diff --git "a/checkpoint-30000/trainer_state.json" "b/checkpoint-30000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-30000/trainer_state.json" @@ -0,0 +1,18034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.6, + "eval_steps": 500, + "global_step": 30000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "grad_norm": 1.5342543125152588, + "learning_rate": 3.6e-07, + "loss": 0.2018, + "step": 10 + }, + { + "grad_norm": 2.5678420066833496, + "learning_rate": 7.6e-07, + "loss": 0.1933, + "step": 20 + }, + { + "grad_norm": 2.4396657943725586, + "learning_rate": 1.16e-06, + "loss": 0.1992, + "step": 30 + }, + { + "grad_norm": 1.4771136045455933, + "learning_rate": 1.56e-06, + "loss": 0.1713, + "step": 40 + }, + { + "grad_norm": 1.3211125135421753, + "learning_rate": 1.96e-06, + "loss": 0.1829, + "step": 50 + }, + { + "grad_norm": 2.2051808834075928, + "learning_rate": 2.36e-06, + "loss": 0.1808, + "step": 60 + }, + { + "grad_norm": 1.1721709966659546, + "learning_rate": 2.7600000000000003e-06, + "loss": 0.1569, + "step": 70 + }, + { + "grad_norm": 1.0180284976959229, + "learning_rate": 3.1600000000000007e-06, + "loss": 0.1559, + "step": 80 + }, + { + "grad_norm": 0.9153556227684021, + "learning_rate": 3.5600000000000002e-06, + "loss": 0.1274, + "step": 90 + }, + { + "grad_norm": 1.1027413606643677, + "learning_rate": 3.96e-06, + "loss": 0.1239, + "step": 100 + }, + { + "grad_norm": 2.09124755859375, + "learning_rate": 4.360000000000001e-06, + "loss": 0.1238, + "step": 110 + }, + { + "grad_norm": 0.8874685168266296, + "learning_rate": 4.76e-06, + "loss": 0.1146, + "step": 120 + }, + { + "grad_norm": 0.3545866012573242, + "learning_rate": 5.1600000000000006e-06, + "loss": 0.0876, + "step": 130 + }, + { + "grad_norm": 0.34454017877578735, + "learning_rate": 5.56e-06, + "loss": 0.0894, + "step": 140 + }, + { + "grad_norm": 0.9030827879905701, + "learning_rate": 5.9600000000000005e-06, + "loss": 0.088, + "step": 150 + }, + { + "grad_norm": 0.43642544746398926, + "learning_rate": 6.360000000000001e-06, + "loss": 0.0724, + "step": 160 + }, + { + "grad_norm": 0.2760602831840515, + "learning_rate": 6.76e-06, + "loss": 0.0763, + "step": 170 + }, + { + "grad_norm": 0.4073124825954437, + "learning_rate": 7.16e-06, + "loss": 0.0807, + "step": 180 + }, + { + "grad_norm": 0.4644203782081604, + "learning_rate": 7.5600000000000005e-06, + "loss": 0.0742, + "step": 190 + }, + { + "grad_norm": 0.8297818899154663, + "learning_rate": 7.96e-06, + "loss": 0.075, + "step": 200 + }, + { + "grad_norm": 0.4910941421985626, + "learning_rate": 8.36e-06, + "loss": 0.0677, + "step": 210 + }, + { + "grad_norm": 0.8020650744438171, + "learning_rate": 8.76e-06, + "loss": 0.0745, + "step": 220 + }, + { + "grad_norm": 1.2653248310089111, + "learning_rate": 9.16e-06, + "loss": 0.0702, + "step": 230 + }, + { + "grad_norm": 1.3739262819290161, + "learning_rate": 9.560000000000002e-06, + "loss": 0.0821, + "step": 240 + }, + { + "grad_norm": 0.5403990745544434, + "learning_rate": 9.96e-06, + "loss": 0.0666, + "step": 250 + }, + { + "grad_norm": 0.6638477444648743, + "learning_rate": 1.036e-05, + "loss": 0.0621, + "step": 260 + }, + { + "grad_norm": 0.33246782422065735, + "learning_rate": 1.076e-05, + "loss": 0.0661, + "step": 270 + }, + { + "grad_norm": 0.547707200050354, + "learning_rate": 1.1160000000000002e-05, + "loss": 0.0618, + "step": 280 + }, + { + "grad_norm": 0.3923631012439728, + "learning_rate": 1.156e-05, + "loss": 0.0663, + "step": 290 + }, + { + "grad_norm": 0.35859954357147217, + "learning_rate": 1.196e-05, + "loss": 0.065, + "step": 300 + }, + { + "grad_norm": 0.32143810391426086, + "learning_rate": 1.236e-05, + "loss": 0.0637, + "step": 310 + }, + { + "grad_norm": 0.5920489430427551, + "learning_rate": 1.276e-05, + "loss": 0.0637, + "step": 320 + }, + { + "grad_norm": 0.46733343601226807, + "learning_rate": 1.316e-05, + "loss": 0.0616, + "step": 330 + }, + { + "grad_norm": 0.700232207775116, + "learning_rate": 1.356e-05, + "loss": 0.0614, + "step": 340 + }, + { + "grad_norm": 0.6487019658088684, + "learning_rate": 1.396e-05, + "loss": 0.0624, + "step": 350 + }, + { + "grad_norm": 0.9349107146263123, + "learning_rate": 1.4360000000000001e-05, + "loss": 0.0588, + "step": 360 + }, + { + "grad_norm": 0.3644509017467499, + "learning_rate": 1.4760000000000001e-05, + "loss": 0.0653, + "step": 370 + }, + { + "grad_norm": 0.33854031562805176, + "learning_rate": 1.5160000000000002e-05, + "loss": 0.0673, + "step": 380 + }, + { + "grad_norm": 0.5820100903511047, + "learning_rate": 1.556e-05, + "loss": 0.0607, + "step": 390 + }, + { + "grad_norm": 0.5228902697563171, + "learning_rate": 1.596e-05, + "loss": 0.0626, + "step": 400 + }, + { + "grad_norm": 0.45191171765327454, + "learning_rate": 1.636e-05, + "loss": 0.063, + "step": 410 + }, + { + "grad_norm": 0.3791556656360626, + "learning_rate": 1.6760000000000002e-05, + "loss": 0.0563, + "step": 420 + }, + { + "grad_norm": 1.0261706113815308, + "learning_rate": 1.7160000000000002e-05, + "loss": 0.0561, + "step": 430 + }, + { + "grad_norm": 1.2060846090316772, + "learning_rate": 1.756e-05, + "loss": 0.0618, + "step": 440 + }, + { + "grad_norm": 0.27181217074394226, + "learning_rate": 1.796e-05, + "loss": 0.0585, + "step": 450 + }, + { + "grad_norm": 0.3692938983440399, + "learning_rate": 1.8360000000000004e-05, + "loss": 0.0577, + "step": 460 + }, + { + "grad_norm": 0.3502022922039032, + "learning_rate": 1.876e-05, + "loss": 0.0586, + "step": 470 + }, + { + "grad_norm": 0.4491177797317505, + "learning_rate": 1.916e-05, + "loss": 0.06, + "step": 480 + }, + { + "grad_norm": 0.3414238393306732, + "learning_rate": 1.956e-05, + "loss": 0.0599, + "step": 490 + }, + { + "grad_norm": 0.37603023648262024, + "learning_rate": 1.9960000000000002e-05, + "loss": 0.057, + "step": 500 + }, + { + "grad_norm": 0.3067108988761902, + "learning_rate": 2.036e-05, + "loss": 0.0643, + "step": 510 + }, + { + "grad_norm": 0.4726731777191162, + "learning_rate": 2.076e-05, + "loss": 0.0628, + "step": 520 + }, + { + "grad_norm": 0.3966815769672394, + "learning_rate": 2.116e-05, + "loss": 0.0653, + "step": 530 + }, + { + "grad_norm": 0.44207993149757385, + "learning_rate": 2.1560000000000004e-05, + "loss": 0.0582, + "step": 540 + }, + { + "grad_norm": 0.3891861140727997, + "learning_rate": 2.196e-05, + "loss": 0.0638, + "step": 550 + }, + { + "grad_norm": 0.6533893346786499, + "learning_rate": 2.236e-05, + "loss": 0.066, + "step": 560 + }, + { + "grad_norm": 0.30303889513015747, + "learning_rate": 2.2760000000000002e-05, + "loss": 0.0581, + "step": 570 + }, + { + "grad_norm": 0.4397026002407074, + "learning_rate": 2.3160000000000002e-05, + "loss": 0.0591, + "step": 580 + }, + { + "grad_norm": 0.29636865854263306, + "learning_rate": 2.356e-05, + "loss": 0.0536, + "step": 590 + }, + { + "grad_norm": 0.5023694634437561, + "learning_rate": 2.396e-05, + "loss": 0.061, + "step": 600 + }, + { + "grad_norm": 0.7895945310592651, + "learning_rate": 2.4360000000000004e-05, + "loss": 0.0546, + "step": 610 + }, + { + "grad_norm": 0.41812124848365784, + "learning_rate": 2.476e-05, + "loss": 0.0554, + "step": 620 + }, + { + "grad_norm": 0.23504601418972015, + "learning_rate": 2.516e-05, + "loss": 0.0534, + "step": 630 + }, + { + "grad_norm": 0.36915016174316406, + "learning_rate": 2.556e-05, + "loss": 0.055, + "step": 640 + }, + { + "grad_norm": 0.6620070338249207, + "learning_rate": 2.5960000000000002e-05, + "loss": 0.0536, + "step": 650 + }, + { + "grad_norm": 0.5696082711219788, + "learning_rate": 2.6360000000000002e-05, + "loss": 0.0683, + "step": 660 + }, + { + "grad_norm": 0.30002257227897644, + "learning_rate": 2.676e-05, + "loss": 0.0525, + "step": 670 + }, + { + "grad_norm": 0.32420310378074646, + "learning_rate": 2.716e-05, + "loss": 0.0543, + "step": 680 + }, + { + "grad_norm": 0.5361897945404053, + "learning_rate": 2.7560000000000004e-05, + "loss": 0.0528, + "step": 690 + }, + { + "grad_norm": 0.2880547046661377, + "learning_rate": 2.7960000000000003e-05, + "loss": 0.0574, + "step": 700 + }, + { + "grad_norm": 0.5170421004295349, + "learning_rate": 2.8360000000000003e-05, + "loss": 0.054, + "step": 710 + }, + { + "grad_norm": 0.465273916721344, + "learning_rate": 2.8760000000000002e-05, + "loss": 0.0566, + "step": 720 + }, + { + "grad_norm": 0.280321329832077, + "learning_rate": 2.9160000000000005e-05, + "loss": 0.0524, + "step": 730 + }, + { + "grad_norm": 0.54316246509552, + "learning_rate": 2.9559999999999998e-05, + "loss": 0.0477, + "step": 740 + }, + { + "grad_norm": 0.7218356132507324, + "learning_rate": 2.9959999999999998e-05, + "loss": 0.0553, + "step": 750 + }, + { + "grad_norm": 0.3057969808578491, + "learning_rate": 3.036e-05, + "loss": 0.0525, + "step": 760 + }, + { + "grad_norm": 0.29656973481178284, + "learning_rate": 3.076e-05, + "loss": 0.0519, + "step": 770 + }, + { + "grad_norm": 0.3371675908565521, + "learning_rate": 3.116e-05, + "loss": 0.059, + "step": 780 + }, + { + "grad_norm": 0.2772464156150818, + "learning_rate": 3.156e-05, + "loss": 0.049, + "step": 790 + }, + { + "grad_norm": 0.3676070272922516, + "learning_rate": 3.196e-05, + "loss": 0.0494, + "step": 800 + }, + { + "grad_norm": 0.28127241134643555, + "learning_rate": 3.236e-05, + "loss": 0.051, + "step": 810 + }, + { + "grad_norm": 0.4703880250453949, + "learning_rate": 3.2760000000000005e-05, + "loss": 0.0512, + "step": 820 + }, + { + "grad_norm": 0.2652164101600647, + "learning_rate": 3.316e-05, + "loss": 0.0612, + "step": 830 + }, + { + "grad_norm": 0.3020285367965698, + "learning_rate": 3.3560000000000004e-05, + "loss": 0.0546, + "step": 840 + }, + { + "grad_norm": 0.34790706634521484, + "learning_rate": 3.396e-05, + "loss": 0.0542, + "step": 850 + }, + { + "grad_norm": 0.40061137080192566, + "learning_rate": 3.436e-05, + "loss": 0.0522, + "step": 860 + }, + { + "grad_norm": 0.536561906337738, + "learning_rate": 3.4760000000000006e-05, + "loss": 0.0608, + "step": 870 + }, + { + "grad_norm": 0.365639328956604, + "learning_rate": 3.516e-05, + "loss": 0.0501, + "step": 880 + }, + { + "grad_norm": 1.0557208061218262, + "learning_rate": 3.5560000000000005e-05, + "loss": 0.0531, + "step": 890 + }, + { + "grad_norm": 0.276216059923172, + "learning_rate": 3.596e-05, + "loss": 0.0504, + "step": 900 + }, + { + "grad_norm": 0.8172704577445984, + "learning_rate": 3.636e-05, + "loss": 0.0511, + "step": 910 + }, + { + "grad_norm": 0.4355914294719696, + "learning_rate": 3.676e-05, + "loss": 0.0492, + "step": 920 + }, + { + "grad_norm": 0.4342634081840515, + "learning_rate": 3.716e-05, + "loss": 0.0547, + "step": 930 + }, + { + "grad_norm": 0.2861712574958801, + "learning_rate": 3.756e-05, + "loss": 0.0533, + "step": 940 + }, + { + "grad_norm": 0.6106705069541931, + "learning_rate": 3.796e-05, + "loss": 0.0454, + "step": 950 + }, + { + "grad_norm": 0.30710768699645996, + "learning_rate": 3.836e-05, + "loss": 0.0526, + "step": 960 + }, + { + "grad_norm": 0.4669855535030365, + "learning_rate": 3.876e-05, + "loss": 0.0551, + "step": 970 + }, + { + "grad_norm": 0.3846342861652374, + "learning_rate": 3.9160000000000005e-05, + "loss": 0.0473, + "step": 980 + }, + { + "grad_norm": 0.4343486726284027, + "learning_rate": 3.956e-05, + "loss": 0.0488, + "step": 990 + }, + { + "grad_norm": 0.2789783477783203, + "learning_rate": 3.9960000000000004e-05, + "loss": 0.0483, + "step": 1000 + }, + { + "grad_norm": 0.31235837936401367, + "learning_rate": 4.0360000000000007e-05, + "loss": 0.0483, + "step": 1010 + }, + { + "grad_norm": 0.32143115997314453, + "learning_rate": 4.076e-05, + "loss": 0.0653, + "step": 1020 + }, + { + "grad_norm": 0.4385811686515808, + "learning_rate": 4.1160000000000006e-05, + "loss": 0.0614, + "step": 1030 + }, + { + "grad_norm": 0.5145010948181152, + "learning_rate": 4.156e-05, + "loss": 0.0629, + "step": 1040 + }, + { + "grad_norm": 0.3946899175643921, + "learning_rate": 4.196e-05, + "loss": 0.062, + "step": 1050 + }, + { + "grad_norm": 0.3191198706626892, + "learning_rate": 4.236e-05, + "loss": 0.0576, + "step": 1060 + }, + { + "grad_norm": 0.4591895639896393, + "learning_rate": 4.276e-05, + "loss": 0.0629, + "step": 1070 + }, + { + "grad_norm": 0.35052725672721863, + "learning_rate": 4.316e-05, + "loss": 0.0617, + "step": 1080 + }, + { + "grad_norm": 0.3962028920650482, + "learning_rate": 4.356e-05, + "loss": 0.0561, + "step": 1090 + }, + { + "grad_norm": 0.3900878131389618, + "learning_rate": 4.396e-05, + "loss": 0.0541, + "step": 1100 + }, + { + "grad_norm": 0.42446741461753845, + "learning_rate": 4.436e-05, + "loss": 0.0653, + "step": 1110 + }, + { + "grad_norm": 0.4316879212856293, + "learning_rate": 4.4760000000000005e-05, + "loss": 0.052, + "step": 1120 + }, + { + "grad_norm": 0.3290082514286041, + "learning_rate": 4.516e-05, + "loss": 0.0603, + "step": 1130 + }, + { + "grad_norm": 0.44073694944381714, + "learning_rate": 4.5560000000000004e-05, + "loss": 0.0687, + "step": 1140 + }, + { + "grad_norm": 0.34532594680786133, + "learning_rate": 4.596e-05, + "loss": 0.0638, + "step": 1150 + }, + { + "grad_norm": 0.4461052715778351, + "learning_rate": 4.636e-05, + "loss": 0.0585, + "step": 1160 + }, + { + "grad_norm": 1.0743989944458008, + "learning_rate": 4.6760000000000006e-05, + "loss": 0.0564, + "step": 1170 + }, + { + "grad_norm": 0.32037389278411865, + "learning_rate": 4.716e-05, + "loss": 0.0542, + "step": 1180 + }, + { + "grad_norm": 0.2531489431858063, + "learning_rate": 4.7560000000000005e-05, + "loss": 0.0586, + "step": 1190 + }, + { + "grad_norm": 0.3873251676559448, + "learning_rate": 4.796e-05, + "loss": 0.0665, + "step": 1200 + }, + { + "grad_norm": 0.34096768498420715, + "learning_rate": 4.836e-05, + "loss": 0.0618, + "step": 1210 + }, + { + "grad_norm": 0.39880308508872986, + "learning_rate": 4.876e-05, + "loss": 0.0612, + "step": 1220 + }, + { + "grad_norm": 0.47188231348991394, + "learning_rate": 4.9160000000000004e-05, + "loss": 0.0608, + "step": 1230 + }, + { + "grad_norm": 0.2783375084400177, + "learning_rate": 4.956e-05, + "loss": 0.0584, + "step": 1240 + }, + { + "grad_norm": 0.5556674599647522, + "learning_rate": 4.996e-05, + "loss": 0.0561, + "step": 1250 + }, + { + "grad_norm": 0.31111642718315125, + "learning_rate": 5.0360000000000006e-05, + "loss": 0.0551, + "step": 1260 + }, + { + "grad_norm": 0.3899536430835724, + "learning_rate": 5.076000000000001e-05, + "loss": 0.0527, + "step": 1270 + }, + { + "grad_norm": 0.4737466871738434, + "learning_rate": 5.1160000000000005e-05, + "loss": 0.0586, + "step": 1280 + }, + { + "grad_norm": 0.2724415361881256, + "learning_rate": 5.1559999999999994e-05, + "loss": 0.0581, + "step": 1290 + }, + { + "grad_norm": 0.2751433253288269, + "learning_rate": 5.196e-05, + "loss": 0.0544, + "step": 1300 + }, + { + "grad_norm": 0.3903752565383911, + "learning_rate": 5.236e-05, + "loss": 0.0592, + "step": 1310 + }, + { + "grad_norm": 0.316282719373703, + "learning_rate": 5.2759999999999996e-05, + "loss": 0.0575, + "step": 1320 + }, + { + "grad_norm": 0.2514778673648834, + "learning_rate": 5.316e-05, + "loss": 0.0512, + "step": 1330 + }, + { + "grad_norm": 0.5850709080696106, + "learning_rate": 5.356e-05, + "loss": 0.0485, + "step": 1340 + }, + { + "grad_norm": 0.48276272416114807, + "learning_rate": 5.396e-05, + "loss": 0.0609, + "step": 1350 + }, + { + "grad_norm": 0.3128260672092438, + "learning_rate": 5.436e-05, + "loss": 0.0586, + "step": 1360 + }, + { + "grad_norm": 0.41716521978378296, + "learning_rate": 5.476e-05, + "loss": 0.0498, + "step": 1370 + }, + { + "grad_norm": 0.24611888825893402, + "learning_rate": 5.516e-05, + "loss": 0.0519, + "step": 1380 + }, + { + "grad_norm": 0.4254191517829895, + "learning_rate": 5.556e-05, + "loss": 0.0561, + "step": 1390 + }, + { + "grad_norm": 0.46240922808647156, + "learning_rate": 5.596e-05, + "loss": 0.0516, + "step": 1400 + }, + { + "grad_norm": 0.521133303642273, + "learning_rate": 5.636e-05, + "loss": 0.0545, + "step": 1410 + }, + { + "grad_norm": 0.56767338514328, + "learning_rate": 5.6760000000000005e-05, + "loss": 0.0593, + "step": 1420 + }, + { + "grad_norm": 0.598982572555542, + "learning_rate": 5.716e-05, + "loss": 0.0466, + "step": 1430 + }, + { + "grad_norm": 0.40146076679229736, + "learning_rate": 5.7560000000000005e-05, + "loss": 0.0556, + "step": 1440 + }, + { + "grad_norm": 0.5066417455673218, + "learning_rate": 5.796e-05, + "loss": 0.0598, + "step": 1450 + }, + { + "grad_norm": 0.6498299241065979, + "learning_rate": 5.8360000000000004e-05, + "loss": 0.0596, + "step": 1460 + }, + { + "grad_norm": 0.386627733707428, + "learning_rate": 5.876000000000001e-05, + "loss": 0.0587, + "step": 1470 + }, + { + "grad_norm": 0.5395021438598633, + "learning_rate": 5.916e-05, + "loss": 0.0512, + "step": 1480 + }, + { + "grad_norm": 0.33391037583351135, + "learning_rate": 5.9560000000000006e-05, + "loss": 0.055, + "step": 1490 + }, + { + "grad_norm": 0.3497184216976166, + "learning_rate": 5.996e-05, + "loss": 0.0556, + "step": 1500 + }, + { + "grad_norm": 0.4519764482975006, + "learning_rate": 6.0360000000000005e-05, + "loss": 0.051, + "step": 1510 + }, + { + "grad_norm": 0.3308703899383545, + "learning_rate": 6.076000000000001e-05, + "loss": 0.0581, + "step": 1520 + }, + { + "grad_norm": 0.3152478039264679, + "learning_rate": 6.116e-05, + "loss": 0.0599, + "step": 1530 + }, + { + "grad_norm": 0.2613620162010193, + "learning_rate": 6.156e-05, + "loss": 0.0668, + "step": 1540 + }, + { + "grad_norm": 0.6287797093391418, + "learning_rate": 6.196000000000001e-05, + "loss": 0.0602, + "step": 1550 + }, + { + "grad_norm": 1.6549569368362427, + "learning_rate": 6.236e-05, + "loss": 0.0697, + "step": 1560 + }, + { + "grad_norm": 0.4459933638572693, + "learning_rate": 6.276e-05, + "loss": 0.0587, + "step": 1570 + }, + { + "grad_norm": 0.46182262897491455, + "learning_rate": 6.316000000000001e-05, + "loss": 0.0655, + "step": 1580 + }, + { + "grad_norm": 0.3341088891029358, + "learning_rate": 6.356000000000001e-05, + "loss": 0.0644, + "step": 1590 + }, + { + "grad_norm": 0.3770160377025604, + "learning_rate": 6.396e-05, + "loss": 0.0583, + "step": 1600 + }, + { + "grad_norm": 0.4884144961833954, + "learning_rate": 6.436e-05, + "loss": 0.0537, + "step": 1610 + }, + { + "grad_norm": 1.0907537937164307, + "learning_rate": 6.476e-05, + "loss": 0.0648, + "step": 1620 + }, + { + "grad_norm": 0.43327778577804565, + "learning_rate": 6.515999999999999e-05, + "loss": 0.0886, + "step": 1630 + }, + { + "grad_norm": 0.5364894270896912, + "learning_rate": 6.556e-05, + "loss": 0.0668, + "step": 1640 + }, + { + "grad_norm": 0.3677687644958496, + "learning_rate": 6.596e-05, + "loss": 0.0646, + "step": 1650 + }, + { + "grad_norm": 0.32623112201690674, + "learning_rate": 6.636e-05, + "loss": 0.0651, + "step": 1660 + }, + { + "grad_norm": 0.45597919821739197, + "learning_rate": 6.676e-05, + "loss": 0.0597, + "step": 1670 + }, + { + "grad_norm": 0.34211739897727966, + "learning_rate": 6.716e-05, + "loss": 0.061, + "step": 1680 + }, + { + "grad_norm": 0.30173254013061523, + "learning_rate": 6.756e-05, + "loss": 0.0608, + "step": 1690 + }, + { + "grad_norm": 0.583385169506073, + "learning_rate": 6.796e-05, + "loss": 0.0679, + "step": 1700 + }, + { + "grad_norm": 0.36351579427719116, + "learning_rate": 6.836e-05, + "loss": 0.0661, + "step": 1710 + }, + { + "grad_norm": 0.2929661273956299, + "learning_rate": 6.876e-05, + "loss": 0.0566, + "step": 1720 + }, + { + "grad_norm": 0.39732998609542847, + "learning_rate": 6.916000000000001e-05, + "loss": 0.0582, + "step": 1730 + }, + { + "grad_norm": 0.44213226437568665, + "learning_rate": 6.956e-05, + "loss": 0.0611, + "step": 1740 + }, + { + "grad_norm": 0.28369325399398804, + "learning_rate": 6.996e-05, + "loss": 0.0635, + "step": 1750 + }, + { + "grad_norm": 0.43503010272979736, + "learning_rate": 7.036e-05, + "loss": 0.0582, + "step": 1760 + }, + { + "grad_norm": 0.3156523108482361, + "learning_rate": 7.076000000000001e-05, + "loss": 0.0577, + "step": 1770 + }, + { + "grad_norm": 0.5001112222671509, + "learning_rate": 7.116e-05, + "loss": 0.066, + "step": 1780 + }, + { + "grad_norm": 0.3080451786518097, + "learning_rate": 7.156e-05, + "loss": 0.0566, + "step": 1790 + }, + { + "grad_norm": 0.831169843673706, + "learning_rate": 7.196000000000001e-05, + "loss": 0.0584, + "step": 1800 + }, + { + "grad_norm": 0.3193128705024719, + "learning_rate": 7.236e-05, + "loss": 0.0551, + "step": 1810 + }, + { + "grad_norm": 0.37571078538894653, + "learning_rate": 7.276e-05, + "loss": 0.0613, + "step": 1820 + }, + { + "grad_norm": 0.4342571496963501, + "learning_rate": 7.316000000000001e-05, + "loss": 0.0644, + "step": 1830 + }, + { + "grad_norm": 0.3834092915058136, + "learning_rate": 7.356000000000001e-05, + "loss": 0.0617, + "step": 1840 + }, + { + "grad_norm": 0.4150540828704834, + "learning_rate": 7.396e-05, + "loss": 0.0684, + "step": 1850 + }, + { + "grad_norm": 0.3278968334197998, + "learning_rate": 7.436000000000001e-05, + "loss": 0.0666, + "step": 1860 + }, + { + "grad_norm": 0.48192402720451355, + "learning_rate": 7.476000000000001e-05, + "loss": 0.0577, + "step": 1870 + }, + { + "grad_norm": 0.38662004470825195, + "learning_rate": 7.516e-05, + "loss": 0.0565, + "step": 1880 + }, + { + "grad_norm": 0.6272531747817993, + "learning_rate": 7.556000000000002e-05, + "loss": 0.0617, + "step": 1890 + }, + { + "grad_norm": 0.21432529389858246, + "learning_rate": 7.596000000000001e-05, + "loss": 0.0558, + "step": 1900 + }, + { + "grad_norm": 0.3008495271205902, + "learning_rate": 7.636e-05, + "loss": 0.0631, + "step": 1910 + }, + { + "grad_norm": 0.36212414503097534, + "learning_rate": 7.676e-05, + "loss": 0.0589, + "step": 1920 + }, + { + "grad_norm": 0.3560740649700165, + "learning_rate": 7.716e-05, + "loss": 0.0569, + "step": 1930 + }, + { + "grad_norm": 0.41767972707748413, + "learning_rate": 7.756e-05, + "loss": 0.0594, + "step": 1940 + }, + { + "grad_norm": 0.35767459869384766, + "learning_rate": 7.796e-05, + "loss": 0.0616, + "step": 1950 + }, + { + "grad_norm": 0.31310945749282837, + "learning_rate": 7.836e-05, + "loss": 0.063, + "step": 1960 + }, + { + "grad_norm": 0.37815824151039124, + "learning_rate": 7.876e-05, + "loss": 0.0588, + "step": 1970 + }, + { + "grad_norm": 0.3359051048755646, + "learning_rate": 7.916e-05, + "loss": 0.0639, + "step": 1980 + }, + { + "grad_norm": 0.34196144342422485, + "learning_rate": 7.956e-05, + "loss": 0.0539, + "step": 1990 + }, + { + "grad_norm": 0.3005351126194, + "learning_rate": 7.996e-05, + "loss": 0.0626, + "step": 2000 + }, + { + "grad_norm": 0.36942023038864136, + "learning_rate": 8.036e-05, + "loss": 0.072, + "step": 2010 + }, + { + "grad_norm": 0.37583857774734497, + "learning_rate": 8.076e-05, + "loss": 0.0631, + "step": 2020 + }, + { + "grad_norm": 1.0421243906021118, + "learning_rate": 8.116e-05, + "loss": 0.0562, + "step": 2030 + }, + { + "grad_norm": 0.303426057100296, + "learning_rate": 8.156e-05, + "loss": 0.0653, + "step": 2040 + }, + { + "grad_norm": 0.3465934991836548, + "learning_rate": 8.196000000000001e-05, + "loss": 0.068, + "step": 2050 + }, + { + "grad_norm": 0.4840015769004822, + "learning_rate": 8.236e-05, + "loss": 0.0629, + "step": 2060 + }, + { + "grad_norm": 0.6769399046897888, + "learning_rate": 8.276e-05, + "loss": 0.0634, + "step": 2070 + }, + { + "grad_norm": 0.5165045857429504, + "learning_rate": 8.316000000000001e-05, + "loss": 0.0592, + "step": 2080 + }, + { + "grad_norm": 0.3529292941093445, + "learning_rate": 8.356e-05, + "loss": 0.0635, + "step": 2090 + }, + { + "grad_norm": 0.47214287519454956, + "learning_rate": 8.396e-05, + "loss": 0.0582, + "step": 2100 + }, + { + "grad_norm": 0.2411620169878006, + "learning_rate": 8.436000000000001e-05, + "loss": 0.0689, + "step": 2110 + }, + { + "grad_norm": 0.4558154046535492, + "learning_rate": 8.476000000000001e-05, + "loss": 0.0593, + "step": 2120 + }, + { + "grad_norm": 0.6284074187278748, + "learning_rate": 8.516e-05, + "loss": 0.0618, + "step": 2130 + }, + { + "grad_norm": 0.5316752791404724, + "learning_rate": 8.556e-05, + "loss": 0.0553, + "step": 2140 + }, + { + "grad_norm": 0.4434513747692108, + "learning_rate": 8.596000000000001e-05, + "loss": 0.0629, + "step": 2150 + }, + { + "grad_norm": 0.5474179983139038, + "learning_rate": 8.636e-05, + "loss": 0.0568, + "step": 2160 + }, + { + "grad_norm": 1.0496461391448975, + "learning_rate": 8.676e-05, + "loss": 0.0568, + "step": 2170 + }, + { + "grad_norm": 0.8544120788574219, + "learning_rate": 8.716000000000001e-05, + "loss": 0.061, + "step": 2180 + }, + { + "grad_norm": 0.6749675273895264, + "learning_rate": 8.756000000000001e-05, + "loss": 0.0554, + "step": 2190 + }, + { + "grad_norm": 0.4508029818534851, + "learning_rate": 8.796e-05, + "loss": 0.058, + "step": 2200 + }, + { + "grad_norm": 0.28224825859069824, + "learning_rate": 8.836000000000001e-05, + "loss": 0.0564, + "step": 2210 + }, + { + "grad_norm": 0.4218464195728302, + "learning_rate": 8.876e-05, + "loss": 0.0561, + "step": 2220 + }, + { + "grad_norm": 0.3189133107662201, + "learning_rate": 8.916e-05, + "loss": 0.0512, + "step": 2230 + }, + { + "grad_norm": 0.34293559193611145, + "learning_rate": 8.956e-05, + "loss": 0.055, + "step": 2240 + }, + { + "grad_norm": 0.3806557059288025, + "learning_rate": 8.996e-05, + "loss": 0.059, + "step": 2250 + }, + { + "grad_norm": 1.344613790512085, + "learning_rate": 9.036e-05, + "loss": 0.0558, + "step": 2260 + }, + { + "grad_norm": 0.5414432287216187, + "learning_rate": 9.076e-05, + "loss": 0.0508, + "step": 2270 + }, + { + "grad_norm": 0.4002189040184021, + "learning_rate": 9.116e-05, + "loss": 0.0573, + "step": 2280 + }, + { + "grad_norm": 0.6896014213562012, + "learning_rate": 9.156e-05, + "loss": 0.0691, + "step": 2290 + }, + { + "grad_norm": 0.3006056845188141, + "learning_rate": 9.196000000000001e-05, + "loss": 0.0532, + "step": 2300 + }, + { + "grad_norm": 0.4386611878871918, + "learning_rate": 9.236e-05, + "loss": 0.0599, + "step": 2310 + }, + { + "grad_norm": 0.3190461993217468, + "learning_rate": 9.276e-05, + "loss": 0.0511, + "step": 2320 + }, + { + "grad_norm": 0.47109025716781616, + "learning_rate": 9.316000000000001e-05, + "loss": 0.0625, + "step": 2330 + }, + { + "grad_norm": 0.42380431294441223, + "learning_rate": 9.356e-05, + "loss": 0.0556, + "step": 2340 + }, + { + "grad_norm": 0.4474894404411316, + "learning_rate": 9.396e-05, + "loss": 0.0628, + "step": 2350 + }, + { + "grad_norm": 0.7644767165184021, + "learning_rate": 9.436e-05, + "loss": 0.0603, + "step": 2360 + }, + { + "grad_norm": 0.20851793885231018, + "learning_rate": 9.476000000000001e-05, + "loss": 0.0542, + "step": 2370 + }, + { + "grad_norm": 0.3603307902812958, + "learning_rate": 9.516e-05, + "loss": 0.0588, + "step": 2380 + }, + { + "grad_norm": 0.3683120906352997, + "learning_rate": 9.556e-05, + "loss": 0.0617, + "step": 2390 + }, + { + "grad_norm": 0.42983752489089966, + "learning_rate": 9.596000000000001e-05, + "loss": 0.0583, + "step": 2400 + }, + { + "grad_norm": 0.519270122051239, + "learning_rate": 9.636e-05, + "loss": 0.0569, + "step": 2410 + }, + { + "grad_norm": 0.32160207629203796, + "learning_rate": 9.676e-05, + "loss": 0.0711, + "step": 2420 + }, + { + "grad_norm": 0.42999571561813354, + "learning_rate": 9.716000000000001e-05, + "loss": 0.0647, + "step": 2430 + }, + { + "grad_norm": 0.29067036509513855, + "learning_rate": 9.756000000000001e-05, + "loss": 0.0657, + "step": 2440 + }, + { + "grad_norm": 0.436850905418396, + "learning_rate": 9.796e-05, + "loss": 0.0596, + "step": 2450 + }, + { + "grad_norm": 0.3162844479084015, + "learning_rate": 9.836000000000001e-05, + "loss": 0.0513, + "step": 2460 + }, + { + "grad_norm": 0.35924121737480164, + "learning_rate": 9.876000000000001e-05, + "loss": 0.0538, + "step": 2470 + }, + { + "grad_norm": 0.42179298400878906, + "learning_rate": 9.916e-05, + "loss": 0.0571, + "step": 2480 + }, + { + "grad_norm": 0.33225223422050476, + "learning_rate": 9.956e-05, + "loss": 0.0585, + "step": 2490 + }, + { + "grad_norm": 0.30350151658058167, + "learning_rate": 9.996000000000001e-05, + "loss": 0.0558, + "step": 2500 + }, + { + "grad_norm": 0.34478995203971863, + "learning_rate": 9.999999114196196e-05, + "loss": 0.0521, + "step": 2510 + }, + { + "grad_norm": 0.49204564094543457, + "learning_rate": 9.99999605215876e-05, + "loss": 0.0565, + "step": 2520 + }, + { + "grad_norm": 2.378544330596924, + "learning_rate": 9.999990802953179e-05, + "loss": 0.0519, + "step": 2530 + }, + { + "grad_norm": 0.4369680881500244, + "learning_rate": 9.99998336658175e-05, + "loss": 0.0617, + "step": 2540 + }, + { + "grad_norm": 0.3564000427722931, + "learning_rate": 9.999973743047727e-05, + "loss": 0.0596, + "step": 2550 + }, + { + "grad_norm": 0.3950352668762207, + "learning_rate": 9.999961932355319e-05, + "loss": 0.0605, + "step": 2560 + }, + { + "grad_norm": 0.3422238230705261, + "learning_rate": 9.999947934509693e-05, + "loss": 0.0611, + "step": 2570 + }, + { + "grad_norm": 0.5072673559188843, + "learning_rate": 9.999931749516971e-05, + "loss": 0.0618, + "step": 2580 + }, + { + "grad_norm": 0.43182674050331116, + "learning_rate": 9.999913377384233e-05, + "loss": 0.064, + "step": 2590 + }, + { + "grad_norm": 0.34264642000198364, + "learning_rate": 9.999892818119517e-05, + "loss": 0.0679, + "step": 2600 + }, + { + "grad_norm": 0.2715684175491333, + "learning_rate": 9.999870071731814e-05, + "loss": 0.0637, + "step": 2610 + }, + { + "grad_norm": 0.5193445682525635, + "learning_rate": 9.999845138231076e-05, + "loss": 0.0622, + "step": 2620 + }, + { + "grad_norm": 0.3659679889678955, + "learning_rate": 9.999818017628208e-05, + "loss": 0.0646, + "step": 2630 + }, + { + "grad_norm": 0.3637860119342804, + "learning_rate": 9.999788709935078e-05, + "loss": 0.0701, + "step": 2640 + }, + { + "grad_norm": 0.2345864325761795, + "learning_rate": 9.9997572151645e-05, + "loss": 0.0596, + "step": 2650 + }, + { + "grad_norm": 0.5347143411636353, + "learning_rate": 9.999723533330254e-05, + "loss": 0.0646, + "step": 2660 + }, + { + "grad_norm": 0.36904141306877136, + "learning_rate": 9.999687664447074e-05, + "loss": 0.0622, + "step": 2670 + }, + { + "grad_norm": 1.1589168310165405, + "learning_rate": 9.99964960853065e-05, + "loss": 0.0706, + "step": 2680 + }, + { + "grad_norm": 0.3876209855079651, + "learning_rate": 9.999609365597627e-05, + "loss": 0.0633, + "step": 2690 + }, + { + "grad_norm": 0.5037931203842163, + "learning_rate": 9.99956693566561e-05, + "loss": 0.0564, + "step": 2700 + }, + { + "grad_norm": 0.5626908540725708, + "learning_rate": 9.99952231875316e-05, + "loss": 0.062, + "step": 2710 + }, + { + "grad_norm": 0.4957813620567322, + "learning_rate": 9.999475514879795e-05, + "loss": 0.0769, + "step": 2720 + }, + { + "grad_norm": 0.38109272718429565, + "learning_rate": 9.999426524065984e-05, + "loss": 0.0606, + "step": 2730 + }, + { + "grad_norm": 0.6560131311416626, + "learning_rate": 9.999375346333162e-05, + "loss": 0.0561, + "step": 2740 + }, + { + "grad_norm": 0.3259795010089874, + "learning_rate": 9.999321981703715e-05, + "loss": 0.0615, + "step": 2750 + }, + { + "grad_norm": 0.44143277406692505, + "learning_rate": 9.999266430200985e-05, + "loss": 0.0606, + "step": 2760 + }, + { + "grad_norm": 0.40541836619377136, + "learning_rate": 9.999208691849271e-05, + "loss": 0.0579, + "step": 2770 + }, + { + "grad_norm": 0.3301512897014618, + "learning_rate": 9.999148766673832e-05, + "loss": 0.0651, + "step": 2780 + }, + { + "grad_norm": 0.2767590582370758, + "learning_rate": 9.999086654700881e-05, + "loss": 0.0667, + "step": 2790 + }, + { + "grad_norm": 0.4223048985004425, + "learning_rate": 9.999022355957588e-05, + "loss": 0.0612, + "step": 2800 + }, + { + "grad_norm": 0.3622538447380066, + "learning_rate": 9.998955870472079e-05, + "loss": 0.0754, + "step": 2810 + }, + { + "grad_norm": 0.40687400102615356, + "learning_rate": 9.998887198273437e-05, + "loss": 0.0592, + "step": 2820 + }, + { + "grad_norm": 0.19539569318294525, + "learning_rate": 9.998816339391701e-05, + "loss": 0.0552, + "step": 2830 + }, + { + "grad_norm": 0.3722985088825226, + "learning_rate": 9.998743293857868e-05, + "loss": 0.0599, + "step": 2840 + }, + { + "grad_norm": 0.3133811056613922, + "learning_rate": 9.998668061703891e-05, + "loss": 0.0588, + "step": 2850 + }, + { + "grad_norm": 0.37510183453559875, + "learning_rate": 9.998590642962679e-05, + "loss": 0.0578, + "step": 2860 + }, + { + "grad_norm": 0.24154698848724365, + "learning_rate": 9.998511037668095e-05, + "loss": 0.0501, + "step": 2870 + }, + { + "grad_norm": 0.22546043992042542, + "learning_rate": 9.998429245854964e-05, + "loss": 0.0476, + "step": 2880 + }, + { + "grad_norm": 0.2856214940547943, + "learning_rate": 9.998345267559064e-05, + "loss": 0.0526, + "step": 2890 + }, + { + "grad_norm": 0.2906782627105713, + "learning_rate": 9.998259102817129e-05, + "loss": 0.065, + "step": 2900 + }, + { + "grad_norm": 0.3082631528377533, + "learning_rate": 9.99817075166685e-05, + "loss": 0.0483, + "step": 2910 + }, + { + "grad_norm": 0.5731604099273682, + "learning_rate": 9.998080214146878e-05, + "loss": 0.0533, + "step": 2920 + }, + { + "grad_norm": 0.26483601331710815, + "learning_rate": 9.997987490296813e-05, + "loss": 0.044, + "step": 2930 + }, + { + "grad_norm": 0.21970583498477936, + "learning_rate": 9.99789258015722e-05, + "loss": 0.0517, + "step": 2940 + }, + { + "grad_norm": 0.2651503384113312, + "learning_rate": 9.997795483769611e-05, + "loss": 0.0496, + "step": 2950 + }, + { + "grad_norm": 0.5049136281013489, + "learning_rate": 9.997696201176462e-05, + "loss": 0.051, + "step": 2960 + }, + { + "grad_norm": 0.42929860949516296, + "learning_rate": 9.997594732421203e-05, + "loss": 0.0535, + "step": 2970 + }, + { + "grad_norm": 0.3046446740627289, + "learning_rate": 9.997491077548217e-05, + "loss": 0.0565, + "step": 2980 + }, + { + "grad_norm": 0.24823695421218872, + "learning_rate": 9.997385236602851e-05, + "loss": 0.0554, + "step": 2990 + }, + { + "grad_norm": 0.3062506914138794, + "learning_rate": 9.997277209631399e-05, + "loss": 0.0571, + "step": 3000 + }, + { + "grad_norm": 0.636642336845398, + "learning_rate": 9.997166996681118e-05, + "loss": 0.0504, + "step": 3010 + }, + { + "grad_norm": 0.40153154730796814, + "learning_rate": 9.997054597800218e-05, + "loss": 0.0549, + "step": 3020 + }, + { + "grad_norm": 0.20223447680473328, + "learning_rate": 9.996940013037866e-05, + "loss": 0.0544, + "step": 3030 + }, + { + "grad_norm": 1.0459119081497192, + "learning_rate": 9.996823242444186e-05, + "loss": 0.0644, + "step": 3040 + }, + { + "grad_norm": 0.3025889992713928, + "learning_rate": 9.996704286070258e-05, + "loss": 0.0624, + "step": 3050 + }, + { + "grad_norm": 0.37579914927482605, + "learning_rate": 9.996583143968115e-05, + "loss": 0.0677, + "step": 3060 + }, + { + "grad_norm": 0.24798591434955597, + "learning_rate": 9.99645981619075e-05, + "loss": 0.055, + "step": 3070 + }, + { + "grad_norm": 0.25343555212020874, + "learning_rate": 9.996334302792114e-05, + "loss": 0.0613, + "step": 3080 + }, + { + "grad_norm": 0.4283932149410248, + "learning_rate": 9.996206603827105e-05, + "loss": 0.0714, + "step": 3090 + }, + { + "grad_norm": 0.3279590308666229, + "learning_rate": 9.996076719351587e-05, + "loss": 0.0595, + "step": 3100 + }, + { + "grad_norm": 0.3750005066394806, + "learning_rate": 9.995944649422374e-05, + "loss": 0.06, + "step": 3110 + }, + { + "grad_norm": 0.4188271760940552, + "learning_rate": 9.995810394097239e-05, + "loss": 0.0617, + "step": 3120 + }, + { + "grad_norm": 0.2905043959617615, + "learning_rate": 9.995673953434909e-05, + "loss": 0.0617, + "step": 3130 + }, + { + "grad_norm": 0.42362311482429504, + "learning_rate": 9.995535327495068e-05, + "loss": 0.0598, + "step": 3140 + }, + { + "grad_norm": 0.24688857793807983, + "learning_rate": 9.995394516338355e-05, + "loss": 0.0597, + "step": 3150 + }, + { + "grad_norm": 0.40488284826278687, + "learning_rate": 9.995251520026367e-05, + "loss": 0.0652, + "step": 3160 + }, + { + "grad_norm": 0.2746496796607971, + "learning_rate": 9.995106338621656e-05, + "loss": 0.0633, + "step": 3170 + }, + { + "grad_norm": 0.5197851657867432, + "learning_rate": 9.994958972187726e-05, + "loss": 0.0674, + "step": 3180 + }, + { + "grad_norm": 0.35745927691459656, + "learning_rate": 9.994809420789044e-05, + "loss": 0.0619, + "step": 3190 + }, + { + "grad_norm": 0.3419089913368225, + "learning_rate": 9.994657684491027e-05, + "loss": 0.0648, + "step": 3200 + }, + { + "grad_norm": 0.4323773682117462, + "learning_rate": 9.994503763360048e-05, + "loss": 0.0567, + "step": 3210 + }, + { + "grad_norm": 0.2885483205318451, + "learning_rate": 9.99434765746344e-05, + "loss": 0.0644, + "step": 3220 + }, + { + "grad_norm": 0.4111396372318268, + "learning_rate": 9.994189366869488e-05, + "loss": 0.063, + "step": 3230 + }, + { + "grad_norm": 0.276893675327301, + "learning_rate": 9.994028891647433e-05, + "loss": 0.0579, + "step": 3240 + }, + { + "grad_norm": 0.3733084499835968, + "learning_rate": 9.993866231867475e-05, + "loss": 0.0655, + "step": 3250 + }, + { + "grad_norm": 0.43024298548698425, + "learning_rate": 9.993701387600762e-05, + "loss": 0.0587, + "step": 3260 + }, + { + "grad_norm": 0.8470737338066101, + "learning_rate": 9.993534358919408e-05, + "loss": 0.0649, + "step": 3270 + }, + { + "grad_norm": 0.26634731888771057, + "learning_rate": 9.993365145896473e-05, + "loss": 0.0494, + "step": 3280 + }, + { + "grad_norm": 0.2917919158935547, + "learning_rate": 9.993193748605977e-05, + "loss": 0.0563, + "step": 3290 + }, + { + "grad_norm": 0.2604598104953766, + "learning_rate": 9.993020167122898e-05, + "loss": 0.0531, + "step": 3300 + }, + { + "grad_norm": 0.3339392840862274, + "learning_rate": 9.992844401523164e-05, + "loss": 0.06, + "step": 3310 + }, + { + "grad_norm": 0.35684725642204285, + "learning_rate": 9.992666451883661e-05, + "loss": 0.0625, + "step": 3320 + }, + { + "grad_norm": 0.38751131296157837, + "learning_rate": 9.99248631828223e-05, + "loss": 0.0553, + "step": 3330 + }, + { + "grad_norm": 0.3519878685474396, + "learning_rate": 9.99230400079767e-05, + "loss": 0.0576, + "step": 3340 + }, + { + "grad_norm": 0.5501101016998291, + "learning_rate": 9.992119499509728e-05, + "loss": 0.0574, + "step": 3350 + }, + { + "grad_norm": 1.0512702465057373, + "learning_rate": 9.991932814499114e-05, + "loss": 0.0585, + "step": 3360 + }, + { + "grad_norm": 0.25948435068130493, + "learning_rate": 9.991743945847493e-05, + "loss": 0.0581, + "step": 3370 + }, + { + "grad_norm": 1.9880280494689941, + "learning_rate": 9.991552893637478e-05, + "loss": 0.0808, + "step": 3380 + }, + { + "grad_norm": 1.5339511632919312, + "learning_rate": 9.991359657952644e-05, + "loss": 0.0575, + "step": 3390 + }, + { + "grad_norm": 0.23251667618751526, + "learning_rate": 9.991164238877519e-05, + "loss": 0.0603, + "step": 3400 + }, + { + "grad_norm": 0.43117889761924744, + "learning_rate": 9.990966636497585e-05, + "loss": 0.0606, + "step": 3410 + }, + { + "grad_norm": 0.40009692311286926, + "learning_rate": 9.99076685089928e-05, + "loss": 0.0587, + "step": 3420 + }, + { + "grad_norm": 0.34918004274368286, + "learning_rate": 9.990564882169998e-05, + "loss": 0.0601, + "step": 3430 + }, + { + "grad_norm": 0.302581787109375, + "learning_rate": 9.990360730398088e-05, + "loss": 0.0635, + "step": 3440 + }, + { + "grad_norm": 0.2354021519422531, + "learning_rate": 9.990154395672849e-05, + "loss": 0.0549, + "step": 3450 + }, + { + "grad_norm": 0.29993781447410583, + "learning_rate": 9.989945878084541e-05, + "loss": 0.0509, + "step": 3460 + }, + { + "grad_norm": 0.3091602623462677, + "learning_rate": 9.989735177724378e-05, + "loss": 0.0573, + "step": 3470 + }, + { + "grad_norm": 0.3266462981700897, + "learning_rate": 9.989522294684526e-05, + "loss": 0.0625, + "step": 3480 + }, + { + "grad_norm": 0.25673335790634155, + "learning_rate": 9.989307229058107e-05, + "loss": 0.0578, + "step": 3490 + }, + { + "grad_norm": 0.2643424868583679, + "learning_rate": 9.989089980939202e-05, + "loss": 0.047, + "step": 3500 + }, + { + "grad_norm": 0.19617415964603424, + "learning_rate": 9.988870550422835e-05, + "loss": 0.0539, + "step": 3510 + }, + { + "grad_norm": 0.39058825373649597, + "learning_rate": 9.988648937604999e-05, + "loss": 0.0575, + "step": 3520 + }, + { + "grad_norm": 0.4344352185726166, + "learning_rate": 9.988425142582632e-05, + "loss": 0.0502, + "step": 3530 + }, + { + "grad_norm": 0.29478776454925537, + "learning_rate": 9.98819916545363e-05, + "loss": 0.0517, + "step": 3540 + }, + { + "grad_norm": 0.34580639004707336, + "learning_rate": 9.987971006316844e-05, + "loss": 0.0568, + "step": 3550 + }, + { + "grad_norm": 0.30327799916267395, + "learning_rate": 9.987740665272077e-05, + "loss": 0.0502, + "step": 3560 + }, + { + "grad_norm": 0.312209814786911, + "learning_rate": 9.98750814242009e-05, + "loss": 0.0612, + "step": 3570 + }, + { + "grad_norm": 0.39961788058280945, + "learning_rate": 9.987273437862594e-05, + "loss": 0.0653, + "step": 3580 + }, + { + "grad_norm": 0.47028428316116333, + "learning_rate": 9.987036551702259e-05, + "loss": 0.0688, + "step": 3590 + }, + { + "grad_norm": 0.4095730781555176, + "learning_rate": 9.986797484042706e-05, + "loss": 0.0625, + "step": 3600 + }, + { + "grad_norm": 0.34831205010414124, + "learning_rate": 9.986556234988512e-05, + "loss": 0.0599, + "step": 3610 + }, + { + "grad_norm": 0.318075567483902, + "learning_rate": 9.986312804645205e-05, + "loss": 0.0525, + "step": 3620 + }, + { + "grad_norm": 1.0986396074295044, + "learning_rate": 9.986067193119273e-05, + "loss": 0.0527, + "step": 3630 + }, + { + "grad_norm": 0.35214728116989136, + "learning_rate": 9.985819400518153e-05, + "loss": 0.0552, + "step": 3640 + }, + { + "grad_norm": 0.36618295311927795, + "learning_rate": 9.985569426950239e-05, + "loss": 0.0585, + "step": 3650 + }, + { + "grad_norm": 1.2941402196884155, + "learning_rate": 9.985317272524876e-05, + "loss": 0.0556, + "step": 3660 + }, + { + "grad_norm": 0.5228000283241272, + "learning_rate": 9.985062937352366e-05, + "loss": 0.0655, + "step": 3670 + }, + { + "grad_norm": 0.31802839040756226, + "learning_rate": 9.984806421543966e-05, + "loss": 0.0587, + "step": 3680 + }, + { + "grad_norm": 0.19141189754009247, + "learning_rate": 9.984547725211881e-05, + "loss": 0.0524, + "step": 3690 + }, + { + "grad_norm": 0.2859116792678833, + "learning_rate": 9.984286848469276e-05, + "loss": 0.0517, + "step": 3700 + }, + { + "grad_norm": 0.20641356706619263, + "learning_rate": 9.984023791430266e-05, + "loss": 0.0536, + "step": 3710 + }, + { + "grad_norm": 0.17230287194252014, + "learning_rate": 9.983758554209924e-05, + "loss": 0.0528, + "step": 3720 + }, + { + "grad_norm": 0.27197426557540894, + "learning_rate": 9.983491136924268e-05, + "loss": 0.0472, + "step": 3730 + }, + { + "grad_norm": 0.21751855313777924, + "learning_rate": 9.983221539690282e-05, + "loss": 0.0509, + "step": 3740 + }, + { + "grad_norm": 0.28505557775497437, + "learning_rate": 9.982949762625892e-05, + "loss": 0.0627, + "step": 3750 + }, + { + "grad_norm": 0.4560917019844055, + "learning_rate": 9.982675805849986e-05, + "loss": 0.0492, + "step": 3760 + }, + { + "grad_norm": 0.42408081889152527, + "learning_rate": 9.982399669482399e-05, + "loss": 0.0605, + "step": 3770 + }, + { + "grad_norm": 0.35707083344459534, + "learning_rate": 9.982121353643924e-05, + "loss": 0.0573, + "step": 3780 + }, + { + "grad_norm": 0.3529580235481262, + "learning_rate": 9.981840858456306e-05, + "loss": 0.0533, + "step": 3790 + }, + { + "grad_norm": 0.3122136890888214, + "learning_rate": 9.981558184042243e-05, + "loss": 0.0668, + "step": 3800 + }, + { + "grad_norm": 0.1438124179840088, + "learning_rate": 9.981273330525387e-05, + "loss": 0.0495, + "step": 3810 + }, + { + "grad_norm": 0.3667554557323456, + "learning_rate": 9.980986298030341e-05, + "loss": 0.0548, + "step": 3820 + }, + { + "grad_norm": 1.043173909187317, + "learning_rate": 9.980697086682662e-05, + "loss": 0.0523, + "step": 3830 + }, + { + "grad_norm": 0.3902455270290375, + "learning_rate": 9.980405696608866e-05, + "loss": 0.0501, + "step": 3840 + }, + { + "grad_norm": 0.2659913897514343, + "learning_rate": 9.980112127936409e-05, + "loss": 0.0583, + "step": 3850 + }, + { + "grad_norm": 0.35857316851615906, + "learning_rate": 9.979816380793717e-05, + "loss": 0.0508, + "step": 3860 + }, + { + "grad_norm": 0.245192289352417, + "learning_rate": 9.979518455310151e-05, + "loss": 0.0496, + "step": 3870 + }, + { + "grad_norm": 0.5779495239257812, + "learning_rate": 9.97921835161604e-05, + "loss": 0.0488, + "step": 3880 + }, + { + "grad_norm": 0.35414496064186096, + "learning_rate": 9.978916069842656e-05, + "loss": 0.0576, + "step": 3890 + }, + { + "grad_norm": 0.4379914700984955, + "learning_rate": 9.97861161012223e-05, + "loss": 0.0546, + "step": 3900 + }, + { + "grad_norm": 0.8089884519577026, + "learning_rate": 9.978304972587942e-05, + "loss": 0.0553, + "step": 3910 + }, + { + "grad_norm": 0.3096292316913605, + "learning_rate": 9.977996157373925e-05, + "loss": 0.0567, + "step": 3920 + }, + { + "grad_norm": 0.24275760352611542, + "learning_rate": 9.977685164615265e-05, + "loss": 0.0581, + "step": 3930 + }, + { + "grad_norm": 0.36570072174072266, + "learning_rate": 9.977371994448002e-05, + "loss": 0.0615, + "step": 3940 + }, + { + "grad_norm": 0.4931832253932953, + "learning_rate": 9.977056647009127e-05, + "loss": 0.0541, + "step": 3950 + }, + { + "grad_norm": 0.2548185884952545, + "learning_rate": 9.976739122436582e-05, + "loss": 0.0618, + "step": 3960 + }, + { + "grad_norm": 0.38659340143203735, + "learning_rate": 9.976419420869265e-05, + "loss": 0.0552, + "step": 3970 + }, + { + "grad_norm": 0.2776116132736206, + "learning_rate": 9.976097542447025e-05, + "loss": 0.0563, + "step": 3980 + }, + { + "grad_norm": 0.3018476068973541, + "learning_rate": 9.97577348731066e-05, + "loss": 0.0531, + "step": 3990 + }, + { + "grad_norm": 0.3590131402015686, + "learning_rate": 9.975447255601927e-05, + "loss": 0.0501, + "step": 4000 + }, + { + "grad_norm": 0.24539364874362946, + "learning_rate": 9.975118847463525e-05, + "loss": 0.0465, + "step": 4010 + }, + { + "grad_norm": 0.3582766652107239, + "learning_rate": 9.974788263039114e-05, + "loss": 0.0531, + "step": 4020 + }, + { + "grad_norm": 0.28320595622062683, + "learning_rate": 9.974455502473303e-05, + "loss": 0.0624, + "step": 4030 + }, + { + "grad_norm": 0.5343993306159973, + "learning_rate": 9.974120565911652e-05, + "loss": 0.0505, + "step": 4040 + }, + { + "grad_norm": 0.3147805333137512, + "learning_rate": 9.973783453500674e-05, + "loss": 0.0571, + "step": 4050 + }, + { + "grad_norm": 0.9093114137649536, + "learning_rate": 9.973444165387835e-05, + "loss": 0.0595, + "step": 4060 + }, + { + "grad_norm": 1.652077317237854, + "learning_rate": 9.973102701721549e-05, + "loss": 0.0554, + "step": 4070 + }, + { + "grad_norm": 0.24093550443649292, + "learning_rate": 9.972759062651184e-05, + "loss": 0.061, + "step": 4080 + }, + { + "grad_norm": 0.24438557028770447, + "learning_rate": 9.972413248327059e-05, + "loss": 0.0677, + "step": 4090 + }, + { + "grad_norm": 0.3446384370326996, + "learning_rate": 9.972065258900447e-05, + "loss": 0.063, + "step": 4100 + }, + { + "grad_norm": 0.252673864364624, + "learning_rate": 9.971715094523569e-05, + "loss": 0.0692, + "step": 4110 + }, + { + "grad_norm": 0.40909671783447266, + "learning_rate": 9.971362755349598e-05, + "loss": 0.0665, + "step": 4120 + }, + { + "grad_norm": 0.41660276055336, + "learning_rate": 9.971008241532662e-05, + "loss": 0.0675, + "step": 4130 + }, + { + "grad_norm": 0.5259755849838257, + "learning_rate": 9.970651553227835e-05, + "loss": 0.0611, + "step": 4140 + }, + { + "grad_norm": 0.32138922810554504, + "learning_rate": 9.970292690591143e-05, + "loss": 0.0661, + "step": 4150 + }, + { + "grad_norm": 0.3288574516773224, + "learning_rate": 9.969931653779569e-05, + "loss": 0.0577, + "step": 4160 + }, + { + "grad_norm": 0.358282208442688, + "learning_rate": 9.969568442951038e-05, + "loss": 0.0567, + "step": 4170 + }, + { + "grad_norm": 0.2919079065322876, + "learning_rate": 9.969203058264436e-05, + "loss": 0.0536, + "step": 4180 + }, + { + "grad_norm": 0.39442670345306396, + "learning_rate": 9.96883549987959e-05, + "loss": 0.0567, + "step": 4190 + }, + { + "grad_norm": 0.2576538622379303, + "learning_rate": 9.968465767957287e-05, + "loss": 0.0603, + "step": 4200 + }, + { + "grad_norm": 0.21659307181835175, + "learning_rate": 9.968093862659256e-05, + "loss": 0.0631, + "step": 4210 + }, + { + "grad_norm": 0.2989773750305176, + "learning_rate": 9.967719784148182e-05, + "loss": 0.0662, + "step": 4220 + }, + { + "grad_norm": 0.32037344574928284, + "learning_rate": 9.967343532587702e-05, + "loss": 0.0739, + "step": 4230 + }, + { + "grad_norm": 0.6197174191474915, + "learning_rate": 9.966965108142399e-05, + "loss": 0.0658, + "step": 4240 + }, + { + "grad_norm": 0.47274309396743774, + "learning_rate": 9.96658451097781e-05, + "loss": 0.0642, + "step": 4250 + }, + { + "grad_norm": 0.2118692249059677, + "learning_rate": 9.966201741260419e-05, + "loss": 0.0645, + "step": 4260 + }, + { + "grad_norm": 0.3241516649723053, + "learning_rate": 9.965816799157665e-05, + "loss": 0.0632, + "step": 4270 + }, + { + "grad_norm": 0.40076926350593567, + "learning_rate": 9.965429684837935e-05, + "loss": 0.0544, + "step": 4280 + }, + { + "grad_norm": 0.3264576196670532, + "learning_rate": 9.965040398470562e-05, + "loss": 0.0603, + "step": 4290 + }, + { + "grad_norm": 0.3481746315956116, + "learning_rate": 9.964648940225838e-05, + "loss": 0.0554, + "step": 4300 + }, + { + "grad_norm": 0.49234458804130554, + "learning_rate": 9.964255310274997e-05, + "loss": 0.0606, + "step": 4310 + }, + { + "grad_norm": 0.4657320976257324, + "learning_rate": 9.963859508790228e-05, + "loss": 0.0781, + "step": 4320 + }, + { + "grad_norm": 0.6024658679962158, + "learning_rate": 9.963461535944664e-05, + "loss": 0.0596, + "step": 4330 + }, + { + "grad_norm": 0.28868165612220764, + "learning_rate": 9.963061391912399e-05, + "loss": 0.0496, + "step": 4340 + }, + { + "grad_norm": 0.20309802889823914, + "learning_rate": 9.962659076868463e-05, + "loss": 0.0463, + "step": 4350 + }, + { + "grad_norm": 0.20622700452804565, + "learning_rate": 9.962254590988846e-05, + "loss": 0.0527, + "step": 4360 + }, + { + "grad_norm": 0.5116778016090393, + "learning_rate": 9.961847934450481e-05, + "loss": 0.0563, + "step": 4370 + }, + { + "grad_norm": 0.39127907156944275, + "learning_rate": 9.961439107431257e-05, + "loss": 0.0579, + "step": 4380 + }, + { + "grad_norm": 0.38995954394340515, + "learning_rate": 9.961028110110006e-05, + "loss": 0.0555, + "step": 4390 + }, + { + "grad_norm": 0.30136775970458984, + "learning_rate": 9.960614942666513e-05, + "loss": 0.0634, + "step": 4400 + }, + { + "grad_norm": 0.3027201294898987, + "learning_rate": 9.960199605281511e-05, + "loss": 0.0561, + "step": 4410 + }, + { + "grad_norm": 0.4213874042034149, + "learning_rate": 9.959782098136683e-05, + "loss": 0.0653, + "step": 4420 + }, + { + "grad_norm": 0.2917952239513397, + "learning_rate": 9.959362421414662e-05, + "loss": 0.0601, + "step": 4430 + }, + { + "grad_norm": 0.2129185050725937, + "learning_rate": 9.958940575299027e-05, + "loss": 0.0594, + "step": 4440 + }, + { + "grad_norm": 0.3271077275276184, + "learning_rate": 9.95851655997431e-05, + "loss": 0.0635, + "step": 4450 + }, + { + "grad_norm": 0.26671102643013, + "learning_rate": 9.958090375625986e-05, + "loss": 0.0568, + "step": 4460 + }, + { + "grad_norm": 0.3331860303878784, + "learning_rate": 9.957662022440486e-05, + "loss": 0.0543, + "step": 4470 + }, + { + "grad_norm": 0.2610504925251007, + "learning_rate": 9.957231500605187e-05, + "loss": 0.0572, + "step": 4480 + }, + { + "grad_norm": 0.3193032443523407, + "learning_rate": 9.95679881030841e-05, + "loss": 0.0567, + "step": 4490 + }, + { + "grad_norm": 0.2753559648990631, + "learning_rate": 9.95636395173943e-05, + "loss": 0.0554, + "step": 4500 + }, + { + "grad_norm": 0.2880411446094513, + "learning_rate": 9.95592692508847e-05, + "loss": 0.0756, + "step": 4510 + }, + { + "grad_norm": 0.24522718787193298, + "learning_rate": 9.9554877305467e-05, + "loss": 0.0576, + "step": 4520 + }, + { + "grad_norm": 0.20306415855884552, + "learning_rate": 9.955046368306237e-05, + "loss": 0.0545, + "step": 4530 + }, + { + "grad_norm": 0.4749528169631958, + "learning_rate": 9.954602838560153e-05, + "loss": 0.0627, + "step": 4540 + }, + { + "grad_norm": 0.44332918524742126, + "learning_rate": 9.954157141502456e-05, + "loss": 0.0646, + "step": 4550 + }, + { + "grad_norm": 0.353924959897995, + "learning_rate": 9.953709277328112e-05, + "loss": 0.0609, + "step": 4560 + }, + { + "grad_norm": 0.23054833710193634, + "learning_rate": 9.953259246233032e-05, + "loss": 0.0558, + "step": 4570 + }, + { + "grad_norm": 0.26510316133499146, + "learning_rate": 9.952807048414077e-05, + "loss": 0.0523, + "step": 4580 + }, + { + "grad_norm": 0.25476324558258057, + "learning_rate": 9.95235268406905e-05, + "loss": 0.0576, + "step": 4590 + }, + { + "grad_norm": 0.6492944955825806, + "learning_rate": 9.951896153396708e-05, + "loss": 0.0549, + "step": 4600 + }, + { + "grad_norm": 0.2787790596485138, + "learning_rate": 9.95143745659675e-05, + "loss": 0.059, + "step": 4610 + }, + { + "grad_norm": 0.21355631947517395, + "learning_rate": 9.95097659386983e-05, + "loss": 0.0493, + "step": 4620 + }, + { + "grad_norm": 0.34450557827949524, + "learning_rate": 9.950513565417542e-05, + "loss": 0.0525, + "step": 4630 + }, + { + "grad_norm": 0.2980755865573883, + "learning_rate": 9.95004837144243e-05, + "loss": 0.0572, + "step": 4640 + }, + { + "grad_norm": 0.44951826333999634, + "learning_rate": 9.949581012147988e-05, + "loss": 0.0492, + "step": 4650 + }, + { + "grad_norm": 0.3013516664505005, + "learning_rate": 9.949111487738653e-05, + "loss": 0.0604, + "step": 4660 + }, + { + "grad_norm": 0.34380707144737244, + "learning_rate": 9.948639798419813e-05, + "loss": 0.0566, + "step": 4670 + }, + { + "grad_norm": 0.3113146424293518, + "learning_rate": 9.948165944397799e-05, + "loss": 0.0524, + "step": 4680 + }, + { + "grad_norm": 0.30695992708206177, + "learning_rate": 9.94768992587989e-05, + "loss": 0.0561, + "step": 4690 + }, + { + "grad_norm": 0.19974786043167114, + "learning_rate": 9.947211743074313e-05, + "loss": 0.0511, + "step": 4700 + }, + { + "grad_norm": 0.27647677063941956, + "learning_rate": 9.946731396190246e-05, + "loss": 0.054, + "step": 4710 + }, + { + "grad_norm": 0.35592901706695557, + "learning_rate": 9.946248885437803e-05, + "loss": 0.048, + "step": 4720 + }, + { + "grad_norm": 0.24948430061340332, + "learning_rate": 9.945764211028053e-05, + "loss": 0.0596, + "step": 4730 + }, + { + "grad_norm": 0.2906743586063385, + "learning_rate": 9.94527737317301e-05, + "loss": 0.0481, + "step": 4740 + }, + { + "grad_norm": 0.23265224695205688, + "learning_rate": 9.944788372085631e-05, + "loss": 0.0479, + "step": 4750 + }, + { + "grad_norm": 0.3830947279930115, + "learning_rate": 9.944297207979825e-05, + "loss": 0.051, + "step": 4760 + }, + { + "grad_norm": 0.2622953951358795, + "learning_rate": 9.943803881070441e-05, + "loss": 0.0558, + "step": 4770 + }, + { + "grad_norm": 0.43968266248703003, + "learning_rate": 9.943308391573278e-05, + "loss": 0.0573, + "step": 4780 + }, + { + "grad_norm": 0.34006908535957336, + "learning_rate": 9.942810739705079e-05, + "loss": 0.0476, + "step": 4790 + }, + { + "grad_norm": 0.2849588990211487, + "learning_rate": 9.942310925683538e-05, + "loss": 0.0468, + "step": 4800 + }, + { + "grad_norm": 0.3808932304382324, + "learning_rate": 9.941808949727285e-05, + "loss": 0.0518, + "step": 4810 + }, + { + "grad_norm": 0.3541061282157898, + "learning_rate": 9.941304812055903e-05, + "loss": 0.0535, + "step": 4820 + }, + { + "grad_norm": 0.3605745732784271, + "learning_rate": 9.940798512889921e-05, + "loss": 0.0472, + "step": 4830 + }, + { + "grad_norm": 0.3603658974170685, + "learning_rate": 9.94029005245081e-05, + "loss": 0.0499, + "step": 4840 + }, + { + "grad_norm": 0.3298913240432739, + "learning_rate": 9.939779430960988e-05, + "loss": 0.0484, + "step": 4850 + }, + { + "grad_norm": 0.3747410178184509, + "learning_rate": 9.939266648643817e-05, + "loss": 0.0462, + "step": 4860 + }, + { + "grad_norm": 0.2815413773059845, + "learning_rate": 9.938751705723607e-05, + "loss": 0.0497, + "step": 4870 + }, + { + "grad_norm": 0.299408495426178, + "learning_rate": 9.938234602425613e-05, + "loss": 0.0467, + "step": 4880 + }, + { + "grad_norm": 0.32164788246154785, + "learning_rate": 9.93771533897603e-05, + "loss": 0.056, + "step": 4890 + }, + { + "grad_norm": 0.23385289311408997, + "learning_rate": 9.937193915602004e-05, + "loss": 0.0424, + "step": 4900 + }, + { + "grad_norm": 0.2980406582355499, + "learning_rate": 9.936670332531621e-05, + "loss": 0.0498, + "step": 4910 + }, + { + "grad_norm": 0.3253431022167206, + "learning_rate": 9.936144589993916e-05, + "loss": 0.0511, + "step": 4920 + }, + { + "grad_norm": 0.19928674399852753, + "learning_rate": 9.935616688218867e-05, + "loss": 0.0508, + "step": 4930 + }, + { + "grad_norm": 0.24933956563472748, + "learning_rate": 9.935086627437395e-05, + "loss": 0.0482, + "step": 4940 + }, + { + "grad_norm": 0.6691576242446899, + "learning_rate": 9.934554407881366e-05, + "loss": 0.0547, + "step": 4950 + }, + { + "grad_norm": 0.4286210834980011, + "learning_rate": 9.934020029783593e-05, + "loss": 0.0505, + "step": 4960 + }, + { + "grad_norm": 0.3404936194419861, + "learning_rate": 9.933483493377829e-05, + "loss": 0.0547, + "step": 4970 + }, + { + "grad_norm": 0.45622265338897705, + "learning_rate": 9.932944798898774e-05, + "loss": 0.0531, + "step": 4980 + }, + { + "grad_norm": 0.2808411121368408, + "learning_rate": 9.932403946582072e-05, + "loss": 0.0489, + "step": 4990 + }, + { + "grad_norm": 0.30106040835380554, + "learning_rate": 9.93186093666431e-05, + "loss": 0.0529, + "step": 5000 + }, + { + "grad_norm": 0.2851671278476715, + "learning_rate": 9.931315769383018e-05, + "loss": 0.0543, + "step": 5010 + }, + { + "grad_norm": 0.299526184797287, + "learning_rate": 9.930768444976672e-05, + "loss": 0.0527, + "step": 5020 + }, + { + "grad_norm": 0.3189558684825897, + "learning_rate": 9.93021896368469e-05, + "loss": 0.0533, + "step": 5030 + }, + { + "grad_norm": 0.21386253833770752, + "learning_rate": 9.929667325747431e-05, + "loss": 0.0582, + "step": 5040 + }, + { + "grad_norm": 0.27085790038108826, + "learning_rate": 9.929113531406205e-05, + "loss": 0.0616, + "step": 5050 + }, + { + "grad_norm": 0.26358920335769653, + "learning_rate": 9.928557580903257e-05, + "loss": 0.047, + "step": 5060 + }, + { + "grad_norm": 0.2666590213775635, + "learning_rate": 9.927999474481779e-05, + "loss": 0.0641, + "step": 5070 + }, + { + "grad_norm": 0.29679474234580994, + "learning_rate": 9.927439212385907e-05, + "loss": 0.0649, + "step": 5080 + }, + { + "grad_norm": 0.3643966615200043, + "learning_rate": 9.926876794860718e-05, + "loss": 0.0525, + "step": 5090 + }, + { + "grad_norm": 0.3897665739059448, + "learning_rate": 9.926312222152235e-05, + "loss": 0.052, + "step": 5100 + }, + { + "grad_norm": 0.351213663816452, + "learning_rate": 9.925745494507414e-05, + "loss": 0.0566, + "step": 5110 + }, + { + "grad_norm": 0.4341650903224945, + "learning_rate": 9.925176612174169e-05, + "loss": 0.0625, + "step": 5120 + }, + { + "grad_norm": 0.2726602554321289, + "learning_rate": 9.924605575401346e-05, + "loss": 0.0609, + "step": 5130 + }, + { + "grad_norm": 0.30280399322509766, + "learning_rate": 9.924032384438733e-05, + "loss": 0.056, + "step": 5140 + }, + { + "grad_norm": 0.3153168559074402, + "learning_rate": 9.923457039537066e-05, + "loss": 0.0497, + "step": 5150 + }, + { + "grad_norm": 0.21572822332382202, + "learning_rate": 9.92287954094802e-05, + "loss": 0.064, + "step": 5160 + }, + { + "grad_norm": 0.23348617553710938, + "learning_rate": 9.922299888924212e-05, + "loss": 0.0531, + "step": 5170 + }, + { + "grad_norm": 0.14465846121311188, + "learning_rate": 9.921718083719203e-05, + "loss": 0.0519, + "step": 5180 + }, + { + "grad_norm": 0.34029722213745117, + "learning_rate": 9.921134125587491e-05, + "loss": 0.0643, + "step": 5190 + }, + { + "grad_norm": 0.36133241653442383, + "learning_rate": 9.920548014784523e-05, + "loss": 0.0448, + "step": 5200 + }, + { + "grad_norm": 0.36014851927757263, + "learning_rate": 9.919959751566681e-05, + "loss": 0.0527, + "step": 5210 + }, + { + "grad_norm": 0.22264505922794342, + "learning_rate": 9.919369336191291e-05, + "loss": 0.0537, + "step": 5220 + }, + { + "grad_norm": 0.20799975097179413, + "learning_rate": 9.918776768916625e-05, + "loss": 0.0475, + "step": 5230 + }, + { + "grad_norm": 0.30014392733573914, + "learning_rate": 9.918182050001888e-05, + "loss": 0.0484, + "step": 5240 + }, + { + "grad_norm": 0.3485787510871887, + "learning_rate": 9.91758517970723e-05, + "loss": 0.059, + "step": 5250 + }, + { + "grad_norm": 0.31363630294799805, + "learning_rate": 9.916986158293744e-05, + "loss": 0.0513, + "step": 5260 + }, + { + "grad_norm": 0.2687159776687622, + "learning_rate": 9.916384986023463e-05, + "loss": 0.0535, + "step": 5270 + }, + { + "grad_norm": 0.5354437232017517, + "learning_rate": 9.915781663159359e-05, + "loss": 0.0627, + "step": 5280 + }, + { + "grad_norm": 0.3529573380947113, + "learning_rate": 9.915176189965346e-05, + "loss": 0.0545, + "step": 5290 + }, + { + "grad_norm": 0.22913996875286102, + "learning_rate": 9.914568566706279e-05, + "loss": 0.0526, + "step": 5300 + }, + { + "grad_norm": 0.23836082220077515, + "learning_rate": 9.913958793647953e-05, + "loss": 0.049, + "step": 5310 + }, + { + "grad_norm": 0.24388466775417328, + "learning_rate": 9.913346871057102e-05, + "loss": 0.0541, + "step": 5320 + }, + { + "grad_norm": 0.43231871724128723, + "learning_rate": 9.912732799201403e-05, + "loss": 0.0558, + "step": 5330 + }, + { + "grad_norm": 0.24662280082702637, + "learning_rate": 9.912116578349474e-05, + "loss": 0.0568, + "step": 5340 + }, + { + "grad_norm": 0.18286211788654327, + "learning_rate": 9.911498208770867e-05, + "loss": 0.0458, + "step": 5350 + }, + { + "grad_norm": 0.20792986452579498, + "learning_rate": 9.910877690736078e-05, + "loss": 0.0518, + "step": 5360 + }, + { + "grad_norm": 0.3829275369644165, + "learning_rate": 9.910255024516546e-05, + "loss": 0.0495, + "step": 5370 + }, + { + "grad_norm": 0.21728557348251343, + "learning_rate": 9.909630210384644e-05, + "loss": 0.0518, + "step": 5380 + }, + { + "grad_norm": 0.2345687448978424, + "learning_rate": 9.909003248613688e-05, + "loss": 0.0477, + "step": 5390 + }, + { + "grad_norm": 0.21150833368301392, + "learning_rate": 9.908374139477934e-05, + "loss": 0.0552, + "step": 5400 + }, + { + "grad_norm": 0.2656225264072418, + "learning_rate": 9.907742883252571e-05, + "loss": 0.053, + "step": 5410 + }, + { + "grad_norm": 0.2412857711315155, + "learning_rate": 9.907109480213736e-05, + "loss": 0.0513, + "step": 5420 + }, + { + "grad_norm": 0.32271242141723633, + "learning_rate": 9.906473930638498e-05, + "loss": 0.0485, + "step": 5430 + }, + { + "grad_norm": 0.28578418493270874, + "learning_rate": 9.90583623480487e-05, + "loss": 0.0581, + "step": 5440 + }, + { + "grad_norm": 0.3096124529838562, + "learning_rate": 9.905196392991802e-05, + "loss": 0.0462, + "step": 5450 + }, + { + "grad_norm": 0.20730197429656982, + "learning_rate": 9.904554405479183e-05, + "loss": 0.0491, + "step": 5460 + }, + { + "grad_norm": 0.32392096519470215, + "learning_rate": 9.903910272547838e-05, + "loss": 0.0479, + "step": 5470 + }, + { + "grad_norm": 0.26954641938209534, + "learning_rate": 9.903263994479533e-05, + "loss": 0.0483, + "step": 5480 + }, + { + "grad_norm": 0.26622581481933594, + "learning_rate": 9.902615571556972e-05, + "loss": 0.0485, + "step": 5490 + }, + { + "grad_norm": 0.2812046706676483, + "learning_rate": 9.9019650040638e-05, + "loss": 0.0527, + "step": 5500 + }, + { + "grad_norm": 0.2909301221370697, + "learning_rate": 9.901312292284594e-05, + "loss": 0.0509, + "step": 5510 + }, + { + "grad_norm": 0.23271426558494568, + "learning_rate": 9.900657436504875e-05, + "loss": 0.0496, + "step": 5520 + }, + { + "grad_norm": 0.4343194365501404, + "learning_rate": 9.900000437011097e-05, + "loss": 0.0557, + "step": 5530 + }, + { + "grad_norm": 0.15935465693473816, + "learning_rate": 9.899341294090653e-05, + "loss": 0.0478, + "step": 5540 + }, + { + "grad_norm": 0.21886824071407318, + "learning_rate": 9.898680008031877e-05, + "loss": 0.0471, + "step": 5550 + }, + { + "grad_norm": 0.3611968457698822, + "learning_rate": 9.898016579124037e-05, + "loss": 0.051, + "step": 5560 + }, + { + "grad_norm": 0.357333779335022, + "learning_rate": 9.89735100765734e-05, + "loss": 0.05, + "step": 5570 + }, + { + "grad_norm": 0.31987226009368896, + "learning_rate": 9.89668329392293e-05, + "loss": 0.0567, + "step": 5580 + }, + { + "grad_norm": 0.34940192103385925, + "learning_rate": 9.896013438212885e-05, + "loss": 0.0544, + "step": 5590 + }, + { + "grad_norm": 0.3193245530128479, + "learning_rate": 9.895341440820225e-05, + "loss": 0.0594, + "step": 5600 + }, + { + "grad_norm": 0.21757572889328003, + "learning_rate": 9.894667302038902e-05, + "loss": 0.0573, + "step": 5610 + }, + { + "grad_norm": 0.3589266240596771, + "learning_rate": 9.893991022163812e-05, + "loss": 0.0577, + "step": 5620 + }, + { + "grad_norm": 0.3821682035923004, + "learning_rate": 9.893312601490779e-05, + "loss": 0.0495, + "step": 5630 + }, + { + "grad_norm": 0.26858869194984436, + "learning_rate": 9.892632040316568e-05, + "loss": 0.0572, + "step": 5640 + }, + { + "grad_norm": 0.4632846415042877, + "learning_rate": 9.891949338938877e-05, + "loss": 0.0673, + "step": 5650 + }, + { + "grad_norm": 0.2718506157398224, + "learning_rate": 9.89126449765635e-05, + "loss": 0.0568, + "step": 5660 + }, + { + "grad_norm": 0.39689314365386963, + "learning_rate": 9.890577516768551e-05, + "loss": 0.0534, + "step": 5670 + }, + { + "grad_norm": 0.2447853982448578, + "learning_rate": 9.889888396575995e-05, + "loss": 0.0603, + "step": 5680 + }, + { + "grad_norm": 0.22035516798496246, + "learning_rate": 9.889197137380125e-05, + "loss": 0.0555, + "step": 5690 + }, + { + "grad_norm": 0.20806258916854858, + "learning_rate": 9.888503739483321e-05, + "loss": 0.0515, + "step": 5700 + }, + { + "grad_norm": 0.260668009519577, + "learning_rate": 9.887808203188897e-05, + "loss": 0.0573, + "step": 5710 + }, + { + "grad_norm": 0.39202260971069336, + "learning_rate": 9.887110528801108e-05, + "loss": 0.0547, + "step": 5720 + }, + { + "grad_norm": 0.553096354007721, + "learning_rate": 9.886410716625137e-05, + "loss": 0.0487, + "step": 5730 + }, + { + "grad_norm": 0.4173396825790405, + "learning_rate": 9.885708766967108e-05, + "loss": 0.0581, + "step": 5740 + }, + { + "grad_norm": 0.2851225435733795, + "learning_rate": 9.885004680134076e-05, + "loss": 0.0525, + "step": 5750 + }, + { + "grad_norm": 0.28869113326072693, + "learning_rate": 9.884298456434033e-05, + "loss": 0.0573, + "step": 5760 + }, + { + "grad_norm": 0.6250273585319519, + "learning_rate": 9.883590096175905e-05, + "loss": 0.0547, + "step": 5770 + }, + { + "grad_norm": 0.35810384154319763, + "learning_rate": 9.882879599669554e-05, + "loss": 0.062, + "step": 5780 + }, + { + "grad_norm": 0.24264571070671082, + "learning_rate": 9.882166967225775e-05, + "loss": 0.0527, + "step": 5790 + }, + { + "grad_norm": 0.26370278000831604, + "learning_rate": 9.881452199156296e-05, + "loss": 0.0616, + "step": 5800 + }, + { + "grad_norm": 0.2879207730293274, + "learning_rate": 9.880735295773782e-05, + "loss": 0.0507, + "step": 5810 + }, + { + "grad_norm": 0.3015146851539612, + "learning_rate": 9.88001625739183e-05, + "loss": 0.0551, + "step": 5820 + }, + { + "grad_norm": 0.3228124976158142, + "learning_rate": 9.879295084324976e-05, + "loss": 0.0553, + "step": 5830 + }, + { + "grad_norm": 0.3922732174396515, + "learning_rate": 9.87857177688868e-05, + "loss": 0.0566, + "step": 5840 + }, + { + "grad_norm": 0.6140612959861755, + "learning_rate": 9.877846335399343e-05, + "loss": 0.0535, + "step": 5850 + }, + { + "grad_norm": 0.2479388415813446, + "learning_rate": 9.8771187601743e-05, + "loss": 0.057, + "step": 5860 + }, + { + "grad_norm": 0.21422210335731506, + "learning_rate": 9.876389051531813e-05, + "loss": 0.0543, + "step": 5870 + }, + { + "grad_norm": 0.2666412591934204, + "learning_rate": 9.875657209791088e-05, + "loss": 0.0586, + "step": 5880 + }, + { + "grad_norm": 0.16914033889770508, + "learning_rate": 9.874923235272248e-05, + "loss": 0.0503, + "step": 5890 + }, + { + "grad_norm": 0.2822473645210266, + "learning_rate": 9.874187128296369e-05, + "loss": 0.0646, + "step": 5900 + }, + { + "grad_norm": 0.29717889428138733, + "learning_rate": 9.873448889185439e-05, + "loss": 0.0593, + "step": 5910 + }, + { + "grad_norm": 0.26817578077316284, + "learning_rate": 9.872708518262397e-05, + "loss": 0.0553, + "step": 5920 + }, + { + "grad_norm": 0.30797094106674194, + "learning_rate": 9.871966015851101e-05, + "loss": 0.0516, + "step": 5930 + }, + { + "grad_norm": 0.2785377502441406, + "learning_rate": 9.87122138227635e-05, + "loss": 0.0566, + "step": 5940 + }, + { + "grad_norm": 0.3575267493724823, + "learning_rate": 9.87047461786387e-05, + "loss": 0.0542, + "step": 5950 + }, + { + "grad_norm": 0.3271638751029968, + "learning_rate": 9.869725722940323e-05, + "loss": 0.0524, + "step": 5960 + }, + { + "grad_norm": 0.34594056010246277, + "learning_rate": 9.868974697833299e-05, + "loss": 0.056, + "step": 5970 + }, + { + "grad_norm": 0.3843018114566803, + "learning_rate": 9.868221542871324e-05, + "loss": 0.0671, + "step": 5980 + }, + { + "grad_norm": 0.3426145613193512, + "learning_rate": 9.867466258383853e-05, + "loss": 0.0661, + "step": 5990 + }, + { + "grad_norm": 0.2671982944011688, + "learning_rate": 9.866708844701272e-05, + "loss": 0.0703, + "step": 6000 + }, + { + "grad_norm": 0.44629666209220886, + "learning_rate": 9.865949302154899e-05, + "loss": 0.0578, + "step": 6010 + }, + { + "grad_norm": 0.3208904266357422, + "learning_rate": 9.865187631076987e-05, + "loss": 0.06, + "step": 6020 + }, + { + "grad_norm": 0.507607638835907, + "learning_rate": 9.864423831800712e-05, + "loss": 0.058, + "step": 6030 + }, + { + "grad_norm": 0.2494739294052124, + "learning_rate": 9.863657904660191e-05, + "loss": 0.0558, + "step": 6040 + }, + { + "grad_norm": 0.2869704067707062, + "learning_rate": 9.862889849990462e-05, + "loss": 0.0605, + "step": 6050 + }, + { + "grad_norm": 0.3025570809841156, + "learning_rate": 9.8621196681275e-05, + "loss": 0.0563, + "step": 6060 + }, + { + "grad_norm": 0.32025548815727234, + "learning_rate": 9.861347359408211e-05, + "loss": 0.0502, + "step": 6070 + }, + { + "grad_norm": 0.21850061416625977, + "learning_rate": 9.860572924170426e-05, + "loss": 0.0589, + "step": 6080 + }, + { + "grad_norm": 0.3711686134338379, + "learning_rate": 9.85979636275291e-05, + "loss": 0.0539, + "step": 6090 + }, + { + "grad_norm": 0.2814316749572754, + "learning_rate": 9.859017675495357e-05, + "loss": 0.0616, + "step": 6100 + }, + { + "grad_norm": 0.25092363357543945, + "learning_rate": 9.858236862738392e-05, + "loss": 0.0528, + "step": 6110 + }, + { + "grad_norm": 0.3003692924976349, + "learning_rate": 9.85745392482357e-05, + "loss": 0.0527, + "step": 6120 + }, + { + "grad_norm": 0.5318816304206848, + "learning_rate": 9.856668862093372e-05, + "loss": 0.0493, + "step": 6130 + }, + { + "grad_norm": 0.399995356798172, + "learning_rate": 9.855881674891214e-05, + "loss": 0.0536, + "step": 6140 + }, + { + "grad_norm": 1.2692540884017944, + "learning_rate": 9.855092363561437e-05, + "loss": 0.0617, + "step": 6150 + }, + { + "grad_norm": 0.4828345477581024, + "learning_rate": 9.854300928449312e-05, + "loss": 0.072, + "step": 6160 + }, + { + "grad_norm": 0.33370307087898254, + "learning_rate": 9.85350736990104e-05, + "loss": 0.0652, + "step": 6170 + }, + { + "grad_norm": 0.3561568558216095, + "learning_rate": 9.852711688263752e-05, + "loss": 0.055, + "step": 6180 + }, + { + "grad_norm": 0.26626092195510864, + "learning_rate": 9.851913883885503e-05, + "loss": 0.0626, + "step": 6190 + }, + { + "grad_norm": 0.23566456139087677, + "learning_rate": 9.851113957115283e-05, + "loss": 0.0538, + "step": 6200 + }, + { + "grad_norm": 0.35528701543807983, + "learning_rate": 9.850311908303005e-05, + "loss": 0.0557, + "step": 6210 + }, + { + "grad_norm": 0.33952996134757996, + "learning_rate": 9.849507737799514e-05, + "loss": 0.0506, + "step": 6220 + }, + { + "grad_norm": 0.1993565708398819, + "learning_rate": 9.84870144595658e-05, + "loss": 0.0529, + "step": 6230 + }, + { + "grad_norm": 0.3411701023578644, + "learning_rate": 9.847893033126903e-05, + "loss": 0.0522, + "step": 6240 + }, + { + "grad_norm": 0.2291853278875351, + "learning_rate": 9.847082499664111e-05, + "loss": 0.0584, + "step": 6250 + }, + { + "grad_norm": 0.2774844467639923, + "learning_rate": 9.846269845922758e-05, + "loss": 0.0499, + "step": 6260 + }, + { + "grad_norm": 0.2034483551979065, + "learning_rate": 9.845455072258326e-05, + "loss": 0.0456, + "step": 6270 + }, + { + "grad_norm": 0.2315664291381836, + "learning_rate": 9.844638179027226e-05, + "loss": 0.0535, + "step": 6280 + }, + { + "grad_norm": 0.252745121717453, + "learning_rate": 9.843819166586795e-05, + "loss": 0.0463, + "step": 6290 + }, + { + "grad_norm": 0.34486016631126404, + "learning_rate": 9.842998035295294e-05, + "loss": 0.0506, + "step": 6300 + }, + { + "grad_norm": 0.4860289394855499, + "learning_rate": 9.842174785511918e-05, + "loss": 0.0494, + "step": 6310 + }, + { + "grad_norm": 0.28997802734375, + "learning_rate": 9.841349417596779e-05, + "loss": 0.0551, + "step": 6320 + }, + { + "grad_norm": 0.6759776473045349, + "learning_rate": 9.840521931910926e-05, + "loss": 0.0519, + "step": 6330 + }, + { + "grad_norm": 0.4370477497577667, + "learning_rate": 9.839692328816327e-05, + "loss": 0.0684, + "step": 6340 + }, + { + "grad_norm": 0.2716401517391205, + "learning_rate": 9.838860608675879e-05, + "loss": 0.0555, + "step": 6350 + }, + { + "grad_norm": 0.3338410556316376, + "learning_rate": 9.838026771853406e-05, + "loss": 0.0592, + "step": 6360 + }, + { + "grad_norm": 0.2383304089307785, + "learning_rate": 9.837190818713655e-05, + "loss": 0.0516, + "step": 6370 + }, + { + "grad_norm": 0.6331921815872192, + "learning_rate": 9.836352749622299e-05, + "loss": 0.0508, + "step": 6380 + }, + { + "grad_norm": 0.355459600687027, + "learning_rate": 9.835512564945941e-05, + "loss": 0.0518, + "step": 6390 + }, + { + "grad_norm": 0.24321119487285614, + "learning_rate": 9.834670265052104e-05, + "loss": 0.0567, + "step": 6400 + }, + { + "grad_norm": 0.44104352593421936, + "learning_rate": 9.833825850309239e-05, + "loss": 0.0396, + "step": 6410 + }, + { + "grad_norm": 0.3218808174133301, + "learning_rate": 9.832979321086723e-05, + "loss": 0.0503, + "step": 6420 + }, + { + "grad_norm": 0.2789342701435089, + "learning_rate": 9.832130677754854e-05, + "loss": 0.0469, + "step": 6430 + }, + { + "grad_norm": 0.24555006623268127, + "learning_rate": 9.831279920684861e-05, + "loss": 0.0515, + "step": 6440 + }, + { + "grad_norm": 0.2657032012939453, + "learning_rate": 9.830427050248891e-05, + "loss": 0.0456, + "step": 6450 + }, + { + "grad_norm": 0.24394527077674866, + "learning_rate": 9.829572066820021e-05, + "loss": 0.049, + "step": 6460 + }, + { + "grad_norm": 0.2874397039413452, + "learning_rate": 9.828714970772247e-05, + "loss": 0.0484, + "step": 6470 + }, + { + "grad_norm": 0.3289159834384918, + "learning_rate": 9.827855762480493e-05, + "loss": 0.048, + "step": 6480 + }, + { + "grad_norm": 0.29053834080696106, + "learning_rate": 9.826994442320608e-05, + "loss": 0.0524, + "step": 6490 + }, + { + "grad_norm": 0.1769585907459259, + "learning_rate": 9.82613101066936e-05, + "loss": 0.0459, + "step": 6500 + }, + { + "grad_norm": 0.29631030559539795, + "learning_rate": 9.825265467904445e-05, + "loss": 0.0449, + "step": 6510 + }, + { + "grad_norm": 0.21309903264045715, + "learning_rate": 9.82439781440448e-05, + "loss": 0.0529, + "step": 6520 + }, + { + "grad_norm": 0.22263571619987488, + "learning_rate": 9.823528050549006e-05, + "loss": 0.0485, + "step": 6530 + }, + { + "grad_norm": 0.26140663027763367, + "learning_rate": 9.822656176718487e-05, + "loss": 0.0596, + "step": 6540 + }, + { + "grad_norm": 0.28320685029029846, + "learning_rate": 9.82178219329431e-05, + "loss": 0.0485, + "step": 6550 + }, + { + "grad_norm": 0.22510777413845062, + "learning_rate": 9.820906100658789e-05, + "loss": 0.0535, + "step": 6560 + }, + { + "grad_norm": 0.3903724253177643, + "learning_rate": 9.82002789919515e-05, + "loss": 0.0583, + "step": 6570 + }, + { + "grad_norm": 0.26812875270843506, + "learning_rate": 9.819147589287554e-05, + "loss": 0.049, + "step": 6580 + }, + { + "grad_norm": 0.2502971291542053, + "learning_rate": 9.818265171321074e-05, + "loss": 0.0637, + "step": 6590 + }, + { + "grad_norm": 0.4208245873451233, + "learning_rate": 9.817380645681714e-05, + "loss": 0.0643, + "step": 6600 + }, + { + "grad_norm": 0.22509707510471344, + "learning_rate": 9.816494012756392e-05, + "loss": 0.0564, + "step": 6610 + }, + { + "grad_norm": 0.8730741739273071, + "learning_rate": 9.815605272932953e-05, + "loss": 0.0548, + "step": 6620 + }, + { + "grad_norm": 0.3300253450870514, + "learning_rate": 9.814714426600162e-05, + "loss": 0.0644, + "step": 6630 + }, + { + "grad_norm": 0.2607601284980774, + "learning_rate": 9.813821474147704e-05, + "loss": 0.0511, + "step": 6640 + }, + { + "grad_norm": 0.21411798894405365, + "learning_rate": 9.812926415966189e-05, + "loss": 0.0641, + "step": 6650 + }, + { + "grad_norm": 0.2826286852359772, + "learning_rate": 9.812029252447143e-05, + "loss": 0.0558, + "step": 6660 + }, + { + "grad_norm": 0.28230929374694824, + "learning_rate": 9.811129983983018e-05, + "loss": 0.0535, + "step": 6670 + }, + { + "grad_norm": 0.2508173882961273, + "learning_rate": 9.810228610967185e-05, + "loss": 0.0589, + "step": 6680 + }, + { + "grad_norm": 0.3340100646018982, + "learning_rate": 9.809325133793934e-05, + "loss": 0.0619, + "step": 6690 + }, + { + "grad_norm": 0.49923238158226013, + "learning_rate": 9.808419552858477e-05, + "loss": 0.0564, + "step": 6700 + }, + { + "grad_norm": 0.2616688907146454, + "learning_rate": 9.807511868556944e-05, + "loss": 0.0607, + "step": 6710 + }, + { + "grad_norm": 0.4825271666049957, + "learning_rate": 9.806602081286392e-05, + "loss": 0.0609, + "step": 6720 + }, + { + "grad_norm": 0.38764438033103943, + "learning_rate": 9.805690191444785e-05, + "loss": 0.0599, + "step": 6730 + }, + { + "grad_norm": 0.2188865691423416, + "learning_rate": 9.804776199431022e-05, + "loss": 0.0667, + "step": 6740 + }, + { + "grad_norm": 0.2728814482688904, + "learning_rate": 9.80386010564491e-05, + "loss": 0.0533, + "step": 6750 + }, + { + "grad_norm": 0.3174012303352356, + "learning_rate": 9.802941910487183e-05, + "loss": 0.0518, + "step": 6760 + }, + { + "grad_norm": 0.214443638920784, + "learning_rate": 9.802021614359487e-05, + "loss": 0.052, + "step": 6770 + }, + { + "grad_norm": 0.29440051317214966, + "learning_rate": 9.801099217664394e-05, + "loss": 0.0616, + "step": 6780 + }, + { + "grad_norm": 0.3439861238002777, + "learning_rate": 9.800174720805388e-05, + "loss": 0.0546, + "step": 6790 + }, + { + "grad_norm": 0.35434988141059875, + "learning_rate": 9.799248124186878e-05, + "loss": 0.0574, + "step": 6800 + }, + { + "grad_norm": 0.3358229398727417, + "learning_rate": 9.798319428214187e-05, + "loss": 0.0465, + "step": 6810 + }, + { + "grad_norm": 0.36372849345207214, + "learning_rate": 9.79738863329356e-05, + "loss": 0.0581, + "step": 6820 + }, + { + "grad_norm": 0.4100535809993744, + "learning_rate": 9.79645573983216e-05, + "loss": 0.0578, + "step": 6830 + }, + { + "grad_norm": 0.33580055832862854, + "learning_rate": 9.795520748238062e-05, + "loss": 0.0611, + "step": 6840 + }, + { + "grad_norm": 0.3214091956615448, + "learning_rate": 9.794583658920264e-05, + "loss": 0.0583, + "step": 6850 + }, + { + "grad_norm": 0.23918569087982178, + "learning_rate": 9.793644472288683e-05, + "loss": 0.0557, + "step": 6860 + }, + { + "grad_norm": 0.310507208108902, + "learning_rate": 9.79270318875415e-05, + "loss": 0.0641, + "step": 6870 + }, + { + "grad_norm": 0.5224227905273438, + "learning_rate": 9.791759808728416e-05, + "loss": 0.0639, + "step": 6880 + }, + { + "grad_norm": 0.29302531480789185, + "learning_rate": 9.790814332624143e-05, + "loss": 0.0637, + "step": 6890 + }, + { + "grad_norm": 0.33376821875572205, + "learning_rate": 9.789866760854919e-05, + "loss": 0.0602, + "step": 6900 + }, + { + "grad_norm": 0.23674370348453522, + "learning_rate": 9.788917093835242e-05, + "loss": 0.0522, + "step": 6910 + }, + { + "grad_norm": 0.248178631067276, + "learning_rate": 9.787965331980528e-05, + "loss": 0.057, + "step": 6920 + }, + { + "grad_norm": 0.4613815248012543, + "learning_rate": 9.787011475707111e-05, + "loss": 0.0603, + "step": 6930 + }, + { + "grad_norm": 0.6173457503318787, + "learning_rate": 9.786055525432241e-05, + "loss": 0.0456, + "step": 6940 + }, + { + "grad_norm": 0.22543007135391235, + "learning_rate": 9.78509748157408e-05, + "loss": 0.0518, + "step": 6950 + }, + { + "grad_norm": 0.37629908323287964, + "learning_rate": 9.784137344551713e-05, + "loss": 0.0535, + "step": 6960 + }, + { + "grad_norm": 0.33666786551475525, + "learning_rate": 9.783175114785134e-05, + "loss": 0.059, + "step": 6970 + }, + { + "grad_norm": 0.3158324360847473, + "learning_rate": 9.782210792695254e-05, + "loss": 0.0643, + "step": 6980 + }, + { + "grad_norm": 0.25573205947875977, + "learning_rate": 9.781244378703904e-05, + "loss": 0.0558, + "step": 6990 + }, + { + "grad_norm": 0.3298552930355072, + "learning_rate": 9.780275873233824e-05, + "loss": 0.0579, + "step": 7000 + }, + { + "grad_norm": 0.31629255414009094, + "learning_rate": 9.77930527670867e-05, + "loss": 0.0638, + "step": 7010 + }, + { + "grad_norm": 0.4163645803928375, + "learning_rate": 9.778332589553016e-05, + "loss": 0.0611, + "step": 7020 + }, + { + "grad_norm": 0.3224683701992035, + "learning_rate": 9.777357812192349e-05, + "loss": 0.0647, + "step": 7030 + }, + { + "grad_norm": 0.29395633935928345, + "learning_rate": 9.776380945053066e-05, + "loss": 0.0543, + "step": 7040 + }, + { + "grad_norm": 0.32856321334838867, + "learning_rate": 9.775401988562486e-05, + "loss": 0.0554, + "step": 7050 + }, + { + "grad_norm": 0.258163183927536, + "learning_rate": 9.774420943148835e-05, + "loss": 0.0524, + "step": 7060 + }, + { + "grad_norm": 0.3040701150894165, + "learning_rate": 9.773437809241257e-05, + "loss": 0.0544, + "step": 7070 + }, + { + "grad_norm": 0.30010950565338135, + "learning_rate": 9.772452587269808e-05, + "loss": 0.069, + "step": 7080 + }, + { + "grad_norm": 0.38734909892082214, + "learning_rate": 9.771465277665457e-05, + "loss": 0.0597, + "step": 7090 + }, + { + "grad_norm": 0.2585156559944153, + "learning_rate": 9.770475880860089e-05, + "loss": 0.0534, + "step": 7100 + }, + { + "grad_norm": 0.21464261412620544, + "learning_rate": 9.769484397286495e-05, + "loss": 0.0562, + "step": 7110 + }, + { + "grad_norm": 0.20491628348827362, + "learning_rate": 9.768490827378388e-05, + "loss": 0.0504, + "step": 7120 + }, + { + "grad_norm": 0.19044701755046844, + "learning_rate": 9.767495171570387e-05, + "loss": 0.0524, + "step": 7130 + }, + { + "grad_norm": 0.25236910581588745, + "learning_rate": 9.766497430298027e-05, + "loss": 0.0694, + "step": 7140 + }, + { + "grad_norm": 0.29156821966171265, + "learning_rate": 9.765497603997754e-05, + "loss": 0.065, + "step": 7150 + }, + { + "grad_norm": 0.4211452603340149, + "learning_rate": 9.764495693106926e-05, + "loss": 0.0584, + "step": 7160 + }, + { + "grad_norm": 0.24101591110229492, + "learning_rate": 9.76349169806381e-05, + "loss": 0.0546, + "step": 7170 + }, + { + "grad_norm": 0.30891934037208557, + "learning_rate": 9.762485619307591e-05, + "loss": 0.0548, + "step": 7180 + }, + { + "grad_norm": 0.40987032651901245, + "learning_rate": 9.76147745727836e-05, + "loss": 0.0598, + "step": 7190 + }, + { + "grad_norm": 0.4604049026966095, + "learning_rate": 9.760467212417124e-05, + "loss": 0.0517, + "step": 7200 + }, + { + "grad_norm": 0.29977038502693176, + "learning_rate": 9.759454885165796e-05, + "loss": 0.0598, + "step": 7210 + }, + { + "grad_norm": 0.26485520601272583, + "learning_rate": 9.7584404759672e-05, + "loss": 0.0604, + "step": 7220 + }, + { + "grad_norm": 0.2938430607318878, + "learning_rate": 9.75742398526508e-05, + "loss": 0.056, + "step": 7230 + }, + { + "grad_norm": 0.175235778093338, + "learning_rate": 9.756405413504077e-05, + "loss": 0.0551, + "step": 7240 + }, + { + "grad_norm": 0.2755601108074188, + "learning_rate": 9.755384761129752e-05, + "loss": 0.0588, + "step": 7250 + }, + { + "grad_norm": 0.1937788873910904, + "learning_rate": 9.754362028588572e-05, + "loss": 0.0476, + "step": 7260 + }, + { + "grad_norm": 0.2519840896129608, + "learning_rate": 9.753337216327917e-05, + "loss": 0.052, + "step": 7270 + }, + { + "grad_norm": 0.3645058870315552, + "learning_rate": 9.752310324796073e-05, + "loss": 0.0526, + "step": 7280 + }, + { + "grad_norm": 0.49549588561058044, + "learning_rate": 9.751281354442235e-05, + "loss": 0.0551, + "step": 7290 + }, + { + "grad_norm": 0.22502219676971436, + "learning_rate": 9.750250305716513e-05, + "loss": 0.0528, + "step": 7300 + }, + { + "grad_norm": 0.28197038173675537, + "learning_rate": 9.749217179069923e-05, + "loss": 0.0566, + "step": 7310 + }, + { + "grad_norm": 0.28005310893058777, + "learning_rate": 9.748181974954386e-05, + "loss": 0.0469, + "step": 7320 + }, + { + "grad_norm": 0.2546914219856262, + "learning_rate": 9.74714469382274e-05, + "loss": 0.0541, + "step": 7330 + }, + { + "grad_norm": 0.33449283242225647, + "learning_rate": 9.746105336128724e-05, + "loss": 0.0604, + "step": 7340 + }, + { + "grad_norm": 0.42341873049736023, + "learning_rate": 9.745063902326988e-05, + "loss": 0.0509, + "step": 7350 + }, + { + "grad_norm": 0.28345221281051636, + "learning_rate": 9.744020392873093e-05, + "loss": 0.0516, + "step": 7360 + }, + { + "grad_norm": 0.20774313807487488, + "learning_rate": 9.742974808223504e-05, + "loss": 0.0532, + "step": 7370 + }, + { + "grad_norm": 0.2298946976661682, + "learning_rate": 9.741927148835596e-05, + "loss": 0.0507, + "step": 7380 + }, + { + "grad_norm": 0.2024466097354889, + "learning_rate": 9.740877415167648e-05, + "loss": 0.0543, + "step": 7390 + }, + { + "grad_norm": 0.24586723744869232, + "learning_rate": 9.739825607678854e-05, + "loss": 0.0519, + "step": 7400 + }, + { + "grad_norm": 0.2650355100631714, + "learning_rate": 9.738771726829308e-05, + "loss": 0.0553, + "step": 7410 + }, + { + "grad_norm": 0.2684073746204376, + "learning_rate": 9.737715773080011e-05, + "loss": 0.0643, + "step": 7420 + }, + { + "grad_norm": 0.2560785710811615, + "learning_rate": 9.736657746892877e-05, + "loss": 0.0601, + "step": 7430 + }, + { + "grad_norm": 0.25257155299186707, + "learning_rate": 9.735597648730718e-05, + "loss": 0.0559, + "step": 7440 + }, + { + "grad_norm": 0.5449967384338379, + "learning_rate": 9.734535479057262e-05, + "loss": 0.0574, + "step": 7450 + }, + { + "grad_norm": 0.2706824541091919, + "learning_rate": 9.733471238337136e-05, + "loss": 0.0509, + "step": 7460 + }, + { + "grad_norm": 0.24612566828727722, + "learning_rate": 9.732404927035871e-05, + "loss": 0.0533, + "step": 7470 + }, + { + "grad_norm": 0.32935768365859985, + "learning_rate": 9.731336545619915e-05, + "loss": 0.0574, + "step": 7480 + }, + { + "grad_norm": 0.30474749207496643, + "learning_rate": 9.73026609455661e-05, + "loss": 0.0537, + "step": 7490 + }, + { + "grad_norm": 0.436855673789978, + "learning_rate": 9.72919357431421e-05, + "loss": 0.0475, + "step": 7500 + }, + { + "grad_norm": 0.2664956748485565, + "learning_rate": 9.72811898536187e-05, + "loss": 0.0473, + "step": 7510 + }, + { + "grad_norm": 0.35856911540031433, + "learning_rate": 9.72704232816965e-05, + "loss": 0.0498, + "step": 7520 + }, + { + "grad_norm": 0.23607347905635834, + "learning_rate": 9.725963603208519e-05, + "loss": 0.0508, + "step": 7530 + }, + { + "grad_norm": 0.34169915318489075, + "learning_rate": 9.724882810950348e-05, + "loss": 0.0553, + "step": 7540 + }, + { + "grad_norm": 0.24430648982524872, + "learning_rate": 9.723799951867913e-05, + "loss": 0.0526, + "step": 7550 + }, + { + "grad_norm": 0.3952392339706421, + "learning_rate": 9.722715026434889e-05, + "loss": 0.0578, + "step": 7560 + }, + { + "grad_norm": 0.44756636023521423, + "learning_rate": 9.721628035125866e-05, + "loss": 0.0633, + "step": 7570 + }, + { + "grad_norm": 0.3241315186023712, + "learning_rate": 9.720538978416325e-05, + "loss": 0.0654, + "step": 7580 + }, + { + "grad_norm": 0.18763107061386108, + "learning_rate": 9.71944785678266e-05, + "loss": 0.0573, + "step": 7590 + }, + { + "grad_norm": 1.1229244470596313, + "learning_rate": 9.718354670702161e-05, + "loss": 0.0669, + "step": 7600 + }, + { + "grad_norm": 0.41748327016830444, + "learning_rate": 9.717259420653029e-05, + "loss": 0.0618, + "step": 7610 + }, + { + "grad_norm": 0.3326422870159149, + "learning_rate": 9.716162107114361e-05, + "loss": 0.0598, + "step": 7620 + }, + { + "grad_norm": 0.1885843276977539, + "learning_rate": 9.71506273056616e-05, + "loss": 0.0585, + "step": 7630 + }, + { + "grad_norm": 0.40534964203834534, + "learning_rate": 9.713961291489331e-05, + "loss": 0.0595, + "step": 7640 + }, + { + "grad_norm": 0.2412186861038208, + "learning_rate": 9.71285779036568e-05, + "loss": 0.0548, + "step": 7650 + }, + { + "grad_norm": 0.32689639925956726, + "learning_rate": 9.711752227677916e-05, + "loss": 0.0518, + "step": 7660 + }, + { + "grad_norm": 0.25141337513923645, + "learning_rate": 9.710644603909649e-05, + "loss": 0.0576, + "step": 7670 + }, + { + "grad_norm": 0.35604891180992126, + "learning_rate": 9.709534919545393e-05, + "loss": 0.0599, + "step": 7680 + }, + { + "grad_norm": 0.2833408713340759, + "learning_rate": 9.708423175070563e-05, + "loss": 0.0585, + "step": 7690 + }, + { + "grad_norm": 0.19502434134483337, + "learning_rate": 9.707309370971468e-05, + "loss": 0.0475, + "step": 7700 + }, + { + "grad_norm": 0.29581671953201294, + "learning_rate": 9.70619350773533e-05, + "loss": 0.0603, + "step": 7710 + }, + { + "grad_norm": 0.28610464930534363, + "learning_rate": 9.70507558585026e-05, + "loss": 0.0588, + "step": 7720 + }, + { + "grad_norm": 0.3562217950820923, + "learning_rate": 9.703955605805281e-05, + "loss": 0.0569, + "step": 7730 + }, + { + "grad_norm": 0.23887154459953308, + "learning_rate": 9.702833568090306e-05, + "loss": 0.0567, + "step": 7740 + }, + { + "grad_norm": 0.3056223690509796, + "learning_rate": 9.701709473196154e-05, + "loss": 0.0505, + "step": 7750 + }, + { + "grad_norm": 0.3595779538154602, + "learning_rate": 9.700583321614541e-05, + "loss": 0.0605, + "step": 7760 + }, + { + "grad_norm": 0.37052369117736816, + "learning_rate": 9.699455113838085e-05, + "loss": 0.0607, + "step": 7770 + }, + { + "grad_norm": 0.2955126166343689, + "learning_rate": 9.698324850360303e-05, + "loss": 0.06, + "step": 7780 + }, + { + "grad_norm": 0.338073194026947, + "learning_rate": 9.69719253167561e-05, + "loss": 0.0556, + "step": 7790 + }, + { + "grad_norm": 0.27446600794792175, + "learning_rate": 9.69605815827932e-05, + "loss": 0.0546, + "step": 7800 + }, + { + "grad_norm": 0.23636142909526825, + "learning_rate": 9.694921730667647e-05, + "loss": 0.058, + "step": 7810 + }, + { + "grad_norm": 0.24378602206707, + "learning_rate": 9.693783249337705e-05, + "loss": 0.0476, + "step": 7820 + }, + { + "grad_norm": 0.32347503304481506, + "learning_rate": 9.692642714787501e-05, + "loss": 0.0517, + "step": 7830 + }, + { + "grad_norm": 0.3591710925102234, + "learning_rate": 9.691500127515945e-05, + "loss": 0.047, + "step": 7840 + }, + { + "grad_norm": 0.25681620836257935, + "learning_rate": 9.690355488022844e-05, + "loss": 0.0529, + "step": 7850 + }, + { + "grad_norm": 0.2505233585834503, + "learning_rate": 9.689208796808902e-05, + "loss": 0.0568, + "step": 7860 + }, + { + "grad_norm": 0.2618549168109894, + "learning_rate": 9.68806005437572e-05, + "loss": 0.0605, + "step": 7870 + }, + { + "grad_norm": 0.3010663092136383, + "learning_rate": 9.686909261225796e-05, + "loss": 0.0579, + "step": 7880 + }, + { + "grad_norm": 0.28273433446884155, + "learning_rate": 9.685756417862531e-05, + "loss": 0.0533, + "step": 7890 + }, + { + "grad_norm": 0.27602753043174744, + "learning_rate": 9.684601524790212e-05, + "loss": 0.0604, + "step": 7900 + }, + { + "grad_norm": 0.22656139731407166, + "learning_rate": 9.68344458251403e-05, + "loss": 0.0563, + "step": 7910 + }, + { + "grad_norm": 0.4990251660346985, + "learning_rate": 9.682285591540072e-05, + "loss": 0.0668, + "step": 7920 + }, + { + "grad_norm": 0.29815471172332764, + "learning_rate": 9.681124552375322e-05, + "loss": 0.0525, + "step": 7930 + }, + { + "grad_norm": 0.2810400128364563, + "learning_rate": 9.679961465527654e-05, + "loss": 0.0593, + "step": 7940 + }, + { + "grad_norm": 0.3660327196121216, + "learning_rate": 9.678796331505843e-05, + "loss": 0.053, + "step": 7950 + }, + { + "grad_norm": 0.2246003895998001, + "learning_rate": 9.677629150819558e-05, + "loss": 0.0514, + "step": 7960 + }, + { + "grad_norm": 0.23118118941783905, + "learning_rate": 9.676459923979366e-05, + "loss": 0.056, + "step": 7970 + }, + { + "grad_norm": 0.3197435736656189, + "learning_rate": 9.675288651496723e-05, + "loss": 0.0525, + "step": 7980 + }, + { + "grad_norm": 0.18230006098747253, + "learning_rate": 9.674115333883986e-05, + "loss": 0.0541, + "step": 7990 + }, + { + "grad_norm": 0.19630298018455505, + "learning_rate": 9.672939971654402e-05, + "loss": 0.0471, + "step": 8000 + }, + { + "grad_norm": 0.26768818497657776, + "learning_rate": 9.671762565322117e-05, + "loss": 0.0507, + "step": 8010 + }, + { + "grad_norm": 0.1955055147409439, + "learning_rate": 9.670583115402166e-05, + "loss": 0.053, + "step": 8020 + }, + { + "grad_norm": 0.29596009850502014, + "learning_rate": 9.669401622410482e-05, + "loss": 0.0553, + "step": 8030 + }, + { + "grad_norm": 0.2619599401950836, + "learning_rate": 9.668218086863887e-05, + "loss": 0.055, + "step": 8040 + }, + { + "grad_norm": 0.1859939694404602, + "learning_rate": 9.667032509280103e-05, + "loss": 0.0547, + "step": 8050 + }, + { + "grad_norm": 0.238547220826149, + "learning_rate": 9.665844890177743e-05, + "loss": 0.0504, + "step": 8060 + }, + { + "grad_norm": 0.2322293519973755, + "learning_rate": 9.664655230076309e-05, + "loss": 0.0544, + "step": 8070 + }, + { + "grad_norm": 0.2987595796585083, + "learning_rate": 9.663463529496199e-05, + "loss": 0.0637, + "step": 8080 + }, + { + "grad_norm": 0.3090727627277374, + "learning_rate": 9.662269788958705e-05, + "loss": 0.056, + "step": 8090 + }, + { + "grad_norm": 0.2017315924167633, + "learning_rate": 9.661074008986008e-05, + "loss": 0.0543, + "step": 8100 + }, + { + "grad_norm": 0.25973206758499146, + "learning_rate": 9.659876190101184e-05, + "loss": 0.0619, + "step": 8110 + }, + { + "grad_norm": 0.21100321412086487, + "learning_rate": 9.6586763328282e-05, + "loss": 0.0584, + "step": 8120 + }, + { + "grad_norm": 0.2853195071220398, + "learning_rate": 9.657474437691913e-05, + "loss": 0.0659, + "step": 8130 + }, + { + "grad_norm": 0.1989116221666336, + "learning_rate": 9.656270505218073e-05, + "loss": 0.055, + "step": 8140 + }, + { + "grad_norm": 0.2167089581489563, + "learning_rate": 9.655064535933321e-05, + "loss": 0.0526, + "step": 8150 + }, + { + "grad_norm": 0.3178420960903168, + "learning_rate": 9.653856530365189e-05, + "loss": 0.0555, + "step": 8160 + }, + { + "grad_norm": 0.338356614112854, + "learning_rate": 9.6526464890421e-05, + "loss": 0.0528, + "step": 8170 + }, + { + "grad_norm": 0.2995891869068146, + "learning_rate": 9.651434412493367e-05, + "loss": 0.0507, + "step": 8180 + }, + { + "grad_norm": 0.39911699295043945, + "learning_rate": 9.650220301249195e-05, + "loss": 0.0568, + "step": 8190 + }, + { + "grad_norm": 0.26076415181159973, + "learning_rate": 9.649004155840675e-05, + "loss": 0.0468, + "step": 8200 + }, + { + "grad_norm": 0.3769339919090271, + "learning_rate": 9.647785976799792e-05, + "loss": 0.0474, + "step": 8210 + }, + { + "grad_norm": 0.25200337171554565, + "learning_rate": 9.646565764659417e-05, + "loss": 0.048, + "step": 8220 + }, + { + "grad_norm": 0.2387138307094574, + "learning_rate": 9.645343519953314e-05, + "loss": 0.0489, + "step": 8230 + }, + { + "grad_norm": 1.2830387353897095, + "learning_rate": 9.644119243216135e-05, + "loss": 0.0479, + "step": 8240 + }, + { + "grad_norm": 0.1845233142375946, + "learning_rate": 9.642892934983417e-05, + "loss": 0.0493, + "step": 8250 + }, + { + "grad_norm": 0.25130486488342285, + "learning_rate": 9.641664595791593e-05, + "loss": 0.0441, + "step": 8260 + }, + { + "grad_norm": 0.4373854100704193, + "learning_rate": 9.640434226177977e-05, + "loss": 0.049, + "step": 8270 + }, + { + "grad_norm": 0.37242013216018677, + "learning_rate": 9.639201826680777e-05, + "loss": 0.0564, + "step": 8280 + }, + { + "grad_norm": 0.3660130798816681, + "learning_rate": 9.637967397839083e-05, + "loss": 0.0457, + "step": 8290 + }, + { + "grad_norm": 0.25453826785087585, + "learning_rate": 9.63673094019288e-05, + "loss": 0.0493, + "step": 8300 + }, + { + "grad_norm": 0.23056882619857788, + "learning_rate": 9.635492454283035e-05, + "loss": 0.0483, + "step": 8310 + }, + { + "grad_norm": 0.18137189745903015, + "learning_rate": 9.634251940651302e-05, + "loss": 0.0389, + "step": 8320 + }, + { + "grad_norm": 0.27250951528549194, + "learning_rate": 9.633009399840327e-05, + "loss": 0.0497, + "step": 8330 + }, + { + "grad_norm": 0.16131381690502167, + "learning_rate": 9.631764832393639e-05, + "loss": 0.0418, + "step": 8340 + }, + { + "grad_norm": 0.26916640996932983, + "learning_rate": 9.630518238855651e-05, + "loss": 0.0539, + "step": 8350 + }, + { + "grad_norm": 0.24256230890750885, + "learning_rate": 9.629269619771668e-05, + "loss": 0.0494, + "step": 8360 + }, + { + "grad_norm": 0.22105772793293, + "learning_rate": 9.628018975687878e-05, + "loss": 0.0458, + "step": 8370 + }, + { + "grad_norm": 0.2685621976852417, + "learning_rate": 9.626766307151355e-05, + "loss": 0.0501, + "step": 8380 + }, + { + "grad_norm": 0.3519449830055237, + "learning_rate": 9.62551161471006e-05, + "loss": 0.0465, + "step": 8390 + }, + { + "grad_norm": 0.43870192766189575, + "learning_rate": 9.624254898912837e-05, + "loss": 0.0496, + "step": 8400 + }, + { + "grad_norm": 0.25902122259140015, + "learning_rate": 9.622996160309414e-05, + "loss": 0.0543, + "step": 8410 + }, + { + "grad_norm": 0.6716800332069397, + "learning_rate": 9.62173539945041e-05, + "loss": 0.0506, + "step": 8420 + }, + { + "grad_norm": 0.2965831458568573, + "learning_rate": 9.620472616887323e-05, + "loss": 0.0514, + "step": 8430 + }, + { + "grad_norm": 0.2312847375869751, + "learning_rate": 9.619207813172536e-05, + "loss": 0.0499, + "step": 8440 + }, + { + "grad_norm": 0.23909732699394226, + "learning_rate": 9.617940988859318e-05, + "loss": 0.0481, + "step": 8450 + }, + { + "grad_norm": 0.3774909973144531, + "learning_rate": 9.616672144501821e-05, + "loss": 0.0482, + "step": 8460 + }, + { + "grad_norm": 0.1937675029039383, + "learning_rate": 9.615401280655081e-05, + "loss": 0.0442, + "step": 8470 + }, + { + "grad_norm": 0.31886494159698486, + "learning_rate": 9.614128397875017e-05, + "loss": 0.0487, + "step": 8480 + }, + { + "grad_norm": 0.3426457345485687, + "learning_rate": 9.612853496718429e-05, + "loss": 0.0519, + "step": 8490 + }, + { + "grad_norm": 0.355590283870697, + "learning_rate": 9.611576577743004e-05, + "loss": 0.0531, + "step": 8500 + }, + { + "grad_norm": 0.33898794651031494, + "learning_rate": 9.610297641507313e-05, + "loss": 0.0516, + "step": 8510 + }, + { + "grad_norm": 0.21016043424606323, + "learning_rate": 9.6090166885708e-05, + "loss": 0.0418, + "step": 8520 + }, + { + "grad_norm": 0.17833904922008514, + "learning_rate": 9.607733719493798e-05, + "loss": 0.0525, + "step": 8530 + }, + { + "grad_norm": 0.23229184746742249, + "learning_rate": 9.606448734837527e-05, + "loss": 0.0472, + "step": 8540 + }, + { + "grad_norm": 1.4789303541183472, + "learning_rate": 9.605161735164079e-05, + "loss": 0.0542, + "step": 8550 + }, + { + "grad_norm": 0.18022626638412476, + "learning_rate": 9.60387272103643e-05, + "loss": 0.0505, + "step": 8560 + }, + { + "grad_norm": 0.3230234980583191, + "learning_rate": 9.60258169301844e-05, + "loss": 0.0687, + "step": 8570 + }, + { + "grad_norm": 0.2461729645729065, + "learning_rate": 9.601288651674851e-05, + "loss": 0.0548, + "step": 8580 + }, + { + "grad_norm": 0.25226449966430664, + "learning_rate": 9.59999359757128e-05, + "loss": 0.0501, + "step": 8590 + }, + { + "grad_norm": 0.2582034468650818, + "learning_rate": 9.598696531274227e-05, + "loss": 0.0536, + "step": 8600 + }, + { + "grad_norm": 0.2656032145023346, + "learning_rate": 9.597397453351076e-05, + "loss": 0.0539, + "step": 8610 + }, + { + "grad_norm": 0.32256147265434265, + "learning_rate": 9.596096364370087e-05, + "loss": 0.0574, + "step": 8620 + }, + { + "grad_norm": 0.27658355236053467, + "learning_rate": 9.594793264900399e-05, + "loss": 0.0524, + "step": 8630 + }, + { + "grad_norm": 0.29951712489128113, + "learning_rate": 9.593488155512032e-05, + "loss": 0.0597, + "step": 8640 + }, + { + "grad_norm": 0.23224188387393951, + "learning_rate": 9.592181036775886e-05, + "loss": 0.0599, + "step": 8650 + }, + { + "grad_norm": 0.2726594805717468, + "learning_rate": 9.590871909263741e-05, + "loss": 0.0535, + "step": 8660 + }, + { + "grad_norm": 0.36063140630722046, + "learning_rate": 9.589560773548252e-05, + "loss": 0.0516, + "step": 8670 + }, + { + "grad_norm": 0.19559621810913086, + "learning_rate": 9.588247630202952e-05, + "loss": 0.0484, + "step": 8680 + }, + { + "grad_norm": 0.3248080909252167, + "learning_rate": 9.586932479802258e-05, + "loss": 0.0543, + "step": 8690 + }, + { + "grad_norm": 0.26646092534065247, + "learning_rate": 9.585615322921462e-05, + "loss": 0.0562, + "step": 8700 + }, + { + "grad_norm": 0.2323741763830185, + "learning_rate": 9.58429616013673e-05, + "loss": 0.0585, + "step": 8710 + }, + { + "grad_norm": 0.3219522535800934, + "learning_rate": 9.58297499202511e-05, + "loss": 0.0647, + "step": 8720 + }, + { + "grad_norm": 0.2698144018650055, + "learning_rate": 9.581651819164525e-05, + "loss": 0.0504, + "step": 8730 + }, + { + "grad_norm": 0.35338547825813293, + "learning_rate": 9.580326642133777e-05, + "loss": 0.0545, + "step": 8740 + }, + { + "grad_norm": 0.39096197485923767, + "learning_rate": 9.578999461512544e-05, + "loss": 0.0559, + "step": 8750 + }, + { + "grad_norm": 0.2849591374397278, + "learning_rate": 9.577670277881378e-05, + "loss": 0.0501, + "step": 8760 + }, + { + "grad_norm": 0.39986085891723633, + "learning_rate": 9.57633909182171e-05, + "loss": 0.0564, + "step": 8770 + }, + { + "grad_norm": 0.2610582411289215, + "learning_rate": 9.575005903915847e-05, + "loss": 0.0596, + "step": 8780 + }, + { + "grad_norm": 0.5612221956253052, + "learning_rate": 9.573670714746972e-05, + "loss": 0.0536, + "step": 8790 + }, + { + "grad_norm": 0.22693412005901337, + "learning_rate": 9.572333524899136e-05, + "loss": 0.0523, + "step": 8800 + }, + { + "grad_norm": 0.22233878076076508, + "learning_rate": 9.570994334957278e-05, + "loss": 0.0557, + "step": 8810 + }, + { + "grad_norm": 0.395352303981781, + "learning_rate": 9.569653145507201e-05, + "loss": 0.0533, + "step": 8820 + }, + { + "grad_norm": 0.2697340250015259, + "learning_rate": 9.568309957135587e-05, + "loss": 0.0527, + "step": 8830 + }, + { + "grad_norm": 0.4027293920516968, + "learning_rate": 9.566964770429995e-05, + "loss": 0.0534, + "step": 8840 + }, + { + "grad_norm": 0.2686888873577118, + "learning_rate": 9.565617585978853e-05, + "loss": 0.0498, + "step": 8850 + }, + { + "grad_norm": 0.36583271622657776, + "learning_rate": 9.564268404371466e-05, + "loss": 0.0508, + "step": 8860 + }, + { + "grad_norm": 0.3749639689922333, + "learning_rate": 9.562917226198013e-05, + "loss": 0.0529, + "step": 8870 + }, + { + "grad_norm": 0.2547416090965271, + "learning_rate": 9.561564052049543e-05, + "loss": 0.0545, + "step": 8880 + }, + { + "grad_norm": 0.2867489755153656, + "learning_rate": 9.560208882517982e-05, + "loss": 0.0526, + "step": 8890 + }, + { + "grad_norm": 0.2949729859828949, + "learning_rate": 9.558851718196127e-05, + "loss": 0.0581, + "step": 8900 + }, + { + "grad_norm": 0.3401259481906891, + "learning_rate": 9.557492559677646e-05, + "loss": 0.0517, + "step": 8910 + }, + { + "grad_norm": 0.40360990166664124, + "learning_rate": 9.556131407557082e-05, + "loss": 0.0623, + "step": 8920 + }, + { + "grad_norm": 0.47348034381866455, + "learning_rate": 9.554768262429853e-05, + "loss": 0.0527, + "step": 8930 + }, + { + "grad_norm": 0.25452175736427307, + "learning_rate": 9.553403124892239e-05, + "loss": 0.0565, + "step": 8940 + }, + { + "grad_norm": 0.2991984188556671, + "learning_rate": 9.552035995541402e-05, + "loss": 0.0496, + "step": 8950 + }, + { + "grad_norm": 0.3472542464733124, + "learning_rate": 9.550666874975368e-05, + "loss": 0.0493, + "step": 8960 + }, + { + "grad_norm": 0.29348689317703247, + "learning_rate": 9.549295763793038e-05, + "loss": 0.0442, + "step": 8970 + }, + { + "grad_norm": 0.22654491662979126, + "learning_rate": 9.547922662594183e-05, + "loss": 0.0545, + "step": 8980 + }, + { + "grad_norm": 0.3419937789440155, + "learning_rate": 9.546547571979443e-05, + "loss": 0.063, + "step": 8990 + }, + { + "grad_norm": 1.2934268712997437, + "learning_rate": 9.545170492550331e-05, + "loss": 0.0528, + "step": 9000 + }, + { + "grad_norm": 0.6618008613586426, + "learning_rate": 9.543791424909226e-05, + "loss": 0.0513, + "step": 9010 + }, + { + "grad_norm": 0.24241536855697632, + "learning_rate": 9.542410369659382e-05, + "loss": 0.0544, + "step": 9020 + }, + { + "grad_norm": 0.28125348687171936, + "learning_rate": 9.541027327404916e-05, + "loss": 0.0591, + "step": 9030 + }, + { + "grad_norm": 0.3858105540275574, + "learning_rate": 9.539642298750821e-05, + "loss": 0.0584, + "step": 9040 + }, + { + "grad_norm": 0.20200857520103455, + "learning_rate": 9.538255284302954e-05, + "loss": 0.0506, + "step": 9050 + }, + { + "grad_norm": 0.2534528076648712, + "learning_rate": 9.536866284668043e-05, + "loss": 0.0582, + "step": 9060 + }, + { + "grad_norm": 0.2641616761684418, + "learning_rate": 9.53547530045368e-05, + "loss": 0.049, + "step": 9070 + }, + { + "grad_norm": 0.22964267432689667, + "learning_rate": 9.534082332268335e-05, + "loss": 0.0525, + "step": 9080 + }, + { + "grad_norm": 0.30588415265083313, + "learning_rate": 9.532687380721334e-05, + "loss": 0.0617, + "step": 9090 + }, + { + "grad_norm": 0.33312100172042847, + "learning_rate": 9.53129044642288e-05, + "loss": 0.0607, + "step": 9100 + }, + { + "grad_norm": 0.26640376448631287, + "learning_rate": 9.529891529984039e-05, + "loss": 0.0532, + "step": 9110 + }, + { + "grad_norm": 0.30497097969055176, + "learning_rate": 9.528490632016743e-05, + "loss": 0.059, + "step": 9120 + }, + { + "grad_norm": 0.322510302066803, + "learning_rate": 9.52708775313379e-05, + "loss": 0.0597, + "step": 9130 + }, + { + "grad_norm": 0.2359931319952011, + "learning_rate": 9.525682893948853e-05, + "loss": 0.0467, + "step": 9140 + }, + { + "grad_norm": 0.32413962483406067, + "learning_rate": 9.524276055076463e-05, + "loss": 0.0581, + "step": 9150 + }, + { + "grad_norm": 0.3484371602535248, + "learning_rate": 9.522867237132016e-05, + "loss": 0.0473, + "step": 9160 + }, + { + "grad_norm": 0.27642977237701416, + "learning_rate": 9.52145644073178e-05, + "loss": 0.0531, + "step": 9170 + }, + { + "grad_norm": 0.28197118639945984, + "learning_rate": 9.520043666492884e-05, + "loss": 0.0471, + "step": 9180 + }, + { + "grad_norm": 0.4332337975502014, + "learning_rate": 9.518628915033323e-05, + "loss": 0.0463, + "step": 9190 + }, + { + "grad_norm": 0.2410987913608551, + "learning_rate": 9.517212186971957e-05, + "loss": 0.0482, + "step": 9200 + }, + { + "grad_norm": 0.5694149732589722, + "learning_rate": 9.515793482928515e-05, + "loss": 0.0633, + "step": 9210 + }, + { + "grad_norm": 0.42111533880233765, + "learning_rate": 9.514372803523582e-05, + "loss": 0.0477, + "step": 9220 + }, + { + "grad_norm": 0.25322848558425903, + "learning_rate": 9.512950149378613e-05, + "loss": 0.047, + "step": 9230 + }, + { + "grad_norm": 0.3368905484676361, + "learning_rate": 9.511525521115925e-05, + "loss": 0.0519, + "step": 9240 + }, + { + "grad_norm": 0.24848178029060364, + "learning_rate": 9.510098919358698e-05, + "loss": 0.045, + "step": 9250 + }, + { + "grad_norm": 0.19530996680259705, + "learning_rate": 9.508670344730979e-05, + "loss": 0.0434, + "step": 9260 + }, + { + "grad_norm": 0.2947063446044922, + "learning_rate": 9.507239797857672e-05, + "loss": 0.0448, + "step": 9270 + }, + { + "grad_norm": 0.24613305926322937, + "learning_rate": 9.505807279364548e-05, + "loss": 0.0484, + "step": 9280 + }, + { + "grad_norm": 0.32375675439834595, + "learning_rate": 9.504372789878239e-05, + "loss": 0.053, + "step": 9290 + }, + { + "grad_norm": 0.3004557490348816, + "learning_rate": 9.502936330026239e-05, + "loss": 0.0547, + "step": 9300 + }, + { + "grad_norm": 0.29393112659454346, + "learning_rate": 9.501497900436906e-05, + "loss": 0.0457, + "step": 9310 + }, + { + "grad_norm": 0.40138012170791626, + "learning_rate": 9.500057501739457e-05, + "loss": 0.0588, + "step": 9320 + }, + { + "grad_norm": 0.29981282353401184, + "learning_rate": 9.49861513456397e-05, + "loss": 0.0524, + "step": 9330 + }, + { + "grad_norm": 0.1960965096950531, + "learning_rate": 9.497170799541388e-05, + "loss": 0.0538, + "step": 9340 + }, + { + "grad_norm": 0.3299374580383301, + "learning_rate": 9.495724497303508e-05, + "loss": 0.0513, + "step": 9350 + }, + { + "grad_norm": 0.31664595007896423, + "learning_rate": 9.494276228482998e-05, + "loss": 0.0458, + "step": 9360 + }, + { + "grad_norm": 0.32792413234710693, + "learning_rate": 9.492825993713374e-05, + "loss": 0.0548, + "step": 9370 + }, + { + "grad_norm": 0.24962927401065826, + "learning_rate": 9.491373793629023e-05, + "loss": 0.0514, + "step": 9380 + }, + { + "grad_norm": 0.3771499991416931, + "learning_rate": 9.489919628865182e-05, + "loss": 0.0508, + "step": 9390 + }, + { + "grad_norm": 0.23603038489818573, + "learning_rate": 9.488463500057955e-05, + "loss": 0.0508, + "step": 9400 + }, + { + "grad_norm": 0.29124873876571655, + "learning_rate": 9.487005407844302e-05, + "loss": 0.0492, + "step": 9410 + }, + { + "grad_norm": 0.2587064206600189, + "learning_rate": 9.485545352862039e-05, + "loss": 0.0495, + "step": 9420 + }, + { + "grad_norm": 0.5276822447776794, + "learning_rate": 9.48408333574985e-05, + "loss": 0.0454, + "step": 9430 + }, + { + "grad_norm": 0.2821282148361206, + "learning_rate": 9.482619357147264e-05, + "loss": 0.0438, + "step": 9440 + }, + { + "grad_norm": 0.4382314682006836, + "learning_rate": 9.481153417694679e-05, + "loss": 0.0442, + "step": 9450 + }, + { + "grad_norm": 0.2198871672153473, + "learning_rate": 9.479685518033347e-05, + "loss": 0.0423, + "step": 9460 + }, + { + "grad_norm": 0.3655402660369873, + "learning_rate": 9.478215658805375e-05, + "loss": 0.0566, + "step": 9470 + }, + { + "grad_norm": 0.2774282693862915, + "learning_rate": 9.47674384065373e-05, + "loss": 0.0426, + "step": 9480 + }, + { + "grad_norm": 0.23369228839874268, + "learning_rate": 9.475270064222237e-05, + "loss": 0.0475, + "step": 9490 + }, + { + "grad_norm": 0.271768182516098, + "learning_rate": 9.473794330155572e-05, + "loss": 0.0415, + "step": 9500 + }, + { + "grad_norm": 0.3560444414615631, + "learning_rate": 9.472316639099276e-05, + "loss": 0.0435, + "step": 9510 + }, + { + "grad_norm": 0.2580602169036865, + "learning_rate": 9.470836991699739e-05, + "loss": 0.046, + "step": 9520 + }, + { + "grad_norm": 0.3152103126049042, + "learning_rate": 9.469355388604208e-05, + "loss": 0.0442, + "step": 9530 + }, + { + "grad_norm": 0.3154265880584717, + "learning_rate": 9.467871830460787e-05, + "loss": 0.051, + "step": 9540 + }, + { + "grad_norm": 0.33570346236228943, + "learning_rate": 9.466386317918436e-05, + "loss": 0.047, + "step": 9550 + }, + { + "grad_norm": 0.32015907764434814, + "learning_rate": 9.464898851626969e-05, + "loss": 0.051, + "step": 9560 + }, + { + "grad_norm": 0.2775513529777527, + "learning_rate": 9.463409432237051e-05, + "loss": 0.056, + "step": 9570 + }, + { + "grad_norm": 0.27707451581954956, + "learning_rate": 9.461918060400209e-05, + "loss": 0.0485, + "step": 9580 + }, + { + "grad_norm": 0.2520845830440521, + "learning_rate": 9.460424736768816e-05, + "loss": 0.047, + "step": 9590 + }, + { + "grad_norm": 0.23821496963500977, + "learning_rate": 9.458929461996105e-05, + "loss": 0.0471, + "step": 9600 + }, + { + "grad_norm": 0.32992157340049744, + "learning_rate": 9.457432236736158e-05, + "loss": 0.0468, + "step": 9610 + }, + { + "grad_norm": 0.34224632382392883, + "learning_rate": 9.455933061643916e-05, + "loss": 0.0531, + "step": 9620 + }, + { + "grad_norm": 0.5947456359863281, + "learning_rate": 9.454431937375164e-05, + "loss": 0.0484, + "step": 9630 + }, + { + "grad_norm": 0.3298882842063904, + "learning_rate": 9.452928864586547e-05, + "loss": 0.0465, + "step": 9640 + }, + { + "grad_norm": 0.23984675109386444, + "learning_rate": 9.451423843935563e-05, + "loss": 0.0482, + "step": 9650 + }, + { + "grad_norm": 0.3991466760635376, + "learning_rate": 9.449916876080553e-05, + "loss": 0.0489, + "step": 9660 + }, + { + "grad_norm": 0.3024240732192993, + "learning_rate": 9.44840796168072e-05, + "loss": 0.0498, + "step": 9670 + }, + { + "grad_norm": 0.3000549376010895, + "learning_rate": 9.446897101396115e-05, + "loss": 0.0511, + "step": 9680 + }, + { + "grad_norm": 0.3498810827732086, + "learning_rate": 9.445384295887638e-05, + "loss": 0.0497, + "step": 9690 + }, + { + "grad_norm": 0.2871325612068176, + "learning_rate": 9.443869545817043e-05, + "loss": 0.0501, + "step": 9700 + }, + { + "grad_norm": 0.3635595142841339, + "learning_rate": 9.442352851846929e-05, + "loss": 0.0509, + "step": 9710 + }, + { + "grad_norm": 0.32764551043510437, + "learning_rate": 9.440834214640755e-05, + "loss": 0.053, + "step": 9720 + }, + { + "grad_norm": 0.2293274700641632, + "learning_rate": 9.439313634862823e-05, + "loss": 0.0551, + "step": 9730 + }, + { + "grad_norm": 0.205005943775177, + "learning_rate": 9.437791113178282e-05, + "loss": 0.0486, + "step": 9740 + }, + { + "grad_norm": 0.23215121030807495, + "learning_rate": 9.43626665025314e-05, + "loss": 0.0467, + "step": 9750 + }, + { + "grad_norm": 0.24552595615386963, + "learning_rate": 9.434740246754248e-05, + "loss": 0.0528, + "step": 9760 + }, + { + "grad_norm": 0.31673070788383484, + "learning_rate": 9.433211903349304e-05, + "loss": 0.0542, + "step": 9770 + }, + { + "grad_norm": 0.2940168082714081, + "learning_rate": 9.431681620706858e-05, + "loss": 0.0494, + "step": 9780 + }, + { + "grad_norm": 0.2104993611574173, + "learning_rate": 9.43014939949631e-05, + "loss": 0.0512, + "step": 9790 + }, + { + "grad_norm": 0.19203343987464905, + "learning_rate": 9.428615240387904e-05, + "loss": 0.0484, + "step": 9800 + }, + { + "grad_norm": 0.28214403986930847, + "learning_rate": 9.427079144052732e-05, + "loss": 0.0488, + "step": 9810 + }, + { + "grad_norm": 0.23665781319141388, + "learning_rate": 9.425541111162739e-05, + "loss": 0.0504, + "step": 9820 + }, + { + "grad_norm": 0.2723560631275177, + "learning_rate": 9.424001142390709e-05, + "loss": 0.047, + "step": 9830 + }, + { + "grad_norm": 0.24549894034862518, + "learning_rate": 9.422459238410277e-05, + "loss": 0.0545, + "step": 9840 + }, + { + "grad_norm": 0.23627550899982452, + "learning_rate": 9.420915399895926e-05, + "loss": 0.0492, + "step": 9850 + }, + { + "grad_norm": 0.23891566693782806, + "learning_rate": 9.419369627522981e-05, + "loss": 0.0451, + "step": 9860 + }, + { + "grad_norm": 0.3533182442188263, + "learning_rate": 9.417821921967618e-05, + "loss": 0.0448, + "step": 9870 + }, + { + "grad_norm": 0.2844022214412689, + "learning_rate": 9.416272283906855e-05, + "loss": 0.0525, + "step": 9880 + }, + { + "grad_norm": 0.23856426775455475, + "learning_rate": 9.414720714018554e-05, + "loss": 0.043, + "step": 9890 + }, + { + "grad_norm": 0.5825778841972351, + "learning_rate": 9.413167212981427e-05, + "loss": 0.0489, + "step": 9900 + }, + { + "grad_norm": 0.299273818731308, + "learning_rate": 9.41161178147503e-05, + "loss": 0.0458, + "step": 9910 + }, + { + "grad_norm": 0.22647003829479218, + "learning_rate": 9.410054420179755e-05, + "loss": 0.0514, + "step": 9920 + }, + { + "grad_norm": 0.2706452012062073, + "learning_rate": 9.408495129776851e-05, + "loss": 0.0485, + "step": 9930 + }, + { + "grad_norm": 0.2235209047794342, + "learning_rate": 9.406933910948403e-05, + "loss": 0.0525, + "step": 9940 + }, + { + "grad_norm": 0.2879578173160553, + "learning_rate": 9.40537076437734e-05, + "loss": 0.0419, + "step": 9950 + }, + { + "grad_norm": 0.20402956008911133, + "learning_rate": 9.403805690747436e-05, + "loss": 0.0429, + "step": 9960 + }, + { + "grad_norm": 0.21465864777565002, + "learning_rate": 9.402238690743308e-05, + "loss": 0.0489, + "step": 9970 + }, + { + "grad_norm": 0.2903194725513458, + "learning_rate": 9.400669765050413e-05, + "loss": 0.0512, + "step": 9980 + }, + { + "grad_norm": 0.2027258276939392, + "learning_rate": 9.399098914355055e-05, + "loss": 0.0487, + "step": 9990 + }, + { + "grad_norm": 0.32303106784820557, + "learning_rate": 9.397526139344378e-05, + "loss": 0.0567, + "step": 10000 + }, + { + "grad_norm": 0.27272990345954895, + "learning_rate": 9.395951440706362e-05, + "loss": 0.0485, + "step": 10010 + }, + { + "grad_norm": 0.19187656044960022, + "learning_rate": 9.394374819129839e-05, + "loss": 0.0474, + "step": 10020 + }, + { + "grad_norm": 0.2766871452331543, + "learning_rate": 9.392796275304474e-05, + "loss": 0.0468, + "step": 10030 + }, + { + "grad_norm": 0.2867666184902191, + "learning_rate": 9.391215809920778e-05, + "loss": 0.0533, + "step": 10040 + }, + { + "grad_norm": 0.3950456380844116, + "learning_rate": 9.389633423670099e-05, + "loss": 0.0501, + "step": 10050 + }, + { + "grad_norm": 0.2603629529476166, + "learning_rate": 9.388049117244626e-05, + "loss": 0.0488, + "step": 10060 + }, + { + "grad_norm": 0.2541274428367615, + "learning_rate": 9.386462891337389e-05, + "loss": 0.0426, + "step": 10070 + }, + { + "grad_norm": 0.25304803252220154, + "learning_rate": 9.384874746642257e-05, + "loss": 0.0503, + "step": 10080 + }, + { + "grad_norm": 0.45924946665763855, + "learning_rate": 9.383284683853937e-05, + "loss": 0.0471, + "step": 10090 + }, + { + "grad_norm": 0.1458349972963333, + "learning_rate": 9.381692703667981e-05, + "loss": 0.0468, + "step": 10100 + }, + { + "grad_norm": 0.3060206472873688, + "learning_rate": 9.380098806780771e-05, + "loss": 0.045, + "step": 10110 + }, + { + "grad_norm": 0.3223259747028351, + "learning_rate": 9.378502993889533e-05, + "loss": 0.0493, + "step": 10120 + }, + { + "grad_norm": 0.2873098850250244, + "learning_rate": 9.376905265692329e-05, + "loss": 0.0551, + "step": 10130 + }, + { + "grad_norm": 0.5747036337852478, + "learning_rate": 9.37530562288806e-05, + "loss": 0.047, + "step": 10140 + }, + { + "grad_norm": 0.3002784550189972, + "learning_rate": 9.373704066176465e-05, + "loss": 0.0523, + "step": 10150 + }, + { + "grad_norm": 0.17516814172267914, + "learning_rate": 9.372100596258118e-05, + "loss": 0.0489, + "step": 10160 + }, + { + "grad_norm": 0.5741003751754761, + "learning_rate": 9.370495213834433e-05, + "loss": 0.0556, + "step": 10170 + }, + { + "grad_norm": 0.32274115085601807, + "learning_rate": 9.368887919607657e-05, + "loss": 0.0595, + "step": 10180 + }, + { + "grad_norm": 0.26641854643821716, + "learning_rate": 9.367278714280876e-05, + "loss": 0.0537, + "step": 10190 + }, + { + "grad_norm": 0.2620307207107544, + "learning_rate": 9.36566759855801e-05, + "loss": 0.0529, + "step": 10200 + }, + { + "grad_norm": 0.17607788741588593, + "learning_rate": 9.36405457314382e-05, + "loss": 0.0507, + "step": 10210 + }, + { + "grad_norm": 0.21855588257312775, + "learning_rate": 9.36243963874389e-05, + "loss": 0.0486, + "step": 10220 + }, + { + "grad_norm": 0.3885951042175293, + "learning_rate": 9.360822796064655e-05, + "loss": 0.0494, + "step": 10230 + }, + { + "grad_norm": 0.21305544674396515, + "learning_rate": 9.359204045813372e-05, + "loss": 0.0507, + "step": 10240 + }, + { + "grad_norm": 0.3113187253475189, + "learning_rate": 9.357583388698141e-05, + "loss": 0.0495, + "step": 10250 + }, + { + "grad_norm": 0.31736746430397034, + "learning_rate": 9.35596082542789e-05, + "loss": 0.0427, + "step": 10260 + }, + { + "grad_norm": 0.36639708280563354, + "learning_rate": 9.354336356712383e-05, + "loss": 0.0528, + "step": 10270 + }, + { + "grad_norm": 0.3673812747001648, + "learning_rate": 9.35270998326222e-05, + "loss": 0.0559, + "step": 10280 + }, + { + "grad_norm": 0.376943439245224, + "learning_rate": 9.351081705788831e-05, + "loss": 0.049, + "step": 10290 + }, + { + "grad_norm": 0.357707142829895, + "learning_rate": 9.349451525004477e-05, + "loss": 0.049, + "step": 10300 + }, + { + "grad_norm": 0.3553984761238098, + "learning_rate": 9.347819441622261e-05, + "loss": 0.0458, + "step": 10310 + }, + { + "grad_norm": 0.20998893678188324, + "learning_rate": 9.346185456356105e-05, + "loss": 0.0479, + "step": 10320 + }, + { + "grad_norm": 0.21264870464801788, + "learning_rate": 9.344549569920774e-05, + "loss": 0.0481, + "step": 10330 + }, + { + "grad_norm": 0.27864503860473633, + "learning_rate": 9.342911783031858e-05, + "loss": 0.0481, + "step": 10340 + }, + { + "grad_norm": 0.2538957893848419, + "learning_rate": 9.341272096405782e-05, + "loss": 0.058, + "step": 10350 + }, + { + "grad_norm": 0.26466071605682373, + "learning_rate": 9.3396305107598e-05, + "loss": 0.0448, + "step": 10360 + }, + { + "grad_norm": 0.2584731876850128, + "learning_rate": 9.337987026811998e-05, + "loss": 0.0461, + "step": 10370 + }, + { + "grad_norm": 0.22136348485946655, + "learning_rate": 9.33634164528129e-05, + "loss": 0.0484, + "step": 10380 + }, + { + "grad_norm": 0.5950494408607483, + "learning_rate": 9.334694366887424e-05, + "loss": 0.0463, + "step": 10390 + }, + { + "grad_norm": 0.3114577531814575, + "learning_rate": 9.333045192350973e-05, + "loss": 0.048, + "step": 10400 + }, + { + "grad_norm": 0.35795795917510986, + "learning_rate": 9.331394122393345e-05, + "loss": 0.0451, + "step": 10410 + }, + { + "grad_norm": 0.26375913619995117, + "learning_rate": 9.329741157736771e-05, + "loss": 0.0448, + "step": 10420 + }, + { + "grad_norm": 0.2766830325126648, + "learning_rate": 9.328086299104317e-05, + "loss": 0.044, + "step": 10430 + }, + { + "grad_norm": 0.32745814323425293, + "learning_rate": 9.326429547219872e-05, + "loss": 0.0454, + "step": 10440 + }, + { + "grad_norm": 0.28437820076942444, + "learning_rate": 9.324770902808155e-05, + "loss": 0.0487, + "step": 10450 + }, + { + "grad_norm": 0.2531982362270355, + "learning_rate": 9.323110366594717e-05, + "loss": 0.0411, + "step": 10460 + }, + { + "grad_norm": 0.21777601540088654, + "learning_rate": 9.32144793930593e-05, + "loss": 0.0466, + "step": 10470 + }, + { + "grad_norm": 0.23847635090351105, + "learning_rate": 9.319783621668996e-05, + "loss": 0.0444, + "step": 10480 + }, + { + "grad_norm": 0.4019499123096466, + "learning_rate": 9.318117414411947e-05, + "loss": 0.0443, + "step": 10490 + }, + { + "grad_norm": 0.17550282180309296, + "learning_rate": 9.316449318263635e-05, + "loss": 0.044, + "step": 10500 + }, + { + "grad_norm": 0.26180991530418396, + "learning_rate": 9.314779333953744e-05, + "loss": 0.0511, + "step": 10510 + }, + { + "grad_norm": 0.17288275063037872, + "learning_rate": 9.313107462212781e-05, + "loss": 0.0411, + "step": 10520 + }, + { + "grad_norm": 0.30768224596977234, + "learning_rate": 9.311433703772082e-05, + "loss": 0.0474, + "step": 10530 + }, + { + "grad_norm": 0.20746031403541565, + "learning_rate": 9.3097580593638e-05, + "loss": 0.044, + "step": 10540 + }, + { + "grad_norm": 0.3185042440891266, + "learning_rate": 9.308080529720926e-05, + "loss": 0.0427, + "step": 10550 + }, + { + "grad_norm": 0.23041144013404846, + "learning_rate": 9.306401115577264e-05, + "loss": 0.0437, + "step": 10560 + }, + { + "grad_norm": 0.21200716495513916, + "learning_rate": 9.304719817667447e-05, + "loss": 0.0447, + "step": 10570 + }, + { + "grad_norm": 0.21086765825748444, + "learning_rate": 9.303036636726934e-05, + "loss": 0.0462, + "step": 10580 + }, + { + "grad_norm": 0.30576106905937195, + "learning_rate": 9.301351573492003e-05, + "loss": 0.0497, + "step": 10590 + }, + { + "grad_norm": 0.3868655860424042, + "learning_rate": 9.299664628699758e-05, + "loss": 0.0565, + "step": 10600 + }, + { + "grad_norm": 0.29961860179901123, + "learning_rate": 9.297975803088129e-05, + "loss": 0.0623, + "step": 10610 + }, + { + "grad_norm": 0.3361785113811493, + "learning_rate": 9.296285097395864e-05, + "loss": 0.0607, + "step": 10620 + }, + { + "grad_norm": 0.36018431186676025, + "learning_rate": 9.294592512362533e-05, + "loss": 0.0477, + "step": 10630 + }, + { + "grad_norm": 0.3523446023464203, + "learning_rate": 9.292898048728533e-05, + "loss": 0.0611, + "step": 10640 + }, + { + "grad_norm": 0.4470199942588806, + "learning_rate": 9.29120170723508e-05, + "loss": 0.0616, + "step": 10650 + }, + { + "grad_norm": 0.359570175409317, + "learning_rate": 9.28950348862421e-05, + "loss": 0.051, + "step": 10660 + }, + { + "grad_norm": 0.18625225126743317, + "learning_rate": 9.287803393638781e-05, + "loss": 0.0505, + "step": 10670 + }, + { + "grad_norm": 0.4705457091331482, + "learning_rate": 9.286101423022474e-05, + "loss": 0.0441, + "step": 10680 + }, + { + "grad_norm": 0.20678521692752838, + "learning_rate": 9.284397577519788e-05, + "loss": 0.0455, + "step": 10690 + }, + { + "grad_norm": 0.34289947152137756, + "learning_rate": 9.282691857876043e-05, + "loss": 0.0663, + "step": 10700 + }, + { + "grad_norm": 0.28105267882347107, + "learning_rate": 9.280984264837377e-05, + "loss": 0.0548, + "step": 10710 + }, + { + "grad_norm": 0.2858068645000458, + "learning_rate": 9.279274799150752e-05, + "loss": 0.0555, + "step": 10720 + }, + { + "grad_norm": 0.210941880941391, + "learning_rate": 9.277563461563945e-05, + "loss": 0.0463, + "step": 10730 + }, + { + "grad_norm": 0.2831188440322876, + "learning_rate": 9.275850252825555e-05, + "loss": 0.0503, + "step": 10740 + }, + { + "grad_norm": 0.27556106448173523, + "learning_rate": 9.274135173684994e-05, + "loss": 0.0576, + "step": 10750 + }, + { + "grad_norm": 0.26079273223876953, + "learning_rate": 9.272418224892498e-05, + "loss": 0.0581, + "step": 10760 + }, + { + "grad_norm": 0.4629969894886017, + "learning_rate": 9.27069940719912e-05, + "loss": 0.0584, + "step": 10770 + }, + { + "grad_norm": 0.22222499549388885, + "learning_rate": 9.268978721356727e-05, + "loss": 0.0539, + "step": 10780 + }, + { + "grad_norm": 0.2190624177455902, + "learning_rate": 9.267256168118008e-05, + "loss": 0.0491, + "step": 10790 + }, + { + "grad_norm": 0.35115647315979004, + "learning_rate": 9.265531748236463e-05, + "loss": 0.0552, + "step": 10800 + }, + { + "grad_norm": 0.3602045476436615, + "learning_rate": 9.263805462466416e-05, + "loss": 0.0494, + "step": 10810 + }, + { + "grad_norm": 0.29872798919677734, + "learning_rate": 9.262077311562998e-05, + "loss": 0.0477, + "step": 10820 + }, + { + "grad_norm": 0.23612548410892487, + "learning_rate": 9.260347296282165e-05, + "loss": 0.052, + "step": 10830 + }, + { + "grad_norm": 0.30279484391212463, + "learning_rate": 9.258615417380683e-05, + "loss": 0.0519, + "step": 10840 + }, + { + "grad_norm": 0.25993409752845764, + "learning_rate": 9.256881675616133e-05, + "loss": 0.0503, + "step": 10850 + }, + { + "grad_norm": 0.3406111001968384, + "learning_rate": 9.255146071746917e-05, + "loss": 0.0445, + "step": 10860 + }, + { + "grad_norm": 0.2732725441455841, + "learning_rate": 9.253408606532241e-05, + "loss": 0.0488, + "step": 10870 + }, + { + "grad_norm": 0.25960537791252136, + "learning_rate": 9.251669280732137e-05, + "loss": 0.0496, + "step": 10880 + }, + { + "grad_norm": 0.2867802679538727, + "learning_rate": 9.249928095107441e-05, + "loss": 0.0503, + "step": 10890 + }, + { + "grad_norm": 0.26875922083854675, + "learning_rate": 9.248185050419811e-05, + "loss": 0.0505, + "step": 10900 + }, + { + "grad_norm": 0.44912880659103394, + "learning_rate": 9.24644014743171e-05, + "loss": 0.0505, + "step": 10910 + }, + { + "grad_norm": 0.2535572350025177, + "learning_rate": 9.24469338690642e-05, + "loss": 0.0485, + "step": 10920 + }, + { + "grad_norm": 0.31861644983291626, + "learning_rate": 9.242944769608033e-05, + "loss": 0.0539, + "step": 10930 + }, + { + "grad_norm": 0.2584599256515503, + "learning_rate": 9.241194296301454e-05, + "loss": 0.0534, + "step": 10940 + }, + { + "grad_norm": 0.2617528736591339, + "learning_rate": 9.239441967752397e-05, + "loss": 0.0475, + "step": 10950 + }, + { + "grad_norm": 0.45682376623153687, + "learning_rate": 9.237687784727393e-05, + "loss": 0.0425, + "step": 10960 + }, + { + "grad_norm": 0.34504735469818115, + "learning_rate": 9.235931747993781e-05, + "loss": 0.0465, + "step": 10970 + }, + { + "grad_norm": 0.3561752438545227, + "learning_rate": 9.234173858319707e-05, + "loss": 0.0443, + "step": 10980 + }, + { + "grad_norm": 0.22578829526901245, + "learning_rate": 9.23241411647414e-05, + "loss": 0.0444, + "step": 10990 + }, + { + "grad_norm": 0.21679706871509552, + "learning_rate": 9.230652523226841e-05, + "loss": 0.0465, + "step": 11000 + }, + { + "grad_norm": 0.2539629638195038, + "learning_rate": 9.2288890793484e-05, + "loss": 0.0466, + "step": 11010 + }, + { + "grad_norm": 0.25012898445129395, + "learning_rate": 9.227123785610199e-05, + "loss": 0.0477, + "step": 11020 + }, + { + "grad_norm": 0.32265061140060425, + "learning_rate": 9.225356642784443e-05, + "loss": 0.0477, + "step": 11030 + }, + { + "grad_norm": 0.3530280292034149, + "learning_rate": 9.223587651644139e-05, + "loss": 0.0475, + "step": 11040 + }, + { + "grad_norm": 0.16473746299743652, + "learning_rate": 9.221816812963104e-05, + "loss": 0.0422, + "step": 11050 + }, + { + "grad_norm": 0.27623602747917175, + "learning_rate": 9.22004412751596e-05, + "loss": 0.0539, + "step": 11060 + }, + { + "grad_norm": 0.27629002928733826, + "learning_rate": 9.218269596078146e-05, + "loss": 0.0459, + "step": 11070 + }, + { + "grad_norm": 0.36106985807418823, + "learning_rate": 9.216493219425895e-05, + "loss": 0.0487, + "step": 11080 + }, + { + "grad_norm": 0.20527558028697968, + "learning_rate": 9.21471499833626e-05, + "loss": 0.048, + "step": 11090 + }, + { + "grad_norm": 0.24303168058395386, + "learning_rate": 9.212934933587094e-05, + "loss": 0.0528, + "step": 11100 + }, + { + "grad_norm": 0.27419960498809814, + "learning_rate": 9.211153025957056e-05, + "loss": 0.0524, + "step": 11110 + }, + { + "grad_norm": 0.4332587718963623, + "learning_rate": 9.209369276225614e-05, + "loss": 0.0544, + "step": 11120 + }, + { + "grad_norm": 0.2555220127105713, + "learning_rate": 9.207583685173042e-05, + "loss": 0.0563, + "step": 11130 + }, + { + "grad_norm": 0.37728455662727356, + "learning_rate": 9.205796253580417e-05, + "loss": 0.0489, + "step": 11140 + }, + { + "grad_norm": 0.2955632209777832, + "learning_rate": 9.204006982229621e-05, + "loss": 0.0391, + "step": 11150 + }, + { + "grad_norm": 0.30774274468421936, + "learning_rate": 9.202215871903346e-05, + "loss": 0.0551, + "step": 11160 + }, + { + "grad_norm": 0.3260641098022461, + "learning_rate": 9.20042292338508e-05, + "loss": 0.0497, + "step": 11170 + }, + { + "grad_norm": 0.3156026303768158, + "learning_rate": 9.198628137459123e-05, + "loss": 0.0532, + "step": 11180 + }, + { + "grad_norm": 0.20249497890472412, + "learning_rate": 9.196831514910572e-05, + "loss": 0.0573, + "step": 11190 + }, + { + "grad_norm": 0.18463073670864105, + "learning_rate": 9.195033056525332e-05, + "loss": 0.052, + "step": 11200 + }, + { + "grad_norm": 0.3264400362968445, + "learning_rate": 9.193232763090114e-05, + "loss": 0.0518, + "step": 11210 + }, + { + "grad_norm": 0.2558581233024597, + "learning_rate": 9.191430635392422e-05, + "loss": 0.0499, + "step": 11220 + }, + { + "grad_norm": 0.5275874137878418, + "learning_rate": 9.18962667422057e-05, + "loss": 0.0548, + "step": 11230 + }, + { + "grad_norm": 0.23690290749073029, + "learning_rate": 9.187820880363671e-05, + "loss": 0.05, + "step": 11240 + }, + { + "grad_norm": 0.24714899063110352, + "learning_rate": 9.186013254611643e-05, + "loss": 0.0422, + "step": 11250 + }, + { + "grad_norm": 0.2586497962474823, + "learning_rate": 9.1842037977552e-05, + "loss": 0.0458, + "step": 11260 + }, + { + "grad_norm": 0.2291662096977234, + "learning_rate": 9.182392510585862e-05, + "loss": 0.0466, + "step": 11270 + }, + { + "grad_norm": 0.219789057970047, + "learning_rate": 9.180579393895946e-05, + "loss": 0.0588, + "step": 11280 + }, + { + "grad_norm": 0.4060150682926178, + "learning_rate": 9.178764448478572e-05, + "loss": 0.0537, + "step": 11290 + }, + { + "grad_norm": 0.32612964510917664, + "learning_rate": 9.176947675127658e-05, + "loss": 0.0527, + "step": 11300 + }, + { + "grad_norm": 0.41145622730255127, + "learning_rate": 9.175129074637924e-05, + "loss": 0.0516, + "step": 11310 + }, + { + "grad_norm": 0.16977590322494507, + "learning_rate": 9.173308647804884e-05, + "loss": 0.0522, + "step": 11320 + }, + { + "grad_norm": 0.28435587882995605, + "learning_rate": 9.171486395424859e-05, + "loss": 0.0455, + "step": 11330 + }, + { + "grad_norm": 0.26183009147644043, + "learning_rate": 9.16966231829496e-05, + "loss": 0.0465, + "step": 11340 + }, + { + "grad_norm": 0.4000638425350189, + "learning_rate": 9.167836417213105e-05, + "loss": 0.0653, + "step": 11350 + }, + { + "grad_norm": 0.35159870982170105, + "learning_rate": 9.166008692978001e-05, + "loss": 0.0526, + "step": 11360 + }, + { + "grad_norm": 0.25777462124824524, + "learning_rate": 9.164179146389158e-05, + "loss": 0.0424, + "step": 11370 + }, + { + "grad_norm": 0.2794339954853058, + "learning_rate": 9.162347778246882e-05, + "loss": 0.0478, + "step": 11380 + }, + { + "grad_norm": 0.3287450671195984, + "learning_rate": 9.160514589352276e-05, + "loss": 0.0519, + "step": 11390 + }, + { + "grad_norm": 0.35843923687934875, + "learning_rate": 9.15867958050724e-05, + "loss": 0.0549, + "step": 11400 + }, + { + "grad_norm": 0.30936941504478455, + "learning_rate": 9.156842752514466e-05, + "loss": 0.0476, + "step": 11410 + }, + { + "grad_norm": 0.3429429233074188, + "learning_rate": 9.155004106177447e-05, + "loss": 0.0607, + "step": 11420 + }, + { + "grad_norm": 0.3377288281917572, + "learning_rate": 9.153163642300471e-05, + "loss": 0.0504, + "step": 11430 + }, + { + "grad_norm": 0.25106170773506165, + "learning_rate": 9.151321361688616e-05, + "loss": 0.051, + "step": 11440 + }, + { + "grad_norm": 0.2530759572982788, + "learning_rate": 9.149477265147762e-05, + "loss": 0.0456, + "step": 11450 + }, + { + "grad_norm": 0.3569735288619995, + "learning_rate": 9.147631353484574e-05, + "loss": 0.0565, + "step": 11460 + }, + { + "grad_norm": 0.3010083734989166, + "learning_rate": 9.145783627506522e-05, + "loss": 0.047, + "step": 11470 + }, + { + "grad_norm": 0.1985643357038498, + "learning_rate": 9.143934088021861e-05, + "loss": 0.0471, + "step": 11480 + }, + { + "grad_norm": 0.35397574305534363, + "learning_rate": 9.142082735839645e-05, + "loss": 0.0511, + "step": 11490 + }, + { + "grad_norm": 0.3255859911441803, + "learning_rate": 9.140229571769715e-05, + "loss": 0.047, + "step": 11500 + }, + { + "grad_norm": 1.3405965566635132, + "learning_rate": 9.138374596622709e-05, + "loss": 0.046, + "step": 11510 + }, + { + "grad_norm": 0.2239931970834732, + "learning_rate": 9.136517811210059e-05, + "loss": 0.0478, + "step": 11520 + }, + { + "grad_norm": 0.15112195909023285, + "learning_rate": 9.134659216343984e-05, + "loss": 0.0435, + "step": 11530 + }, + { + "grad_norm": 0.5903862118721008, + "learning_rate": 9.132798812837494e-05, + "loss": 0.052, + "step": 11540 + }, + { + "grad_norm": 0.2350093573331833, + "learning_rate": 9.130936601504396e-05, + "loss": 0.0497, + "step": 11550 + }, + { + "grad_norm": 0.38323184847831726, + "learning_rate": 9.129072583159284e-05, + "loss": 0.0451, + "step": 11560 + }, + { + "grad_norm": 0.23426315188407898, + "learning_rate": 9.127206758617542e-05, + "loss": 0.0446, + "step": 11570 + }, + { + "grad_norm": 0.22941850125789642, + "learning_rate": 9.125339128695346e-05, + "loss": 0.0507, + "step": 11580 + }, + { + "grad_norm": 0.24355117976665497, + "learning_rate": 9.123469694209659e-05, + "loss": 0.0505, + "step": 11590 + }, + { + "grad_norm": 0.2510841488838196, + "learning_rate": 9.121598455978239e-05, + "loss": 0.0489, + "step": 11600 + }, + { + "grad_norm": 0.22843404114246368, + "learning_rate": 9.119725414819624e-05, + "loss": 0.055, + "step": 11610 + }, + { + "grad_norm": 0.41791489720344543, + "learning_rate": 9.117850571553149e-05, + "loss": 0.0577, + "step": 11620 + }, + { + "grad_norm": 0.2923620045185089, + "learning_rate": 9.115973926998935e-05, + "loss": 0.0604, + "step": 11630 + }, + { + "grad_norm": 0.230786994099617, + "learning_rate": 9.114095481977888e-05, + "loss": 0.0467, + "step": 11640 + }, + { + "grad_norm": 0.31672289967536926, + "learning_rate": 9.112215237311703e-05, + "loss": 0.0489, + "step": 11650 + }, + { + "grad_norm": 0.2930147349834442, + "learning_rate": 9.110333193822867e-05, + "loss": 0.0537, + "step": 11660 + }, + { + "grad_norm": 0.11846616119146347, + "learning_rate": 9.108449352334645e-05, + "loss": 0.0415, + "step": 11670 + }, + { + "grad_norm": 0.2827041447162628, + "learning_rate": 9.106563713671094e-05, + "loss": 0.0482, + "step": 11680 + }, + { + "grad_norm": 0.27395281195640564, + "learning_rate": 9.104676278657061e-05, + "loss": 0.0475, + "step": 11690 + }, + { + "grad_norm": 0.3264710009098053, + "learning_rate": 9.102787048118169e-05, + "loss": 0.0532, + "step": 11700 + }, + { + "grad_norm": 1.0123249292373657, + "learning_rate": 9.100896022880834e-05, + "loss": 0.0465, + "step": 11710 + }, + { + "grad_norm": 0.21104976534843445, + "learning_rate": 9.099003203772254e-05, + "loss": 0.0484, + "step": 11720 + }, + { + "grad_norm": 0.289313942193985, + "learning_rate": 9.097108591620413e-05, + "loss": 0.0419, + "step": 11730 + }, + { + "grad_norm": 0.22602401673793793, + "learning_rate": 9.095212187254078e-05, + "loss": 0.0508, + "step": 11740 + }, + { + "grad_norm": 0.30888116359710693, + "learning_rate": 9.093313991502801e-05, + "loss": 0.0501, + "step": 11750 + }, + { + "grad_norm": 0.3105355203151703, + "learning_rate": 9.091414005196917e-05, + "loss": 0.0516, + "step": 11760 + }, + { + "grad_norm": 0.1861780881881714, + "learning_rate": 9.089512229167545e-05, + "loss": 0.0485, + "step": 11770 + }, + { + "grad_norm": 0.34521734714508057, + "learning_rate": 9.087608664246587e-05, + "loss": 0.0489, + "step": 11780 + }, + { + "grad_norm": 0.3040093183517456, + "learning_rate": 9.085703311266727e-05, + "loss": 0.045, + "step": 11790 + }, + { + "grad_norm": 0.3086695373058319, + "learning_rate": 9.083796171061429e-05, + "loss": 0.0493, + "step": 11800 + }, + { + "grad_norm": 0.25285810232162476, + "learning_rate": 9.081887244464941e-05, + "loss": 0.0575, + "step": 11810 + }, + { + "grad_norm": 0.2789866626262665, + "learning_rate": 9.079976532312297e-05, + "loss": 0.0522, + "step": 11820 + }, + { + "grad_norm": 0.33461615443229675, + "learning_rate": 9.078064035439301e-05, + "loss": 0.0569, + "step": 11830 + }, + { + "grad_norm": 0.20290158689022064, + "learning_rate": 9.07614975468255e-05, + "loss": 0.0462, + "step": 11840 + }, + { + "grad_norm": 0.8556491136550903, + "learning_rate": 9.074233690879412e-05, + "loss": 0.0502, + "step": 11850 + }, + { + "grad_norm": 0.36448901891708374, + "learning_rate": 9.072315844868038e-05, + "loss": 0.0574, + "step": 11860 + }, + { + "grad_norm": 0.23859399557113647, + "learning_rate": 9.07039621748736e-05, + "loss": 0.0463, + "step": 11870 + }, + { + "grad_norm": 0.22027674317359924, + "learning_rate": 9.06847480957709e-05, + "loss": 0.0623, + "step": 11880 + }, + { + "grad_norm": 0.21024280786514282, + "learning_rate": 9.066551621977713e-05, + "loss": 0.0526, + "step": 11890 + }, + { + "grad_norm": 0.17751123011112213, + "learning_rate": 9.064626655530501e-05, + "loss": 0.0469, + "step": 11900 + }, + { + "grad_norm": 0.35309866070747375, + "learning_rate": 9.062699911077497e-05, + "loss": 0.0464, + "step": 11910 + }, + { + "grad_norm": 0.28122296929359436, + "learning_rate": 9.060771389461524e-05, + "loss": 0.0428, + "step": 11920 + }, + { + "grad_norm": 0.26177695393562317, + "learning_rate": 9.058841091526187e-05, + "loss": 0.0507, + "step": 11930 + }, + { + "grad_norm": 0.22181791067123413, + "learning_rate": 9.056909018115858e-05, + "loss": 0.0478, + "step": 11940 + }, + { + "grad_norm": 0.2118104100227356, + "learning_rate": 9.054975170075697e-05, + "loss": 0.0429, + "step": 11950 + }, + { + "grad_norm": 0.23057784140110016, + "learning_rate": 9.053039548251631e-05, + "loss": 0.0455, + "step": 11960 + }, + { + "grad_norm": 0.23574604094028473, + "learning_rate": 9.051102153490368e-05, + "loss": 0.0476, + "step": 11970 + }, + { + "grad_norm": 0.28718382120132446, + "learning_rate": 9.04916298663939e-05, + "loss": 0.0488, + "step": 11980 + }, + { + "grad_norm": 0.33691275119781494, + "learning_rate": 9.047222048546955e-05, + "loss": 0.0497, + "step": 11990 + }, + { + "grad_norm": 0.3617194890975952, + "learning_rate": 9.045279340062097e-05, + "loss": 0.0563, + "step": 12000 + }, + { + "grad_norm": 0.31874698400497437, + "learning_rate": 9.043334862034618e-05, + "loss": 0.0522, + "step": 12010 + }, + { + "grad_norm": 0.22822734713554382, + "learning_rate": 9.041388615315102e-05, + "loss": 0.0439, + "step": 12020 + }, + { + "grad_norm": 0.4050627052783966, + "learning_rate": 9.039440600754905e-05, + "loss": 0.0451, + "step": 12030 + }, + { + "grad_norm": 0.4195640981197357, + "learning_rate": 9.037490819206151e-05, + "loss": 0.0484, + "step": 12040 + }, + { + "grad_norm": 0.3148546516895294, + "learning_rate": 9.035539271521744e-05, + "loss": 0.0503, + "step": 12050 + }, + { + "grad_norm": 0.22198393940925598, + "learning_rate": 9.033585958555356e-05, + "loss": 0.0478, + "step": 12060 + }, + { + "grad_norm": 0.1886405348777771, + "learning_rate": 9.031630881161431e-05, + "loss": 0.057, + "step": 12070 + }, + { + "grad_norm": 0.3791317343711853, + "learning_rate": 9.029674040195186e-05, + "loss": 0.0455, + "step": 12080 + }, + { + "grad_norm": 0.2935805320739746, + "learning_rate": 9.027715436512613e-05, + "loss": 0.0417, + "step": 12090 + }, + { + "grad_norm": 0.31982192397117615, + "learning_rate": 9.02575507097047e-05, + "loss": 0.0529, + "step": 12100 + }, + { + "grad_norm": 0.29975026845932007, + "learning_rate": 9.023792944426286e-05, + "loss": 0.0452, + "step": 12110 + }, + { + "grad_norm": 0.30609676241874695, + "learning_rate": 9.021829057738364e-05, + "loss": 0.0451, + "step": 12120 + }, + { + "grad_norm": 0.20191465318202972, + "learning_rate": 9.019863411765775e-05, + "loss": 0.0453, + "step": 12130 + }, + { + "grad_norm": 0.3398725688457489, + "learning_rate": 9.017896007368357e-05, + "loss": 0.0418, + "step": 12140 + }, + { + "grad_norm": 0.15289364755153656, + "learning_rate": 9.015926845406722e-05, + "loss": 0.0443, + "step": 12150 + }, + { + "grad_norm": 0.408047080039978, + "learning_rate": 9.013955926742245e-05, + "loss": 0.0421, + "step": 12160 + }, + { + "grad_norm": 0.19557738304138184, + "learning_rate": 9.011983252237077e-05, + "loss": 0.0512, + "step": 12170 + }, + { + "grad_norm": 0.34096524119377136, + "learning_rate": 9.01000882275413e-05, + "loss": 0.0548, + "step": 12180 + }, + { + "grad_norm": 0.3546806871891022, + "learning_rate": 9.008032639157088e-05, + "loss": 0.0442, + "step": 12190 + }, + { + "grad_norm": 0.42631158232688904, + "learning_rate": 9.006054702310401e-05, + "loss": 0.0521, + "step": 12200 + }, + { + "grad_norm": 0.2328377068042755, + "learning_rate": 9.004075013079283e-05, + "loss": 0.0551, + "step": 12210 + }, + { + "grad_norm": 0.2771594226360321, + "learning_rate": 9.00209357232972e-05, + "loss": 0.0506, + "step": 12220 + }, + { + "grad_norm": 0.26295754313468933, + "learning_rate": 9.000110380928461e-05, + "loss": 0.0492, + "step": 12230 + }, + { + "grad_norm": 0.22268828749656677, + "learning_rate": 8.998125439743021e-05, + "loss": 0.0462, + "step": 12240 + }, + { + "grad_norm": 0.29290592670440674, + "learning_rate": 8.996138749641682e-05, + "loss": 0.0581, + "step": 12250 + }, + { + "grad_norm": 0.26483315229415894, + "learning_rate": 8.994150311493488e-05, + "loss": 0.0553, + "step": 12260 + }, + { + "grad_norm": 0.27006685733795166, + "learning_rate": 8.992160126168247e-05, + "loss": 0.0497, + "step": 12270 + }, + { + "grad_norm": 0.26884204149246216, + "learning_rate": 8.99016819453654e-05, + "loss": 0.051, + "step": 12280 + }, + { + "grad_norm": 0.44962790608406067, + "learning_rate": 8.988174517469702e-05, + "loss": 0.0522, + "step": 12290 + }, + { + "grad_norm": 0.38280707597732544, + "learning_rate": 8.986179095839835e-05, + "loss": 0.0444, + "step": 12300 + }, + { + "grad_norm": 0.35236504673957825, + "learning_rate": 8.984181930519804e-05, + "loss": 0.0468, + "step": 12310 + }, + { + "grad_norm": 0.23648720979690552, + "learning_rate": 8.982183022383237e-05, + "loss": 0.05, + "step": 12320 + }, + { + "grad_norm": 0.2600039839744568, + "learning_rate": 8.980182372304525e-05, + "loss": 0.0453, + "step": 12330 + }, + { + "grad_norm": 0.3298589885234833, + "learning_rate": 8.97817998115882e-05, + "loss": 0.0483, + "step": 12340 + }, + { + "grad_norm": 0.18765972554683685, + "learning_rate": 8.976175849822038e-05, + "loss": 0.0465, + "step": 12350 + }, + { + "grad_norm": 0.2042599469423294, + "learning_rate": 8.97416997917085e-05, + "loss": 0.0526, + "step": 12360 + }, + { + "grad_norm": 0.269914448261261, + "learning_rate": 8.972162370082695e-05, + "loss": 0.0492, + "step": 12370 + }, + { + "grad_norm": 0.24950291216373444, + "learning_rate": 8.97015302343577e-05, + "loss": 0.0521, + "step": 12380 + }, + { + "grad_norm": 0.2607584297657013, + "learning_rate": 8.968141940109027e-05, + "loss": 0.0497, + "step": 12390 + }, + { + "grad_norm": 0.3150050938129425, + "learning_rate": 8.966129120982188e-05, + "loss": 0.0555, + "step": 12400 + }, + { + "grad_norm": 0.28229111433029175, + "learning_rate": 8.964114566935724e-05, + "loss": 0.0553, + "step": 12410 + }, + { + "grad_norm": 0.7515745759010315, + "learning_rate": 8.962098278850871e-05, + "loss": 0.0483, + "step": 12420 + }, + { + "grad_norm": 0.3274504244327545, + "learning_rate": 8.960080257609622e-05, + "loss": 0.0558, + "step": 12430 + }, + { + "grad_norm": 0.2590310573577881, + "learning_rate": 8.95806050409473e-05, + "loss": 0.0529, + "step": 12440 + }, + { + "grad_norm": 0.17778275907039642, + "learning_rate": 8.9560390191897e-05, + "loss": 0.045, + "step": 12450 + }, + { + "grad_norm": 0.2603387236595154, + "learning_rate": 8.9540158037788e-05, + "loss": 0.0523, + "step": 12460 + }, + { + "grad_norm": 0.19239230453968048, + "learning_rate": 8.951990858747054e-05, + "loss": 0.0448, + "step": 12470 + }, + { + "grad_norm": 0.4179021418094635, + "learning_rate": 8.94996418498024e-05, + "loss": 0.0444, + "step": 12480 + }, + { + "grad_norm": 0.23133867979049683, + "learning_rate": 8.947935783364896e-05, + "loss": 0.0474, + "step": 12490 + }, + { + "grad_norm": 0.2757193446159363, + "learning_rate": 8.945905654788311e-05, + "loss": 0.0525, + "step": 12500 + }, + { + "grad_norm": 0.2726050913333893, + "learning_rate": 8.943873800138535e-05, + "loss": 0.057, + "step": 12510 + }, + { + "grad_norm": 0.3149632513523102, + "learning_rate": 8.94184022030437e-05, + "loss": 0.0488, + "step": 12520 + }, + { + "grad_norm": 0.3149303197860718, + "learning_rate": 8.939804916175372e-05, + "loss": 0.0483, + "step": 12530 + }, + { + "grad_norm": 0.37101179361343384, + "learning_rate": 8.93776788864185e-05, + "loss": 0.0494, + "step": 12540 + }, + { + "grad_norm": 0.3884722590446472, + "learning_rate": 8.935729138594873e-05, + "loss": 0.0466, + "step": 12550 + }, + { + "grad_norm": 0.3030417561531067, + "learning_rate": 8.933688666926258e-05, + "loss": 0.0455, + "step": 12560 + }, + { + "grad_norm": 0.2256031632423401, + "learning_rate": 8.931646474528575e-05, + "loss": 0.0485, + "step": 12570 + }, + { + "grad_norm": 0.3653827905654907, + "learning_rate": 8.929602562295151e-05, + "loss": 0.0485, + "step": 12580 + }, + { + "grad_norm": 0.2592281401157379, + "learning_rate": 8.92755693112006e-05, + "loss": 0.0499, + "step": 12590 + }, + { + "grad_norm": 0.25693824887275696, + "learning_rate": 8.925509581898136e-05, + "loss": 0.0454, + "step": 12600 + }, + { + "grad_norm": 0.2483212947845459, + "learning_rate": 8.923460515524951e-05, + "loss": 0.0422, + "step": 12610 + }, + { + "grad_norm": 0.4972546100616455, + "learning_rate": 8.921409732896842e-05, + "loss": 0.0462, + "step": 12620 + }, + { + "grad_norm": 0.3080849349498749, + "learning_rate": 8.919357234910887e-05, + "loss": 0.0445, + "step": 12630 + }, + { + "grad_norm": 0.28636589646339417, + "learning_rate": 8.917303022464923e-05, + "loss": 0.0572, + "step": 12640 + }, + { + "grad_norm": 0.32691898941993713, + "learning_rate": 8.915247096457531e-05, + "loss": 0.0455, + "step": 12650 + }, + { + "grad_norm": 0.24564863741397858, + "learning_rate": 8.91318945778804e-05, + "loss": 0.0587, + "step": 12660 + }, + { + "grad_norm": 0.23709703981876373, + "learning_rate": 8.911130107356534e-05, + "loss": 0.0549, + "step": 12670 + }, + { + "grad_norm": 0.2828724980354309, + "learning_rate": 8.90906904606384e-05, + "loss": 0.0536, + "step": 12680 + }, + { + "grad_norm": 0.20537249743938446, + "learning_rate": 8.90700627481154e-05, + "loss": 0.0509, + "step": 12690 + }, + { + "grad_norm": 0.2688646912574768, + "learning_rate": 8.904941794501957e-05, + "loss": 0.0486, + "step": 12700 + }, + { + "grad_norm": 0.31358954310417175, + "learning_rate": 8.902875606038166e-05, + "loss": 0.0475, + "step": 12710 + }, + { + "grad_norm": 0.3412231504917145, + "learning_rate": 8.900807710323989e-05, + "loss": 0.0471, + "step": 12720 + }, + { + "grad_norm": 0.20012524724006653, + "learning_rate": 8.898738108263993e-05, + "loss": 0.0514, + "step": 12730 + }, + { + "grad_norm": 0.23452356457710266, + "learning_rate": 8.896666800763491e-05, + "loss": 0.0514, + "step": 12740 + }, + { + "grad_norm": 0.2950536906719208, + "learning_rate": 8.894593788728546e-05, + "loss": 0.0475, + "step": 12750 + }, + { + "grad_norm": 0.3884913921356201, + "learning_rate": 8.892519073065961e-05, + "loss": 0.0636, + "step": 12760 + }, + { + "grad_norm": 0.37784627079963684, + "learning_rate": 8.89044265468329e-05, + "loss": 0.0567, + "step": 12770 + }, + { + "grad_norm": 0.34963589906692505, + "learning_rate": 8.888364534488827e-05, + "loss": 0.0564, + "step": 12780 + }, + { + "grad_norm": 0.31827372312545776, + "learning_rate": 8.886284713391613e-05, + "loss": 0.0539, + "step": 12790 + }, + { + "grad_norm": 0.3780674934387207, + "learning_rate": 8.884203192301431e-05, + "loss": 0.0541, + "step": 12800 + }, + { + "grad_norm": 0.3073527216911316, + "learning_rate": 8.88211997212881e-05, + "loss": 0.0599, + "step": 12810 + }, + { + "grad_norm": 0.2417912483215332, + "learning_rate": 8.880035053785023e-05, + "loss": 0.0535, + "step": 12820 + }, + { + "grad_norm": 0.23326820135116577, + "learning_rate": 8.877948438182083e-05, + "loss": 0.0557, + "step": 12830 + }, + { + "grad_norm": 0.20759133994579315, + "learning_rate": 8.875860126232745e-05, + "loss": 0.0494, + "step": 12840 + }, + { + "grad_norm": 0.32590991258621216, + "learning_rate": 8.87377011885051e-05, + "loss": 0.0476, + "step": 12850 + }, + { + "grad_norm": 0.18976378440856934, + "learning_rate": 8.871678416949617e-05, + "loss": 0.0533, + "step": 12860 + }, + { + "grad_norm": 0.3496427834033966, + "learning_rate": 8.869585021445046e-05, + "loss": 0.0486, + "step": 12870 + }, + { + "grad_norm": 0.28008535504341125, + "learning_rate": 8.867489933252521e-05, + "loss": 0.0472, + "step": 12880 + }, + { + "grad_norm": 0.2698345184326172, + "learning_rate": 8.865393153288504e-05, + "loss": 0.0523, + "step": 12890 + }, + { + "grad_norm": 0.3421589136123657, + "learning_rate": 8.8632946824702e-05, + "loss": 0.0601, + "step": 12900 + }, + { + "grad_norm": 0.21093851327896118, + "learning_rate": 8.86119452171555e-05, + "loss": 0.044, + "step": 12910 + }, + { + "grad_norm": 0.19494196772575378, + "learning_rate": 8.859092671943234e-05, + "loss": 0.0413, + "step": 12920 + }, + { + "grad_norm": 0.3115912079811096, + "learning_rate": 8.856989134072676e-05, + "loss": 0.0468, + "step": 12930 + }, + { + "grad_norm": 0.27316781878471375, + "learning_rate": 8.85488390902403e-05, + "loss": 0.0491, + "step": 12940 + }, + { + "grad_norm": 0.3065856099128723, + "learning_rate": 8.852776997718199e-05, + "loss": 0.0501, + "step": 12950 + }, + { + "grad_norm": 0.2173897922039032, + "learning_rate": 8.850668401076812e-05, + "loss": 0.0462, + "step": 12960 + }, + { + "grad_norm": 0.2506040632724762, + "learning_rate": 8.848558120022246e-05, + "loss": 0.0479, + "step": 12970 + }, + { + "grad_norm": 0.30096593499183655, + "learning_rate": 8.846446155477603e-05, + "loss": 0.044, + "step": 12980 + }, + { + "grad_norm": 0.17942823469638824, + "learning_rate": 8.844332508366735e-05, + "loss": 0.047, + "step": 12990 + }, + { + "grad_norm": 0.2948220670223236, + "learning_rate": 8.84221717961422e-05, + "loss": 0.0462, + "step": 13000 + }, + { + "grad_norm": 0.3185938894748688, + "learning_rate": 8.840100170145374e-05, + "loss": 0.0593, + "step": 13010 + }, + { + "grad_norm": 0.34361276030540466, + "learning_rate": 8.837981480886249e-05, + "loss": 0.056, + "step": 13020 + }, + { + "grad_norm": 0.39633557200431824, + "learning_rate": 8.835861112763633e-05, + "loss": 0.0586, + "step": 13030 + }, + { + "grad_norm": 0.33521145582199097, + "learning_rate": 8.833739066705044e-05, + "loss": 0.0501, + "step": 13040 + }, + { + "grad_norm": 0.26364269852638245, + "learning_rate": 8.831615343638742e-05, + "loss": 0.0447, + "step": 13050 + }, + { + "grad_norm": 0.23465600609779358, + "learning_rate": 8.829489944493711e-05, + "loss": 0.0481, + "step": 13060 + }, + { + "grad_norm": 0.2553117275238037, + "learning_rate": 8.827362870199675e-05, + "loss": 0.058, + "step": 13070 + }, + { + "grad_norm": 0.24219797551631927, + "learning_rate": 8.825234121687089e-05, + "loss": 0.0504, + "step": 13080 + }, + { + "grad_norm": 0.3645480275154114, + "learning_rate": 8.823103699887139e-05, + "loss": 0.0521, + "step": 13090 + }, + { + "grad_norm": 0.33535656332969666, + "learning_rate": 8.820971605731745e-05, + "loss": 0.0516, + "step": 13100 + }, + { + "grad_norm": 0.24338792264461517, + "learning_rate": 8.818837840153556e-05, + "loss": 0.0518, + "step": 13110 + }, + { + "grad_norm": 0.250099778175354, + "learning_rate": 8.816702404085952e-05, + "loss": 0.0483, + "step": 13120 + }, + { + "grad_norm": 0.27779555320739746, + "learning_rate": 8.814565298463048e-05, + "loss": 0.0585, + "step": 13130 + }, + { + "grad_norm": 0.3290708661079407, + "learning_rate": 8.812426524219688e-05, + "loss": 0.0585, + "step": 13140 + }, + { + "grad_norm": 0.2673681974411011, + "learning_rate": 8.81028608229144e-05, + "loss": 0.0548, + "step": 13150 + }, + { + "grad_norm": 0.21974056959152222, + "learning_rate": 8.808143973614611e-05, + "loss": 0.0448, + "step": 13160 + }, + { + "grad_norm": 0.1769345998764038, + "learning_rate": 8.806000199126228e-05, + "loss": 0.0478, + "step": 13170 + }, + { + "grad_norm": 0.24667759239673615, + "learning_rate": 8.803854759764052e-05, + "loss": 0.0495, + "step": 13180 + }, + { + "grad_norm": 0.1820092350244522, + "learning_rate": 8.801707656466572e-05, + "loss": 0.0492, + "step": 13190 + }, + { + "grad_norm": 0.2761339545249939, + "learning_rate": 8.799558890173003e-05, + "loss": 0.056, + "step": 13200 + }, + { + "grad_norm": 0.1854458451271057, + "learning_rate": 8.79740846182329e-05, + "loss": 0.0397, + "step": 13210 + }, + { + "grad_norm": 0.3285755515098572, + "learning_rate": 8.7952563723581e-05, + "loss": 0.0443, + "step": 13220 + }, + { + "grad_norm": 0.20458106696605682, + "learning_rate": 8.793102622718834e-05, + "loss": 0.0483, + "step": 13230 + }, + { + "grad_norm": 0.2993568778038025, + "learning_rate": 8.790947213847613e-05, + "loss": 0.051, + "step": 13240 + }, + { + "grad_norm": 0.3012157380580902, + "learning_rate": 8.788790146687286e-05, + "loss": 0.0467, + "step": 13250 + }, + { + "grad_norm": 0.2973864674568176, + "learning_rate": 8.786631422181429e-05, + "loss": 0.0502, + "step": 13260 + }, + { + "grad_norm": 0.19456258416175842, + "learning_rate": 8.78447104127434e-05, + "loss": 0.0478, + "step": 13270 + }, + { + "grad_norm": 0.42257118225097656, + "learning_rate": 8.782309004911042e-05, + "loss": 0.0451, + "step": 13280 + }, + { + "grad_norm": 0.32617780566215515, + "learning_rate": 8.780145314037286e-05, + "loss": 0.0533, + "step": 13290 + }, + { + "grad_norm": 0.3191179037094116, + "learning_rate": 8.777979969599542e-05, + "loss": 0.0529, + "step": 13300 + }, + { + "grad_norm": 0.5732515454292297, + "learning_rate": 8.775812972545006e-05, + "loss": 0.0527, + "step": 13310 + }, + { + "grad_norm": 0.3744677007198334, + "learning_rate": 8.773644323821596e-05, + "loss": 0.0574, + "step": 13320 + }, + { + "grad_norm": 0.27403494715690613, + "learning_rate": 8.771474024377953e-05, + "loss": 0.0478, + "step": 13330 + }, + { + "grad_norm": 0.23771624267101288, + "learning_rate": 8.769302075163438e-05, + "loss": 0.0482, + "step": 13340 + }, + { + "grad_norm": 0.25402507185935974, + "learning_rate": 8.767128477128137e-05, + "loss": 0.0501, + "step": 13350 + }, + { + "grad_norm": 0.2281847596168518, + "learning_rate": 8.764953231222854e-05, + "loss": 0.0519, + "step": 13360 + }, + { + "grad_norm": 0.20658376812934875, + "learning_rate": 8.762776338399119e-05, + "loss": 0.0514, + "step": 13370 + }, + { + "grad_norm": 0.3388976454734802, + "learning_rate": 8.760597799609176e-05, + "loss": 0.052, + "step": 13380 + }, + { + "grad_norm": 0.3172014653682709, + "learning_rate": 8.758417615805992e-05, + "loss": 0.0509, + "step": 13390 + }, + { + "grad_norm": 0.4778355360031128, + "learning_rate": 8.756235787943254e-05, + "loss": 0.0553, + "step": 13400 + }, + { + "grad_norm": 0.37995707988739014, + "learning_rate": 8.754052316975367e-05, + "loss": 0.0496, + "step": 13410 + }, + { + "grad_norm": 0.22134019434452057, + "learning_rate": 8.751867203857455e-05, + "loss": 0.0528, + "step": 13420 + }, + { + "grad_norm": 0.2964901924133301, + "learning_rate": 8.749680449545363e-05, + "loss": 0.0456, + "step": 13430 + }, + { + "grad_norm": 0.28125977516174316, + "learning_rate": 8.747492054995649e-05, + "loss": 0.0526, + "step": 13440 + }, + { + "grad_norm": 0.3229845464229584, + "learning_rate": 8.745302021165595e-05, + "loss": 0.0574, + "step": 13450 + }, + { + "grad_norm": 0.752109706401825, + "learning_rate": 8.743110349013192e-05, + "loss": 0.0521, + "step": 13460 + }, + { + "grad_norm": 0.19487060606479645, + "learning_rate": 8.740917039497153e-05, + "loss": 0.0571, + "step": 13470 + }, + { + "grad_norm": 0.2073637694120407, + "learning_rate": 8.738722093576906e-05, + "loss": 0.054, + "step": 13480 + }, + { + "grad_norm": 0.2958093583583832, + "learning_rate": 8.736525512212597e-05, + "loss": 0.0534, + "step": 13490 + }, + { + "grad_norm": 0.20417754352092743, + "learning_rate": 8.734327296365084e-05, + "loss": 0.0417, + "step": 13500 + }, + { + "grad_norm": 0.2741698622703552, + "learning_rate": 8.732127446995939e-05, + "loss": 0.0535, + "step": 13510 + }, + { + "grad_norm": 0.29916131496429443, + "learning_rate": 8.729925965067454e-05, + "loss": 0.0553, + "step": 13520 + }, + { + "grad_norm": 0.3421539068222046, + "learning_rate": 8.72772285154263e-05, + "loss": 0.0569, + "step": 13530 + }, + { + "grad_norm": 0.30390986800193787, + "learning_rate": 8.725518107385187e-05, + "loss": 0.0507, + "step": 13540 + }, + { + "grad_norm": 0.22314341366291046, + "learning_rate": 8.72331173355955e-05, + "loss": 0.0445, + "step": 13550 + }, + { + "grad_norm": 0.24384461343288422, + "learning_rate": 8.721103731030867e-05, + "loss": 0.0493, + "step": 13560 + }, + { + "grad_norm": 0.32606133818626404, + "learning_rate": 8.718894100764989e-05, + "loss": 0.0454, + "step": 13570 + }, + { + "grad_norm": 0.45664507150650024, + "learning_rate": 8.716682843728485e-05, + "loss": 0.0462, + "step": 13580 + }, + { + "grad_norm": 0.2946050763130188, + "learning_rate": 8.714469960888634e-05, + "loss": 0.0519, + "step": 13590 + }, + { + "grad_norm": 0.19385796785354614, + "learning_rate": 8.712255453213427e-05, + "loss": 0.0587, + "step": 13600 + }, + { + "grad_norm": 0.25244462490081787, + "learning_rate": 8.710039321671563e-05, + "loss": 0.0419, + "step": 13610 + }, + { + "grad_norm": 0.3114321231842041, + "learning_rate": 8.707821567232456e-05, + "loss": 0.0547, + "step": 13620 + }, + { + "grad_norm": 0.26362699270248413, + "learning_rate": 8.705602190866225e-05, + "loss": 0.0452, + "step": 13630 + }, + { + "grad_norm": 0.4614866375923157, + "learning_rate": 8.703381193543701e-05, + "loss": 0.0436, + "step": 13640 + }, + { + "grad_norm": 0.23748990893363953, + "learning_rate": 8.701158576236423e-05, + "loss": 0.0508, + "step": 13650 + }, + { + "grad_norm": 0.36542969942092896, + "learning_rate": 8.69893433991664e-05, + "loss": 0.0543, + "step": 13660 + }, + { + "grad_norm": 0.22369742393493652, + "learning_rate": 8.69670848555731e-05, + "loss": 0.0466, + "step": 13670 + }, + { + "grad_norm": 0.2049321085214615, + "learning_rate": 8.694481014132096e-05, + "loss": 0.0519, + "step": 13680 + }, + { + "grad_norm": 0.39369475841522217, + "learning_rate": 8.69225192661537e-05, + "loss": 0.0495, + "step": 13690 + }, + { + "grad_norm": 0.269054651260376, + "learning_rate": 8.690021223982208e-05, + "loss": 0.0535, + "step": 13700 + }, + { + "grad_norm": 0.379645973443985, + "learning_rate": 8.687788907208398e-05, + "loss": 0.0493, + "step": 13710 + }, + { + "grad_norm": 0.20228365063667297, + "learning_rate": 8.685554977270431e-05, + "loss": 0.0488, + "step": 13720 + }, + { + "grad_norm": 0.25368914008140564, + "learning_rate": 8.683319435145503e-05, + "loss": 0.0437, + "step": 13730 + }, + { + "grad_norm": 0.3224157691001892, + "learning_rate": 8.681082281811517e-05, + "loss": 0.0485, + "step": 13740 + }, + { + "grad_norm": 0.28033843636512756, + "learning_rate": 8.67884351824708e-05, + "loss": 0.0493, + "step": 13750 + }, + { + "grad_norm": 0.2731296420097351, + "learning_rate": 8.676603145431501e-05, + "loss": 0.058, + "step": 13760 + }, + { + "grad_norm": 0.21485047042369843, + "learning_rate": 8.674361164344799e-05, + "loss": 0.0431, + "step": 13770 + }, + { + "grad_norm": 0.20864133536815643, + "learning_rate": 8.672117575967688e-05, + "loss": 0.0498, + "step": 13780 + }, + { + "grad_norm": 0.24016423523426056, + "learning_rate": 8.669872381281595e-05, + "loss": 0.0474, + "step": 13790 + }, + { + "grad_norm": 0.4311399757862091, + "learning_rate": 8.667625581268639e-05, + "loss": 0.0479, + "step": 13800 + }, + { + "grad_norm": 0.173959881067276, + "learning_rate": 8.665377176911651e-05, + "loss": 0.0498, + "step": 13810 + }, + { + "grad_norm": 0.4298970699310303, + "learning_rate": 8.663127169194159e-05, + "loss": 0.0521, + "step": 13820 + }, + { + "grad_norm": 0.2917996346950531, + "learning_rate": 8.660875559100389e-05, + "loss": 0.0519, + "step": 13830 + }, + { + "grad_norm": 0.2162090539932251, + "learning_rate": 8.658622347615274e-05, + "loss": 0.0527, + "step": 13840 + }, + { + "grad_norm": 0.3259277641773224, + "learning_rate": 8.656367535724448e-05, + "loss": 0.045, + "step": 13850 + }, + { + "grad_norm": 0.26229017972946167, + "learning_rate": 8.65411112441424e-05, + "loss": 0.0493, + "step": 13860 + }, + { + "grad_norm": 0.2914193868637085, + "learning_rate": 8.651853114671679e-05, + "loss": 0.0514, + "step": 13870 + }, + { + "grad_norm": 0.2730576992034912, + "learning_rate": 8.649593507484499e-05, + "loss": 0.0482, + "step": 13880 + }, + { + "grad_norm": 0.3319849967956543, + "learning_rate": 8.647332303841126e-05, + "loss": 0.0572, + "step": 13890 + }, + { + "grad_norm": 0.21871210634708405, + "learning_rate": 8.645069504730689e-05, + "loss": 0.0454, + "step": 13900 + }, + { + "grad_norm": 0.2906345725059509, + "learning_rate": 8.64280511114301e-05, + "loss": 0.048, + "step": 13910 + }, + { + "grad_norm": 0.23607976734638214, + "learning_rate": 8.640539124068617e-05, + "loss": 0.0505, + "step": 13920 + }, + { + "grad_norm": 0.48682305216789246, + "learning_rate": 8.638271544498727e-05, + "loss": 0.0465, + "step": 13930 + }, + { + "grad_norm": 0.2255808264017105, + "learning_rate": 8.636002373425257e-05, + "loss": 0.049, + "step": 13940 + }, + { + "grad_norm": 0.2710997760295868, + "learning_rate": 8.633731611840817e-05, + "loss": 0.0493, + "step": 13950 + }, + { + "grad_norm": 0.24385470151901245, + "learning_rate": 8.631459260738717e-05, + "loss": 0.0521, + "step": 13960 + }, + { + "grad_norm": 0.24469460546970367, + "learning_rate": 8.62918532111296e-05, + "loss": 0.0618, + "step": 13970 + }, + { + "grad_norm": 0.5234110951423645, + "learning_rate": 8.626909793958248e-05, + "loss": 0.0535, + "step": 13980 + }, + { + "grad_norm": 0.2695081830024719, + "learning_rate": 8.624632680269969e-05, + "loss": 0.0529, + "step": 13990 + }, + { + "grad_norm": 0.39833858609199524, + "learning_rate": 8.622353981044212e-05, + "loss": 0.0517, + "step": 14000 + }, + { + "grad_norm": 0.622788667678833, + "learning_rate": 8.620073697277757e-05, + "loss": 0.0484, + "step": 14010 + }, + { + "grad_norm": 0.27222877740859985, + "learning_rate": 8.617791829968079e-05, + "loss": 0.0563, + "step": 14020 + }, + { + "grad_norm": 0.33116415143013, + "learning_rate": 8.615508380113344e-05, + "loss": 0.0495, + "step": 14030 + }, + { + "grad_norm": 0.2767364978790283, + "learning_rate": 8.613223348712408e-05, + "loss": 0.0402, + "step": 14040 + }, + { + "grad_norm": 0.19616127014160156, + "learning_rate": 8.610936736764824e-05, + "loss": 0.0406, + "step": 14050 + }, + { + "grad_norm": 0.19590343534946442, + "learning_rate": 8.608648545270833e-05, + "loss": 0.047, + "step": 14060 + }, + { + "grad_norm": 0.281523734331131, + "learning_rate": 8.606358775231366e-05, + "loss": 0.0449, + "step": 14070 + }, + { + "grad_norm": 0.2226816564798355, + "learning_rate": 8.60406742764805e-05, + "loss": 0.0476, + "step": 14080 + }, + { + "grad_norm": 0.25387731194496155, + "learning_rate": 8.601774503523195e-05, + "loss": 0.0435, + "step": 14090 + }, + { + "grad_norm": 0.2613195478916168, + "learning_rate": 8.599480003859805e-05, + "loss": 0.0561, + "step": 14100 + }, + { + "grad_norm": 0.21515342593193054, + "learning_rate": 8.597183929661573e-05, + "loss": 0.0485, + "step": 14110 + }, + { + "grad_norm": 0.15978682041168213, + "learning_rate": 8.594886281932879e-05, + "loss": 0.043, + "step": 14120 + }, + { + "grad_norm": 0.3208857476711273, + "learning_rate": 8.59258706167879e-05, + "loss": 0.0504, + "step": 14130 + }, + { + "grad_norm": 0.34434574842453003, + "learning_rate": 8.590286269905068e-05, + "loss": 0.0526, + "step": 14140 + }, + { + "grad_norm": 0.23682771623134613, + "learning_rate": 8.587983907618154e-05, + "loss": 0.0443, + "step": 14150 + }, + { + "grad_norm": 0.4380292296409607, + "learning_rate": 8.585679975825178e-05, + "loss": 0.059, + "step": 14160 + }, + { + "grad_norm": 0.3615966737270355, + "learning_rate": 8.583374475533962e-05, + "loss": 0.051, + "step": 14170 + }, + { + "grad_norm": 0.29449954628944397, + "learning_rate": 8.581067407753009e-05, + "loss": 0.0506, + "step": 14180 + }, + { + "grad_norm": 0.34454676508903503, + "learning_rate": 8.578758773491507e-05, + "loss": 0.0501, + "step": 14190 + }, + { + "grad_norm": 0.22112718224525452, + "learning_rate": 8.576448573759332e-05, + "loss": 0.0493, + "step": 14200 + }, + { + "grad_norm": 0.3391239047050476, + "learning_rate": 8.574136809567044e-05, + "loss": 0.0486, + "step": 14210 + }, + { + "grad_norm": 0.3071008026599884, + "learning_rate": 8.57182348192589e-05, + "loss": 0.0535, + "step": 14220 + }, + { + "grad_norm": 0.22822785377502441, + "learning_rate": 8.569508591847792e-05, + "loss": 0.0507, + "step": 14230 + }, + { + "grad_norm": 0.2732331156730652, + "learning_rate": 8.567192140345367e-05, + "loss": 0.0516, + "step": 14240 + }, + { + "grad_norm": 0.27169573307037354, + "learning_rate": 8.564874128431906e-05, + "loss": 0.0567, + "step": 14250 + }, + { + "grad_norm": 0.3751930594444275, + "learning_rate": 8.562554557121389e-05, + "loss": 0.0503, + "step": 14260 + }, + { + "grad_norm": 0.24979856610298157, + "learning_rate": 8.560233427428475e-05, + "loss": 0.0578, + "step": 14270 + }, + { + "grad_norm": 0.29438623785972595, + "learning_rate": 8.557910740368503e-05, + "loss": 0.0519, + "step": 14280 + }, + { + "grad_norm": 0.2992285192012787, + "learning_rate": 8.555586496957495e-05, + "loss": 0.0412, + "step": 14290 + }, + { + "grad_norm": 0.20691318809986115, + "learning_rate": 8.553260698212155e-05, + "loss": 0.048, + "step": 14300 + }, + { + "grad_norm": 0.2899720370769501, + "learning_rate": 8.550933345149868e-05, + "loss": 0.0452, + "step": 14310 + }, + { + "grad_norm": 0.32470715045928955, + "learning_rate": 8.548604438788696e-05, + "loss": 0.0603, + "step": 14320 + }, + { + "grad_norm": 0.2646409273147583, + "learning_rate": 8.546273980147383e-05, + "loss": 0.0476, + "step": 14330 + }, + { + "grad_norm": 0.3365819752216339, + "learning_rate": 8.543941970245348e-05, + "loss": 0.0516, + "step": 14340 + }, + { + "grad_norm": 0.47422313690185547, + "learning_rate": 8.541608410102693e-05, + "loss": 0.0539, + "step": 14350 + }, + { + "grad_norm": 0.2389652281999588, + "learning_rate": 8.539273300740195e-05, + "loss": 0.0461, + "step": 14360 + }, + { + "grad_norm": 0.19776326417922974, + "learning_rate": 8.536936643179313e-05, + "loss": 0.0498, + "step": 14370 + }, + { + "grad_norm": 0.25964123010635376, + "learning_rate": 8.534598438442179e-05, + "loss": 0.047, + "step": 14380 + }, + { + "grad_norm": 0.2804294526576996, + "learning_rate": 8.532258687551603e-05, + "loss": 0.0558, + "step": 14390 + }, + { + "grad_norm": 0.22603479027748108, + "learning_rate": 8.529917391531071e-05, + "loss": 0.0474, + "step": 14400 + }, + { + "grad_norm": 0.24700140953063965, + "learning_rate": 8.527574551404747e-05, + "loss": 0.043, + "step": 14410 + }, + { + "grad_norm": 0.24795480072498322, + "learning_rate": 8.525230168197468e-05, + "loss": 0.0427, + "step": 14420 + }, + { + "grad_norm": 0.30368635058403015, + "learning_rate": 8.522884242934745e-05, + "loss": 0.0417, + "step": 14430 + }, + { + "grad_norm": 0.27224308252334595, + "learning_rate": 8.520536776642768e-05, + "loss": 0.0429, + "step": 14440 + }, + { + "grad_norm": 0.23862966895103455, + "learning_rate": 8.5181877703484e-05, + "loss": 0.0452, + "step": 14450 + }, + { + "grad_norm": 0.8169718980789185, + "learning_rate": 8.51583722507917e-05, + "loss": 0.0445, + "step": 14460 + }, + { + "grad_norm": 0.25850367546081543, + "learning_rate": 8.513485141863293e-05, + "loss": 0.0488, + "step": 14470 + }, + { + "grad_norm": 0.2603486180305481, + "learning_rate": 8.511131521729647e-05, + "loss": 0.0433, + "step": 14480 + }, + { + "grad_norm": 0.2850497364997864, + "learning_rate": 8.508776365707787e-05, + "loss": 0.0478, + "step": 14490 + }, + { + "grad_norm": 0.1778164654970169, + "learning_rate": 8.506419674827934e-05, + "loss": 0.0547, + "step": 14500 + }, + { + "grad_norm": 0.2103332132101059, + "learning_rate": 8.50406145012099e-05, + "loss": 0.0493, + "step": 14510 + }, + { + "grad_norm": 0.19832666218280792, + "learning_rate": 8.501701692618519e-05, + "loss": 0.0423, + "step": 14520 + }, + { + "grad_norm": 0.2751433253288269, + "learning_rate": 8.499340403352761e-05, + "loss": 0.0589, + "step": 14530 + }, + { + "grad_norm": 0.2020551860332489, + "learning_rate": 8.496977583356623e-05, + "loss": 0.0467, + "step": 14540 + }, + { + "grad_norm": 0.20921841263771057, + "learning_rate": 8.494613233663684e-05, + "loss": 0.046, + "step": 14550 + }, + { + "grad_norm": 0.20969167351722717, + "learning_rate": 8.492247355308189e-05, + "loss": 0.0431, + "step": 14560 + }, + { + "grad_norm": 0.28741535544395447, + "learning_rate": 8.489879949325056e-05, + "loss": 0.0423, + "step": 14570 + }, + { + "grad_norm": 0.429153710603714, + "learning_rate": 8.487511016749868e-05, + "loss": 0.0481, + "step": 14580 + }, + { + "grad_norm": 0.24654732644557953, + "learning_rate": 8.485140558618874e-05, + "loss": 0.053, + "step": 14590 + }, + { + "grad_norm": 0.6828926801681519, + "learning_rate": 8.482768575968995e-05, + "loss": 0.0539, + "step": 14600 + }, + { + "grad_norm": 0.2726950943470001, + "learning_rate": 8.480395069837818e-05, + "loss": 0.0422, + "step": 14610 + }, + { + "grad_norm": 0.2835889160633087, + "learning_rate": 8.478020041263595e-05, + "loss": 0.0447, + "step": 14620 + }, + { + "grad_norm": 0.30381447076797485, + "learning_rate": 8.475643491285242e-05, + "loss": 0.0463, + "step": 14630 + }, + { + "grad_norm": 0.2802698314189911, + "learning_rate": 8.473265420942345e-05, + "loss": 0.0385, + "step": 14640 + }, + { + "grad_norm": 0.3217429220676422, + "learning_rate": 8.470885831275151e-05, + "loss": 0.0455, + "step": 14650 + }, + { + "grad_norm": 0.22375494241714478, + "learning_rate": 8.468504723324574e-05, + "loss": 0.0474, + "step": 14660 + }, + { + "grad_norm": 0.3101956248283386, + "learning_rate": 8.466122098132193e-05, + "loss": 0.0516, + "step": 14670 + }, + { + "grad_norm": 0.18240323662757874, + "learning_rate": 8.463737956740245e-05, + "loss": 0.0435, + "step": 14680 + }, + { + "grad_norm": 0.3390025496482849, + "learning_rate": 8.461352300191639e-05, + "loss": 0.0501, + "step": 14690 + }, + { + "grad_norm": 0.36219367384910583, + "learning_rate": 8.45896512952994e-05, + "loss": 0.0516, + "step": 14700 + }, + { + "grad_norm": 0.25984299182891846, + "learning_rate": 8.456576445799377e-05, + "loss": 0.0559, + "step": 14710 + }, + { + "grad_norm": 0.19381801784038544, + "learning_rate": 8.454186250044844e-05, + "loss": 0.0492, + "step": 14720 + }, + { + "grad_norm": 0.23419189453125, + "learning_rate": 8.451794543311892e-05, + "loss": 0.0567, + "step": 14730 + }, + { + "grad_norm": 0.18677419424057007, + "learning_rate": 8.449401326646736e-05, + "loss": 0.0528, + "step": 14740 + }, + { + "grad_norm": 0.4057401120662689, + "learning_rate": 8.447006601096248e-05, + "loss": 0.053, + "step": 14750 + }, + { + "grad_norm": 0.17777937650680542, + "learning_rate": 8.444610367707964e-05, + "loss": 0.0486, + "step": 14760 + }, + { + "grad_norm": 0.41457298398017883, + "learning_rate": 8.442212627530078e-05, + "loss": 0.0514, + "step": 14770 + }, + { + "grad_norm": 0.2982412874698639, + "learning_rate": 8.439813381611441e-05, + "loss": 0.0488, + "step": 14780 + }, + { + "grad_norm": 0.23484905064105988, + "learning_rate": 8.437412631001567e-05, + "loss": 0.0506, + "step": 14790 + }, + { + "grad_norm": 0.3151070773601532, + "learning_rate": 8.435010376750626e-05, + "loss": 0.0546, + "step": 14800 + }, + { + "grad_norm": 0.2916225790977478, + "learning_rate": 8.432606619909442e-05, + "loss": 0.0456, + "step": 14810 + }, + { + "grad_norm": 0.22510762512683868, + "learning_rate": 8.430201361529506e-05, + "loss": 0.0527, + "step": 14820 + }, + { + "grad_norm": 0.24267661571502686, + "learning_rate": 8.427794602662954e-05, + "loss": 0.0543, + "step": 14830 + }, + { + "grad_norm": 0.30495402216911316, + "learning_rate": 8.425386344362586e-05, + "loss": 0.0565, + "step": 14840 + }, + { + "grad_norm": 0.27054378390312195, + "learning_rate": 8.422976587681859e-05, + "loss": 0.0561, + "step": 14850 + }, + { + "grad_norm": 0.28912046551704407, + "learning_rate": 8.42056533367488e-05, + "loss": 0.0463, + "step": 14860 + }, + { + "grad_norm": 0.2332920879125595, + "learning_rate": 8.41815258339641e-05, + "loss": 0.0449, + "step": 14870 + }, + { + "grad_norm": 0.5779340863227844, + "learning_rate": 8.415738337901874e-05, + "loss": 0.0535, + "step": 14880 + }, + { + "grad_norm": 0.2679862976074219, + "learning_rate": 8.413322598247342e-05, + "loss": 0.0506, + "step": 14890 + }, + { + "grad_norm": 0.40776756405830383, + "learning_rate": 8.41090536548954e-05, + "loss": 0.0481, + "step": 14900 + }, + { + "grad_norm": 0.2507895529270172, + "learning_rate": 8.408486640685849e-05, + "loss": 0.0475, + "step": 14910 + }, + { + "grad_norm": 0.24625568091869354, + "learning_rate": 8.4060664248943e-05, + "loss": 0.0468, + "step": 14920 + }, + { + "grad_norm": 0.3376091718673706, + "learning_rate": 8.40364471917358e-05, + "loss": 0.0475, + "step": 14930 + }, + { + "grad_norm": 0.32235807180404663, + "learning_rate": 8.401221524583024e-05, + "loss": 0.0563, + "step": 14940 + }, + { + "grad_norm": 0.3673151433467865, + "learning_rate": 8.398796842182619e-05, + "loss": 0.0607, + "step": 14950 + }, + { + "grad_norm": 0.251372367143631, + "learning_rate": 8.396370673033006e-05, + "loss": 0.0504, + "step": 14960 + }, + { + "grad_norm": 0.2154448926448822, + "learning_rate": 8.39394301819547e-05, + "loss": 0.049, + "step": 14970 + }, + { + "grad_norm": 0.20977380871772766, + "learning_rate": 8.391513878731949e-05, + "loss": 0.0456, + "step": 14980 + }, + { + "grad_norm": 0.3409726321697235, + "learning_rate": 8.389083255705037e-05, + "loss": 0.0485, + "step": 14990 + }, + { + "grad_norm": 0.6285210847854614, + "learning_rate": 8.386651150177968e-05, + "loss": 0.0558, + "step": 15000 + }, + { + "grad_norm": 0.22602251172065735, + "learning_rate": 8.384217563214627e-05, + "loss": 0.0464, + "step": 15010 + }, + { + "grad_norm": 0.2808140814304352, + "learning_rate": 8.381782495879546e-05, + "loss": 0.0459, + "step": 15020 + }, + { + "grad_norm": 0.2126108855009079, + "learning_rate": 8.37934594923791e-05, + "loss": 0.0467, + "step": 15030 + }, + { + "grad_norm": 0.39669474959373474, + "learning_rate": 8.376907924355546e-05, + "loss": 0.0573, + "step": 15040 + }, + { + "grad_norm": 0.31097763776779175, + "learning_rate": 8.374468422298926e-05, + "loss": 0.0583, + "step": 15050 + }, + { + "grad_norm": 0.5381788611412048, + "learning_rate": 8.372027444135176e-05, + "loss": 0.0591, + "step": 15060 + }, + { + "grad_norm": 0.22570443153381348, + "learning_rate": 8.36958499093206e-05, + "loss": 0.0478, + "step": 15070 + }, + { + "grad_norm": 0.23430952429771423, + "learning_rate": 8.367141063757988e-05, + "loss": 0.0474, + "step": 15080 + }, + { + "grad_norm": 0.1796354204416275, + "learning_rate": 8.364695663682022e-05, + "loss": 0.0518, + "step": 15090 + }, + { + "grad_norm": 0.2122570127248764, + "learning_rate": 8.362248791773858e-05, + "loss": 0.0479, + "step": 15100 + }, + { + "grad_norm": 0.3130631148815155, + "learning_rate": 8.359800449103842e-05, + "loss": 0.0509, + "step": 15110 + }, + { + "grad_norm": 0.28079769015312195, + "learning_rate": 8.357350636742967e-05, + "loss": 0.0525, + "step": 15120 + }, + { + "grad_norm": 0.2873387038707733, + "learning_rate": 8.354899355762858e-05, + "loss": 0.049, + "step": 15130 + }, + { + "grad_norm": 0.1790793389081955, + "learning_rate": 8.352446607235791e-05, + "loss": 0.0394, + "step": 15140 + }, + { + "grad_norm": 0.4895153343677521, + "learning_rate": 8.349992392234683e-05, + "loss": 0.0545, + "step": 15150 + }, + { + "grad_norm": 0.20546014606952667, + "learning_rate": 8.347536711833088e-05, + "loss": 0.0491, + "step": 15160 + }, + { + "grad_norm": 0.26383155584335327, + "learning_rate": 8.345079567105206e-05, + "loss": 0.0497, + "step": 15170 + }, + { + "grad_norm": 0.2700273394584656, + "learning_rate": 8.342620959125875e-05, + "loss": 0.0501, + "step": 15180 + }, + { + "grad_norm": 0.28833016753196716, + "learning_rate": 8.340160888970574e-05, + "loss": 0.0471, + "step": 15190 + }, + { + "grad_norm": 0.23644377291202545, + "learning_rate": 8.337699357715422e-05, + "loss": 0.0577, + "step": 15200 + }, + { + "grad_norm": 0.16979679465293884, + "learning_rate": 8.335236366437174e-05, + "loss": 0.047, + "step": 15210 + }, + { + "grad_norm": 0.4880257844924927, + "learning_rate": 8.332771916213227e-05, + "loss": 0.0521, + "step": 15220 + }, + { + "grad_norm": 0.30424660444259644, + "learning_rate": 8.330306008121618e-05, + "loss": 0.0462, + "step": 15230 + }, + { + "grad_norm": 0.25751152634620667, + "learning_rate": 8.327838643241014e-05, + "loss": 0.0486, + "step": 15240 + }, + { + "grad_norm": 0.20192283391952515, + "learning_rate": 8.32536982265073e-05, + "loss": 0.0619, + "step": 15250 + }, + { + "grad_norm": 0.3020082712173462, + "learning_rate": 8.32289954743071e-05, + "loss": 0.0564, + "step": 15260 + }, + { + "grad_norm": 0.6136850118637085, + "learning_rate": 8.320427818661532e-05, + "loss": 0.0494, + "step": 15270 + }, + { + "grad_norm": 0.3004221022129059, + "learning_rate": 8.317954637424419e-05, + "loss": 0.0478, + "step": 15280 + }, + { + "grad_norm": 0.2937559187412262, + "learning_rate": 8.315480004801221e-05, + "loss": 0.0524, + "step": 15290 + }, + { + "grad_norm": 0.41092967987060547, + "learning_rate": 8.31300392187443e-05, + "loss": 0.0501, + "step": 15300 + }, + { + "grad_norm": 0.20701701939105988, + "learning_rate": 8.31052638972717e-05, + "loss": 0.0506, + "step": 15310 + }, + { + "grad_norm": 0.37486034631729126, + "learning_rate": 8.30804740944319e-05, + "loss": 0.0466, + "step": 15320 + }, + { + "grad_norm": 0.3188874423503876, + "learning_rate": 8.305566982106887e-05, + "loss": 0.0493, + "step": 15330 + }, + { + "grad_norm": 0.33146581053733826, + "learning_rate": 8.303085108803282e-05, + "loss": 0.0537, + "step": 15340 + }, + { + "grad_norm": 0.19872762262821198, + "learning_rate": 8.300601790618032e-05, + "loss": 0.051, + "step": 15350 + }, + { + "grad_norm": 0.3138678967952728, + "learning_rate": 8.298117028637423e-05, + "loss": 0.05, + "step": 15360 + }, + { + "grad_norm": 0.2681148052215576, + "learning_rate": 8.295630823948375e-05, + "loss": 0.0549, + "step": 15370 + }, + { + "grad_norm": 0.28993672132492065, + "learning_rate": 8.293143177638438e-05, + "loss": 0.0471, + "step": 15380 + }, + { + "grad_norm": 0.2755443751811981, + "learning_rate": 8.290654090795795e-05, + "loss": 0.0461, + "step": 15390 + }, + { + "grad_norm": 0.2955344319343567, + "learning_rate": 8.288163564509253e-05, + "loss": 0.0505, + "step": 15400 + }, + { + "grad_norm": 0.27887973189353943, + "learning_rate": 8.285671599868257e-05, + "loss": 0.0408, + "step": 15410 + }, + { + "grad_norm": 0.47870227694511414, + "learning_rate": 8.283178197962875e-05, + "loss": 0.0548, + "step": 15420 + }, + { + "grad_norm": 0.2513674199581146, + "learning_rate": 8.280683359883808e-05, + "loss": 0.0573, + "step": 15430 + }, + { + "grad_norm": 0.23418642580509186, + "learning_rate": 8.278187086722377e-05, + "loss": 0.0474, + "step": 15440 + }, + { + "grad_norm": 0.3008323311805725, + "learning_rate": 8.275689379570544e-05, + "loss": 0.0537, + "step": 15450 + }, + { + "grad_norm": 0.15333348512649536, + "learning_rate": 8.273190239520887e-05, + "loss": 0.0492, + "step": 15460 + }, + { + "grad_norm": 0.25308820605278015, + "learning_rate": 8.270689667666616e-05, + "loss": 0.0402, + "step": 15470 + }, + { + "grad_norm": 0.2593865990638733, + "learning_rate": 8.268187665101563e-05, + "loss": 0.0534, + "step": 15480 + }, + { + "grad_norm": 0.2491397261619568, + "learning_rate": 8.265684232920194e-05, + "loss": 0.0497, + "step": 15490 + }, + { + "grad_norm": 0.292310893535614, + "learning_rate": 8.263179372217592e-05, + "loss": 0.0444, + "step": 15500 + }, + { + "grad_norm": 0.2212711125612259, + "learning_rate": 8.260673084089468e-05, + "loss": 0.0496, + "step": 15510 + }, + { + "grad_norm": 0.4310258626937866, + "learning_rate": 8.25816536963216e-05, + "loss": 0.0461, + "step": 15520 + }, + { + "grad_norm": 0.19070219993591309, + "learning_rate": 8.255656229942626e-05, + "loss": 0.0432, + "step": 15530 + }, + { + "grad_norm": 0.351921409368515, + "learning_rate": 8.253145666118447e-05, + "loss": 0.0479, + "step": 15540 + }, + { + "grad_norm": 0.23553447425365448, + "learning_rate": 8.250633679257834e-05, + "loss": 0.0421, + "step": 15550 + }, + { + "grad_norm": 0.386153906583786, + "learning_rate": 8.24812027045961e-05, + "loss": 0.0531, + "step": 15560 + }, + { + "grad_norm": 0.24609176814556122, + "learning_rate": 8.245605440823226e-05, + "loss": 0.0502, + "step": 15570 + }, + { + "grad_norm": 0.45675793290138245, + "learning_rate": 8.243089191448759e-05, + "loss": 0.0526, + "step": 15580 + }, + { + "grad_norm": 0.22116690874099731, + "learning_rate": 8.240571523436896e-05, + "loss": 0.0449, + "step": 15590 + }, + { + "grad_norm": 0.26174142956733704, + "learning_rate": 8.238052437888955e-05, + "loss": 0.0434, + "step": 15600 + }, + { + "grad_norm": 0.14099019765853882, + "learning_rate": 8.235531935906866e-05, + "loss": 0.0542, + "step": 15610 + }, + { + "grad_norm": 0.1783047616481781, + "learning_rate": 8.233010018593185e-05, + "loss": 0.0427, + "step": 15620 + }, + { + "grad_norm": 0.24036858975887299, + "learning_rate": 8.230486687051082e-05, + "loss": 0.046, + "step": 15630 + }, + { + "grad_norm": 0.21231655776500702, + "learning_rate": 8.22796194238435e-05, + "loss": 0.0483, + "step": 15640 + }, + { + "grad_norm": 0.26541340351104736, + "learning_rate": 8.225435785697397e-05, + "loss": 0.046, + "step": 15650 + }, + { + "grad_norm": 0.3607955574989319, + "learning_rate": 8.222908218095252e-05, + "loss": 0.0521, + "step": 15660 + }, + { + "grad_norm": 0.326739639043808, + "learning_rate": 8.220379240683553e-05, + "loss": 0.0482, + "step": 15670 + }, + { + "grad_norm": 0.243887796998024, + "learning_rate": 8.217848854568566e-05, + "loss": 0.04, + "step": 15680 + }, + { + "grad_norm": 0.22504648566246033, + "learning_rate": 8.215317060857167e-05, + "loss": 0.0485, + "step": 15690 + }, + { + "grad_norm": 0.5301767587661743, + "learning_rate": 8.212783860656847e-05, + "loss": 0.0481, + "step": 15700 + }, + { + "grad_norm": 0.3889665901660919, + "learning_rate": 8.210249255075717e-05, + "loss": 0.0496, + "step": 15710 + }, + { + "grad_norm": 0.17289559543132782, + "learning_rate": 8.207713245222496e-05, + "loss": 0.0423, + "step": 15720 + }, + { + "grad_norm": 0.37787705659866333, + "learning_rate": 8.205175832206522e-05, + "loss": 0.0529, + "step": 15730 + }, + { + "grad_norm": 0.32995274662971497, + "learning_rate": 8.202637017137747e-05, + "loss": 0.0444, + "step": 15740 + }, + { + "grad_norm": 0.22739599645137787, + "learning_rate": 8.200096801126734e-05, + "loss": 0.0475, + "step": 15750 + }, + { + "grad_norm": 0.15813696384429932, + "learning_rate": 8.197555185284658e-05, + "loss": 0.0467, + "step": 15760 + }, + { + "grad_norm": 0.2213938981294632, + "learning_rate": 8.19501217072331e-05, + "loss": 0.0484, + "step": 15770 + }, + { + "grad_norm": 0.21349665522575378, + "learning_rate": 8.192467758555091e-05, + "loss": 0.049, + "step": 15780 + }, + { + "grad_norm": 0.24196334183216095, + "learning_rate": 8.189921949893014e-05, + "loss": 0.0549, + "step": 15790 + }, + { + "grad_norm": 1.1211202144622803, + "learning_rate": 8.187374745850702e-05, + "loss": 0.0519, + "step": 15800 + }, + { + "grad_norm": 2.476567029953003, + "learning_rate": 8.184826147542387e-05, + "loss": 0.0431, + "step": 15810 + }, + { + "grad_norm": 0.2487039715051651, + "learning_rate": 8.18227615608291e-05, + "loss": 0.0498, + "step": 15820 + }, + { + "grad_norm": 0.22460639476776123, + "learning_rate": 8.179724772587732e-05, + "loss": 0.0496, + "step": 15830 + }, + { + "grad_norm": 0.343122273683548, + "learning_rate": 8.177171998172908e-05, + "loss": 0.0478, + "step": 15840 + }, + { + "grad_norm": 0.3638080656528473, + "learning_rate": 8.174617833955112e-05, + "loss": 0.0487, + "step": 15850 + }, + { + "grad_norm": 0.17759139835834503, + "learning_rate": 8.172062281051618e-05, + "loss": 0.0458, + "step": 15860 + }, + { + "grad_norm": 0.3953992426395416, + "learning_rate": 8.169505340580316e-05, + "loss": 0.0502, + "step": 15870 + }, + { + "grad_norm": 0.3295949399471283, + "learning_rate": 8.166947013659696e-05, + "loss": 0.0504, + "step": 15880 + }, + { + "grad_norm": 0.21673160791397095, + "learning_rate": 8.164387301408858e-05, + "loss": 0.0488, + "step": 15890 + }, + { + "grad_norm": 0.26426154375076294, + "learning_rate": 8.161826204947507e-05, + "loss": 0.0405, + "step": 15900 + }, + { + "grad_norm": 0.23368501663208008, + "learning_rate": 8.159263725395954e-05, + "loss": 0.0456, + "step": 15910 + }, + { + "grad_norm": 0.22134003043174744, + "learning_rate": 8.156699863875115e-05, + "loss": 0.0457, + "step": 15920 + }, + { + "grad_norm": 0.15985193848609924, + "learning_rate": 8.154134621506508e-05, + "loss": 0.0512, + "step": 15930 + }, + { + "grad_norm": 0.22489361464977264, + "learning_rate": 8.15156799941226e-05, + "loss": 0.0466, + "step": 15940 + }, + { + "grad_norm": 0.1829720437526703, + "learning_rate": 8.148999998715096e-05, + "loss": 0.0446, + "step": 15950 + }, + { + "grad_norm": 0.2119031697511673, + "learning_rate": 8.146430620538348e-05, + "loss": 0.0445, + "step": 15960 + }, + { + "grad_norm": 0.22281813621520996, + "learning_rate": 8.143859866005949e-05, + "loss": 0.0451, + "step": 15970 + }, + { + "grad_norm": 0.16508421301841736, + "learning_rate": 8.141287736242438e-05, + "loss": 0.0424, + "step": 15980 + }, + { + "grad_norm": 0.3511927127838135, + "learning_rate": 8.138714232372945e-05, + "loss": 0.0476, + "step": 15990 + }, + { + "grad_norm": 0.26984071731567383, + "learning_rate": 8.136139355523213e-05, + "loss": 0.0455, + "step": 16000 + }, + { + "grad_norm": 0.32539334893226624, + "learning_rate": 8.13356310681958e-05, + "loss": 0.0473, + "step": 16010 + }, + { + "grad_norm": 0.2754739820957184, + "learning_rate": 8.130985487388982e-05, + "loss": 0.0408, + "step": 16020 + }, + { + "grad_norm": 0.24549642205238342, + "learning_rate": 8.12840649835896e-05, + "loss": 0.0465, + "step": 16030 + }, + { + "grad_norm": 0.3046579360961914, + "learning_rate": 8.125826140857653e-05, + "loss": 0.0408, + "step": 16040 + }, + { + "grad_norm": 0.15913325548171997, + "learning_rate": 8.123244416013792e-05, + "loss": 0.0388, + "step": 16050 + }, + { + "grad_norm": 0.260062575340271, + "learning_rate": 8.120661324956716e-05, + "loss": 0.0404, + "step": 16060 + }, + { + "grad_norm": 0.2077709436416626, + "learning_rate": 8.118076868816353e-05, + "loss": 0.0405, + "step": 16070 + }, + { + "grad_norm": 0.4824502170085907, + "learning_rate": 8.115491048723234e-05, + "loss": 0.0461, + "step": 16080 + }, + { + "grad_norm": 0.3155345022678375, + "learning_rate": 8.112903865808486e-05, + "loss": 0.046, + "step": 16090 + }, + { + "grad_norm": 0.2631441652774811, + "learning_rate": 8.110315321203826e-05, + "loss": 0.0454, + "step": 16100 + }, + { + "grad_norm": 0.28251734375953674, + "learning_rate": 8.107725416041575e-05, + "loss": 0.0451, + "step": 16110 + }, + { + "grad_norm": 0.2745240032672882, + "learning_rate": 8.105134151454645e-05, + "loss": 0.0534, + "step": 16120 + }, + { + "grad_norm": 0.4143787622451782, + "learning_rate": 8.102541528576544e-05, + "loss": 0.0474, + "step": 16130 + }, + { + "grad_norm": 0.3000306487083435, + "learning_rate": 8.099947548541371e-05, + "loss": 0.0439, + "step": 16140 + }, + { + "grad_norm": 0.18583226203918457, + "learning_rate": 8.097352212483824e-05, + "loss": 0.0464, + "step": 16150 + }, + { + "grad_norm": 0.2741362154483795, + "learning_rate": 8.094755521539187e-05, + "loss": 0.0468, + "step": 16160 + }, + { + "grad_norm": 0.2543449401855469, + "learning_rate": 8.092157476843346e-05, + "loss": 0.0427, + "step": 16170 + }, + { + "grad_norm": 0.3731735646724701, + "learning_rate": 8.089558079532769e-05, + "loss": 0.0426, + "step": 16180 + }, + { + "grad_norm": 0.21744950115680695, + "learning_rate": 8.086957330744522e-05, + "loss": 0.044, + "step": 16190 + }, + { + "grad_norm": 0.3252987861633301, + "learning_rate": 8.084355231616265e-05, + "loss": 0.0462, + "step": 16200 + }, + { + "grad_norm": 0.23801885545253754, + "learning_rate": 8.08175178328624e-05, + "loss": 0.0516, + "step": 16210 + }, + { + "grad_norm": 0.5206947326660156, + "learning_rate": 8.079146986893284e-05, + "loss": 0.0438, + "step": 16220 + }, + { + "grad_norm": 0.3513336479663849, + "learning_rate": 8.076540843576826e-05, + "loss": 0.0561, + "step": 16230 + }, + { + "grad_norm": 0.3107638657093048, + "learning_rate": 8.073933354476879e-05, + "loss": 0.051, + "step": 16240 + }, + { + "grad_norm": 0.42434656620025635, + "learning_rate": 8.07132452073405e-05, + "loss": 0.0503, + "step": 16250 + }, + { + "grad_norm": 0.25476688146591187, + "learning_rate": 8.068714343489529e-05, + "loss": 0.0497, + "step": 16260 + }, + { + "grad_norm": 0.19115327298641205, + "learning_rate": 8.066102823885097e-05, + "loss": 0.0476, + "step": 16270 + }, + { + "grad_norm": 0.17581267654895782, + "learning_rate": 8.063489963063125e-05, + "loss": 0.0434, + "step": 16280 + }, + { + "grad_norm": 0.17504671216011047, + "learning_rate": 8.06087576216656e-05, + "loss": 0.0442, + "step": 16290 + }, + { + "grad_norm": 0.24983634054660797, + "learning_rate": 8.058260222338946e-05, + "loss": 0.0432, + "step": 16300 + }, + { + "grad_norm": 0.30978214740753174, + "learning_rate": 8.05564334472441e-05, + "loss": 0.0506, + "step": 16310 + }, + { + "grad_norm": 0.36007800698280334, + "learning_rate": 8.053025130467664e-05, + "loss": 0.0459, + "step": 16320 + }, + { + "grad_norm": 0.21634210646152496, + "learning_rate": 8.050405580714e-05, + "loss": 0.0485, + "step": 16330 + }, + { + "grad_norm": 0.22532159090042114, + "learning_rate": 8.047784696609301e-05, + "loss": 0.0462, + "step": 16340 + }, + { + "grad_norm": 0.22443729639053345, + "learning_rate": 8.045162479300029e-05, + "loss": 0.0409, + "step": 16350 + }, + { + "grad_norm": 0.20016464591026306, + "learning_rate": 8.042538929933232e-05, + "loss": 0.0458, + "step": 16360 + }, + { + "grad_norm": 0.21439528465270996, + "learning_rate": 8.039914049656539e-05, + "loss": 0.0449, + "step": 16370 + }, + { + "grad_norm": 0.35193634033203125, + "learning_rate": 8.037287839618163e-05, + "loss": 0.0464, + "step": 16380 + }, + { + "grad_norm": 0.1642657071352005, + "learning_rate": 8.034660300966898e-05, + "loss": 0.0462, + "step": 16390 + }, + { + "grad_norm": 0.2502826452255249, + "learning_rate": 8.032031434852115e-05, + "loss": 0.0482, + "step": 16400 + }, + { + "grad_norm": 0.2697552740573883, + "learning_rate": 8.029401242423773e-05, + "loss": 0.0478, + "step": 16410 + }, + { + "grad_norm": 0.2180778682231903, + "learning_rate": 8.026769724832406e-05, + "loss": 0.0465, + "step": 16420 + }, + { + "grad_norm": 0.24381598830223083, + "learning_rate": 8.02413688322913e-05, + "loss": 0.0432, + "step": 16430 + }, + { + "grad_norm": 0.2162577509880066, + "learning_rate": 8.021502718765639e-05, + "loss": 0.0455, + "step": 16440 + }, + { + "grad_norm": 0.5961136221885681, + "learning_rate": 8.018867232594204e-05, + "loss": 0.0472, + "step": 16450 + }, + { + "grad_norm": 0.28763073682785034, + "learning_rate": 8.016230425867679e-05, + "loss": 0.0434, + "step": 16460 + }, + { + "grad_norm": 0.31491199135780334, + "learning_rate": 8.013592299739494e-05, + "loss": 0.0449, + "step": 16470 + }, + { + "grad_norm": 0.2644865810871124, + "learning_rate": 8.01095285536365e-05, + "loss": 0.0504, + "step": 16480 + }, + { + "grad_norm": 0.23660212755203247, + "learning_rate": 8.008312093894735e-05, + "loss": 0.0505, + "step": 16490 + }, + { + "grad_norm": 0.24279020726680756, + "learning_rate": 8.005670016487903e-05, + "loss": 0.044, + "step": 16500 + }, + { + "grad_norm": 0.23892426490783691, + "learning_rate": 8.003026624298892e-05, + "loss": 0.0449, + "step": 16510 + }, + { + "grad_norm": 0.2215159386396408, + "learning_rate": 8.00038191848401e-05, + "loss": 0.0472, + "step": 16520 + }, + { + "grad_norm": 0.27212369441986084, + "learning_rate": 7.997735900200142e-05, + "loss": 0.0451, + "step": 16530 + }, + { + "grad_norm": 0.26074928045272827, + "learning_rate": 7.995088570604745e-05, + "loss": 0.0426, + "step": 16540 + }, + { + "grad_norm": 0.29725363850593567, + "learning_rate": 7.992439930855852e-05, + "loss": 0.0433, + "step": 16550 + }, + { + "grad_norm": 0.3003670275211334, + "learning_rate": 7.989789982112067e-05, + "loss": 0.0515, + "step": 16560 + }, + { + "grad_norm": 0.2422982007265091, + "learning_rate": 7.987138725532572e-05, + "loss": 0.0479, + "step": 16570 + }, + { + "grad_norm": 0.2696502208709717, + "learning_rate": 7.98448616227711e-05, + "loss": 0.0562, + "step": 16580 + }, + { + "grad_norm": 0.2325446456670761, + "learning_rate": 7.981832293506004e-05, + "loss": 0.0456, + "step": 16590 + }, + { + "grad_norm": 0.28643763065338135, + "learning_rate": 7.97917712038015e-05, + "loss": 0.0453, + "step": 16600 + }, + { + "grad_norm": 0.3607753813266754, + "learning_rate": 7.976520644061007e-05, + "loss": 0.043, + "step": 16610 + }, + { + "grad_norm": 0.4180326759815216, + "learning_rate": 7.973862865710611e-05, + "loss": 0.0416, + "step": 16620 + }, + { + "grad_norm": 0.2718431055545807, + "learning_rate": 7.971203786491563e-05, + "loss": 0.0447, + "step": 16630 + }, + { + "grad_norm": 0.2836783528327942, + "learning_rate": 7.968543407567034e-05, + "loss": 0.0505, + "step": 16640 + }, + { + "grad_norm": 0.2728825807571411, + "learning_rate": 7.965881730100765e-05, + "loss": 0.0498, + "step": 16650 + }, + { + "grad_norm": 0.25224876403808594, + "learning_rate": 7.963218755257064e-05, + "loss": 0.0441, + "step": 16660 + }, + { + "grad_norm": 0.2630907893180847, + "learning_rate": 7.960554484200807e-05, + "loss": 0.0548, + "step": 16670 + }, + { + "grad_norm": 0.4577437937259674, + "learning_rate": 7.957888918097436e-05, + "loss": 0.0463, + "step": 16680 + }, + { + "grad_norm": 0.2750319838523865, + "learning_rate": 7.955222058112962e-05, + "loss": 0.0487, + "step": 16690 + }, + { + "grad_norm": 0.22793051600456238, + "learning_rate": 7.952553905413955e-05, + "loss": 0.046, + "step": 16700 + }, + { + "grad_norm": 0.2269093096256256, + "learning_rate": 7.949884461167563e-05, + "loss": 0.0452, + "step": 16710 + }, + { + "grad_norm": 0.28408628702163696, + "learning_rate": 7.947213726541484e-05, + "loss": 0.0563, + "step": 16720 + }, + { + "grad_norm": 0.2279474437236786, + "learning_rate": 7.944541702703996e-05, + "loss": 0.049, + "step": 16730 + }, + { + "grad_norm": 0.2661546468734741, + "learning_rate": 7.941868390823928e-05, + "loss": 0.0443, + "step": 16740 + }, + { + "grad_norm": 0.22768355906009674, + "learning_rate": 7.939193792070677e-05, + "loss": 0.0453, + "step": 16750 + }, + { + "grad_norm": 0.24266907572746277, + "learning_rate": 7.936517907614207e-05, + "loss": 0.0423, + "step": 16760 + }, + { + "grad_norm": 0.15507596731185913, + "learning_rate": 7.933840738625036e-05, + "loss": 0.0424, + "step": 16770 + }, + { + "grad_norm": 0.24810543656349182, + "learning_rate": 7.931162286274252e-05, + "loss": 0.0477, + "step": 16780 + }, + { + "grad_norm": 0.36286166310310364, + "learning_rate": 7.928482551733503e-05, + "loss": 0.042, + "step": 16790 + }, + { + "grad_norm": 0.30641835927963257, + "learning_rate": 7.925801536174991e-05, + "loss": 0.0466, + "step": 16800 + }, + { + "grad_norm": 0.21632890403270721, + "learning_rate": 7.923119240771484e-05, + "loss": 0.0423, + "step": 16810 + }, + { + "grad_norm": 0.204432412981987, + "learning_rate": 7.920435666696313e-05, + "loss": 0.0424, + "step": 16820 + }, + { + "grad_norm": 0.299988329410553, + "learning_rate": 7.917750815123362e-05, + "loss": 0.0431, + "step": 16830 + }, + { + "grad_norm": 0.2159128189086914, + "learning_rate": 7.915064687227075e-05, + "loss": 0.0424, + "step": 16840 + }, + { + "grad_norm": 0.19131851196289062, + "learning_rate": 7.912377284182458e-05, + "loss": 0.0442, + "step": 16850 + }, + { + "grad_norm": 0.3547931909561157, + "learning_rate": 7.90968860716507e-05, + "loss": 0.0506, + "step": 16860 + }, + { + "grad_norm": 0.21776549518108368, + "learning_rate": 7.906998657351033e-05, + "loss": 0.0447, + "step": 16870 + }, + { + "grad_norm": 0.29826438426971436, + "learning_rate": 7.90430743591702e-05, + "loss": 0.0455, + "step": 16880 + }, + { + "grad_norm": 0.28762292861938477, + "learning_rate": 7.901614944040263e-05, + "loss": 0.0432, + "step": 16890 + }, + { + "grad_norm": 0.9428495764732361, + "learning_rate": 7.89892118289855e-05, + "loss": 0.0427, + "step": 16900 + }, + { + "grad_norm": 0.2525760233402252, + "learning_rate": 7.896226153670223e-05, + "loss": 0.0428, + "step": 16910 + }, + { + "grad_norm": 0.36690130829811096, + "learning_rate": 7.89352985753418e-05, + "loss": 0.0427, + "step": 16920 + }, + { + "grad_norm": 0.2508055567741394, + "learning_rate": 7.890832295669871e-05, + "loss": 0.0482, + "step": 16930 + }, + { + "grad_norm": 0.42276278138160706, + "learning_rate": 7.888133469257303e-05, + "loss": 0.0429, + "step": 16940 + }, + { + "grad_norm": 0.16246840357780457, + "learning_rate": 7.885433379477034e-05, + "loss": 0.0507, + "step": 16950 + }, + { + "grad_norm": 0.2189026027917862, + "learning_rate": 7.882732027510174e-05, + "loss": 0.0463, + "step": 16960 + }, + { + "grad_norm": 0.26692453026771545, + "learning_rate": 7.880029414538386e-05, + "loss": 0.0444, + "step": 16970 + }, + { + "grad_norm": 0.25436022877693176, + "learning_rate": 7.877325541743888e-05, + "loss": 0.0487, + "step": 16980 + }, + { + "grad_norm": 0.282093346118927, + "learning_rate": 7.874620410309443e-05, + "loss": 0.0514, + "step": 16990 + }, + { + "grad_norm": 0.42680996656417847, + "learning_rate": 7.871914021418368e-05, + "loss": 0.047, + "step": 17000 + }, + { + "grad_norm": 0.16526643931865692, + "learning_rate": 7.869206376254528e-05, + "loss": 0.0522, + "step": 17010 + }, + { + "grad_norm": 0.23375287652015686, + "learning_rate": 7.86649747600234e-05, + "loss": 0.0467, + "step": 17020 + }, + { + "grad_norm": 0.2718370854854584, + "learning_rate": 7.86378732184677e-05, + "loss": 0.0409, + "step": 17030 + }, + { + "grad_norm": 0.33902549743652344, + "learning_rate": 7.86107591497333e-05, + "loss": 0.0374, + "step": 17040 + }, + { + "grad_norm": 0.21357154846191406, + "learning_rate": 7.858363256568083e-05, + "loss": 0.0433, + "step": 17050 + }, + { + "grad_norm": 0.4044388234615326, + "learning_rate": 7.855649347817637e-05, + "loss": 0.0476, + "step": 17060 + }, + { + "grad_norm": 0.2756783366203308, + "learning_rate": 7.85293418990915e-05, + "loss": 0.0397, + "step": 17070 + }, + { + "grad_norm": 0.3501099944114685, + "learning_rate": 7.85021778403032e-05, + "loss": 0.0427, + "step": 17080 + }, + { + "grad_norm": 0.2302849292755127, + "learning_rate": 7.847500131369397e-05, + "loss": 0.0451, + "step": 17090 + }, + { + "grad_norm": 0.26437583565711975, + "learning_rate": 7.844781233115176e-05, + "loss": 0.0425, + "step": 17100 + }, + { + "grad_norm": 0.3003065884113312, + "learning_rate": 7.842061090456996e-05, + "loss": 0.0431, + "step": 17110 + }, + { + "grad_norm": 0.2548016309738159, + "learning_rate": 7.839339704584736e-05, + "loss": 0.0436, + "step": 17120 + }, + { + "grad_norm": 0.20567747950553894, + "learning_rate": 7.836617076688828e-05, + "loss": 0.0454, + "step": 17130 + }, + { + "grad_norm": 0.27523770928382874, + "learning_rate": 7.833893207960236e-05, + "loss": 0.0425, + "step": 17140 + }, + { + "grad_norm": 0.19109442830085754, + "learning_rate": 7.831168099590478e-05, + "loss": 0.045, + "step": 17150 + }, + { + "grad_norm": 0.1960945427417755, + "learning_rate": 7.828441752771607e-05, + "loss": 0.043, + "step": 17160 + }, + { + "grad_norm": 0.2562422752380371, + "learning_rate": 7.825714168696219e-05, + "loss": 0.0465, + "step": 17170 + }, + { + "grad_norm": 0.36145254969596863, + "learning_rate": 7.82298534855745e-05, + "loss": 0.0479, + "step": 17180 + }, + { + "grad_norm": 0.20211783051490784, + "learning_rate": 7.820255293548984e-05, + "loss": 0.0385, + "step": 17190 + }, + { + "grad_norm": 0.3345087170600891, + "learning_rate": 7.817524004865035e-05, + "loss": 0.0447, + "step": 17200 + }, + { + "grad_norm": 0.3105199635028839, + "learning_rate": 7.814791483700366e-05, + "loss": 0.0419, + "step": 17210 + }, + { + "grad_norm": 0.5463982224464417, + "learning_rate": 7.81205773125027e-05, + "loss": 0.0453, + "step": 17220 + }, + { + "grad_norm": 0.3952043652534485, + "learning_rate": 7.809322748710585e-05, + "loss": 0.0444, + "step": 17230 + }, + { + "grad_norm": 0.25046324729919434, + "learning_rate": 7.806586537277685e-05, + "loss": 0.0466, + "step": 17240 + }, + { + "grad_norm": 0.19596698880195618, + "learning_rate": 7.803849098148484e-05, + "loss": 0.0474, + "step": 17250 + }, + { + "grad_norm": 0.27947261929512024, + "learning_rate": 7.801110432520427e-05, + "loss": 0.0413, + "step": 17260 + }, + { + "grad_norm": 0.21539688110351562, + "learning_rate": 7.798370541591504e-05, + "loss": 0.043, + "step": 17270 + }, + { + "grad_norm": 0.2552337348461151, + "learning_rate": 7.795629426560232e-05, + "loss": 0.0406, + "step": 17280 + }, + { + "grad_norm": 0.24017810821533203, + "learning_rate": 7.792887088625671e-05, + "loss": 0.0453, + "step": 17290 + }, + { + "grad_norm": 0.23487193882465363, + "learning_rate": 7.790143528987412e-05, + "loss": 0.0416, + "step": 17300 + }, + { + "grad_norm": 0.12699605524539948, + "learning_rate": 7.787398748845578e-05, + "loss": 0.0409, + "step": 17310 + }, + { + "grad_norm": 0.19278188049793243, + "learning_rate": 7.784652749400834e-05, + "loss": 0.0371, + "step": 17320 + }, + { + "grad_norm": 0.2033521980047226, + "learning_rate": 7.78190553185437e-05, + "loss": 0.0401, + "step": 17330 + }, + { + "grad_norm": 0.25979408621788025, + "learning_rate": 7.779157097407915e-05, + "loss": 0.0454, + "step": 17340 + }, + { + "grad_norm": 0.2472003698348999, + "learning_rate": 7.776407447263725e-05, + "loss": 0.0411, + "step": 17350 + }, + { + "grad_norm": 0.19804500043392181, + "learning_rate": 7.773656582624593e-05, + "loss": 0.0504, + "step": 17360 + }, + { + "grad_norm": 0.24102944135665894, + "learning_rate": 7.770904504693837e-05, + "loss": 0.0428, + "step": 17370 + }, + { + "grad_norm": 0.31512728333473206, + "learning_rate": 7.768151214675314e-05, + "loss": 0.0441, + "step": 17380 + }, + { + "grad_norm": 0.16371285915374756, + "learning_rate": 7.765396713773402e-05, + "loss": 0.0416, + "step": 17390 + }, + { + "grad_norm": 0.22041228413581848, + "learning_rate": 7.762641003193017e-05, + "loss": 0.0456, + "step": 17400 + }, + { + "grad_norm": 0.22067894041538239, + "learning_rate": 7.759884084139597e-05, + "loss": 0.0369, + "step": 17410 + }, + { + "grad_norm": 0.1927630752325058, + "learning_rate": 7.757125957819116e-05, + "loss": 0.0433, + "step": 17420 + }, + { + "grad_norm": 0.39044779539108276, + "learning_rate": 7.754366625438069e-05, + "loss": 0.0375, + "step": 17430 + }, + { + "grad_norm": 0.4189770519733429, + "learning_rate": 7.751606088203483e-05, + "loss": 0.047, + "step": 17440 + }, + { + "grad_norm": 0.30137109756469727, + "learning_rate": 7.74884434732291e-05, + "loss": 0.0449, + "step": 17450 + }, + { + "grad_norm": 0.21325665712356567, + "learning_rate": 7.746081404004431e-05, + "loss": 0.0421, + "step": 17460 + }, + { + "grad_norm": 0.18796317279338837, + "learning_rate": 7.743317259456649e-05, + "loss": 0.0366, + "step": 17470 + }, + { + "grad_norm": 0.36630943417549133, + "learning_rate": 7.740551914888696e-05, + "loss": 0.0422, + "step": 17480 + }, + { + "grad_norm": 0.26569893956184387, + "learning_rate": 7.737785371510229e-05, + "loss": 0.0386, + "step": 17490 + }, + { + "grad_norm": 0.2646424174308777, + "learning_rate": 7.735017630531426e-05, + "loss": 0.0371, + "step": 17500 + }, + { + "grad_norm": 0.22034171223640442, + "learning_rate": 7.732248693162991e-05, + "loss": 0.0436, + "step": 17510 + }, + { + "grad_norm": 0.22573381662368774, + "learning_rate": 7.729478560616153e-05, + "loss": 0.0435, + "step": 17520 + }, + { + "grad_norm": 0.22049105167388916, + "learning_rate": 7.726707234102659e-05, + "loss": 0.0456, + "step": 17530 + }, + { + "grad_norm": 0.21868786215782166, + "learning_rate": 7.723934714834785e-05, + "loss": 0.0461, + "step": 17540 + }, + { + "grad_norm": 0.3446011245250702, + "learning_rate": 7.721161004025323e-05, + "loss": 0.0394, + "step": 17550 + }, + { + "grad_norm": 2.71877384185791, + "learning_rate": 7.718386102887588e-05, + "loss": 0.0499, + "step": 17560 + }, + { + "grad_norm": 0.2756086587905884, + "learning_rate": 7.715610012635418e-05, + "loss": 0.0478, + "step": 17570 + }, + { + "grad_norm": 0.24354103207588196, + "learning_rate": 7.712832734483166e-05, + "loss": 0.047, + "step": 17580 + }, + { + "grad_norm": 0.18910571932792664, + "learning_rate": 7.710054269645713e-05, + "loss": 0.0482, + "step": 17590 + }, + { + "grad_norm": 0.18501418828964233, + "learning_rate": 7.707274619338449e-05, + "loss": 0.0509, + "step": 17600 + }, + { + "grad_norm": 0.19310568273067474, + "learning_rate": 7.704493784777289e-05, + "loss": 0.0479, + "step": 17610 + }, + { + "grad_norm": 0.201610267162323, + "learning_rate": 7.701711767178665e-05, + "loss": 0.048, + "step": 17620 + }, + { + "grad_norm": 0.31395599246025085, + "learning_rate": 7.698928567759525e-05, + "loss": 0.0369, + "step": 17630 + }, + { + "grad_norm": 0.298191100358963, + "learning_rate": 7.696144187737335e-05, + "loss": 0.043, + "step": 17640 + }, + { + "grad_norm": 0.17872561514377594, + "learning_rate": 7.69335862833008e-05, + "loss": 0.0412, + "step": 17650 + }, + { + "grad_norm": 0.30401352047920227, + "learning_rate": 7.690571890756256e-05, + "loss": 0.0518, + "step": 17660 + }, + { + "grad_norm": 0.3510412275791168, + "learning_rate": 7.687783976234877e-05, + "loss": 0.0436, + "step": 17670 + }, + { + "grad_norm": 0.450725793838501, + "learning_rate": 7.684994885985472e-05, + "loss": 0.0456, + "step": 17680 + }, + { + "grad_norm": 0.20090939104557037, + "learning_rate": 7.682204621228083e-05, + "loss": 0.0405, + "step": 17690 + }, + { + "grad_norm": 0.17499704658985138, + "learning_rate": 7.679413183183267e-05, + "loss": 0.0513, + "step": 17700 + }, + { + "grad_norm": 0.22829179465770721, + "learning_rate": 7.676620573072093e-05, + "loss": 0.0477, + "step": 17710 + }, + { + "grad_norm": 0.2406584620475769, + "learning_rate": 7.673826792116145e-05, + "loss": 0.0448, + "step": 17720 + }, + { + "grad_norm": 0.14450183510780334, + "learning_rate": 7.671031841537519e-05, + "loss": 0.0399, + "step": 17730 + }, + { + "grad_norm": 0.3748242259025574, + "learning_rate": 7.668235722558819e-05, + "loss": 0.0523, + "step": 17740 + }, + { + "grad_norm": 0.2726356089115143, + "learning_rate": 7.665438436403161e-05, + "loss": 0.0548, + "step": 17750 + }, + { + "grad_norm": 0.24410803616046906, + "learning_rate": 7.662639984294178e-05, + "loss": 0.0496, + "step": 17760 + }, + { + "grad_norm": 0.22637571394443512, + "learning_rate": 7.659840367456002e-05, + "loss": 0.0456, + "step": 17770 + }, + { + "grad_norm": 0.25263461470603943, + "learning_rate": 7.657039587113287e-05, + "loss": 0.0513, + "step": 17780 + }, + { + "grad_norm": 0.22148913145065308, + "learning_rate": 7.654237644491182e-05, + "loss": 0.0378, + "step": 17790 + }, + { + "grad_norm": 0.3080918788909912, + "learning_rate": 7.651434540815358e-05, + "loss": 0.0407, + "step": 17800 + }, + { + "grad_norm": 0.23077724874019623, + "learning_rate": 7.648630277311986e-05, + "loss": 0.0527, + "step": 17810 + }, + { + "grad_norm": 0.21719354391098022, + "learning_rate": 7.645824855207744e-05, + "loss": 0.0475, + "step": 17820 + }, + { + "grad_norm": 0.2413601130247116, + "learning_rate": 7.643018275729821e-05, + "loss": 0.045, + "step": 17830 + }, + { + "grad_norm": 0.39304617047309875, + "learning_rate": 7.640210540105911e-05, + "loss": 0.0464, + "step": 17840 + }, + { + "grad_norm": 0.2689058482646942, + "learning_rate": 7.637401649564213e-05, + "loss": 0.0439, + "step": 17850 + }, + { + "grad_norm": 0.2948184609413147, + "learning_rate": 7.63459160533343e-05, + "loss": 0.0469, + "step": 17860 + }, + { + "grad_norm": 0.40297508239746094, + "learning_rate": 7.631780408642769e-05, + "loss": 0.0477, + "step": 17870 + }, + { + "grad_norm": 0.7311241626739502, + "learning_rate": 7.628968060721946e-05, + "loss": 0.0441, + "step": 17880 + }, + { + "grad_norm": 0.16293588280677795, + "learning_rate": 7.626154562801176e-05, + "loss": 0.042, + "step": 17890 + }, + { + "grad_norm": 0.2799907624721527, + "learning_rate": 7.623339916111177e-05, + "loss": 0.0426, + "step": 17900 + }, + { + "grad_norm": 0.2653927505016327, + "learning_rate": 7.620524121883174e-05, + "loss": 0.0425, + "step": 17910 + }, + { + "grad_norm": 0.2996769845485687, + "learning_rate": 7.617707181348892e-05, + "loss": 0.0411, + "step": 17920 + }, + { + "grad_norm": 0.2243075966835022, + "learning_rate": 7.614889095740554e-05, + "loss": 0.0406, + "step": 17930 + }, + { + "grad_norm": 0.19755291938781738, + "learning_rate": 7.612069866290885e-05, + "loss": 0.0414, + "step": 17940 + }, + { + "grad_norm": 0.164434552192688, + "learning_rate": 7.609249494233115e-05, + "loss": 0.0484, + "step": 17950 + }, + { + "grad_norm": 0.35055506229400635, + "learning_rate": 7.60642798080097e-05, + "loss": 0.0473, + "step": 17960 + }, + { + "grad_norm": 0.48704639077186584, + "learning_rate": 7.603605327228673e-05, + "loss": 0.0395, + "step": 17970 + }, + { + "grad_norm": 0.2701720595359802, + "learning_rate": 7.600781534750953e-05, + "loss": 0.0485, + "step": 17980 + }, + { + "grad_norm": 0.27761179208755493, + "learning_rate": 7.597956604603029e-05, + "loss": 0.0352, + "step": 17990 + }, + { + "grad_norm": 0.18493492901325226, + "learning_rate": 7.595130538020622e-05, + "loss": 0.0454, + "step": 18000 + }, + { + "grad_norm": 0.3538404703140259, + "learning_rate": 7.592303336239952e-05, + "loss": 0.0495, + "step": 18010 + }, + { + "grad_norm": 0.3108192980289459, + "learning_rate": 7.589475000497733e-05, + "loss": 0.0464, + "step": 18020 + }, + { + "grad_norm": 0.21092292666435242, + "learning_rate": 7.586645532031171e-05, + "loss": 0.0451, + "step": 18030 + }, + { + "grad_norm": 0.11811036616563797, + "learning_rate": 7.583814932077975e-05, + "loss": 0.0437, + "step": 18040 + }, + { + "grad_norm": 0.23175688087940216, + "learning_rate": 7.580983201876343e-05, + "loss": 0.0465, + "step": 18050 + }, + { + "grad_norm": 0.2576209604740143, + "learning_rate": 7.578150342664973e-05, + "loss": 0.0404, + "step": 18060 + }, + { + "grad_norm": 0.23884275555610657, + "learning_rate": 7.575316355683055e-05, + "loss": 0.0421, + "step": 18070 + }, + { + "grad_norm": 0.2180229276418686, + "learning_rate": 7.572481242170266e-05, + "loss": 0.0367, + "step": 18080 + }, + { + "grad_norm": 0.317882776260376, + "learning_rate": 7.569645003366786e-05, + "loss": 0.0436, + "step": 18090 + }, + { + "grad_norm": 0.36856502294540405, + "learning_rate": 7.566807640513278e-05, + "loss": 0.0463, + "step": 18100 + }, + { + "grad_norm": 0.20523519814014435, + "learning_rate": 7.563969154850905e-05, + "loss": 0.0399, + "step": 18110 + }, + { + "grad_norm": 0.4998184144496918, + "learning_rate": 7.561129547621314e-05, + "loss": 0.0445, + "step": 18120 + }, + { + "grad_norm": 0.3613101541996002, + "learning_rate": 7.558288820066648e-05, + "loss": 0.0424, + "step": 18130 + }, + { + "grad_norm": 0.441201776266098, + "learning_rate": 7.555446973429534e-05, + "loss": 0.0405, + "step": 18140 + }, + { + "grad_norm": 0.28131234645843506, + "learning_rate": 7.552604008953097e-05, + "loss": 0.046, + "step": 18150 + }, + { + "grad_norm": 0.33133038878440857, + "learning_rate": 7.549759927880942e-05, + "loss": 0.0391, + "step": 18160 + }, + { + "grad_norm": 0.3760335445404053, + "learning_rate": 7.54691473145717e-05, + "loss": 0.0447, + "step": 18170 + }, + { + "grad_norm": 0.3404892683029175, + "learning_rate": 7.544068420926365e-05, + "loss": 0.0413, + "step": 18180 + }, + { + "grad_norm": 0.3099025785923004, + "learning_rate": 7.5412209975336e-05, + "loss": 0.0458, + "step": 18190 + }, + { + "grad_norm": 0.2101154327392578, + "learning_rate": 7.538372462524435e-05, + "loss": 0.0393, + "step": 18200 + }, + { + "grad_norm": 0.26981544494628906, + "learning_rate": 7.535522817144915e-05, + "loss": 0.0479, + "step": 18210 + }, + { + "grad_norm": 0.2628600001335144, + "learning_rate": 7.532672062641575e-05, + "loss": 0.0431, + "step": 18220 + }, + { + "grad_norm": 0.25093576312065125, + "learning_rate": 7.529820200261427e-05, + "loss": 0.0447, + "step": 18230 + }, + { + "grad_norm": 0.3456785976886749, + "learning_rate": 7.526967231251977e-05, + "loss": 0.0492, + "step": 18240 + }, + { + "grad_norm": 0.3906562626361847, + "learning_rate": 7.524113156861206e-05, + "loss": 0.0519, + "step": 18250 + }, + { + "grad_norm": 0.2727521061897278, + "learning_rate": 7.521257978337587e-05, + "loss": 0.0475, + "step": 18260 + }, + { + "grad_norm": 0.21381136775016785, + "learning_rate": 7.518401696930072e-05, + "loss": 0.0529, + "step": 18270 + }, + { + "grad_norm": 0.2852931022644043, + "learning_rate": 7.515544313888092e-05, + "loss": 0.0403, + "step": 18280 + }, + { + "grad_norm": 0.18755605816841125, + "learning_rate": 7.512685830461568e-05, + "loss": 0.0457, + "step": 18290 + }, + { + "grad_norm": 0.29005900025367737, + "learning_rate": 7.509826247900892e-05, + "loss": 0.0462, + "step": 18300 + }, + { + "grad_norm": 0.290880024433136, + "learning_rate": 7.50696556745695e-05, + "loss": 0.0513, + "step": 18310 + }, + { + "grad_norm": 0.3242936432361603, + "learning_rate": 7.504103790381095e-05, + "loss": 0.0471, + "step": 18320 + }, + { + "grad_norm": 0.28991928696632385, + "learning_rate": 7.501240917925167e-05, + "loss": 0.0482, + "step": 18330 + }, + { + "grad_norm": 0.33945515751838684, + "learning_rate": 7.498376951341483e-05, + "loss": 0.0473, + "step": 18340 + }, + { + "grad_norm": 0.24618077278137207, + "learning_rate": 7.495511891882844e-05, + "loss": 0.0471, + "step": 18350 + }, + { + "grad_norm": 0.18991342186927795, + "learning_rate": 7.492645740802517e-05, + "loss": 0.0456, + "step": 18360 + }, + { + "grad_norm": 0.33570000529289246, + "learning_rate": 7.489778499354259e-05, + "loss": 0.0484, + "step": 18370 + }, + { + "grad_norm": 0.35356444120407104, + "learning_rate": 7.486910168792298e-05, + "loss": 0.0528, + "step": 18380 + }, + { + "grad_norm": 0.24751172959804535, + "learning_rate": 7.484040750371337e-05, + "loss": 0.0453, + "step": 18390 + }, + { + "grad_norm": 0.34632375836372375, + "learning_rate": 7.48117024534656e-05, + "loss": 0.0591, + "step": 18400 + }, + { + "grad_norm": 0.3029097020626068, + "learning_rate": 7.478298654973621e-05, + "loss": 0.0503, + "step": 18410 + }, + { + "grad_norm": 0.175028994679451, + "learning_rate": 7.475425980508654e-05, + "loss": 0.0563, + "step": 18420 + }, + { + "grad_norm": 0.2320028692483902, + "learning_rate": 7.47255222320826e-05, + "loss": 0.0466, + "step": 18430 + }, + { + "grad_norm": 0.35278230905532837, + "learning_rate": 7.469677384329522e-05, + "loss": 0.0478, + "step": 18440 + }, + { + "grad_norm": 0.31250959634780884, + "learning_rate": 7.46680146512999e-05, + "loss": 0.0545, + "step": 18450 + }, + { + "grad_norm": 0.24490231275558472, + "learning_rate": 7.463924466867692e-05, + "loss": 0.0498, + "step": 18460 + }, + { + "grad_norm": 0.31729283928871155, + "learning_rate": 7.461046390801121e-05, + "loss": 0.0518, + "step": 18470 + }, + { + "grad_norm": 0.25371986627578735, + "learning_rate": 7.458167238189248e-05, + "loss": 0.0464, + "step": 18480 + }, + { + "grad_norm": 0.23125293850898743, + "learning_rate": 7.455287010291509e-05, + "loss": 0.041, + "step": 18490 + }, + { + "grad_norm": 0.18254122138023376, + "learning_rate": 7.452405708367818e-05, + "loss": 0.0533, + "step": 18500 + }, + { + "grad_norm": 0.3204014003276825, + "learning_rate": 7.449523333678549e-05, + "loss": 0.0493, + "step": 18510 + }, + { + "grad_norm": 0.31590405106544495, + "learning_rate": 7.446639887484555e-05, + "loss": 0.0445, + "step": 18520 + }, + { + "grad_norm": 0.3014664351940155, + "learning_rate": 7.443755371047152e-05, + "loss": 0.0459, + "step": 18530 + }, + { + "grad_norm": 0.1890948861837387, + "learning_rate": 7.440869785628124e-05, + "loss": 0.043, + "step": 18540 + }, + { + "grad_norm": 0.3611214756965637, + "learning_rate": 7.437983132489727e-05, + "loss": 0.0483, + "step": 18550 + }, + { + "grad_norm": 0.2811214029788971, + "learning_rate": 7.435095412894678e-05, + "loss": 0.055, + "step": 18560 + }, + { + "grad_norm": 0.2529371976852417, + "learning_rate": 7.432206628106165e-05, + "loss": 0.0445, + "step": 18570 + }, + { + "grad_norm": 0.34955793619155884, + "learning_rate": 7.429316779387842e-05, + "loss": 0.0454, + "step": 18580 + }, + { + "grad_norm": 0.22038641571998596, + "learning_rate": 7.426425868003825e-05, + "loss": 0.0482, + "step": 18590 + }, + { + "grad_norm": 0.2581322193145752, + "learning_rate": 7.423533895218698e-05, + "loss": 0.0544, + "step": 18600 + }, + { + "grad_norm": 0.25325170159339905, + "learning_rate": 7.420640862297507e-05, + "loss": 0.0435, + "step": 18610 + }, + { + "grad_norm": 0.2403404414653778, + "learning_rate": 7.417746770505766e-05, + "loss": 0.0515, + "step": 18620 + }, + { + "grad_norm": 0.32526129484176636, + "learning_rate": 7.414851621109444e-05, + "loss": 0.0667, + "step": 18630 + }, + { + "grad_norm": 0.2808336019515991, + "learning_rate": 7.411955415374982e-05, + "loss": 0.0473, + "step": 18640 + }, + { + "grad_norm": 0.303888738155365, + "learning_rate": 7.409058154569278e-05, + "loss": 0.0477, + "step": 18650 + }, + { + "grad_norm": 0.31710758805274963, + "learning_rate": 7.406159839959689e-05, + "loss": 0.0518, + "step": 18660 + }, + { + "grad_norm": 0.2196469008922577, + "learning_rate": 7.403260472814039e-05, + "loss": 0.0499, + "step": 18670 + }, + { + "grad_norm": 0.21000899374485016, + "learning_rate": 7.400360054400611e-05, + "loss": 0.0473, + "step": 18680 + }, + { + "grad_norm": 0.3968193829059601, + "learning_rate": 7.397458585988143e-05, + "loss": 0.0455, + "step": 18690 + }, + { + "grad_norm": 0.35527318716049194, + "learning_rate": 7.394556068845841e-05, + "loss": 0.0385, + "step": 18700 + }, + { + "grad_norm": 0.15870428085327148, + "learning_rate": 7.39165250424336e-05, + "loss": 0.0458, + "step": 18710 + }, + { + "grad_norm": 0.3662666082382202, + "learning_rate": 7.38874789345082e-05, + "loss": 0.0401, + "step": 18720 + }, + { + "grad_norm": 0.29475975036621094, + "learning_rate": 7.385842237738794e-05, + "loss": 0.0612, + "step": 18730 + }, + { + "grad_norm": 0.15673398971557617, + "learning_rate": 7.382935538378318e-05, + "loss": 0.0421, + "step": 18740 + }, + { + "grad_norm": 0.1619872748851776, + "learning_rate": 7.38002779664088e-05, + "loss": 0.0418, + "step": 18750 + }, + { + "grad_norm": 0.32073190808296204, + "learning_rate": 7.377119013798422e-05, + "loss": 0.0432, + "step": 18760 + }, + { + "grad_norm": 0.509111762046814, + "learning_rate": 7.374209191123349e-05, + "loss": 0.0548, + "step": 18770 + }, + { + "grad_norm": 0.34130600094795227, + "learning_rate": 7.371298329888512e-05, + "loss": 0.0487, + "step": 18780 + }, + { + "grad_norm": 0.24560610949993134, + "learning_rate": 7.368386431367223e-05, + "loss": 0.0463, + "step": 18790 + }, + { + "grad_norm": 0.2692168354988098, + "learning_rate": 7.365473496833246e-05, + "loss": 0.0435, + "step": 18800 + }, + { + "grad_norm": 0.18307925760746002, + "learning_rate": 7.362559527560797e-05, + "loss": 0.0442, + "step": 18810 + }, + { + "grad_norm": 0.22160214185714722, + "learning_rate": 7.359644524824545e-05, + "loss": 0.0468, + "step": 18820 + }, + { + "grad_norm": 0.3439023792743683, + "learning_rate": 7.35672848989961e-05, + "loss": 0.0436, + "step": 18830 + }, + { + "grad_norm": 0.22338168323040009, + "learning_rate": 7.353811424061565e-05, + "loss": 0.0446, + "step": 18840 + }, + { + "grad_norm": 0.2597607672214508, + "learning_rate": 7.350893328586435e-05, + "loss": 0.0425, + "step": 18850 + }, + { + "grad_norm": 0.26414936780929565, + "learning_rate": 7.347974204750696e-05, + "loss": 0.0518, + "step": 18860 + }, + { + "grad_norm": 0.265909880399704, + "learning_rate": 7.345054053831266e-05, + "loss": 0.0478, + "step": 18870 + }, + { + "grad_norm": 0.20775741338729858, + "learning_rate": 7.342132877105523e-05, + "loss": 0.0451, + "step": 18880 + }, + { + "grad_norm": 0.35120683908462524, + "learning_rate": 7.339210675851288e-05, + "loss": 0.0417, + "step": 18890 + }, + { + "grad_norm": 0.26934367418289185, + "learning_rate": 7.336287451346831e-05, + "loss": 0.0533, + "step": 18900 + }, + { + "grad_norm": 0.18519669771194458, + "learning_rate": 7.333363204870868e-05, + "loss": 0.0376, + "step": 18910 + }, + { + "grad_norm": 0.23976537585258484, + "learning_rate": 7.330437937702566e-05, + "loss": 0.0383, + "step": 18920 + }, + { + "grad_norm": 0.2723686695098877, + "learning_rate": 7.327511651121534e-05, + "loss": 0.0467, + "step": 18930 + }, + { + "grad_norm": 0.3505902588367462, + "learning_rate": 7.324584346407833e-05, + "loss": 0.0389, + "step": 18940 + }, + { + "grad_norm": 0.21419118344783783, + "learning_rate": 7.321656024841962e-05, + "loss": 0.0383, + "step": 18950 + }, + { + "grad_norm": 0.24488846957683563, + "learning_rate": 7.31872668770487e-05, + "loss": 0.0397, + "step": 18960 + }, + { + "grad_norm": 0.22103920578956604, + "learning_rate": 7.315796336277949e-05, + "loss": 0.0411, + "step": 18970 + }, + { + "grad_norm": 0.18622983992099762, + "learning_rate": 7.312864971843031e-05, + "loss": 0.0431, + "step": 18980 + }, + { + "grad_norm": 0.1754983365535736, + "learning_rate": 7.309932595682398e-05, + "loss": 0.0405, + "step": 18990 + }, + { + "grad_norm": 0.2521551847457886, + "learning_rate": 7.30699920907877e-05, + "loss": 0.0471, + "step": 19000 + }, + { + "grad_norm": 0.2516734004020691, + "learning_rate": 7.304064813315308e-05, + "loss": 0.0445, + "step": 19010 + }, + { + "grad_norm": 0.25409579277038574, + "learning_rate": 7.30112940967562e-05, + "loss": 0.0418, + "step": 19020 + }, + { + "grad_norm": 0.2805182933807373, + "learning_rate": 7.298192999443747e-05, + "loss": 0.0396, + "step": 19030 + }, + { + "grad_norm": 0.16406716406345367, + "learning_rate": 7.295255583904179e-05, + "loss": 0.0496, + "step": 19040 + }, + { + "grad_norm": 0.2151872217655182, + "learning_rate": 7.29231716434184e-05, + "loss": 0.0419, + "step": 19050 + }, + { + "grad_norm": 0.17130610346794128, + "learning_rate": 7.289377742042092e-05, + "loss": 0.0441, + "step": 19060 + }, + { + "grad_norm": 0.16771681606769562, + "learning_rate": 7.286437318290741e-05, + "loss": 0.0453, + "step": 19070 + }, + { + "grad_norm": 1.2136824131011963, + "learning_rate": 7.283495894374027e-05, + "loss": 0.0412, + "step": 19080 + }, + { + "grad_norm": 0.20082734525203705, + "learning_rate": 7.280553471578629e-05, + "loss": 0.0468, + "step": 19090 + }, + { + "grad_norm": 0.24355126917362213, + "learning_rate": 7.277610051191663e-05, + "loss": 0.0449, + "step": 19100 + }, + { + "grad_norm": 0.21989169716835022, + "learning_rate": 7.27466563450068e-05, + "loss": 0.0398, + "step": 19110 + }, + { + "grad_norm": 0.1578790545463562, + "learning_rate": 7.271720222793668e-05, + "loss": 0.0408, + "step": 19120 + }, + { + "grad_norm": 0.3163919150829315, + "learning_rate": 7.268773817359054e-05, + "loss": 0.0495, + "step": 19130 + }, + { + "grad_norm": 0.211142435669899, + "learning_rate": 7.265826419485689e-05, + "loss": 0.0368, + "step": 19140 + }, + { + "grad_norm": 0.26274731755256653, + "learning_rate": 7.26287803046287e-05, + "loss": 0.0337, + "step": 19150 + }, + { + "grad_norm": 0.18850970268249512, + "learning_rate": 7.259928651580322e-05, + "loss": 0.0425, + "step": 19160 + }, + { + "grad_norm": 0.2350223958492279, + "learning_rate": 7.256978284128202e-05, + "loss": 0.0377, + "step": 19170 + }, + { + "grad_norm": 0.22486239671707153, + "learning_rate": 7.254026929397101e-05, + "loss": 0.047, + "step": 19180 + }, + { + "grad_norm": 0.2848038673400879, + "learning_rate": 7.251074588678044e-05, + "loss": 0.0626, + "step": 19190 + }, + { + "grad_norm": 0.7751398682594299, + "learning_rate": 7.248121263262482e-05, + "loss": 0.0461, + "step": 19200 + }, + { + "grad_norm": 0.23727691173553467, + "learning_rate": 7.245166954442304e-05, + "loss": 0.0459, + "step": 19210 + }, + { + "grad_norm": 0.2860068678855896, + "learning_rate": 7.24221166350982e-05, + "loss": 0.0423, + "step": 19220 + }, + { + "grad_norm": 0.3130536675453186, + "learning_rate": 7.239255391757777e-05, + "loss": 0.0484, + "step": 19230 + }, + { + "grad_norm": 0.31119194626808167, + "learning_rate": 7.236298140479351e-05, + "loss": 0.0504, + "step": 19240 + }, + { + "grad_norm": 0.37345564365386963, + "learning_rate": 7.233339910968141e-05, + "loss": 0.0476, + "step": 19250 + }, + { + "grad_norm": 0.24428381025791168, + "learning_rate": 7.230380704518178e-05, + "loss": 0.0424, + "step": 19260 + }, + { + "grad_norm": 0.40783414244651794, + "learning_rate": 7.227420522423919e-05, + "loss": 0.0364, + "step": 19270 + }, + { + "grad_norm": 0.1878710240125656, + "learning_rate": 7.224459365980249e-05, + "loss": 0.0451, + "step": 19280 + }, + { + "grad_norm": 0.3155026435852051, + "learning_rate": 7.221497236482479e-05, + "loss": 0.0538, + "step": 19290 + }, + { + "grad_norm": 0.2403191328048706, + "learning_rate": 7.218534135226345e-05, + "loss": 0.0419, + "step": 19300 + }, + { + "grad_norm": 0.3125084638595581, + "learning_rate": 7.215570063508005e-05, + "loss": 0.0465, + "step": 19310 + }, + { + "grad_norm": 0.3434067964553833, + "learning_rate": 7.21260502262405e-05, + "loss": 0.0575, + "step": 19320 + }, + { + "grad_norm": 0.6436324715614319, + "learning_rate": 7.209639013871487e-05, + "loss": 0.0595, + "step": 19330 + }, + { + "grad_norm": 0.23739971220493317, + "learning_rate": 7.206672038547746e-05, + "loss": 0.0611, + "step": 19340 + }, + { + "grad_norm": 0.28187182545661926, + "learning_rate": 7.203704097950687e-05, + "loss": 0.0528, + "step": 19350 + }, + { + "grad_norm": 0.21036197245121002, + "learning_rate": 7.200735193378587e-05, + "loss": 0.0403, + "step": 19360 + }, + { + "grad_norm": 0.22275292873382568, + "learning_rate": 7.197765326130145e-05, + "loss": 0.0457, + "step": 19370 + }, + { + "grad_norm": 0.2646772265434265, + "learning_rate": 7.194794497504481e-05, + "loss": 0.0565, + "step": 19380 + }, + { + "grad_norm": 0.2773742377758026, + "learning_rate": 7.19182270880114e-05, + "loss": 0.0473, + "step": 19390 + }, + { + "grad_norm": 0.9727538228034973, + "learning_rate": 7.188849961320079e-05, + "loss": 0.0443, + "step": 19400 + }, + { + "grad_norm": 0.24029754102230072, + "learning_rate": 7.185876256361681e-05, + "loss": 0.0418, + "step": 19410 + }, + { + "grad_norm": 0.21285182237625122, + "learning_rate": 7.182901595226745e-05, + "loss": 0.0408, + "step": 19420 + }, + { + "grad_norm": 0.38008591532707214, + "learning_rate": 7.179925979216491e-05, + "loss": 0.0572, + "step": 19430 + }, + { + "grad_norm": 0.4972408413887024, + "learning_rate": 7.17694940963255e-05, + "loss": 0.0447, + "step": 19440 + }, + { + "grad_norm": 0.20722796022891998, + "learning_rate": 7.17397188777698e-05, + "loss": 0.0417, + "step": 19450 + }, + { + "grad_norm": 0.2823748290538788, + "learning_rate": 7.170993414952245e-05, + "loss": 0.0444, + "step": 19460 + }, + { + "grad_norm": 0.23613905906677246, + "learning_rate": 7.168013992461233e-05, + "loss": 0.0515, + "step": 19470 + }, + { + "grad_norm": 0.19227184355258942, + "learning_rate": 7.165033621607245e-05, + "loss": 0.0431, + "step": 19480 + }, + { + "grad_norm": 0.9318661689758301, + "learning_rate": 7.162052303693995e-05, + "loss": 0.0557, + "step": 19490 + }, + { + "grad_norm": 0.29035326838493347, + "learning_rate": 7.159070040025614e-05, + "loss": 0.048, + "step": 19500 + }, + { + "grad_norm": 0.35574012994766235, + "learning_rate": 7.156086831906646e-05, + "loss": 0.054, + "step": 19510 + }, + { + "grad_norm": 0.24983002245426178, + "learning_rate": 7.153102680642044e-05, + "loss": 0.0403, + "step": 19520 + }, + { + "grad_norm": 0.21753616631031036, + "learning_rate": 7.150117587537182e-05, + "loss": 0.044, + "step": 19530 + }, + { + "grad_norm": 0.26731398701667786, + "learning_rate": 7.147131553897838e-05, + "loss": 0.0455, + "step": 19540 + }, + { + "grad_norm": 0.27705609798431396, + "learning_rate": 7.144144581030205e-05, + "loss": 0.0471, + "step": 19550 + }, + { + "grad_norm": 0.2204599529504776, + "learning_rate": 7.141156670240889e-05, + "loss": 0.057, + "step": 19560 + }, + { + "grad_norm": 0.25218528509140015, + "learning_rate": 7.138167822836899e-05, + "loss": 0.0461, + "step": 19570 + }, + { + "grad_norm": 0.2473684698343277, + "learning_rate": 7.135178040125661e-05, + "loss": 0.052, + "step": 19580 + }, + { + "grad_norm": 0.21554124355316162, + "learning_rate": 7.13218732341501e-05, + "loss": 0.0431, + "step": 19590 + }, + { + "grad_norm": 0.24597714841365814, + "learning_rate": 7.129195674013181e-05, + "loss": 0.0518, + "step": 19600 + }, + { + "grad_norm": 0.7426440715789795, + "learning_rate": 7.126203093228829e-05, + "loss": 0.0443, + "step": 19610 + }, + { + "grad_norm": 0.24221934378147125, + "learning_rate": 7.123209582371006e-05, + "loss": 0.0439, + "step": 19620 + }, + { + "grad_norm": 0.3561791181564331, + "learning_rate": 7.120215142749179e-05, + "loss": 0.0465, + "step": 19630 + }, + { + "grad_norm": 0.17972509562969208, + "learning_rate": 7.117219775673218e-05, + "loss": 0.0442, + "step": 19640 + }, + { + "grad_norm": 0.2032284438610077, + "learning_rate": 7.114223482453394e-05, + "loss": 0.0477, + "step": 19650 + }, + { + "grad_norm": 0.25220125913619995, + "learning_rate": 7.11122626440039e-05, + "loss": 0.0406, + "step": 19660 + }, + { + "grad_norm": 0.4258060157299042, + "learning_rate": 7.10822812282529e-05, + "loss": 0.0454, + "step": 19670 + }, + { + "grad_norm": 0.2057989239692688, + "learning_rate": 7.105229059039586e-05, + "loss": 0.0462, + "step": 19680 + }, + { + "grad_norm": 0.2687794268131256, + "learning_rate": 7.102229074355168e-05, + "loss": 0.0507, + "step": 19690 + }, + { + "grad_norm": 0.30194294452667236, + "learning_rate": 7.099228170084332e-05, + "loss": 0.0455, + "step": 19700 + }, + { + "grad_norm": 0.1386367380619049, + "learning_rate": 7.096226347539771e-05, + "loss": 0.0453, + "step": 19710 + }, + { + "grad_norm": 0.3091317117214203, + "learning_rate": 7.093223608034592e-05, + "loss": 0.0466, + "step": 19720 + }, + { + "grad_norm": 0.2166784554719925, + "learning_rate": 7.090219952882291e-05, + "loss": 0.0438, + "step": 19730 + }, + { + "grad_norm": 0.25268250703811646, + "learning_rate": 7.087215383396768e-05, + "loss": 0.0431, + "step": 19740 + }, + { + "grad_norm": 0.16843193769454956, + "learning_rate": 7.084209900892325e-05, + "loss": 0.0414, + "step": 19750 + }, + { + "grad_norm": 1.5761910676956177, + "learning_rate": 7.081203506683663e-05, + "loss": 0.0494, + "step": 19760 + }, + { + "grad_norm": 0.18809546530246735, + "learning_rate": 7.078196202085879e-05, + "loss": 0.0409, + "step": 19770 + }, + { + "grad_norm": 0.20938237011432648, + "learning_rate": 7.075187988414473e-05, + "loss": 0.0523, + "step": 19780 + }, + { + "grad_norm": 0.28376442193984985, + "learning_rate": 7.072178866985334e-05, + "loss": 0.0519, + "step": 19790 + }, + { + "grad_norm": 0.22412905097007751, + "learning_rate": 7.069168839114764e-05, + "loss": 0.0433, + "step": 19800 + }, + { + "grad_norm": 0.3303901255130768, + "learning_rate": 7.066157906119441e-05, + "loss": 0.0475, + "step": 19810 + }, + { + "grad_norm": 0.2787971496582031, + "learning_rate": 7.063146069316456e-05, + "loss": 0.0416, + "step": 19820 + }, + { + "grad_norm": 0.23016560077667236, + "learning_rate": 7.060133330023287e-05, + "loss": 0.0354, + "step": 19830 + }, + { + "grad_norm": 0.2872569262981415, + "learning_rate": 7.057119689557807e-05, + "loss": 0.0441, + "step": 19840 + }, + { + "grad_norm": 0.3543080985546112, + "learning_rate": 7.054105149238287e-05, + "loss": 0.054, + "step": 19850 + }, + { + "grad_norm": 0.2781756818294525, + "learning_rate": 7.051089710383387e-05, + "loss": 0.0521, + "step": 19860 + }, + { + "grad_norm": 0.3082110285758972, + "learning_rate": 7.048073374312166e-05, + "loss": 0.0439, + "step": 19870 + }, + { + "grad_norm": 0.3483620285987854, + "learning_rate": 7.045056142344067e-05, + "loss": 0.0488, + "step": 19880 + }, + { + "grad_norm": 0.17022645473480225, + "learning_rate": 7.042038015798934e-05, + "loss": 0.0461, + "step": 19890 + }, + { + "grad_norm": 0.3602137565612793, + "learning_rate": 7.039018995996995e-05, + "loss": 0.0624, + "step": 19900 + }, + { + "grad_norm": 0.3290145993232727, + "learning_rate": 7.035999084258876e-05, + "loss": 0.0438, + "step": 19910 + }, + { + "grad_norm": 0.3379499018192291, + "learning_rate": 7.032978281905583e-05, + "loss": 0.0444, + "step": 19920 + }, + { + "grad_norm": 0.22548800706863403, + "learning_rate": 7.029956590258522e-05, + "loss": 0.0469, + "step": 19930 + }, + { + "grad_norm": 0.253935843706131, + "learning_rate": 7.026934010639483e-05, + "loss": 0.0416, + "step": 19940 + }, + { + "grad_norm": 0.32323718070983887, + "learning_rate": 7.023910544370645e-05, + "loss": 0.0428, + "step": 19950 + }, + { + "grad_norm": 0.25240588188171387, + "learning_rate": 7.020886192774573e-05, + "loss": 0.0411, + "step": 19960 + }, + { + "grad_norm": 0.16038109362125397, + "learning_rate": 7.017860957174222e-05, + "loss": 0.0445, + "step": 19970 + }, + { + "grad_norm": 0.23036697506904602, + "learning_rate": 7.014834838892934e-05, + "loss": 0.0425, + "step": 19980 + }, + { + "grad_norm": 0.20752739906311035, + "learning_rate": 7.011807839254435e-05, + "loss": 0.0418, + "step": 19990 + }, + { + "grad_norm": 0.15357288718223572, + "learning_rate": 7.008779959582837e-05, + "loss": 0.0454, + "step": 20000 + }, + { + "grad_norm": 0.22830070555210114, + "learning_rate": 7.005751201202639e-05, + "loss": 0.045, + "step": 20010 + }, + { + "grad_norm": 0.2798432409763336, + "learning_rate": 7.002721565438724e-05, + "loss": 0.0425, + "step": 20020 + }, + { + "grad_norm": 0.1822013258934021, + "learning_rate": 6.999691053616353e-05, + "loss": 0.0437, + "step": 20030 + }, + { + "grad_norm": 0.21344102919101715, + "learning_rate": 6.996659667061178e-05, + "loss": 0.0575, + "step": 20040 + }, + { + "grad_norm": 0.30875828862190247, + "learning_rate": 6.993627407099229e-05, + "loss": 0.0509, + "step": 20050 + }, + { + "grad_norm": 0.2052740752696991, + "learning_rate": 6.990594275056921e-05, + "loss": 0.0394, + "step": 20060 + }, + { + "grad_norm": 0.281326562166214, + "learning_rate": 6.987560272261049e-05, + "loss": 0.0505, + "step": 20070 + }, + { + "grad_norm": 0.3128502666950226, + "learning_rate": 6.984525400038788e-05, + "loss": 0.0433, + "step": 20080 + }, + { + "grad_norm": 0.27932634949684143, + "learning_rate": 6.981489659717693e-05, + "loss": 0.0453, + "step": 20090 + }, + { + "grad_norm": 0.3599071204662323, + "learning_rate": 6.978453052625703e-05, + "loss": 0.0422, + "step": 20100 + }, + { + "grad_norm": 0.2411578893661499, + "learning_rate": 6.975415580091131e-05, + "loss": 0.0397, + "step": 20110 + }, + { + "grad_norm": 0.251193642616272, + "learning_rate": 6.972377243442672e-05, + "loss": 0.0444, + "step": 20120 + }, + { + "grad_norm": 0.17321524024009705, + "learning_rate": 6.969338044009396e-05, + "loss": 0.0429, + "step": 20130 + }, + { + "grad_norm": 0.30823197960853577, + "learning_rate": 6.966297983120753e-05, + "loss": 0.039, + "step": 20140 + }, + { + "grad_norm": 1.0327293872833252, + "learning_rate": 6.963257062106571e-05, + "loss": 0.0424, + "step": 20150 + }, + { + "grad_norm": 0.19960245490074158, + "learning_rate": 6.960215282297048e-05, + "loss": 0.0428, + "step": 20160 + }, + { + "grad_norm": 0.23490236699581146, + "learning_rate": 6.957172645022765e-05, + "loss": 0.048, + "step": 20170 + }, + { + "grad_norm": 0.23590052127838135, + "learning_rate": 6.954129151614673e-05, + "loss": 0.044, + "step": 20180 + }, + { + "grad_norm": 0.1640266627073288, + "learning_rate": 6.9510848034041e-05, + "loss": 0.0402, + "step": 20190 + }, + { + "grad_norm": 0.1811748892068863, + "learning_rate": 6.94803960172275e-05, + "loss": 0.0393, + "step": 20200 + }, + { + "grad_norm": 0.14477819204330444, + "learning_rate": 6.944993547902692e-05, + "loss": 0.0418, + "step": 20210 + }, + { + "grad_norm": 0.17939502000808716, + "learning_rate": 6.941946643276379e-05, + "loss": 0.0389, + "step": 20220 + }, + { + "grad_norm": 0.21192972362041473, + "learning_rate": 6.938898889176626e-05, + "loss": 0.051, + "step": 20230 + }, + { + "grad_norm": 0.3237829804420471, + "learning_rate": 6.935850286936627e-05, + "loss": 0.0443, + "step": 20240 + }, + { + "grad_norm": 0.26890817284584045, + "learning_rate": 6.932800837889943e-05, + "loss": 0.0444, + "step": 20250 + }, + { + "grad_norm": 0.17918992042541504, + "learning_rate": 6.92975054337051e-05, + "loss": 0.0426, + "step": 20260 + }, + { + "grad_norm": 0.23302538692951202, + "learning_rate": 6.926699404712627e-05, + "loss": 0.0463, + "step": 20270 + }, + { + "grad_norm": 0.25081199407577515, + "learning_rate": 6.923647423250967e-05, + "loss": 0.04, + "step": 20280 + }, + { + "grad_norm": 0.4809999465942383, + "learning_rate": 6.920594600320567e-05, + "loss": 0.0508, + "step": 20290 + }, + { + "grad_norm": 0.13893239200115204, + "learning_rate": 6.91754093725684e-05, + "loss": 0.0443, + "step": 20300 + }, + { + "grad_norm": 0.6014142036437988, + "learning_rate": 6.914486435395561e-05, + "loss": 0.0447, + "step": 20310 + }, + { + "grad_norm": 0.288357675075531, + "learning_rate": 6.911431096072871e-05, + "loss": 0.0451, + "step": 20320 + }, + { + "grad_norm": 0.29664796590805054, + "learning_rate": 6.90837492062528e-05, + "loss": 0.0404, + "step": 20330 + }, + { + "grad_norm": 0.25249770283699036, + "learning_rate": 6.905317910389664e-05, + "loss": 0.0464, + "step": 20340 + }, + { + "grad_norm": 0.20473140478134155, + "learning_rate": 6.902260066703262e-05, + "loss": 0.0491, + "step": 20350 + }, + { + "grad_norm": 0.18926258385181427, + "learning_rate": 6.89920139090368e-05, + "loss": 0.0448, + "step": 20360 + }, + { + "grad_norm": 0.2342446744441986, + "learning_rate": 6.896141884328886e-05, + "loss": 0.0465, + "step": 20370 + }, + { + "grad_norm": 0.3188340663909912, + "learning_rate": 6.893081548317211e-05, + "loss": 0.0585, + "step": 20380 + }, + { + "grad_norm": 0.3619644343852997, + "learning_rate": 6.890020384207353e-05, + "loss": 0.0436, + "step": 20390 + }, + { + "grad_norm": 0.27550095319747925, + "learning_rate": 6.886958393338366e-05, + "loss": 0.0398, + "step": 20400 + }, + { + "grad_norm": 0.21320457756519318, + "learning_rate": 6.883895577049668e-05, + "loss": 0.0466, + "step": 20410 + }, + { + "grad_norm": 0.23654574155807495, + "learning_rate": 6.880831936681045e-05, + "loss": 0.0398, + "step": 20420 + }, + { + "grad_norm": 0.16719819605350494, + "learning_rate": 6.87776747357263e-05, + "loss": 0.0428, + "step": 20430 + }, + { + "grad_norm": 0.2041938304901123, + "learning_rate": 6.874702189064927e-05, + "loss": 0.0468, + "step": 20440 + }, + { + "grad_norm": 0.22052635252475739, + "learning_rate": 6.871636084498796e-05, + "loss": 0.0412, + "step": 20450 + }, + { + "grad_norm": 0.20660695433616638, + "learning_rate": 6.868569161215453e-05, + "loss": 0.0543, + "step": 20460 + }, + { + "grad_norm": 0.3968670964241028, + "learning_rate": 6.865501420556477e-05, + "loss": 0.0481, + "step": 20470 + }, + { + "grad_norm": 0.17547956109046936, + "learning_rate": 6.862432863863802e-05, + "loss": 0.0402, + "step": 20480 + }, + { + "grad_norm": 0.20360998809337616, + "learning_rate": 6.859363492479716e-05, + "loss": 0.0381, + "step": 20490 + }, + { + "grad_norm": 0.18338605761528015, + "learning_rate": 6.856293307746868e-05, + "loss": 0.0415, + "step": 20500 + }, + { + "grad_norm": 0.3309321701526642, + "learning_rate": 6.853222311008263e-05, + "loss": 0.0455, + "step": 20510 + }, + { + "grad_norm": 0.21988047659397125, + "learning_rate": 6.850150503607256e-05, + "loss": 0.0385, + "step": 20520 + }, + { + "grad_norm": 0.232944056391716, + "learning_rate": 6.847077886887562e-05, + "loss": 0.0427, + "step": 20530 + }, + { + "grad_norm": 0.16992206871509552, + "learning_rate": 6.844004462193248e-05, + "loss": 0.0426, + "step": 20540 + }, + { + "grad_norm": 0.23553824424743652, + "learning_rate": 6.840930230868734e-05, + "loss": 0.0475, + "step": 20550 + }, + { + "grad_norm": 0.2665756642818451, + "learning_rate": 6.837855194258791e-05, + "loss": 0.0442, + "step": 20560 + }, + { + "grad_norm": 0.27930334210395813, + "learning_rate": 6.834779353708548e-05, + "loss": 0.0423, + "step": 20570 + }, + { + "grad_norm": 0.2339128702878952, + "learning_rate": 6.831702710563478e-05, + "loss": 0.0401, + "step": 20580 + }, + { + "grad_norm": 0.33976343274116516, + "learning_rate": 6.828625266169416e-05, + "loss": 0.0417, + "step": 20590 + }, + { + "grad_norm": 1.0606943368911743, + "learning_rate": 6.825547021872534e-05, + "loss": 0.0453, + "step": 20600 + }, + { + "grad_norm": 0.367786705493927, + "learning_rate": 6.822467979019365e-05, + "loss": 0.0482, + "step": 20610 + }, + { + "grad_norm": 0.2882705330848694, + "learning_rate": 6.819388138956783e-05, + "loss": 0.0411, + "step": 20620 + }, + { + "grad_norm": 0.21982760727405548, + "learning_rate": 6.816307503032019e-05, + "loss": 0.0408, + "step": 20630 + }, + { + "grad_norm": 0.2309218943119049, + "learning_rate": 6.813226072592645e-05, + "loss": 0.0438, + "step": 20640 + }, + { + "grad_norm": 0.19745425879955292, + "learning_rate": 6.810143848986583e-05, + "loss": 0.0381, + "step": 20650 + }, + { + "grad_norm": 0.27891165018081665, + "learning_rate": 6.807060833562106e-05, + "loss": 0.0402, + "step": 20660 + }, + { + "grad_norm": 0.1885381042957306, + "learning_rate": 6.803977027667825e-05, + "loss": 0.0384, + "step": 20670 + }, + { + "grad_norm": 0.14481784403324127, + "learning_rate": 6.800892432652704e-05, + "loss": 0.0422, + "step": 20680 + }, + { + "grad_norm": 0.30975160002708435, + "learning_rate": 6.797807049866049e-05, + "loss": 0.0428, + "step": 20690 + }, + { + "grad_norm": 0.19779956340789795, + "learning_rate": 6.794720880657512e-05, + "loss": 0.0444, + "step": 20700 + }, + { + "grad_norm": 0.2485109120607376, + "learning_rate": 6.791633926377085e-05, + "loss": 0.0416, + "step": 20710 + }, + { + "grad_norm": 0.2975394129753113, + "learning_rate": 6.788546188375114e-05, + "loss": 0.0486, + "step": 20720 + }, + { + "grad_norm": 0.2515464723110199, + "learning_rate": 6.785457668002274e-05, + "loss": 0.0538, + "step": 20730 + }, + { + "grad_norm": 1.8464007377624512, + "learning_rate": 6.78236836660959e-05, + "loss": 0.0382, + "step": 20740 + }, + { + "grad_norm": 0.23969878256320953, + "learning_rate": 6.779278285548427e-05, + "loss": 0.055, + "step": 20750 + }, + { + "grad_norm": 0.2477761209011078, + "learning_rate": 6.776187426170493e-05, + "loss": 0.0472, + "step": 20760 + }, + { + "grad_norm": 0.2502649426460266, + "learning_rate": 6.773095789827837e-05, + "loss": 0.0511, + "step": 20770 + }, + { + "grad_norm": 0.21304060518741608, + "learning_rate": 6.770003377872841e-05, + "loss": 0.0502, + "step": 20780 + }, + { + "grad_norm": 0.6063156723976135, + "learning_rate": 6.766910191658233e-05, + "loss": 0.0407, + "step": 20790 + }, + { + "grad_norm": 0.2997325360774994, + "learning_rate": 6.763816232537079e-05, + "loss": 0.052, + "step": 20800 + }, + { + "grad_norm": 0.21107950806617737, + "learning_rate": 6.76072150186278e-05, + "loss": 0.0402, + "step": 20810 + }, + { + "grad_norm": 0.27426570653915405, + "learning_rate": 6.757626000989077e-05, + "loss": 0.0538, + "step": 20820 + }, + { + "grad_norm": 0.26422345638275146, + "learning_rate": 6.754529731270048e-05, + "loss": 0.0453, + "step": 20830 + }, + { + "grad_norm": 0.26599356532096863, + "learning_rate": 6.751432694060105e-05, + "loss": 0.052, + "step": 20840 + }, + { + "grad_norm": 0.17751742899417877, + "learning_rate": 6.748334890713999e-05, + "loss": 0.046, + "step": 20850 + }, + { + "grad_norm": 0.27679508924484253, + "learning_rate": 6.745236322586813e-05, + "loss": 0.0516, + "step": 20860 + }, + { + "grad_norm": 0.23525691032409668, + "learning_rate": 6.742136991033967e-05, + "loss": 0.0456, + "step": 20870 + }, + { + "grad_norm": 0.33716991543769836, + "learning_rate": 6.739036897411215e-05, + "loss": 0.0464, + "step": 20880 + }, + { + "grad_norm": 0.3383137881755829, + "learning_rate": 6.73593604307464e-05, + "loss": 0.0493, + "step": 20890 + }, + { + "grad_norm": 0.328839510679245, + "learning_rate": 6.732834429380663e-05, + "loss": 0.0471, + "step": 20900 + }, + { + "grad_norm": 0.2623497545719147, + "learning_rate": 6.729732057686036e-05, + "loss": 0.0448, + "step": 20910 + }, + { + "grad_norm": 0.1810142993927002, + "learning_rate": 6.726628929347839e-05, + "loss": 0.0424, + "step": 20920 + }, + { + "grad_norm": 0.2022293210029602, + "learning_rate": 6.723525045723488e-05, + "loss": 0.0431, + "step": 20930 + }, + { + "grad_norm": 0.1750342696905136, + "learning_rate": 6.720420408170728e-05, + "loss": 0.0481, + "step": 20940 + }, + { + "grad_norm": 0.30105656385421753, + "learning_rate": 6.71731501804763e-05, + "loss": 0.0485, + "step": 20950 + }, + { + "grad_norm": 0.26268884539604187, + "learning_rate": 6.714208876712602e-05, + "loss": 0.0426, + "step": 20960 + }, + { + "grad_norm": 0.24731552600860596, + "learning_rate": 6.71110198552437e-05, + "loss": 0.0429, + "step": 20970 + }, + { + "grad_norm": 0.4312298893928528, + "learning_rate": 6.707994345841996e-05, + "loss": 0.0415, + "step": 20980 + }, + { + "grad_norm": 0.23767593502998352, + "learning_rate": 6.70488595902487e-05, + "loss": 0.042, + "step": 20990 + }, + { + "grad_norm": 0.23684239387512207, + "learning_rate": 6.7017768264327e-05, + "loss": 0.0425, + "step": 21000 + }, + { + "grad_norm": 0.26782703399658203, + "learning_rate": 6.698666949425532e-05, + "loss": 0.0494, + "step": 21010 + }, + { + "grad_norm": 0.30706843733787537, + "learning_rate": 6.69555632936373e-05, + "loss": 0.0436, + "step": 21020 + }, + { + "grad_norm": 0.22835667431354523, + "learning_rate": 6.692444967607983e-05, + "loss": 0.0492, + "step": 21030 + }, + { + "grad_norm": 0.2796635031700134, + "learning_rate": 6.689332865519309e-05, + "loss": 0.0486, + "step": 21040 + }, + { + "grad_norm": 0.20086221396923065, + "learning_rate": 6.686220024459045e-05, + "loss": 0.05, + "step": 21050 + }, + { + "grad_norm": 0.21580913662910461, + "learning_rate": 6.683106445788856e-05, + "loss": 0.0514, + "step": 21060 + }, + { + "grad_norm": 0.248008593916893, + "learning_rate": 6.679992130870723e-05, + "loss": 0.0405, + "step": 21070 + }, + { + "grad_norm": 0.17854763567447662, + "learning_rate": 6.676877081066957e-05, + "loss": 0.0408, + "step": 21080 + }, + { + "grad_norm": 0.2149074673652649, + "learning_rate": 6.673761297740186e-05, + "loss": 0.0481, + "step": 21090 + }, + { + "grad_norm": 0.2767467498779297, + "learning_rate": 6.670644782253358e-05, + "loss": 0.0484, + "step": 21100 + }, + { + "grad_norm": 0.2187749594449997, + "learning_rate": 6.667527535969744e-05, + "loss": 0.0424, + "step": 21110 + }, + { + "grad_norm": 0.13549676537513733, + "learning_rate": 6.664409560252933e-05, + "loss": 0.0435, + "step": 21120 + }, + { + "grad_norm": 0.27149125933647156, + "learning_rate": 6.661290856466833e-05, + "loss": 0.0458, + "step": 21130 + }, + { + "grad_norm": 0.2530698776245117, + "learning_rate": 6.658171425975672e-05, + "loss": 0.0441, + "step": 21140 + }, + { + "grad_norm": 0.42095255851745605, + "learning_rate": 6.655051270143994e-05, + "loss": 0.0456, + "step": 21150 + }, + { + "grad_norm": 0.16760483384132385, + "learning_rate": 6.651930390336662e-05, + "loss": 0.0387, + "step": 21160 + }, + { + "grad_norm": 0.19523708522319794, + "learning_rate": 6.648808787918854e-05, + "loss": 0.0433, + "step": 21170 + }, + { + "grad_norm": 0.295576810836792, + "learning_rate": 6.645686464256065e-05, + "loss": 0.0422, + "step": 21180 + }, + { + "grad_norm": 0.33571192622184753, + "learning_rate": 6.642563420714106e-05, + "loss": 0.0456, + "step": 21190 + }, + { + "grad_norm": 0.2586629092693329, + "learning_rate": 6.639439658659102e-05, + "loss": 0.0438, + "step": 21200 + }, + { + "grad_norm": 0.2044009268283844, + "learning_rate": 6.636315179457492e-05, + "loss": 0.0437, + "step": 21210 + }, + { + "grad_norm": 0.2863611876964569, + "learning_rate": 6.633189984476031e-05, + "loss": 0.0468, + "step": 21220 + }, + { + "grad_norm": 0.3650352358818054, + "learning_rate": 6.630064075081785e-05, + "loss": 0.0481, + "step": 21230 + }, + { + "grad_norm": 0.2593759298324585, + "learning_rate": 6.626937452642132e-05, + "loss": 0.041, + "step": 21240 + }, + { + "grad_norm": 0.31575798988342285, + "learning_rate": 6.623810118524765e-05, + "loss": 0.0413, + "step": 21250 + }, + { + "grad_norm": 0.26852625608444214, + "learning_rate": 6.620682074097682e-05, + "loss": 0.0532, + "step": 21260 + }, + { + "grad_norm": 0.2606206238269806, + "learning_rate": 6.6175533207292e-05, + "loss": 0.0448, + "step": 21270 + }, + { + "grad_norm": 0.29901719093322754, + "learning_rate": 6.61442385978794e-05, + "loss": 0.0497, + "step": 21280 + }, + { + "grad_norm": 0.21151910722255707, + "learning_rate": 6.611293692642836e-05, + "loss": 0.0445, + "step": 21290 + }, + { + "grad_norm": 0.19500747323036194, + "learning_rate": 6.608162820663128e-05, + "loss": 0.0421, + "step": 21300 + }, + { + "grad_norm": 0.2103758007287979, + "learning_rate": 6.605031245218365e-05, + "loss": 0.0452, + "step": 21310 + }, + { + "grad_norm": 0.42192313075065613, + "learning_rate": 6.601898967678408e-05, + "loss": 0.0442, + "step": 21320 + }, + { + "grad_norm": 0.2420249879360199, + "learning_rate": 6.598765989413419e-05, + "loss": 0.0512, + "step": 21330 + }, + { + "grad_norm": 0.2971647381782532, + "learning_rate": 6.595632311793867e-05, + "loss": 0.0418, + "step": 21340 + }, + { + "grad_norm": 0.24438196420669556, + "learning_rate": 6.592497936190534e-05, + "loss": 0.044, + "step": 21350 + }, + { + "grad_norm": 0.4021906554698944, + "learning_rate": 6.5893628639745e-05, + "loss": 0.0434, + "step": 21360 + }, + { + "grad_norm": 0.29024767875671387, + "learning_rate": 6.586227096517151e-05, + "loss": 0.0463, + "step": 21370 + }, + { + "grad_norm": 0.2133564054965973, + "learning_rate": 6.58309063519018e-05, + "loss": 0.0412, + "step": 21380 + }, + { + "grad_norm": 0.24389421939849854, + "learning_rate": 6.579953481365582e-05, + "loss": 0.0382, + "step": 21390 + }, + { + "grad_norm": 0.30771970748901367, + "learning_rate": 6.576815636415652e-05, + "loss": 0.0465, + "step": 21400 + }, + { + "grad_norm": 0.21605297923088074, + "learning_rate": 6.573677101712991e-05, + "loss": 0.0506, + "step": 21410 + }, + { + "grad_norm": 0.20667588710784912, + "learning_rate": 6.570537878630502e-05, + "loss": 0.0412, + "step": 21420 + }, + { + "grad_norm": 0.24841894209384918, + "learning_rate": 6.56739796854139e-05, + "loss": 0.0412, + "step": 21430 + }, + { + "grad_norm": 3.990542411804199, + "learning_rate": 6.564257372819153e-05, + "loss": 0.0398, + "step": 21440 + }, + { + "grad_norm": 0.24460607767105103, + "learning_rate": 6.561116092837596e-05, + "loss": 0.0491, + "step": 21450 + }, + { + "grad_norm": 0.4791257083415985, + "learning_rate": 6.557974129970825e-05, + "loss": 0.0384, + "step": 21460 + }, + { + "grad_norm": 0.3347814083099365, + "learning_rate": 6.554831485593239e-05, + "loss": 0.0498, + "step": 21470 + }, + { + "grad_norm": 0.20645277202129364, + "learning_rate": 6.551688161079534e-05, + "loss": 0.038, + "step": 21480 + }, + { + "grad_norm": 0.3636249899864197, + "learning_rate": 6.548544157804712e-05, + "loss": 0.0438, + "step": 21490 + }, + { + "grad_norm": 0.202219620347023, + "learning_rate": 6.545399477144065e-05, + "loss": 0.0408, + "step": 21500 + }, + { + "grad_norm": 0.22704562544822693, + "learning_rate": 6.542254120473184e-05, + "loss": 0.0436, + "step": 21510 + }, + { + "grad_norm": 0.30232369899749756, + "learning_rate": 6.539108089167953e-05, + "loss": 0.0425, + "step": 21520 + }, + { + "grad_norm": 0.31295961141586304, + "learning_rate": 6.535961384604554e-05, + "loss": 0.0411, + "step": 21530 + }, + { + "grad_norm": 0.21097449958324432, + "learning_rate": 6.532814008159461e-05, + "loss": 0.0397, + "step": 21540 + }, + { + "grad_norm": 0.24861901998519897, + "learning_rate": 6.529665961209446e-05, + "loss": 0.0419, + "step": 21550 + }, + { + "grad_norm": 0.3834035396575928, + "learning_rate": 6.52651724513157e-05, + "loss": 0.0442, + "step": 21560 + }, + { + "grad_norm": 0.2872314155101776, + "learning_rate": 6.523367861303189e-05, + "loss": 0.041, + "step": 21570 + }, + { + "grad_norm": 0.2485586404800415, + "learning_rate": 6.520217811101951e-05, + "loss": 0.04, + "step": 21580 + }, + { + "grad_norm": 0.31485962867736816, + "learning_rate": 6.517067095905793e-05, + "loss": 0.0465, + "step": 21590 + }, + { + "grad_norm": 0.18059074878692627, + "learning_rate": 6.513915717092948e-05, + "loss": 0.04, + "step": 21600 + }, + { + "grad_norm": 0.25693851709365845, + "learning_rate": 6.510763676041935e-05, + "loss": 0.0482, + "step": 21610 + }, + { + "grad_norm": 0.14031554758548737, + "learning_rate": 6.50761097413156e-05, + "loss": 0.0411, + "step": 21620 + }, + { + "grad_norm": 0.2882936894893646, + "learning_rate": 6.504457612740928e-05, + "loss": 0.0441, + "step": 21630 + }, + { + "grad_norm": 0.2750907242298126, + "learning_rate": 6.501303593249423e-05, + "loss": 0.0411, + "step": 21640 + }, + { + "grad_norm": 0.21626485884189606, + "learning_rate": 6.49814891703672e-05, + "loss": 0.0395, + "step": 21650 + }, + { + "grad_norm": 0.5418786406517029, + "learning_rate": 6.494993585482783e-05, + "loss": 0.0471, + "step": 21660 + }, + { + "grad_norm": 0.2148239016532898, + "learning_rate": 6.49183759996786e-05, + "loss": 0.0386, + "step": 21670 + }, + { + "grad_norm": 0.2557004988193512, + "learning_rate": 6.488680961872486e-05, + "loss": 0.0382, + "step": 21680 + }, + { + "grad_norm": 0.2030767947435379, + "learning_rate": 6.485523672577485e-05, + "loss": 0.0406, + "step": 21690 + }, + { + "grad_norm": 0.22912780940532684, + "learning_rate": 6.482365733463959e-05, + "loss": 0.0427, + "step": 21700 + }, + { + "grad_norm": 0.18666468560695648, + "learning_rate": 6.4792071459133e-05, + "loss": 0.0401, + "step": 21710 + }, + { + "grad_norm": 0.28178051114082336, + "learning_rate": 6.476047911307179e-05, + "loss": 0.0515, + "step": 21720 + }, + { + "grad_norm": 0.323742151260376, + "learning_rate": 6.472888031027556e-05, + "loss": 0.0416, + "step": 21730 + }, + { + "grad_norm": 0.23193258047103882, + "learning_rate": 6.469727506456666e-05, + "loss": 0.0441, + "step": 21740 + }, + { + "grad_norm": 0.39936506748199463, + "learning_rate": 6.466566338977034e-05, + "loss": 0.042, + "step": 21750 + }, + { + "grad_norm": 0.2593773901462555, + "learning_rate": 6.46340452997146e-05, + "loss": 0.0489, + "step": 21760 + }, + { + "grad_norm": 0.19762325286865234, + "learning_rate": 6.460242080823025e-05, + "loss": 0.0489, + "step": 21770 + }, + { + "grad_norm": 0.2608921229839325, + "learning_rate": 6.457078992915093e-05, + "loss": 0.0526, + "step": 21780 + }, + { + "grad_norm": 0.3807969391345978, + "learning_rate": 6.453915267631308e-05, + "loss": 0.0455, + "step": 21790 + }, + { + "grad_norm": 0.3065604269504547, + "learning_rate": 6.450750906355589e-05, + "loss": 0.0494, + "step": 21800 + }, + { + "grad_norm": 0.21343564987182617, + "learning_rate": 6.447585910472134e-05, + "loss": 0.0377, + "step": 21810 + }, + { + "grad_norm": 0.15237128734588623, + "learning_rate": 6.444420281365424e-05, + "loss": 0.0415, + "step": 21820 + }, + { + "grad_norm": 0.27995485067367554, + "learning_rate": 6.441254020420208e-05, + "loss": 0.051, + "step": 21830 + }, + { + "grad_norm": 0.19680631160736084, + "learning_rate": 6.438087129021522e-05, + "loss": 0.0458, + "step": 21840 + }, + { + "grad_norm": 0.24782589077949524, + "learning_rate": 6.434919608554663e-05, + "loss": 0.0476, + "step": 21850 + }, + { + "grad_norm": 0.2683826684951782, + "learning_rate": 6.43175146040522e-05, + "loss": 0.045, + "step": 21860 + }, + { + "grad_norm": 0.5573546886444092, + "learning_rate": 6.428582685959045e-05, + "loss": 0.0507, + "step": 21870 + }, + { + "grad_norm": 0.25505775213241577, + "learning_rate": 6.42541328660227e-05, + "loss": 0.0422, + "step": 21880 + }, + { + "grad_norm": 0.23029331862926483, + "learning_rate": 6.422243263721293e-05, + "loss": 0.043, + "step": 21890 + }, + { + "grad_norm": 0.3023585081100464, + "learning_rate": 6.419072618702794e-05, + "loss": 0.0428, + "step": 21900 + }, + { + "grad_norm": 0.2617660462856293, + "learning_rate": 6.41590135293372e-05, + "loss": 0.0458, + "step": 21910 + }, + { + "grad_norm": 0.24395586550235748, + "learning_rate": 6.41272946780129e-05, + "loss": 0.0447, + "step": 21920 + }, + { + "grad_norm": 0.1870013177394867, + "learning_rate": 6.409556964692996e-05, + "loss": 0.041, + "step": 21930 + }, + { + "grad_norm": 0.2870008647441864, + "learning_rate": 6.406383844996595e-05, + "loss": 0.0442, + "step": 21940 + }, + { + "grad_norm": 0.2589300274848938, + "learning_rate": 6.403210110100121e-05, + "loss": 0.0392, + "step": 21950 + }, + { + "grad_norm": 0.21494160592556, + "learning_rate": 6.400035761391871e-05, + "loss": 0.0401, + "step": 21960 + }, + { + "grad_norm": 0.20223873853683472, + "learning_rate": 6.396860800260416e-05, + "loss": 0.0438, + "step": 21970 + }, + { + "grad_norm": 0.2523810565471649, + "learning_rate": 6.393685228094588e-05, + "loss": 0.0403, + "step": 21980 + }, + { + "grad_norm": 0.3039255440235138, + "learning_rate": 6.390509046283493e-05, + "loss": 0.0394, + "step": 21990 + }, + { + "grad_norm": 0.2716519832611084, + "learning_rate": 6.387332256216498e-05, + "loss": 0.0415, + "step": 22000 + }, + { + "grad_norm": 0.19715377688407898, + "learning_rate": 6.384154859283243e-05, + "loss": 0.0399, + "step": 22010 + }, + { + "grad_norm": 0.4142388105392456, + "learning_rate": 6.380976856873627e-05, + "loss": 0.0445, + "step": 22020 + }, + { + "grad_norm": 0.24471324682235718, + "learning_rate": 6.377798250377815e-05, + "loss": 0.044, + "step": 22030 + }, + { + "grad_norm": 0.6157432794570923, + "learning_rate": 6.374619041186238e-05, + "loss": 0.0433, + "step": 22040 + }, + { + "grad_norm": 0.1925642043352127, + "learning_rate": 6.371439230689592e-05, + "loss": 0.0402, + "step": 22050 + }, + { + "grad_norm": 0.20039434731006622, + "learning_rate": 6.368258820278833e-05, + "loss": 0.0406, + "step": 22060 + }, + { + "grad_norm": 0.23832319676876068, + "learning_rate": 6.365077811345182e-05, + "loss": 0.0414, + "step": 22070 + }, + { + "grad_norm": 0.2361021637916565, + "learning_rate": 6.361896205280117e-05, + "loss": 0.0374, + "step": 22080 + }, + { + "grad_norm": 0.20611290633678436, + "learning_rate": 6.358714003475384e-05, + "loss": 0.0468, + "step": 22090 + }, + { + "grad_norm": 0.2925374507904053, + "learning_rate": 6.355531207322983e-05, + "loss": 0.0431, + "step": 22100 + }, + { + "grad_norm": 0.21961981058120728, + "learning_rate": 6.35234781821518e-05, + "loss": 0.052, + "step": 22110 + }, + { + "grad_norm": 0.287398099899292, + "learning_rate": 6.349163837544497e-05, + "loss": 0.0479, + "step": 22120 + }, + { + "grad_norm": 0.39964717626571655, + "learning_rate": 6.345979266703714e-05, + "loss": 0.046, + "step": 22130 + }, + { + "grad_norm": 0.21497699618339539, + "learning_rate": 6.342794107085872e-05, + "loss": 0.0434, + "step": 22140 + }, + { + "grad_norm": 0.2133675068616867, + "learning_rate": 6.339608360084265e-05, + "loss": 0.0404, + "step": 22150 + }, + { + "grad_norm": 0.22311994433403015, + "learning_rate": 6.336422027092452e-05, + "loss": 0.0398, + "step": 22160 + }, + { + "grad_norm": 0.5785652995109558, + "learning_rate": 6.333235109504241e-05, + "loss": 0.0373, + "step": 22170 + }, + { + "grad_norm": 0.8199707269668579, + "learning_rate": 6.330047608713695e-05, + "loss": 0.0427, + "step": 22180 + }, + { + "grad_norm": 0.1417238861322403, + "learning_rate": 6.32685952611514e-05, + "loss": 0.0394, + "step": 22190 + }, + { + "grad_norm": 0.15882174670696259, + "learning_rate": 6.323670863103149e-05, + "loss": 0.0374, + "step": 22200 + }, + { + "grad_norm": 0.18418791890144348, + "learning_rate": 6.32048162107255e-05, + "loss": 0.0474, + "step": 22210 + }, + { + "grad_norm": 0.24710609018802643, + "learning_rate": 6.317291801418431e-05, + "loss": 0.0382, + "step": 22220 + }, + { + "grad_norm": 0.19000141322612762, + "learning_rate": 6.31410140553612e-05, + "loss": 0.0351, + "step": 22230 + }, + { + "grad_norm": 0.3168812096118927, + "learning_rate": 6.310910434821211e-05, + "loss": 0.0417, + "step": 22240 + }, + { + "grad_norm": 0.23559285700321198, + "learning_rate": 6.307718890669541e-05, + "loss": 0.0372, + "step": 22250 + }, + { + "grad_norm": 0.2305229753255844, + "learning_rate": 6.304526774477198e-05, + "loss": 0.0394, + "step": 22260 + }, + { + "grad_norm": 0.16326230764389038, + "learning_rate": 6.301334087640526e-05, + "loss": 0.034, + "step": 22270 + }, + { + "grad_norm": 0.15433450043201447, + "learning_rate": 6.298140831556112e-05, + "loss": 0.0417, + "step": 22280 + }, + { + "grad_norm": 0.1992369294166565, + "learning_rate": 6.294947007620794e-05, + "loss": 0.0391, + "step": 22290 + }, + { + "grad_norm": 0.19086134433746338, + "learning_rate": 6.291752617231661e-05, + "loss": 0.0373, + "step": 22300 + }, + { + "grad_norm": 0.20376954972743988, + "learning_rate": 6.288557661786047e-05, + "loss": 0.0369, + "step": 22310 + }, + { + "grad_norm": 0.26466166973114014, + "learning_rate": 6.285362142681534e-05, + "loss": 0.0442, + "step": 22320 + }, + { + "grad_norm": 0.21691085398197174, + "learning_rate": 6.28216606131595e-05, + "loss": 0.036, + "step": 22330 + }, + { + "grad_norm": 0.1116667240858078, + "learning_rate": 6.27896941908737e-05, + "loss": 0.0409, + "step": 22340 + }, + { + "grad_norm": 0.24572524428367615, + "learning_rate": 6.275772217394115e-05, + "loss": 0.0447, + "step": 22350 + }, + { + "grad_norm": 0.14841492474079132, + "learning_rate": 6.27257445763475e-05, + "loss": 0.0378, + "step": 22360 + }, + { + "grad_norm": 0.15640942752361298, + "learning_rate": 6.269376141208082e-05, + "loss": 0.0475, + "step": 22370 + }, + { + "grad_norm": 0.28329700231552124, + "learning_rate": 6.266177269513162e-05, + "loss": 0.041, + "step": 22380 + }, + { + "grad_norm": 0.22019749879837036, + "learning_rate": 6.262977843949289e-05, + "loss": 0.0326, + "step": 22390 + }, + { + "grad_norm": 0.2789740264415741, + "learning_rate": 6.259777865915997e-05, + "loss": 0.0419, + "step": 22400 + }, + { + "grad_norm": 0.43104198575019836, + "learning_rate": 6.256577336813069e-05, + "loss": 0.0487, + "step": 22410 + }, + { + "grad_norm": 0.2059200406074524, + "learning_rate": 6.253376258040521e-05, + "loss": 0.0426, + "step": 22420 + }, + { + "grad_norm": 0.1832166612148285, + "learning_rate": 6.250174630998616e-05, + "loss": 0.0364, + "step": 22430 + }, + { + "grad_norm": 0.1470152586698532, + "learning_rate": 6.246972457087854e-05, + "loss": 0.0371, + "step": 22440 + }, + { + "grad_norm": 0.275521844625473, + "learning_rate": 6.243769737708974e-05, + "loss": 0.0407, + "step": 22450 + }, + { + "grad_norm": 0.23849360644817352, + "learning_rate": 6.240566474262955e-05, + "loss": 0.0408, + "step": 22460 + }, + { + "grad_norm": 0.2471248209476471, + "learning_rate": 6.237362668151012e-05, + "loss": 0.0433, + "step": 22470 + }, + { + "grad_norm": 0.2630729377269745, + "learning_rate": 6.2341583207746e-05, + "loss": 0.0419, + "step": 22480 + }, + { + "grad_norm": 0.33960890769958496, + "learning_rate": 6.23095343353541e-05, + "loss": 0.0385, + "step": 22490 + }, + { + "grad_norm": 0.16153781116008759, + "learning_rate": 6.227748007835366e-05, + "loss": 0.0366, + "step": 22500 + }, + { + "grad_norm": 0.24375514686107635, + "learning_rate": 6.224542045076632e-05, + "loss": 0.0361, + "step": 22510 + }, + { + "grad_norm": 0.13429024815559387, + "learning_rate": 6.221335546661606e-05, + "loss": 0.0382, + "step": 22520 + }, + { + "grad_norm": 0.24162112176418304, + "learning_rate": 6.218128513992917e-05, + "loss": 0.0431, + "step": 22530 + }, + { + "grad_norm": 0.15499433875083923, + "learning_rate": 6.214920948473433e-05, + "loss": 0.0329, + "step": 22540 + }, + { + "grad_norm": 0.13873636722564697, + "learning_rate": 6.211712851506249e-05, + "loss": 0.0427, + "step": 22550 + }, + { + "grad_norm": 0.2822945713996887, + "learning_rate": 6.208504224494696e-05, + "loss": 0.0453, + "step": 22560 + }, + { + "grad_norm": 0.24575242400169373, + "learning_rate": 6.205295068842338e-05, + "loss": 0.0405, + "step": 22570 + }, + { + "grad_norm": 0.13659469783306122, + "learning_rate": 6.202085385952968e-05, + "loss": 0.038, + "step": 22580 + }, + { + "grad_norm": 0.19783326983451843, + "learning_rate": 6.19887517723061e-05, + "loss": 0.0328, + "step": 22590 + }, + { + "grad_norm": 0.1814902275800705, + "learning_rate": 6.19566444407952e-05, + "loss": 0.034, + "step": 22600 + }, + { + "grad_norm": 0.18354730308055878, + "learning_rate": 6.192453187904178e-05, + "loss": 0.036, + "step": 22610 + }, + { + "grad_norm": 0.17712073028087616, + "learning_rate": 6.189241410109303e-05, + "loss": 0.0344, + "step": 22620 + }, + { + "grad_norm": 0.16313575208187103, + "learning_rate": 6.186029112099829e-05, + "loss": 0.0351, + "step": 22630 + }, + { + "grad_norm": 0.24425248801708221, + "learning_rate": 6.182816295280926e-05, + "loss": 0.045, + "step": 22640 + }, + { + "grad_norm": 0.27053943276405334, + "learning_rate": 6.179602961057993e-05, + "loss": 0.0375, + "step": 22650 + }, + { + "grad_norm": 0.27765899896621704, + "learning_rate": 6.176389110836647e-05, + "loss": 0.039, + "step": 22660 + }, + { + "grad_norm": 0.2883318066596985, + "learning_rate": 6.173174746022738e-05, + "loss": 0.0401, + "step": 22670 + }, + { + "grad_norm": 0.3245389461517334, + "learning_rate": 6.169959868022338e-05, + "loss": 0.0388, + "step": 22680 + }, + { + "grad_norm": 0.2832232415676117, + "learning_rate": 6.166744478241742e-05, + "loss": 0.0386, + "step": 22690 + }, + { + "grad_norm": 0.24988260865211487, + "learning_rate": 6.163528578087473e-05, + "loss": 0.0394, + "step": 22700 + }, + { + "grad_norm": 0.2935691475868225, + "learning_rate": 6.160312168966274e-05, + "loss": 0.044, + "step": 22710 + }, + { + "grad_norm": 0.23527897894382477, + "learning_rate": 6.157095252285112e-05, + "loss": 0.038, + "step": 22720 + }, + { + "grad_norm": 0.28962263464927673, + "learning_rate": 6.153877829451174e-05, + "loss": 0.0423, + "step": 22730 + }, + { + "grad_norm": 0.27447453141212463, + "learning_rate": 6.150659901871872e-05, + "loss": 0.0446, + "step": 22740 + }, + { + "grad_norm": 0.2491799145936966, + "learning_rate": 6.147441470954836e-05, + "loss": 0.0494, + "step": 22750 + }, + { + "grad_norm": 0.30069226026535034, + "learning_rate": 6.144222538107918e-05, + "loss": 0.0518, + "step": 22760 + }, + { + "grad_norm": 0.20076125860214233, + "learning_rate": 6.141003104739188e-05, + "loss": 0.0436, + "step": 22770 + }, + { + "grad_norm": 0.28307682275772095, + "learning_rate": 6.137783172256935e-05, + "loss": 0.0406, + "step": 22780 + }, + { + "grad_norm": 0.16537494957447052, + "learning_rate": 6.134562742069669e-05, + "loss": 0.0426, + "step": 22790 + }, + { + "grad_norm": 0.2814267575740814, + "learning_rate": 6.131341815586111e-05, + "loss": 0.0376, + "step": 22800 + }, + { + "grad_norm": 0.26762107014656067, + "learning_rate": 6.128120394215209e-05, + "loss": 0.044, + "step": 22810 + }, + { + "grad_norm": 0.2902795672416687, + "learning_rate": 6.124898479366116e-05, + "loss": 0.0492, + "step": 22820 + }, + { + "grad_norm": 0.2982543110847473, + "learning_rate": 6.121676072448214e-05, + "loss": 0.0382, + "step": 22830 + }, + { + "grad_norm": 0.20312157273292542, + "learning_rate": 6.118453174871089e-05, + "loss": 0.051, + "step": 22840 + }, + { + "grad_norm": 0.21027754247188568, + "learning_rate": 6.115229788044548e-05, + "loss": 0.0547, + "step": 22850 + }, + { + "grad_norm": 0.23976849019527435, + "learning_rate": 6.112005913378607e-05, + "loss": 0.0401, + "step": 22860 + }, + { + "grad_norm": 0.24301859736442566, + "learning_rate": 6.1087815522835e-05, + "loss": 0.0461, + "step": 22870 + }, + { + "grad_norm": 0.23062682151794434, + "learning_rate": 6.105556706169673e-05, + "loss": 0.0498, + "step": 22880 + }, + { + "grad_norm": 0.18718381226062775, + "learning_rate": 6.102331376447782e-05, + "loss": 0.0435, + "step": 22890 + }, + { + "grad_norm": 0.19916605949401855, + "learning_rate": 6.099105564528697e-05, + "loss": 0.0523, + "step": 22900 + }, + { + "grad_norm": 0.5237333178520203, + "learning_rate": 6.095879271823495e-05, + "loss": 0.0453, + "step": 22910 + }, + { + "grad_norm": 0.16829967498779297, + "learning_rate": 6.09265249974347e-05, + "loss": 0.0469, + "step": 22920 + }, + { + "grad_norm": 0.20390939712524414, + "learning_rate": 6.089425249700118e-05, + "loss": 0.0442, + "step": 22930 + }, + { + "grad_norm": 0.32718709111213684, + "learning_rate": 6.0861975231051505e-05, + "loss": 0.0418, + "step": 22940 + }, + { + "grad_norm": 0.3874143064022064, + "learning_rate": 6.082969321370484e-05, + "loss": 0.0483, + "step": 22950 + }, + { + "grad_norm": 0.32879629731178284, + "learning_rate": 6.079740645908243e-05, + "loss": 0.0408, + "step": 22960 + }, + { + "grad_norm": 0.32888078689575195, + "learning_rate": 6.07651149813076e-05, + "loss": 0.0415, + "step": 22970 + }, + { + "grad_norm": 0.1920211911201477, + "learning_rate": 6.073281879450574e-05, + "loss": 0.0462, + "step": 22980 + }, + { + "grad_norm": 0.36741429567337036, + "learning_rate": 6.070051791280431e-05, + "loss": 0.046, + "step": 22990 + }, + { + "grad_norm": 0.19778962433338165, + "learning_rate": 6.066821235033281e-05, + "loss": 0.04, + "step": 23000 + }, + { + "grad_norm": 0.23873370885849, + "learning_rate": 6.063590212122277e-05, + "loss": 0.042, + "step": 23010 + }, + { + "grad_norm": 0.28695908188819885, + "learning_rate": 6.0603587239607794e-05, + "loss": 0.0414, + "step": 23020 + }, + { + "grad_norm": 0.36111414432525635, + "learning_rate": 6.057126771962353e-05, + "loss": 0.0488, + "step": 23030 + }, + { + "grad_norm": 0.22798055410385132, + "learning_rate": 6.05389435754076e-05, + "loss": 0.0423, + "step": 23040 + }, + { + "grad_norm": 0.3858727812767029, + "learning_rate": 6.0506614821099705e-05, + "loss": 0.0494, + "step": 23050 + }, + { + "grad_norm": 0.23563183844089508, + "learning_rate": 6.047428147084153e-05, + "loss": 0.042, + "step": 23060 + }, + { + "grad_norm": 0.23517069220542908, + "learning_rate": 6.044194353877679e-05, + "loss": 0.0375, + "step": 23070 + }, + { + "grad_norm": 0.21315550804138184, + "learning_rate": 6.04096010390512e-05, + "loss": 0.04, + "step": 23080 + }, + { + "grad_norm": 0.27107566595077515, + "learning_rate": 6.0377253985812445e-05, + "loss": 0.038, + "step": 23090 + }, + { + "grad_norm": 0.18437093496322632, + "learning_rate": 6.0344902393210235e-05, + "loss": 0.0447, + "step": 23100 + }, + { + "grad_norm": 0.22087988257408142, + "learning_rate": 6.031254627539625e-05, + "loss": 0.047, + "step": 23110 + }, + { + "grad_norm": 0.2387913167476654, + "learning_rate": 6.028018564652418e-05, + "loss": 0.042, + "step": 23120 + }, + { + "grad_norm": 0.19238406419754028, + "learning_rate": 6.024782052074964e-05, + "loss": 0.0395, + "step": 23130 + }, + { + "grad_norm": 0.4068948030471802, + "learning_rate": 6.021545091223026e-05, + "loss": 0.0396, + "step": 23140 + }, + { + "grad_norm": 0.18338733911514282, + "learning_rate": 6.0183076835125586e-05, + "loss": 0.0467, + "step": 23150 + }, + { + "grad_norm": 0.249656543135643, + "learning_rate": 6.015069830359714e-05, + "loss": 0.0399, + "step": 23160 + }, + { + "grad_norm": 0.248051255941391, + "learning_rate": 6.0118315331808384e-05, + "loss": 0.0409, + "step": 23170 + }, + { + "grad_norm": 0.20149396359920502, + "learning_rate": 6.0085927933924745e-05, + "loss": 0.0395, + "step": 23180 + }, + { + "grad_norm": 0.1738068014383316, + "learning_rate": 6.0053536124113585e-05, + "loss": 0.043, + "step": 23190 + }, + { + "grad_norm": 0.22969108819961548, + "learning_rate": 6.0021139916544144e-05, + "loss": 0.038, + "step": 23200 + }, + { + "grad_norm": 0.2316327542066574, + "learning_rate": 5.998873932538764e-05, + "loss": 0.0396, + "step": 23210 + }, + { + "grad_norm": 0.20128203928470612, + "learning_rate": 5.995633436481721e-05, + "loss": 0.0389, + "step": 23220 + }, + { + "grad_norm": 0.18468771874904633, + "learning_rate": 5.9923925049007855e-05, + "loss": 0.0398, + "step": 23230 + }, + { + "grad_norm": 0.18566346168518066, + "learning_rate": 5.989151139213656e-05, + "loss": 0.0392, + "step": 23240 + }, + { + "grad_norm": 0.18482564389705658, + "learning_rate": 5.9859093408382094e-05, + "loss": 0.038, + "step": 23250 + }, + { + "grad_norm": 0.31279876828193665, + "learning_rate": 5.982667111192523e-05, + "loss": 0.0391, + "step": 23260 + }, + { + "grad_norm": 0.27162909507751465, + "learning_rate": 5.979424451694857e-05, + "loss": 0.0447, + "step": 23270 + }, + { + "grad_norm": 0.24170444905757904, + "learning_rate": 5.97618136376366e-05, + "loss": 0.0409, + "step": 23280 + }, + { + "grad_norm": 0.23025526106357574, + "learning_rate": 5.97293784881757e-05, + "loss": 0.05, + "step": 23290 + }, + { + "grad_norm": 0.2679314911365509, + "learning_rate": 5.969693908275411e-05, + "loss": 0.0415, + "step": 23300 + }, + { + "grad_norm": 0.29347410798072815, + "learning_rate": 5.9664495435561916e-05, + "loss": 0.0455, + "step": 23310 + }, + { + "grad_norm": 0.19619110226631165, + "learning_rate": 5.963204756079108e-05, + "loss": 0.0398, + "step": 23320 + }, + { + "grad_norm": 0.3856193721294403, + "learning_rate": 5.959959547263538e-05, + "loss": 0.0398, + "step": 23330 + }, + { + "grad_norm": 0.2663048207759857, + "learning_rate": 5.95671391852905e-05, + "loss": 0.0425, + "step": 23340 + }, + { + "grad_norm": 0.6768359541893005, + "learning_rate": 5.95346787129539e-05, + "loss": 0.0466, + "step": 23350 + }, + { + "grad_norm": 0.15548175573349, + "learning_rate": 5.950221406982489e-05, + "loss": 0.041, + "step": 23360 + }, + { + "grad_norm": 0.2706908881664276, + "learning_rate": 5.9469745270104625e-05, + "loss": 0.0471, + "step": 23370 + }, + { + "grad_norm": 0.22814902663230896, + "learning_rate": 5.9437272327996064e-05, + "loss": 0.0456, + "step": 23380 + }, + { + "grad_norm": 0.18509364128112793, + "learning_rate": 5.940479525770395e-05, + "loss": 0.0455, + "step": 23390 + }, + { + "grad_norm": 0.18338188529014587, + "learning_rate": 5.9372314073434876e-05, + "loss": 0.041, + "step": 23400 + }, + { + "grad_norm": 0.28042495250701904, + "learning_rate": 5.93398287893972e-05, + "loss": 0.0448, + "step": 23410 + }, + { + "grad_norm": 0.16222524642944336, + "learning_rate": 5.930733941980111e-05, + "loss": 0.0343, + "step": 23420 + }, + { + "grad_norm": 0.28714415431022644, + "learning_rate": 5.927484597885855e-05, + "loss": 0.0408, + "step": 23430 + }, + { + "grad_norm": 0.2066669911146164, + "learning_rate": 5.924234848078325e-05, + "loss": 0.0394, + "step": 23440 + }, + { + "grad_norm": 0.20917826890945435, + "learning_rate": 5.9209846939790715e-05, + "loss": 0.0419, + "step": 23450 + }, + { + "grad_norm": 0.18547774851322174, + "learning_rate": 5.917734137009824e-05, + "loss": 0.0374, + "step": 23460 + }, + { + "grad_norm": 0.17115826904773712, + "learning_rate": 5.914483178592485e-05, + "loss": 0.0406, + "step": 23470 + }, + { + "grad_norm": 0.1512717455625534, + "learning_rate": 5.911231820149136e-05, + "loss": 0.0352, + "step": 23480 + }, + { + "grad_norm": 0.1681210845708847, + "learning_rate": 5.907980063102031e-05, + "loss": 0.0423, + "step": 23490 + }, + { + "grad_norm": 0.1712867021560669, + "learning_rate": 5.904727908873597e-05, + "loss": 0.0465, + "step": 23500 + }, + { + "grad_norm": 0.19941763579845428, + "learning_rate": 5.90147535888644e-05, + "loss": 0.0442, + "step": 23510 + }, + { + "grad_norm": 0.21636046469211578, + "learning_rate": 5.898222414563333e-05, + "loss": 0.0497, + "step": 23520 + }, + { + "grad_norm": 0.21234756708145142, + "learning_rate": 5.8949690773272256e-05, + "loss": 0.0343, + "step": 23530 + }, + { + "grad_norm": 0.20401786267757416, + "learning_rate": 5.891715348601239e-05, + "loss": 0.0424, + "step": 23540 + }, + { + "grad_norm": 0.2129133939743042, + "learning_rate": 5.888461229808663e-05, + "loss": 0.0394, + "step": 23550 + }, + { + "grad_norm": 0.20955686271190643, + "learning_rate": 5.885206722372959e-05, + "loss": 0.0414, + "step": 23560 + }, + { + "grad_norm": 0.16619403660297394, + "learning_rate": 5.881951827717763e-05, + "loss": 0.0351, + "step": 23570 + }, + { + "grad_norm": 0.1975027322769165, + "learning_rate": 5.878696547266872e-05, + "loss": 0.0387, + "step": 23580 + }, + { + "grad_norm": 0.24111974239349365, + "learning_rate": 5.875440882444259e-05, + "loss": 0.0454, + "step": 23590 + }, + { + "grad_norm": 0.30113711953163147, + "learning_rate": 5.8721848346740617e-05, + "loss": 0.0428, + "step": 23600 + }, + { + "grad_norm": 0.17772018909454346, + "learning_rate": 5.868928405380585e-05, + "loss": 0.0405, + "step": 23610 + }, + { + "grad_norm": 0.1777854710817337, + "learning_rate": 5.865671595988304e-05, + "loss": 0.038, + "step": 23620 + }, + { + "grad_norm": 0.18491272628307343, + "learning_rate": 5.862414407921857e-05, + "loss": 0.0363, + "step": 23630 + }, + { + "grad_norm": 0.22004805505275726, + "learning_rate": 5.8591568426060475e-05, + "loss": 0.046, + "step": 23640 + }, + { + "grad_norm": 0.2650536000728607, + "learning_rate": 5.8558989014658464e-05, + "loss": 0.0415, + "step": 23650 + }, + { + "grad_norm": 0.15616431832313538, + "learning_rate": 5.8526405859263855e-05, + "loss": 0.0409, + "step": 23660 + }, + { + "grad_norm": 0.2005353718996048, + "learning_rate": 5.8493818974129667e-05, + "loss": 0.0384, + "step": 23670 + }, + { + "grad_norm": 0.21041050553321838, + "learning_rate": 5.846122837351048e-05, + "loss": 0.048, + "step": 23680 + }, + { + "grad_norm": 0.2210773527622223, + "learning_rate": 5.842863407166255e-05, + "loss": 0.0413, + "step": 23690 + }, + { + "grad_norm": 0.2451685070991516, + "learning_rate": 5.83960360828437e-05, + "loss": 0.0424, + "step": 23700 + }, + { + "grad_norm": 0.3020227551460266, + "learning_rate": 5.836343442131343e-05, + "loss": 0.0417, + "step": 23710 + }, + { + "grad_norm": 0.15161046385765076, + "learning_rate": 5.833082910133281e-05, + "loss": 0.0424, + "step": 23720 + }, + { + "grad_norm": 0.2967725992202759, + "learning_rate": 5.82982201371645e-05, + "loss": 0.0482, + "step": 23730 + }, + { + "grad_norm": 0.31087496876716614, + "learning_rate": 5.826560754307279e-05, + "loss": 0.0432, + "step": 23740 + }, + { + "grad_norm": 0.19199047982692719, + "learning_rate": 5.8232991333323516e-05, + "loss": 0.0485, + "step": 23750 + }, + { + "grad_norm": 0.2515293061733246, + "learning_rate": 5.8200371522184116e-05, + "loss": 0.0472, + "step": 23760 + }, + { + "grad_norm": 0.18196968734264374, + "learning_rate": 5.816774812392362e-05, + "loss": 0.0427, + "step": 23770 + }, + { + "grad_norm": 0.24861595034599304, + "learning_rate": 5.8135121152812614e-05, + "loss": 0.0444, + "step": 23780 + }, + { + "grad_norm": 0.24612261354923248, + "learning_rate": 5.8102490623123204e-05, + "loss": 0.047, + "step": 23790 + }, + { + "grad_norm": 0.19741815328598022, + "learning_rate": 5.806985654912914e-05, + "loss": 0.0464, + "step": 23800 + }, + { + "grad_norm": 0.2319311797618866, + "learning_rate": 5.8037218945105644e-05, + "loss": 0.0455, + "step": 23810 + }, + { + "grad_norm": 0.3774605989456177, + "learning_rate": 5.800457782532953e-05, + "loss": 0.0454, + "step": 23820 + }, + { + "grad_norm": 0.32694074511528015, + "learning_rate": 5.797193320407912e-05, + "loss": 0.0429, + "step": 23830 + }, + { + "grad_norm": 0.33883824944496155, + "learning_rate": 5.793928509563431e-05, + "loss": 0.0443, + "step": 23840 + }, + { + "grad_norm": 0.25539833307266235, + "learning_rate": 5.790663351427644e-05, + "loss": 0.0431, + "step": 23850 + }, + { + "grad_norm": 0.16766542196273804, + "learning_rate": 5.7873978474288484e-05, + "loss": 0.044, + "step": 23860 + }, + { + "grad_norm": 0.2735307216644287, + "learning_rate": 5.78413199899548e-05, + "loss": 0.0482, + "step": 23870 + }, + { + "grad_norm": 0.27288922667503357, + "learning_rate": 5.780865807556136e-05, + "loss": 0.0428, + "step": 23880 + }, + { + "grad_norm": 0.20512767136096954, + "learning_rate": 5.777599274539559e-05, + "loss": 0.0524, + "step": 23890 + }, + { + "grad_norm": 0.21579177677631378, + "learning_rate": 5.77433240137464e-05, + "loss": 0.0478, + "step": 23900 + }, + { + "grad_norm": 0.2850676476955414, + "learning_rate": 5.771065189490422e-05, + "loss": 0.0478, + "step": 23910 + }, + { + "grad_norm": 0.2298383265733719, + "learning_rate": 5.767797640316095e-05, + "loss": 0.0538, + "step": 23920 + }, + { + "grad_norm": 0.20415417850017548, + "learning_rate": 5.764529755280994e-05, + "loss": 0.0462, + "step": 23930 + }, + { + "grad_norm": 0.2318338006734848, + "learning_rate": 5.761261535814603e-05, + "loss": 0.0432, + "step": 23940 + }, + { + "grad_norm": 0.30676454305648804, + "learning_rate": 5.757992983346554e-05, + "loss": 0.0412, + "step": 23950 + }, + { + "grad_norm": 0.31899306178092957, + "learning_rate": 5.754724099306621e-05, + "loss": 0.0577, + "step": 23960 + }, + { + "grad_norm": 0.23821024596691132, + "learning_rate": 5.751454885124727e-05, + "loss": 0.0504, + "step": 23970 + }, + { + "grad_norm": 0.18881940841674805, + "learning_rate": 5.748185342230935e-05, + "loss": 0.0446, + "step": 23980 + }, + { + "grad_norm": 0.19262129068374634, + "learning_rate": 5.744915472055456e-05, + "loss": 0.0492, + "step": 23990 + }, + { + "grad_norm": 0.2729032337665558, + "learning_rate": 5.741645276028642e-05, + "loss": 0.0424, + "step": 24000 + }, + { + "grad_norm": 0.2035289853811264, + "learning_rate": 5.7383747555809873e-05, + "loss": 0.0387, + "step": 24010 + }, + { + "grad_norm": 0.28312230110168457, + "learning_rate": 5.7351039121431294e-05, + "loss": 0.0418, + "step": 24020 + }, + { + "grad_norm": 0.34070152044296265, + "learning_rate": 5.7318327471458454e-05, + "loss": 0.047, + "step": 24030 + }, + { + "grad_norm": 0.27926838397979736, + "learning_rate": 5.728561262020054e-05, + "loss": 0.0457, + "step": 24040 + }, + { + "grad_norm": 0.23850099742412567, + "learning_rate": 5.725289458196814e-05, + "loss": 0.0462, + "step": 24050 + }, + { + "grad_norm": 0.23028390109539032, + "learning_rate": 5.722017337107324e-05, + "loss": 0.0484, + "step": 24060 + }, + { + "grad_norm": 0.17565253376960754, + "learning_rate": 5.7187449001829206e-05, + "loss": 0.0403, + "step": 24070 + }, + { + "grad_norm": 0.2969692349433899, + "learning_rate": 5.715472148855081e-05, + "loss": 0.0492, + "step": 24080 + }, + { + "grad_norm": 0.39781543612480164, + "learning_rate": 5.712199084555414e-05, + "loss": 0.0417, + "step": 24090 + }, + { + "grad_norm": 0.29589900374412537, + "learning_rate": 5.7089257087156734e-05, + "loss": 0.0506, + "step": 24100 + }, + { + "grad_norm": 0.22286370396614075, + "learning_rate": 5.70565202276774e-05, + "loss": 0.0398, + "step": 24110 + }, + { + "grad_norm": 0.25703203678131104, + "learning_rate": 5.70237802814364e-05, + "loss": 0.0379, + "step": 24120 + }, + { + "grad_norm": 0.22883641719818115, + "learning_rate": 5.6991037262755277e-05, + "loss": 0.0462, + "step": 24130 + }, + { + "grad_norm": 0.17019672691822052, + "learning_rate": 5.6958291185956924e-05, + "loss": 0.0467, + "step": 24140 + }, + { + "grad_norm": 0.5325593948364258, + "learning_rate": 5.692554206536561e-05, + "loss": 0.0432, + "step": 24150 + }, + { + "grad_norm": 0.18293291330337524, + "learning_rate": 5.689278991530692e-05, + "loss": 0.0477, + "step": 24160 + }, + { + "grad_norm": 0.3359695374965668, + "learning_rate": 5.686003475010773e-05, + "loss": 0.0427, + "step": 24170 + }, + { + "grad_norm": 0.19805194437503815, + "learning_rate": 5.682727658409628e-05, + "loss": 0.0473, + "step": 24180 + }, + { + "grad_norm": 0.3593471944332123, + "learning_rate": 5.67945154316021e-05, + "loss": 0.041, + "step": 24190 + }, + { + "grad_norm": 0.2109519988298416, + "learning_rate": 5.676175130695602e-05, + "loss": 0.0394, + "step": 24200 + }, + { + "grad_norm": 0.22293725609779358, + "learning_rate": 5.672898422449022e-05, + "loss": 0.0412, + "step": 24210 + }, + { + "grad_norm": 0.21390534937381744, + "learning_rate": 5.66962141985381e-05, + "loss": 0.0446, + "step": 24220 + }, + { + "grad_norm": 0.20379029214382172, + "learning_rate": 5.666344124343439e-05, + "loss": 0.0437, + "step": 24230 + }, + { + "grad_norm": 0.29900509119033813, + "learning_rate": 5.66306653735151e-05, + "loss": 0.0379, + "step": 24240 + }, + { + "grad_norm": 0.20484600961208344, + "learning_rate": 5.659788660311749e-05, + "loss": 0.0468, + "step": 24250 + }, + { + "grad_norm": 0.2844574451446533, + "learning_rate": 5.656510494658014e-05, + "loss": 0.0425, + "step": 24260 + }, + { + "grad_norm": 0.2159847617149353, + "learning_rate": 5.6532320418242835e-05, + "loss": 0.0402, + "step": 24270 + }, + { + "grad_norm": 0.2227267324924469, + "learning_rate": 5.649953303244665e-05, + "loss": 0.0419, + "step": 24280 + }, + { + "grad_norm": 0.20816011726856232, + "learning_rate": 5.646674280353389e-05, + "loss": 0.0483, + "step": 24290 + }, + { + "grad_norm": 0.1941169798374176, + "learning_rate": 5.6433949745848135e-05, + "loss": 0.0485, + "step": 24300 + }, + { + "grad_norm": 0.2996561825275421, + "learning_rate": 5.6401153873734166e-05, + "loss": 0.0462, + "step": 24310 + }, + { + "grad_norm": 0.5130030512809753, + "learning_rate": 5.6368355201538036e-05, + "loss": 0.043, + "step": 24320 + }, + { + "grad_norm": 0.23379091918468475, + "learning_rate": 5.633555374360697e-05, + "loss": 0.0377, + "step": 24330 + }, + { + "grad_norm": 0.18415650725364685, + "learning_rate": 5.630274951428944e-05, + "loss": 0.0443, + "step": 24340 + }, + { + "grad_norm": 0.18176016211509705, + "learning_rate": 5.626994252793517e-05, + "loss": 0.0489, + "step": 24350 + }, + { + "grad_norm": 0.3070266544818878, + "learning_rate": 5.6237132798895006e-05, + "loss": 0.0382, + "step": 24360 + }, + { + "grad_norm": 0.25669872760772705, + "learning_rate": 5.620432034152106e-05, + "loss": 0.0431, + "step": 24370 + }, + { + "grad_norm": 0.2144925892353058, + "learning_rate": 5.6171505170166604e-05, + "loss": 0.0384, + "step": 24380 + }, + { + "grad_norm": 0.5197767615318298, + "learning_rate": 5.613868729918612e-05, + "loss": 0.0399, + "step": 24390 + }, + { + "grad_norm": 0.3641085624694824, + "learning_rate": 5.6105866742935256e-05, + "loss": 0.0422, + "step": 24400 + }, + { + "grad_norm": 0.287893146276474, + "learning_rate": 5.607304351577083e-05, + "loss": 0.0404, + "step": 24410 + }, + { + "grad_norm": 0.2331518828868866, + "learning_rate": 5.604021763205085e-05, + "loss": 0.0413, + "step": 24420 + }, + { + "grad_norm": 0.2216121107339859, + "learning_rate": 5.600738910613445e-05, + "loss": 0.0435, + "step": 24430 + }, + { + "grad_norm": 0.27576881647109985, + "learning_rate": 5.597455795238198e-05, + "loss": 0.0344, + "step": 24440 + }, + { + "grad_norm": 0.19578826427459717, + "learning_rate": 5.5941724185154876e-05, + "loss": 0.0436, + "step": 24450 + }, + { + "grad_norm": 0.15418802201747894, + "learning_rate": 5.5908887818815746e-05, + "loss": 0.0441, + "step": 24460 + }, + { + "grad_norm": 0.2561571002006531, + "learning_rate": 5.587604886772834e-05, + "loss": 0.0394, + "step": 24470 + }, + { + "grad_norm": 0.23851102590560913, + "learning_rate": 5.584320734625752e-05, + "loss": 0.0414, + "step": 24480 + }, + { + "grad_norm": 0.25298061966896057, + "learning_rate": 5.581036326876928e-05, + "loss": 0.0381, + "step": 24490 + }, + { + "grad_norm": 0.20589864253997803, + "learning_rate": 5.5777516649630756e-05, + "loss": 0.0426, + "step": 24500 + }, + { + "grad_norm": 0.19082482159137726, + "learning_rate": 5.574466750321017e-05, + "loss": 0.0367, + "step": 24510 + }, + { + "grad_norm": 0.1936592310667038, + "learning_rate": 5.571181584387683e-05, + "loss": 0.0435, + "step": 24520 + }, + { + "grad_norm": 0.2812248766422272, + "learning_rate": 5.5678961686001196e-05, + "loss": 0.0391, + "step": 24530 + }, + { + "grad_norm": 0.3352706730365753, + "learning_rate": 5.5646105043954785e-05, + "loss": 0.0405, + "step": 24540 + }, + { + "grad_norm": 0.22116193175315857, + "learning_rate": 5.5613245932110205e-05, + "loss": 0.043, + "step": 24550 + }, + { + "grad_norm": 0.24838396906852722, + "learning_rate": 5.5580384364841157e-05, + "loss": 0.0363, + "step": 24560 + }, + { + "grad_norm": 0.22631792724132538, + "learning_rate": 5.554752035652237e-05, + "loss": 0.0401, + "step": 24570 + }, + { + "grad_norm": 0.2836701273918152, + "learning_rate": 5.5514653921529736e-05, + "loss": 0.038, + "step": 24580 + }, + { + "grad_norm": 0.20426374673843384, + "learning_rate": 5.5481785074240114e-05, + "loss": 0.0359, + "step": 24590 + }, + { + "grad_norm": 0.3109476566314697, + "learning_rate": 5.544891382903146e-05, + "loss": 0.0341, + "step": 24600 + }, + { + "grad_norm": 0.273123562335968, + "learning_rate": 5.541604020028277e-05, + "loss": 0.0456, + "step": 24610 + }, + { + "grad_norm": 0.2094653695821762, + "learning_rate": 5.5383164202374105e-05, + "loss": 0.0367, + "step": 24620 + }, + { + "grad_norm": 0.2390955686569214, + "learning_rate": 5.5350285849686515e-05, + "loss": 0.0463, + "step": 24630 + }, + { + "grad_norm": 0.284551739692688, + "learning_rate": 5.531740515660213e-05, + "loss": 0.0451, + "step": 24640 + }, + { + "grad_norm": 0.27283188700675964, + "learning_rate": 5.528452213750408e-05, + "loss": 0.0406, + "step": 24650 + }, + { + "grad_norm": 0.26220133900642395, + "learning_rate": 5.52516368067765e-05, + "loss": 0.0395, + "step": 24660 + }, + { + "grad_norm": 0.21715177595615387, + "learning_rate": 5.5218749178804563e-05, + "loss": 0.0399, + "step": 24670 + }, + { + "grad_norm": 0.2182496190071106, + "learning_rate": 5.518585926797443e-05, + "loss": 0.0359, + "step": 24680 + }, + { + "grad_norm": 0.2362930029630661, + "learning_rate": 5.5152967088673265e-05, + "loss": 0.0426, + "step": 24690 + }, + { + "grad_norm": 0.23282302916049957, + "learning_rate": 5.512007265528924e-05, + "loss": 0.0444, + "step": 24700 + }, + { + "grad_norm": 0.19850479066371918, + "learning_rate": 5.508717598221148e-05, + "loss": 0.0398, + "step": 24710 + }, + { + "grad_norm": 0.24651411175727844, + "learning_rate": 5.505427708383012e-05, + "loss": 0.0375, + "step": 24720 + }, + { + "grad_norm": 0.2699449956417084, + "learning_rate": 5.502137597453624e-05, + "loss": 0.0397, + "step": 24730 + }, + { + "grad_norm": 0.35601842403411865, + "learning_rate": 5.4988472668721914e-05, + "loss": 0.0393, + "step": 24740 + }, + { + "grad_norm": 0.20890773832798004, + "learning_rate": 5.4955567180780164e-05, + "loss": 0.0432, + "step": 24750 + }, + { + "grad_norm": 0.2932673692703247, + "learning_rate": 5.492265952510497e-05, + "loss": 0.0477, + "step": 24760 + }, + { + "grad_norm": 0.34370917081832886, + "learning_rate": 5.4889749716091254e-05, + "loss": 0.0423, + "step": 24770 + }, + { + "grad_norm": 0.1946830302476883, + "learning_rate": 5.4856837768134885e-05, + "loss": 0.042, + "step": 24780 + }, + { + "grad_norm": 0.20745769143104553, + "learning_rate": 5.482392369563267e-05, + "loss": 0.041, + "step": 24790 + }, + { + "grad_norm": 0.23703566193580627, + "learning_rate": 5.4791007512982354e-05, + "loss": 0.0348, + "step": 24800 + }, + { + "grad_norm": 0.23097532987594604, + "learning_rate": 5.4758089234582566e-05, + "loss": 0.0389, + "step": 24810 + }, + { + "grad_norm": 0.2957744896411896, + "learning_rate": 5.47251688748329e-05, + "loss": 0.0385, + "step": 24820 + }, + { + "grad_norm": 0.22601066529750824, + "learning_rate": 5.4692246448133843e-05, + "loss": 0.0421, + "step": 24830 + }, + { + "grad_norm": 0.3413033187389374, + "learning_rate": 5.465932196888676e-05, + "loss": 0.0398, + "step": 24840 + }, + { + "grad_norm": 0.192387655377388, + "learning_rate": 5.462639545149395e-05, + "loss": 0.037, + "step": 24850 + }, + { + "grad_norm": 0.16018234193325043, + "learning_rate": 5.459346691035861e-05, + "loss": 0.0365, + "step": 24860 + }, + { + "grad_norm": 0.25115710496902466, + "learning_rate": 5.4560536359884776e-05, + "loss": 0.0456, + "step": 24870 + }, + { + "grad_norm": 0.18478043377399445, + "learning_rate": 5.45276038144774e-05, + "loss": 0.0373, + "step": 24880 + }, + { + "grad_norm": 0.2164691537618637, + "learning_rate": 5.449466928854229e-05, + "loss": 0.0394, + "step": 24890 + }, + { + "grad_norm": 0.22718556225299835, + "learning_rate": 5.446173279648613e-05, + "loss": 0.0433, + "step": 24900 + }, + { + "grad_norm": 0.19831962883472443, + "learning_rate": 5.442879435271646e-05, + "loss": 0.0397, + "step": 24910 + }, + { + "grad_norm": 0.1635976880788803, + "learning_rate": 5.4395853971641675e-05, + "loss": 0.0391, + "step": 24920 + }, + { + "grad_norm": 0.2628844380378723, + "learning_rate": 5.4362911667671015e-05, + "loss": 0.0407, + "step": 24930 + }, + { + "grad_norm": 0.2720104455947876, + "learning_rate": 5.432996745521458e-05, + "loss": 0.041, + "step": 24940 + }, + { + "grad_norm": 0.22052864730358124, + "learning_rate": 5.4297021348683264e-05, + "loss": 0.0468, + "step": 24950 + }, + { + "grad_norm": 0.3534083366394043, + "learning_rate": 5.426407336248882e-05, + "loss": 0.0422, + "step": 24960 + }, + { + "grad_norm": 0.25173914432525635, + "learning_rate": 5.423112351104382e-05, + "loss": 0.0379, + "step": 24970 + }, + { + "grad_norm": 0.31828242540359497, + "learning_rate": 5.4198171808761654e-05, + "loss": 0.0438, + "step": 24980 + }, + { + "grad_norm": 0.5772734880447388, + "learning_rate": 5.41652182700565e-05, + "loss": 0.0528, + "step": 24990 + }, + { + "grad_norm": 0.227115198969841, + "learning_rate": 5.413226290934338e-05, + "loss": 0.0409, + "step": 25000 + }, + { + "grad_norm": 0.23156395554542542, + "learning_rate": 5.4099305741038056e-05, + "loss": 0.0386, + "step": 25010 + }, + { + "grad_norm": 0.20600104331970215, + "learning_rate": 5.406634677955713e-05, + "loss": 0.0407, + "step": 25020 + }, + { + "grad_norm": 0.2407175898551941, + "learning_rate": 5.403338603931798e-05, + "loss": 0.0498, + "step": 25030 + }, + { + "grad_norm": 0.1975805163383484, + "learning_rate": 5.4000423534738745e-05, + "loss": 0.035, + "step": 25040 + }, + { + "grad_norm": 0.1437540203332901, + "learning_rate": 5.396745928023835e-05, + "loss": 0.0419, + "step": 25050 + }, + { + "grad_norm": 0.20713253319263458, + "learning_rate": 5.3934493290236474e-05, + "loss": 0.0443, + "step": 25060 + }, + { + "grad_norm": 0.3323761224746704, + "learning_rate": 5.390152557915357e-05, + "loss": 0.0375, + "step": 25070 + }, + { + "grad_norm": 0.21251605451107025, + "learning_rate": 5.3868556161410796e-05, + "loss": 0.0423, + "step": 25080 + }, + { + "grad_norm": 0.22305183112621307, + "learning_rate": 5.383558505143015e-05, + "loss": 0.0451, + "step": 25090 + }, + { + "grad_norm": 0.18583405017852783, + "learning_rate": 5.38026122636343e-05, + "loss": 0.0413, + "step": 25100 + }, + { + "grad_norm": 0.2210431843996048, + "learning_rate": 5.3769637812446636e-05, + "loss": 0.0398, + "step": 25110 + }, + { + "grad_norm": 0.26007455587387085, + "learning_rate": 5.373666171229133e-05, + "loss": 0.0398, + "step": 25120 + }, + { + "grad_norm": 0.16124461591243744, + "learning_rate": 5.370368397759324e-05, + "loss": 0.0542, + "step": 25130 + }, + { + "grad_norm": 0.18233254551887512, + "learning_rate": 5.3670704622777944e-05, + "loss": 0.0363, + "step": 25140 + }, + { + "grad_norm": 0.14618948101997375, + "learning_rate": 5.3637723662271744e-05, + "loss": 0.0385, + "step": 25150 + }, + { + "grad_norm": 0.25506430864334106, + "learning_rate": 5.360474111050162e-05, + "loss": 0.037, + "step": 25160 + }, + { + "grad_norm": 0.21190592646598816, + "learning_rate": 5.357175698189527e-05, + "loss": 0.0327, + "step": 25170 + }, + { + "grad_norm": 0.30616626143455505, + "learning_rate": 5.353877129088107e-05, + "loss": 0.0426, + "step": 25180 + }, + { + "grad_norm": 0.1919948160648346, + "learning_rate": 5.350578405188809e-05, + "loss": 0.0434, + "step": 25190 + }, + { + "grad_norm": 0.14712542295455933, + "learning_rate": 5.347279527934603e-05, + "loss": 0.0374, + "step": 25200 + }, + { + "grad_norm": 0.22413775324821472, + "learning_rate": 5.343980498768536e-05, + "loss": 0.0417, + "step": 25210 + }, + { + "grad_norm": 0.3901083469390869, + "learning_rate": 5.340681319133711e-05, + "loss": 0.036, + "step": 25220 + }, + { + "grad_norm": 0.2003306746482849, + "learning_rate": 5.337381990473305e-05, + "loss": 0.0449, + "step": 25230 + }, + { + "grad_norm": 0.27891799807548523, + "learning_rate": 5.3340825142305504e-05, + "loss": 0.0357, + "step": 25240 + }, + { + "grad_norm": 0.4073387682437897, + "learning_rate": 5.330782891848756e-05, + "loss": 0.0369, + "step": 25250 + }, + { + "grad_norm": 0.19623728096485138, + "learning_rate": 5.327483124771285e-05, + "loss": 0.0348, + "step": 25260 + }, + { + "grad_norm": 0.17416293919086456, + "learning_rate": 5.3241832144415706e-05, + "loss": 0.0399, + "step": 25270 + }, + { + "grad_norm": 0.411786288022995, + "learning_rate": 5.320883162303104e-05, + "loss": 0.0397, + "step": 25280 + }, + { + "grad_norm": 0.20922939479351044, + "learning_rate": 5.3175829697994416e-05, + "loss": 0.0431, + "step": 25290 + }, + { + "grad_norm": 0.18980693817138672, + "learning_rate": 5.3142826383741975e-05, + "loss": 0.0412, + "step": 25300 + }, + { + "grad_norm": 0.2563205063343048, + "learning_rate": 5.3109821694710506e-05, + "loss": 0.0452, + "step": 25310 + }, + { + "grad_norm": 0.17645376920700073, + "learning_rate": 5.307681564533736e-05, + "loss": 0.0376, + "step": 25320 + }, + { + "grad_norm": 0.2580287754535675, + "learning_rate": 5.3043808250060535e-05, + "loss": 0.0514, + "step": 25330 + }, + { + "grad_norm": 0.21742255985736847, + "learning_rate": 5.3010799523318566e-05, + "loss": 0.0415, + "step": 25340 + }, + { + "grad_norm": 0.33646151423454285, + "learning_rate": 5.29777894795506e-05, + "loss": 0.0416, + "step": 25350 + }, + { + "grad_norm": 0.2972324788570404, + "learning_rate": 5.294477813319634e-05, + "loss": 0.037, + "step": 25360 + }, + { + "grad_norm": 0.28459399938583374, + "learning_rate": 5.291176549869608e-05, + "loss": 0.0414, + "step": 25370 + }, + { + "grad_norm": 0.3171035349369049, + "learning_rate": 5.287875159049065e-05, + "loss": 0.0407, + "step": 25380 + }, + { + "grad_norm": 0.2789437174797058, + "learning_rate": 5.284573642302149e-05, + "loss": 0.0398, + "step": 25390 + }, + { + "grad_norm": 0.2359062135219574, + "learning_rate": 5.2812720010730543e-05, + "loss": 0.0359, + "step": 25400 + }, + { + "grad_norm": 0.29596444964408875, + "learning_rate": 5.277970236806029e-05, + "loss": 0.0442, + "step": 25410 + }, + { + "grad_norm": 0.21037913858890533, + "learning_rate": 5.2746683509453795e-05, + "loss": 0.0384, + "step": 25420 + }, + { + "grad_norm": 0.3899734318256378, + "learning_rate": 5.271366344935461e-05, + "loss": 0.0482, + "step": 25430 + }, + { + "grad_norm": 0.21373499929904938, + "learning_rate": 5.268064220220683e-05, + "loss": 0.0387, + "step": 25440 + }, + { + "grad_norm": 0.22348617017269135, + "learning_rate": 5.2647619782455094e-05, + "loss": 0.0426, + "step": 25450 + }, + { + "grad_norm": 0.3784096837043762, + "learning_rate": 5.261459620454451e-05, + "loss": 0.0417, + "step": 25460 + }, + { + "grad_norm": 0.27401065826416016, + "learning_rate": 5.258157148292071e-05, + "loss": 0.0471, + "step": 25470 + }, + { + "grad_norm": 0.1970139592885971, + "learning_rate": 5.2548545632029846e-05, + "loss": 0.037, + "step": 25480 + }, + { + "grad_norm": 0.28404489159584045, + "learning_rate": 5.2515518666318534e-05, + "loss": 0.0392, + "step": 25490 + }, + { + "grad_norm": 0.22748690843582153, + "learning_rate": 5.248249060023389e-05, + "loss": 0.0386, + "step": 25500 + }, + { + "grad_norm": 0.32837170362472534, + "learning_rate": 5.2449461448223517e-05, + "loss": 0.0351, + "step": 25510 + }, + { + "grad_norm": 0.23461532592773438, + "learning_rate": 5.2416431224735486e-05, + "loss": 0.0378, + "step": 25520 + }, + { + "grad_norm": 0.19430959224700928, + "learning_rate": 5.238339994421836e-05, + "loss": 0.0449, + "step": 25530 + }, + { + "grad_norm": 0.2498452365398407, + "learning_rate": 5.235036762112111e-05, + "loss": 0.0429, + "step": 25540 + }, + { + "grad_norm": 0.34006017446517944, + "learning_rate": 5.2317334269893206e-05, + "loss": 0.0398, + "step": 25550 + }, + { + "grad_norm": 0.23740608990192413, + "learning_rate": 5.228429990498458e-05, + "loss": 0.0395, + "step": 25560 + }, + { + "grad_norm": 0.37409427762031555, + "learning_rate": 5.225126454084556e-05, + "loss": 0.0409, + "step": 25570 + }, + { + "grad_norm": 0.16856342554092407, + "learning_rate": 5.221822819192694e-05, + "loss": 0.0438, + "step": 25580 + }, + { + "grad_norm": 0.19827355444431305, + "learning_rate": 5.218519087267995e-05, + "loss": 0.0414, + "step": 25590 + }, + { + "grad_norm": 0.24741782248020172, + "learning_rate": 5.2152152597556225e-05, + "loss": 0.0373, + "step": 25600 + }, + { + "grad_norm": 0.21673403680324554, + "learning_rate": 5.211911338100784e-05, + "loss": 0.0447, + "step": 25610 + }, + { + "grad_norm": 0.23472148180007935, + "learning_rate": 5.2086073237487264e-05, + "loss": 0.0459, + "step": 25620 + }, + { + "grad_norm": 1.0803660154342651, + "learning_rate": 5.205303218144739e-05, + "loss": 0.0439, + "step": 25630 + }, + { + "grad_norm": 0.1977989673614502, + "learning_rate": 5.2019990227341494e-05, + "loss": 0.0355, + "step": 25640 + }, + { + "grad_norm": 0.37129589915275574, + "learning_rate": 5.198694738962324e-05, + "loss": 0.046, + "step": 25650 + }, + { + "grad_norm": 0.2951109707355499, + "learning_rate": 5.19539036827467e-05, + "loss": 0.0315, + "step": 25660 + }, + { + "grad_norm": 0.2563914358615875, + "learning_rate": 5.192085912116629e-05, + "loss": 0.0477, + "step": 25670 + }, + { + "grad_norm": 0.33899837732315063, + "learning_rate": 5.188781371933685e-05, + "loss": 0.0384, + "step": 25680 + }, + { + "grad_norm": 0.22883956134319305, + "learning_rate": 5.185476749171356e-05, + "loss": 0.0392, + "step": 25690 + }, + { + "grad_norm": 0.2644627094268799, + "learning_rate": 5.1821720452751945e-05, + "loss": 0.0385, + "step": 25700 + }, + { + "grad_norm": 0.27428728342056274, + "learning_rate": 5.178867261690791e-05, + "loss": 0.0366, + "step": 25710 + }, + { + "grad_norm": 0.2667413353919983, + "learning_rate": 5.1755623998637715e-05, + "loss": 0.0386, + "step": 25720 + }, + { + "grad_norm": 0.48529252409935, + "learning_rate": 5.172257461239794e-05, + "loss": 0.0427, + "step": 25730 + }, + { + "grad_norm": 0.2786993682384491, + "learning_rate": 5.168952447264548e-05, + "loss": 0.0373, + "step": 25740 + }, + { + "grad_norm": 0.21594011783599854, + "learning_rate": 5.165647359383763e-05, + "loss": 0.0412, + "step": 25750 + }, + { + "grad_norm": 0.24051646888256073, + "learning_rate": 5.1623421990431954e-05, + "loss": 0.0402, + "step": 25760 + }, + { + "grad_norm": 0.15194164216518402, + "learning_rate": 5.159036967688635e-05, + "loss": 0.0392, + "step": 25770 + }, + { + "grad_norm": 0.1582792103290558, + "learning_rate": 5.1557316667659015e-05, + "loss": 0.0342, + "step": 25780 + }, + { + "grad_norm": 0.18386130034923553, + "learning_rate": 5.152426297720846e-05, + "loss": 0.0407, + "step": 25790 + }, + { + "grad_norm": 0.18481838703155518, + "learning_rate": 5.1491208619993505e-05, + "loss": 0.0383, + "step": 25800 + }, + { + "grad_norm": 0.26829683780670166, + "learning_rate": 5.145815361047321e-05, + "loss": 0.0372, + "step": 25810 + }, + { + "grad_norm": 0.21088580787181854, + "learning_rate": 5.1425097963107003e-05, + "loss": 0.0336, + "step": 25820 + }, + { + "grad_norm": 0.18597400188446045, + "learning_rate": 5.1392041692354544e-05, + "loss": 0.0394, + "step": 25830 + }, + { + "grad_norm": 0.20213186740875244, + "learning_rate": 5.1358984812675736e-05, + "loss": 0.0445, + "step": 25840 + }, + { + "grad_norm": 0.1789536029100418, + "learning_rate": 5.13259273385308e-05, + "loss": 0.0393, + "step": 25850 + }, + { + "grad_norm": 0.1526263803243637, + "learning_rate": 5.12928692843802e-05, + "loss": 0.0352, + "step": 25860 + }, + { + "grad_norm": 0.21324574947357178, + "learning_rate": 5.125981066468465e-05, + "loss": 0.0406, + "step": 25870 + }, + { + "grad_norm": 0.196905717253685, + "learning_rate": 5.122675149390514e-05, + "loss": 0.0483, + "step": 25880 + }, + { + "grad_norm": 0.16705191135406494, + "learning_rate": 5.119369178650282e-05, + "loss": 0.0384, + "step": 25890 + }, + { + "grad_norm": 0.2670370638370514, + "learning_rate": 5.1160631556939156e-05, + "loss": 0.0363, + "step": 25900 + }, + { + "grad_norm": 0.1765984743833542, + "learning_rate": 5.112757081967584e-05, + "loss": 0.0389, + "step": 25910 + }, + { + "grad_norm": 0.3973836302757263, + "learning_rate": 5.109450958917472e-05, + "loss": 0.0392, + "step": 25920 + }, + { + "grad_norm": 0.17813219130039215, + "learning_rate": 5.106144787989794e-05, + "loss": 0.0392, + "step": 25930 + }, + { + "grad_norm": 0.29686787724494934, + "learning_rate": 5.102838570630778e-05, + "loss": 0.0355, + "step": 25940 + }, + { + "grad_norm": 0.1494239717721939, + "learning_rate": 5.099532308286678e-05, + "loss": 0.0388, + "step": 25950 + }, + { + "grad_norm": 0.14587479829788208, + "learning_rate": 5.0962260024037645e-05, + "loss": 0.0399, + "step": 25960 + }, + { + "grad_norm": 0.1860077828168869, + "learning_rate": 5.0929196544283276e-05, + "loss": 0.0393, + "step": 25970 + }, + { + "grad_norm": 0.26564839482307434, + "learning_rate": 5.089613265806678e-05, + "loss": 0.0366, + "step": 25980 + }, + { + "grad_norm": 0.16036881506443024, + "learning_rate": 5.086306837985141e-05, + "loss": 0.036, + "step": 25990 + }, + { + "grad_norm": 0.18831287324428558, + "learning_rate": 5.0830003724100604e-05, + "loss": 0.0408, + "step": 26000 + }, + { + "grad_norm": 0.2690129280090332, + "learning_rate": 5.0796938705277984e-05, + "loss": 0.0418, + "step": 26010 + }, + { + "grad_norm": 0.27296677231788635, + "learning_rate": 5.076387333784729e-05, + "loss": 0.0374, + "step": 26020 + }, + { + "grad_norm": 0.2256561815738678, + "learning_rate": 5.0730807636272445e-05, + "loss": 0.0382, + "step": 26030 + }, + { + "grad_norm": 0.2510363757610321, + "learning_rate": 5.0697741615017524e-05, + "loss": 0.0377, + "step": 26040 + }, + { + "grad_norm": 0.17991462349891663, + "learning_rate": 5.06646752885467e-05, + "loss": 0.0396, + "step": 26050 + }, + { + "grad_norm": 0.24477890133857727, + "learning_rate": 5.0631608671324324e-05, + "loss": 0.0352, + "step": 26060 + }, + { + "grad_norm": 0.2214019000530243, + "learning_rate": 5.059854177781487e-05, + "loss": 0.0418, + "step": 26070 + }, + { + "grad_norm": 0.43681687116622925, + "learning_rate": 5.05654746224829e-05, + "loss": 0.044, + "step": 26080 + }, + { + "grad_norm": 0.23855137825012207, + "learning_rate": 5.053240721979311e-05, + "loss": 0.0451, + "step": 26090 + }, + { + "grad_norm": 0.3033919036388397, + "learning_rate": 5.049933958421031e-05, + "loss": 0.0424, + "step": 26100 + }, + { + "grad_norm": 0.14140504598617554, + "learning_rate": 5.046627173019943e-05, + "loss": 0.0342, + "step": 26110 + }, + { + "grad_norm": 0.25879359245300293, + "learning_rate": 5.043320367222545e-05, + "loss": 0.0366, + "step": 26120 + }, + { + "grad_norm": 0.2117154896259308, + "learning_rate": 5.040013542475346e-05, + "loss": 0.0397, + "step": 26130 + }, + { + "grad_norm": 0.3721894919872284, + "learning_rate": 5.0367067002248646e-05, + "loss": 0.0405, + "step": 26140 + }, + { + "grad_norm": 0.2217881977558136, + "learning_rate": 5.0333998419176254e-05, + "loss": 0.0404, + "step": 26150 + }, + { + "grad_norm": 0.19205327332019806, + "learning_rate": 5.030092969000161e-05, + "loss": 0.0363, + "step": 26160 + }, + { + "grad_norm": 0.30992475152015686, + "learning_rate": 5.026786082919008e-05, + "loss": 0.0379, + "step": 26170 + }, + { + "grad_norm": 0.24599654972553253, + "learning_rate": 5.0234791851207144e-05, + "loss": 0.0397, + "step": 26180 + }, + { + "grad_norm": 0.264251708984375, + "learning_rate": 5.0201722770518256e-05, + "loss": 0.0404, + "step": 26190 + }, + { + "grad_norm": 0.36400049924850464, + "learning_rate": 5.0168653601589e-05, + "loss": 0.0399, + "step": 26200 + }, + { + "grad_norm": 0.1486753523349762, + "learning_rate": 5.01355843588849e-05, + "loss": 0.0435, + "step": 26210 + }, + { + "grad_norm": 0.26320648193359375, + "learning_rate": 5.010251505687161e-05, + "loss": 0.0394, + "step": 26220 + }, + { + "grad_norm": 0.16862131655216217, + "learning_rate": 5.006944571001474e-05, + "loss": 0.0428, + "step": 26230 + }, + { + "grad_norm": 0.219700425863266, + "learning_rate": 5.003637633277997e-05, + "loss": 0.0368, + "step": 26240 + }, + { + "grad_norm": 0.235765278339386, + "learning_rate": 5.0003306939632946e-05, + "loss": 0.04, + "step": 26250 + }, + { + "grad_norm": 0.21412985026836395, + "learning_rate": 4.997023754503937e-05, + "loss": 0.0451, + "step": 26260 + }, + { + "grad_norm": 0.16325482726097107, + "learning_rate": 4.9937168163464897e-05, + "loss": 0.0381, + "step": 26270 + }, + { + "grad_norm": 0.30272242426872253, + "learning_rate": 4.990409880937519e-05, + "loss": 0.0471, + "step": 26280 + }, + { + "grad_norm": 0.17329923808574677, + "learning_rate": 4.9871029497235936e-05, + "loss": 0.0522, + "step": 26290 + }, + { + "grad_norm": 0.3844163417816162, + "learning_rate": 4.9837960241512766e-05, + "loss": 0.0441, + "step": 26300 + }, + { + "grad_norm": 0.22124320268630981, + "learning_rate": 4.9804891056671265e-05, + "loss": 0.0528, + "step": 26310 + }, + { + "grad_norm": 0.1721300184726715, + "learning_rate": 4.9771821957177074e-05, + "loss": 0.0417, + "step": 26320 + }, + { + "grad_norm": 0.21750026941299438, + "learning_rate": 4.9738752957495694e-05, + "loss": 0.0397, + "step": 26330 + }, + { + "grad_norm": 0.2886511981487274, + "learning_rate": 4.970568407209265e-05, + "loss": 0.0409, + "step": 26340 + }, + { + "grad_norm": 0.22979038953781128, + "learning_rate": 4.967261531543339e-05, + "loss": 0.0411, + "step": 26350 + }, + { + "grad_norm": 0.199431374669075, + "learning_rate": 4.963954670198332e-05, + "loss": 0.0434, + "step": 26360 + }, + { + "grad_norm": 0.2378472238779068, + "learning_rate": 4.960647824620776e-05, + "loss": 0.0416, + "step": 26370 + }, + { + "grad_norm": 0.18406768143177032, + "learning_rate": 4.9573409962571995e-05, + "loss": 0.0375, + "step": 26380 + }, + { + "grad_norm": 0.43017899990081787, + "learning_rate": 4.954034186554121e-05, + "loss": 0.0466, + "step": 26390 + }, + { + "grad_norm": 0.27334004640579224, + "learning_rate": 4.950727396958051e-05, + "loss": 0.0437, + "step": 26400 + }, + { + "grad_norm": 0.2553160786628723, + "learning_rate": 4.947420628915489e-05, + "loss": 0.0401, + "step": 26410 + }, + { + "grad_norm": 0.21064309775829315, + "learning_rate": 4.944113883872932e-05, + "loss": 0.0383, + "step": 26420 + }, + { + "grad_norm": 0.17449435591697693, + "learning_rate": 4.9408071632768624e-05, + "loss": 0.0474, + "step": 26430 + }, + { + "grad_norm": 0.2873331904411316, + "learning_rate": 4.937500468573747e-05, + "loss": 0.0486, + "step": 26440 + }, + { + "grad_norm": 0.1807928830385208, + "learning_rate": 4.934193801210053e-05, + "loss": 0.0407, + "step": 26450 + }, + { + "grad_norm": 0.30449622869491577, + "learning_rate": 4.930887162632224e-05, + "loss": 0.0413, + "step": 26460 + }, + { + "grad_norm": 0.2570914924144745, + "learning_rate": 4.9275805542866994e-05, + "loss": 0.0386, + "step": 26470 + }, + { + "grad_norm": 0.26955723762512207, + "learning_rate": 4.9242739776199006e-05, + "loss": 0.0395, + "step": 26480 + }, + { + "grad_norm": 0.1530124396085739, + "learning_rate": 4.920967434078237e-05, + "loss": 0.0462, + "step": 26490 + }, + { + "grad_norm": 0.2146182805299759, + "learning_rate": 4.917660925108101e-05, + "loss": 0.0469, + "step": 26500 + }, + { + "grad_norm": 0.22419171035289764, + "learning_rate": 4.914354452155876e-05, + "loss": 0.0452, + "step": 26510 + }, + { + "grad_norm": 0.2440042793750763, + "learning_rate": 4.911048016667923e-05, + "loss": 0.0431, + "step": 26520 + }, + { + "grad_norm": 0.25216683745384216, + "learning_rate": 4.907741620090588e-05, + "loss": 0.0451, + "step": 26530 + }, + { + "grad_norm": 0.21394920349121094, + "learning_rate": 4.9044352638702046e-05, + "loss": 0.0434, + "step": 26540 + }, + { + "grad_norm": 0.2091953009366989, + "learning_rate": 4.9011289494530834e-05, + "loss": 0.0384, + "step": 26550 + }, + { + "grad_norm": 0.19359642267227173, + "learning_rate": 4.897822678285516e-05, + "loss": 0.0422, + "step": 26560 + }, + { + "grad_norm": 0.22336170077323914, + "learning_rate": 4.8945164518137835e-05, + "loss": 0.0412, + "step": 26570 + }, + { + "grad_norm": 0.20505818724632263, + "learning_rate": 4.891210271484138e-05, + "loss": 0.0342, + "step": 26580 + }, + { + "grad_norm": 0.19231770932674408, + "learning_rate": 4.8879041387428164e-05, + "loss": 0.0413, + "step": 26590 + }, + { + "grad_norm": 0.26102927327156067, + "learning_rate": 4.8845980550360306e-05, + "loss": 0.0429, + "step": 26600 + }, + { + "grad_norm": 0.22103744745254517, + "learning_rate": 4.8812920218099786e-05, + "loss": 0.0412, + "step": 26610 + }, + { + "grad_norm": 0.21639350056648254, + "learning_rate": 4.8779860405108294e-05, + "loss": 0.0438, + "step": 26620 + }, + { + "grad_norm": 0.2193070948123932, + "learning_rate": 4.8746801125847285e-05, + "loss": 0.0418, + "step": 26630 + }, + { + "grad_norm": 0.2753795385360718, + "learning_rate": 4.8713742394778064e-05, + "loss": 0.0386, + "step": 26640 + }, + { + "grad_norm": 0.18732264637947083, + "learning_rate": 4.868068422636162e-05, + "loss": 0.0346, + "step": 26650 + }, + { + "grad_norm": 0.24149300158023834, + "learning_rate": 4.86476266350587e-05, + "loss": 0.0496, + "step": 26660 + }, + { + "grad_norm": 0.17312058806419373, + "learning_rate": 4.861456963532985e-05, + "loss": 0.0415, + "step": 26670 + }, + { + "grad_norm": 0.2391873151063919, + "learning_rate": 4.8581513241635316e-05, + "loss": 0.0436, + "step": 26680 + }, + { + "grad_norm": 0.2287628948688507, + "learning_rate": 4.8548457468435065e-05, + "loss": 0.0391, + "step": 26690 + }, + { + "grad_norm": 0.2765870690345764, + "learning_rate": 4.851540233018884e-05, + "loss": 0.0535, + "step": 26700 + }, + { + "grad_norm": 0.27770504355430603, + "learning_rate": 4.848234784135608e-05, + "loss": 0.0443, + "step": 26710 + }, + { + "grad_norm": 0.2505369782447815, + "learning_rate": 4.8449294016395916e-05, + "loss": 0.0339, + "step": 26720 + }, + { + "grad_norm": 0.16844360530376434, + "learning_rate": 4.8416240869767246e-05, + "loss": 0.0357, + "step": 26730 + }, + { + "grad_norm": 0.14746582508087158, + "learning_rate": 4.838318841592864e-05, + "loss": 0.0398, + "step": 26740 + }, + { + "grad_norm": 0.3010350465774536, + "learning_rate": 4.835013666933834e-05, + "loss": 0.0487, + "step": 26750 + }, + { + "grad_norm": 0.3764389753341675, + "learning_rate": 4.83170856444543e-05, + "loss": 0.0375, + "step": 26760 + }, + { + "grad_norm": 0.23232802748680115, + "learning_rate": 4.828403535573419e-05, + "loss": 0.0339, + "step": 26770 + }, + { + "grad_norm": 0.23862674832344055, + "learning_rate": 4.825098581763532e-05, + "loss": 0.0404, + "step": 26780 + }, + { + "grad_norm": 0.2287665754556656, + "learning_rate": 4.821793704461465e-05, + "loss": 0.0409, + "step": 26790 + }, + { + "grad_norm": 0.19083739817142487, + "learning_rate": 4.8184889051128884e-05, + "loss": 0.0377, + "step": 26800 + }, + { + "grad_norm": 0.7621665000915527, + "learning_rate": 4.81518418516343e-05, + "loss": 0.0361, + "step": 26810 + }, + { + "grad_norm": 0.34678131341934204, + "learning_rate": 4.811879546058688e-05, + "loss": 0.0359, + "step": 26820 + }, + { + "grad_norm": 0.12013948708772659, + "learning_rate": 4.808574989244224e-05, + "loss": 0.0339, + "step": 26830 + }, + { + "grad_norm": 0.2796230912208557, + "learning_rate": 4.8052705161655644e-05, + "loss": 0.0439, + "step": 26840 + }, + { + "grad_norm": 0.1998787820339203, + "learning_rate": 4.8019661282681936e-05, + "loss": 0.0356, + "step": 26850 + }, + { + "grad_norm": 0.13542300462722778, + "learning_rate": 4.79866182699757e-05, + "loss": 0.0345, + "step": 26860 + }, + { + "grad_norm": 0.17278547585010529, + "learning_rate": 4.795357613799103e-05, + "loss": 0.0349, + "step": 26870 + }, + { + "grad_norm": 0.21349704265594482, + "learning_rate": 4.792053490118166e-05, + "loss": 0.0381, + "step": 26880 + }, + { + "grad_norm": 0.1444307565689087, + "learning_rate": 4.788749457400099e-05, + "loss": 0.0366, + "step": 26890 + }, + { + "grad_norm": 0.26682382822036743, + "learning_rate": 4.7854455170901966e-05, + "loss": 0.0373, + "step": 26900 + }, + { + "grad_norm": 0.1910473257303238, + "learning_rate": 4.7821416706337136e-05, + "loss": 0.0369, + "step": 26910 + }, + { + "grad_norm": 0.23536857962608337, + "learning_rate": 4.7788379194758645e-05, + "loss": 0.0425, + "step": 26920 + }, + { + "grad_norm": 0.241017684340477, + "learning_rate": 4.7755342650618246e-05, + "loss": 0.0397, + "step": 26930 + }, + { + "grad_norm": 0.24927900731563568, + "learning_rate": 4.7722307088367235e-05, + "loss": 0.0497, + "step": 26940 + }, + { + "grad_norm": 0.5537494421005249, + "learning_rate": 4.768927252245646e-05, + "loss": 0.0395, + "step": 26950 + }, + { + "grad_norm": 0.5603659749031067, + "learning_rate": 4.765623896733642e-05, + "loss": 0.036, + "step": 26960 + }, + { + "grad_norm": 0.26369595527648926, + "learning_rate": 4.762320643745708e-05, + "loss": 0.0396, + "step": 26970 + }, + { + "grad_norm": 0.2552984654903412, + "learning_rate": 4.7590174947267966e-05, + "loss": 0.0314, + "step": 26980 + }, + { + "grad_norm": 0.2591046094894409, + "learning_rate": 4.755714451121823e-05, + "loss": 0.042, + "step": 26990 + }, + { + "grad_norm": 0.2559952735900879, + "learning_rate": 4.7524115143756476e-05, + "loss": 0.0353, + "step": 27000 + }, + { + "grad_norm": 0.2619532346725464, + "learning_rate": 4.7491086859330854e-05, + "loss": 0.0332, + "step": 27010 + }, + { + "grad_norm": 0.2252744734287262, + "learning_rate": 4.74580596723891e-05, + "loss": 0.0349, + "step": 27020 + }, + { + "grad_norm": 0.14117932319641113, + "learning_rate": 4.742503359737841e-05, + "loss": 0.0335, + "step": 27030 + }, + { + "grad_norm": 0.12386465817689896, + "learning_rate": 4.739200864874547e-05, + "loss": 0.0443, + "step": 27040 + }, + { + "grad_norm": 0.23630161583423615, + "learning_rate": 4.735898484093658e-05, + "loss": 0.0351, + "step": 27050 + }, + { + "grad_norm": 0.1775914877653122, + "learning_rate": 4.7325962188397444e-05, + "loss": 0.0352, + "step": 27060 + }, + { + "grad_norm": 0.2195945531129837, + "learning_rate": 4.729294070557329e-05, + "loss": 0.0344, + "step": 27070 + }, + { + "grad_norm": 0.2160033881664276, + "learning_rate": 4.7259920406908833e-05, + "loss": 0.0376, + "step": 27080 + }, + { + "grad_norm": 0.2969173192977905, + "learning_rate": 4.7226901306848306e-05, + "loss": 0.0358, + "step": 27090 + }, + { + "grad_norm": 0.283121794462204, + "learning_rate": 4.719388341983535e-05, + "loss": 0.0334, + "step": 27100 + }, + { + "grad_norm": 0.2640632688999176, + "learning_rate": 4.71608667603131e-05, + "loss": 0.0339, + "step": 27110 + }, + { + "grad_norm": 0.28586897253990173, + "learning_rate": 4.712785134272422e-05, + "loss": 0.0384, + "step": 27120 + }, + { + "grad_norm": 0.31412801146507263, + "learning_rate": 4.7094837181510734e-05, + "loss": 0.0323, + "step": 27130 + }, + { + "grad_norm": 0.24479122459888458, + "learning_rate": 4.706182429111414e-05, + "loss": 0.0441, + "step": 27140 + }, + { + "grad_norm": 0.20378798246383667, + "learning_rate": 4.702881268597545e-05, + "loss": 0.0383, + "step": 27150 + }, + { + "grad_norm": 0.2549411356449127, + "learning_rate": 4.699580238053503e-05, + "loss": 0.0354, + "step": 27160 + }, + { + "grad_norm": 0.25235897302627563, + "learning_rate": 4.69627933892327e-05, + "loss": 0.0409, + "step": 27170 + }, + { + "grad_norm": 0.26611122488975525, + "learning_rate": 4.692978572650773e-05, + "loss": 0.043, + "step": 27180 + }, + { + "grad_norm": 0.1793341487646103, + "learning_rate": 4.689677940679881e-05, + "loss": 0.0416, + "step": 27190 + }, + { + "grad_norm": 0.18809224665164948, + "learning_rate": 4.6863774444543984e-05, + "loss": 0.0481, + "step": 27200 + }, + { + "grad_norm": 0.2146705985069275, + "learning_rate": 4.6830770854180784e-05, + "loss": 0.0378, + "step": 27210 + }, + { + "grad_norm": 0.2166173756122589, + "learning_rate": 4.679776865014608e-05, + "loss": 0.0371, + "step": 27220 + }, + { + "grad_norm": 0.16056673228740692, + "learning_rate": 4.676476784687615e-05, + "loss": 0.0351, + "step": 27230 + }, + { + "grad_norm": 0.19471785426139832, + "learning_rate": 4.673176845880669e-05, + "loss": 0.0362, + "step": 27240 + }, + { + "grad_norm": 0.21388252079486847, + "learning_rate": 4.669877050037275e-05, + "loss": 0.0412, + "step": 27250 + }, + { + "grad_norm": 0.2803722023963928, + "learning_rate": 4.666577398600874e-05, + "loss": 0.0351, + "step": 27260 + }, + { + "grad_norm": 0.26729774475097656, + "learning_rate": 4.663277893014846e-05, + "loss": 0.0316, + "step": 27270 + }, + { + "grad_norm": 0.1956341713666916, + "learning_rate": 4.659978534722508e-05, + "loss": 0.039, + "step": 27280 + }, + { + "grad_norm": 0.17879733443260193, + "learning_rate": 4.6566793251671104e-05, + "loss": 0.0357, + "step": 27290 + }, + { + "grad_norm": 0.1992887705564499, + "learning_rate": 4.65338026579184e-05, + "loss": 0.0353, + "step": 27300 + }, + { + "grad_norm": 0.13648714125156403, + "learning_rate": 4.650081358039817e-05, + "loss": 0.0428, + "step": 27310 + }, + { + "grad_norm": 0.25659477710723877, + "learning_rate": 4.646782603354098e-05, + "loss": 0.0374, + "step": 27320 + }, + { + "grad_norm": 0.27785012125968933, + "learning_rate": 4.643484003177666e-05, + "loss": 0.0329, + "step": 27330 + }, + { + "grad_norm": 0.19397379457950592, + "learning_rate": 4.6401855589534454e-05, + "loss": 0.0382, + "step": 27340 + }, + { + "grad_norm": 0.21030788123607635, + "learning_rate": 4.636887272124286e-05, + "loss": 0.0443, + "step": 27350 + }, + { + "grad_norm": 0.23639163374900818, + "learning_rate": 4.6335891441329685e-05, + "loss": 0.0468, + "step": 27360 + }, + { + "grad_norm": 0.16829243302345276, + "learning_rate": 4.630291176422209e-05, + "loss": 0.0383, + "step": 27370 + }, + { + "grad_norm": 0.1876152604818344, + "learning_rate": 4.62699337043465e-05, + "loss": 0.0394, + "step": 27380 + }, + { + "grad_norm": 0.2670457661151886, + "learning_rate": 4.623695727612861e-05, + "loss": 0.0414, + "step": 27390 + }, + { + "grad_norm": 0.1527857780456543, + "learning_rate": 4.620398249399347e-05, + "loss": 0.0363, + "step": 27400 + }, + { + "grad_norm": 0.13354989886283875, + "learning_rate": 4.617100937236535e-05, + "loss": 0.0444, + "step": 27410 + }, + { + "grad_norm": 0.9408333897590637, + "learning_rate": 4.613803792566781e-05, + "loss": 0.0407, + "step": 27420 + }, + { + "grad_norm": 0.2162020206451416, + "learning_rate": 4.61050681683237e-05, + "loss": 0.0401, + "step": 27430 + }, + { + "grad_norm": 0.2288733571767807, + "learning_rate": 4.6072100114755104e-05, + "loss": 0.0467, + "step": 27440 + }, + { + "grad_norm": 0.14125864207744598, + "learning_rate": 4.603913377938336e-05, + "loss": 0.0393, + "step": 27450 + }, + { + "grad_norm": 0.20029211044311523, + "learning_rate": 4.600616917662905e-05, + "loss": 0.0391, + "step": 27460 + }, + { + "grad_norm": 0.230002760887146, + "learning_rate": 4.597320632091206e-05, + "loss": 0.0457, + "step": 27470 + }, + { + "grad_norm": 0.28552472591400146, + "learning_rate": 4.594024522665142e-05, + "loss": 0.042, + "step": 27480 + }, + { + "grad_norm": 0.3058757483959198, + "learning_rate": 4.590728590826543e-05, + "loss": 0.0385, + "step": 27490 + }, + { + "grad_norm": 0.28427189588546753, + "learning_rate": 4.587432838017165e-05, + "loss": 0.0481, + "step": 27500 + }, + { + "grad_norm": 0.2714974582195282, + "learning_rate": 4.58413726567868e-05, + "loss": 0.0471, + "step": 27510 + }, + { + "grad_norm": 0.2471446692943573, + "learning_rate": 4.5808418752526814e-05, + "loss": 0.0466, + "step": 27520 + }, + { + "grad_norm": 0.2502569556236267, + "learning_rate": 4.5775466681806895e-05, + "loss": 0.0451, + "step": 27530 + }, + { + "grad_norm": 0.17959007620811462, + "learning_rate": 4.574251645904136e-05, + "loss": 0.0427, + "step": 27540 + }, + { + "grad_norm": 0.22217662632465363, + "learning_rate": 4.5709568098643776e-05, + "loss": 0.0444, + "step": 27550 + }, + { + "grad_norm": 0.23914866149425507, + "learning_rate": 4.567662161502687e-05, + "loss": 0.0338, + "step": 27560 + }, + { + "grad_norm": 0.211800217628479, + "learning_rate": 4.5643677022602554e-05, + "loss": 0.0405, + "step": 27570 + }, + { + "grad_norm": 0.24929095804691315, + "learning_rate": 4.56107343357819e-05, + "loss": 0.0378, + "step": 27580 + }, + { + "grad_norm": 0.18457183241844177, + "learning_rate": 4.557779356897518e-05, + "loss": 0.0362, + "step": 27590 + }, + { + "grad_norm": 0.334768682718277, + "learning_rate": 4.55448547365918e-05, + "loss": 0.0417, + "step": 27600 + }, + { + "grad_norm": 0.2521315813064575, + "learning_rate": 4.5511917853040315e-05, + "loss": 0.0461, + "step": 27610 + }, + { + "grad_norm": 0.20163139700889587, + "learning_rate": 4.547898293272842e-05, + "loss": 0.0399, + "step": 27620 + }, + { + "grad_norm": 0.24085839092731476, + "learning_rate": 4.544604999006301e-05, + "loss": 0.0386, + "step": 27630 + }, + { + "grad_norm": 0.4989137351512909, + "learning_rate": 4.541311903945004e-05, + "loss": 0.0413, + "step": 27640 + }, + { + "grad_norm": 0.19424307346343994, + "learning_rate": 4.538019009529463e-05, + "loss": 0.0435, + "step": 27650 + }, + { + "grad_norm": 0.20670220255851746, + "learning_rate": 4.5347263172001035e-05, + "loss": 0.0467, + "step": 27660 + }, + { + "grad_norm": 0.23022867739200592, + "learning_rate": 4.531433828397261e-05, + "loss": 0.0444, + "step": 27670 + }, + { + "grad_norm": 0.2970667779445648, + "learning_rate": 4.528141544561179e-05, + "loss": 0.0432, + "step": 27680 + }, + { + "grad_norm": 0.17336441576480865, + "learning_rate": 4.5248494671320186e-05, + "loss": 0.0364, + "step": 27690 + }, + { + "grad_norm": 0.3028905391693115, + "learning_rate": 4.521557597549844e-05, + "loss": 0.043, + "step": 27700 + }, + { + "grad_norm": 0.1581028699874878, + "learning_rate": 4.5182659372546284e-05, + "loss": 0.0379, + "step": 27710 + }, + { + "grad_norm": 0.17496375739574432, + "learning_rate": 4.514974487686261e-05, + "loss": 0.0372, + "step": 27720 + }, + { + "grad_norm": 0.25802016258239746, + "learning_rate": 4.5116832502845306e-05, + "loss": 0.0359, + "step": 27730 + }, + { + "grad_norm": 0.2998349070549011, + "learning_rate": 4.508392226489136e-05, + "loss": 0.0398, + "step": 27740 + }, + { + "grad_norm": 0.1482316106557846, + "learning_rate": 4.505101417739683e-05, + "loss": 0.0373, + "step": 27750 + }, + { + "grad_norm": 0.14447730779647827, + "learning_rate": 4.501810825475685e-05, + "loss": 0.0368, + "step": 27760 + }, + { + "grad_norm": 0.31831878423690796, + "learning_rate": 4.4985204511365556e-05, + "loss": 0.0475, + "step": 27770 + }, + { + "grad_norm": 0.24238106608390808, + "learning_rate": 4.49523029616162e-05, + "loss": 0.0378, + "step": 27780 + }, + { + "grad_norm": 0.1976774036884308, + "learning_rate": 4.491940361990101e-05, + "loss": 0.042, + "step": 27790 + }, + { + "grad_norm": 0.27148017287254333, + "learning_rate": 4.48865065006113e-05, + "loss": 0.0407, + "step": 27800 + }, + { + "grad_norm": 0.2535555958747864, + "learning_rate": 4.485361161813735e-05, + "loss": 0.0425, + "step": 27810 + }, + { + "grad_norm": 0.7500180602073669, + "learning_rate": 4.482071898686854e-05, + "loss": 0.0389, + "step": 27820 + }, + { + "grad_norm": 0.2336490899324417, + "learning_rate": 4.478782862119321e-05, + "loss": 0.0366, + "step": 27830 + }, + { + "grad_norm": 0.27042555809020996, + "learning_rate": 4.475494053549869e-05, + "loss": 0.0433, + "step": 27840 + }, + { + "grad_norm": 0.15648993849754333, + "learning_rate": 4.4722054744171405e-05, + "loss": 0.0381, + "step": 27850 + }, + { + "grad_norm": 0.262460321187973, + "learning_rate": 4.468917126159669e-05, + "loss": 0.0337, + "step": 27860 + }, + { + "grad_norm": 0.1946655958890915, + "learning_rate": 4.465629010215887e-05, + "loss": 0.0346, + "step": 27870 + }, + { + "grad_norm": 0.3208519518375397, + "learning_rate": 4.462341128024133e-05, + "loss": 0.0572, + "step": 27880 + }, + { + "grad_norm": 0.2386903613805771, + "learning_rate": 4.4590534810226356e-05, + "loss": 0.0431, + "step": 27890 + }, + { + "grad_norm": 0.20282046496868134, + "learning_rate": 4.455766070649523e-05, + "loss": 0.037, + "step": 27900 + }, + { + "grad_norm": 0.2272447943687439, + "learning_rate": 4.4524788983428215e-05, + "loss": 0.0428, + "step": 27910 + }, + { + "grad_norm": 0.3156401813030243, + "learning_rate": 4.4491919655404515e-05, + "loss": 0.0423, + "step": 27920 + }, + { + "grad_norm": 0.18539641797542572, + "learning_rate": 4.445905273680228e-05, + "loss": 0.0418, + "step": 27930 + }, + { + "grad_norm": 0.24765874445438385, + "learning_rate": 4.442618824199865e-05, + "loss": 0.0443, + "step": 27940 + }, + { + "grad_norm": 0.9508894085884094, + "learning_rate": 4.4393326185369635e-05, + "loss": 0.0375, + "step": 27950 + }, + { + "grad_norm": 0.2862255275249481, + "learning_rate": 4.436046658129024e-05, + "loss": 0.0455, + "step": 27960 + }, + { + "grad_norm": 0.19788923859596252, + "learning_rate": 4.432760944413433e-05, + "loss": 0.0392, + "step": 27970 + }, + { + "grad_norm": 0.6149610280990601, + "learning_rate": 4.4294754788274796e-05, + "loss": 0.039, + "step": 27980 + }, + { + "grad_norm": 0.24419766664505005, + "learning_rate": 4.426190262808334e-05, + "loss": 0.0411, + "step": 27990 + }, + { + "grad_norm": 0.3051716387271881, + "learning_rate": 4.42290529779306e-05, + "loss": 0.0483, + "step": 28000 + }, + { + "grad_norm": 0.19984018802642822, + "learning_rate": 4.4196205852186176e-05, + "loss": 0.0385, + "step": 28010 + }, + { + "grad_norm": 0.28868523240089417, + "learning_rate": 4.416336126521848e-05, + "loss": 0.0473, + "step": 28020 + }, + { + "grad_norm": 0.3551953136920929, + "learning_rate": 4.4130519231394865e-05, + "loss": 0.0416, + "step": 28030 + }, + { + "grad_norm": 0.2090347856283188, + "learning_rate": 4.409767976508154e-05, + "loss": 0.0401, + "step": 28040 + }, + { + "grad_norm": 0.30839836597442627, + "learning_rate": 4.406484288064363e-05, + "loss": 0.0376, + "step": 28050 + }, + { + "grad_norm": 0.2166980654001236, + "learning_rate": 4.4032008592445055e-05, + "loss": 0.0372, + "step": 28060 + }, + { + "grad_norm": 0.18054412305355072, + "learning_rate": 4.3999176914848714e-05, + "loss": 0.0408, + "step": 28070 + }, + { + "grad_norm": 0.19099491834640503, + "learning_rate": 4.396634786221625e-05, + "loss": 0.0373, + "step": 28080 + }, + { + "grad_norm": 0.3336857557296753, + "learning_rate": 4.3933521448908203e-05, + "loss": 0.041, + "step": 28090 + }, + { + "grad_norm": 0.15106748044490814, + "learning_rate": 4.3900697689284e-05, + "loss": 0.0383, + "step": 28100 + }, + { + "grad_norm": 0.30592405796051025, + "learning_rate": 4.386787659770185e-05, + "loss": 0.0384, + "step": 28110 + }, + { + "grad_norm": 0.20068812370300293, + "learning_rate": 4.383505818851878e-05, + "loss": 0.044, + "step": 28120 + }, + { + "grad_norm": 0.3637959063053131, + "learning_rate": 4.380224247609073e-05, + "loss": 0.0365, + "step": 28130 + }, + { + "grad_norm": 0.3100582957267761, + "learning_rate": 4.3769429474772376e-05, + "loss": 0.0378, + "step": 28140 + }, + { + "grad_norm": 0.23590798676013947, + "learning_rate": 4.373661919891726e-05, + "loss": 0.0471, + "step": 28150 + }, + { + "grad_norm": 0.21717236936092377, + "learning_rate": 4.3703811662877674e-05, + "loss": 0.0443, + "step": 28160 + }, + { + "grad_norm": 0.19387245178222656, + "learning_rate": 4.36710068810048e-05, + "loss": 0.0392, + "step": 28170 + }, + { + "grad_norm": 0.1915985345840454, + "learning_rate": 4.363820486764853e-05, + "loss": 0.0429, + "step": 28180 + }, + { + "grad_norm": 0.20320744812488556, + "learning_rate": 4.360540563715758e-05, + "loss": 0.0477, + "step": 28190 + }, + { + "grad_norm": 0.24676193296909332, + "learning_rate": 4.357260920387946e-05, + "loss": 0.0422, + "step": 28200 + }, + { + "grad_norm": 0.28041350841522217, + "learning_rate": 4.353981558216045e-05, + "loss": 0.0382, + "step": 28210 + }, + { + "grad_norm": 0.2025567591190338, + "learning_rate": 4.3507024786345565e-05, + "loss": 0.0381, + "step": 28220 + }, + { + "grad_norm": 0.37154629826545715, + "learning_rate": 4.3474236830778646e-05, + "loss": 0.0435, + "step": 28230 + }, + { + "grad_norm": 0.1718381643295288, + "learning_rate": 4.344145172980223e-05, + "loss": 0.0373, + "step": 28240 + }, + { + "grad_norm": 0.3268093466758728, + "learning_rate": 4.3408669497757656e-05, + "loss": 0.046, + "step": 28250 + }, + { + "grad_norm": 0.2736709713935852, + "learning_rate": 4.3375890148984976e-05, + "loss": 0.0451, + "step": 28260 + }, + { + "grad_norm": 0.24384810030460358, + "learning_rate": 4.334311369782301e-05, + "loss": 0.0363, + "step": 28270 + }, + { + "grad_norm": 0.24286609888076782, + "learning_rate": 4.3310340158609244e-05, + "loss": 0.0415, + "step": 28280 + }, + { + "grad_norm": 0.12236310541629791, + "learning_rate": 4.327756954568e-05, + "loss": 0.0367, + "step": 28290 + }, + { + "grad_norm": 0.12003287672996521, + "learning_rate": 4.324480187337021e-05, + "loss": 0.0386, + "step": 28300 + }, + { + "grad_norm": 0.14164325594902039, + "learning_rate": 4.3212037156013596e-05, + "loss": 0.0392, + "step": 28310 + }, + { + "grad_norm": 0.2514494061470032, + "learning_rate": 4.317927540794251e-05, + "loss": 0.0386, + "step": 28320 + }, + { + "grad_norm": 0.18796436488628387, + "learning_rate": 4.314651664348812e-05, + "loss": 0.0443, + "step": 28330 + }, + { + "grad_norm": 0.349304735660553, + "learning_rate": 4.3113760876980195e-05, + "loss": 0.0441, + "step": 28340 + }, + { + "grad_norm": 0.16897207498550415, + "learning_rate": 4.308100812274719e-05, + "loss": 0.0391, + "step": 28350 + }, + { + "grad_norm": 0.19208431243896484, + "learning_rate": 4.304825839511632e-05, + "loss": 0.0363, + "step": 28360 + }, + { + "grad_norm": 0.2454347461462021, + "learning_rate": 4.3015511708413395e-05, + "loss": 0.0435, + "step": 28370 + }, + { + "grad_norm": 0.2096932977437973, + "learning_rate": 4.298276807696294e-05, + "loss": 0.0452, + "step": 28380 + }, + { + "grad_norm": 0.174760639667511, + "learning_rate": 4.2950027515088136e-05, + "loss": 0.0424, + "step": 28390 + }, + { + "grad_norm": 0.18972258269786835, + "learning_rate": 4.291729003711083e-05, + "loss": 0.0365, + "step": 28400 + }, + { + "grad_norm": 0.2169981151819229, + "learning_rate": 4.2884555657351455e-05, + "loss": 0.0409, + "step": 28410 + }, + { + "grad_norm": 0.26336580514907837, + "learning_rate": 4.285182439012921e-05, + "loss": 0.0455, + "step": 28420 + }, + { + "grad_norm": 0.3879757225513458, + "learning_rate": 4.281909624976183e-05, + "loss": 0.0414, + "step": 28430 + }, + { + "grad_norm": 0.2518034875392914, + "learning_rate": 4.278637125056568e-05, + "loss": 0.044, + "step": 28440 + }, + { + "grad_norm": 0.3636282682418823, + "learning_rate": 4.2753649406855855e-05, + "loss": 0.0401, + "step": 28450 + }, + { + "grad_norm": 0.2771316170692444, + "learning_rate": 4.2720930732945966e-05, + "loss": 0.0397, + "step": 28460 + }, + { + "grad_norm": 0.2294473797082901, + "learning_rate": 4.268821524314826e-05, + "loss": 0.0363, + "step": 28470 + }, + { + "grad_norm": 0.17498955130577087, + "learning_rate": 4.2655502951773623e-05, + "loss": 0.0445, + "step": 28480 + }, + { + "grad_norm": 0.19324304163455963, + "learning_rate": 4.262279387313152e-05, + "loss": 0.0331, + "step": 28490 + }, + { + "grad_norm": 0.2167862504720688, + "learning_rate": 4.259008802153001e-05, + "loss": 0.0402, + "step": 28500 + }, + { + "grad_norm": 0.1905154436826706, + "learning_rate": 4.2557385411275735e-05, + "loss": 0.0403, + "step": 28510 + }, + { + "grad_norm": 0.2537941038608551, + "learning_rate": 4.252468605667395e-05, + "loss": 0.0426, + "step": 28520 + }, + { + "grad_norm": 0.2611413300037384, + "learning_rate": 4.2491989972028456e-05, + "loss": 0.0413, + "step": 28530 + }, + { + "grad_norm": 0.30923518538475037, + "learning_rate": 4.245929717164161e-05, + "loss": 0.0372, + "step": 28540 + }, + { + "grad_norm": 0.24510693550109863, + "learning_rate": 4.242660766981438e-05, + "loss": 0.0378, + "step": 28550 + }, + { + "grad_norm": 0.17100051045417786, + "learning_rate": 4.2393921480846264e-05, + "loss": 0.0417, + "step": 28560 + }, + { + "grad_norm": 0.21296367049217224, + "learning_rate": 4.236123861903528e-05, + "loss": 0.0473, + "step": 28570 + }, + { + "grad_norm": 0.24200445413589478, + "learning_rate": 4.232855909867806e-05, + "loss": 0.0406, + "step": 28580 + }, + { + "grad_norm": 0.2666257917881012, + "learning_rate": 4.229588293406972e-05, + "loss": 0.0404, + "step": 28590 + }, + { + "grad_norm": 0.26968857645988464, + "learning_rate": 4.226321013950392e-05, + "loss": 0.0324, + "step": 28600 + }, + { + "grad_norm": 0.3065314292907715, + "learning_rate": 4.223054072927286e-05, + "loss": 0.0466, + "step": 28610 + }, + { + "grad_norm": 0.18486309051513672, + "learning_rate": 4.219787471766724e-05, + "loss": 0.0438, + "step": 28620 + }, + { + "grad_norm": 0.18401570618152618, + "learning_rate": 4.216521211897629e-05, + "loss": 0.0356, + "step": 28630 + }, + { + "grad_norm": 0.21826894581317902, + "learning_rate": 4.2132552947487734e-05, + "loss": 0.0441, + "step": 28640 + }, + { + "grad_norm": 0.196818009018898, + "learning_rate": 4.209989721748783e-05, + "loss": 0.0418, + "step": 28650 + }, + { + "grad_norm": 0.19522982835769653, + "learning_rate": 4.206724494326127e-05, + "loss": 0.0397, + "step": 28660 + }, + { + "grad_norm": 0.2141752988100052, + "learning_rate": 4.203459613909126e-05, + "loss": 0.0365, + "step": 28670 + }, + { + "grad_norm": 0.1845782995223999, + "learning_rate": 4.200195081925955e-05, + "loss": 0.0409, + "step": 28680 + }, + { + "grad_norm": 0.162319615483284, + "learning_rate": 4.196930899804627e-05, + "loss": 0.0435, + "step": 28690 + }, + { + "grad_norm": 0.5740032196044922, + "learning_rate": 4.1936670689730044e-05, + "loss": 0.0368, + "step": 28700 + }, + { + "grad_norm": 0.23285657167434692, + "learning_rate": 4.190403590858804e-05, + "loss": 0.0356, + "step": 28710 + }, + { + "grad_norm": 0.12245967239141464, + "learning_rate": 4.187140466889575e-05, + "loss": 0.0396, + "step": 28720 + }, + { + "grad_norm": 0.24291767179965973, + "learning_rate": 4.183877698492724e-05, + "loss": 0.0531, + "step": 28730 + }, + { + "grad_norm": 0.23898065090179443, + "learning_rate": 4.1806152870954935e-05, + "loss": 0.0339, + "step": 28740 + }, + { + "grad_norm": 0.2194547802209854, + "learning_rate": 4.177353234124976e-05, + "loss": 0.043, + "step": 28750 + }, + { + "grad_norm": 0.1570994257926941, + "learning_rate": 4.1740915410081e-05, + "loss": 0.0456, + "step": 28760 + }, + { + "grad_norm": 0.1475139707326889, + "learning_rate": 4.170830209171646e-05, + "loss": 0.032, + "step": 28770 + }, + { + "grad_norm": 0.1709119826555252, + "learning_rate": 4.1675692400422286e-05, + "loss": 0.0396, + "step": 28780 + }, + { + "grad_norm": 0.21874411404132843, + "learning_rate": 4.164308635046304e-05, + "loss": 0.0407, + "step": 28790 + }, + { + "grad_norm": 0.1855204850435257, + "learning_rate": 4.161048395610177e-05, + "loss": 0.0389, + "step": 28800 + }, + { + "grad_norm": 0.18620269000530243, + "learning_rate": 4.157788523159985e-05, + "loss": 0.0337, + "step": 28810 + }, + { + "grad_norm": 0.17994223535060883, + "learning_rate": 4.154529019121706e-05, + "loss": 0.0403, + "step": 28820 + }, + { + "grad_norm": 0.5565552711486816, + "learning_rate": 4.1512698849211564e-05, + "loss": 0.0368, + "step": 28830 + }, + { + "grad_norm": 0.16028566658496857, + "learning_rate": 4.1480111219839954e-05, + "loss": 0.0368, + "step": 28840 + }, + { + "grad_norm": 0.45539984107017517, + "learning_rate": 4.144752731735714e-05, + "loss": 0.0467, + "step": 28850 + }, + { + "grad_norm": 0.1442885547876358, + "learning_rate": 4.1414947156016446e-05, + "loss": 0.0392, + "step": 28860 + }, + { + "grad_norm": 0.33492910861968994, + "learning_rate": 4.138237075006953e-05, + "loss": 0.0448, + "step": 28870 + }, + { + "grad_norm": 0.2017662227153778, + "learning_rate": 4.134979811376645e-05, + "loss": 0.0422, + "step": 28880 + }, + { + "grad_norm": 0.23511157929897308, + "learning_rate": 4.131722926135552e-05, + "loss": 0.0437, + "step": 28890 + }, + { + "grad_norm": 0.18362148106098175, + "learning_rate": 4.1284664207083526e-05, + "loss": 0.0429, + "step": 28900 + }, + { + "grad_norm": 0.3010881245136261, + "learning_rate": 4.125210296519551e-05, + "loss": 0.0396, + "step": 28910 + }, + { + "grad_norm": 0.1588871031999588, + "learning_rate": 4.121954554993484e-05, + "loss": 0.037, + "step": 28920 + }, + { + "grad_norm": 0.2217692881822586, + "learning_rate": 4.118699197554327e-05, + "loss": 0.0387, + "step": 28930 + }, + { + "grad_norm": 0.21976140141487122, + "learning_rate": 4.1154442256260814e-05, + "loss": 0.0395, + "step": 28940 + }, + { + "grad_norm": 0.1833856701850891, + "learning_rate": 4.1121896406325824e-05, + "loss": 0.0426, + "step": 28950 + }, + { + "grad_norm": 0.25323963165283203, + "learning_rate": 4.108935443997499e-05, + "loss": 0.0375, + "step": 28960 + }, + { + "grad_norm": 0.19561076164245605, + "learning_rate": 4.105681637144325e-05, + "loss": 0.0407, + "step": 28970 + }, + { + "grad_norm": 0.17947345972061157, + "learning_rate": 4.102428221496387e-05, + "loss": 0.0362, + "step": 28980 + }, + { + "grad_norm": 0.1231376975774765, + "learning_rate": 4.099175198476838e-05, + "loss": 0.0402, + "step": 28990 + }, + { + "grad_norm": 0.16933660209178925, + "learning_rate": 4.0959225695086635e-05, + "loss": 0.0367, + "step": 29000 + }, + { + "grad_norm": 0.2154293656349182, + "learning_rate": 4.0926703360146724e-05, + "loss": 0.0422, + "step": 29010 + }, + { + "grad_norm": 0.2567761242389679, + "learning_rate": 4.0894184994175e-05, + "loss": 0.0404, + "step": 29020 + }, + { + "grad_norm": 0.2804943323135376, + "learning_rate": 4.086167061139614e-05, + "loss": 0.0392, + "step": 29030 + }, + { + "grad_norm": 0.25781139731407166, + "learning_rate": 4.082916022603303e-05, + "loss": 0.0431, + "step": 29040 + }, + { + "grad_norm": 0.28368085622787476, + "learning_rate": 4.079665385230678e-05, + "loss": 0.0439, + "step": 29050 + }, + { + "grad_norm": 0.2548641562461853, + "learning_rate": 4.076415150443683e-05, + "loss": 0.0438, + "step": 29060 + }, + { + "grad_norm": 0.20477421581745148, + "learning_rate": 4.073165319664079e-05, + "loss": 0.0441, + "step": 29070 + }, + { + "grad_norm": 0.2112181931734085, + "learning_rate": 4.069915894313451e-05, + "loss": 0.0376, + "step": 29080 + }, + { + "grad_norm": 0.19391492009162903, + "learning_rate": 4.066666875813212e-05, + "loss": 0.0374, + "step": 29090 + }, + { + "grad_norm": 0.1752719283103943, + "learning_rate": 4.063418265584589e-05, + "loss": 0.0387, + "step": 29100 + }, + { + "grad_norm": 0.271930068731308, + "learning_rate": 4.060170065048636e-05, + "loss": 0.0379, + "step": 29110 + }, + { + "grad_norm": 0.28461381793022156, + "learning_rate": 4.056922275626227e-05, + "loss": 0.0381, + "step": 29120 + }, + { + "grad_norm": 0.18398123979568481, + "learning_rate": 4.053674898738056e-05, + "loss": 0.0347, + "step": 29130 + }, + { + "grad_norm": 0.14471934735774994, + "learning_rate": 4.050427935804633e-05, + "loss": 0.0387, + "step": 29140 + }, + { + "grad_norm": 0.19377751648426056, + "learning_rate": 4.047181388246293e-05, + "loss": 0.0374, + "step": 29150 + }, + { + "grad_norm": 0.23109528422355652, + "learning_rate": 4.043935257483185e-05, + "loss": 0.0475, + "step": 29160 + }, + { + "grad_norm": 0.33038729429244995, + "learning_rate": 4.0406895449352775e-05, + "loss": 0.0471, + "step": 29170 + }, + { + "grad_norm": 0.20396077632904053, + "learning_rate": 4.0374442520223525e-05, + "loss": 0.0415, + "step": 29180 + }, + { + "grad_norm": 0.26014503836631775, + "learning_rate": 4.0341993801640155e-05, + "loss": 0.0384, + "step": 29190 + }, + { + "grad_norm": 0.20859824120998383, + "learning_rate": 4.030954930779681e-05, + "loss": 0.0422, + "step": 29200 + }, + { + "grad_norm": 0.2075911909341812, + "learning_rate": 4.027710905288582e-05, + "loss": 0.0413, + "step": 29210 + }, + { + "grad_norm": 0.235653817653656, + "learning_rate": 4.0244673051097666e-05, + "loss": 0.0439, + "step": 29220 + }, + { + "grad_norm": 0.2508152723312378, + "learning_rate": 4.021224131662095e-05, + "loss": 0.0412, + "step": 29230 + }, + { + "grad_norm": 0.25134527683258057, + "learning_rate": 4.0179813863642396e-05, + "loss": 0.0424, + "step": 29240 + }, + { + "grad_norm": 0.23619815707206726, + "learning_rate": 4.014739070634691e-05, + "loss": 0.04, + "step": 29250 + }, + { + "grad_norm": 0.2932503819465637, + "learning_rate": 4.011497185891746e-05, + "loss": 0.0403, + "step": 29260 + }, + { + "grad_norm": 0.18264251947402954, + "learning_rate": 4.008255733553513e-05, + "loss": 0.0418, + "step": 29270 + }, + { + "grad_norm": 0.16437993943691254, + "learning_rate": 4.0050147150379194e-05, + "loss": 0.0375, + "step": 29280 + }, + { + "grad_norm": 0.1675713062286377, + "learning_rate": 4.001774131762692e-05, + "loss": 0.0407, + "step": 29290 + }, + { + "grad_norm": 0.23732753098011017, + "learning_rate": 3.9985339851453705e-05, + "loss": 0.0394, + "step": 29300 + }, + { + "grad_norm": 0.6330673694610596, + "learning_rate": 3.9952942766033114e-05, + "loss": 0.046, + "step": 29310 + }, + { + "grad_norm": 0.21205027401447296, + "learning_rate": 3.9920550075536686e-05, + "loss": 0.042, + "step": 29320 + }, + { + "grad_norm": 0.25067654252052307, + "learning_rate": 3.988816179413408e-05, + "loss": 0.0388, + "step": 29330 + }, + { + "grad_norm": 0.28034064173698425, + "learning_rate": 3.985577793599307e-05, + "loss": 0.0411, + "step": 29340 + }, + { + "grad_norm": 0.19013310968875885, + "learning_rate": 3.982339851527942e-05, + "loss": 0.0458, + "step": 29350 + }, + { + "grad_norm": 0.19631612300872803, + "learning_rate": 3.979102354615701e-05, + "loss": 0.0379, + "step": 29360 + }, + { + "grad_norm": 0.21214808523654938, + "learning_rate": 3.975865304278772e-05, + "loss": 0.0441, + "step": 29370 + }, + { + "grad_norm": 0.27955734729766846, + "learning_rate": 3.9726287019331546e-05, + "loss": 0.0412, + "step": 29380 + }, + { + "grad_norm": 0.2193417102098465, + "learning_rate": 3.9693925489946474e-05, + "loss": 0.0346, + "step": 29390 + }, + { + "grad_norm": 0.2135174423456192, + "learning_rate": 3.966156846878852e-05, + "loss": 0.0399, + "step": 29400 + }, + { + "grad_norm": 0.1881139576435089, + "learning_rate": 3.962921597001177e-05, + "loss": 0.0405, + "step": 29410 + }, + { + "grad_norm": 0.3922048807144165, + "learning_rate": 3.9596868007768294e-05, + "loss": 0.0373, + "step": 29420 + }, + { + "grad_norm": 0.19986343383789062, + "learning_rate": 3.9564524596208165e-05, + "loss": 0.0366, + "step": 29430 + }, + { + "grad_norm": 0.26563000679016113, + "learning_rate": 3.953218574947954e-05, + "loss": 0.0354, + "step": 29440 + }, + { + "grad_norm": 0.18783007562160492, + "learning_rate": 3.949985148172849e-05, + "loss": 0.0377, + "step": 29450 + }, + { + "grad_norm": 0.25012075901031494, + "learning_rate": 3.946752180709914e-05, + "loss": 0.0354, + "step": 29460 + }, + { + "grad_norm": 0.18234556913375854, + "learning_rate": 3.943519673973357e-05, + "loss": 0.0365, + "step": 29470 + }, + { + "grad_norm": 0.21702563762664795, + "learning_rate": 3.940287629377189e-05, + "loss": 0.0428, + "step": 29480 + }, + { + "grad_norm": 0.29365813732147217, + "learning_rate": 3.937056048335214e-05, + "loss": 0.0398, + "step": 29490 + }, + { + "grad_norm": 0.22889494895935059, + "learning_rate": 3.933824932261037e-05, + "loss": 0.0455, + "step": 29500 + }, + { + "grad_norm": 0.22234588861465454, + "learning_rate": 3.9305942825680576e-05, + "loss": 0.038, + "step": 29510 + }, + { + "grad_norm": 0.265420526266098, + "learning_rate": 3.9273641006694706e-05, + "loss": 0.0337, + "step": 29520 + }, + { + "grad_norm": 0.4012007415294647, + "learning_rate": 3.924134387978266e-05, + "loss": 0.0425, + "step": 29530 + }, + { + "grad_norm": 0.26083412766456604, + "learning_rate": 3.920905145907234e-05, + "loss": 0.0486, + "step": 29540 + }, + { + "grad_norm": 0.23669490218162537, + "learning_rate": 3.9176763758689525e-05, + "loss": 0.0383, + "step": 29550 + }, + { + "grad_norm": 0.19917179644107819, + "learning_rate": 3.9144480792757925e-05, + "loss": 0.0365, + "step": 29560 + }, + { + "grad_norm": 0.17789767682552338, + "learning_rate": 3.911220257539926e-05, + "loss": 0.037, + "step": 29570 + }, + { + "grad_norm": 0.21543094515800476, + "learning_rate": 3.907992912073308e-05, + "loss": 0.0422, + "step": 29580 + }, + { + "grad_norm": 0.1968764215707779, + "learning_rate": 3.90476604428769e-05, + "loss": 0.0378, + "step": 29590 + }, + { + "grad_norm": 0.18774478137493134, + "learning_rate": 3.901539655594614e-05, + "loss": 0.0357, + "step": 29600 + }, + { + "grad_norm": 0.198426753282547, + "learning_rate": 3.898313747405413e-05, + "loss": 0.0342, + "step": 29610 + }, + { + "grad_norm": 0.21987013518810272, + "learning_rate": 3.895088321131205e-05, + "loss": 0.0361, + "step": 29620 + }, + { + "grad_norm": 0.2543981373310089, + "learning_rate": 3.891863378182907e-05, + "loss": 0.0427, + "step": 29630 + }, + { + "grad_norm": 0.17579501867294312, + "learning_rate": 3.8886389199712145e-05, + "loss": 0.0372, + "step": 29640 + }, + { + "grad_norm": 0.16945664584636688, + "learning_rate": 3.885414947906614e-05, + "loss": 0.0356, + "step": 29650 + }, + { + "grad_norm": 0.19443975389003754, + "learning_rate": 3.882191463399385e-05, + "loss": 0.0368, + "step": 29660 + }, + { + "grad_norm": 0.31935274600982666, + "learning_rate": 3.878968467859585e-05, + "loss": 0.0421, + "step": 29670 + }, + { + "grad_norm": 0.19033153355121613, + "learning_rate": 3.875745962697063e-05, + "loss": 0.041, + "step": 29680 + }, + { + "grad_norm": 0.2068074345588684, + "learning_rate": 3.872523949321454e-05, + "loss": 0.0395, + "step": 29690 + }, + { + "grad_norm": 0.17546594142913818, + "learning_rate": 3.869302429142173e-05, + "loss": 0.0344, + "step": 29700 + }, + { + "grad_norm": 0.1593313217163086, + "learning_rate": 3.866081403568425e-05, + "loss": 0.0358, + "step": 29710 + }, + { + "grad_norm": 0.2145960032939911, + "learning_rate": 3.8628608740091934e-05, + "loss": 0.0398, + "step": 29720 + }, + { + "grad_norm": 0.21726202964782715, + "learning_rate": 3.85964084187325e-05, + "loss": 0.0434, + "step": 29730 + }, + { + "grad_norm": 0.1498306542634964, + "learning_rate": 3.856421308569146e-05, + "loss": 0.0339, + "step": 29740 + }, + { + "grad_norm": 0.22501511871814728, + "learning_rate": 3.853202275505212e-05, + "loss": 0.0389, + "step": 29750 + }, + { + "grad_norm": 0.20161031186580658, + "learning_rate": 3.849983744089565e-05, + "loss": 0.0377, + "step": 29760 + }, + { + "grad_norm": 0.15441448986530304, + "learning_rate": 3.8467657157301e-05, + "loss": 0.0409, + "step": 29770 + }, + { + "grad_norm": 0.1651192605495453, + "learning_rate": 3.8435481918344884e-05, + "loss": 0.041, + "step": 29780 + }, + { + "grad_norm": 0.4408777356147766, + "learning_rate": 3.84033117381019e-05, + "loss": 0.0357, + "step": 29790 + }, + { + "grad_norm": 0.2229592651128769, + "learning_rate": 3.8371146630644335e-05, + "loss": 0.0354, + "step": 29800 + }, + { + "grad_norm": 0.1340201050043106, + "learning_rate": 3.833898661004231e-05, + "loss": 0.0344, + "step": 29810 + }, + { + "grad_norm": 0.14352409541606903, + "learning_rate": 3.830683169036372e-05, + "loss": 0.0323, + "step": 29820 + }, + { + "grad_norm": 0.2811305820941925, + "learning_rate": 3.827468188567422e-05, + "loss": 0.0471, + "step": 29830 + }, + { + "grad_norm": 0.22450487315654755, + "learning_rate": 3.82425372100372e-05, + "loss": 0.0437, + "step": 29840 + }, + { + "grad_norm": 0.24634194374084473, + "learning_rate": 3.821039767751387e-05, + "loss": 0.0419, + "step": 29850 + }, + { + "grad_norm": 0.2428329586982727, + "learning_rate": 3.817826330216315e-05, + "loss": 0.0402, + "step": 29860 + }, + { + "grad_norm": 0.20882609486579895, + "learning_rate": 3.814613409804169e-05, + "loss": 0.037, + "step": 29870 + }, + { + "grad_norm": 0.2825510799884796, + "learning_rate": 3.811401007920388e-05, + "loss": 0.0457, + "step": 29880 + }, + { + "grad_norm": 0.15261554718017578, + "learning_rate": 3.8081891259701895e-05, + "loss": 0.0363, + "step": 29890 + }, + { + "grad_norm": 0.23255451023578644, + "learning_rate": 3.8049777653585584e-05, + "loss": 0.0444, + "step": 29900 + }, + { + "grad_norm": 0.4298957288265228, + "learning_rate": 3.80176692749025e-05, + "loss": 0.0389, + "step": 29910 + }, + { + "grad_norm": 0.25413399934768677, + "learning_rate": 3.798556613769799e-05, + "loss": 0.0381, + "step": 29920 + }, + { + "grad_norm": 0.24871167540550232, + "learning_rate": 3.795346825601502e-05, + "loss": 0.0458, + "step": 29930 + }, + { + "grad_norm": 0.20103898644447327, + "learning_rate": 3.7921375643894305e-05, + "loss": 0.0426, + "step": 29940 + }, + { + "grad_norm": 0.18064256012439728, + "learning_rate": 3.788928831537424e-05, + "loss": 0.0416, + "step": 29950 + }, + { + "grad_norm": 0.2360270470380783, + "learning_rate": 3.7857206284490934e-05, + "loss": 0.0435, + "step": 29960 + }, + { + "grad_norm": 0.2410675287246704, + "learning_rate": 3.7825129565278114e-05, + "loss": 0.0389, + "step": 29970 + }, + { + "grad_norm": 0.19112522900104523, + "learning_rate": 3.779305817176728e-05, + "loss": 0.0412, + "step": 29980 + }, + { + "grad_norm": 0.18204303085803986, + "learning_rate": 3.776099211798751e-05, + "loss": 0.042, + "step": 29990 + }, + { + "grad_norm": 0.1981995701789856, + "learning_rate": 3.7728931417965584e-05, + "loss": 0.045, + "step": 30000 + } + ], + "logging_steps": 10, + "max_steps": 50000, + "num_input_tokens_seen": 0, + "num_train_epochs": 9223372036854775807, + "save_steps": 2500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}