109e-06, | |
| "loss": 0.3067, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.7230538922155688, | |
| "grad_norm": 0.16174565255641937, | |
| "learning_rate": 7.175515557492139e-06, | |
| "loss": 0.2737, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.7245508982035929, | |
| "grad_norm": 0.2620272934436798, | |
| "learning_rate": 7.1648434390966356e-06, | |
| "loss": 0.2914, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.7260479041916168, | |
| "grad_norm": 0.27633848786354065, | |
| "learning_rate": 7.154159168946607e-06, | |
| "loss": 0.4116, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7275449101796407, | |
| "grad_norm": 0.242568239569664, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.357, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.7290419161676647, | |
| "grad_norm": 0.18494291603565216, | |
| "learning_rate": 7.132754413343721e-06, | |
| "loss": 0.4115, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.7305389221556886, | |
| "grad_norm": 0.3592863976955414, | |
| "learning_rate": 7.122034048040586e-06, | |
| "loss": 0.3621, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.7320359281437125, | |
| "grad_norm": 0.19837193191051483, | |
| "learning_rate": 7.111301771281692e-06, | |
| "loss": 0.2239, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.7335329341317365, | |
| "grad_norm": 0.18162252008914948, | |
| "learning_rate": 7.100557643309732e-06, | |
| "loss": 0.2108, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7350299401197605, | |
| "grad_norm": 0.17453975975513458, | |
| "learning_rate": 7.089801724433918e-06, | |
| "loss": 0.2776, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.7365269461077845, | |
| "grad_norm": 0.21002128720283508, | |
| "learning_rate": 7.079034075029651e-06, | |
| "loss": 0.2821, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.7380239520958084, | |
| "grad_norm": 0.2540780007839203, | |
| "learning_rate": 7.0682547555381734e-06, | |
| "loss": 0.368, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.7395209580838323, | |
| "grad_norm": 0.28589752316474915, | |
| "learning_rate": 7.057463826466235e-06, | |
| "loss": 0.3174, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.7410179640718563, | |
| "grad_norm": 0.2610958516597748, | |
| "learning_rate": 7.0466613483857615e-06, | |
| "loss": 0.2885, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.7425149700598802, | |
| "grad_norm": 0.13012602925300598, | |
| "learning_rate": 7.035847381933494e-06, | |
| "loss": 0.1997, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.7440119760479041, | |
| "grad_norm": 0.18959909677505493, | |
| "learning_rate": 7.025021987810664e-06, | |
| "loss": 0.273, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.7455089820359282, | |
| "grad_norm": 0.20305494964122772, | |
| "learning_rate": 7.014185226782655e-06, | |
| "loss": 0.3904, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.7470059880239521, | |
| "grad_norm": 0.2603055536746979, | |
| "learning_rate": 7.003337159678649e-06, | |
| "loss": 0.3786, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.7485029940119761, | |
| "grad_norm": 0.23833239078521729, | |
| "learning_rate": 6.992477847391292e-06, | |
| "loss": 0.2107, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.20404885709285736, | |
| "learning_rate": 6.981607350876357e-06, | |
| "loss": 0.2572, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 0.3029465675354004, | |
| "eval_runtime": 95.9587, | |
| "eval_samples_per_second": 7.67, | |
| "eval_steps_per_second": 0.959, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.7514970059880239, | |
| "grad_norm": 0.23784396052360535, | |
| "learning_rate": 6.970725731152389e-06, | |
| "loss": 0.2849, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.7529940119760479, | |
| "grad_norm": 0.23206007480621338, | |
| "learning_rate": 6.959833049300376e-06, | |
| "loss": 0.3136, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.7544910179640718, | |
| "grad_norm": 0.19017294049263, | |
| "learning_rate": 6.948929366463397e-06, | |
| "loss": 0.2339, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.7559880239520959, | |
| "grad_norm": 0.21896733343601227, | |
| "learning_rate": 6.938014743846285e-06, | |
| "loss": 0.3311, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7574850299401198, | |
| "grad_norm": 0.20357593894004822, | |
| "learning_rate": 6.927089242715277e-06, | |
| "loss": 0.2871, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.7589820359281437, | |
| "grad_norm": 0.2007334679365158, | |
| "learning_rate": 6.916152924397676e-06, | |
| "loss": 0.2615, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.7604790419161677, | |
| "grad_norm": 0.284795880317688, | |
| "learning_rate": 6.905205850281502e-06, | |
| "loss": 0.4047, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.7619760479041916, | |
| "grad_norm": 0.25428882241249084, | |
| "learning_rate": 6.894248081815155e-06, | |
| "loss": 0.3606, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.7634730538922155, | |
| "grad_norm": 0.2195863127708435, | |
| "learning_rate": 6.883279680507057e-06, | |
| "loss": 0.4163, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7649700598802395, | |
| "grad_norm": 0.19675293564796448, | |
| "learning_rate": 6.872300707925319e-06, | |
| "loss": 0.3199, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.7664670658682635, | |
| "grad_norm": 0.16144664585590363, | |
| "learning_rate": 6.861311225697392e-06, | |
| "loss": 0.2514, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.7679640718562875, | |
| "grad_norm": 0.2618183493614197, | |
| "learning_rate": 6.850311295509719e-06, | |
| "loss": 0.298, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.7694610778443114, | |
| "grad_norm": 0.2717672288417816, | |
| "learning_rate": 6.8393009791073895e-06, | |
| "loss": 0.4989, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.7709580838323353, | |
| "grad_norm": 0.24221351742744446, | |
| "learning_rate": 6.828280338293792e-06, | |
| "loss": 0.3629, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7724550898203593, | |
| "grad_norm": 0.18478509783744812, | |
| "learning_rate": 6.817249434930267e-06, | |
| "loss": 0.1947, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.7739520958083832, | |
| "grad_norm": 0.2637442648410797, | |
| "learning_rate": 6.806208330935766e-06, | |
| "loss": 0.4101, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.7754491017964071, | |
| "grad_norm": 0.2035938799381256, | |
| "learning_rate": 6.7951570882864944e-06, | |
| "loss": 0.2898, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.7769461077844312, | |
| "grad_norm": 0.15909993648529053, | |
| "learning_rate": 6.784095769015573e-06, | |
| "loss": 0.2155, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.7784431137724551, | |
| "grad_norm": 0.27301713824272156, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.5023, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7799401197604791, | |
| "grad_norm": 0.2372361719608307, | |
| "learning_rate": 6.761943149023706e-06, | |
| "loss": 0.2363, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.781437125748503, | |
| "grad_norm": 0.2195385992527008, | |
| "learning_rate": 6.750851972650416e-06, | |
| "loss": 0.261, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.7829341317365269, | |
| "grad_norm": 0.22001579403877258, | |
| "learning_rate": 6.739750968350081e-06, | |
| "loss": 0.2957, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.7844311377245509, | |
| "grad_norm": 0.17113597691059113, | |
| "learning_rate": 6.728640198435143e-06, | |
| "loss": 0.2352, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.7859281437125748, | |
| "grad_norm": 0.21069037914276123, | |
| "learning_rate": 6.717519725272859e-06, | |
| "loss": 0.2657, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.7874251497005988, | |
| "grad_norm": 0.22734540700912476, | |
| "learning_rate": 6.706389611284953e-06, | |
| "loss": 0.2924, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.7889221556886228, | |
| "grad_norm": 0.30117544531822205, | |
| "learning_rate": 6.6952499189472665e-06, | |
| "loss": 0.2928, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.7904191616766467, | |
| "grad_norm": 0.1863144040107727, | |
| "learning_rate": 6.684100710789405e-06, | |
| "loss": 0.2685, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.7919161676646707, | |
| "grad_norm": 0.28917059302330017, | |
| "learning_rate": 6.6729420493943875e-06, | |
| "loss": 0.3043, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.7934131736526946, | |
| "grad_norm": 0.24295863509178162, | |
| "learning_rate": 6.6617739973982985e-06, | |
| "loss": 0.3415, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7949101796407185, | |
| "grad_norm": 0.16020585596561432, | |
| "learning_rate": 6.6505966174899326e-06, | |
| "loss": 0.1827, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.7964071856287425, | |
| "grad_norm": 0.23983198404312134, | |
| "learning_rate": 6.639409972410446e-06, | |
| "loss": 0.341, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.7979041916167665, | |
| "grad_norm": 0.1791592538356781, | |
| "learning_rate": 6.628214124952999e-06, | |
| "loss": 0.232, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.7994011976047904, | |
| "grad_norm": 0.22436876595020294, | |
| "learning_rate": 6.617009137962407e-06, | |
| "loss": 0.3309, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.8008982035928144, | |
| "grad_norm": 0.20181064307689667, | |
| "learning_rate": 6.605795074334793e-06, | |
| "loss": 0.4341, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8023952095808383, | |
| "grad_norm": 0.26198679208755493, | |
| "learning_rate": 6.594571997017224e-06, | |
| "loss": 0.3157, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.8038922155688623, | |
| "grad_norm": 0.182569220662117, | |
| "learning_rate": 6.583339969007364e-06, | |
| "loss": 0.1959, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.8053892215568862, | |
| "grad_norm": 0.23903851211071014, | |
| "learning_rate": 6.57209905335312e-06, | |
| "loss": 0.3192, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.8068862275449101, | |
| "grad_norm": 0.26894667744636536, | |
| "learning_rate": 6.560849313152287e-06, | |
| "loss": 0.2271, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.8083832335329342, | |
| "grad_norm": 0.24519504606723785, | |
| "learning_rate": 6.549590811552193e-06, | |
| "loss": 0.2898, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8098802395209581, | |
| "grad_norm": 0.25122636556625366, | |
| "learning_rate": 6.538323611749351e-06, | |
| "loss": 0.4239, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.811377245508982, | |
| "grad_norm": 0.21076048910617828, | |
| "learning_rate": 6.5270477769890906e-06, | |
| "loss": 0.3781, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.812874251497006, | |
| "grad_norm": 0.17968085408210754, | |
| "learning_rate": 6.515763370565218e-06, | |
| "loss": 0.2529, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.8143712574850299, | |
| "grad_norm": 0.37112146615982056, | |
| "learning_rate": 6.504470455819651e-06, | |
| "loss": 0.4063, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.8158682634730539, | |
| "grad_norm": 0.25475284457206726, | |
| "learning_rate": 6.493169096142068e-06, | |
| "loss": 0.2871, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.8173652694610778, | |
| "grad_norm": 0.17944371700286865, | |
| "learning_rate": 6.481859354969549e-06, | |
| "loss": 0.2466, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.8188622754491018, | |
| "grad_norm": 0.22027306258678436, | |
| "learning_rate": 6.470541295786222e-06, | |
| "loss": 0.31, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.8203592814371258, | |
| "grad_norm": 0.21977883577346802, | |
| "learning_rate": 6.4592149821229064e-06, | |
| "loss": 0.2706, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.8218562874251497, | |
| "grad_norm": 0.2781427502632141, | |
| "learning_rate": 6.447880477556757e-06, | |
| "loss": 0.3271, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.8233532934131736, | |
| "grad_norm": 0.1752307116985321, | |
| "learning_rate": 6.436537845710904e-06, | |
| "loss": 0.2063, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8248502994011976, | |
| "grad_norm": 0.18458420038223267, | |
| "learning_rate": 6.425187150254097e-06, | |
| "loss": 0.2665, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.8263473053892215, | |
| "grad_norm": 0.25183022022247314, | |
| "learning_rate": 6.413828454900351e-06, | |
| "loss": 0.3594, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.8278443113772455, | |
| "grad_norm": 0.21946825087070465, | |
| "learning_rate": 6.402461823408584e-06, | |
| "loss": 0.2277, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.8293413173652695, | |
| "grad_norm": 0.2452111393213272, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.3541, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.8308383233532934, | |
| "grad_norm": 0.20177938044071198, | |
| "learning_rate": 6.379705007269046e-06, | |
| "loss": 0.2672, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.8323353293413174, | |
| "grad_norm": 0.23342332243919373, | |
| "learning_rate": 6.368314950360416e-06, | |
| "loss": 0.3052, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.8338323353293413, | |
| "grad_norm": 0.1889418214559555, | |
| "learning_rate": 6.356917212791332e-06, | |
| "loss": 0.1696, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.8353293413173652, | |
| "grad_norm": 0.21418286859989166, | |
| "learning_rate": 6.3455118585398676e-06, | |
| "loss": 0.298, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.8368263473053892, | |
| "grad_norm": 0.28259849548339844, | |
| "learning_rate": 6.334098951626847e-06, | |
| "loss": 0.3329, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.8383233532934131, | |
| "grad_norm": 0.20598739385604858, | |
| "learning_rate": 6.3226785561154914e-06, | |
| "loss": 0.3018, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8398203592814372, | |
| "grad_norm": 0.22180384397506714, | |
| "learning_rate": 6.311250736111058e-06, | |
| "loss": 0.3665, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.8413173652694611, | |
| "grad_norm": 0.25343960523605347, | |
| "learning_rate": 6.299815555760478e-06, | |
| "loss": 0.4075, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.842814371257485, | |
| "grad_norm": 0.20312725007534027, | |
| "learning_rate": 6.288373079251996e-06, | |
| "loss": 0.2432, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.844311377245509, | |
| "grad_norm": 0.19411613047122955, | |
| "learning_rate": 6.276923370814815e-06, | |
| "loss": 0.268, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.8458083832335329, | |
| "grad_norm": 0.2687308192253113, | |
| "learning_rate": 6.265466494718731e-06, | |
| "loss": 0.224, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.8473053892215568, | |
| "grad_norm": 0.25632116198539734, | |
| "learning_rate": 6.254002515273775e-06, | |
| "loss": 0.2666, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.8488023952095808, | |
| "grad_norm": 0.18523173034191132, | |
| "learning_rate": 6.242531496829848e-06, | |
| "loss": 0.2491, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.8502994011976048, | |
| "grad_norm": 0.5198162198066711, | |
| "learning_rate": 6.231053503776363e-06, | |
| "loss": 0.3999, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.8517964071856288, | |
| "grad_norm": 0.302183598279953, | |
| "learning_rate": 6.219568600541886e-06, | |
| "loss": 0.2657, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.8532934131736527, | |
| "grad_norm": 0.25520554184913635, | |
| "learning_rate": 6.208076851593768e-06, | |
| "loss": 0.3349, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8547904191616766, | |
| "grad_norm": 0.2884499132633209, | |
| "learning_rate": 6.1965783214377895e-06, | |
| "loss": 0.2423, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.8562874251497006, | |
| "grad_norm": 0.20894049108028412, | |
| "learning_rate": 6.185073074617793e-06, | |
| "loss": 0.287, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.8577844311377245, | |
| "grad_norm": 0.2774272859096527, | |
| "learning_rate": 6.173561175715323e-06, | |
| "loss": 0.3144, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.8592814371257484, | |
| "grad_norm": 0.20034471154212952, | |
| "learning_rate": 6.1620426893492645e-06, | |
| "loss": 0.2634, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.8607784431137725, | |
| "grad_norm": 0.28622931241989136, | |
| "learning_rate": 6.150517680175482e-06, | |
| "loss": 0.3312, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8622754491017964, | |
| "grad_norm": 0.15074674785137177, | |
| "learning_rate": 6.13898621288645e-06, | |
| "loss": 0.1718, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.8637724550898204, | |
| "grad_norm": 0.24118809401988983, | |
| "learning_rate": 6.127448352210894e-06, | |
| "loss": 0.4585, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.8652694610778443, | |
| "grad_norm": 0.24849183857440948, | |
| "learning_rate": 6.115904162913431e-06, | |
| "loss": 0.2945, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.8667664670658682, | |
| "grad_norm": 0.1957549899816513, | |
| "learning_rate": 6.1043537097941985e-06, | |
| "loss": 0.2029, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.8682634730538922, | |
| "grad_norm": 0.22294564545154572, | |
| "learning_rate": 6.092797057688496e-06, | |
| "loss": 0.2482, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8697604790419161, | |
| "grad_norm": 0.22241006791591644, | |
| "learning_rate": 6.081234271466416e-06, | |
| "loss": 0.2561, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.8712574850299402, | |
| "grad_norm": 0.19968393445014954, | |
| "learning_rate": 6.0696654160324875e-06, | |
| "loss": 0.232, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.8727544910179641, | |
| "grad_norm": 0.23499524593353271, | |
| "learning_rate": 6.058090556325305e-06, | |
| "loss": 0.2794, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.874251497005988, | |
| "grad_norm": 0.23651576042175293, | |
| "learning_rate": 6.046509757317168e-06, | |
| "loss": 0.4183, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.875748502994012, | |
| "grad_norm": 0.1924506276845932, | |
| "learning_rate": 6.034923084013713e-06, | |
| "loss": 0.2618, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.8772455089820359, | |
| "grad_norm": 0.23315298557281494, | |
| "learning_rate": 6.0233306014535505e-06, | |
| "loss": 0.2963, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.8787425149700598, | |
| "grad_norm": 0.19225433468818665, | |
| "learning_rate": 6.0117323747079e-06, | |
| "loss": 0.229, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.8802395209580839, | |
| "grad_norm": 0.1963253617286682, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.2386, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.8817365269461078, | |
| "grad_norm": 0.1923033744096756, | |
| "learning_rate": 5.988518949105862e-06, | |
| "loss": 0.2318, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.8832335329341318, | |
| "grad_norm": 0.20783311128616333, | |
| "learning_rate": 5.976903880551669e-06, | |
| "loss": 0.2633, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8847305389221557, | |
| "grad_norm": 0.23813506960868835, | |
| "learning_rate": 5.965283328415644e-06, | |
| "loss": 0.2994, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.8862275449101796, | |
| "grad_norm": 0.31241193413734436, | |
| "learning_rate": 5.953657357926569e-06, | |
| "loss": 0.3336, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.8877245508982036, | |
| "grad_norm": 0.20485684275627136, | |
| "learning_rate": 5.942026034343636e-06, | |
| "loss": 0.2006, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.8892215568862275, | |
| "grad_norm": 0.23816922307014465, | |
| "learning_rate": 5.930389422956088e-06, | |
| "loss": 0.3098, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.8907185628742516, | |
| "grad_norm": 0.23085439205169678, | |
| "learning_rate": 5.918747589082853e-06, | |
| "loss": 0.2973, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.8922155688622755, | |
| "grad_norm": 0.22914822399616241, | |
| "learning_rate": 5.907100598072166e-06, | |
| "loss": 0.2719, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.8937125748502994, | |
| "grad_norm": 0.2522124946117401, | |
| "learning_rate": 5.895448515301218e-06, | |
| "loss": 0.2547, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.8952095808383234, | |
| "grad_norm": 0.20887671411037445, | |
| "learning_rate": 5.883791406175775e-06, | |
| "loss": 0.2112, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.8967065868263473, | |
| "grad_norm": 0.35924533009529114, | |
| "learning_rate": 5.872129336129821e-06, | |
| "loss": 0.2568, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.8982035928143712, | |
| "grad_norm": 0.18609894812107086, | |
| "learning_rate": 5.860462370625189e-06, | |
| "loss": 0.2106, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8997005988023952, | |
| "grad_norm": 0.19369381666183472, | |
| "learning_rate": 5.848790575151181e-06, | |
| "loss": 0.2701, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.9011976047904192, | |
| "grad_norm": 0.25602883100509644, | |
| "learning_rate": 5.837114015224223e-06, | |
| "loss": 0.4195, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.9026946107784432, | |
| "grad_norm": 0.2999608516693115, | |
| "learning_rate": 5.8254327563874794e-06, | |
| "loss": 0.4362, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.9041916167664671, | |
| "grad_norm": 0.29497092962265015, | |
| "learning_rate": 5.813746864210489e-06, | |
| "loss": 0.3599, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.905688622754491, | |
| "grad_norm": 0.27466148138046265, | |
| "learning_rate": 5.8020564042888015e-06, | |
| "loss": 0.2137, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.907185628742515, | |
| "grad_norm": 0.1898701786994934, | |
| "learning_rate": 5.790361442243605e-06, | |
| "loss": 0.2576, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.9086826347305389, | |
| "grad_norm": 0.2565309405326843, | |
| "learning_rate": 5.778662043721359e-06, | |
| "loss": 0.2865, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.9101796407185628, | |
| "grad_norm": 0.19474193453788757, | |
| "learning_rate": 5.766958274393428e-06, | |
| "loss": 0.2358, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.9116766467065869, | |
| "grad_norm": 0.2232877016067505, | |
| "learning_rate": 5.7552501999557065e-06, | |
| "loss": 0.2392, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.9131736526946108, | |
| "grad_norm": 0.19385279715061188, | |
| "learning_rate": 5.743537886128258e-06, | |
| "loss": 0.196, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9146706586826348, | |
| "grad_norm": 0.15841762721538544, | |
| "learning_rate": 5.731821398654944e-06, | |
| "loss": 0.179, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.9161676646706587, | |
| "grad_norm": 0.17763687670230865, | |
| "learning_rate": 5.72010080330305e-06, | |
| "loss": 0.1782, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.9176646706586826, | |
| "grad_norm": 0.21254192292690277, | |
| "learning_rate": 5.708376165862921e-06, | |
| "loss": 0.219, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.9191616766467066, | |
| "grad_norm": 0.1851448267698288, | |
| "learning_rate": 5.696647552147589e-06, | |
| "loss": 0.1812, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.9206586826347305, | |
| "grad_norm": 0.23409013450145721, | |
| "learning_rate": 5.684915027992415e-06, | |
| "loss": 0.2064, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9221556886227545, | |
| "grad_norm": 0.249491885304451, | |
| "learning_rate": 5.673178659254698e-06, | |
| "loss": 0.2612, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.9236526946107785, | |
| "grad_norm": 0.25312793254852295, | |
| "learning_rate": 5.661438511813324e-06, | |
| "loss": 0.2762, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.9251497005988024, | |
| "grad_norm": 0.23058243095874786, | |
| "learning_rate": 5.64969465156839e-06, | |
| "loss": 0.2876, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.9266467065868264, | |
| "grad_norm": 0.19073981046676636, | |
| "learning_rate": 5.637947144440832e-06, | |
| "loss": 0.2177, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.9281437125748503, | |
| "grad_norm": 0.23557421565055847, | |
| "learning_rate": 5.626196056372056e-06, | |
| "loss": 0.3985, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9296407185628742, | |
| "grad_norm": 0.18695135414600372, | |
| "learning_rate": 5.614441453323571e-06, | |
| "loss": 0.1995, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.9311377245508982, | |
| "grad_norm": 0.18378764390945435, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.2558, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.9326347305389222, | |
| "grad_norm": 0.25633805990219116, | |
| "learning_rate": 5.590921966231788e-06, | |
| "loss": 0.3359, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.9341317365269461, | |
| "grad_norm": 0.27903831005096436, | |
| "learning_rate": 5.579157214208675e-06, | |
| "loss": 0.3387, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.9356287425149701, | |
| "grad_norm": 0.28710755705833435, | |
| "learning_rate": 5.567389211245486e-06, | |
| "loss": 0.2983, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.937125748502994, | |
| "grad_norm": 0.26144298911094666, | |
| "learning_rate": 5.555618023398671e-06, | |
| "loss": 0.2697, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.938622754491018, | |
| "grad_norm": 0.20057110488414764, | |
| "learning_rate": 5.5438437167425675e-06, | |
| "loss": 0.2945, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.9401197604790419, | |
| "grad_norm": 0.25036031007766724, | |
| "learning_rate": 5.532066357369012e-06, | |
| "loss": 0.3688, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.9416167664670658, | |
| "grad_norm": 0.23812642693519592, | |
| "learning_rate": 5.52028601138698e-06, | |
| "loss": 0.2967, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.9431137724550899, | |
| "grad_norm": 0.30330294370651245, | |
| "learning_rate": 5.508502744922212e-06, | |
| "loss": 0.3328, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9446107784431138, | |
| "grad_norm": 0.2676711678504944, | |
| "learning_rate": 5.496716624116836e-06, | |
| "loss": 0.2188, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.9461077844311377, | |
| "grad_norm": 0.37458300590515137, | |
| "learning_rate": 5.484927715129011e-06, | |
| "loss": 0.265, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.9476047904191617, | |
| "grad_norm": 0.2473772168159485, | |
| "learning_rate": 5.4731360841325405e-06, | |
| "loss": 0.2214, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.9491017964071856, | |
| "grad_norm": 0.25858768820762634, | |
| "learning_rate": 5.46134179731651e-06, | |
| "loss": 0.2649, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.9505988023952096, | |
| "grad_norm": 0.25391316413879395, | |
| "learning_rate": 5.449544920884912e-06, | |
| "loss": 0.3284, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9520958083832335, | |
| "grad_norm": 0.1776520311832428, | |
| "learning_rate": 5.437745521056272e-06, | |
| "loss": 0.1544, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.9535928143712575, | |
| "grad_norm": 0.2536100447177887, | |
| "learning_rate": 5.425943664063284e-06, | |
| "loss": 0.2558, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.9550898203592815, | |
| "grad_norm": 0.27622076869010925, | |
| "learning_rate": 5.414139416152435e-06, | |
| "loss": 0.2889, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.9565868263473054, | |
| "grad_norm": 0.19522008299827576, | |
| "learning_rate": 5.402332843583631e-06, | |
| "loss": 0.2173, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.9580838323353293, | |
| "grad_norm": 0.19289469718933105, | |
| "learning_rate": 5.390524012629824e-06, | |
| "loss": 0.3045, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.9595808383233533, | |
| "grad_norm": 0.21840733289718628, | |
| "learning_rate": 5.3787129895766484e-06, | |
| "loss": 0.3464, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.9610778443113772, | |
| "grad_norm": 0.3051481246948242, | |
| "learning_rate": 5.3668998407220385e-06, | |
| "loss": 0.3759, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.9625748502994012, | |
| "grad_norm": 0.21447789669036865, | |
| "learning_rate": 5.3550846323758666e-06, | |
| "loss": 0.2464, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.9640718562874252, | |
| "grad_norm": 0.25622251629829407, | |
| "learning_rate": 5.343267430859559e-06, | |
| "loss": 0.2389, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.9655688622754491, | |
| "grad_norm": 0.2348499894142151, | |
| "learning_rate": 5.331448302505736e-06, | |
| "loss": 0.2629, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.9670658682634731, | |
| "grad_norm": 0.2669554650783539, | |
| "learning_rate": 5.319627313657829e-06, | |
| "loss": 0.2561, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.968562874251497, | |
| "grad_norm": 0.30076301097869873, | |
| "learning_rate": 5.3078045306697154e-06, | |
| "loss": 0.4078, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.9700598802395209, | |
| "grad_norm": 0.23209704458713531, | |
| "learning_rate": 5.295980019905342e-06, | |
| "loss": 0.2687, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.9715568862275449, | |
| "grad_norm": 0.25834420323371887, | |
| "learning_rate": 5.284153847738356e-06, | |
| "loss": 0.2643, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.9730538922155688, | |
| "grad_norm": 0.27238282561302185, | |
| "learning_rate": 5.272326080551729e-06, | |
| "loss": 0.1859, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9745508982035929, | |
| "grad_norm": 0.21913808584213257, | |
| "learning_rate": 5.260496784737386e-06, | |
| "loss": 0.2564, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.9760479041916168, | |
| "grad_norm": 0.3546600043773651, | |
| "learning_rate": 5.248666026695835e-06, | |
| "loss": 0.4955, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.9775449101796407, | |
| "grad_norm": 0.20646998286247253, | |
| "learning_rate": 5.236833872835785e-06, | |
| "loss": 0.2132, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.9790419161676647, | |
| "grad_norm": 0.2146719992160797, | |
| "learning_rate": 5.2250003895737865e-06, | |
| "loss": 0.2016, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.9805389221556886, | |
| "grad_norm": 0.27419614791870117, | |
| "learning_rate": 5.213165643333851e-06, | |
| "loss": 0.3216, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.9820359281437125, | |
| "grad_norm": 0.3328908681869507, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.2583, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.9835329341317365, | |
| "grad_norm": 0.4016421437263489, | |
| "learning_rate": 5.1894926276512824e-06, | |
| "loss": 0.2865, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.9850299401197605, | |
| "grad_norm": 0.22060658037662506, | |
| "learning_rate": 5.177654491090627e-06, | |
| "loss": 0.3583, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.9865269461077845, | |
| "grad_norm": 0.2804822325706482, | |
| "learning_rate": 5.1658153573152405e-06, | |
| "loss": 0.4259, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.9880239520958084, | |
| "grad_norm": 0.35486575961112976, | |
| "learning_rate": 5.153975292780852e-06, | |
| "loss": 0.2605, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9895209580838323, | |
| "grad_norm": 0.32918664813041687, | |
| "learning_rate": 5.1421343639484165e-06, | |
| "loss": 0.3118, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.9910179640718563, | |
| "grad_norm": 0.3175852298736572, | |
| "learning_rate": 5.130292637283735e-06, | |
| "loss": 0.3451, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.9925149700598802, | |
| "grad_norm": 0.23685918748378754, | |
| "learning_rate": 5.118450179257091e-06, | |
| "loss": 0.3018, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.9940119760479041, | |
| "grad_norm": 0.32527291774749756, | |
| "learning_rate": 5.1066070563428736e-06, | |
| "loss": 0.5126, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.9955089820359282, | |
| "grad_norm": 0.2735530734062195, | |
| "learning_rate": 5.0947633350192035e-06, | |
| "loss": 0.2648, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.9970059880239521, | |
| "grad_norm": 0.2354782074689865, | |
| "learning_rate": 5.082919081767558e-06, | |
| "loss": 0.3093, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.9985029940119761, | |
| "grad_norm": 0.24940013885498047, | |
| "learning_rate": 5.071074363072403e-06, | |
| "loss": 0.2115, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.19145430624485016, | |
| "learning_rate": 5.059229245420819e-06, | |
| "loss": 0.2216, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.2865866720676422, | |
| "eval_runtime": 96.0569, | |
| "eval_samples_per_second": 7.662, | |
| "eval_steps_per_second": 0.958, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.001497005988024, | |
| "grad_norm": 0.3310520648956299, | |
| "learning_rate": 5.047383795302119e-06, | |
| "loss": 0.202, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.0029940119760479, | |
| "grad_norm": 0.21326784789562225, | |
| "learning_rate": 5.035538079207488e-06, | |
| "loss": 0.2274, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0044910179640718, | |
| "grad_norm": 0.1706678867340088, | |
| "learning_rate": 5.023692163629603e-06, | |
| "loss": 0.1668, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.0059880239520957, | |
| "grad_norm": 0.2599492073059082, | |
| "learning_rate": 5.01184611506226e-06, | |
| "loss": 0.1964, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.0074850299401197, | |
| "grad_norm": 0.4854053556919098, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4238, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.0089820359281436, | |
| "grad_norm": 0.2861318290233612, | |
| "learning_rate": 4.988153884937742e-06, | |
| "loss": 0.3526, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.0104790419161678, | |
| "grad_norm": 0.3655579388141632, | |
| "learning_rate": 4.9763078363703975e-06, | |
| "loss": 0.3014, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.0119760479041917, | |
| "grad_norm": 0.22971998155117035, | |
| "learning_rate": 4.964461920792512e-06, | |
| "loss": 0.3414, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.0134730538922156, | |
| "grad_norm": 0.3435530364513397, | |
| "learning_rate": 4.952616204697882e-06, | |
| "loss": 0.316, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.0149700598802396, | |
| "grad_norm": 0.15627485513687134, | |
| "learning_rate": 4.940770754579183e-06, | |
| "loss": 0.1875, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.0164670658682635, | |
| "grad_norm": 0.34559714794158936, | |
| "learning_rate": 4.928925636927597e-06, | |
| "loss": 0.3933, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.0179640718562875, | |
| "grad_norm": 0.28742656111717224, | |
| "learning_rate": 4.917080918232444e-06, | |
| "loss": 0.4621, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0194610778443114, | |
| "grad_norm": 0.21044401824474335, | |
| "learning_rate": 4.905236664980797e-06, | |
| "loss": 0.282, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.0209580838323353, | |
| "grad_norm": 0.2607158124446869, | |
| "learning_rate": 4.893392943657127e-06, | |
| "loss": 0.3244, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.001497005988024, | |
| "grad_norm": 0.2220315784215927, | |
| "learning_rate": 4.88154982074291e-06, | |
| "loss": 0.2843, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.0029940119760479, | |
| "grad_norm": 0.2902830243110657, | |
| "learning_rate": 4.8697073627162675e-06, | |
| "loss": 0.3467, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.0044910179640718, | |
| "grad_norm": 0.2681308388710022, | |
| "learning_rate": 4.857865636051586e-06, | |
| "loss": 0.2249, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.0059880239520957, | |
| "grad_norm": 0.2687780559062958, | |
| "learning_rate": 4.846024707219149e-06, | |
| "loss": 0.3296, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.0074850299401197, | |
| "grad_norm": 0.3211975693702698, | |
| "learning_rate": 4.834184642684762e-06, | |
| "loss": 0.2454, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.0089820359281436, | |
| "grad_norm": 0.27423396706581116, | |
| "learning_rate": 4.822345508909376e-06, | |
| "loss": 0.1711, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.0104790419161676, | |
| "grad_norm": 0.17680241167545319, | |
| "learning_rate": 4.810507372348721e-06, | |
| "loss": 0.2281, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.0119760479041917, | |
| "grad_norm": 0.299165278673172, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.2054, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0134730538922156, | |
| "grad_norm": 0.181101456284523, | |
| "learning_rate": 4.786834356666153e-06, | |
| "loss": 0.1524, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.0149700598802396, | |
| "grad_norm": 0.28579655289649963, | |
| "learning_rate": 4.774999610426216e-06, | |
| "loss": 0.3557, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.0164670658682635, | |
| "grad_norm": 0.21840621531009674, | |
| "learning_rate": 4.7631661271642185e-06, | |
| "loss": 0.2558, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.0179640718562875, | |
| "grad_norm": 0.25449657440185547, | |
| "learning_rate": 4.751333973304166e-06, | |
| "loss": 0.4695, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.0194610778443114, | |
| "grad_norm": 0.2505941092967987, | |
| "learning_rate": 4.739503215262614e-06, | |
| "loss": 0.2168, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.0209580838323353, | |
| "grad_norm": 0.1709732562303543, | |
| "learning_rate": 4.727673919448271e-06, | |
| "loss": 0.2167, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.0224550898203593, | |
| "grad_norm": 0.2961471378803253, | |
| "learning_rate": 4.715846152261645e-06, | |
| "loss": 0.2312, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.0239520958083832, | |
| "grad_norm": 0.21817496418952942, | |
| "learning_rate": 4.704019980094659e-06, | |
| "loss": 0.2802, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.0254491017964071, | |
| "grad_norm": 0.2459694743156433, | |
| "learning_rate": 4.692195469330286e-06, | |
| "loss": 0.2981, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.026946107784431, | |
| "grad_norm": 0.24430765211582184, | |
| "learning_rate": 4.680372686342173e-06, | |
| "loss": 0.2718, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.028443113772455, | |
| "grad_norm": 0.27209728956222534, | |
| "learning_rate": 4.668551697494265e-06, | |
| "loss": 0.4779, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.029940119760479, | |
| "grad_norm": 0.2800404727458954, | |
| "learning_rate": 4.656732569140441e-06, | |
| "loss": 0.2695, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.031437125748503, | |
| "grad_norm": 0.21455827355384827, | |
| "learning_rate": 4.644915367624134e-06, | |
| "loss": 0.2927, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.032934131736527, | |
| "grad_norm": 0.3750689625740051, | |
| "learning_rate": 4.6331001592779615e-06, | |
| "loss": 0.3039, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.034431137724551, | |
| "grad_norm": 0.35561925172805786, | |
| "learning_rate": 4.621287010423353e-06, | |
| "loss": 0.3393, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.035928143712575, | |
| "grad_norm": 0.3378138244152069, | |
| "learning_rate": 4.609475987370177e-06, | |
| "loss": 0.3245, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.0374251497005988, | |
| "grad_norm": 0.24566203355789185, | |
| "learning_rate": 4.597667156416371e-06, | |
| "loss": 0.2596, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.0389221556886228, | |
| "grad_norm": 0.17309314012527466, | |
| "learning_rate": 4.585860583847566e-06, | |
| "loss": 0.2294, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.0404191616766467, | |
| "grad_norm": 0.3359794616699219, | |
| "learning_rate": 4.5740563359367164e-06, | |
| "loss": 0.4138, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.0419161676646707, | |
| "grad_norm": 0.24172459542751312, | |
| "learning_rate": 4.562254478943729e-06, | |
| "loss": 0.2675, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.0434131736526946, | |
| "grad_norm": 0.22163425385951996, | |
| "learning_rate": 4.550455079115091e-06, | |
| "loss": 0.4267, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.0449101796407185, | |
| "grad_norm": 0.2985590696334839, | |
| "learning_rate": 4.53865820268349e-06, | |
| "loss": 0.3657, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.0464071856287425, | |
| "grad_norm": 0.2542339861392975, | |
| "learning_rate": 4.52686391586746e-06, | |
| "loss": 0.2772, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.0479041916167664, | |
| "grad_norm": 0.253623902797699, | |
| "learning_rate": 4.51507228487099e-06, | |
| "loss": 0.4212, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.0494011976047903, | |
| "grad_norm": 0.3196866512298584, | |
| "learning_rate": 4.503283375883165e-06, | |
| "loss": 0.2925, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.0508982035928143, | |
| "grad_norm": 0.22368155419826508, | |
| "learning_rate": 4.49149725507779e-06, | |
| "loss": 0.2496, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.0523952095808382, | |
| "grad_norm": 0.32750844955444336, | |
| "learning_rate": 4.479713988613021e-06, | |
| "loss": 0.1986, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.0538922155688624, | |
| "grad_norm": 0.3797385096549988, | |
| "learning_rate": 4.467933642630989e-06, | |
| "loss": 0.2568, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.0553892215568863, | |
| "grad_norm": 0.3641880750656128, | |
| "learning_rate": 4.456156283257433e-06, | |
| "loss": 0.2927, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.0568862275449102, | |
| "grad_norm": 0.24843014776706696, | |
| "learning_rate": 4.44438197660133e-06, | |
| "loss": 0.2483, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.0583832335329342, | |
| "grad_norm": 0.22964823246002197, | |
| "learning_rate": 4.432610788754517e-06, | |
| "loss": 0.2175, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.0598802395209581, | |
| "grad_norm": 0.23855966329574585, | |
| "learning_rate": 4.420842785791326e-06, | |
| "loss": 0.3055, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.061377245508982, | |
| "grad_norm": 0.3656154274940491, | |
| "learning_rate": 4.409078033768214e-06, | |
| "loss": 0.3667, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.062874251497006, | |
| "grad_norm": 0.23534388840198517, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.2423, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.06437125748503, | |
| "grad_norm": 0.29222187399864197, | |
| "learning_rate": 4.3855585466764305e-06, | |
| "loss": 0.2366, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.0658682634730539, | |
| "grad_norm": 0.3136173188686371, | |
| "learning_rate": 4.373803943627946e-06, | |
| "loss": 0.2871, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.0673652694610778, | |
| "grad_norm": 0.24266062676906586, | |
| "learning_rate": 4.362052855559171e-06, | |
| "loss": 0.2742, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.0688622754491017, | |
| "grad_norm": 0.24898295104503632, | |
| "learning_rate": 4.350305348431612e-06, | |
| "loss": 0.3362, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.0703592814371257, | |
| "grad_norm": 0.28294843435287476, | |
| "learning_rate": 4.338561488186678e-06, | |
| "loss": 0.2887, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.0718562874251496, | |
| "grad_norm": 0.28065305948257446, | |
| "learning_rate": 4.326821340745304e-06, | |
| "loss": 0.3043, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.0733532934131738, | |
| "grad_norm": 0.23319000005722046, | |
| "learning_rate": 4.315084972007587e-06, | |
| "loss": 0.214, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.0748502994011977, | |
| "grad_norm": 0.2753102481365204, | |
| "learning_rate": 4.303352447852412e-06, | |
| "loss": 0.2536, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.0763473053892216, | |
| "grad_norm": 0.2678392827510834, | |
| "learning_rate": 4.291623834137082e-06, | |
| "loss": 0.3256, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.0778443113772456, | |
| "grad_norm": 0.39115357398986816, | |
| "learning_rate": 4.279899196696953e-06, | |
| "loss": 0.3188, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.0793413173652695, | |
| "grad_norm": 0.312159925699234, | |
| "learning_rate": 4.268178601345057e-06, | |
| "loss": 0.2588, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.0808383233532934, | |
| "grad_norm": 0.18832477927207947, | |
| "learning_rate": 4.256462113871741e-06, | |
| "loss": 0.2005, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.0823353293413174, | |
| "grad_norm": 0.3424656391143799, | |
| "learning_rate": 4.2447498000442935e-06, | |
| "loss": 0.2885, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.0838323353293413, | |
| "grad_norm": 0.1885494589805603, | |
| "learning_rate": 4.233041725606573e-06, | |
| "loss": 0.1686, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.0853293413173652, | |
| "grad_norm": 0.25505173206329346, | |
| "learning_rate": 4.2213379562786406e-06, | |
| "loss": 0.3603, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.0868263473053892, | |
| "grad_norm": 0.2631552815437317, | |
| "learning_rate": 4.209638557756396e-06, | |
| "loss": 0.3339, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.0883233532934131, | |
| "grad_norm": 0.2027507871389389, | |
| "learning_rate": 4.1979435957111984e-06, | |
| "loss": 0.4054, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.089820359281437, | |
| "grad_norm": 0.24192075431346893, | |
| "learning_rate": 4.186253135789511e-06, | |
| "loss": 0.319, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.091317365269461, | |
| "grad_norm": 0.23616907000541687, | |
| "learning_rate": 4.1745672436125205e-06, | |
| "loss": 0.2824, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.092814371257485, | |
| "grad_norm": 0.18234433233737946, | |
| "learning_rate": 4.162885984775777e-06, | |
| "loss": 0.1787, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.0943113772455089, | |
| "grad_norm": 0.2815669775009155, | |
| "learning_rate": 4.15120942484882e-06, | |
| "loss": 0.389, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.095808383233533, | |
| "grad_norm": 0.255751371383667, | |
| "learning_rate": 4.139537629374814e-06, | |
| "loss": 0.2521, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.097305389221557, | |
| "grad_norm": 0.25240209698677063, | |
| "learning_rate": 4.12787066387018e-06, | |
| "loss": 0.2305, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.098802395209581, | |
| "grad_norm": 0.24336189031600952, | |
| "learning_rate": 4.116208593824227e-06, | |
| "loss": 0.3735, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.1002994011976048, | |
| "grad_norm": 0.25405654311180115, | |
| "learning_rate": 4.104551484698785e-06, | |
| "loss": 0.1988, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.1017964071856288, | |
| "grad_norm": 0.3620491325855255, | |
| "learning_rate": 4.092899401927836e-06, | |
| "loss": 0.4211, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1032934131736527, | |
| "grad_norm": 0.37368297576904297, | |
| "learning_rate": 4.081252410917148e-06, | |
| "loss": 0.346, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.1047904191616766, | |
| "grad_norm": 0.20550665259361267, | |
| "learning_rate": 4.069610577043912e-06, | |
| "loss": 0.2211, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.1062874251497006, | |
| "grad_norm": 0.20395252108573914, | |
| "learning_rate": 4.057973965656365e-06, | |
| "loss": 0.1803, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.1077844311377245, | |
| "grad_norm": 0.16643236577510834, | |
| "learning_rate": 4.046342642073433e-06, | |
| "loss": 0.202, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.1092814371257484, | |
| "grad_norm": 0.32340267300605774, | |
| "learning_rate": 4.034716671584357e-06, | |
| "loss": 0.2943, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.1107784431137724, | |
| "grad_norm": 0.218381866812706, | |
| "learning_rate": 4.0230961194483325e-06, | |
| "loss": 0.2293, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.1122754491017963, | |
| "grad_norm": 0.2537286579608917, | |
| "learning_rate": 4.01148105089414e-06, | |
| "loss": 0.2854, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.1137724550898203, | |
| "grad_norm": 0.24806788563728333, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.3641, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.1152694610778444, | |
| "grad_norm": 0.32125091552734375, | |
| "learning_rate": 3.988267625292102e-06, | |
| "loss": 0.3263, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.1167664670658684, | |
| "grad_norm": 0.23608674108982086, | |
| "learning_rate": 3.976669398546451e-06, | |
| "loss": 0.237, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1182634730538923, | |
| "grad_norm": 0.2589946687221527, | |
| "learning_rate": 3.9650769159862875e-06, | |
| "loss": 0.3088, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.1197604790419162, | |
| "grad_norm": 0.2909379005432129, | |
| "learning_rate": 3.9534902426828325e-06, | |
| "loss": 0.2574, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.1212574850299402, | |
| "grad_norm": 0.1950826793909073, | |
| "learning_rate": 3.941909443674696e-06, | |
| "loss": 0.1856, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.122754491017964, | |
| "grad_norm": 0.21967299282550812, | |
| "learning_rate": 3.930334583967514e-06, | |
| "loss": 0.2938, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.124251497005988, | |
| "grad_norm": 0.225912407040596, | |
| "learning_rate": 3.918765728533586e-06, | |
| "loss": 0.1812, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.125748502994012, | |
| "grad_norm": 0.26577168703079224, | |
| "learning_rate": 3.907202942311506e-06, | |
| "loss": 0.3021, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.127245508982036, | |
| "grad_norm": 0.20895415544509888, | |
| "learning_rate": 3.895646290205803e-06, | |
| "loss": 0.2644, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.1287425149700598, | |
| "grad_norm": 0.2302289605140686, | |
| "learning_rate": 3.884095837086571e-06, | |
| "loss": 0.2647, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.1302395209580838, | |
| "grad_norm": 0.28787434101104736, | |
| "learning_rate": 3.872551647789108e-06, | |
| "loss": 0.3042, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.1317365269461077, | |
| "grad_norm": 0.2890431880950928, | |
| "learning_rate": 3.861013787113553e-06, | |
| "loss": 0.3888, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.1332335329341316, | |
| "grad_norm": 0.2250119298696518, | |
| "learning_rate": 3.849482319824521e-06, | |
| "loss": 0.3427, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.1347305389221556, | |
| "grad_norm": 0.3872185945510864, | |
| "learning_rate": 3.837957310650738e-06, | |
| "loss": 0.4126, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.1362275449101795, | |
| "grad_norm": 0.3162260353565216, | |
| "learning_rate": 3.82643882428468e-06, | |
| "loss": 0.2721, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.1377245508982037, | |
| "grad_norm": 0.2870970368385315, | |
| "learning_rate": 3.81492692538221e-06, | |
| "loss": 0.3195, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.1392215568862276, | |
| "grad_norm": 0.27217966318130493, | |
| "learning_rate": 3.803421678562213e-06, | |
| "loss": 0.394, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.1407185628742516, | |
| "grad_norm": 0.32144343852996826, | |
| "learning_rate": 3.7919231484062334e-06, | |
| "loss": 0.4021, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.1422155688622755, | |
| "grad_norm": 0.2571192979812622, | |
| "learning_rate": 3.7804313994581143e-06, | |
| "loss": 0.3301, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.1437125748502994, | |
| "grad_norm": 0.19559802114963531, | |
| "learning_rate": 3.7689464962236367e-06, | |
| "loss": 0.3152, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.1452095808383234, | |
| "grad_norm": 0.21712540090084076, | |
| "learning_rate": 3.757468503170153e-06, | |
| "loss": 0.2456, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.1467065868263473, | |
| "grad_norm": 0.2776012420654297, | |
| "learning_rate": 3.7459974847262253e-06, | |
| "loss": 0.3182, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.1482035928143712, | |
| "grad_norm": 0.2935360372066498, | |
| "learning_rate": 3.734533505281269e-06, | |
| "loss": 0.2671, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.1497005988023952, | |
| "grad_norm": 0.30140233039855957, | |
| "learning_rate": 3.723076629185186e-06, | |
| "loss": 0.3733, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.151197604790419, | |
| "grad_norm": 0.2267109453678131, | |
| "learning_rate": 3.7116269207480055e-06, | |
| "loss": 0.3015, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.152694610778443, | |
| "grad_norm": 0.24247941374778748, | |
| "learning_rate": 3.700184444239524e-06, | |
| "loss": 0.1761, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.154191616766467, | |
| "grad_norm": 0.14026691019535065, | |
| "learning_rate": 3.6887492638889433e-06, | |
| "loss": 0.152, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.1556886227544911, | |
| "grad_norm": 0.24493476748466492, | |
| "learning_rate": 3.677321443884509e-06, | |
| "loss": 0.2549, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.157185628742515, | |
| "grad_norm": 0.22861488163471222, | |
| "learning_rate": 3.6659010483731543e-06, | |
| "loss": 0.2145, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.158682634730539, | |
| "grad_norm": 0.21037505567073822, | |
| "learning_rate": 3.654488141460134e-06, | |
| "loss": 0.2116, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.160179640718563, | |
| "grad_norm": 0.24431665241718292, | |
| "learning_rate": 3.6430827872086694e-06, | |
| "loss": 0.3154, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.1616766467065869, | |
| "grad_norm": 0.288648784160614, | |
| "learning_rate": 3.6316850496395863e-06, | |
| "loss": 0.4031, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.1631736526946108, | |
| "grad_norm": 0.2301226109266281, | |
| "learning_rate": 3.6202949927309555e-06, | |
| "loss": 0.2984, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.1646706586826348, | |
| "grad_norm": 0.3974512815475464, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.341, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.1661676646706587, | |
| "grad_norm": 0.25364506244659424, | |
| "learning_rate": 3.597538176591417e-06, | |
| "loss": 0.2507, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.1676646706586826, | |
| "grad_norm": 0.1950305849313736, | |
| "learning_rate": 3.5861715450996505e-06, | |
| "loss": 0.1766, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.1691616766467066, | |
| "grad_norm": 0.265399694442749, | |
| "learning_rate": 3.5748128497459044e-06, | |
| "loss": 0.2574, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.1706586826347305, | |
| "grad_norm": 0.3101622462272644, | |
| "learning_rate": 3.563462154289098e-06, | |
| "loss": 0.3014, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.1721556886227544, | |
| "grad_norm": 0.22397872805595398, | |
| "learning_rate": 3.5521195224432436e-06, | |
| "loss": 0.2439, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.1736526946107784, | |
| "grad_norm": 0.22311492264270782, | |
| "learning_rate": 3.5407850178770944e-06, | |
| "loss": 0.1595, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.1751497005988023, | |
| "grad_norm": 0.24088788032531738, | |
| "learning_rate": 3.5294587042137796e-06, | |
| "loss": 0.2284, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.1766467065868262, | |
| "grad_norm": 0.30875658988952637, | |
| "learning_rate": 3.5181406450304536e-06, | |
| "loss": 0.3085, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.1781437125748502, | |
| "grad_norm": 0.2661195993423462, | |
| "learning_rate": 3.506830903857933e-06, | |
| "loss": 0.3207, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.1796407185628743, | |
| "grad_norm": 0.237378790974617, | |
| "learning_rate": 3.49552954418035e-06, | |
| "loss": 0.3883, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.1811377245508983, | |
| "grad_norm": 0.2710888385772705, | |
| "learning_rate": 3.484236629434783e-06, | |
| "loss": 0.2923, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.1826347305389222, | |
| "grad_norm": 0.2072073370218277, | |
| "learning_rate": 3.4729522230109103e-06, | |
| "loss": 0.2178, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.1841317365269461, | |
| "grad_norm": 0.22102674841880798, | |
| "learning_rate": 3.461676388250651e-06, | |
| "loss": 0.2868, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.18562874251497, | |
| "grad_norm": 0.28038355708122253, | |
| "learning_rate": 3.4504091884478076e-06, | |
| "loss": 0.3112, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.187125748502994, | |
| "grad_norm": 0.22115986049175262, | |
| "learning_rate": 3.4391506868477153e-06, | |
| "loss": 0.2361, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.188622754491018, | |
| "grad_norm": 0.24288050830364227, | |
| "learning_rate": 3.4279009466468825e-06, | |
| "loss": 0.2372, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.1901197604790419, | |
| "grad_norm": 0.225479856133461, | |
| "learning_rate": 3.416660030992639e-06, | |
| "loss": 0.2474, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.1916167664670658, | |
| "grad_norm": 0.19721685349941254, | |
| "learning_rate": 3.405428002982779e-06, | |
| "loss": 0.1992, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.1931137724550898, | |
| "grad_norm": 0.27366501092910767, | |
| "learning_rate": 3.3942049256652093e-06, | |
| "loss": 0.3074, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.1946107784431137, | |
| "grad_norm": 0.3281024694442749, | |
| "learning_rate": 3.3829908620375953e-06, | |
| "loss": 0.3376, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.1961077844311376, | |
| "grad_norm": 0.20951387286186218, | |
| "learning_rate": 3.3717858750470046e-06, | |
| "loss": 0.2649, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.1976047904191618, | |
| "grad_norm": 0.2254786491394043, | |
| "learning_rate": 3.3605900275895565e-06, | |
| "loss": 0.3744, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.1991017964071857, | |
| "grad_norm": 0.31041470170021057, | |
| "learning_rate": 3.349403382510068e-06, | |
| "loss": 0.3571, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.2005988023952097, | |
| "grad_norm": 0.3175371289253235, | |
| "learning_rate": 3.3382260026017027e-06, | |
| "loss": 0.2337, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.2020958083832336, | |
| "grad_norm": 0.23739373683929443, | |
| "learning_rate": 3.3270579506056146e-06, | |
| "loss": 0.3272, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.2035928143712575, | |
| "grad_norm": 0.216998890042305, | |
| "learning_rate": 3.3158992892105975e-06, | |
| "loss": 0.2777, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.2050898203592815, | |
| "grad_norm": 0.25838586688041687, | |
| "learning_rate": 3.3047500810527343e-06, | |
| "loss": 0.2233, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.2065868263473054, | |
| "grad_norm": 0.2895229756832123, | |
| "learning_rate": 3.2936103887150484e-06, | |
| "loss": 0.3911, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2080838323353293, | |
| "grad_norm": 0.27420610189437866, | |
| "learning_rate": 3.2824802747271424e-06, | |
| "loss": 0.381, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.2095808383233533, | |
| "grad_norm": 0.24717342853546143, | |
| "learning_rate": 3.271359801564858e-06, | |
| "loss": 0.3571, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.2110778443113772, | |
| "grad_norm": 0.23982776701450348, | |
| "learning_rate": 3.2602490316499197e-06, | |
| "loss": 0.2092, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.2125748502994012, | |
| "grad_norm": 0.25339189171791077, | |
| "learning_rate": 3.2491480273495847e-06, | |
| "loss": 0.263, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.214071856287425, | |
| "grad_norm": 0.18809233605861664, | |
| "learning_rate": 3.2380568509762935e-06, | |
| "loss": 0.2132, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.215568862275449, | |
| "grad_norm": 0.24008889496326447, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.3366, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.217065868263473, | |
| "grad_norm": 0.2413085252046585, | |
| "learning_rate": 3.215904230984428e-06, | |
| "loss": 0.2753, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.218562874251497, | |
| "grad_norm": 0.2766128182411194, | |
| "learning_rate": 3.204842911713506e-06, | |
| "loss": 0.3486, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.220059880239521, | |
| "grad_norm": 0.24891407787799835, | |
| "learning_rate": 3.1937916690642356e-06, | |
| "loss": 0.2632, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.221556886227545, | |
| "grad_norm": 0.2821076214313507, | |
| "learning_rate": 3.182750565069735e-06, | |
| "loss": 0.202, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.223053892215569, | |
| "grad_norm": 0.2556501626968384, | |
| "learning_rate": 3.171719661706211e-06, | |
| "loss": 0.3678, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.2245508982035929, | |
| "grad_norm": 0.304360032081604, | |
| "learning_rate": 3.1606990208926125e-06, | |
| "loss": 0.3085, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.2260479041916168, | |
| "grad_norm": 0.25805479288101196, | |
| "learning_rate": 3.1496887044902815e-06, | |
| "loss": 0.2892, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.2275449101796407, | |
| "grad_norm": 0.2673249840736389, | |
| "learning_rate": 3.1386887743026083e-06, | |
| "loss": 0.2122, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.2290419161676647, | |
| "grad_norm": 0.21123580634593964, | |
| "learning_rate": 3.127699292074683e-06, | |
| "loss": 0.2108, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.2290419161676647, | |
| "eval_loss": 0.27716606855392456, | |
| "eval_runtime": 95.8808, | |
| "eval_samples_per_second": 7.676, | |
| "eval_steps_per_second": 0.96, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.2305389221556886, | |
| "grad_norm": 0.23262470960617065, | |
| "learning_rate": 3.1167203194929447e-06, | |
| "loss": 0.2687, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.2320359281437125, | |
| "grad_norm": 0.23696181178092957, | |
| "learning_rate": 3.1057519181848474e-06, | |
| "loss": 0.2368, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.2335329341317365, | |
| "grad_norm": 0.17337414622306824, | |
| "learning_rate": 3.0947941497184985e-06, | |
| "loss": 0.2181, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.2350299401197604, | |
| "grad_norm": 0.299317330121994, | |
| "learning_rate": 3.0838470756023253e-06, | |
| "loss": 0.2551, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.2365269461077844, | |
| "grad_norm": 0.33490660786628723, | |
| "learning_rate": 3.0729107572847244e-06, | |
| "loss": 0.2473, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.2380239520958083, | |
| "grad_norm": 0.4856211841106415, | |
| "learning_rate": 3.0619852561537165e-06, | |
| "loss": 0.3014, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.2395209580838324, | |
| "grad_norm": 0.26796865463256836, | |
| "learning_rate": 3.0510706335366034e-06, | |
| "loss": 0.2985, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.2410179640718564, | |
| "grad_norm": 0.27012571692466736, | |
| "learning_rate": 3.040166950699626e-06, | |
| "loss": 0.2157, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.2425149700598803, | |
| "grad_norm": 0.2794853150844574, | |
| "learning_rate": 3.0292742688476125e-06, | |
| "loss": 0.3638, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.2440119760479043, | |
| "grad_norm": 0.243549183011055, | |
| "learning_rate": 3.018392649123645e-06, | |
| "loss": 0.2581, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.2455089820359282, | |
| "grad_norm": 0.23559677600860596, | |
| "learning_rate": 3.0075221526087083e-06, | |
| "loss": 0.2066, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.2470059880239521, | |
| "grad_norm": 0.21091359853744507, | |
| "learning_rate": 2.9966628403213528e-06, | |
| "loss": 0.1868, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.248502994011976, | |
| "grad_norm": 0.2931704521179199, | |
| "learning_rate": 2.985814773217346e-06, | |
| "loss": 0.3558, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.3139488399028778, | |
| "learning_rate": 2.9749780121893366e-06, | |
| "loss": 0.3084, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.251497005988024, | |
| "grad_norm": 0.3278771638870239, | |
| "learning_rate": 2.964152618066508e-06, | |
| "loss": 0.3552, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.2529940119760479, | |
| "grad_norm": 0.17210358381271362, | |
| "learning_rate": 2.9533386516142402e-06, | |
| "loss": 0.1554, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.2544910179640718, | |
| "grad_norm": 0.23768174648284912, | |
| "learning_rate": 2.9425361735337655e-06, | |
| "loss": 0.2715, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.2559880239520957, | |
| "grad_norm": 0.26038113236427307, | |
| "learning_rate": 2.93174524446183e-06, | |
| "loss": 0.2104, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.2574850299401197, | |
| "grad_norm": 0.2509795129299164, | |
| "learning_rate": 2.920965924970352e-06, | |
| "loss": 0.2998, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.2589820359281436, | |
| "grad_norm": 0.18279989063739777, | |
| "learning_rate": 2.910198275566085e-06, | |
| "loss": 0.1925, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.2604790419161676, | |
| "grad_norm": 0.2960183620452881, | |
| "learning_rate": 2.899442356690271e-06, | |
| "loss": 0.2593, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.2619760479041915, | |
| "grad_norm": 0.3304743766784668, | |
| "learning_rate": 2.8886982287183092e-06, | |
| "loss": 0.283, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.2634730538922156, | |
| "grad_norm": 0.31268325448036194, | |
| "learning_rate": 2.8779659519594173e-06, | |
| "loss": 0.3402, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.2649700598802396, | |
| "grad_norm": 0.3021641671657562, | |
| "learning_rate": 2.8672455866562797e-06, | |
| "loss": 0.2746, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.2664670658682635, | |
| "grad_norm": 0.2867257297039032, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.4056, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.2679640718562875, | |
| "grad_norm": 0.26254844665527344, | |
| "learning_rate": 2.8458408310533948e-06, | |
| "loss": 0.1884, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.2694610778443114, | |
| "grad_norm": 0.25362345576286316, | |
| "learning_rate": 2.835156560903365e-06, | |
| "loss": 0.3639, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.2709580838323353, | |
| "grad_norm": 0.2275954782962799, | |
| "learning_rate": 2.824484442507863e-06, | |
| "loss": 0.2712, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.2724550898203593, | |
| "grad_norm": 0.30598387122154236, | |
| "learning_rate": 2.813824535771892e-06, | |
| "loss": 0.324, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.2739520958083832, | |
| "grad_norm": 0.20137560367584229, | |
| "learning_rate": 2.803176900531915e-06, | |
| "loss": 0.2698, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.2754491017964071, | |
| "grad_norm": 0.2496221363544464, | |
| "learning_rate": 2.7925415965555126e-06, | |
| "loss": 0.2704, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.276946107784431, | |
| "grad_norm": 0.20142346620559692, | |
| "learning_rate": 2.78191868354104e-06, | |
| "loss": 0.3, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.278443113772455, | |
| "grad_norm": 0.303371787071228, | |
| "learning_rate": 2.771308221117309e-06, | |
| "loss": 0.3695, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.2799401197604792, | |
| "grad_norm": 0.30154815316200256, | |
| "learning_rate": 2.760710268843234e-06, | |
| "loss": 0.1902, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.281437125748503, | |
| "grad_norm": 0.24857012927532196, | |
| "learning_rate": 2.7501248862075163e-06, | |
| "loss": 0.3152, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.282934131736527, | |
| "grad_norm": 0.21458423137664795, | |
| "learning_rate": 2.7395521326282913e-06, | |
| "loss": 0.1814, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.284431137724551, | |
| "grad_norm": 0.27455374598503113, | |
| "learning_rate": 2.7289920674528142e-06, | |
| "loss": 0.3064, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.285928143712575, | |
| "grad_norm": 0.2155693769454956, | |
| "learning_rate": 2.718444749957109e-06, | |
| "loss": 0.2616, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.2874251497005988, | |
| "grad_norm": 0.2787468433380127, | |
| "learning_rate": 2.7079102393456503e-06, | |
| "loss": 0.4756, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.2889221556886228, | |
| "grad_norm": 0.24366062879562378, | |
| "learning_rate": 2.69738859475102e-06, | |
| "loss": 0.3375, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.2904191616766467, | |
| "grad_norm": 0.3134368658065796, | |
| "learning_rate": 2.6868798752335867e-06, | |
| "loss": 0.305, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.2919161676646707, | |
| "grad_norm": 0.15974688529968262, | |
| "learning_rate": 2.6763841397811576e-06, | |
| "loss": 0.1503, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.2934131736526946, | |
| "grad_norm": 0.2607382833957672, | |
| "learning_rate": 2.6659014473086665e-06, | |
| "loss": 0.2365, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.2949101796407185, | |
| "grad_norm": 0.23047217726707458, | |
| "learning_rate": 2.655431856657833e-06, | |
| "loss": 0.215, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.2964071856287425, | |
| "grad_norm": 0.23638281226158142, | |
| "learning_rate": 2.6449754265968263e-06, | |
| "loss": 0.3338, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.2979041916167664, | |
| "grad_norm": 0.25799238681793213, | |
| "learning_rate": 2.6345322158199503e-06, | |
| "loss": 0.2688, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.2994011976047903, | |
| "grad_norm": 0.22320988774299622, | |
| "learning_rate": 2.6241022829473e-06, | |
| "loss": 0.3521, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.3008982035928143, | |
| "grad_norm": 0.2436802089214325, | |
| "learning_rate": 2.6136856865244443e-06, | |
| "loss": 0.2838, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.3023952095808382, | |
| "grad_norm": 0.27055466175079346, | |
| "learning_rate": 2.603282485022085e-06, | |
| "loss": 0.3206, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.3038922155688621, | |
| "grad_norm": 0.3011034429073334, | |
| "learning_rate": 2.592892736835742e-06, | |
| "loss": 0.2213, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.3053892215568863, | |
| "grad_norm": 0.3129655718803406, | |
| "learning_rate": 2.5825165002854124e-06, | |
| "loss": 0.2309, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.3068862275449102, | |
| "grad_norm": 0.308241605758667, | |
| "learning_rate": 2.5721538336152553e-06, | |
| "loss": 0.3263, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.3083832335329342, | |
| "grad_norm": 0.3574126660823822, | |
| "learning_rate": 2.5618047949932524e-06, | |
| "loss": 0.198, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.3098802395209581, | |
| "grad_norm": 0.2755787968635559, | |
| "learning_rate": 2.5514694425108968e-06, | |
| "loss": 0.251, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.311377245508982, | |
| "grad_norm": 0.3657899796962738, | |
| "learning_rate": 2.5411478341828475e-06, | |
| "loss": 0.2594, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.312874251497006, | |
| "grad_norm": 0.2408738136291504, | |
| "learning_rate": 2.5308400279466262e-06, | |
| "loss": 0.3013, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.31437125748503, | |
| "grad_norm": 0.21882514655590057, | |
| "learning_rate": 2.5205460816622684e-06, | |
| "loss": 0.2877, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.3158682634730539, | |
| "grad_norm": 0.22941765189170837, | |
| "learning_rate": 2.5102660531120204e-06, | |
| "loss": 0.1921, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.3173652694610778, | |
| "grad_norm": 0.25530606508255005, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.2747, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.3188622754491017, | |
| "grad_norm": 0.2586192786693573, | |
| "learning_rate": 2.4897479799518797e-06, | |
| "loss": 0.2441, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.3203592814371259, | |
| "grad_norm": 0.2807331383228302, | |
| "learning_rate": 2.479510050514561e-06, | |
| "loss": 0.285, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.3218562874251498, | |
| "grad_norm": 0.2594211995601654, | |
| "learning_rate": 2.469286269155848e-06, | |
| "loss": 0.2154, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.3233532934131738, | |
| "grad_norm": 0.3582344055175781, | |
| "learning_rate": 2.4590766932641353e-06, | |
| "loss": 0.3159, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.3248502994011977, | |
| "grad_norm": 0.33185476064682007, | |
| "learning_rate": 2.4488813801480717e-06, | |
| "loss": 0.3023, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.3263473053892216, | |
| "grad_norm": 0.20739290118217468, | |
| "learning_rate": 2.438700387036253e-06, | |
| "loss": 0.2176, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.3278443113772456, | |
| "grad_norm": 0.30346420407295227, | |
| "learning_rate": 2.4285337710768843e-06, | |
| "loss": 0.2937, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.3293413173652695, | |
| "grad_norm": 0.2638108432292938, | |
| "learning_rate": 2.4183815893374817e-06, | |
| "loss": 0.2442, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 1.3308383233532934, | |
| "grad_norm": 0.23496340215206146, | |
| "learning_rate": 2.4082438988045253e-06, | |
| "loss": 0.2105, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.3323353293413174, | |
| "grad_norm": 0.4059658944606781, | |
| "learning_rate": 2.3981207563831633e-06, | |
| "loss": 0.4157, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.3338323353293413, | |
| "grad_norm": 0.28790730237960815, | |
| "learning_rate": 2.388012218896873e-06, | |
| "loss": 0.4796, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.3353293413173652, | |
| "grad_norm": 0.26836466789245605, | |
| "learning_rate": 2.3779183430871596e-06, | |
| "loss": 0.2747, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 1.3368263473053892, | |
| "grad_norm": 0.26263725757598877, | |
| "learning_rate": 2.3678391856132203e-06, | |
| "loss": 0.2096, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 1.3383233532934131, | |
| "grad_norm": 0.32726365327835083, | |
| "learning_rate": 2.3577748030516443e-06, | |
| "loss": 0.373, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.339820359281437, | |
| "grad_norm": 0.295310378074646, | |
| "learning_rate": 2.3477252518960764e-06, | |
| "loss": 0.2601, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 1.341317365269461, | |
| "grad_norm": 0.26851561665534973, | |
| "learning_rate": 2.3376905885569185e-06, | |
| "loss": 0.2078, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.342814371257485, | |
| "grad_norm": 0.33125850558280945, | |
| "learning_rate": 2.3276708693609947e-06, | |
| "loss": 0.3367, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 1.3443113772455089, | |
| "grad_norm": 0.23281192779541016, | |
| "learning_rate": 2.3176661505512534e-06, | |
| "loss": 0.2787, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.3458083832335328, | |
| "grad_norm": 0.5017257928848267, | |
| "learning_rate": 2.3076764882864333e-06, | |
| "loss": 0.3425, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 1.347305389221557, | |
| "grad_norm": 0.249893456697464, | |
| "learning_rate": 2.2977019386407653e-06, | |
| "loss": 0.3495, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 1.348802395209581, | |
| "grad_norm": 0.3614089787006378, | |
| "learning_rate": 2.2877425576036467e-06, | |
| "loss": 0.3012, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.3502994011976048, | |
| "grad_norm": 0.26093000173568726, | |
| "learning_rate": 2.2777984010793264e-06, | |
| "loss": 0.3139, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.3517964071856288, | |
| "grad_norm": 0.28172314167022705, | |
| "learning_rate": 2.267869524886603e-06, | |
| "loss": 0.3505, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.3532934131736527, | |
| "grad_norm": 0.5635291934013367, | |
| "learning_rate": 2.2579559847584924e-06, | |
| "loss": 0.3933, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 1.3547904191616766, | |
| "grad_norm": 0.2114657163619995, | |
| "learning_rate": 2.2480578363419363e-06, | |
| "loss": 0.1635, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 1.3562874251497006, | |
| "grad_norm": 0.28377217054367065, | |
| "learning_rate": 2.238175135197471e-06, | |
| "loss": 0.2308, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.3577844311377245, | |
| "grad_norm": 0.22767485678195953, | |
| "learning_rate": 2.2283079367989303e-06, | |
| "loss": 0.1877, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 1.3592814371257484, | |
| "grad_norm": 0.4151133894920349, | |
| "learning_rate": 2.2184562965331203e-06, | |
| "loss": 0.3405, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 1.3607784431137724, | |
| "grad_norm": 0.17500266432762146, | |
| "learning_rate": 2.2086202696995248e-06, | |
| "loss": 0.1802, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 1.3622754491017965, | |
| "grad_norm": 0.2124902904033661, | |
| "learning_rate": 2.1987999115099763e-06, | |
| "loss": 0.2348, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.3637724550898205, | |
| "grad_norm": 0.21619191765785217, | |
| "learning_rate": 2.1889952770883644e-06, | |
| "loss": 0.2918, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.3652694610778444, | |
| "grad_norm": 0.2912244498729706, | |
| "learning_rate": 2.17920642147031e-06, | |
| "loss": 0.2907, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 1.3667664670658684, | |
| "grad_norm": 0.3516407608985901, | |
| "learning_rate": 2.169433399602872e-06, | |
| "loss": 0.3611, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 1.3682634730538923, | |
| "grad_norm": 0.23837457597255707, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.3391, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 1.3697604790419162, | |
| "grad_norm": 0.3627464175224304, | |
| "learning_rate": 2.1499350764633513e-06, | |
| "loss": 0.309, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 1.3712574850299402, | |
| "grad_norm": 0.23673026263713837, | |
| "learning_rate": 2.140209884639759e-06, | |
| "loss": 0.2918, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.372754491017964, | |
| "grad_norm": 0.27986136078834534, | |
| "learning_rate": 2.130500745463136e-06, | |
| "loss": 0.3377, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.374251497005988, | |
| "grad_norm": 0.2681880295276642, | |
| "learning_rate": 2.120807713433074e-06, | |
| "loss": 0.2247, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 1.375748502994012, | |
| "grad_norm": 0.21259628236293793, | |
| "learning_rate": 2.1111308429587446e-06, | |
| "loss": 0.2602, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 1.377245508982036, | |
| "grad_norm": 0.2214302122592926, | |
| "learning_rate": 2.1014701883586087e-06, | |
| "loss": 0.2005, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 1.3787425149700598, | |
| "grad_norm": 0.17799615859985352, | |
| "learning_rate": 2.091825803860095e-06, | |
| "loss": 0.2805, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.3802395209580838, | |
| "grad_norm": 0.23212796449661255, | |
| "learning_rate": 2.082197743599314e-06, | |
| "loss": 0.2076, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 1.3817365269461077, | |
| "grad_norm": 0.28191059827804565, | |
| "learning_rate": 2.072586061620735e-06, | |
| "loss": 0.2368, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 1.3832335329341316, | |
| "grad_norm": 0.2307487428188324, | |
| "learning_rate": 2.0629908118769004e-06, | |
| "loss": 0.3296, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 1.3847305389221556, | |
| "grad_norm": 0.17868001759052277, | |
| "learning_rate": 2.0534120482281087e-06, | |
| "loss": 0.1473, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 1.3862275449101795, | |
| "grad_norm": 0.24826651811599731, | |
| "learning_rate": 2.043849824442124e-06, | |
| "loss": 0.296, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.3877245508982035, | |
| "grad_norm": 0.22485694289207458, | |
| "learning_rate": 2.034304194193861e-06, | |
| "loss": 0.2683, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 1.3892215568862276, | |
| "grad_norm": 0.2466598004102707, | |
| "learning_rate": 2.024775211065098e-06, | |
| "loss": 0.3431, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 1.3907185628742516, | |
| "grad_norm": 0.23691681027412415, | |
| "learning_rate": 2.0152629285441668e-06, | |
| "loss": 0.3196, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 1.3922155688622755, | |
| "grad_norm": 0.22169752418994904, | |
| "learning_rate": 2.0057674000256556e-06, | |
| "loss": 0.2795, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 1.3937125748502994, | |
| "grad_norm": 0.21425652503967285, | |
| "learning_rate": 1.996288678810105e-06, | |
| "loss": 0.2318, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.3952095808383234, | |
| "grad_norm": 0.346021831035614, | |
| "learning_rate": 1.9868268181037186e-06, | |
| "loss": 0.3006, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 1.3967065868263473, | |
| "grad_norm": 0.29640334844589233, | |
| "learning_rate": 1.9773818710180514e-06, | |
| "loss": 0.4372, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 1.3982035928143712, | |
| "grad_norm": 0.3824538588523865, | |
| "learning_rate": 1.967953890569723e-06, | |
| "loss": 0.4661, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 1.3997005988023952, | |
| "grad_norm": 0.27732986211776733, | |
| "learning_rate": 1.958542929680117e-06, | |
| "loss": 0.221, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 1.401197604790419, | |
| "grad_norm": 0.35698726773262024, | |
| "learning_rate": 1.9491490411750745e-06, | |
| "loss": 0.3821, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.402694610778443, | |
| "grad_norm": 0.3157179653644562, | |
| "learning_rate": 1.9397722777846153e-06, | |
| "loss": 0.3621, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 1.4041916167664672, | |
| "grad_norm": 0.23464487493038177, | |
| "learning_rate": 1.9304126921426235e-06, | |
| "loss": 0.3528, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.4056886227544911, | |
| "grad_norm": 0.26018989086151123, | |
| "learning_rate": 1.921070336786568e-06, | |
| "loss": 0.2786, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 1.407185628742515, | |
| "grad_norm": 0.2732929289340973, | |
| "learning_rate": 1.9117452641571934e-06, | |
| "loss": 0.328, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 1.408682634730539, | |
| "grad_norm": 0.3107425570487976, | |
| "learning_rate": 1.9024375265982386e-06, | |
| "loss": 0.2317, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.410179640718563, | |
| "grad_norm": 0.2310229390859604, | |
| "learning_rate": 1.893147176356131e-06, | |
| "loss": 0.3121, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 1.4116766467065869, | |
| "grad_norm": 0.24809664487838745, | |
| "learning_rate": 1.8838742655797053e-06, | |
| "loss": 0.3083, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 1.4131736526946108, | |
| "grad_norm": 0.36014804244041443, | |
| "learning_rate": 1.8746188463198983e-06, | |
| "loss": 0.2166, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 1.4146706586826348, | |
| "grad_norm": 0.20624969899654388, | |
| "learning_rate": 1.865380970529469e-06, | |
| "loss": 0.2533, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 1.4161676646706587, | |
| "grad_norm": 0.2667752504348755, | |
| "learning_rate": 1.8561606900626938e-06, | |
| "loss": 0.306, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4176646706586826, | |
| "grad_norm": 0.27965497970581055, | |
| "learning_rate": 1.8469580566750911e-06, | |
| "loss": 0.2644, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 1.4191616766467066, | |
| "grad_norm": 0.2576799690723419, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.2376, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 1.4206586826347305, | |
| "grad_norm": 0.383226215839386, | |
| "learning_rate": 1.8286059376638748e-06, | |
| "loss": 0.3097, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 1.4221556886227544, | |
| "grad_norm": 0.26679062843322754, | |
| "learning_rate": 1.8194565550548477e-06, | |
| "loss": 0.2423, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 1.4236526946107784, | |
| "grad_norm": 0.321591854095459, | |
| "learning_rate": 1.810325025553578e-06, | |
| "loss": 0.3952, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.4251497005988023, | |
| "grad_norm": 0.26733914017677307, | |
| "learning_rate": 1.8012114004174048e-06, | |
| "loss": 0.3133, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.4266467065868262, | |
| "grad_norm": 0.2287786900997162, | |
| "learning_rate": 1.7921157308031567e-06, | |
| "loss": 0.2384, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 1.4281437125748502, | |
| "grad_norm": 0.2414834201335907, | |
| "learning_rate": 1.7830380677668836e-06, | |
| "loss": 0.2552, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.4296407185628741, | |
| "grad_norm": 0.32560208439826965, | |
| "learning_rate": 1.7739784622635514e-06, | |
| "loss": 0.3029, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 1.4311377245508983, | |
| "grad_norm": 0.344568133354187, | |
| "learning_rate": 1.764936965146773e-06, | |
| "loss": 0.3252, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.4326347305389222, | |
| "grad_norm": 0.3280235528945923, | |
| "learning_rate": 1.7559136271685079e-06, | |
| "loss": 0.2814, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 1.4341317365269461, | |
| "grad_norm": 0.3319287896156311, | |
| "learning_rate": 1.746908498978791e-06, | |
| "loss": 0.2855, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 1.43562874251497, | |
| "grad_norm": 0.26920896768569946, | |
| "learning_rate": 1.7379216311254339e-06, | |
| "loss": 0.2206, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 1.437125748502994, | |
| "grad_norm": 0.2052461802959442, | |
| "learning_rate": 1.7289530740537569e-06, | |
| "loss": 0.2528, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 1.438622754491018, | |
| "grad_norm": 0.33078446984291077, | |
| "learning_rate": 1.72000287810629e-06, | |
| "loss": 0.2986, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.4401197604790419, | |
| "grad_norm": 0.3203030228614807, | |
| "learning_rate": 1.7110710935225055e-06, | |
| "loss": 0.2862, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.4416167664670658, | |
| "grad_norm": 0.25592538714408875, | |
| "learning_rate": 1.7021577704385218e-06, | |
| "loss": 0.2645, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 1.4431137724550898, | |
| "grad_norm": 0.26727235317230225, | |
| "learning_rate": 1.6932629588868332e-06, | |
| "loss": 0.2634, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 1.4446107784431137, | |
| "grad_norm": 0.19937850534915924, | |
| "learning_rate": 1.6843867087960252e-06, | |
| "loss": 0.2024, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 1.4461077844311379, | |
| "grad_norm": 0.21263986825942993, | |
| "learning_rate": 1.6755290699904881e-06, | |
| "loss": 0.2111, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.4476047904191618, | |
| "grad_norm": 0.20202231407165527, | |
| "learning_rate": 1.6666900921901497e-06, | |
| "loss": 0.2371, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 1.4491017964071857, | |
| "grad_norm": 0.43511268496513367, | |
| "learning_rate": 1.6578698250101828e-06, | |
| "loss": 0.3341, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 1.4505988023952097, | |
| "grad_norm": 0.3771866261959076, | |
| "learning_rate": 1.6490683179607403e-06, | |
| "loss": 0.5046, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 1.4520958083832336, | |
| "grad_norm": 0.25945380330085754, | |
| "learning_rate": 1.6402856204466611e-06, | |
| "loss": 0.2812, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.4535928143712575, | |
| "grad_norm": 0.24568547308444977, | |
| "learning_rate": 1.6315217817672142e-06, | |
| "loss": 0.1704, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.4550898203592815, | |
| "grad_norm": 0.26437002420425415, | |
| "learning_rate": 1.6227768511157976e-06, | |
| "loss": 0.1989, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 1.4565868263473054, | |
| "grad_norm": 0.3020990788936615, | |
| "learning_rate": 1.6140508775796832e-06, | |
| "loss": 0.2377, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.4580838323353293, | |
| "grad_norm": 0.15546730160713196, | |
| "learning_rate": 1.6053439101397257e-06, | |
| "loss": 0.1476, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.4595808383233533, | |
| "grad_norm": 0.22115832567214966, | |
| "learning_rate": 1.5966559976701e-06, | |
| "loss": 0.3021, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 1.4610778443113772, | |
| "grad_norm": 0.1683429330587387, | |
| "learning_rate": 1.5879871889380155e-06, | |
| "loss": 0.1581, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.4625748502994012, | |
| "grad_norm": 0.2293458878993988, | |
| "learning_rate": 1.5793375326034539e-06, | |
| "loss": 0.2103, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 1.464071856287425, | |
| "grad_norm": 0.2591915428638458, | |
| "learning_rate": 1.5707070772188843e-06, | |
| "loss": 0.2493, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 1.465568862275449, | |
| "grad_norm": 0.3226713538169861, | |
| "learning_rate": 1.5620958712290023e-06, | |
| "loss": 0.3839, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 1.467065868263473, | |
| "grad_norm": 0.3980361223220825, | |
| "learning_rate": 1.5535039629704467e-06, | |
| "loss": 0.2999, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 1.468562874251497, | |
| "grad_norm": 0.2781156003475189, | |
| "learning_rate": 1.5449314006715394e-06, | |
| "loss": 0.2316, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.4700598802395208, | |
| "grad_norm": 0.332856684923172, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.3408, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 1.471556886227545, | |
| "grad_norm": 0.28347525000572205, | |
| "learning_rate": 1.5278445063227038e-06, | |
| "loss": 0.3741, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 1.473053892215569, | |
| "grad_norm": 0.2651091516017914, | |
| "learning_rate": 1.5193302701853674e-06, | |
| "loss": 0.3159, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 1.4745508982035929, | |
| "grad_norm": 0.2445843368768692, | |
| "learning_rate": 1.5108355718323236e-06, | |
| "loss": 0.2487, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 1.4760479041916168, | |
| "grad_norm": 0.31026923656463623, | |
| "learning_rate": 1.502360458946232e-06, | |
| "loss": 0.3436, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.4775449101796407, | |
| "grad_norm": 0.4750896990299225, | |
| "learning_rate": 1.4939049790998095e-06, | |
| "loss": 0.3669, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 1.4790419161676647, | |
| "grad_norm": 0.2514893114566803, | |
| "learning_rate": 1.4854691797555753e-06, | |
| "loss": 0.2849, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 1.4790419161676647, | |
| "eval_loss": 0.27240613102912903, | |
| "eval_runtime": 96.0291, | |
| "eval_samples_per_second": 7.664, | |
| "eval_steps_per_second": 0.958, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 1.4805389221556886, | |
| "grad_norm": 0.23577263951301575, | |
| "learning_rate": 1.4770531082655704e-06, | |
| "loss": 0.2443, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 1.4820359281437125, | |
| "grad_norm": 0.29413002729415894, | |
| "learning_rate": 1.4686568118711054e-06, | |
| "loss": 0.3396, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 1.4835329341317365, | |
| "grad_norm": 0.2968040704727173, | |
| "learning_rate": 1.4602803377024833e-06, | |
| "loss": 0.2327, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.4850299401197604, | |
| "grad_norm": 0.29157525300979614, | |
| "learning_rate": 1.451923732778745e-06, | |
| "loss": 0.2629, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 1.4865269461077844, | |
| "grad_norm": 0.2867380976676941, | |
| "learning_rate": 1.4435870440073968e-06, | |
| "loss": 0.4332, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 1.4880239520958085, | |
| "grad_norm": 0.22759690880775452, | |
| "learning_rate": 1.435270318184156e-06, | |
| "loss": 0.2434, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 1.4895209580838324, | |
| "grad_norm": 0.22279328107833862, | |
| "learning_rate": 1.4269736019926778e-06, | |
| "loss": 0.2133, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 1.4910179640718564, | |
| "grad_norm": 0.21220585703849792, | |
| "learning_rate": 1.418696942004304e-06, | |
| "loss": 0.1834, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.4925149700598803, | |
| "grad_norm": 0.22208869457244873, | |
| "learning_rate": 1.410440384677791e-06, | |
| "loss": 0.3176, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 1.4940119760479043, | |
| "grad_norm": 0.25137487053871155, | |
| "learning_rate": 1.4022039763590595e-06, | |
| "loss": 0.303, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 1.4955089820359282, | |
| "grad_norm": 0.3287590444087982, | |
| "learning_rate": 1.3939877632809279e-06, | |
| "loss": 0.2652, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 1.4970059880239521, | |
| "grad_norm": 0.22760730981826782, | |
| "learning_rate": 1.3857917915628516e-06, | |
| "loss": 0.2777, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 1.498502994011976, | |
| "grad_norm": 0.20680049061775208, | |
| "learning_rate": 1.3776161072106703e-06, | |
| "loss": 0.2284, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.4575501084327698, | |
| "learning_rate": 1.369460756116342e-06, | |
| "loss": 0.31, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 1.501497005988024, | |
| "grad_norm": 0.26314109563827515, | |
| "learning_rate": 1.3613257840576954e-06, | |
| "loss": 0.2972, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 1.5029940119760479, | |
| "grad_norm": 0.25917747616767883, | |
| "learning_rate": 1.3532112366981598e-06, | |
| "loss": 0.3022, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 1.5044910179640718, | |
| "grad_norm": 0.22802290320396423, | |
| "learning_rate": 1.3451171595865226e-06, | |
| "loss": 0.2922, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 1.5059880239520957, | |
| "grad_norm": 0.2178899347782135, | |
| "learning_rate": 1.3370435981566622e-06, | |
| "loss": 0.2445, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5074850299401197, | |
| "grad_norm": 0.39115995168685913, | |
| "learning_rate": 1.3289905977273027e-06, | |
| "loss": 0.3144, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 1.5089820359281436, | |
| "grad_norm": 0.23431116342544556, | |
| "learning_rate": 1.3209582035017487e-06, | |
| "loss": 0.2239, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 1.5104790419161676, | |
| "grad_norm": 0.2912103235721588, | |
| "learning_rate": 1.312946460567644e-06, | |
| "loss": 0.3606, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 1.5119760479041915, | |
| "grad_norm": 0.20403842628002167, | |
| "learning_rate": 1.3049554138967052e-06, | |
| "loss": 0.1988, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 1.5134730538922154, | |
| "grad_norm": 0.1640235036611557, | |
| "learning_rate": 1.2969851083444834e-06, | |
| "loss": 0.1336, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.5149700598802394, | |
| "grad_norm": 0.22775056958198547, | |
| "learning_rate": 1.2890355886500971e-06, | |
| "loss": 0.1849, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 1.5164670658682635, | |
| "grad_norm": 0.315218061208725, | |
| "learning_rate": 1.2811068994359992e-06, | |
| "loss": 0.2266, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 1.5179640718562875, | |
| "grad_norm": 0.361473023891449, | |
| "learning_rate": 1.273199085207706e-06, | |
| "loss": 0.3429, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 1.5194610778443114, | |
| "grad_norm": 0.22169454395771027, | |
| "learning_rate": 1.2653121903535653e-06, | |
| "loss": 0.1868, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 1.5209580838323353, | |
| "grad_norm": 0.20558999478816986, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.2067, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.5224550898203593, | |
| "grad_norm": 0.37300822138786316, | |
| "learning_rate": 1.2496013357337416e-06, | |
| "loss": 0.2873, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 1.5239520958083832, | |
| "grad_norm": 0.24679654836654663, | |
| "learning_rate": 1.2417774641566298e-06, | |
| "loss": 0.2406, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 1.5254491017964071, | |
| "grad_norm": 0.25380274653434753, | |
| "learning_rate": 1.233974688330315e-06, | |
| "loss": 0.284, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 1.5269461077844313, | |
| "grad_norm": 0.24675562977790833, | |
| "learning_rate": 1.2261930520535403e-06, | |
| "loss": 0.2421, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 1.5284431137724552, | |
| "grad_norm": 0.23980629444122314, | |
| "learning_rate": 1.2184325990063822e-06, | |
| "loss": 0.2885, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.5299401197604792, | |
| "grad_norm": 0.24391433596611023, | |
| "learning_rate": 1.210693372750017e-06, | |
| "loss": 0.1621, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 1.531437125748503, | |
| "grad_norm": 0.2686309218406677, | |
| "learning_rate": 1.202975416726464e-06, | |
| "loss": 0.2907, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 1.532934131736527, | |
| "grad_norm": 0.45841044187545776, | |
| "learning_rate": 1.1952787742583549e-06, | |
| "loss": 0.2877, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 1.534431137724551, | |
| "grad_norm": 0.1831907480955124, | |
| "learning_rate": 1.1876034885486764e-06, | |
| "loss": 0.2025, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 1.535928143712575, | |
| "grad_norm": 0.28029313683509827, | |
| "learning_rate": 1.1799496026805413e-06, | |
| "loss": 0.2588, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.5374251497005988, | |
| "grad_norm": 0.2655869424343109, | |
| "learning_rate": 1.1723171596169353e-06, | |
| "loss": 0.2087, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 1.5389221556886228, | |
| "grad_norm": 0.3924802243709564, | |
| "learning_rate": 1.1647062022004845e-06, | |
| "loss": 0.2952, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 1.5404191616766467, | |
| "grad_norm": 0.3049009144306183, | |
| "learning_rate": 1.157116773153208e-06, | |
| "loss": 0.2123, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 1.5419161676646707, | |
| "grad_norm": 0.23748943209648132, | |
| "learning_rate": 1.1495489150762851e-06, | |
| "loss": 0.2867, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 1.5434131736526946, | |
| "grad_norm": 0.18197742104530334, | |
| "learning_rate": 1.1420026704498077e-06, | |
| "loss": 0.2095, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.5449101796407185, | |
| "grad_norm": 0.2282951921224594, | |
| "learning_rate": 1.1344780816325512e-06, | |
| "loss": 0.1673, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 1.5464071856287425, | |
| "grad_norm": 0.26313188672065735, | |
| "learning_rate": 1.1269751908617277e-06, | |
| "loss": 0.3961, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 1.5479041916167664, | |
| "grad_norm": 0.2284393161535263, | |
| "learning_rate": 1.1194940402527566e-06, | |
| "loss": 0.1948, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 1.5494011976047903, | |
| "grad_norm": 0.28718557953834534, | |
| "learning_rate": 1.112034671799025e-06, | |
| "loss": 0.3257, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 1.5508982035928143, | |
| "grad_norm": 0.23936651647090912, | |
| "learning_rate": 1.1045971273716476e-06, | |
| "loss": 0.1952, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.5523952095808382, | |
| "grad_norm": 0.3323695957660675, | |
| "learning_rate": 1.0971814487192429e-06, | |
| "loss": 0.3468, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 1.5538922155688621, | |
| "grad_norm": 0.26825106143951416, | |
| "learning_rate": 1.089787677467683e-06, | |
| "loss": 0.2717, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 1.555389221556886, | |
| "grad_norm": 0.34908008575439453, | |
| "learning_rate": 1.0824158551198783e-06, | |
| "loss": 0.3218, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 1.55688622754491, | |
| "grad_norm": 0.27587297558784485, | |
| "learning_rate": 1.075066023055527e-06, | |
| "loss": 0.2816, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 1.5583832335329342, | |
| "grad_norm": 0.2504119277000427, | |
| "learning_rate": 1.0677382225308969e-06, | |
| "loss": 0.3119, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.5598802395209581, | |
| "grad_norm": 0.3709030747413635, | |
| "learning_rate": 1.0604324946785826e-06, | |
| "loss": 0.2381, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 1.561377245508982, | |
| "grad_norm": 0.27703753113746643, | |
| "learning_rate": 1.0531488805072848e-06, | |
| "loss": 0.1919, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 1.562874251497006, | |
| "grad_norm": 0.27848151326179504, | |
| "learning_rate": 1.0458874209015708e-06, | |
| "loss": 0.2071, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 1.56437125748503, | |
| "grad_norm": 0.2879417836666107, | |
| "learning_rate": 1.0386481566216532e-06, | |
| "loss": 0.3086, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 1.5658682634730539, | |
| "grad_norm": 0.21330009400844574, | |
| "learning_rate": 1.0314311283031531e-06, | |
| "loss": 0.1942, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.5673652694610778, | |
| "grad_norm": 0.27636274695396423, | |
| "learning_rate": 1.0242363764568808e-06, | |
| "loss": 0.2228, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 1.568862275449102, | |
| "grad_norm": 0.5699312686920166, | |
| "learning_rate": 1.0170639414685985e-06, | |
| "loss": 0.3452, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 1.5703592814371259, | |
| "grad_norm": 0.27370771765708923, | |
| "learning_rate": 1.0099138635988026e-06, | |
| "loss": 0.2548, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 1.5718562874251498, | |
| "grad_norm": 0.24189740419387817, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.2262, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 1.5733532934131738, | |
| "grad_norm": 0.2238190919160843, | |
| "learning_rate": 9.956809396289519e-07, | |
| "loss": 0.2553, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.5748502994011977, | |
| "grad_norm": 0.2638553977012634, | |
| "learning_rate": 9.885981734215094e-07, | |
| "loss": 0.3157, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 1.5763473053892216, | |
| "grad_norm": 0.22409968078136444, | |
| "learning_rate": 9.815379241173295e-07, | |
| "loss": 0.3396, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 1.5778443113772456, | |
| "grad_norm": 0.25611743330955505, | |
| "learning_rate": 9.745002313471847e-07, | |
| "loss": 0.2865, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 1.5793413173652695, | |
| "grad_norm": 0.2451084852218628, | |
| "learning_rate": 9.67485134615232e-07, | |
| "loss": 0.292, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 1.5808383233532934, | |
| "grad_norm": 0.26699185371398926, | |
| "learning_rate": 9.60492673298794e-07, | |
| "loss": 0.297, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.5823353293413174, | |
| "grad_norm": 0.23169200122356415, | |
| "learning_rate": 9.535228866481295e-07, | |
| "loss": 0.2808, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 1.5838323353293413, | |
| "grad_norm": 0.37093454599380493, | |
| "learning_rate": 9.465758137862264e-07, | |
| "loss": 0.2549, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 1.5853293413173652, | |
| "grad_norm": 0.4046109914779663, | |
| "learning_rate": 9.396514937085682e-07, | |
| "loss": 0.2644, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 1.5868263473053892, | |
| "grad_norm": 0.24264554679393768, | |
| "learning_rate": 9.327499652829292e-07, | |
| "loss": 0.1966, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 1.5883233532934131, | |
| "grad_norm": 0.35896986722946167, | |
| "learning_rate": 9.258712672491416e-07, | |
| "loss": 0.4137, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.589820359281437, | |
| "grad_norm": 0.3543803095817566, | |
| "learning_rate": 9.190154382188921e-07, | |
| "loss": 0.2922, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 1.591317365269461, | |
| "grad_norm": 0.18955141305923462, | |
| "learning_rate": 9.121825166754927e-07, | |
| "loss": 0.1885, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 1.592814371257485, | |
| "grad_norm": 0.2830042839050293, | |
| "learning_rate": 9.053725409736752e-07, | |
| "loss": 0.3324, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 1.5943113772455089, | |
| "grad_norm": 0.2516660988330841, | |
| "learning_rate": 8.98585549339368e-07, | |
| "loss": 0.3912, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 1.5958083832335328, | |
| "grad_norm": 0.25523948669433594, | |
| "learning_rate": 8.918215798694879e-07, | |
| "loss": 0.2574, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.5973053892215567, | |
| "grad_norm": 0.3023395538330078, | |
| "learning_rate": 8.850806705317183e-07, | |
| "loss": 0.2945, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 1.5988023952095807, | |
| "grad_norm": 0.2778310775756836, | |
| "learning_rate": 8.783628591643056e-07, | |
| "loss": 0.3833, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 1.6002994011976048, | |
| "grad_norm": 0.27976202964782715, | |
| "learning_rate": 8.716681834758411e-07, | |
| "loss": 0.2099, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 1.6017964071856288, | |
| "grad_norm": 0.33491119742393494, | |
| "learning_rate": 8.649966810450472e-07, | |
| "loss": 0.3326, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 1.6032934131736527, | |
| "grad_norm": 0.24264518916606903, | |
| "learning_rate": 8.583483893205746e-07, | |
| "loss": 0.2515, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.6047904191616766, | |
| "grad_norm": 0.24761344492435455, | |
| "learning_rate": 8.517233456207819e-07, | |
| "loss": 0.1778, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 1.6062874251497006, | |
| "grad_norm": 0.30928072333335876, | |
| "learning_rate": 8.451215871335355e-07, | |
| "loss": 0.3494, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 1.6077844311377245, | |
| "grad_norm": 0.2120644450187683, | |
| "learning_rate": 8.38543150915993e-07, | |
| "loss": 0.1977, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 1.6092814371257484, | |
| "grad_norm": 0.36671245098114014, | |
| "learning_rate": 8.31988073894403e-07, | |
| "loss": 0.2415, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 1.6107784431137726, | |
| "grad_norm": 0.2658866047859192, | |
| "learning_rate": 8.254563928638892e-07, | |
| "loss": 0.2486, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6122754491017965, | |
| "grad_norm": 0.21688513457775116, | |
| "learning_rate": 8.189481444882524e-07, | |
| "loss": 0.243, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 1.6137724550898205, | |
| "grad_norm": 0.24383318424224854, | |
| "learning_rate": 8.124633652997571e-07, | |
| "loss": 0.2836, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 1.6152694610778444, | |
| "grad_norm": 0.21168118715286255, | |
| "learning_rate": 8.060020916989331e-07, | |
| "loss": 0.2159, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 1.6167664670658684, | |
| "grad_norm": 0.235800638794899, | |
| "learning_rate": 7.995643599543645e-07, | |
| "loss": 0.3152, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 1.6182634730538923, | |
| "grad_norm": 0.20235656201839447, | |
| "learning_rate": 7.931502062024949e-07, | |
| "loss": 0.2067, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.6197604790419162, | |
| "grad_norm": 0.21362897753715515, | |
| "learning_rate": 7.86759666447412e-07, | |
| "loss": 0.2366, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 1.6212574850299402, | |
| "grad_norm": 0.297219455242157, | |
| "learning_rate": 7.803927765606595e-07, | |
| "loss": 0.2066, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 1.622754491017964, | |
| "grad_norm": 0.1866680532693863, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.1817, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 1.624251497005988, | |
| "grad_norm": 0.25540104508399963, | |
| "learning_rate": 7.677300892143485e-07, | |
| "loss": 0.3215, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 1.625748502994012, | |
| "grad_norm": 0.2844352424144745, | |
| "learning_rate": 7.614343628333104e-07, | |
| "loss": 0.3789, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.627245508982036, | |
| "grad_norm": 0.32128477096557617, | |
| "learning_rate": 7.55162428477243e-07, | |
| "loss": 0.3061, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 1.6287425149700598, | |
| "grad_norm": 0.19842462241649628, | |
| "learning_rate": 7.489143213519301e-07, | |
| "loss": 0.2446, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 1.6302395209580838, | |
| "grad_norm": 0.21685732901096344, | |
| "learning_rate": 7.426900765294043e-07, | |
| "loss": 0.1991, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 1.6317365269461077, | |
| "grad_norm": 0.3464147746562958, | |
| "learning_rate": 7.364897289477585e-07, | |
| "loss": 0.2764, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 1.6332335329341316, | |
| "grad_norm": 0.30749276280403137, | |
| "learning_rate": 7.303133134109391e-07, | |
| "loss": 0.1874, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.6347305389221556, | |
| "grad_norm": 0.310250461101532, | |
| "learning_rate": 7.241608645885629e-07, | |
| "loss": 0.3505, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 1.6362275449101795, | |
| "grad_norm": 0.3631021976470947, | |
| "learning_rate": 7.180324170157094e-07, | |
| "loss": 0.4384, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 1.6377245508982035, | |
| "grad_norm": 0.21250741183757782, | |
| "learning_rate": 7.119280050927407e-07, | |
| "loss": 0.213, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 1.6392215568862274, | |
| "grad_norm": 0.2978675365447998, | |
| "learning_rate": 7.058476630850935e-07, | |
| "loss": 0.2757, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 1.6407185628742516, | |
| "grad_norm": 0.2608672082424164, | |
| "learning_rate": 6.997914251231036e-07, | |
| "loss": 0.2849, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.6422155688622755, | |
| "grad_norm": 0.3758143484592438, | |
| "learning_rate": 6.937593252017983e-07, | |
| "loss": 0.2907, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 1.6437125748502994, | |
| "grad_norm": 0.2593882083892822, | |
| "learning_rate": 6.87751397180716e-07, | |
| "loss": 0.2183, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 1.6452095808383234, | |
| "grad_norm": 0.46381810307502747, | |
| "learning_rate": 6.817676747837104e-07, | |
| "loss": 0.4085, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 1.6467065868263473, | |
| "grad_norm": 0.32167354226112366, | |
| "learning_rate": 6.758081915987669e-07, | |
| "loss": 0.1879, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 1.6482035928143712, | |
| "grad_norm": 0.4099140167236328, | |
| "learning_rate": 6.698729810778065e-07, | |
| "loss": 0.3501, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.6497005988023952, | |
| "grad_norm": 0.234484001994133, | |
| "learning_rate": 6.639620765365074e-07, | |
| "loss": 0.1925, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 1.6511976047904193, | |
| "grad_norm": 0.22274383902549744, | |
| "learning_rate": 6.580755111541076e-07, | |
| "loss": 0.2334, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 1.6526946107784433, | |
| "grad_norm": 0.27316105365753174, | |
| "learning_rate": 6.522133179732271e-07, | |
| "loss": 0.2915, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 1.6541916167664672, | |
| "grad_norm": 0.2743278443813324, | |
| "learning_rate": 6.463755298996799e-07, | |
| "loss": 0.3385, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 1.6556886227544911, | |
| "grad_norm": 0.24365045130252838, | |
| "learning_rate": 6.405621797022848e-07, | |
| "loss": 0.2271, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.657185628742515, | |
| "grad_norm": 0.2597195506095886, | |
| "learning_rate": 6.347733000126899e-07, | |
| "loss": 0.2802, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 1.658682634730539, | |
| "grad_norm": 0.3242190182209015, | |
| "learning_rate": 6.290089233251811e-07, | |
| "loss": 0.2547, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 1.660179640718563, | |
| "grad_norm": 0.23406442999839783, | |
| "learning_rate": 6.232690819965065e-07, | |
| "loss": 0.2242, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 1.6616766467065869, | |
| "grad_norm": 0.32079535722732544, | |
| "learning_rate": 6.175538082456883e-07, | |
| "loss": 0.3538, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 1.6631736526946108, | |
| "grad_norm": 0.32577863335609436, | |
| "learning_rate": 6.118631341538489e-07, | |
| "loss": 0.3625, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.6646706586826348, | |
| "grad_norm": 0.3936420977115631, | |
| "learning_rate": 6.061970916640236e-07, | |
| "loss": 0.2731, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 1.6661676646706587, | |
| "grad_norm": 0.34293246269226074, | |
| "learning_rate": 6.005557125809896e-07, | |
| "loss": 0.3033, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 1.6676646706586826, | |
| "grad_norm": 0.20506998896598816, | |
| "learning_rate": 5.949390285710777e-07, | |
| "loss": 0.294, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 1.6691616766467066, | |
| "grad_norm": 0.17009401321411133, | |
| "learning_rate": 5.893470711620036e-07, | |
| "loss": 0.1798, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 1.6706586826347305, | |
| "grad_norm": 0.2312851995229721, | |
| "learning_rate": 5.837798717426846e-07, | |
| "loss": 0.2047, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.6721556886227544, | |
| "grad_norm": 0.16008791327476501, | |
| "learning_rate": 5.782374615630682e-07, | |
| "loss": 0.142, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 1.6736526946107784, | |
| "grad_norm": 0.20815108716487885, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.2526, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 1.6751497005988023, | |
| "grad_norm": 0.22949810326099396, | |
| "learning_rate": 5.672271332268098e-07, | |
| "loss": 0.2564, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 1.6766467065868262, | |
| "grad_norm": 0.3489370048046112, | |
| "learning_rate": 5.617592768736269e-07, | |
| "loss": 0.2188, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 1.6781437125748502, | |
| "grad_norm": 0.2736472487449646, | |
| "learning_rate": 5.563163333667098e-07, | |
| "loss": 0.308, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 1.6796407185628741, | |
| "grad_norm": 0.2761566936969757, | |
| "learning_rate": 5.508983332585316e-07, | |
| "loss": 0.1934, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 1.681137724550898, | |
| "grad_norm": 0.21854543685913086, | |
| "learning_rate": 5.455053069615456e-07, | |
| "loss": 0.2165, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 1.6826347305389222, | |
| "grad_norm": 0.32581812143325806, | |
| "learning_rate": 5.401372847480285e-07, | |
| "loss": 0.3124, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 1.6841317365269461, | |
| "grad_norm": 0.3002076745033264, | |
| "learning_rate": 5.347942967498965e-07, | |
| "loss": 0.36, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 1.68562874251497, | |
| "grad_norm": 0.22961673140525818, | |
| "learning_rate": 5.294763729585484e-07, | |
| "loss": 0.2958, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.687125748502994, | |
| "grad_norm": 0.2980080842971802, | |
| "learning_rate": 5.241835432246888e-07, | |
| "loss": 0.183, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 1.688622754491018, | |
| "grad_norm": 0.34977275133132935, | |
| "learning_rate": 5.18915837258166e-07, | |
| "loss": 0.3158, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 1.6901197604790419, | |
| "grad_norm": 0.2116621881723404, | |
| "learning_rate": 5.136732846278003e-07, | |
| "loss": 0.3113, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 1.6916167664670658, | |
| "grad_norm": 0.3168148398399353, | |
| "learning_rate": 5.084559147612244e-07, | |
| "loss": 0.3469, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 1.69311377245509, | |
| "grad_norm": 0.3314928412437439, | |
| "learning_rate": 5.032637569447091e-07, | |
| "loss": 0.3255, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 1.694610778443114, | |
| "grad_norm": 0.2442367523908615, | |
| "learning_rate": 4.980968403230097e-07, | |
| "loss": 0.2865, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 1.6961077844311379, | |
| "grad_norm": 0.23955686390399933, | |
| "learning_rate": 4.929551938991945e-07, | |
| "loss": 0.3061, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 1.6976047904191618, | |
| "grad_norm": 0.20683260262012482, | |
| "learning_rate": 4.87838846534483e-07, | |
| "loss": 0.1759, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 1.6991017964071857, | |
| "grad_norm": 0.27983346581459045, | |
| "learning_rate": 4.827478269480895e-07, | |
| "loss": 0.2992, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 1.7005988023952097, | |
| "grad_norm": 0.33870527148246765, | |
| "learning_rate": 4.776821637170525e-07, | |
| "loss": 0.3379, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7020958083832336, | |
| "grad_norm": 0.2928355038166046, | |
| "learning_rate": 4.726418852760839e-07, | |
| "loss": 0.3088, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 1.7035928143712575, | |
| "grad_norm": 0.3086561858654022, | |
| "learning_rate": 4.6762701991740434e-07, | |
| "loss": 0.394, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 1.7050898203592815, | |
| "grad_norm": 0.2599639594554901, | |
| "learning_rate": 4.626375957905821e-07, | |
| "loss": 0.2423, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 1.7065868263473054, | |
| "grad_norm": 0.24536360800266266, | |
| "learning_rate": 4.576736409023813e-07, | |
| "loss": 0.2244, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 1.7080838323353293, | |
| "grad_norm": 0.27678626775741577, | |
| "learning_rate": 4.5273518311660103e-07, | |
| "loss": 0.2654, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 1.7095808383233533, | |
| "grad_norm": 0.25134310126304626, | |
| "learning_rate": 4.4782225015391754e-07, | |
| "loss": 0.2651, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 1.7110778443113772, | |
| "grad_norm": 0.2866244316101074, | |
| "learning_rate": 4.429348695917329e-07, | |
| "loss": 0.362, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 1.7125748502994012, | |
| "grad_norm": 0.3232477605342865, | |
| "learning_rate": 4.3807306886401555e-07, | |
| "loss": 0.3016, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 1.714071856287425, | |
| "grad_norm": 0.21404534578323364, | |
| "learning_rate": 4.3323687526115045e-07, | |
| "loss": 0.238, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 1.715568862275449, | |
| "grad_norm": 0.28791582584381104, | |
| "learning_rate": 4.284263159297819e-07, | |
| "loss": 0.2013, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.717065868263473, | |
| "grad_norm": 0.33721908926963806, | |
| "learning_rate": 4.2364141787266613e-07, | |
| "loss": 0.278, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 1.718562874251497, | |
| "grad_norm": 0.3573349714279175, | |
| "learning_rate": 4.1888220794851386e-07, | |
| "loss": 0.2282, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 1.7200598802395208, | |
| "grad_norm": 0.2762373685836792, | |
| "learning_rate": 4.141487128718452e-07, | |
| "loss": 0.3332, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 1.7215568862275448, | |
| "grad_norm": 0.2202342003583908, | |
| "learning_rate": 4.0944095921283347e-07, | |
| "loss": 0.2471, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 1.7230538922155687, | |
| "grad_norm": 0.49206018447875977, | |
| "learning_rate": 4.0475897339716466e-07, | |
| "loss": 0.2495, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 1.7245508982035929, | |
| "grad_norm": 0.25945720076560974, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.2775, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 1.7260479041916168, | |
| "grad_norm": 0.2269635945558548, | |
| "learning_rate": 3.9547241027523164e-07, | |
| "loss": 0.344, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 1.7275449101796407, | |
| "grad_norm": 0.1602826565504074, | |
| "learning_rate": 3.908678850965425e-07, | |
| "loss": 0.1716, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 1.7290419161676647, | |
| "grad_norm": 0.4009440839290619, | |
| "learning_rate": 3.862892320160483e-07, | |
| "loss": 0.3302, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 1.7290419161676647, | |
| "eval_loss": 0.27044132351875305, | |
| "eval_runtime": 96.2127, | |
| "eval_samples_per_second": 7.65, | |
| "eval_steps_per_second": 0.956, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 1.7305389221556886, | |
| "grad_norm": 0.3490601181983948, | |
| "learning_rate": 3.8173647673476366e-07, | |
| "loss": 0.4149, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.7320359281437125, | |
| "grad_norm": 0.32757532596588135, | |
| "learning_rate": 3.7720964480832847e-07, | |
| "loss": 0.3498, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 1.7335329341317365, | |
| "grad_norm": 0.17421932518482208, | |
| "learning_rate": 3.727087616468739e-07, | |
| "loss": 0.1484, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 1.7350299401197606, | |
| "grad_norm": 0.2285478264093399, | |
| "learning_rate": 3.682338525148699e-07, | |
| "loss": 0.2416, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 1.7365269461077846, | |
| "grad_norm": 0.2338390350341797, | |
| "learning_rate": 3.6378494253099307e-07, | |
| "loss": 0.1934, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 1.7380239520958085, | |
| "grad_norm": 0.34419965744018555, | |
| "learning_rate": 3.5936205666797675e-07, | |
| "loss": 0.3123, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.7395209580838324, | |
| "grad_norm": 0.2365407645702362, | |
| "learning_rate": 3.549652197524783e-07, | |
| "loss": 0.3097, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 1.7410179640718564, | |
| "grad_norm": 0.36558160185813904, | |
| "learning_rate": 3.505944564649344e-07, | |
| "loss": 0.3637, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 1.7425149700598803, | |
| "grad_norm": 0.261536180973053, | |
| "learning_rate": 3.462497913394258e-07, | |
| "loss": 0.2309, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 1.7440119760479043, | |
| "grad_norm": 0.21218152344226837, | |
| "learning_rate": 3.419312487635362e-07, | |
| "loss": 0.2562, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 1.7455089820359282, | |
| "grad_norm": 0.29205673933029175, | |
| "learning_rate": 3.3763885297822153e-07, | |
| "loss": 0.4012, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.7470059880239521, | |
| "grad_norm": 0.24798201024532318, | |
| "learning_rate": 3.333726280776656e-07, | |
| "loss": 0.2164, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 1.748502994011976, | |
| "grad_norm": 0.2931423485279083, | |
| "learning_rate": 3.2913259800915196e-07, | |
| "loss": 0.2651, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.16753946244716644, | |
| "learning_rate": 3.2491878657292643e-07, | |
| "loss": 0.2041, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 1.751497005988024, | |
| "grad_norm": 0.22982138395309448, | |
| "learning_rate": 3.2073121742206117e-07, | |
| "loss": 0.1714, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 1.7529940119760479, | |
| "grad_norm": 0.23351234197616577, | |
| "learning_rate": 3.165699140623285e-07, | |
| "loss": 0.2545, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 1.7544910179640718, | |
| "grad_norm": 0.37126511335372925, | |
| "learning_rate": 3.1243489985206097e-07, | |
| "loss": 0.3697, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 1.7559880239520957, | |
| "grad_norm": 0.23918622732162476, | |
| "learning_rate": 3.0832619800202746e-07, | |
| "loss": 0.2375, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 1.7574850299401197, | |
| "grad_norm": 0.25686293840408325, | |
| "learning_rate": 3.0424383157529716e-07, | |
| "loss": 0.2728, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 1.7589820359281436, | |
| "grad_norm": 0.2703104019165039, | |
| "learning_rate": 3.001878234871147e-07, | |
| "loss": 0.27, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 1.7604790419161676, | |
| "grad_norm": 0.20858930051326752, | |
| "learning_rate": 2.961581965047672e-07, | |
| "loss": 0.2294, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.7619760479041915, | |
| "grad_norm": 0.2548951506614685, | |
| "learning_rate": 2.921549732474599e-07, | |
| "loss": 0.3003, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 1.7634730538922154, | |
| "grad_norm": 0.3152521252632141, | |
| "learning_rate": 2.8817817618618846e-07, | |
| "loss": 0.3112, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 1.7649700598802394, | |
| "grad_norm": 0.35076630115509033, | |
| "learning_rate": 2.842278276436128e-07, | |
| "loss": 0.2925, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 1.7664670658682635, | |
| "grad_norm": 0.3196788430213928, | |
| "learning_rate": 2.803039497939281e-07, | |
| "loss": 0.2372, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 1.7679640718562875, | |
| "grad_norm": 0.20753467082977295, | |
| "learning_rate": 2.7640656466274785e-07, | |
| "loss": 0.2861, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 1.7694610778443114, | |
| "grad_norm": 0.31822919845581055, | |
| "learning_rate": 2.7253569412697244e-07, | |
| "loss": 0.3193, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 1.7709580838323353, | |
| "grad_norm": 0.21876436471939087, | |
| "learning_rate": 2.686913599146723e-07, | |
| "loss": 0.2215, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 1.7724550898203593, | |
| "grad_norm": 0.23750582337379456, | |
| "learning_rate": 2.648735836049615e-07, | |
| "loss": 0.2909, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 1.7739520958083832, | |
| "grad_norm": 0.2619275450706482, | |
| "learning_rate": 2.6108238662788057e-07, | |
| "loss": 0.3952, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 1.7754491017964071, | |
| "grad_norm": 0.25464171171188354, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.2432, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.7769461077844313, | |
| "grad_norm": 0.31740161776542664, | |
| "learning_rate": 2.5357981564566647e-07, | |
| "loss": 0.4162, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 1.7784431137724552, | |
| "grad_norm": 0.24966104328632355, | |
| "learning_rate": 2.4986848375415653e-07, | |
| "loss": 0.229, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 1.7799401197604792, | |
| "grad_norm": 0.39699026942253113, | |
| "learning_rate": 2.4618381542228565e-07, | |
| "loss": 0.4622, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 1.781437125748503, | |
| "grad_norm": 0.2294144630432129, | |
| "learning_rate": 2.4252583133292927e-07, | |
| "loss": 0.2457, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 1.782934131736527, | |
| "grad_norm": 0.1940593421459198, | |
| "learning_rate": 2.3889455201917655e-07, | |
| "loss": 0.2306, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 1.784431137724551, | |
| "grad_norm": 0.4276283085346222, | |
| "learning_rate": 2.3528999786421758e-07, | |
| "loss": 0.2661, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 1.785928143712575, | |
| "grad_norm": 0.2756398618221283, | |
| "learning_rate": 2.3171218910122695e-07, | |
| "loss": 0.2456, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 1.7874251497005988, | |
| "grad_norm": 0.22976361215114594, | |
| "learning_rate": 2.2816114581325377e-07, | |
| "loss": 0.2645, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 1.7889221556886228, | |
| "grad_norm": 0.2517649233341217, | |
| "learning_rate": 2.2463688793310345e-07, | |
| "loss": 0.3377, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 1.7904191616766467, | |
| "grad_norm": 0.1688610017299652, | |
| "learning_rate": 2.2113943524323167e-07, | |
| "loss": 0.1488, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.7919161676646707, | |
| "grad_norm": 0.2724323868751526, | |
| "learning_rate": 2.1766880737562833e-07, | |
| "loss": 0.252, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 1.7934131736526946, | |
| "grad_norm": 0.229396253824234, | |
| "learning_rate": 2.1422502381171163e-07, | |
| "loss": 0.1885, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 1.7949101796407185, | |
| "grad_norm": 0.276480108499527, | |
| "learning_rate": 2.1080810388221406e-07, | |
| "loss": 0.2015, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 1.7964071856287425, | |
| "grad_norm": 0.27435392141342163, | |
| "learning_rate": 2.0741806676707887e-07, | |
| "loss": 0.2593, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 1.7979041916167664, | |
| "grad_norm": 0.41313275694847107, | |
| "learning_rate": 2.0405493149534828e-07, | |
| "loss": 0.3106, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 1.7994011976047903, | |
| "grad_norm": 0.2007514089345932, | |
| "learning_rate": 2.007187169450603e-07, | |
| "loss": 0.2197, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 1.8008982035928143, | |
| "grad_norm": 0.2099570631980896, | |
| "learning_rate": 1.9740944184313882e-07, | |
| "loss": 0.1886, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 1.8023952095808382, | |
| "grad_norm": 0.22008772194385529, | |
| "learning_rate": 1.941271247652915e-07, | |
| "loss": 0.2284, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 1.8038922155688621, | |
| "grad_norm": 0.3029489815235138, | |
| "learning_rate": 1.908717841359048e-07, | |
| "loss": 0.3553, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 1.805389221556886, | |
| "grad_norm": 0.24110634624958038, | |
| "learning_rate": 1.8764343822793962e-07, | |
| "loss": 0.2374, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.80688622754491, | |
| "grad_norm": 0.2744743824005127, | |
| "learning_rate": 1.8444210516283035e-07, | |
| "loss": 0.2079, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 1.8083832335329342, | |
| "grad_norm": 0.23340438306331635, | |
| "learning_rate": 1.8126780291038037e-07, | |
| "loss": 0.2714, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 1.8098802395209581, | |
| "grad_norm": 0.30072230100631714, | |
| "learning_rate": 1.7812054928866617e-07, | |
| "loss": 0.2633, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 1.811377245508982, | |
| "grad_norm": 0.30632907152175903, | |
| "learning_rate": 1.7500036196392956e-07, | |
| "loss": 0.2352, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 1.812874251497006, | |
| "grad_norm": 0.23265916109085083, | |
| "learning_rate": 1.7190725845048827e-07, | |
| "loss": 0.2612, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.81437125748503, | |
| "grad_norm": 0.22589431703090668, | |
| "learning_rate": 1.688412561106284e-07, | |
| "loss": 0.2423, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 1.8158682634730539, | |
| "grad_norm": 0.24091245234012604, | |
| "learning_rate": 1.6580237215451378e-07, | |
| "loss": 0.2122, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 1.8173652694610778, | |
| "grad_norm": 0.3324667513370514, | |
| "learning_rate": 1.6279062364008446e-07, | |
| "loss": 0.4563, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 1.818862275449102, | |
| "grad_norm": 0.33297663927078247, | |
| "learning_rate": 1.5980602747296513e-07, | |
| "loss": 0.2925, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 1.8203592814371259, | |
| "grad_norm": 0.2548763155937195, | |
| "learning_rate": 1.5684860040636573e-07, | |
| "loss": 0.232, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.8218562874251498, | |
| "grad_norm": 0.2788844406604767, | |
| "learning_rate": 1.5391835904099316e-07, | |
| "loss": 0.2535, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 1.8233532934131738, | |
| "grad_norm": 0.3000269830226898, | |
| "learning_rate": 1.510153198249531e-07, | |
| "loss": 0.2428, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 1.8248502994011977, | |
| "grad_norm": 0.25358203053474426, | |
| "learning_rate": 1.4813949905365833e-07, | |
| "loss": 0.2758, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 1.8263473053892216, | |
| "grad_norm": 0.22794456779956818, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.1812, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 1.8278443113772456, | |
| "grad_norm": 0.29520854353904724, | |
| "learning_rate": 1.424695772629553e-07, | |
| "loss": 0.2238, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 1.8293413173652695, | |
| "grad_norm": 0.23672978579998016, | |
| "learning_rate": 1.3967550807009677e-07, | |
| "loss": 0.2423, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 1.8308383233532934, | |
| "grad_norm": 0.1825004518032074, | |
| "learning_rate": 1.3690872097490481e-07, | |
| "loss": 0.2453, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 1.8323353293413174, | |
| "grad_norm": 0.22814802825450897, | |
| "learning_rate": 1.3416923150798123e-07, | |
| "loss": 0.2367, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 1.8338323353293413, | |
| "grad_norm": 0.23552152514457703, | |
| "learning_rate": 1.3145705504669592e-07, | |
| "loss": 0.2364, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 1.8353293413173652, | |
| "grad_norm": 0.2058655321598053, | |
| "learning_rate": 1.2877220681510927e-07, | |
| "loss": 0.1397, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.8368263473053892, | |
| "grad_norm": 0.243401437997818, | |
| "learning_rate": 1.261147018838782e-07, | |
| "loss": 0.2931, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 1.8383233532934131, | |
| "grad_norm": 0.2716439366340637, | |
| "learning_rate": 1.2348455517017855e-07, | |
| "loss": 0.2844, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 1.839820359281437, | |
| "grad_norm": 0.349772185087204, | |
| "learning_rate": 1.208817814376162e-07, | |
| "loss": 0.312, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 1.841317365269461, | |
| "grad_norm": 0.21398629248142242, | |
| "learning_rate": 1.1830639529614774e-07, | |
| "loss": 0.1563, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 1.842814371257485, | |
| "grad_norm": 0.22315660119056702, | |
| "learning_rate": 1.157584112019966e-07, | |
| "loss": 0.2364, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 1.8443113772455089, | |
| "grad_norm": 0.27086567878723145, | |
| "learning_rate": 1.1323784345757205e-07, | |
| "loss": 0.3122, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 1.8458083832335328, | |
| "grad_norm": 0.32250702381134033, | |
| "learning_rate": 1.1074470621138866e-07, | |
| "loss": 0.2097, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 1.8473053892215567, | |
| "grad_norm": 0.40964534878730774, | |
| "learning_rate": 1.0827901345798919e-07, | |
| "loss": 0.1931, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 1.8488023952095807, | |
| "grad_norm": 0.26256802678108215, | |
| "learning_rate": 1.0584077903786238e-07, | |
| "loss": 0.3613, | |
| "step": 1249 | |
| }, | |
| { | |
| "epoch": 1.8502994011976048, | |
| "grad_norm": 0.3172236979007721, | |
| "learning_rate": 1.0343001663736918e-07, | |
| "loss": 0.2944, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.8517964071856288, | |
| "grad_norm": 0.2462015599012375, | |
| "learning_rate": 1.0104673978866164e-07, | |
| "loss": 0.2264, | |
| "step": 1251 | |
| }, | |
| { | |
| "epoch": 1.8532934131736527, | |
| "grad_norm": 0.2893504798412323, | |
| "learning_rate": 9.869096186961025e-08, | |
| "loss": 0.3066, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 1.8547904191616766, | |
| "grad_norm": 0.33563166856765747, | |
| "learning_rate": 9.636269610372895e-08, | |
| "loss": 0.4775, | |
| "step": 1253 | |
| }, | |
| { | |
| "epoch": 1.8562874251497006, | |
| "grad_norm": 0.36685824394226074, | |
| "learning_rate": 9.406195556009745e-08, | |
| "loss": 0.3233, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 1.8577844311377245, | |
| "grad_norm": 0.35795897245407104, | |
| "learning_rate": 9.178875315329183e-08, | |
| "loss": 0.4117, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 1.8592814371257484, | |
| "grad_norm": 0.21380247175693512, | |
| "learning_rate": 8.954310164331015e-08, | |
| "loss": 0.263, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 1.8607784431137726, | |
| "grad_norm": 0.29042747616767883, | |
| "learning_rate": 8.732501363550028e-08, | |
| "loss": 0.2964, | |
| "step": 1257 | |
| }, | |
| { | |
| "epoch": 1.8622754491017965, | |
| "grad_norm": 0.3003634512424469, | |
| "learning_rate": 8.513450158049109e-08, | |
| "loss": 0.3321, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 1.8637724550898205, | |
| "grad_norm": 0.2523651719093323, | |
| "learning_rate": 8.29715777741208e-08, | |
| "loss": 0.2319, | |
| "step": 1259 | |
| }, | |
| { | |
| "epoch": 1.8652694610778444, | |
| "grad_norm": 0.30345892906188965, | |
| "learning_rate": 8.08362543573682e-08, | |
| "loss": 0.1789, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.8667664670658684, | |
| "grad_norm": 0.1946685016155243, | |
| "learning_rate": 7.872854331628599e-08, | |
| "loss": 0.2507, | |
| "step": 1261 | |
| }, | |
| { | |
| "epoch": 1.8682634730538923, | |
| "grad_norm": 0.28906092047691345, | |
| "learning_rate": 7.664845648193087e-08, | |
| "loss": 0.3652, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 1.8697604790419162, | |
| "grad_norm": 0.20177105069160461, | |
| "learning_rate": 7.459600553029966e-08, | |
| "loss": 0.2312, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 1.8712574850299402, | |
| "grad_norm": 0.3082532584667206, | |
| "learning_rate": 7.257120198226219e-08, | |
| "loss": 0.2132, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 1.872754491017964, | |
| "grad_norm": 0.25730371475219727, | |
| "learning_rate": 7.057405720349853e-08, | |
| "loss": 0.2497, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 1.874251497005988, | |
| "grad_norm": 0.21774837374687195, | |
| "learning_rate": 6.860458240443179e-08, | |
| "loss": 0.2716, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 1.875748502994012, | |
| "grad_norm": 0.2254839390516281, | |
| "learning_rate": 6.666278864016884e-08, | |
| "loss": 0.2995, | |
| "step": 1267 | |
| }, | |
| { | |
| "epoch": 1.877245508982036, | |
| "grad_norm": 0.23497092723846436, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.2641, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 1.8787425149700598, | |
| "grad_norm": 0.28167974948883057, | |
| "learning_rate": 6.286228765951807e-08, | |
| "loss": 0.3233, | |
| "step": 1269 | |
| }, | |
| { | |
| "epoch": 1.8802395209580838, | |
| "grad_norm": 0.29625046253204346, | |
| "learning_rate": 6.100360177619946e-08, | |
| "loss": 0.2986, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.8817365269461077, | |
| "grad_norm": 0.408067524433136, | |
| "learning_rate": 5.917263959370312e-08, | |
| "loss": 0.3761, | |
| "step": 1271 | |
| }, | |
| { | |
| "epoch": 1.8832335329341316, | |
| "grad_norm": 0.3008968234062195, | |
| "learning_rate": 5.736941138963281e-08, | |
| "loss": 0.3179, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 1.8847305389221556, | |
| "grad_norm": 0.188938170671463, | |
| "learning_rate": 5.559392728591517e-08, | |
| "loss": 0.1488, | |
| "step": 1273 | |
| }, | |
| { | |
| "epoch": 1.8862275449101795, | |
| "grad_norm": 0.3134278655052185, | |
| "learning_rate": 5.384619724874307e-08, | |
| "loss": 0.3308, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 1.8877245508982035, | |
| "grad_norm": 0.259612113237381, | |
| "learning_rate": 5.2126231088519e-08, | |
| "loss": 0.195, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.8892215568862274, | |
| "grad_norm": 0.22781744599342346, | |
| "learning_rate": 5.0434038459801213e-08, | |
| "loss": 0.2226, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 1.8907185628742516, | |
| "grad_norm": 0.1933296024799347, | |
| "learning_rate": 4.876962886124936e-08, | |
| "loss": 0.1725, | |
| "step": 1277 | |
| }, | |
| { | |
| "epoch": 1.8922155688622755, | |
| "grad_norm": 0.2313513308763504, | |
| "learning_rate": 4.713301163556894e-08, | |
| "loss": 0.2243, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 1.8937125748502994, | |
| "grad_norm": 0.2517513036727905, | |
| "learning_rate": 4.5524195969461895e-08, | |
| "loss": 0.3134, | |
| "step": 1279 | |
| }, | |
| { | |
| "epoch": 1.8952095808383234, | |
| "grad_norm": 0.2580326795578003, | |
| "learning_rate": 4.394319089357335e-08, | |
| "loss": 0.266, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.8967065868263473, | |
| "grad_norm": 0.2893156409263611, | |
| "learning_rate": 4.239000528244164e-08, | |
| "loss": 0.2997, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 1.8982035928143712, | |
| "grad_norm": 0.2750832736492157, | |
| "learning_rate": 4.086464785444777e-08, | |
| "loss": 0.2071, | |
| "step": 1282 | |
| }, | |
| { | |
| "epoch": 1.8997005988023952, | |
| "grad_norm": 0.38243815302848816, | |
| "learning_rate": 3.936712717176716e-08, | |
| "loss": 0.3132, | |
| "step": 1283 | |
| }, | |
| { | |
| "epoch": 1.9011976047904193, | |
| "grad_norm": 0.34103187918663025, | |
| "learning_rate": 3.7897451640321326e-08, | |
| "loss": 0.3574, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 1.9026946107784433, | |
| "grad_norm": 0.22488778829574585, | |
| "learning_rate": 3.645562950973014e-08, | |
| "loss": 0.2411, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 1.9041916167664672, | |
| "grad_norm": 0.24086658656597137, | |
| "learning_rate": 3.504166887326688e-08, | |
| "loss": 0.2503, | |
| "step": 1286 | |
| }, | |
| { | |
| "epoch": 1.9056886227544911, | |
| "grad_norm": 0.2731955945491791, | |
| "learning_rate": 3.365557766781047e-08, | |
| "loss": 0.2224, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 1.907185628742515, | |
| "grad_norm": 0.34961998462677, | |
| "learning_rate": 3.229736367380498e-08, | |
| "loss": 0.4016, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 1.908682634730539, | |
| "grad_norm": 0.18793629109859467, | |
| "learning_rate": 3.0967034515211323e-08, | |
| "loss": 0.1871, | |
| "step": 1289 | |
| }, | |
| { | |
| "epoch": 1.910179640718563, | |
| "grad_norm": 0.24469394981861115, | |
| "learning_rate": 2.966459765946672e-08, | |
| "loss": 0.3008, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.9116766467065869, | |
| "grad_norm": 0.2874278128147125, | |
| "learning_rate": 2.8390060417443632e-08, | |
| "loss": 0.3207, | |
| "step": 1291 | |
| }, | |
| { | |
| "epoch": 1.9131736526946108, | |
| "grad_norm": 0.3286723494529724, | |
| "learning_rate": 2.714342994340646e-08, | |
| "loss": 0.2332, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 1.9146706586826348, | |
| "grad_norm": 0.24583743512630463, | |
| "learning_rate": 2.592471323497381e-08, | |
| "loss": 0.2854, | |
| "step": 1293 | |
| }, | |
| { | |
| "epoch": 1.9161676646706587, | |
| "grad_norm": 0.37019965052604675, | |
| "learning_rate": 2.4733917133077378e-08, | |
| "loss": 0.3804, | |
| "step": 1294 | |
| }, | |
| { | |
| "epoch": 1.9176646706586826, | |
| "grad_norm": 0.242394357919693, | |
| "learning_rate": 2.3571048321923694e-08, | |
| "loss": 0.1835, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 1.9191616766467066, | |
| "grad_norm": 0.27056631445884705, | |
| "learning_rate": 2.2436113328958565e-08, | |
| "loss": 0.3272, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 1.9206586826347305, | |
| "grad_norm": 0.2912746071815491, | |
| "learning_rate": 2.1329118524827662e-08, | |
| "loss": 0.3609, | |
| "step": 1297 | |
| }, | |
| { | |
| "epoch": 1.9221556886227544, | |
| "grad_norm": 0.267817884683609, | |
| "learning_rate": 2.0250070123342124e-08, | |
| "loss": 0.3424, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 1.9236526946107784, | |
| "grad_norm": 0.250742644071579, | |
| "learning_rate": 1.9198974181444675e-08, | |
| "loss": 0.237, | |
| "step": 1299 | |
| }, | |
| { | |
| "epoch": 1.9251497005988023, | |
| "grad_norm": 0.281894713640213, | |
| "learning_rate": 1.8175836599173545e-08, | |
| "loss": 0.2917, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.9266467065868262, | |
| "grad_norm": 0.20631025731563568, | |
| "learning_rate": 1.7180663119630846e-08, | |
| "loss": 0.195, | |
| "step": 1301 | |
| }, | |
| { | |
| "epoch": 1.9281437125748502, | |
| "grad_norm": 0.2538008689880371, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.2873, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 1.9296407185628741, | |
| "grad_norm": 0.26172444224357605, | |
| "learning_rate": 1.5274230656263656e-08, | |
| "loss": 0.1752, | |
| "step": 1303 | |
| }, | |
| { | |
| "epoch": 1.931137724550898, | |
| "grad_norm": 0.3013996183872223, | |
| "learning_rate": 1.4362982373675171e-08, | |
| "loss": 0.3118, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 1.9326347305389222, | |
| "grad_norm": 0.29984888434410095, | |
| "learning_rate": 1.347971959622496e-08, | |
| "loss": 0.3903, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 1.9341317365269461, | |
| "grad_norm": 0.3161168098449707, | |
| "learning_rate": 1.2624447281867625e-08, | |
| "loss": 0.2794, | |
| "step": 1306 | |
| }, | |
| { | |
| "epoch": 1.93562874251497, | |
| "grad_norm": 0.20692503452301025, | |
| "learning_rate": 1.1797170231439004e-08, | |
| "loss": 0.2728, | |
| "step": 1307 | |
| }, | |
| { | |
| "epoch": 1.937125748502994, | |
| "grad_norm": 0.2914261221885681, | |
| "learning_rate": 1.0997893088632306e-08, | |
| "loss": 0.2823, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 1.938622754491018, | |
| "grad_norm": 0.2750595808029175, | |
| "learning_rate": 1.0226620339969795e-08, | |
| "loss": 0.2931, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 1.9401197604790419, | |
| "grad_norm": 0.28438398241996765, | |
| "learning_rate": 9.48335631477948e-09, | |
| "loss": 0.2434, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.9416167664670658, | |
| "grad_norm": 0.6122963428497314, | |
| "learning_rate": 8.768105185170683e-09, | |
| "loss": 0.3048, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 1.94311377245509, | |
| "grad_norm": 0.25166431069374084, | |
| "learning_rate": 8.080870966008513e-09, | |
| "loss": 0.3405, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 1.944610778443114, | |
| "grad_norm": 0.1666012555360794, | |
| "learning_rate": 7.421657514893321e-09, | |
| "loss": 0.1841, | |
| "step": 1313 | |
| }, | |
| { | |
| "epoch": 1.9461077844311379, | |
| "grad_norm": 0.2463662028312683, | |
| "learning_rate": 6.79046853214016e-09, | |
| "loss": 0.2044, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 1.9476047904191618, | |
| "grad_norm": 0.21893510222434998, | |
| "learning_rate": 6.187307560754363e-09, | |
| "loss": 0.3512, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 1.9491017964071857, | |
| "grad_norm": 0.3041020333766937, | |
| "learning_rate": 5.612177986414891e-09, | |
| "loss": 0.2655, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 1.9505988023952097, | |
| "grad_norm": 0.2836143672466278, | |
| "learning_rate": 5.065083037453234e-09, | |
| "loss": 0.503, | |
| "step": 1317 | |
| }, | |
| { | |
| "epoch": 1.9520958083832336, | |
| "grad_norm": 0.2756477892398834, | |
| "learning_rate": 4.546025784837316e-09, | |
| "loss": 0.2959, | |
| "step": 1318 | |
| }, | |
| { | |
| "epoch": 1.9535928143712575, | |
| "grad_norm": 0.23872525990009308, | |
| "learning_rate": 4.055009142152066e-09, | |
| "loss": 0.2107, | |
| "step": 1319 | |
| }, | |
| { | |
| "epoch": 1.9550898203592815, | |
| "grad_norm": 0.33056017756462097, | |
| "learning_rate": 3.5920358655844312e-09, | |
| "loss": 0.2467, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.9565868263473054, | |
| "grad_norm": 0.27123406529426575, | |
| "learning_rate": 3.1571085539089384e-09, | |
| "loss": 0.2182, | |
| "step": 1321 | |
| }, | |
| { | |
| "epoch": 1.9580838323353293, | |
| "grad_norm": 0.2840729355812073, | |
| "learning_rate": 2.7502296484699374e-09, | |
| "loss": 0.327, | |
| "step": 1322 | |
| }, | |
| { | |
| "epoch": 1.9595808383233533, | |
| "grad_norm": 0.3229871690273285, | |
| "learning_rate": 2.371401433170495e-09, | |
| "loss": 0.3329, | |
| "step": 1323 | |
| }, | |
| { | |
| "epoch": 1.9610778443113772, | |
| "grad_norm": 0.23311461508274078, | |
| "learning_rate": 2.0206260344590724e-09, | |
| "loss": 0.1865, | |
| "step": 1324 | |
| }, | |
| { | |
| "epoch": 1.9625748502994012, | |
| "grad_norm": 0.20590674877166748, | |
| "learning_rate": 1.6979054213173141e-09, | |
| "loss": 0.2345, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.964071856287425, | |
| "grad_norm": 0.2407190203666687, | |
| "learning_rate": 1.4032414052478348e-09, | |
| "loss": 0.2488, | |
| "step": 1326 | |
| }, | |
| { | |
| "epoch": 1.965568862275449, | |
| "grad_norm": 0.27954548597335815, | |
| "learning_rate": 1.136635640267003e-09, | |
| "loss": 0.3113, | |
| "step": 1327 | |
| }, | |
| { | |
| "epoch": 1.967065868263473, | |
| "grad_norm": 0.2925473153591156, | |
| "learning_rate": 8.980896228932834e-10, | |
| "loss": 0.3184, | |
| "step": 1328 | |
| }, | |
| { | |
| "epoch": 1.968562874251497, | |
| "grad_norm": 0.23581089079380035, | |
| "learning_rate": 6.876046921389102e-10, | |
| "loss": 0.2076, | |
| "step": 1329 | |
| }, | |
| { | |
| "epoch": 1.9700598802395208, | |
| "grad_norm": 0.21226383745670319, | |
| "learning_rate": 5.051820295032262e-10, | |
| "loss": 0.2688, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.9715568862275448, | |
| "grad_norm": 0.2070101648569107, | |
| "learning_rate": 3.508226589660213e-10, | |
| "loss": 0.2151, | |
| "step": 1331 | |
| }, | |
| { | |
| "epoch": 1.9730538922155687, | |
| "grad_norm": 0.19046247005462646, | |
| "learning_rate": 2.2452744698087114e-10, | |
| "loss": 0.2351, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 1.9745508982035929, | |
| "grad_norm": 0.21798643469810486, | |
| "learning_rate": 1.2629710247180626e-10, | |
| "loss": 0.1911, | |
| "step": 1333 | |
| }, | |
| { | |
| "epoch": 1.9760479041916168, | |
| "grad_norm": 0.2827717959880829, | |
| "learning_rate": 5.613217682720606e-11, | |
| "loss": 0.2928, | |
| "step": 1334 | |
| }, | |
| { | |
| "epoch": 1.9775449101796407, | |
| "grad_norm": 0.259986937046051, | |
| "learning_rate": 1.4033063899243637e-11, | |
| "loss": 0.2227, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 1.9790419161676647, | |
| "grad_norm": 0.2563554644584656, | |
| "learning_rate": 0.0, | |
| "loss": 0.2769, | |
| "step": 1336 | |
| }, | |
| { | |
| "epoch": 1.9790419161676647, | |
| "eval_loss": 0.27016571164131165, | |
| "eval_runtime": 95.929, | |
| "eval_samples_per_second": 7.672, | |
| "eval_steps_per_second": 0.959, | |
| "step": 1336 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1336, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 334, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.390317365432418e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |