diff --git "a/trainer_state.json" "b/trainer_state.json" --- "a/trainer_state.json" +++ "b/trainer_state.json" @@ -2,12016 +2,36016 @@ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, - "epoch": 1.0, + "epoch": 0.85, "eval_steps": 500, - "global_step": 20000, + "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { - "grad_norm": 4.4084038734436035, - "learning_rate": 9e-07, - "loss": 1.3469, + "grad_norm": 4.447365760803223, + "learning_rate": 3.0000000000000004e-07, + "loss": 1.3484, "step": 10 }, { - "grad_norm": 2.854828119277954, - "learning_rate": 1.9e-06, - "loss": 1.3141, + "grad_norm": 4.113630771636963, + "learning_rate": 6.333333333333333e-07, + "loss": 1.3434, "step": 20 }, { - "grad_norm": 1.38284432888031, - "learning_rate": 2.9e-06, - "loss": 1.2086, + "grad_norm": 3.8941195011138916, + "learning_rate": 9.666666666666668e-07, + "loss": 1.3238, "step": 30 }, { - "grad_norm": 0.6830054521560669, - "learning_rate": 3.9e-06, - "loss": 1.1273, + "grad_norm": 2.6234889030456543, + "learning_rate": 1.3e-06, + "loss": 1.2652, "step": 40 }, { - "grad_norm": 0.3296288847923279, - "learning_rate": 4.9000000000000005e-06, - "loss": 1.0875, + "grad_norm": 1.3288973569869995, + "learning_rate": 1.6333333333333333e-06, + "loss": 1.1887, "step": 50 }, { - "grad_norm": 0.3340240716934204, - "learning_rate": 5.9e-06, - "loss": 1.075, + "grad_norm": 0.8369770646095276, + "learning_rate": 1.9666666666666668e-06, + "loss": 1.1305, "step": 60 }, { - "grad_norm": 0.2849108874797821, - "learning_rate": 6.900000000000001e-06, - "loss": 1.0703, + "grad_norm": 0.4795992970466614, + "learning_rate": 2.3e-06, + "loss": 1.0973, "step": 70 }, { - "grad_norm": 0.31928372383117676, - "learning_rate": 7.9e-06, - "loss": 1.0582, + "grad_norm": 0.2965182065963745, + "learning_rate": 2.6333333333333337e-06, + "loss": 1.0801, "step": 80 }, { - "grad_norm": 0.3880952298641205, - "learning_rate": 8.9e-06, - "loss": 1.0555, + "grad_norm": 0.32636141777038574, + "learning_rate": 2.966666666666667e-06, + "loss": 1.0785, "step": 90 }, { - "grad_norm": 0.30215585231781006, - "learning_rate": 9.900000000000002e-06, - "loss": 1.048, + "grad_norm": 0.3556094169616699, + "learning_rate": 3.3e-06, + "loss": 1.075, "step": 100 }, { - "grad_norm": 0.43695512413978577, - "learning_rate": 1.09e-05, - "loss": 1.043, + "grad_norm": 0.35541146993637085, + "learning_rate": 3.633333333333334e-06, + "loss": 1.0656, "step": 110 }, { - "grad_norm": 0.4662553668022156, - "learning_rate": 1.19e-05, - "loss": 1.0406, + "grad_norm": 0.3391725420951843, + "learning_rate": 3.966666666666667e-06, + "loss": 1.0664, "step": 120 }, { - "grad_norm": 0.7485678195953369, - "learning_rate": 1.29e-05, - "loss": 1.0172, + "grad_norm": 0.4171597361564636, + "learning_rate": 4.2999999999999995e-06, + "loss": 1.0477, "step": 130 }, { - "grad_norm": 0.7978800535202026, - "learning_rate": 1.3900000000000002e-05, - "loss": 1.0072, + "grad_norm": 0.3247183859348297, + "learning_rate": 4.633333333333334e-06, + "loss": 1.0523, "step": 140 }, { - "grad_norm": 0.7703062295913696, - "learning_rate": 1.49e-05, - "loss": 0.9766, + "grad_norm": 0.4718841016292572, + "learning_rate": 4.966666666666667e-06, + "loss": 1.0461, "step": 150 }, { - "grad_norm": 1.0931109189987183, - "learning_rate": 1.59e-05, - "loss": 0.9051, + "grad_norm": 0.4053254723548889, + "learning_rate": 5.3e-06, + "loss": 1.0422, "step": 160 }, { - "grad_norm": 1.3214606046676636, - "learning_rate": 1.69e-05, - "loss": 0.8164, + "grad_norm": 0.39676275849342346, + "learning_rate": 5.633333333333333e-06, + "loss": 1.043, "step": 170 }, { - "grad_norm": 1.6509478092193604, - "learning_rate": 1.79e-05, - "loss": 0.7283, + "grad_norm": 0.3670744299888611, + "learning_rate": 5.9666666666666666e-06, + "loss": 1.0371, "step": 180 }, { - "grad_norm": 1.649338722229004, - "learning_rate": 1.8900000000000002e-05, - "loss": 0.6459, + "grad_norm": 0.542593777179718, + "learning_rate": 6.300000000000001e-06, + "loss": 1.0336, "step": 190 }, { - "grad_norm": 1.5137853622436523, - "learning_rate": 1.9900000000000003e-05, - "loss": 0.5609, + "grad_norm": 0.5725336670875549, + "learning_rate": 6.633333333333333e-06, + "loss": 1.0227, "step": 200 }, { - "grad_norm": 1.676445722579956, - "learning_rate": 2.09e-05, - "loss": 0.5043, + "grad_norm": 0.6448169350624084, + "learning_rate": 6.966666666666667e-06, + "loss": 1.0158, "step": 210 }, { - "grad_norm": 1.5472042560577393, - "learning_rate": 2.19e-05, - "loss": 0.4341, + "grad_norm": 0.5190443992614746, + "learning_rate": 7.2999999999999996e-06, + "loss": 1.0049, "step": 220 }, { - "grad_norm": 2.0228428840637207, - "learning_rate": 2.29e-05, - "loss": 0.38, + "grad_norm": 0.6504555344581604, + "learning_rate": 7.633333333333334e-06, + "loss": 0.9838, "step": 230 }, { - "grad_norm": 2.0979020595550537, - "learning_rate": 2.39e-05, - "loss": 0.3448, + "grad_norm": 0.8528154492378235, + "learning_rate": 7.966666666666666e-06, + "loss": 0.9516, "step": 240 }, { - "grad_norm": 1.9566534757614136, - "learning_rate": 2.4900000000000002e-05, - "loss": 0.3183, + "grad_norm": 1.0889426469802856, + "learning_rate": 8.3e-06, + "loss": 0.8994, "step": 250 }, { - "grad_norm": 1.620675802230835, - "learning_rate": 2.5900000000000003e-05, - "loss": 0.2805, + "grad_norm": 1.0057774782180786, + "learning_rate": 8.633333333333334e-06, + "loss": 0.8383, "step": 260 }, { - "grad_norm": 2.3286774158477783, - "learning_rate": 2.6900000000000003e-05, - "loss": 0.2455, + "grad_norm": 1.1228306293487549, + "learning_rate": 8.966666666666668e-06, + "loss": 0.7645, "step": 270 }, { - "grad_norm": 1.9605939388275146, - "learning_rate": 2.7900000000000004e-05, - "loss": 0.2306, + "grad_norm": 1.5069864988327026, + "learning_rate": 9.3e-06, + "loss": 0.7127, "step": 280 }, { - "grad_norm": 2.058300733566284, - "learning_rate": 2.8899999999999998e-05, - "loss": 0.2136, + "grad_norm": 1.491326093673706, + "learning_rate": 9.633333333333335e-06, + "loss": 0.6578, "step": 290 }, { - "grad_norm": 2.3196494579315186, - "learning_rate": 2.9900000000000002e-05, - "loss": 0.1968, + "grad_norm": 1.7207450866699219, + "learning_rate": 9.966666666666667e-06, + "loss": 0.6078, "step": 300 }, { - "grad_norm": 2.2974038124084473, - "learning_rate": 3.09e-05, - "loss": 0.1902, + "grad_norm": 1.2145884037017822, + "learning_rate": 1.03e-05, + "loss": 0.5693, "step": 310 }, { - "grad_norm": 2.038285732269287, - "learning_rate": 3.19e-05, - "loss": 0.161, + "grad_norm": 1.5747157335281372, + "learning_rate": 1.0633333333333334e-05, + "loss": 0.5098, "step": 320 }, { - "grad_norm": 1.532688856124878, - "learning_rate": 3.29e-05, - "loss": 0.1528, + "grad_norm": 1.5645968914031982, + "learning_rate": 1.0966666666666666e-05, + "loss": 0.4669, "step": 330 }, { - "grad_norm": 2.092379331588745, - "learning_rate": 3.3900000000000004e-05, - "loss": 0.143, + "grad_norm": 1.6364047527313232, + "learning_rate": 1.13e-05, + "loss": 0.435, "step": 340 }, { - "grad_norm": 2.0367798805236816, - "learning_rate": 3.49e-05, - "loss": 0.1339, + "grad_norm": 1.8076460361480713, + "learning_rate": 1.1633333333333334e-05, + "loss": 0.4071, "step": 350 }, { - "grad_norm": 2.3606696128845215, - "learning_rate": 3.59e-05, - "loss": 0.1321, + "grad_norm": 1.8036781549453735, + "learning_rate": 1.1966666666666668e-05, + "loss": 0.3798, "step": 360 }, { - "grad_norm": 1.742854118347168, - "learning_rate": 3.69e-05, - "loss": 0.1226, + "grad_norm": 1.761277437210083, + "learning_rate": 1.23e-05, + "loss": 0.3501, "step": 370 }, { - "grad_norm": 2.3252644538879395, - "learning_rate": 3.79e-05, - "loss": 0.1165, + "grad_norm": 1.606987476348877, + "learning_rate": 1.2633333333333333e-05, + "loss": 0.3208, "step": 380 }, { - "grad_norm": 2.4858038425445557, - "learning_rate": 3.8900000000000004e-05, - "loss": 0.1215, + "grad_norm": 1.7726079225540161, + "learning_rate": 1.2966666666666669e-05, + "loss": 0.3004, "step": 390 }, { - "grad_norm": 2.094243049621582, - "learning_rate": 3.99e-05, - "loss": 0.1252, + "grad_norm": 2.3330554962158203, + "learning_rate": 1.3300000000000001e-05, + "loss": 0.2878, "step": 400 }, { - "grad_norm": 1.9110815525054932, - "learning_rate": 4.09e-05, - "loss": 0.1219, + "grad_norm": 1.6834126710891724, + "learning_rate": 1.3633333333333334e-05, + "loss": 0.2703, "step": 410 }, { - "grad_norm": 2.131819725036621, - "learning_rate": 4.19e-05, - "loss": 0.1087, + "grad_norm": 1.9400217533111572, + "learning_rate": 1.3966666666666666e-05, + "loss": 0.2336, "step": 420 }, { - "grad_norm": 2.0112192630767822, - "learning_rate": 4.29e-05, - "loss": 0.1132, + "grad_norm": 1.739591360092163, + "learning_rate": 1.43e-05, + "loss": 0.2212, "step": 430 }, { - "grad_norm": 2.2110557556152344, - "learning_rate": 4.39e-05, - "loss": 0.1113, + "grad_norm": 2.0412356853485107, + "learning_rate": 1.4633333333333334e-05, + "loss": 0.2075, "step": 440 }, { - "grad_norm": 2.171856164932251, - "learning_rate": 4.49e-05, - "loss": 0.1198, + "grad_norm": 2.1918587684631348, + "learning_rate": 1.4966666666666668e-05, + "loss": 0.2056, "step": 450 }, { - "grad_norm": 1.8911032676696777, - "learning_rate": 4.5900000000000004e-05, - "loss": 0.1119, + "grad_norm": 2.2246336936950684, + "learning_rate": 1.53e-05, + "loss": 0.1877, "step": 460 }, { - "grad_norm": 2.165710926055908, - "learning_rate": 4.69e-05, - "loss": 0.1167, + "grad_norm": 2.368511199951172, + "learning_rate": 1.563333333333333e-05, + "loss": 0.187, "step": 470 }, { - "grad_norm": 2.065707206726074, - "learning_rate": 4.79e-05, - "loss": 0.1074, + "grad_norm": 2.1554298400878906, + "learning_rate": 1.5966666666666667e-05, + "loss": 0.1694, "step": 480 }, { - "grad_norm": 1.9315952062606812, - "learning_rate": 4.89e-05, - "loss": 0.1016, + "grad_norm": 2.1426496505737305, + "learning_rate": 1.63e-05, + "loss": 0.1537, "step": 490 }, { - "grad_norm": 2.0874738693237305, - "learning_rate": 4.99e-05, - "loss": 0.0999, + "grad_norm": 2.1911087036132812, + "learning_rate": 1.6633333333333336e-05, + "loss": 0.147, "step": 500 }, { - "grad_norm": 1.8715001344680786, - "learning_rate": 5.0900000000000004e-05, - "loss": 0.0998, + "grad_norm": 2.10931134223938, + "learning_rate": 1.6966666666666668e-05, + "loss": 0.1434, "step": 510 }, { - "grad_norm": 1.6038424968719482, - "learning_rate": 5.19e-05, - "loss": 0.0966, + "grad_norm": 2.0908308029174805, + "learning_rate": 1.73e-05, + "loss": 0.1382, "step": 520 }, { - "grad_norm": 1.9660725593566895, - "learning_rate": 5.2900000000000005e-05, - "loss": 0.0977, + "grad_norm": 2.3200180530548096, + "learning_rate": 1.7633333333333336e-05, + "loss": 0.1346, "step": 530 }, { - "grad_norm": 1.8537325859069824, - "learning_rate": 5.390000000000001e-05, - "loss": 0.1056, + "grad_norm": 2.0884504318237305, + "learning_rate": 1.796666666666667e-05, + "loss": 0.1432, "step": 540 }, { - "grad_norm": 1.6274349689483643, - "learning_rate": 5.4900000000000006e-05, - "loss": 0.094, + "grad_norm": 2.292975425720215, + "learning_rate": 1.83e-05, + "loss": 0.1266, "step": 550 }, { - "grad_norm": 1.6068252325057983, - "learning_rate": 5.590000000000001e-05, - "loss": 0.1028, + "grad_norm": 1.8442728519439697, + "learning_rate": 1.8633333333333333e-05, + "loss": 0.1323, "step": 560 }, { - "grad_norm": 1.5526419878005981, - "learning_rate": 5.69e-05, - "loss": 0.1021, + "grad_norm": 2.1122641563415527, + "learning_rate": 1.896666666666667e-05, + "loss": 0.1322, "step": 570 }, { - "grad_norm": 1.9871398210525513, - "learning_rate": 5.79e-05, - "loss": 0.0982, + "grad_norm": 2.50030255317688, + "learning_rate": 1.93e-05, + "loss": 0.1261, "step": 580 }, { - "grad_norm": 1.8751955032348633, - "learning_rate": 5.89e-05, - "loss": 0.0894, + "grad_norm": 2.5016114711761475, + "learning_rate": 1.9633333333333334e-05, + "loss": 0.1184, "step": 590 }, { - "grad_norm": 1.8565336465835571, - "learning_rate": 5.99e-05, - "loss": 0.0947, + "grad_norm": 2.4425137042999268, + "learning_rate": 1.9966666666666666e-05, + "loss": 0.1247, "step": 600 }, { - "grad_norm": 1.7967112064361572, - "learning_rate": 6.09e-05, - "loss": 0.0876, + "grad_norm": 1.9957345724105835, + "learning_rate": 2.0300000000000002e-05, + "loss": 0.1137, "step": 610 }, { - "grad_norm": 1.6690943241119385, - "learning_rate": 6.19e-05, - "loss": 0.0907, + "grad_norm": 2.0271806716918945, + "learning_rate": 2.0633333333333335e-05, + "loss": 0.1167, "step": 620 }, { - "grad_norm": 1.8858799934387207, - "learning_rate": 6.29e-05, - "loss": 0.0909, + "grad_norm": 2.0384092330932617, + "learning_rate": 2.0966666666666667e-05, + "loss": 0.1129, "step": 630 }, { - "grad_norm": 1.6637696027755737, - "learning_rate": 6.390000000000001e-05, - "loss": 0.0874, + "grad_norm": 2.214675188064575, + "learning_rate": 2.13e-05, + "loss": 0.1092, "step": 640 }, { - "grad_norm": 1.7460479736328125, - "learning_rate": 6.49e-05, - "loss": 0.09, + "grad_norm": 2.3753225803375244, + "learning_rate": 2.1633333333333332e-05, + "loss": 0.1095, "step": 650 }, { - "grad_norm": 1.7727495431900024, - "learning_rate": 6.59e-05, - "loss": 0.092, + "grad_norm": 1.9375733137130737, + "learning_rate": 2.1966666666666668e-05, + "loss": 0.1152, "step": 660 }, { - "grad_norm": 1.4950904846191406, - "learning_rate": 6.690000000000001e-05, - "loss": 0.0923, + "grad_norm": 1.9821141958236694, + "learning_rate": 2.23e-05, + "loss": 0.1122, "step": 670 }, { - "grad_norm": 1.44473397731781, - "learning_rate": 6.790000000000001e-05, - "loss": 0.0891, + "grad_norm": 1.9919532537460327, + "learning_rate": 2.2633333333333336e-05, + "loss": 0.1044, "step": 680 }, { - "grad_norm": 1.5126936435699463, - "learning_rate": 6.89e-05, - "loss": 0.0874, + "grad_norm": 2.376053810119629, + "learning_rate": 2.2966666666666668e-05, + "loss": 0.1048, "step": 690 }, { - "grad_norm": 1.394876480102539, - "learning_rate": 6.99e-05, - "loss": 0.0849, + "grad_norm": 1.8856782913208008, + "learning_rate": 2.3300000000000004e-05, + "loss": 0.1028, "step": 700 }, { - "grad_norm": 1.681939721107483, - "learning_rate": 7.09e-05, - "loss": 0.0966, + "grad_norm": 2.3740177154541016, + "learning_rate": 2.3633333333333336e-05, + "loss": 0.1176, "step": 710 }, { - "grad_norm": 1.6944371461868286, - "learning_rate": 7.19e-05, - "loss": 0.0864, + "grad_norm": 2.156780958175659, + "learning_rate": 2.396666666666667e-05, + "loss": 0.1048, "step": 720 }, { - "grad_norm": 1.6403831243515015, - "learning_rate": 7.29e-05, - "loss": 0.0764, + "grad_norm": 2.4900104999542236, + "learning_rate": 2.43e-05, + "loss": 0.0942, "step": 730 }, { - "grad_norm": 1.59284245967865, - "learning_rate": 7.390000000000001e-05, - "loss": 0.0874, + "grad_norm": 2.1864829063415527, + "learning_rate": 2.4633333333333334e-05, + "loss": 0.1058, "step": 740 }, { - "grad_norm": 1.707362174987793, - "learning_rate": 7.49e-05, - "loss": 0.0837, + "grad_norm": 2.5174381732940674, + "learning_rate": 2.496666666666667e-05, + "loss": 0.1036, "step": 750 }, { - "grad_norm": 1.5193947553634644, - "learning_rate": 7.59e-05, - "loss": 0.0939, + "grad_norm": 2.1998066902160645, + "learning_rate": 2.5300000000000002e-05, + "loss": 0.1084, "step": 760 }, { - "grad_norm": 1.5924936532974243, - "learning_rate": 7.69e-05, - "loss": 0.0903, + "grad_norm": 2.03684663772583, + "learning_rate": 2.5633333333333338e-05, + "loss": 0.1058, "step": 770 }, { - "grad_norm": 1.5755540132522583, - "learning_rate": 7.790000000000001e-05, - "loss": 0.0854, + "grad_norm": 2.7059671878814697, + "learning_rate": 2.5966666666666667e-05, + "loss": 0.1025, "step": 780 }, { - "grad_norm": 1.28225576877594, - "learning_rate": 7.890000000000001e-05, - "loss": 0.0902, + "grad_norm": 2.069469928741455, + "learning_rate": 2.6300000000000002e-05, + "loss": 0.1068, "step": 790 }, { - "grad_norm": 1.3005540370941162, - "learning_rate": 7.99e-05, - "loss": 0.0885, + "grad_norm": 1.898769497871399, + "learning_rate": 2.663333333333333e-05, + "loss": 0.102, "step": 800 }, { - "grad_norm": 1.4294734001159668, - "learning_rate": 8.090000000000001e-05, - "loss": 0.084, + "grad_norm": 1.8516987562179565, + "learning_rate": 2.6966666666666667e-05, + "loss": 0.0965, "step": 810 }, { - "grad_norm": 1.44032883644104, - "learning_rate": 8.19e-05, - "loss": 0.0816, + "grad_norm": 1.918587565422058, + "learning_rate": 2.7300000000000003e-05, + "loss": 0.0971, "step": 820 }, { - "grad_norm": 1.2144129276275635, - "learning_rate": 8.29e-05, - "loss": 0.0845, + "grad_norm": 2.0393455028533936, + "learning_rate": 2.7633333333333332e-05, + "loss": 0.0989, "step": 830 }, { - "grad_norm": 1.4295352697372437, - "learning_rate": 8.39e-05, - "loss": 0.0838, + "grad_norm": 1.5016096830368042, + "learning_rate": 2.7966666666666668e-05, + "loss": 0.0954, "step": 840 }, { - "grad_norm": 1.4729937314987183, - "learning_rate": 8.49e-05, - "loss": 0.0849, + "grad_norm": 1.993768572807312, + "learning_rate": 2.83e-05, + "loss": 0.0997, "step": 850 }, { - "grad_norm": 1.2805378437042236, - "learning_rate": 8.59e-05, - "loss": 0.0899, + "grad_norm": 2.1159729957580566, + "learning_rate": 2.8633333333333336e-05, + "loss": 0.1051, "step": 860 }, { - "grad_norm": 1.2634493112564087, - "learning_rate": 8.69e-05, - "loss": 0.0815, + "grad_norm": 1.8566912412643433, + "learning_rate": 2.8966666666666668e-05, + "loss": 0.0952, "step": 870 }, { - "grad_norm": 1.3772571086883545, - "learning_rate": 8.790000000000001e-05, - "loss": 0.0818, + "grad_norm": 2.3972713947296143, + "learning_rate": 2.93e-05, + "loss": 0.0976, "step": 880 }, { - "grad_norm": 1.173386573791504, - "learning_rate": 8.89e-05, - "loss": 0.0792, + "grad_norm": 1.9326415061950684, + "learning_rate": 2.9633333333333336e-05, + "loss": 0.0942, "step": 890 }, { - "grad_norm": 1.140397548675537, - "learning_rate": 8.99e-05, - "loss": 0.0738, + "grad_norm": 1.6511321067810059, + "learning_rate": 2.9966666666666672e-05, + "loss": 0.0877, "step": 900 }, { - "grad_norm": 1.2523930072784424, - "learning_rate": 9.090000000000001e-05, - "loss": 0.0728, + "grad_norm": 1.6019006967544556, + "learning_rate": 3.03e-05, + "loss": 0.0835, "step": 910 }, { - "grad_norm": 1.2446190118789673, - "learning_rate": 9.190000000000001e-05, - "loss": 0.0733, + "grad_norm": 2.0079538822174072, + "learning_rate": 3.063333333333334e-05, + "loss": 0.0883, "step": 920 }, { - "grad_norm": 1.123772144317627, - "learning_rate": 9.290000000000001e-05, - "loss": 0.0752, + "grad_norm": 1.8352253437042236, + "learning_rate": 3.096666666666666e-05, + "loss": 0.0894, "step": 930 }, { - "grad_norm": 1.2798773050308228, - "learning_rate": 9.39e-05, - "loss": 0.0718, + "grad_norm": 1.7566516399383545, + "learning_rate": 3.13e-05, + "loss": 0.0823, "step": 940 }, { - "grad_norm": 1.1665390729904175, - "learning_rate": 9.49e-05, - "loss": 0.08, + "grad_norm": 1.7244350910186768, + "learning_rate": 3.1633333333333334e-05, + "loss": 0.0904, "step": 950 }, { - "grad_norm": 1.1190357208251953, - "learning_rate": 9.59e-05, - "loss": 0.0757, + "grad_norm": 1.6203480958938599, + "learning_rate": 3.196666666666667e-05, + "loss": 0.0886, "step": 960 }, { - "grad_norm": 1.172494888305664, - "learning_rate": 9.69e-05, - "loss": 0.0768, + "grad_norm": 2.22615122795105, + "learning_rate": 3.2300000000000006e-05, + "loss": 0.0892, "step": 970 }, { - "grad_norm": 1.2636840343475342, - "learning_rate": 9.790000000000001e-05, - "loss": 0.0777, + "grad_norm": 2.079116106033325, + "learning_rate": 3.263333333333333e-05, + "loss": 0.0886, "step": 980 }, { - "grad_norm": 1.0782413482666016, - "learning_rate": 9.89e-05, - "loss": 0.0781, + "grad_norm": 1.7760223150253296, + "learning_rate": 3.296666666666667e-05, + "loss": 0.0894, "step": 990 }, { - "grad_norm": 1.1838284730911255, - "learning_rate": 9.99e-05, - "loss": 0.0818, + "grad_norm": 1.8316891193389893, + "learning_rate": 3.33e-05, + "loss": 0.0936, "step": 1000 }, { - "grad_norm": 1.178709626197815, - "learning_rate": 9.999994463727085e-05, - "loss": 0.0745, + "grad_norm": 1.8911479711532593, + "learning_rate": 3.3633333333333335e-05, + "loss": 0.0873, "step": 1010 }, { - "grad_norm": 1.0647282600402832, - "learning_rate": 9.999975326009292e-05, - "loss": 0.0764, + "grad_norm": 2.0155932903289795, + "learning_rate": 3.396666666666667e-05, + "loss": 0.0896, "step": 1020 }, { - "grad_norm": 1.127692461013794, - "learning_rate": 9.999942518549879e-05, - "loss": 0.0814, + "grad_norm": 1.7159051895141602, + "learning_rate": 3.430000000000001e-05, + "loss": 0.0939, "step": 1030 }, { - "grad_norm": 1.0108965635299683, - "learning_rate": 9.999896041438544e-05, - "loss": 0.0792, + "grad_norm": 1.8582090139389038, + "learning_rate": 3.463333333333333e-05, + "loss": 0.087, "step": 1040 }, { - "grad_norm": 0.9025714993476868, - "learning_rate": 9.999835894802353e-05, - "loss": 0.0722, + "grad_norm": 1.7398709058761597, + "learning_rate": 3.496666666666667e-05, + "loss": 0.0805, "step": 1050 }, { - "grad_norm": 1.064939022064209, - "learning_rate": 9.999762078805743e-05, - "loss": 0.0717, + "grad_norm": 1.660385251045227, + "learning_rate": 3.53e-05, + "loss": 0.0823, "step": 1060 }, { - "grad_norm": 1.1429728269577026, - "learning_rate": 9.999674593650526e-05, - "loss": 0.069, + "grad_norm": 1.6338180303573608, + "learning_rate": 3.563333333333334e-05, + "loss": 0.0803, "step": 1070 }, { - "grad_norm": 1.0148142576217651, - "learning_rate": 9.99957343957588e-05, - "loss": 0.0643, + "grad_norm": 1.898735523223877, + "learning_rate": 3.596666666666667e-05, + "loss": 0.0771, "step": 1080 }, { - "grad_norm": 0.9264249205589294, - "learning_rate": 9.99945861685836e-05, - "loss": 0.0679, + "grad_norm": 1.6656705141067505, + "learning_rate": 3.63e-05, + "loss": 0.0797, "step": 1090 }, { - "grad_norm": 1.1930052042007446, - "learning_rate": 9.999330125811884e-05, - "loss": 0.0737, + "grad_norm": 2.257603883743286, + "learning_rate": 3.6633333333333334e-05, + "loss": 0.085, "step": 1100 }, { - "grad_norm": 1.004508137702942, - "learning_rate": 9.999187966787744e-05, - "loss": 0.0643, + "grad_norm": 1.6739178895950317, + "learning_rate": 3.6966666666666666e-05, + "loss": 0.0746, "step": 1110 }, { - "grad_norm": 0.9657332897186279, - "learning_rate": 9.999032140174595e-05, - "loss": 0.0749, + "grad_norm": 1.684378981590271, + "learning_rate": 3.73e-05, + "loss": 0.0842, "step": 1120 }, { - "grad_norm": 0.8985906839370728, - "learning_rate": 9.998862646398464e-05, - "loss": 0.0746, + "grad_norm": 1.809085488319397, + "learning_rate": 3.763333333333334e-05, + "loss": 0.0831, "step": 1130 }, { - "grad_norm": 1.174508810043335, - "learning_rate": 9.998679485922739e-05, - "loss": 0.068, + "grad_norm": 1.951213002204895, + "learning_rate": 3.796666666666667e-05, + "loss": 0.0807, "step": 1140 }, { - "grad_norm": 1.1320881843566895, - "learning_rate": 9.998482659248174e-05, - "loss": 0.0749, + "grad_norm": 1.5780582427978516, + "learning_rate": 3.83e-05, + "loss": 0.086, "step": 1150 }, { - "grad_norm": 0.8892739415168762, - "learning_rate": 9.998272166912883e-05, - "loss": 0.065, + "grad_norm": 1.7449209690093994, + "learning_rate": 3.8633333333333335e-05, + "loss": 0.0762, "step": 1160 }, { - "grad_norm": 1.0125502347946167, - "learning_rate": 9.998048009492347e-05, - "loss": 0.0676, + "grad_norm": 1.9996048212051392, + "learning_rate": 3.896666666666667e-05, + "loss": 0.0846, "step": 1170 }, { - "grad_norm": 0.9409289956092834, - "learning_rate": 9.997810187599403e-05, - "loss": 0.0691, + "grad_norm": 1.4134843349456787, + "learning_rate": 3.9300000000000007e-05, + "loss": 0.0837, "step": 1180 }, { - "grad_norm": 1.0810723304748535, - "learning_rate": 9.997558701884249e-05, - "loss": 0.0672, + "grad_norm": 1.4462107419967651, + "learning_rate": 3.963333333333333e-05, + "loss": 0.0787, "step": 1190 }, { - "grad_norm": 0.9851813316345215, - "learning_rate": 9.997293553034433e-05, - "loss": 0.0711, + "grad_norm": 1.7750245332717896, + "learning_rate": 3.996666666666667e-05, + "loss": 0.0836, "step": 1200 }, { - "grad_norm": 0.9922870993614197, - "learning_rate": 9.997014741774866e-05, - "loss": 0.0646, + "grad_norm": 1.5897971391677856, + "learning_rate": 4.0300000000000004e-05, + "loss": 0.0752, "step": 1210 }, { - "grad_norm": 0.8688313364982605, - "learning_rate": 9.996722268867803e-05, - "loss": 0.0598, + "grad_norm": 1.4849587678909302, + "learning_rate": 4.0633333333333336e-05, + "loss": 0.0714, "step": 1220 }, { - "grad_norm": 0.7204305529594421, - "learning_rate": 9.996416135112858e-05, - "loss": 0.061, + "grad_norm": 1.1623440980911255, + "learning_rate": 4.096666666666667e-05, + "loss": 0.0743, "step": 1230 }, { - "grad_norm": 0.9210626482963562, - "learning_rate": 9.996096341346988e-05, - "loss": 0.0651, + "grad_norm": 1.7267963886260986, + "learning_rate": 4.13e-05, + "loss": 0.084, "step": 1240 }, { - "grad_norm": 0.894849419593811, - "learning_rate": 9.995762888444495e-05, - "loss": 0.0677, + "grad_norm": 1.3627574443817139, + "learning_rate": 4.1633333333333333e-05, + "loss": 0.0821, "step": 1250 }, { - "grad_norm": 0.8269813656806946, - "learning_rate": 9.995415777317027e-05, - "loss": 0.0586, + "grad_norm": 1.644944429397583, + "learning_rate": 4.196666666666667e-05, + "loss": 0.0707, "step": 1260 }, { - "grad_norm": 0.705187976360321, - "learning_rate": 9.995055008913574e-05, - "loss": 0.0716, + "grad_norm": 1.4931832551956177, + "learning_rate": 4.23e-05, + "loss": 0.085, "step": 1270 }, { - "grad_norm": 0.9853078722953796, - "learning_rate": 9.994680584220463e-05, - "loss": 0.0643, + "grad_norm": 1.327480435371399, + "learning_rate": 4.263333333333334e-05, + "loss": 0.0788, "step": 1280 }, { - "grad_norm": 0.8162582516670227, - "learning_rate": 9.994292504261355e-05, - "loss": 0.0662, + "grad_norm": 1.8330137729644775, + "learning_rate": 4.296666666666666e-05, + "loss": 0.0811, "step": 1290 }, { - "grad_norm": 0.975835919380188, - "learning_rate": 9.993890770097247e-05, - "loss": 0.0635, + "grad_norm": 1.6781295537948608, + "learning_rate": 4.33e-05, + "loss": 0.0784, "step": 1300 }, { - "grad_norm": 1.0302761793136597, - "learning_rate": 9.993475382826467e-05, - "loss": 0.0572, + "grad_norm": 1.6285682916641235, + "learning_rate": 4.3633333333333335e-05, + "loss": 0.0698, "step": 1310 }, { - "grad_norm": 0.7187492251396179, - "learning_rate": 9.993046343584664e-05, - "loss": 0.0697, + "grad_norm": 1.512878179550171, + "learning_rate": 4.396666666666667e-05, + "loss": 0.0841, "step": 1320 }, { - "grad_norm": 0.7942295670509338, - "learning_rate": 9.992603653544816e-05, - "loss": 0.0577, + "grad_norm": 1.525999903678894, + "learning_rate": 4.43e-05, + "loss": 0.0721, "step": 1330 }, { - "grad_norm": 0.8199202418327332, - "learning_rate": 9.992147313917222e-05, - "loss": 0.0529, + "grad_norm": 1.5048669576644897, + "learning_rate": 4.463333333333334e-05, + "loss": 0.0683, "step": 1340 }, { - "grad_norm": 0.8115194439888, - "learning_rate": 9.991677325949497e-05, - "loss": 0.0624, + "grad_norm": 1.3727238178253174, + "learning_rate": 4.496666666666667e-05, + "loss": 0.0772, "step": 1350 }, { - "grad_norm": 0.6720649600028992, - "learning_rate": 9.991193690926568e-05, - "loss": 0.0646, + "grad_norm": 1.4466592073440552, + "learning_rate": 4.53e-05, + "loss": 0.0815, "step": 1360 }, { - "grad_norm": 0.9936930537223816, - "learning_rate": 9.990696410170678e-05, - "loss": 0.0564, + "grad_norm": 1.4270600080490112, + "learning_rate": 4.5633333333333336e-05, + "loss": 0.0721, "step": 1370 }, { - "grad_norm": 0.7325035929679871, - "learning_rate": 9.990185485041371e-05, - "loss": 0.0564, + "grad_norm": 1.2492244243621826, + "learning_rate": 4.596666666666667e-05, + "loss": 0.0702, "step": 1380 }, { - "grad_norm": 0.7739620804786682, - "learning_rate": 9.989660916935498e-05, - "loss": 0.0522, + "grad_norm": 1.3557374477386475, + "learning_rate": 4.630000000000001e-05, + "loss": 0.0668, "step": 1390 }, { - "grad_norm": 0.6889415979385376, - "learning_rate": 9.989122707287208e-05, - "loss": 0.0494, + "grad_norm": 1.397212266921997, + "learning_rate": 4.663333333333333e-05, + "loss": 0.0652, "step": 1400 }, { - "grad_norm": 0.7498676776885986, - "learning_rate": 9.988570857567945e-05, - "loss": 0.0511, + "grad_norm": 1.514648675918579, + "learning_rate": 4.696666666666667e-05, + "loss": 0.0676, "step": 1410 }, { - "grad_norm": 0.7862585783004761, - "learning_rate": 9.988005369286446e-05, - "loss": 0.0514, + "grad_norm": 1.2394548654556274, + "learning_rate": 4.73e-05, + "loss": 0.0707, "step": 1420 }, { - "grad_norm": 0.86005699634552, - "learning_rate": 9.987426243988734e-05, - "loss": 0.0594, + "grad_norm": 1.5269253253936768, + "learning_rate": 4.763333333333334e-05, + "loss": 0.0804, "step": 1430 }, { - "grad_norm": 0.7341459393501282, - "learning_rate": 9.986833483258114e-05, - "loss": 0.0524, + "grad_norm": 1.300212025642395, + "learning_rate": 4.796666666666667e-05, + "loss": 0.0718, "step": 1440 }, { - "grad_norm": 0.6387414336204529, - "learning_rate": 9.986227088715173e-05, - "loss": 0.0502, + "grad_norm": 1.1953885555267334, + "learning_rate": 4.83e-05, + "loss": 0.0677, "step": 1450 }, { - "grad_norm": 0.6452542543411255, - "learning_rate": 9.98560706201777e-05, - "loss": 0.0552, + "grad_norm": 1.2278605699539185, + "learning_rate": 4.8633333333333334e-05, + "loss": 0.0732, "step": 1460 }, { - "grad_norm": 0.6964207291603088, - "learning_rate": 9.984973404861036e-05, - "loss": 0.053, + "grad_norm": 1.4869178533554077, + "learning_rate": 4.8966666666666667e-05, + "loss": 0.0659, "step": 1470 }, { - "grad_norm": 0.9187918305397034, - "learning_rate": 9.984326118977361e-05, - "loss": 0.0523, + "grad_norm": 1.6627095937728882, + "learning_rate": 4.93e-05, + "loss": 0.0665, "step": 1480 }, { - "grad_norm": 0.5924795269966125, - "learning_rate": 9.983665206136406e-05, - "loss": 0.0572, + "grad_norm": 1.6575415134429932, + "learning_rate": 4.963333333333334e-05, + "loss": 0.0716, "step": 1490 }, { - "grad_norm": 0.7473819851875305, - "learning_rate": 9.982990668145075e-05, - "loss": 0.0536, + "grad_norm": 1.3267849683761597, + "learning_rate": 4.996666666666667e-05, + "loss": 0.0676, "step": 1500 }, { - "grad_norm": 0.7378273606300354, - "learning_rate": 9.982302506847534e-05, - "loss": 0.0563, + "grad_norm": 1.5192252397537231, + "learning_rate": 5.03e-05, + "loss": 0.0719, "step": 1510 }, { - "grad_norm": 0.7275499105453491, - "learning_rate": 9.981600724125189e-05, - "loss": 0.054, + "grad_norm": 1.2676864862442017, + "learning_rate": 5.0633333333333335e-05, + "loss": 0.0687, "step": 1520 }, { - "grad_norm": 0.8484076261520386, - "learning_rate": 9.980885321896685e-05, - "loss": 0.0517, + "grad_norm": 1.3717687129974365, + "learning_rate": 5.0966666666666674e-05, + "loss": 0.0649, "step": 1530 }, { - "grad_norm": 0.7533107995986938, - "learning_rate": 9.980156302117905e-05, - "loss": 0.0555, + "grad_norm": 1.2322230339050293, + "learning_rate": 5.130000000000001e-05, + "loss": 0.0697, "step": 1540 }, { - "grad_norm": 0.7630091905593872, - "learning_rate": 9.979413666781963e-05, - "loss": 0.0537, + "grad_norm": 1.248760461807251, + "learning_rate": 5.163333333333333e-05, + "loss": 0.0665, "step": 1550 }, { - "grad_norm": 0.6897077560424805, - "learning_rate": 9.978657417919193e-05, - "loss": 0.0524, + "grad_norm": 1.2173956632614136, + "learning_rate": 5.196666666666667e-05, + "loss": 0.0676, "step": 1560 }, { - "grad_norm": 0.8400923013687134, - "learning_rate": 9.977887557597153e-05, - "loss": 0.0539, + "grad_norm": 1.379817008972168, + "learning_rate": 5.2300000000000004e-05, + "loss": 0.0686, "step": 1570 }, { - "grad_norm": 0.9766692519187927, - "learning_rate": 9.97710408792061e-05, - "loss": 0.0511, + "grad_norm": 1.5441510677337646, + "learning_rate": 5.2633333333333336e-05, + "loss": 0.0648, "step": 1580 }, { - "grad_norm": 0.8205817341804504, - "learning_rate": 9.976307011031542e-05, - "loss": 0.0521, + "grad_norm": 1.224876880645752, + "learning_rate": 5.296666666666666e-05, + "loss": 0.0643, "step": 1590 }, { - "grad_norm": 0.8941313624382019, - "learning_rate": 9.975496329109126e-05, - "loss": 0.0572, + "grad_norm": 1.2213428020477295, + "learning_rate": 5.330000000000001e-05, + "loss": 0.0721, "step": 1600 }, { - "grad_norm": 0.68316650390625, - "learning_rate": 9.974672044369732e-05, - "loss": 0.0495, + "grad_norm": 1.4989992380142212, + "learning_rate": 5.3633333333333334e-05, + "loss": 0.0666, "step": 1610 }, { - "grad_norm": 0.7722634077072144, - "learning_rate": 9.97383415906693e-05, - "loss": 0.048, + "grad_norm": 1.3955914974212646, + "learning_rate": 5.3966666666666666e-05, + "loss": 0.0659, "step": 1620 }, { - "grad_norm": 0.7082585692405701, - "learning_rate": 9.97298267549146e-05, - "loss": 0.0501, + "grad_norm": 1.2911821603775024, + "learning_rate": 5.4300000000000005e-05, + "loss": 0.0669, "step": 1630 }, { - "grad_norm": 0.9164435863494873, - "learning_rate": 9.972117595971249e-05, - "loss": 0.0517, + "grad_norm": 1.4699312448501587, + "learning_rate": 5.463333333333334e-05, + "loss": 0.0681, "step": 1640 }, { - "grad_norm": 0.6809993386268616, - "learning_rate": 9.971238922871391e-05, - "loss": 0.0471, + "grad_norm": 1.3759241104125977, + "learning_rate": 5.496666666666666e-05, + "loss": 0.0634, "step": 1650 }, { - "grad_norm": 0.5702314376831055, - "learning_rate": 9.970346658594142e-05, - "loss": 0.0449, + "grad_norm": 1.4657407999038696, + "learning_rate": 5.530000000000001e-05, + "loss": 0.063, "step": 1660 }, { - "grad_norm": 0.773526668548584, - "learning_rate": 9.969440805578923e-05, - "loss": 0.0498, + "grad_norm": 1.4505670070648193, + "learning_rate": 5.5633333333333335e-05, + "loss": 0.0694, "step": 1670 }, { - "grad_norm": 0.7197460532188416, - "learning_rate": 9.968521366302298e-05, - "loss": 0.0467, + "grad_norm": 1.1750441789627075, + "learning_rate": 5.596666666666667e-05, + "loss": 0.0645, "step": 1680 }, { - "grad_norm": 0.6625840067863464, - "learning_rate": 9.967588343277981e-05, - "loss": 0.045, + "grad_norm": 1.2160885334014893, + "learning_rate": 5.63e-05, + "loss": 0.0613, "step": 1690 }, { - "grad_norm": 0.8736388087272644, - "learning_rate": 9.966641739056818e-05, - "loss": 0.0545, + "grad_norm": 1.3564035892486572, + "learning_rate": 5.663333333333334e-05, + "loss": 0.073, "step": 1700 }, { - "grad_norm": 0.750792384147644, - "learning_rate": 9.965681556226793e-05, - "loss": 0.0473, + "grad_norm": 1.61124849319458, + "learning_rate": 5.696666666666667e-05, + "loss": 0.0643, "step": 1710 }, { - "grad_norm": 0.786637008190155, - "learning_rate": 9.964707797413006e-05, - "loss": 0.0507, + "grad_norm": 1.247754693031311, + "learning_rate": 5.73e-05, + "loss": 0.0692, "step": 1720 }, { - "grad_norm": 0.6412456631660461, - "learning_rate": 9.963720465277679e-05, - "loss": 0.0487, + "grad_norm": 1.3943097591400146, + "learning_rate": 5.7633333333333336e-05, + "loss": 0.0687, "step": 1730 }, { - "grad_norm": 0.7418937087059021, - "learning_rate": 9.96271956252014e-05, - "loss": 0.0466, + "grad_norm": 0.9708131551742554, + "learning_rate": 5.796666666666667e-05, + "loss": 0.0641, "step": 1740 }, { - "grad_norm": 0.654487669467926, - "learning_rate": 9.961705091876816e-05, - "loss": 0.0437, + "grad_norm": 1.123645544052124, + "learning_rate": 5.83e-05, + "loss": 0.0604, "step": 1750 }, { - "grad_norm": 0.5268855690956116, - "learning_rate": 9.960677056121235e-05, - "loss": 0.0425, + "grad_norm": 1.007624626159668, + "learning_rate": 5.863333333333334e-05, + "loss": 0.0587, "step": 1760 }, { - "grad_norm": 0.6897957921028137, - "learning_rate": 9.959635458064005e-05, - "loss": 0.0524, + "grad_norm": 1.0768353939056396, + "learning_rate": 5.896666666666667e-05, + "loss": 0.0704, "step": 1770 }, { - "grad_norm": 0.7622124552726746, - "learning_rate": 9.958580300552815e-05, - "loss": 0.0486, + "grad_norm": 1.0121560096740723, + "learning_rate": 5.93e-05, + "loss": 0.0667, "step": 1780 }, { - "grad_norm": 0.6903027296066284, - "learning_rate": 9.957511586472426e-05, - "loss": 0.0484, + "grad_norm": 1.04828679561615, + "learning_rate": 5.9633333333333344e-05, + "loss": 0.0649, "step": 1790 }, { - "grad_norm": 0.6610746383666992, - "learning_rate": 9.956429318744662e-05, - "loss": 0.0501, + "grad_norm": 1.2127963304519653, + "learning_rate": 5.996666666666667e-05, + "loss": 0.0674, "step": 1800 }, { - "grad_norm": 0.6257076263427734, - "learning_rate": 9.955333500328404e-05, - "loss": 0.0461, + "grad_norm": 1.0731089115142822, + "learning_rate": 6.03e-05, + "loss": 0.0611, "step": 1810 }, { - "grad_norm": 0.6198020577430725, - "learning_rate": 9.95422413421957e-05, - "loss": 0.0471, + "grad_norm": 1.4043693542480469, + "learning_rate": 6.063333333333333e-05, + "loss": 0.0637, "step": 1820 }, { - "grad_norm": 0.5837280750274658, - "learning_rate": 9.953101223451133e-05, - "loss": 0.0458, + "grad_norm": 1.1265013217926025, + "learning_rate": 6.0966666666666674e-05, + "loss": 0.0623, "step": 1830 }, { - "grad_norm": 0.5286070704460144, - "learning_rate": 9.951964771093085e-05, - "loss": 0.0439, + "grad_norm": 0.9502959847450256, + "learning_rate": 6.13e-05, + "loss": 0.0603, "step": 1840 }, { - "grad_norm": 0.700628399848938, - "learning_rate": 9.950814780252442e-05, - "loss": 0.0449, + "grad_norm": 1.1003267765045166, + "learning_rate": 6.163333333333333e-05, + "loss": 0.0608, "step": 1850 }, { - "grad_norm": 0.659744918346405, - "learning_rate": 9.949651254073236e-05, - "loss": 0.0422, + "grad_norm": 1.1360594034194946, + "learning_rate": 6.196666666666668e-05, + "loss": 0.0576, "step": 1860 }, { - "grad_norm": 0.6888746619224548, - "learning_rate": 9.948474195736504e-05, - "loss": 0.0437, + "grad_norm": 1.1677144765853882, + "learning_rate": 6.23e-05, + "loss": 0.0589, "step": 1870 }, { - "grad_norm": 0.7431710958480835, - "learning_rate": 9.947283608460277e-05, - "loss": 0.0444, + "grad_norm": 1.2256553173065186, + "learning_rate": 6.263333333333333e-05, + "loss": 0.0589, "step": 1880 }, { - "grad_norm": 0.5975768566131592, - "learning_rate": 9.946079495499577e-05, - "loss": 0.0428, + "grad_norm": 1.42429518699646, + "learning_rate": 6.296666666666667e-05, + "loss": 0.0573, "step": 1890 }, { - "grad_norm": 0.7167385816574097, - "learning_rate": 9.944861860146401e-05, - "loss": 0.0478, + "grad_norm": 1.1541980504989624, + "learning_rate": 6.330000000000001e-05, + "loss": 0.0656, "step": 1900 }, { - "grad_norm": 0.8195943236351013, - "learning_rate": 9.943630705729719e-05, - "loss": 0.0462, + "grad_norm": 1.4295943975448608, + "learning_rate": 6.363333333333334e-05, + "loss": 0.065, "step": 1910 }, { - "grad_norm": 0.6716237664222717, - "learning_rate": 9.942386035615459e-05, - "loss": 0.0403, + "grad_norm": 1.0897581577301025, + "learning_rate": 6.396666666666667e-05, + "loss": 0.0579, "step": 1920 }, { - "grad_norm": 0.609624981880188, - "learning_rate": 9.941127853206503e-05, - "loss": 0.0417, + "grad_norm": 1.1042475700378418, + "learning_rate": 6.43e-05, + "loss": 0.0593, "step": 1930 }, { - "grad_norm": 0.5453461408615112, - "learning_rate": 9.939856161942673e-05, - "loss": 0.0513, + "grad_norm": 0.8818796277046204, + "learning_rate": 6.463333333333334e-05, + "loss": 0.0703, "step": 1940 }, { - "grad_norm": 0.6557628512382507, - "learning_rate": 9.938570965300724e-05, - "loss": 0.0426, + "grad_norm": 1.3052315711975098, + "learning_rate": 6.496666666666667e-05, + "loss": 0.0585, "step": 1950 }, { - "grad_norm": 0.7052960991859436, - "learning_rate": 9.937272266794335e-05, - "loss": 0.0427, + "grad_norm": 1.0372258424758911, + "learning_rate": 6.53e-05, + "loss": 0.0589, "step": 1960 }, { - "grad_norm": 0.7427675724029541, - "learning_rate": 9.935960069974096e-05, - "loss": 0.0437, + "grad_norm": 1.0483773946762085, + "learning_rate": 6.563333333333333e-05, + "loss": 0.0598, "step": 1970 }, { - "grad_norm": 0.7904492616653442, - "learning_rate": 9.934634378427506e-05, - "loss": 0.0474, + "grad_norm": 1.1204289197921753, + "learning_rate": 6.596666666666667e-05, + "loss": 0.0634, "step": 1980 }, { - "grad_norm": 0.6603449583053589, - "learning_rate": 9.933295195778954e-05, - "loss": 0.041, + "grad_norm": 1.2750037908554077, + "learning_rate": 6.630000000000001e-05, + "loss": 0.0561, "step": 1990 }, { - "grad_norm": 0.6774188876152039, - "learning_rate": 9.931942525689715e-05, - "loss": 0.0419, + "grad_norm": 1.0252948999404907, + "learning_rate": 6.663333333333333e-05, + "loss": 0.0576, "step": 2000 }, { - "grad_norm": 0.8625200390815735, - "learning_rate": 9.930576371857936e-05, - "loss": 0.0433, + "grad_norm": 1.13008713722229, + "learning_rate": 6.696666666666666e-05, + "loss": 0.0593, "step": 2010 }, { - "grad_norm": 0.6208633184432983, - "learning_rate": 9.929196738018629e-05, - "loss": 0.0434, + "grad_norm": 0.8467022776603699, + "learning_rate": 6.730000000000001e-05, + "loss": 0.0589, "step": 2020 }, { - "grad_norm": 0.7620671391487122, - "learning_rate": 9.927803627943662e-05, - "loss": 0.0396, + "grad_norm": 1.1267976760864258, + "learning_rate": 6.763333333333334e-05, + "loss": 0.0542, "step": 2030 }, { - "grad_norm": 0.7727156281471252, - "learning_rate": 9.926397045441744e-05, - "loss": 0.0438, + "grad_norm": 1.2363266944885254, + "learning_rate": 6.796666666666666e-05, + "loss": 0.0604, "step": 2040 }, { - "grad_norm": 0.7673093676567078, - "learning_rate": 9.924976994358417e-05, - "loss": 0.0463, + "grad_norm": 0.9148569703102112, + "learning_rate": 6.83e-05, + "loss": 0.0618, "step": 2050 }, { - "grad_norm": 0.6097151041030884, - "learning_rate": 9.923543478576048e-05, - "loss": 0.0424, + "grad_norm": 1.1480687856674194, + "learning_rate": 6.863333333333334e-05, + "loss": 0.0583, "step": 2060 }, { - "grad_norm": 0.8347402215003967, - "learning_rate": 9.922096502013813e-05, - "loss": 0.0433, + "grad_norm": 1.2516300678253174, + "learning_rate": 6.896666666666667e-05, + "loss": 0.0597, "step": 2070 }, { - "grad_norm": 0.7491519451141357, - "learning_rate": 9.92063606862769e-05, - "loss": 0.0463, + "grad_norm": 1.0247350931167603, + "learning_rate": 6.93e-05, + "loss": 0.0614, "step": 2080 }, { - "grad_norm": 0.7769274711608887, - "learning_rate": 9.919162182410453e-05, - "loss": 0.0468, + "grad_norm": 0.9880638122558594, + "learning_rate": 6.963333333333334e-05, + "loss": 0.0626, "step": 2090 }, { - "grad_norm": 0.5936280488967896, - "learning_rate": 9.917674847391645e-05, - "loss": 0.0429, + "grad_norm": 1.1071597337722778, + "learning_rate": 6.996666666666667e-05, + "loss": 0.0567, "step": 2100 }, { - "grad_norm": 0.6929172873497009, - "learning_rate": 9.916174067637584e-05, - "loss": 0.0431, + "grad_norm": 1.1150048971176147, + "learning_rate": 7.03e-05, + "loss": 0.0572, "step": 2110 }, { - "grad_norm": 0.6833594441413879, - "learning_rate": 9.914659847251348e-05, - "loss": 0.0417, + "grad_norm": 0.9690060615539551, + "learning_rate": 7.063333333333333e-05, + "loss": 0.057, "step": 2120 }, { - "grad_norm": 0.7112385630607605, - "learning_rate": 9.913132190372753e-05, - "loss": 0.044, + "grad_norm": 1.1471099853515625, + "learning_rate": 7.096666666666667e-05, + "loss": 0.0619, "step": 2130 }, { - "grad_norm": 0.48244747519493103, - "learning_rate": 9.911591101178359e-05, - "loss": 0.0437, + "grad_norm": 0.8427419662475586, + "learning_rate": 7.13e-05, + "loss": 0.0607, "step": 2140 }, { - "grad_norm": 0.7037522792816162, - "learning_rate": 9.910036583881443e-05, - "loss": 0.0414, + "grad_norm": 0.9163537621498108, + "learning_rate": 7.163333333333334e-05, + "loss": 0.0556, "step": 2150 }, { - "grad_norm": 0.6704358458518982, - "learning_rate": 9.908468642731995e-05, - "loss": 0.0366, + "grad_norm": 0.9024648070335388, + "learning_rate": 7.196666666666668e-05, + "loss": 0.0492, "step": 2160 }, { - "grad_norm": 0.5901470184326172, - "learning_rate": 9.906887282016707e-05, - "loss": 0.0446, + "grad_norm": 0.9732857346534729, + "learning_rate": 7.23e-05, + "loss": 0.0583, "step": 2170 }, { - "grad_norm": 0.6559775471687317, - "learning_rate": 9.90529250605896e-05, - "loss": 0.0422, + "grad_norm": 1.0352048873901367, + "learning_rate": 7.263333333333334e-05, + "loss": 0.0569, "step": 2180 }, { - "grad_norm": 0.8128564357757568, - "learning_rate": 9.903684319218809e-05, - "loss": 0.0444, + "grad_norm": 0.9434832334518433, + "learning_rate": 7.296666666666667e-05, + "loss": 0.0593, "step": 2190 }, { - "grad_norm": 0.5437960624694824, - "learning_rate": 9.902062725892976e-05, - "loss": 0.044, + "grad_norm": 1.045331597328186, + "learning_rate": 7.33e-05, + "loss": 0.0536, "step": 2200 }, { - "grad_norm": 0.8856530785560608, - "learning_rate": 9.900427730514834e-05, - "loss": 0.0466, + "grad_norm": 1.0376135110855103, + "learning_rate": 7.363333333333334e-05, + "loss": 0.0538, "step": 2210 }, { - "grad_norm": 0.5385216474533081, - "learning_rate": 9.8987793375544e-05, - "loss": 0.0451, + "grad_norm": 0.9841770529747009, + "learning_rate": 7.396666666666667e-05, + "loss": 0.0554, "step": 2220 }, { - "grad_norm": 0.6559721231460571, - "learning_rate": 9.897117551518318e-05, - "loss": 0.0437, + "grad_norm": 0.9631847739219666, + "learning_rate": 7.43e-05, + "loss": 0.0547, "step": 2230 }, { - "grad_norm": 0.8500446081161499, - "learning_rate": 9.895442376949844e-05, - "loss": 0.0429, + "grad_norm": 1.181673526763916, + "learning_rate": 7.463333333333334e-05, + "loss": 0.0556, "step": 2240 }, { - "grad_norm": 0.538585364818573, - "learning_rate": 9.893753818428845e-05, - "loss": 0.0475, + "grad_norm": 1.006360411643982, + "learning_rate": 7.496666666666667e-05, + "loss": 0.0609, "step": 2250 }, { - "grad_norm": 0.608948826789856, - "learning_rate": 9.892051880571773e-05, - "loss": 0.0468, + "grad_norm": 1.2000733613967896, + "learning_rate": 7.53e-05, + "loss": 0.0627, "step": 2260 }, { - "grad_norm": 0.6505542993545532, - "learning_rate": 9.890336568031663e-05, - "loss": 0.0467, + "grad_norm": 0.965882420539856, + "learning_rate": 7.563333333333333e-05, + "loss": 0.0627, "step": 2270 }, { - "grad_norm": 0.5538124442100525, - "learning_rate": 9.888607885498113e-05, - "loss": 0.0409, + "grad_norm": 0.6869267821311951, + "learning_rate": 7.596666666666668e-05, + "loss": 0.0553, "step": 2280 }, { - "grad_norm": 0.6162983775138855, - "learning_rate": 9.886865837697275e-05, - "loss": 0.0385, + "grad_norm": 0.8393562436103821, + "learning_rate": 7.630000000000001e-05, + "loss": 0.0528, "step": 2290 }, { - "grad_norm": 0.5613241195678711, - "learning_rate": 9.88511042939184e-05, - "loss": 0.0403, + "grad_norm": 1.001491665840149, + "learning_rate": 7.663333333333333e-05, + "loss": 0.0559, "step": 2300 }, { - "grad_norm": 0.6898803114891052, - "learning_rate": 9.883341665381028e-05, - "loss": 0.0406, + "grad_norm": 0.9473748803138733, + "learning_rate": 7.696666666666668e-05, + "loss": 0.0555, "step": 2310 }, { - "grad_norm": 0.7443089485168457, - "learning_rate": 9.881559550500575e-05, - "loss": 0.04, + "grad_norm": 0.9534969925880432, + "learning_rate": 7.730000000000001e-05, + "loss": 0.054, "step": 2320 }, { - "grad_norm": 0.6974295377731323, - "learning_rate": 9.879764089622712e-05, - "loss": 0.0449, + "grad_norm": 0.9704079627990723, + "learning_rate": 7.763333333333334e-05, + "loss": 0.0593, "step": 2330 }, { - "grad_norm": 0.6060269474983215, - "learning_rate": 9.87795528765616e-05, - "loss": 0.0416, + "grad_norm": 0.875954806804657, + "learning_rate": 7.796666666666666e-05, + "loss": 0.0575, "step": 2340 }, { - "grad_norm": 0.6247705221176147, - "learning_rate": 9.876133149546118e-05, - "loss": 0.0404, + "grad_norm": 0.9904036521911621, + "learning_rate": 7.83e-05, + "loss": 0.0561, "step": 2350 }, { - "grad_norm": 0.5524375438690186, - "learning_rate": 9.874297680274238e-05, - "loss": 0.0429, + "grad_norm": 0.8696838021278381, + "learning_rate": 7.863333333333334e-05, + "loss": 0.0603, "step": 2360 }, { - "grad_norm": 0.6544674634933472, - "learning_rate": 9.872448884858624e-05, - "loss": 0.0419, + "grad_norm": 0.8562178611755371, + "learning_rate": 7.896666666666667e-05, + "loss": 0.0576, "step": 2370 }, { - "grad_norm": 0.7922749519348145, - "learning_rate": 9.870586768353815e-05, - "loss": 0.0403, + "grad_norm": 1.0243654251098633, + "learning_rate": 7.93e-05, + "loss": 0.0552, "step": 2380 }, { - "grad_norm": 0.5878010392189026, - "learning_rate": 9.868711335850764e-05, - "loss": 0.0413, + "grad_norm": 0.7713583111763, + "learning_rate": 7.963333333333334e-05, + "loss": 0.0552, "step": 2390 }, { - "grad_norm": 0.577335774898529, - "learning_rate": 9.866822592476833e-05, - "loss": 0.0403, + "grad_norm": 0.8817041516304016, + "learning_rate": 7.996666666666667e-05, + "loss": 0.0546, "step": 2400 }, { - "grad_norm": 0.6214291453361511, - "learning_rate": 9.86492054339577e-05, - "loss": 0.036, + "grad_norm": 0.9232308268547058, + "learning_rate": 8.030000000000001e-05, + "loss": 0.0494, "step": 2410 }, { - "grad_norm": 0.6292899250984192, - "learning_rate": 9.863005193807711e-05, - "loss": 0.0387, + "grad_norm": 1.128440260887146, + "learning_rate": 8.063333333333333e-05, + "loss": 0.0532, "step": 2420 }, { - "grad_norm": 0.6217591166496277, - "learning_rate": 9.861076548949143e-05, - "loss": 0.0386, + "grad_norm": 0.7509755492210388, + "learning_rate": 8.096666666666667e-05, + "loss": 0.0525, "step": 2430 }, { - "grad_norm": 0.642356812953949, - "learning_rate": 9.859134614092912e-05, - "loss": 0.0381, + "grad_norm": 0.9969519376754761, + "learning_rate": 8.13e-05, + "loss": 0.051, "step": 2440 }, { - "grad_norm": 0.615687906742096, - "learning_rate": 9.857179394548191e-05, - "loss": 0.0387, + "grad_norm": 0.7783822417259216, + "learning_rate": 8.163333333333334e-05, + "loss": 0.0519, "step": 2450 }, { - "grad_norm": 0.6285825967788696, - "learning_rate": 9.855210895660477e-05, - "loss": 0.0422, + "grad_norm": 1.017827033996582, + "learning_rate": 8.196666666666668e-05, + "loss": 0.0562, "step": 2460 }, { - "grad_norm": 0.5457389950752258, - "learning_rate": 9.853229122811568e-05, - "loss": 0.036, + "grad_norm": 0.8367951512336731, + "learning_rate": 8.23e-05, + "loss": 0.047, "step": 2470 }, { - "grad_norm": 0.7624576091766357, - "learning_rate": 9.851234081419559e-05, - "loss": 0.0456, + "grad_norm": 0.8985480070114136, + "learning_rate": 8.263333333333334e-05, + "loss": 0.0575, "step": 2480 }, { - "grad_norm": 0.6762944459915161, - "learning_rate": 9.849225776938814e-05, - "loss": 0.039, + "grad_norm": 0.9090971946716309, + "learning_rate": 8.296666666666667e-05, + "loss": 0.0517, "step": 2490 }, { - "grad_norm": 0.607412576675415, - "learning_rate": 9.847204214859964e-05, - "loss": 0.0407, + "grad_norm": 0.7974523901939392, + "learning_rate": 8.33e-05, + "loss": 0.0523, "step": 2500 }, { - "grad_norm": 0.5367524027824402, - "learning_rate": 9.845169400709879e-05, - "loss": 0.0409, + "grad_norm": 0.9519146680831909, + "learning_rate": 8.363333333333334e-05, + "loss": 0.0542, "step": 2510 }, { - "grad_norm": 0.6632624864578247, - "learning_rate": 9.843121340051664e-05, - "loss": 0.0377, + "grad_norm": 0.7397735118865967, + "learning_rate": 8.396666666666667e-05, + "loss": 0.0515, "step": 2520 }, { - "grad_norm": 0.6622194647789001, - "learning_rate": 9.841060038484641e-05, - "loss": 0.0369, + "grad_norm": 0.9653218984603882, + "learning_rate": 8.43e-05, + "loss": 0.0488, "step": 2530 }, { - "grad_norm": 0.4699934124946594, - "learning_rate": 9.838985501644328e-05, - "loss": 0.0346, + "grad_norm": 0.673168957233429, + "learning_rate": 8.463333333333335e-05, + "loss": 0.0474, "step": 2540 }, { - "grad_norm": 0.7941675186157227, - "learning_rate": 9.83689773520243e-05, - "loss": 0.0417, + "grad_norm": 0.7991240620613098, + "learning_rate": 8.496666666666667e-05, + "loss": 0.0538, "step": 2550 }, { - "grad_norm": 0.5687143802642822, - "learning_rate": 9.834796744866819e-05, - "loss": 0.0358, + "grad_norm": 0.8494511842727661, + "learning_rate": 8.53e-05, + "loss": 0.0466, "step": 2560 }, { - "grad_norm": 0.5689958333969116, - "learning_rate": 9.832682536381525e-05, - "loss": 0.0353, + "grad_norm": 0.9272388815879822, + "learning_rate": 8.563333333333333e-05, + "loss": 0.0473, "step": 2570 }, { - "grad_norm": 0.7229174971580505, - "learning_rate": 9.830555115526711e-05, - "loss": 0.0398, + "grad_norm": 0.8365880846977234, + "learning_rate": 8.596666666666668e-05, + "loss": 0.0538, "step": 2580 }, { - "grad_norm": 0.6610782742500305, - "learning_rate": 9.828414488118667e-05, - "loss": 0.0373, + "grad_norm": 0.7655223608016968, + "learning_rate": 8.63e-05, + "loss": 0.0499, "step": 2590 }, { - "grad_norm": 0.6783902049064636, - "learning_rate": 9.826260660009785e-05, - "loss": 0.0378, + "grad_norm": 0.6641157269477844, + "learning_rate": 8.663333333333333e-05, + "loss": 0.0503, "step": 2600 }, { - "grad_norm": 0.5838411450386047, - "learning_rate": 9.824093637088547e-05, - "loss": 0.0374, + "grad_norm": 0.8131173849105835, + "learning_rate": 8.696666666666668e-05, + "loss": 0.0503, "step": 2610 }, { - "grad_norm": 0.5158525109291077, - "learning_rate": 9.821913425279514e-05, - "loss": 0.0346, + "grad_norm": 0.916050374507904, + "learning_rate": 8.730000000000001e-05, + "loss": 0.0476, "step": 2620 }, { - "grad_norm": 0.5010576248168945, - "learning_rate": 9.8197200305433e-05, - "loss": 0.0367, + "grad_norm": 0.7481029629707336, + "learning_rate": 8.763333333333334e-05, + "loss": 0.0505, "step": 2630 }, { - "grad_norm": 0.5873668193817139, - "learning_rate": 9.817513458876564e-05, - "loss": 0.0389, + "grad_norm": 0.817760705947876, + "learning_rate": 8.796666666666667e-05, + "loss": 0.0494, "step": 2640 }, { - "grad_norm": 0.669737696647644, - "learning_rate": 9.815293716311987e-05, - "loss": 0.0391, + "grad_norm": 0.8094144463539124, + "learning_rate": 8.83e-05, + "loss": 0.0501, "step": 2650 }, { - "grad_norm": 0.6051730513572693, - "learning_rate": 9.813060808918262e-05, - "loss": 0.0358, + "grad_norm": 0.9813537001609802, + "learning_rate": 8.863333333333334e-05, + "loss": 0.0477, "step": 2660 }, { - "grad_norm": 0.6769428253173828, - "learning_rate": 9.810814742800069e-05, - "loss": 0.0376, + "grad_norm": 0.6867929697036743, + "learning_rate": 8.896666666666667e-05, + "loss": 0.0512, "step": 2670 }, { - "grad_norm": 0.5549777150154114, - "learning_rate": 9.808555524098074e-05, - "loss": 0.035, + "grad_norm": 0.9549563527107239, + "learning_rate": 8.93e-05, + "loss": 0.0481, "step": 2680 }, { - "grad_norm": 0.6324145197868347, - "learning_rate": 9.806283158988887e-05, - "loss": 0.037, + "grad_norm": 0.7506797909736633, + "learning_rate": 8.963333333333333e-05, + "loss": 0.0494, "step": 2690 }, { - "grad_norm": 0.522504448890686, - "learning_rate": 9.803997653685072e-05, - "loss": 0.0353, + "grad_norm": 0.8433989882469177, + "learning_rate": 8.996666666666667e-05, + "loss": 0.049, "step": 2700 }, { - "grad_norm": 0.6257413029670715, - "learning_rate": 9.801699014435112e-05, - "loss": 0.0398, + "grad_norm": 0.8604899644851685, + "learning_rate": 9.030000000000001e-05, + "loss": 0.0532, "step": 2710 }, { - "grad_norm": 0.6243777871131897, - "learning_rate": 9.799387247523398e-05, - "loss": 0.0329, + "grad_norm": 0.7850534915924072, + "learning_rate": 9.063333333333333e-05, + "loss": 0.0455, "step": 2720 }, { - "grad_norm": 0.5824418663978577, - "learning_rate": 9.797062359270215e-05, - "loss": 0.0368, + "grad_norm": 0.7254740595817566, + "learning_rate": 9.096666666666666e-05, + "loss": 0.0494, "step": 2730 }, { - "grad_norm": 0.5653295516967773, - "learning_rate": 9.794724356031715e-05, - "loss": 0.038, + "grad_norm": 0.7519999146461487, + "learning_rate": 9.130000000000001e-05, + "loss": 0.0493, "step": 2740 }, { - "grad_norm": 0.5860871076583862, - "learning_rate": 9.792373244199913e-05, - "loss": 0.0366, + "grad_norm": 0.8911005258560181, + "learning_rate": 9.163333333333334e-05, + "loss": 0.0487, "step": 2750 }, { - "grad_norm": 0.6080830693244934, - "learning_rate": 9.790009030202658e-05, - "loss": 0.0391, + "grad_norm": 0.896418571472168, + "learning_rate": 9.196666666666666e-05, + "loss": 0.0514, "step": 2760 }, { - "grad_norm": 0.5480026602745056, - "learning_rate": 9.78763172050362e-05, - "loss": 0.0342, + "grad_norm": 0.7974010109901428, + "learning_rate": 9.230000000000001e-05, + "loss": 0.0445, "step": 2770 }, { - "grad_norm": 0.4915301203727722, - "learning_rate": 9.785241321602274e-05, - "loss": 0.0382, + "grad_norm": 0.6575005054473877, + "learning_rate": 9.263333333333334e-05, + "loss": 0.0492, "step": 2780 }, { - "grad_norm": 0.5721039772033691, - "learning_rate": 9.782837840033879e-05, - "loss": 0.0387, + "grad_norm": 0.7433968782424927, + "learning_rate": 9.296666666666667e-05, + "loss": 0.0504, "step": 2790 }, { - "grad_norm": 0.6181113123893738, - "learning_rate": 9.780421282369461e-05, - "loss": 0.0343, + "grad_norm": 0.7195668816566467, + "learning_rate": 9.33e-05, + "loss": 0.0458, "step": 2800 }, { - "grad_norm": 0.5936803817749023, - "learning_rate": 9.777991655215797e-05, - "loss": 0.0325, + "grad_norm": 0.8323925733566284, + "learning_rate": 9.363333333333334e-05, + "loss": 0.0436, "step": 2810 }, { - "grad_norm": 0.6910586357116699, - "learning_rate": 9.775548965215394e-05, - "loss": 0.0314, + "grad_norm": 0.7287455797195435, + "learning_rate": 9.396666666666667e-05, + "loss": 0.0429, "step": 2820 }, { - "grad_norm": 0.5371761918067932, - "learning_rate": 9.773093219046474e-05, - "loss": 0.036, + "grad_norm": 0.8642809987068176, + "learning_rate": 9.43e-05, + "loss": 0.0489, "step": 2830 }, { - "grad_norm": 0.4973508417606354, - "learning_rate": 9.770624423422954e-05, - "loss": 0.0376, + "grad_norm": 0.8033154606819153, + "learning_rate": 9.463333333333333e-05, + "loss": 0.0501, "step": 2840 }, { - "grad_norm": 0.6510023474693298, - "learning_rate": 9.768142585094426e-05, - "loss": 0.0353, + "grad_norm": 0.67922443151474, + "learning_rate": 9.496666666666667e-05, + "loss": 0.0468, "step": 2850 }, { - "grad_norm": 0.6040503978729248, - "learning_rate": 9.765647710846142e-05, - "loss": 0.0347, + "grad_norm": 0.7234774231910706, + "learning_rate": 9.53e-05, + "loss": 0.0449, "step": 2860 }, { - "grad_norm": 0.5976170301437378, - "learning_rate": 9.763139807498991e-05, - "loss": 0.034, + "grad_norm": 0.7744899392127991, + "learning_rate": 9.563333333333334e-05, + "loss": 0.0439, "step": 2870 }, { - "grad_norm": 0.6797834038734436, - "learning_rate": 9.760618881909487e-05, - "loss": 0.0339, + "grad_norm": 0.8154613375663757, + "learning_rate": 9.596666666666668e-05, + "loss": 0.044, "step": 2880 }, { - "grad_norm": 0.5329861044883728, - "learning_rate": 9.758084940969744e-05, - "loss": 0.0317, + "grad_norm": 0.7204713821411133, + "learning_rate": 9.63e-05, + "loss": 0.0428, "step": 2890 }, { - "grad_norm": 0.6726340055465698, - "learning_rate": 9.755537991607459e-05, - "loss": 0.0319, + "grad_norm": 0.8785443305969238, + "learning_rate": 9.663333333333334e-05, + "loss": 0.0428, "step": 2900 }, { - "grad_norm": 0.608772873878479, - "learning_rate": 9.752978040785895e-05, - "loss": 0.0313, + "grad_norm": 0.849808931350708, + "learning_rate": 9.696666666666667e-05, + "loss": 0.0449, "step": 2910 }, { - "grad_norm": 0.7140541076660156, - "learning_rate": 9.750405095503859e-05, - "loss": 0.0311, + "grad_norm": 0.8286228775978088, + "learning_rate": 9.730000000000001e-05, + "loss": 0.0433, "step": 2920 }, { - "grad_norm": 0.5499521493911743, - "learning_rate": 9.747819162795686e-05, - "loss": 0.0367, + "grad_norm": 0.6806231737136841, + "learning_rate": 9.763333333333334e-05, + "loss": 0.0496, "step": 2930 }, { - "grad_norm": 0.6744058728218079, - "learning_rate": 9.745220249731217e-05, - "loss": 0.0407, + "grad_norm": 0.6595783233642578, + "learning_rate": 9.796666666666667e-05, + "loss": 0.0501, "step": 2940 }, { - "grad_norm": 0.7178553342819214, - "learning_rate": 9.742608363415781e-05, - "loss": 0.041, + "grad_norm": 0.8414114117622375, + "learning_rate": 9.83e-05, + "loss": 0.0576, "step": 2950 }, { - "grad_norm": 0.6183580160140991, - "learning_rate": 9.739983510990176e-05, - "loss": 0.0394, + "grad_norm": 0.7837775349617004, + "learning_rate": 9.863333333333334e-05, + "loss": 0.0589, "step": 2960 }, { - "grad_norm": 0.553803026676178, - "learning_rate": 9.737345699630647e-05, - "loss": 0.0329, + "grad_norm": 0.830162525177002, + "learning_rate": 9.896666666666667e-05, + "loss": 0.0487, "step": 2970 }, { - "grad_norm": 0.4599069654941559, - "learning_rate": 9.734694936548869e-05, - "loss": 0.0313, + "grad_norm": 0.6498530507087708, + "learning_rate": 9.93e-05, + "loss": 0.0458, "step": 2980 }, { - "grad_norm": 0.5753762125968933, - "learning_rate": 9.732031228991932e-05, - "loss": 0.0304, + "grad_norm": 0.624728798866272, + "learning_rate": 9.963333333333333e-05, + "loss": 0.0443, "step": 2990 }, { - "grad_norm": 0.6017200946807861, - "learning_rate": 9.729354584242302e-05, - "loss": 0.0343, + "grad_norm": 0.878516435623169, + "learning_rate": 9.996666666666668e-05, + "loss": 0.0484, "step": 3000 }, { - "grad_norm": 0.6461474299430847, - "learning_rate": 9.726665009617832e-05, - "loss": 0.0379, + "grad_norm": 0.7892113924026489, + "learning_rate": 9.999999384858465e-05, + "loss": 0.0531, "step": 3010 }, { - "grad_norm": 0.5760048627853394, - "learning_rate": 9.723962512471714e-05, - "loss": 0.0342, + "grad_norm": 0.8021721243858337, + "learning_rate": 9.999997258443473e-05, + "loss": 0.0465, "step": 3020 }, { - "grad_norm": 0.621891438961029, - "learning_rate": 9.72124710019247e-05, - "loss": 0.0316, + "grad_norm": 0.7913960218429565, + "learning_rate": 9.999993613161331e-05, + "loss": 0.0446, "step": 3030 }, { - "grad_norm": 0.524480402469635, - "learning_rate": 9.718518780203934e-05, - "loss": 0.0292, + "grad_norm": 0.6336753964424133, + "learning_rate": 9.999988449013146e-05, + "loss": 0.042, "step": 3040 }, { - "grad_norm": 0.6318420171737671, - "learning_rate": 9.715777559965228e-05, - "loss": 0.0316, + "grad_norm": 0.683625340461731, + "learning_rate": 9.99998176600049e-05, + "loss": 0.0439, "step": 3050 }, { - "grad_norm": 0.566210150718689, - "learning_rate": 9.713023446970746e-05, - "loss": 0.0309, + "grad_norm": 0.7004140019416809, + "learning_rate": 9.999973564125389e-05, + "loss": 0.0442, "step": 3060 }, { - "grad_norm": 0.704804003238678, - "learning_rate": 9.710256448750126e-05, - "loss": 0.0332, + "grad_norm": 0.8021495342254639, + "learning_rate": 9.999963843390335e-05, + "loss": 0.0466, "step": 3070 }, { - "grad_norm": 0.7099645733833313, - "learning_rate": 9.707476572868235e-05, - "loss": 0.0333, + "grad_norm": 0.6171932816505432, + "learning_rate": 9.999952603798282e-05, + "loss": 0.0455, "step": 3080 }, { - "grad_norm": 0.6047917604446411, - "learning_rate": 9.704683826925149e-05, - "loss": 0.0314, + "grad_norm": 0.7242380380630493, + "learning_rate": 9.999939845352646e-05, + "loss": 0.0442, "step": 3090 }, { - "grad_norm": 0.7446325421333313, - "learning_rate": 9.701878218556129e-05, - "loss": 0.035, + "grad_norm": 0.7642882466316223, + "learning_rate": 9.999925568057298e-05, + "loss": 0.0478, "step": 3100 }, { - "grad_norm": 0.6377901434898376, - "learning_rate": 9.699059755431598e-05, - "loss": 0.0387, + "grad_norm": 0.8305861949920654, + "learning_rate": 9.999909771916578e-05, + "loss": 0.0511, "step": 3110 }, { - "grad_norm": 0.5415878295898438, - "learning_rate": 9.696228445257132e-05, - "loss": 0.0299, + "grad_norm": 0.616356372833252, + "learning_rate": 9.999892456935285e-05, + "loss": 0.0431, "step": 3120 }, { - "grad_norm": 0.6148945093154907, - "learning_rate": 9.693384295773419e-05, - "loss": 0.0391, + "grad_norm": 0.7295675277709961, + "learning_rate": 9.999873623118679e-05, + "loss": 0.0532, "step": 3130 }, { - "grad_norm": 0.578273594379425, - "learning_rate": 9.690527314756259e-05, - "loss": 0.0305, + "grad_norm": 0.7463228702545166, + "learning_rate": 9.999853270472479e-05, + "loss": 0.0433, "step": 3140 }, { - "grad_norm": 0.5767846703529358, - "learning_rate": 9.687657510016527e-05, - "loss": 0.0327, + "grad_norm": 0.7920631766319275, + "learning_rate": 9.999831399002871e-05, + "loss": 0.0467, "step": 3150 }, { - "grad_norm": 0.532577633857727, - "learning_rate": 9.684774889400161e-05, - "loss": 0.0276, + "grad_norm": 0.7888407707214355, + "learning_rate": 9.999808008716494e-05, + "loss": 0.0403, "step": 3160 }, { - "grad_norm": 0.5682797431945801, - "learning_rate": 9.681879460788135e-05, - "loss": 0.0295, + "grad_norm": 0.7514359951019287, + "learning_rate": 9.999783099620459e-05, + "loss": 0.0434, "step": 3170 }, { - "grad_norm": 0.6185994744300842, - "learning_rate": 9.67897123209644e-05, - "loss": 0.0306, + "grad_norm": 0.6265234351158142, + "learning_rate": 9.999756671722328e-05, + "loss": 0.0438, "step": 3180 }, { - "grad_norm": 0.6122663617134094, - "learning_rate": 9.676050211276062e-05, - "loss": 0.0323, + "grad_norm": 0.7002662420272827, + "learning_rate": 9.99972872503013e-05, + "loss": 0.0444, "step": 3190 }, { - "grad_norm": 0.6145076751708984, - "learning_rate": 9.673116406312962e-05, - "loss": 0.0289, + "grad_norm": 0.770563006401062, + "learning_rate": 9.999699259552359e-05, + "loss": 0.0407, "step": 3200 }, { - "grad_norm": 0.7365098595619202, - "learning_rate": 9.67016982522805e-05, - "loss": 0.0264, + "grad_norm": 0.8374990820884705, + "learning_rate": 9.99966827529796e-05, + "loss": 0.0393, "step": 3210 }, { - "grad_norm": 0.6660006642341614, - "learning_rate": 9.667210476077164e-05, - "loss": 0.0307, + "grad_norm": 0.8018321394920349, + "learning_rate": 9.999635772276348e-05, + "loss": 0.0442, "step": 3220 }, { - "grad_norm": 0.43615856766700745, - "learning_rate": 9.664238366951055e-05, - "loss": 0.0347, + "grad_norm": 0.5608826279640198, + "learning_rate": 9.999601750497396e-05, + "loss": 0.0481, "step": 3230 }, { - "grad_norm": 0.5423561334609985, - "learning_rate": 9.661253505975355e-05, - "loss": 0.0286, + "grad_norm": 0.6996321678161621, + "learning_rate": 9.99956620997144e-05, + "loss": 0.0416, "step": 3240 }, { - "grad_norm": 0.6123466491699219, - "learning_rate": 9.658255901310557e-05, - "loss": 0.0297, + "grad_norm": 0.7144750952720642, + "learning_rate": 9.999529150709275e-05, + "loss": 0.0431, "step": 3250 }, { - "grad_norm": 0.5747650861740112, - "learning_rate": 9.655245561152e-05, - "loss": 0.0311, + "grad_norm": 0.6433903574943542, + "learning_rate": 9.999490572722158e-05, + "loss": 0.0451, "step": 3260 }, { - "grad_norm": 0.552060604095459, - "learning_rate": 9.65222249372984e-05, - "loss": 0.0287, + "grad_norm": 0.740658164024353, + "learning_rate": 9.99945047602181e-05, + "loss": 0.0429, "step": 3270 }, { - "grad_norm": 0.5556790828704834, - "learning_rate": 9.649186707309026e-05, - "loss": 0.0274, + "grad_norm": 0.6363276839256287, + "learning_rate": 9.99940886062041e-05, + "loss": 0.0405, "step": 3280 }, { - "grad_norm": 0.5429115295410156, - "learning_rate": 9.646138210189283e-05, - "loss": 0.0297, + "grad_norm": 0.6644964814186096, + "learning_rate": 9.999365726530599e-05, + "loss": 0.0423, "step": 3290 }, { - "grad_norm": 0.6687198281288147, - "learning_rate": 9.643077010705087e-05, - "loss": 0.029, + "grad_norm": 0.6375532746315002, + "learning_rate": 9.999321073765481e-05, + "loss": 0.0425, "step": 3300 }, { - "grad_norm": 0.5589491724967957, - "learning_rate": 9.640003117225637e-05, - "loss": 0.0296, + "grad_norm": 0.620246946811676, + "learning_rate": 9.99927490233862e-05, + "loss": 0.0428, "step": 3310 }, { - "grad_norm": 0.5250815749168396, - "learning_rate": 9.636916538154846e-05, - "loss": 0.0313, + "grad_norm": 0.7844054102897644, + "learning_rate": 9.999227212264043e-05, + "loss": 0.0453, "step": 3320 }, { - "grad_norm": 0.5884456038475037, - "learning_rate": 9.633817281931296e-05, - "loss": 0.0279, + "grad_norm": 0.6290953755378723, + "learning_rate": 9.999178003556236e-05, + "loss": 0.0407, "step": 3330 }, { - "grad_norm": 0.5253452658653259, - "learning_rate": 9.630705357028242e-05, - "loss": 0.0297, + "grad_norm": 0.6946619749069214, + "learning_rate": 9.999127276230146e-05, + "loss": 0.0421, "step": 3340 }, { - "grad_norm": 0.44990262389183044, - "learning_rate": 9.627580771953563e-05, - "loss": 0.0293, + "grad_norm": 0.6346951723098755, + "learning_rate": 9.999075030301184e-05, + "loss": 0.0422, "step": 3350 }, { - "grad_norm": 0.5788272023200989, - "learning_rate": 9.624443535249759e-05, - "loss": 0.028, + "grad_norm": 0.6657125353813171, + "learning_rate": 9.999021265785221e-05, + "loss": 0.0401, "step": 3360 }, { - "grad_norm": 0.4931959807872772, - "learning_rate": 9.621293655493913e-05, - "loss": 0.0282, + "grad_norm": 0.4986710548400879, + "learning_rate": 9.998965982698589e-05, + "loss": 0.0399, "step": 3370 }, { - "grad_norm": 0.49001696705818176, - "learning_rate": 9.618131141297675e-05, - "loss": 0.025, + "grad_norm": 0.6020874381065369, + "learning_rate": 9.998909181058082e-05, + "loss": 0.0376, "step": 3380 }, { - "grad_norm": 0.6005767583847046, - "learning_rate": 9.614956001307242e-05, - "loss": 0.0282, + "grad_norm": 0.6792619228363037, + "learning_rate": 9.998850860880953e-05, + "loss": 0.0412, "step": 3390 }, { - "grad_norm": 0.5819389224052429, - "learning_rate": 9.611768244203321e-05, - "loss": 0.0287, + "grad_norm": 0.8668169379234314, + "learning_rate": 9.998791022184922e-05, + "loss": 0.0421, "step": 3400 }, { - "grad_norm": 0.5611231923103333, - "learning_rate": 9.60856787870112e-05, - "loss": 0.0315, + "grad_norm": 0.6989952921867371, + "learning_rate": 9.99872966498816e-05, + "loss": 0.0458, "step": 3410 }, { - "grad_norm": 0.49898526072502136, - "learning_rate": 9.605354913550318e-05, - "loss": 0.0267, + "grad_norm": 0.5354453921318054, + "learning_rate": 9.998666789309313e-05, + "loss": 0.04, "step": 3420 }, { - "grad_norm": 0.5152934789657593, - "learning_rate": 9.602129357535037e-05, - "loss": 0.0248, + "grad_norm": 0.5989109873771667, + "learning_rate": 9.998602395167475e-05, + "loss": 0.0372, "step": 3430 }, { - "grad_norm": 0.4927721917629242, - "learning_rate": 9.598891219473825e-05, - "loss": 0.0272, + "grad_norm": 0.620912492275238, + "learning_rate": 9.998536482582213e-05, + "loss": 0.0412, "step": 3440 }, { - "grad_norm": 0.5922147035598755, - "learning_rate": 9.595640508219625e-05, - "loss": 0.0225, + "grad_norm": 0.7610539793968201, + "learning_rate": 9.998469051573544e-05, + "loss": 0.0356, "step": 3450 }, { - "grad_norm": 0.6173586845397949, - "learning_rate": 9.592377232659761e-05, - "loss": 0.0269, + "grad_norm": 0.7006726264953613, + "learning_rate": 9.998400102161954e-05, + "loss": 0.0404, "step": 3460 }, { - "grad_norm": 0.6940552592277527, - "learning_rate": 9.589101401715904e-05, - "loss": 0.0314, + "grad_norm": 0.7211143970489502, + "learning_rate": 9.998329634368388e-05, + "loss": 0.0461, "step": 3470 }, { - "grad_norm": 0.5696406364440918, - "learning_rate": 9.585813024344045e-05, - "loss": 0.0271, + "grad_norm": 0.8158736228942871, + "learning_rate": 9.998257648214253e-05, + "loss": 0.0409, "step": 3480 }, { - "grad_norm": 0.5202314853668213, - "learning_rate": 9.58251210953449e-05, - "loss": 0.0247, + "grad_norm": 0.5517433881759644, + "learning_rate": 9.998184143721417e-05, + "loss": 0.0382, "step": 3490 }, { - "grad_norm": 0.5222875475883484, - "learning_rate": 9.579198666311809e-05, - "loss": 0.0238, + "grad_norm": 0.6554803252220154, + "learning_rate": 9.998109120912206e-05, + "loss": 0.0373, "step": 3500 }, { - "grad_norm": 0.6505278944969177, - "learning_rate": 9.575872703734832e-05, - "loss": 0.0295, + "grad_norm": 0.6897868514060974, + "learning_rate": 9.998032579809411e-05, + "loss": 0.0427, "step": 3510 }, { - "grad_norm": 0.619454026222229, - "learning_rate": 9.572534230896611e-05, - "loss": 0.0268, + "grad_norm": 0.6962120532989502, + "learning_rate": 9.997954520436286e-05, + "loss": 0.0397, "step": 3520 }, { - "grad_norm": 0.48343542218208313, - "learning_rate": 9.569183256924403e-05, - "loss": 0.0269, + "grad_norm": 0.6989943981170654, + "learning_rate": 9.997874942816538e-05, + "loss": 0.0404, "step": 3530 }, { - "grad_norm": 0.4461105763912201, - "learning_rate": 9.565819790979646e-05, - "loss": 0.0255, + "grad_norm": 0.6061151623725891, + "learning_rate": 9.997793846974345e-05, + "loss": 0.0389, "step": 3540 }, { - "grad_norm": 0.5143640637397766, - "learning_rate": 9.562443842257925e-05, - "loss": 0.0257, + "grad_norm": 0.5280351638793945, + "learning_rate": 9.997711232934341e-05, + "loss": 0.0383, "step": 3550 }, { - "grad_norm": 0.5106943249702454, - "learning_rate": 9.559055419988956e-05, - "loss": 0.0265, + "grad_norm": 0.728611946105957, + "learning_rate": 9.99762710072162e-05, + "loss": 0.0406, "step": 3560 }, { - "grad_norm": 0.6030021905899048, - "learning_rate": 9.555654533436557e-05, - "loss": 0.0269, + "grad_norm": 0.7230589389801025, + "learning_rate": 9.997541450361743e-05, + "loss": 0.0413, "step": 3570 }, { - "grad_norm": 0.5566527843475342, - "learning_rate": 9.552241191898621e-05, - "loss": 0.0279, + "grad_norm": 0.7686134576797485, + "learning_rate": 9.997454281880723e-05, + "loss": 0.042, "step": 3580 }, { - "grad_norm": 0.4313097894191742, - "learning_rate": 9.548815404707092e-05, - "loss": 0.0246, + "grad_norm": 0.5007238984107971, + "learning_rate": 9.997365595305044e-05, + "loss": 0.0371, "step": 3590 }, { - "grad_norm": 0.6710801720619202, - "learning_rate": 9.545377181227942e-05, - "loss": 0.0242, + "grad_norm": 0.6437061429023743, + "learning_rate": 9.997275390661644e-05, + "loss": 0.0365, "step": 3600 }, { - "grad_norm": 0.4369684159755707, - "learning_rate": 9.541926530861145e-05, - "loss": 0.0238, + "grad_norm": 0.5456061363220215, + "learning_rate": 9.997183667977926e-05, + "loss": 0.0364, "step": 3610 }, { - "grad_norm": 0.48853105306625366, - "learning_rate": 9.538463463040645e-05, - "loss": 0.0226, + "grad_norm": 0.6077585220336914, + "learning_rate": 9.997090427281752e-05, + "loss": 0.0349, "step": 3620 }, { - "grad_norm": 0.5221312642097473, - "learning_rate": 9.534987987234337e-05, - "loss": 0.0267, + "grad_norm": 0.7311477065086365, + "learning_rate": 9.996995668601448e-05, + "loss": 0.0396, "step": 3630 }, { - "grad_norm": 0.5919408798217773, - "learning_rate": 9.53150011294404e-05, - "loss": 0.0234, + "grad_norm": 0.6024607419967651, + "learning_rate": 9.996899391965798e-05, + "loss": 0.0357, "step": 3640 }, { - "grad_norm": 0.5544537305831909, - "learning_rate": 9.527999849705471e-05, - "loss": 0.0229, + "grad_norm": 0.673121988773346, + "learning_rate": 9.996801597404048e-05, + "loss": 0.0357, "step": 3650 }, { - "grad_norm": 0.5028519034385681, - "learning_rate": 9.524487207088213e-05, - "loss": 0.0229, + "grad_norm": 0.5982141494750977, + "learning_rate": 9.996702284945905e-05, + "loss": 0.0382, "step": 3660 }, { - "grad_norm": 0.5428268313407898, - "learning_rate": 9.520962194695698e-05, - "loss": 0.0234, + "grad_norm": 0.512007474899292, + "learning_rate": 9.996601454621539e-05, + "loss": 0.0369, "step": 3670 }, { - "grad_norm": 0.4685244858264923, - "learning_rate": 9.517424822165175e-05, - "loss": 0.023, + "grad_norm": 0.49710696935653687, + "learning_rate": 9.996499106461577e-05, + "loss": 0.0364, "step": 3680 }, { - "grad_norm": 0.6033672094345093, - "learning_rate": 9.513875099167685e-05, - "loss": 0.028, + "grad_norm": 0.7000653743743896, + "learning_rate": 9.996395240497112e-05, + "loss": 0.0429, "step": 3690 }, { - "grad_norm": 0.5230802297592163, - "learning_rate": 9.510313035408035e-05, - "loss": 0.0211, + "grad_norm": 0.5541830062866211, + "learning_rate": 9.996289856759696e-05, + "loss": 0.0352, "step": 3700 }, { - "grad_norm": 0.5911836624145508, - "learning_rate": 9.506738640624775e-05, - "loss": 0.0226, + "grad_norm": 0.7469106912612915, + "learning_rate": 9.996182955281342e-05, + "loss": 0.0377, "step": 3710 }, { - "grad_norm": 0.4973941445350647, - "learning_rate": 9.50315192459016e-05, - "loss": 0.0233, + "grad_norm": 0.7084099650382996, + "learning_rate": 9.996074536094519e-05, + "loss": 0.0392, "step": 3720 }, { - "grad_norm": 0.5748140811920166, - "learning_rate": 9.499552897110136e-05, - "loss": 0.0258, + "grad_norm": 0.6888859272003174, + "learning_rate": 9.995964599232168e-05, + "loss": 0.0427, "step": 3730 }, { - "grad_norm": 0.5241221785545349, - "learning_rate": 9.495941568024304e-05, - "loss": 0.025, + "grad_norm": 0.699871838092804, + "learning_rate": 9.995853144727683e-05, + "loss": 0.0404, "step": 3740 }, { - "grad_norm": 0.42789241671562195, - "learning_rate": 9.492317947205904e-05, - "loss": 0.0231, + "grad_norm": 0.7385432124137878, + "learning_rate": 9.99574017261492e-05, + "loss": 0.0394, "step": 3750 }, { - "grad_norm": 0.6186679601669312, - "learning_rate": 9.488682044561775e-05, - "loss": 0.0259, + "grad_norm": 0.7044745683670044, + "learning_rate": 9.995625682928198e-05, + "loss": 0.0404, "step": 3760 }, { - "grad_norm": 0.6295249462127686, - "learning_rate": 9.485033870032335e-05, - "loss": 0.0204, + "grad_norm": 0.7506102919578552, + "learning_rate": 9.995509675702295e-05, + "loss": 0.0344, "step": 3770 }, { - "grad_norm": 0.5628687739372253, - "learning_rate": 9.481373433591556e-05, - "loss": 0.0252, + "grad_norm": 0.6009000539779663, + "learning_rate": 9.995392150972451e-05, + "loss": 0.0399, "step": 3780 }, { - "grad_norm": 0.4300912916660309, - "learning_rate": 9.47770074524693e-05, - "loss": 0.0224, + "grad_norm": 0.59146648645401, + "learning_rate": 9.995273108774366e-05, + "loss": 0.0378, "step": 3790 }, { - "grad_norm": 0.5620883703231812, - "learning_rate": 9.474015815039446e-05, - "loss": 0.0224, + "grad_norm": 0.5564849972724915, + "learning_rate": 9.995152549144205e-05, + "loss": 0.0372, "step": 3800 }, { - "grad_norm": 0.4041639268398285, - "learning_rate": 9.470318653043565e-05, - "loss": 0.0275, + "grad_norm": 0.6562843322753906, + "learning_rate": 9.995030472118587e-05, + "loss": 0.0424, "step": 3810 }, { - "grad_norm": 0.6825946569442749, - "learning_rate": 9.466609269367185e-05, - "loss": 0.0288, + "grad_norm": 0.7454749345779419, + "learning_rate": 9.9949068777346e-05, + "loss": 0.0441, "step": 3820 }, { - "grad_norm": 0.528630256652832, - "learning_rate": 9.46288767415162e-05, - "loss": 0.0222, + "grad_norm": 0.5691495537757874, + "learning_rate": 9.994781766029786e-05, + "loss": 0.035, "step": 3830 }, { - "grad_norm": 0.5332393646240234, - "learning_rate": 9.459153877571567e-05, - "loss": 0.0231, + "grad_norm": 0.5864580869674683, + "learning_rate": 9.994655137042151e-05, + "loss": 0.036, "step": 3840 }, { - "grad_norm": 0.41178372502326965, - "learning_rate": 9.455407889835087e-05, - "loss": 0.0234, + "grad_norm": 0.5260812640190125, + "learning_rate": 9.99452699081016e-05, + "loss": 0.036, "step": 3850 }, { - "grad_norm": 0.4013999402523041, - "learning_rate": 9.451649721183564e-05, - "loss": 0.026, + "grad_norm": 0.509746789932251, + "learning_rate": 9.994397327372743e-05, + "loss": 0.0392, "step": 3860 }, { - "grad_norm": 0.5963751673698425, - "learning_rate": 9.447879381891692e-05, - "loss": 0.0207, + "grad_norm": 0.6681617498397827, + "learning_rate": 9.994266146769286e-05, + "loss": 0.0347, "step": 3870 }, { - "grad_norm": 0.46209827065467834, - "learning_rate": 9.444096882267428e-05, - "loss": 0.0217, + "grad_norm": 0.5469800233840942, + "learning_rate": 9.994133449039642e-05, + "loss": 0.0354, "step": 3880 }, { - "grad_norm": 0.5197754502296448, - "learning_rate": 9.440302232651988e-05, - "loss": 0.0265, + "grad_norm": 0.6444597840309143, + "learning_rate": 9.993999234224118e-05, + "loss": 0.0408, "step": 3890 }, { - "grad_norm": 0.6915350556373596, - "learning_rate": 9.436495443419795e-05, - "loss": 0.0297, + "grad_norm": 0.5558236241340637, + "learning_rate": 9.993863502363485e-05, + "loss": 0.0438, "step": 3900 }, { - "grad_norm": 0.48685312271118164, - "learning_rate": 9.432676524978466e-05, - "loss": 0.0233, + "grad_norm": 0.5329928994178772, + "learning_rate": 9.993726253498976e-05, + "loss": 0.0362, "step": 3910 }, { - "grad_norm": 0.47423553466796875, - "learning_rate": 9.42884548776878e-05, - "loss": 0.0288, + "grad_norm": 0.6277582049369812, + "learning_rate": 9.993587487672282e-05, + "loss": 0.0422, "step": 3920 }, { - "grad_norm": 0.4048287868499756, - "learning_rate": 9.425002342264646e-05, - "loss": 0.0211, + "grad_norm": 0.655707836151123, + "learning_rate": 9.993447204925558e-05, + "loss": 0.0342, "step": 3930 }, { - "grad_norm": 0.44486644864082336, - "learning_rate": 9.421147098973077e-05, - "loss": 0.0238, + "grad_norm": 0.5586263537406921, + "learning_rate": 9.993305405301416e-05, + "loss": 0.0367, "step": 3940 }, { - "grad_norm": 0.5044620633125305, - "learning_rate": 9.41727976843416e-05, - "loss": 0.023, + "grad_norm": 0.44315865635871887, + "learning_rate": 9.993162088842935e-05, + "loss": 0.0353, "step": 3950 }, { - "grad_norm": 0.6554492712020874, - "learning_rate": 9.413400361221029e-05, - "loss": 0.0224, + "grad_norm": 0.6656477451324463, + "learning_rate": 9.993017255593646e-05, + "loss": 0.0343, "step": 3960 }, { - "grad_norm": 0.5063493251800537, - "learning_rate": 9.409508887939835e-05, - "loss": 0.02, + "grad_norm": 0.5894978046417236, + "learning_rate": 9.992870905597548e-05, + "loss": 0.0318, "step": 3970 }, { - "grad_norm": 0.5956079959869385, - "learning_rate": 9.40560535922972e-05, - "loss": 0.02, + "grad_norm": 0.6920962333679199, + "learning_rate": 9.9927230388991e-05, + "loss": 0.0327, "step": 3980 }, { - "grad_norm": 0.6585922837257385, - "learning_rate": 9.40168978576278e-05, - "loss": 0.0224, + "grad_norm": 0.6823915839195251, + "learning_rate": 9.992573655543215e-05, + "loss": 0.035, "step": 3990 }, { - "grad_norm": 0.4550802707672119, - "learning_rate": 9.397762178244043e-05, - "loss": 0.0235, + "grad_norm": 0.6368980407714844, + "learning_rate": 9.992422755575277e-05, + "loss": 0.0364, "step": 4000 }, { - "grad_norm": 0.4042211174964905, - "learning_rate": 9.393822547411439e-05, - "loss": 0.0208, + "grad_norm": 0.570302426815033, + "learning_rate": 9.992270339041123e-05, + "loss": 0.0337, "step": 4010 }, { - "grad_norm": 0.4356859028339386, - "learning_rate": 9.389870904035769e-05, - "loss": 0.0199, + "grad_norm": 0.5436093807220459, + "learning_rate": 9.992116405987053e-05, + "loss": 0.0319, "step": 4020 }, { - "grad_norm": 0.45227131247520447, - "learning_rate": 9.385907258920672e-05, - "loss": 0.0227, + "grad_norm": 0.6327980756759644, + "learning_rate": 9.991960956459828e-05, + "loss": 0.0352, "step": 4030 }, { - "grad_norm": 0.4879400134086609, - "learning_rate": 9.381931622902607e-05, - "loss": 0.0192, + "grad_norm": 0.4797522723674774, + "learning_rate": 9.991803990506669e-05, + "loss": 0.0307, "step": 4040 }, { - "grad_norm": 0.48629337549209595, - "learning_rate": 9.377944006850807e-05, - "loss": 0.0225, + "grad_norm": 0.5248384475708008, + "learning_rate": 9.991645508175258e-05, + "loss": 0.0346, "step": 4050 }, { - "grad_norm": 0.4865284562110901, - "learning_rate": 9.373944421667265e-05, - "loss": 0.0212, + "grad_norm": 0.6256198883056641, + "learning_rate": 9.99148550951374e-05, + "loss": 0.0334, "step": 4060 }, { - "grad_norm": 0.6001793742179871, - "learning_rate": 9.369932878286691e-05, - "loss": 0.0184, + "grad_norm": 0.6489942073822021, + "learning_rate": 9.991323994570716e-05, + "loss": 0.0299, "step": 4070 }, { - "grad_norm": 0.5513994097709656, - "learning_rate": 9.365909387676494e-05, - "loss": 0.021, + "grad_norm": 0.6390136480331421, + "learning_rate": 9.99116096339525e-05, + "loss": 0.0326, "step": 4080 }, { - "grad_norm": 0.6248122453689575, - "learning_rate": 9.361873960836744e-05, - "loss": 0.0208, + "grad_norm": 0.7176806926727295, + "learning_rate": 9.990996416036869e-05, + "loss": 0.033, "step": 4090 }, { - "grad_norm": 0.4238356947898865, - "learning_rate": 9.357826608800142e-05, - "loss": 0.0182, + "grad_norm": 0.6497689485549927, + "learning_rate": 9.990830352545555e-05, + "loss": 0.0303, "step": 4100 }, { - "grad_norm": 0.4596106708049774, - "learning_rate": 9.353767342631994e-05, - "loss": 0.023, + "grad_norm": 0.5362087488174438, + "learning_rate": 9.990662772971756e-05, + "loss": 0.0361, "step": 4110 }, { - "grad_norm": 0.39670488238334656, - "learning_rate": 9.34969617343018e-05, - "loss": 0.023, + "grad_norm": 0.5707340240478516, + "learning_rate": 9.990493677366376e-05, + "loss": 0.0361, "step": 4120 }, { - "grad_norm": 0.5592398047447205, - "learning_rate": 9.345613112325122e-05, - "loss": 0.0237, + "grad_norm": 0.6157304644584656, + "learning_rate": 9.990323065780786e-05, + "loss": 0.0359, "step": 4130 }, { - "grad_norm": 0.5025811195373535, - "learning_rate": 9.34151817047975e-05, - "loss": 0.0207, + "grad_norm": 0.5689452886581421, + "learning_rate": 9.990150938266808e-05, + "loss": 0.0322, "step": 4140 }, { - "grad_norm": 0.5580839514732361, - "learning_rate": 9.33741135908948e-05, - "loss": 0.0224, + "grad_norm": 0.6773049831390381, + "learning_rate": 9.989977294876733e-05, + "loss": 0.0347, "step": 4150 }, { - "grad_norm": 0.5201109051704407, - "learning_rate": 9.33329268938218e-05, - "loss": 0.0206, + "grad_norm": 0.5795975923538208, + "learning_rate": 9.989802135663308e-05, + "loss": 0.0324, "step": 4160 }, { - "grad_norm": 0.513161838054657, - "learning_rate": 9.329162172618132e-05, - "loss": 0.0231, + "grad_norm": 0.5491712093353271, + "learning_rate": 9.989625460679743e-05, + "loss": 0.0355, "step": 4170 }, { - "grad_norm": 0.4309871792793274, - "learning_rate": 9.325019820090013e-05, - "loss": 0.022, + "grad_norm": 0.5472167730331421, + "learning_rate": 9.989447269979706e-05, + "loss": 0.0334, "step": 4180 }, { - "grad_norm": 0.43937987089157104, - "learning_rate": 9.320865643122855e-05, - "loss": 0.0233, + "grad_norm": 0.6765989065170288, + "learning_rate": 9.989267563617328e-05, + "loss": 0.0356, "step": 4190 }, { - "grad_norm": 0.4912075400352478, - "learning_rate": 9.316699653074023e-05, - "loss": 0.0182, + "grad_norm": 0.6238303184509277, + "learning_rate": 9.989086341647198e-05, + "loss": 0.0296, "step": 4200 }, { - "grad_norm": 0.4246988892555237, - "learning_rate": 9.312521861333172e-05, - "loss": 0.0192, + "grad_norm": 0.6904252767562866, + "learning_rate": 9.988903604124366e-05, + "loss": 0.0306, "step": 4210 }, { - "grad_norm": 0.39220497012138367, - "learning_rate": 9.308332279322224e-05, - "loss": 0.0194, + "grad_norm": 0.6524732708930969, + "learning_rate": 9.988719351104343e-05, + "loss": 0.0312, "step": 4220 }, { - "grad_norm": 0.5391877293586731, - "learning_rate": 9.304130918495338e-05, - "loss": 0.02, + "grad_norm": 0.5553227663040161, + "learning_rate": 9.9885335826431e-05, + "loss": 0.0311, "step": 4230 }, { - "grad_norm": 0.4661605954170227, - "learning_rate": 9.299917790338874e-05, - "loss": 0.0201, + "grad_norm": 0.5986655950546265, + "learning_rate": 9.988346298797071e-05, + "loss": 0.0299, "step": 4240 }, { - "grad_norm": 0.3973395824432373, - "learning_rate": 9.295692906371363e-05, - "loss": 0.0213, + "grad_norm": 0.6228980422019958, + "learning_rate": 9.988157499623146e-05, + "loss": 0.0316, "step": 4250 }, { - "grad_norm": 0.4798072278499603, - "learning_rate": 9.291456278143476e-05, - "loss": 0.02, + "grad_norm": 0.5520809292793274, + "learning_rate": 9.987967185178677e-05, + "loss": 0.0301, "step": 4260 }, { - "grad_norm": 0.6149327158927917, - "learning_rate": 9.287207917237994e-05, - "loss": 0.0203, + "grad_norm": 0.6654737591743469, + "learning_rate": 9.987775355521476e-05, + "loss": 0.0308, "step": 4270 }, { - "grad_norm": 0.5336242318153381, - "learning_rate": 9.282947835269773e-05, - "loss": 0.0215, + "grad_norm": 0.5442163348197937, + "learning_rate": 9.987582010709817e-05, + "loss": 0.0319, "step": 4280 }, { - "grad_norm": 0.5178649425506592, - "learning_rate": 9.278676043885715e-05, - "loss": 0.0194, + "grad_norm": 0.6658127903938293, + "learning_rate": 9.987387150802431e-05, + "loss": 0.029, "step": 4290 }, { - "grad_norm": 0.4814247190952301, - "learning_rate": 9.274392554764733e-05, - "loss": 0.0171, + "grad_norm": 0.5804133415222168, + "learning_rate": 9.987190775858517e-05, + "loss": 0.0277, "step": 4300 }, { - "grad_norm": 0.32441624999046326, - "learning_rate": 9.270097379617723e-05, - "loss": 0.0184, + "grad_norm": 0.5722190141677856, + "learning_rate": 9.98699288593772e-05, + "loss": 0.0292, "step": 4310 }, { - "grad_norm": 0.5904157757759094, - "learning_rate": 9.26579053018753e-05, - "loss": 0.0204, + "grad_norm": 0.623306393623352, + "learning_rate": 9.986793481100161e-05, + "loss": 0.0313, "step": 4320 }, { - "grad_norm": 0.45750078558921814, - "learning_rate": 9.261472018248918e-05, - "loss": 0.0191, + "grad_norm": 0.6122274994850159, + "learning_rate": 9.986592561406412e-05, + "loss": 0.0292, "step": 4330 }, { - "grad_norm": 0.5417255759239197, - "learning_rate": 9.25714185560853e-05, - "loss": 0.0208, + "grad_norm": 0.5425803661346436, + "learning_rate": 9.986390126917503e-05, + "loss": 0.0313, "step": 4340 }, { - "grad_norm": 0.41097697615623474, - "learning_rate": 9.252800054104868e-05, - "loss": 0.0238, + "grad_norm": 0.511246919631958, + "learning_rate": 9.986186177694933e-05, + "loss": 0.0349, "step": 4350 }, { - "grad_norm": 0.42212045192718506, - "learning_rate": 9.248446625608252e-05, - "loss": 0.0196, + "grad_norm": 0.5783056020736694, + "learning_rate": 9.985980713800656e-05, + "loss": 0.0285, "step": 4360 }, { - "grad_norm": 0.5629220008850098, - "learning_rate": 9.244081582020789e-05, - "loss": 0.0208, + "grad_norm": 0.810947060585022, + "learning_rate": 9.985773735297084e-05, + "loss": 0.0315, "step": 4370 }, { - "grad_norm": 0.4684229791164398, - "learning_rate": 9.239704935276339e-05, - "loss": 0.0226, + "grad_norm": 0.5040915608406067, + "learning_rate": 9.985565242247092e-05, + "loss": 0.0327, "step": 4380 }, { - "grad_norm": 0.43924668431282043, - "learning_rate": 9.235316697340489e-05, - "loss": 0.0247, + "grad_norm": 0.4829008877277374, + "learning_rate": 9.985355234714016e-05, + "loss": 0.0362, "step": 4390 }, { - "grad_norm": 0.5363799929618835, - "learning_rate": 9.230916880210512e-05, - "loss": 0.0193, + "grad_norm": 0.49304136633872986, + "learning_rate": 9.985143712761652e-05, + "loss": 0.0288, "step": 4400 }, { - "grad_norm": 0.4902143180370331, - "learning_rate": 9.226505495915342e-05, - "loss": 0.0204, + "grad_norm": 0.5399729013442993, + "learning_rate": 9.984930676454252e-05, + "loss": 0.0298, "step": 4410 }, { - "grad_norm": 0.4934154450893402, - "learning_rate": 9.222082556515536e-05, - "loss": 0.0204, + "grad_norm": 0.6986041069030762, + "learning_rate": 9.984716125856532e-05, + "loss": 0.0308, "step": 4420 }, { - "grad_norm": 0.5217421650886536, - "learning_rate": 9.217648074103242e-05, - "loss": 0.0187, + "grad_norm": 0.5839313864707947, + "learning_rate": 9.984500061033667e-05, + "loss": 0.0286, "step": 4430 }, { - "grad_norm": 0.5157290101051331, - "learning_rate": 9.213202060802161e-05, - "loss": 0.0193, + "grad_norm": 0.6318590044975281, + "learning_rate": 9.984282482051293e-05, + "loss": 0.0291, "step": 4440 }, { - "grad_norm": 0.48323342204093933, - "learning_rate": 9.208744528767528e-05, - "loss": 0.0222, + "grad_norm": 0.7042383551597595, + "learning_rate": 9.9840633889755e-05, + "loss": 0.032, "step": 4450 }, { - "grad_norm": 0.46333587169647217, - "learning_rate": 9.204275490186064e-05, - "loss": 0.0188, + "grad_norm": 0.6632933616638184, + "learning_rate": 9.983842781872848e-05, + "loss": 0.0283, "step": 4460 }, { - "grad_norm": 0.48506951332092285, - "learning_rate": 9.199794957275949e-05, - "loss": 0.0181, + "grad_norm": 0.454008013010025, + "learning_rate": 9.98362066081035e-05, + "loss": 0.0273, "step": 4470 }, { - "grad_norm": 0.547903835773468, - "learning_rate": 9.19530294228679e-05, - "loss": 0.0201, + "grad_norm": 0.6959115862846375, + "learning_rate": 9.983397025855479e-05, + "loss": 0.0294, "step": 4480 }, { - "grad_norm": 0.5428500771522522, - "learning_rate": 9.190799457499583e-05, - "loss": 0.023, + "grad_norm": 0.6951610445976257, + "learning_rate": 9.983171877076171e-05, + "loss": 0.033, "step": 4490 }, { - "grad_norm": 0.4775592088699341, - "learning_rate": 9.186284515226686e-05, - "loss": 0.0187, + "grad_norm": 0.45543181896209717, + "learning_rate": 9.98294521454082e-05, + "loss": 0.0274, "step": 4500 }, { - "grad_norm": 0.515190839767456, - "learning_rate": 9.181758127811777e-05, - "loss": 0.0174, + "grad_norm": 0.6101294755935669, + "learning_rate": 9.98271703831828e-05, + "loss": 0.0253, "step": 4510 }, { - "grad_norm": 0.523912250995636, - "learning_rate": 9.177220307629825e-05, - "loss": 0.0183, + "grad_norm": 0.5760356187820435, + "learning_rate": 9.982487348477865e-05, + "loss": 0.0264, "step": 4520 }, { - "grad_norm": 0.46641191840171814, - "learning_rate": 9.172671067087059e-05, - "loss": 0.0218, + "grad_norm": 0.6294093132019043, + "learning_rate": 9.982256145089347e-05, + "loss": 0.0304, "step": 4530 }, { - "grad_norm": 0.4439499080181122, - "learning_rate": 9.16811041862093e-05, - "loss": 0.0194, + "grad_norm": 0.539822518825531, + "learning_rate": 9.982023428222962e-05, + "loss": 0.0283, "step": 4540 }, { - "grad_norm": 0.39287662506103516, - "learning_rate": 9.163538374700076e-05, - "loss": 0.0174, + "grad_norm": 0.3482538163661957, + "learning_rate": 9.981789197949403e-05, + "loss": 0.0259, "step": 4550 }, { - "grad_norm": 0.49717986583709717, - "learning_rate": 9.158954947824287e-05, - "loss": 0.0192, + "grad_norm": 0.5884767770767212, + "learning_rate": 9.98155345433982e-05, + "loss": 0.0272, "step": 4560 }, { - "grad_norm": 0.4734252691268921, - "learning_rate": 9.154360150524482e-05, - "loss": 0.0177, + "grad_norm": 0.5896508097648621, + "learning_rate": 9.981316197465831e-05, + "loss": 0.0259, "step": 4570 }, { - "grad_norm": 0.37044256925582886, - "learning_rate": 9.14975399536266e-05, - "loss": 0.0216, + "grad_norm": 0.6355373859405518, + "learning_rate": 9.981077427399504e-05, + "loss": 0.0302, "step": 4580 }, { - "grad_norm": 0.36238211393356323, - "learning_rate": 9.14513649493187e-05, - "loss": 0.0241, + "grad_norm": 0.6076304912567139, + "learning_rate": 9.980837144213371e-05, + "loss": 0.0329, "step": 4590 }, { - "grad_norm": 0.3750474750995636, - "learning_rate": 9.140507661856187e-05, - "loss": 0.0188, + "grad_norm": 0.44014647603034973, + "learning_rate": 9.980595347980426e-05, + "loss": 0.0264, "step": 4600 }, { - "grad_norm": 0.6232521533966064, - "learning_rate": 9.135867508790661e-05, - "loss": 0.0232, + "grad_norm": 0.6942553520202637, + "learning_rate": 9.980352038774119e-05, + "loss": 0.031, "step": 4610 }, { - "grad_norm": 0.40693479776382446, - "learning_rate": 9.131216048421291e-05, - "loss": 0.0191, + "grad_norm": 0.5383680462837219, + "learning_rate": 9.98010721666836e-05, + "loss": 0.026, "step": 4620 }, { - "grad_norm": 0.46917590498924255, - "learning_rate": 9.126553293464998e-05, - "loss": 0.0192, + "grad_norm": 0.5941911339759827, + "learning_rate": 9.979860881737523e-05, + "loss": 0.0269, "step": 4630 }, { - "grad_norm": 0.6374298930168152, - "learning_rate": 9.121879256669572e-05, - "loss": 0.0209, + "grad_norm": 0.6166422367095947, + "learning_rate": 9.979613034056434e-05, + "loss": 0.0285, "step": 4640 }, { - "grad_norm": 0.49459344148635864, - "learning_rate": 9.117193950813652e-05, - "loss": 0.0191, + "grad_norm": 0.5569292902946472, + "learning_rate": 9.979363673700386e-05, + "loss": 0.0264, "step": 4650 }, { - "grad_norm": 0.3625936508178711, - "learning_rate": 9.112497388706685e-05, - "loss": 0.02, + "grad_norm": 0.4750020205974579, + "learning_rate": 9.979112800745124e-05, + "loss": 0.0277, "step": 4660 }, { - "grad_norm": 0.5517452955245972, - "learning_rate": 9.10778958318889e-05, - "loss": 0.0178, + "grad_norm": 0.6898394823074341, + "learning_rate": 9.978860415266861e-05, + "loss": 0.0251, "step": 4670 }, { - "grad_norm": 0.5170832276344299, - "learning_rate": 9.103070547131232e-05, - "loss": 0.0172, + "grad_norm": 0.630106508731842, + "learning_rate": 9.978606517342262e-05, + "loss": 0.0247, "step": 4680 }, { - "grad_norm": 0.5125618577003479, - "learning_rate": 9.098340293435375e-05, - "loss": 0.0172, + "grad_norm": 0.5730926990509033, + "learning_rate": 9.978351107048456e-05, + "loss": 0.0252, "step": 4690 }, { - "grad_norm": 0.6093714833259583, - "learning_rate": 9.093598835033649e-05, - "loss": 0.0253, + "grad_norm": 0.6251030564308167, + "learning_rate": 9.978094184463029e-05, + "loss": 0.0341, "step": 4700 }, { - "grad_norm": 0.48880383372306824, - "learning_rate": 9.088846184889021e-05, - "loss": 0.02, + "grad_norm": 0.49815332889556885, + "learning_rate": 9.977835749664029e-05, + "loss": 0.0274, "step": 4710 }, { - "grad_norm": 0.3739651143550873, - "learning_rate": 9.084082355995057e-05, - "loss": 0.0172, + "grad_norm": 0.611866295337677, + "learning_rate": 9.97757580272996e-05, + "loss": 0.0235, "step": 4720 }, { - "grad_norm": 0.41044220328330994, - "learning_rate": 9.079307361375882e-05, - "loss": 0.0178, + "grad_norm": 0.5598731637001038, + "learning_rate": 9.977314343739786e-05, + "loss": 0.0244, "step": 4730 }, { - "grad_norm": 0.42307567596435547, - "learning_rate": 9.074521214086149e-05, - "loss": 0.0184, + "grad_norm": 0.6545037627220154, + "learning_rate": 9.977051372772934e-05, + "loss": 0.0256, "step": 4740 }, { - "grad_norm": 0.39302605390548706, - "learning_rate": 9.069723927211001e-05, - "loss": 0.0175, + "grad_norm": 0.5057151913642883, + "learning_rate": 9.976786889909286e-05, + "loss": 0.0241, "step": 4750 }, { - "grad_norm": 0.4822540581226349, - "learning_rate": 9.064915513866037e-05, - "loss": 0.0205, + "grad_norm": 0.7073130011558533, + "learning_rate": 9.976520895229185e-05, + "loss": 0.0272, "step": 4760 }, { - "grad_norm": 0.4448596239089966, - "learning_rate": 9.060095987197279e-05, - "loss": 0.0246, + "grad_norm": 0.5498185157775879, + "learning_rate": 9.976253388813433e-05, + "loss": 0.0321, "step": 4770 }, { - "grad_norm": 0.3715053200721741, - "learning_rate": 9.055265360381126e-05, - "loss": 0.0182, + "grad_norm": 0.5412503480911255, + "learning_rate": 9.975984370743293e-05, + "loss": 0.0247, "step": 4780 }, { - "grad_norm": 0.3781905770301819, - "learning_rate": 9.050423646624326e-05, - "loss": 0.0218, + "grad_norm": 0.5492538213729858, + "learning_rate": 9.975713841100485e-05, + "loss": 0.0282, "step": 4790 }, { - "grad_norm": 0.4780808091163635, - "learning_rate": 9.045570859163943e-05, - "loss": 0.0203, + "grad_norm": 0.5769220590591431, + "learning_rate": 9.975441799967187e-05, + "loss": 0.0274, "step": 4800 }, { - "grad_norm": 0.45449933409690857, - "learning_rate": 9.04070701126731e-05, - "loss": 0.0168, + "grad_norm": 0.5793959498405457, + "learning_rate": 9.975168247426039e-05, + "loss": 0.023, "step": 4810 }, { - "grad_norm": 0.455399751663208, - "learning_rate": 9.035832116232001e-05, - "loss": 0.0192, + "grad_norm": 0.5048704147338867, + "learning_rate": 9.974893183560139e-05, + "loss": 0.0256, "step": 4820 }, { - "grad_norm": 0.569148600101471, - "learning_rate": 9.030946187385796e-05, - "loss": 0.0191, + "grad_norm": 0.5510044097900391, + "learning_rate": 9.974616608453045e-05, + "loss": 0.0242, "step": 4830 }, { - "grad_norm": 0.39798951148986816, - "learning_rate": 9.026049238086635e-05, - "loss": 0.0183, + "grad_norm": 0.645194411277771, + "learning_rate": 9.974338522188772e-05, + "loss": 0.0241, "step": 4840 }, { - "grad_norm": 0.4741237759590149, - "learning_rate": 9.021141281722591e-05, - "loss": 0.0185, + "grad_norm": 0.5308319330215454, + "learning_rate": 9.974058924851797e-05, + "loss": 0.0248, "step": 4850 }, { - "grad_norm": 0.44035014510154724, - "learning_rate": 9.01622233171183e-05, - "loss": 0.0169, + "grad_norm": 0.5941885709762573, + "learning_rate": 9.973777816527051e-05, + "loss": 0.0224, "step": 4860 }, { - "grad_norm": 0.4502478241920471, - "learning_rate": 9.011292401502574e-05, - "loss": 0.0207, + "grad_norm": 0.5658865571022034, + "learning_rate": 9.973495197299931e-05, + "loss": 0.027, "step": 4870 }, { - "grad_norm": 0.38435491919517517, - "learning_rate": 9.006351504573063e-05, - "loss": 0.0183, + "grad_norm": 0.5737472772598267, + "learning_rate": 9.973211067256287e-05, + "loss": 0.025, "step": 4880 }, { - "grad_norm": 0.36374086141586304, - "learning_rate": 9.001399654431519e-05, - "loss": 0.0195, + "grad_norm": 0.552737832069397, + "learning_rate": 9.97292542648243e-05, + "loss": 0.0261, "step": 4890 }, { - "grad_norm": 0.4774911105632782, - "learning_rate": 8.996436864616116e-05, - "loss": 0.0177, + "grad_norm": 0.6258673071861267, + "learning_rate": 9.972638275065131e-05, + "loss": 0.0243, "step": 4900 }, { - "grad_norm": 0.3256385624408722, - "learning_rate": 8.991463148694925e-05, - "loss": 0.0166, + "grad_norm": 0.553909957408905, + "learning_rate": 9.972349613091621e-05, + "loss": 0.0224, "step": 4910 }, { - "grad_norm": 0.5121222734451294, - "learning_rate": 8.986478520265902e-05, - "loss": 0.0191, + "grad_norm": 0.4020157754421234, + "learning_rate": 9.972059440649584e-05, + "loss": 0.0248, "step": 4920 }, { - "grad_norm": 0.5058743953704834, - "learning_rate": 8.981482992956827e-05, - "loss": 0.0202, + "grad_norm": 0.4582313001155853, + "learning_rate": 9.971767757827168e-05, + "loss": 0.0257, "step": 4930 }, { - "grad_norm": 0.436591774225235, - "learning_rate": 8.976476580425282e-05, - "loss": 0.0165, + "grad_norm": 0.5173747539520264, + "learning_rate": 9.971474564712982e-05, + "loss": 0.0214, "step": 4940 }, { - "grad_norm": 0.4705229699611664, - "learning_rate": 8.971459296358606e-05, - "loss": 0.0195, + "grad_norm": 0.5914600491523743, + "learning_rate": 9.971179861396084e-05, + "loss": 0.0265, "step": 4950 }, { - "grad_norm": 0.423905611038208, - "learning_rate": 8.966431154473864e-05, - "loss": 0.0201, + "grad_norm": 0.4407144784927368, + "learning_rate": 9.970883647966003e-05, + "loss": 0.0279, "step": 4960 }, { - "grad_norm": 0.4129910469055176, - "learning_rate": 8.961392168517803e-05, - "loss": 0.0149, + "grad_norm": 0.6293507218360901, + "learning_rate": 9.970585924512717e-05, + "loss": 0.0272, "step": 4970 }, { - "grad_norm": 0.45044130086898804, - "learning_rate": 8.956342352266821e-05, - "loss": 0.0209, + "grad_norm": 0.5919846296310425, + "learning_rate": 9.970286691126669e-05, + "loss": 0.0345, "step": 4980 }, { - "grad_norm": 0.43892666697502136, - "learning_rate": 8.95128171952692e-05, - "loss": 0.0204, + "grad_norm": 0.5908362865447998, + "learning_rate": 9.969985947898756e-05, + "loss": 0.031, "step": 4990 }, { - "grad_norm": 0.4579017758369446, - "learning_rate": 8.946210284133676e-05, - "loss": 0.0172, + "grad_norm": 0.5150045156478882, + "learning_rate": 9.969683694920337e-05, + "loss": 0.0265, "step": 5000 }, { - "grad_norm": 0.38334211707115173, - "learning_rate": 8.941128059952201e-05, - "loss": 0.0176, + "grad_norm": 0.538543701171875, + "learning_rate": 9.969379932283228e-05, + "loss": 0.0271, "step": 5010 }, { - "grad_norm": 0.4932248592376709, - "learning_rate": 8.936035060877102e-05, - "loss": 0.0197, + "grad_norm": 0.6217741966247559, + "learning_rate": 9.969074660079704e-05, + "loss": 0.0288, "step": 5020 }, { - "grad_norm": 0.4217004179954529, - "learning_rate": 8.930931300832443e-05, - "loss": 0.0159, + "grad_norm": 0.4673011600971222, + "learning_rate": 9.968767878402501e-05, + "loss": 0.0237, "step": 5030 }, { - "grad_norm": 0.5817981958389282, - "learning_rate": 8.925816793771711e-05, - "loss": 0.0162, + "grad_norm": 0.4540856182575226, + "learning_rate": 9.968459587344808e-05, + "loss": 0.0228, "step": 5040 }, { - "grad_norm": 0.43254992365837097, - "learning_rate": 8.92069155367777e-05, - "loss": 0.0193, + "grad_norm": 0.537109375, + "learning_rate": 9.968149787000278e-05, + "loss": 0.025, "step": 5050 }, { - "grad_norm": 0.5035724639892578, - "learning_rate": 8.915555594562834e-05, - "loss": 0.0208, + "grad_norm": 0.5948236584663391, + "learning_rate": 9.967838477463018e-05, + "loss": 0.027, "step": 5060 }, { - "grad_norm": 0.4568447172641754, - "learning_rate": 8.910408930468416e-05, - "loss": 0.0165, + "grad_norm": 0.4020276963710785, + "learning_rate": 9.967525658827597e-05, + "loss": 0.0221, "step": 5070 }, { - "grad_norm": 0.43123510479927063, - "learning_rate": 8.905251575465303e-05, - "loss": 0.0158, + "grad_norm": 0.3882424831390381, + "learning_rate": 9.967211331189042e-05, + "loss": 0.021, "step": 5080 }, { - "grad_norm": 0.4527429938316345, - "learning_rate": 8.900083543653502e-05, - "loss": 0.0169, + "grad_norm": 0.5875966548919678, + "learning_rate": 9.966895494642834e-05, + "loss": 0.0225, "step": 5090 }, { - "grad_norm": 0.41953304409980774, - "learning_rate": 8.894904849162218e-05, - "loss": 0.0184, + "grad_norm": 0.592677116394043, + "learning_rate": 9.96657814928492e-05, + "loss": 0.0237, "step": 5100 }, { - "grad_norm": 0.415479838848114, - "learning_rate": 8.889715506149802e-05, - "loss": 0.0184, + "grad_norm": 0.5306417346000671, + "learning_rate": 9.966259295211697e-05, + "loss": 0.0244, "step": 5110 }, { - "grad_norm": 0.3731771409511566, - "learning_rate": 8.884515528803722e-05, - "loss": 0.0154, + "grad_norm": 0.4629195034503937, + "learning_rate": 9.965938932520028e-05, + "loss": 0.0206, "step": 5120 }, { - "grad_norm": 0.41724300384521484, - "learning_rate": 8.879304931340517e-05, - "loss": 0.0163, + "grad_norm": 0.5218473076820374, + "learning_rate": 9.965617061307229e-05, + "loss": 0.0219, "step": 5130 }, { - "grad_norm": 0.47969841957092285, - "learning_rate": 8.874083728005759e-05, - "loss": 0.0177, + "grad_norm": 0.5671105980873108, + "learning_rate": 9.965293681671077e-05, + "loss": 0.0231, "step": 5140 }, { - "grad_norm": 0.34264758229255676, - "learning_rate": 8.868851933074021e-05, - "loss": 0.0182, + "grad_norm": 0.49504706263542175, + "learning_rate": 9.964968793709804e-05, + "loss": 0.023, "step": 5150 }, { - "grad_norm": 0.4063286781311035, - "learning_rate": 8.863609560848829e-05, - "loss": 0.0182, + "grad_norm": 0.411983460187912, + "learning_rate": 9.964642397522106e-05, + "loss": 0.0228, "step": 5160 }, { - "grad_norm": 0.42877396941185, - "learning_rate": 8.85835662566263e-05, - "loss": 0.0165, + "grad_norm": 0.5020577907562256, + "learning_rate": 9.96431449320713e-05, + "loss": 0.0207, "step": 5170 }, { - "grad_norm": 0.4081937372684479, - "learning_rate": 8.853093141876747e-05, - "loss": 0.0148, + "grad_norm": 0.4675840139389038, + "learning_rate": 9.963985080864486e-05, + "loss": 0.0198, "step": 5180 }, { - "grad_norm": 0.40108522772789, - "learning_rate": 8.847819123881343e-05, - "loss": 0.0174, + "grad_norm": 0.4841030240058899, + "learning_rate": 9.96365416059424e-05, + "loss": 0.0228, "step": 5190 }, { - "grad_norm": 0.4415458142757416, - "learning_rate": 8.842534586095383e-05, - "loss": 0.0165, + "grad_norm": 0.39482370018959045, + "learning_rate": 9.963321732496919e-05, + "loss": 0.0213, "step": 5200 }, { - "grad_norm": 0.48253053426742554, - "learning_rate": 8.837239542966593e-05, - "loss": 0.0212, + "grad_norm": 0.5138295888900757, + "learning_rate": 9.962987796673506e-05, + "loss": 0.0259, "step": 5210 }, { - "grad_norm": 0.4505639970302582, - "learning_rate": 8.831934008971417e-05, - "loss": 0.0211, + "grad_norm": 0.41514965891838074, + "learning_rate": 9.962652353225438e-05, + "loss": 0.0262, "step": 5220 }, { - "grad_norm": 0.4596928656101227, - "learning_rate": 8.826617998614982e-05, - "loss": 0.0182, + "grad_norm": 0.5538020133972168, + "learning_rate": 9.962315402254619e-05, + "loss": 0.0232, "step": 5230 }, { - "grad_norm": 0.4037300646305084, - "learning_rate": 8.821291526431056e-05, - "loss": 0.0146, + "grad_norm": 0.47969186305999756, + "learning_rate": 9.9619769438634e-05, + "loss": 0.0201, "step": 5240 }, { - "grad_norm": 0.3879428505897522, - "learning_rate": 8.815954606982015e-05, - "loss": 0.0173, + "grad_norm": 0.4588451683521271, + "learning_rate": 9.9616369781546e-05, + "loss": 0.0222, "step": 5250 }, { - "grad_norm": 0.4592052102088928, - "learning_rate": 8.810607254858789e-05, - "loss": 0.0158, + "grad_norm": 0.5873843431472778, + "learning_rate": 9.961295505231491e-05, + "loss": 0.0216, "step": 5260 }, { - "grad_norm": 0.43258506059646606, - "learning_rate": 8.805249484680838e-05, - "loss": 0.0176, + "grad_norm": 0.4506019353866577, + "learning_rate": 9.960952525197804e-05, + "loss": 0.023, "step": 5270 }, { - "grad_norm": 0.5278488993644714, - "learning_rate": 8.799881311096096e-05, - "loss": 0.0175, + "grad_norm": 0.5607607960700989, + "learning_rate": 9.960608038157724e-05, + "loss": 0.0223, "step": 5280 }, { - "grad_norm": 0.536786675453186, - "learning_rate": 8.794502748780949e-05, - "loss": 0.0191, + "grad_norm": 0.6128898859024048, + "learning_rate": 9.960262044215901e-05, + "loss": 0.0238, "step": 5290 }, { - "grad_norm": 0.39231422543525696, - "learning_rate": 8.78911381244018e-05, - "loss": 0.0213, + "grad_norm": 0.4695815443992615, + "learning_rate": 9.959914543477435e-05, + "loss": 0.0265, "step": 5300 }, { - "grad_norm": 0.3669500946998596, - "learning_rate": 8.783714516806933e-05, - "loss": 0.0227, + "grad_norm": 0.43593689799308777, + "learning_rate": 9.959565536047892e-05, + "loss": 0.0275, "step": 5310 }, { - "grad_norm": 0.4094989001750946, - "learning_rate": 8.77830487664268e-05, - "loss": 0.0239, + "grad_norm": 0.6567972898483276, + "learning_rate": 9.959215022033288e-05, + "loss": 0.0283, "step": 5320 }, { - "grad_norm": 0.3868635296821594, - "learning_rate": 8.772884906737167e-05, - "loss": 0.0174, + "grad_norm": 0.38720253109931946, + "learning_rate": 9.9588630015401e-05, + "loss": 0.0216, "step": 5330 }, { - "grad_norm": 0.4408529996871948, - "learning_rate": 8.767454621908387e-05, - "loss": 0.02, + "grad_norm": 0.5306717753410339, + "learning_rate": 9.958509474675264e-05, + "loss": 0.0249, "step": 5340 }, { - "grad_norm": 0.4143405258655548, - "learning_rate": 8.76201403700253e-05, - "loss": 0.0188, + "grad_norm": 0.5195200443267822, + "learning_rate": 9.958154441546171e-05, + "loss": 0.0234, "step": 5350 }, { - "grad_norm": 0.4451104402542114, - "learning_rate": 8.756563166893949e-05, - "loss": 0.0157, + "grad_norm": 0.411095529794693, + "learning_rate": 9.957797902260673e-05, + "loss": 0.02, "step": 5360 }, { - "grad_norm": 0.45531290769577026, - "learning_rate": 8.751102026485113e-05, - "loss": 0.0164, + "grad_norm": 0.5348077416419983, + "learning_rate": 9.957439856927073e-05, + "loss": 0.0221, "step": 5370 }, { - "grad_norm": 0.4830329418182373, - "learning_rate": 8.745630630706571e-05, - "loss": 0.0164, + "grad_norm": 0.5276638269424438, + "learning_rate": 9.957080305654139e-05, + "loss": 0.0228, "step": 5380 }, { - "grad_norm": 0.390863835811615, - "learning_rate": 8.740148994516912e-05, - "loss": 0.0157, + "grad_norm": 0.40877729654312134, + "learning_rate": 9.956719248551092e-05, + "loss": 0.0211, "step": 5390 }, { - "grad_norm": 0.32199040055274963, - "learning_rate": 8.73465713290272e-05, - "loss": 0.0147, + "grad_norm": 0.5705676674842834, + "learning_rate": 9.956356685727612e-05, + "loss": 0.0192, "step": 5400 }, { - "grad_norm": 0.4165448844432831, - "learning_rate": 8.729155060878533e-05, - "loss": 0.0162, + "grad_norm": 0.4318673610687256, + "learning_rate": 9.955992617293836e-05, + "loss": 0.0204, "step": 5410 }, { - "grad_norm": 0.3976683020591736, - "learning_rate": 8.723642793486809e-05, - "loss": 0.0155, + "grad_norm": 0.41563689708709717, + "learning_rate": 9.955627043360358e-05, + "loss": 0.02, "step": 5420 }, { - "grad_norm": 0.38279062509536743, - "learning_rate": 8.718120345797873e-05, - "loss": 0.0187, + "grad_norm": 0.4354114234447479, + "learning_rate": 9.955259964038231e-05, + "loss": 0.0231, "step": 5430 }, { - "grad_norm": 0.3204955458641052, - "learning_rate": 8.712587732909889e-05, - "loss": 0.0173, + "grad_norm": 0.345611572265625, + "learning_rate": 9.954891379438962e-05, + "loss": 0.0206, "step": 5440 }, { - "grad_norm": 0.4341621398925781, - "learning_rate": 8.707044969948806e-05, - "loss": 0.0164, + "grad_norm": 0.5379743576049805, + "learning_rate": 9.954521289674519e-05, + "loss": 0.0203, "step": 5450 }, { - "grad_norm": 0.38356325030326843, - "learning_rate": 8.701492072068329e-05, - "loss": 0.0171, + "grad_norm": 0.43512454628944397, + "learning_rate": 9.954149694857325e-05, + "loss": 0.0216, "step": 5460 }, { - "grad_norm": 0.45178765058517456, - "learning_rate": 8.695929054449869e-05, - "loss": 0.0159, + "grad_norm": 0.4377577006816864, + "learning_rate": 9.953776595100258e-05, + "loss": 0.0204, "step": 5470 }, { - "grad_norm": 0.34176012873649597, - "learning_rate": 8.690355932302501e-05, - "loss": 0.0191, + "grad_norm": 0.4804694950580597, + "learning_rate": 9.95340199051666e-05, + "loss": 0.0231, "step": 5480 }, { - "grad_norm": 0.4089793562889099, - "learning_rate": 8.684772720862931e-05, - "loss": 0.0171, + "grad_norm": 0.48580053448677063, + "learning_rate": 9.953025881220325e-05, + "loss": 0.0211, "step": 5490 }, { - "grad_norm": 0.48740607500076294, - "learning_rate": 8.679179435395446e-05, - "loss": 0.0156, + "grad_norm": 0.4967275559902191, + "learning_rate": 9.952648267325504e-05, + "loss": 0.0192, "step": 5500 }, { - "grad_norm": 0.45742067694664, - "learning_rate": 8.673576091191874e-05, - "loss": 0.018, + "grad_norm": 0.5465275645256042, + "learning_rate": 9.952269148946905e-05, + "loss": 0.0218, "step": 5510 }, { - "grad_norm": 0.334735244512558, - "learning_rate": 8.667962703571541e-05, - "loss": 0.0194, + "grad_norm": 0.46749478578567505, + "learning_rate": 9.951888526199697e-05, + "loss": 0.0234, "step": 5520 }, { - "grad_norm": 0.40391767024993896, - "learning_rate": 8.662339287881238e-05, - "loss": 0.0168, + "grad_norm": 0.4764584004878998, + "learning_rate": 9.951506399199501e-05, + "loss": 0.0203, "step": 5530 }, { - "grad_norm": 0.46624624729156494, - "learning_rate": 8.656705859495169e-05, - "loss": 0.0178, + "grad_norm": 0.47356438636779785, + "learning_rate": 9.951122768062399e-05, + "loss": 0.0216, "step": 5540 }, { - "grad_norm": 0.33784207701683044, - "learning_rate": 8.651062433814912e-05, - "loss": 0.0173, + "grad_norm": 0.40681713819503784, + "learning_rate": 9.950737632904927e-05, + "loss": 0.0212, "step": 5550 }, { - "grad_norm": 0.4161977469921112, - "learning_rate": 8.645409026269375e-05, - "loss": 0.0162, + "grad_norm": 0.43655475974082947, + "learning_rate": 9.950350993844077e-05, + "loss": 0.02, "step": 5560 }, { - "grad_norm": 0.4053378701210022, - "learning_rate": 8.639745652314759e-05, - "loss": 0.0168, + "grad_norm": 0.5888035893440247, + "learning_rate": 9.949962850997303e-05, + "loss": 0.0214, "step": 5570 }, { - "grad_norm": 0.3393418490886688, - "learning_rate": 8.634072327434515e-05, - "loss": 0.0149, + "grad_norm": 0.5284451246261597, + "learning_rate": 9.949573204482512e-05, + "loss": 0.0196, "step": 5580 }, { - "grad_norm": 0.40152478218078613, - "learning_rate": 8.628389067139294e-05, - "loss": 0.0163, + "grad_norm": 0.5056226253509521, + "learning_rate": 9.949182054418064e-05, + "loss": 0.021, "step": 5590 }, { - "grad_norm": 0.3247140347957611, - "learning_rate": 8.622695886966911e-05, - "loss": 0.0154, + "grad_norm": 0.42620813846588135, + "learning_rate": 9.948789400922787e-05, + "loss": 0.0197, "step": 5600 }, { - "grad_norm": 0.48909687995910645, - "learning_rate": 8.616992802482308e-05, - "loss": 0.0143, + "grad_norm": 0.43998464941978455, + "learning_rate": 9.948395244115953e-05, + "loss": 0.0187, "step": 5610 }, { - "grad_norm": 0.4496266841888428, - "learning_rate": 8.611279829277496e-05, - "loss": 0.0148, + "grad_norm": 0.5457742214202881, + "learning_rate": 9.9479995841173e-05, + "loss": 0.0189, "step": 5620 }, { - "grad_norm": 0.4135376513004303, - "learning_rate": 8.605556982971528e-05, - "loss": 0.0142, + "grad_norm": 0.5391003489494324, + "learning_rate": 9.947602421047017e-05, + "loss": 0.0186, "step": 5630 }, { - "grad_norm": 0.47809404134750366, - "learning_rate": 8.599824279210447e-05, - "loss": 0.0142, + "grad_norm": 0.43806758522987366, + "learning_rate": 9.947203755025753e-05, + "loss": 0.0177, "step": 5640 }, { - "grad_norm": 0.5461645126342773, - "learning_rate": 8.594081733667243e-05, - "loss": 0.0149, + "grad_norm": 0.4712775647640228, + "learning_rate": 9.946803586174611e-05, + "loss": 0.0176, "step": 5650 }, { - "grad_norm": 0.358334481716156, - "learning_rate": 8.58832936204182e-05, - "loss": 0.0227, + "grad_norm": 0.49622803926467896, + "learning_rate": 9.946401914615151e-05, + "loss": 0.0262, "step": 5660 }, { - "grad_norm": 0.5356438755989075, - "learning_rate": 8.582567180060942e-05, - "loss": 0.0153, + "grad_norm": 0.652975857257843, + "learning_rate": 9.945998740469394e-05, + "loss": 0.0185, "step": 5670 }, { - "grad_norm": 0.5985286831855774, - "learning_rate": 8.576795203478194e-05, - "loss": 0.0173, + "grad_norm": 0.4902481436729431, + "learning_rate": 9.945594063859809e-05, + "loss": 0.0224, "step": 5680 }, { - "grad_norm": 0.44569990038871765, - "learning_rate": 8.571013448073939e-05, - "loss": 0.0193, + "grad_norm": 0.3783179223537445, + "learning_rate": 9.94518788490933e-05, + "loss": 0.0179, "step": 5690 }, { - "grad_norm": 0.4444361627101898, - "learning_rate": 8.565221929655275e-05, - "loss": 0.0216, + "grad_norm": 0.5471696257591248, + "learning_rate": 9.944780203741341e-05, + "loss": 0.0202, "step": 5700 }, { - "grad_norm": 0.3684118986129761, - "learning_rate": 8.559420664055992e-05, - "loss": 0.0207, + "grad_norm": 0.4779660701751709, + "learning_rate": 9.944371020479686e-05, + "loss": 0.019, "step": 5710 }, { - "grad_norm": 0.45600274205207825, - "learning_rate": 8.553609667136532e-05, - "loss": 0.0202, + "grad_norm": 0.428937703371048, + "learning_rate": 9.943960335248662e-05, + "loss": 0.0209, "step": 5720 }, { - "grad_norm": 0.45347586274147034, - "learning_rate": 8.547788954783936e-05, - "loss": 0.02, + "grad_norm": 0.39229315519332886, + "learning_rate": 9.943548148173027e-05, + "loss": 0.0189, "step": 5730 }, { - "grad_norm": 0.477226585149765, - "learning_rate": 8.541958542911808e-05, - "loss": 0.0232, + "grad_norm": 0.5156890749931335, + "learning_rate": 9.943134459377992e-05, + "loss": 0.0226, "step": 5740 }, { - "grad_norm": 0.4014797508716583, - "learning_rate": 8.536118447460275e-05, - "loss": 0.0206, + "grad_norm": 0.3894418478012085, + "learning_rate": 9.942719268989222e-05, + "loss": 0.0221, "step": 5750 }, { - "grad_norm": 0.3849544823169708, - "learning_rate": 8.530268684395932e-05, - "loss": 0.0188, + "grad_norm": 0.3629615008831024, + "learning_rate": 9.942302577132844e-05, + "loss": 0.019, "step": 5760 }, { - "grad_norm": 0.43162935972213745, - "learning_rate": 8.524409269711807e-05, - "loss": 0.0216, + "grad_norm": 0.42278769612312317, + "learning_rate": 9.941884383935438e-05, + "loss": 0.0217, "step": 5770 }, { - "grad_norm": 0.3942539691925049, - "learning_rate": 8.51854021942732e-05, - "loss": 0.0196, + "grad_norm": 0.4613569974899292, + "learning_rate": 9.941464689524039e-05, + "loss": 0.0214, "step": 5780 }, { - "grad_norm": 0.4294084906578064, - "learning_rate": 8.512661549588227e-05, - "loss": 0.0192, + "grad_norm": 0.4308469295501709, + "learning_rate": 9.941043494026139e-05, + "loss": 0.0219, "step": 5790 }, { - "grad_norm": 0.45174744725227356, - "learning_rate": 8.506773276266588e-05, - "loss": 0.0182, + "grad_norm": 0.43162721395492554, + "learning_rate": 9.940620797569685e-05, + "loss": 0.0207, "step": 5800 }, { - "grad_norm": 0.39539965987205505, - "learning_rate": 8.500875415560721e-05, - "loss": 0.016, + "grad_norm": 0.4408177435398102, + "learning_rate": 9.940196600283082e-05, + "loss": 0.0187, "step": 5810 }, { - "grad_norm": 0.46286967396736145, - "learning_rate": 8.494967983595144e-05, - "loss": 0.0175, + "grad_norm": 0.46690818667411804, + "learning_rate": 9.939770902295192e-05, + "loss": 0.0207, "step": 5820 }, { - "grad_norm": 0.46430596709251404, - "learning_rate": 8.489050996520558e-05, - "loss": 0.0182, + "grad_norm": 0.5194782614707947, + "learning_rate": 9.939343703735329e-05, + "loss": 0.0216, "step": 5830 }, { - "grad_norm": 0.45418673753738403, - "learning_rate": 8.483124470513775e-05, - "loss": 0.0173, + "grad_norm": 0.4704713225364685, + "learning_rate": 9.938915004733264e-05, + "loss": 0.0202, "step": 5840 }, { - "grad_norm": 0.34853413701057434, - "learning_rate": 8.477188421777692e-05, - "loss": 0.0172, + "grad_norm": 0.430652916431427, + "learning_rate": 9.938484805419224e-05, + "loss": 0.0204, "step": 5850 }, { - "grad_norm": 0.39042121171951294, - "learning_rate": 8.47124286654124e-05, - "loss": 0.018, + "grad_norm": 0.43421781063079834, + "learning_rate": 9.938053105923894e-05, + "loss": 0.0215, "step": 5860 }, { - "grad_norm": 0.40436068177223206, - "learning_rate": 8.465287821059341e-05, - "loss": 0.0186, + "grad_norm": 0.45059508085250854, + "learning_rate": 9.937619906378413e-05, + "loss": 0.0226, "step": 5870 }, { - "grad_norm": 0.46700358390808105, - "learning_rate": 8.45932330161286e-05, - "loss": 0.017, + "grad_norm": 0.47542697191238403, + "learning_rate": 9.937185206914374e-05, + "loss": 0.0207, "step": 5880 }, { - "grad_norm": 0.42868727445602417, - "learning_rate": 8.453349324508567e-05, - "loss": 0.015, + "grad_norm": 0.3854386806488037, + "learning_rate": 9.936749007663829e-05, + "loss": 0.0184, "step": 5890 }, { - "grad_norm": 0.49199414253234863, - "learning_rate": 8.447365906079088e-05, - "loss": 0.0168, + "grad_norm": 0.583702564239502, + "learning_rate": 9.93631130875928e-05, + "loss": 0.0204, "step": 5900 }, { - "grad_norm": 0.35933196544647217, - "learning_rate": 8.441373062682856e-05, - "loss": 0.017, + "grad_norm": 0.401313841342926, + "learning_rate": 9.935872110333692e-05, + "loss": 0.0205, "step": 5910 }, { - "grad_norm": 0.34801632165908813, - "learning_rate": 8.43537081070408e-05, - "loss": 0.0151, + "grad_norm": 0.37732264399528503, + "learning_rate": 9.935431412520484e-05, + "loss": 0.0183, "step": 5920 }, { - "grad_norm": 0.3106619715690613, - "learning_rate": 8.429359166552689e-05, - "loss": 0.0131, + "grad_norm": 0.46058157086372375, + "learning_rate": 9.934989215453523e-05, + "loss": 0.0161, "step": 5930 }, { - "grad_norm": 0.3804507255554199, - "learning_rate": 8.423338146664284e-05, - "loss": 0.0173, + "grad_norm": 0.46219077706336975, + "learning_rate": 9.934545519267139e-05, + "loss": 0.0201, "step": 5940 }, { - "grad_norm": 0.3459800183773041, - "learning_rate": 8.417307767500107e-05, - "loss": 0.0133, + "grad_norm": 0.41395333409309387, + "learning_rate": 9.934100324096117e-05, + "loss": 0.0165, "step": 5950 }, { - "grad_norm": 0.40529242157936096, - "learning_rate": 8.411268045546983e-05, - "loss": 0.0134, + "grad_norm": 0.4123416543006897, + "learning_rate": 9.933653630075692e-05, + "loss": 0.0166, "step": 5960 }, { - "grad_norm": 0.40981313586235046, - "learning_rate": 8.405218997317281e-05, - "loss": 0.0172, + "grad_norm": 0.4856509566307068, + "learning_rate": 9.93320543734156e-05, + "loss": 0.0207, "step": 5970 }, { - "grad_norm": 0.36945050954818726, - "learning_rate": 8.399160639348869e-05, - "loss": 0.0148, + "grad_norm": 0.4036096930503845, + "learning_rate": 9.932755746029871e-05, + "loss": 0.0184, "step": 5980 }, { - "grad_norm": 0.48703187704086304, - "learning_rate": 8.393092988205065e-05, - "loss": 0.0172, + "grad_norm": 0.5678353309631348, + "learning_rate": 9.932304556277228e-05, + "loss": 0.021, "step": 5990 }, { - "grad_norm": 0.45176467299461365, - "learning_rate": 8.387016060474597e-05, - "loss": 0.0182, + "grad_norm": 0.5616645812988281, + "learning_rate": 9.93185186822069e-05, + "loss": 0.022, "step": 6000 }, { - "grad_norm": 0.3710270822048187, - "learning_rate": 8.380929872771551e-05, - "loss": 0.0139, + "grad_norm": 0.45179468393325806, + "learning_rate": 9.931397681997773e-05, + "loss": 0.0172, "step": 6010 }, { - "grad_norm": 0.423194020986557, - "learning_rate": 8.374834441735335e-05, - "loss": 0.0141, + "grad_norm": 0.4580582082271576, + "learning_rate": 9.930941997746446e-05, + "loss": 0.0167, "step": 6020 }, { - "grad_norm": 0.40545573830604553, - "learning_rate": 8.368729784030622e-05, - "loss": 0.0136, + "grad_norm": 0.3947714865207672, + "learning_rate": 9.930484815605134e-05, + "loss": 0.0163, "step": 6030 }, { - "grad_norm": 0.2971954345703125, - "learning_rate": 8.362615916347315e-05, - "loss": 0.0158, + "grad_norm": 0.3337259590625763, + "learning_rate": 9.930026135712717e-05, + "loss": 0.0186, "step": 6040 }, { - "grad_norm": 0.3459785580635071, - "learning_rate": 8.356492855400493e-05, - "loss": 0.0151, + "grad_norm": 0.33686500787734985, + "learning_rate": 9.92956595820853e-05, + "loss": 0.0181, "step": 6050 }, { - "grad_norm": 0.3283204734325409, - "learning_rate": 8.350360617930371e-05, - "loss": 0.0135, + "grad_norm": 0.39745450019836426, + "learning_rate": 9.929104283232362e-05, + "loss": 0.0167, "step": 6060 }, { - "grad_norm": 0.4347304701805115, - "learning_rate": 8.344219220702255e-05, - "loss": 0.0155, + "grad_norm": 0.41633832454681396, + "learning_rate": 9.92864111092446e-05, + "loss": 0.0187, "step": 6070 }, { - "grad_norm": 0.4429846405982971, - "learning_rate": 8.338068680506485e-05, - "loss": 0.0165, + "grad_norm": 0.48885807394981384, + "learning_rate": 9.92817644142552e-05, + "loss": 0.0199, "step": 6080 }, { - "grad_norm": 0.45707541704177856, - "learning_rate": 8.33190901415841e-05, - "loss": 0.0132, + "grad_norm": 0.4180903434753418, + "learning_rate": 9.927710274876698e-05, + "loss": 0.0161, "step": 6090 }, { - "grad_norm": 0.4395197927951813, - "learning_rate": 8.325740238498317e-05, - "loss": 0.0168, + "grad_norm": 0.3967433571815491, + "learning_rate": 9.927242611419603e-05, + "loss": 0.0203, "step": 6100 }, { - "grad_norm": 0.4128543436527252, - "learning_rate": 8.319562370391406e-05, - "loss": 0.0157, + "grad_norm": 0.4237320125102997, + "learning_rate": 9.926773451196301e-05, + "loss": 0.0197, "step": 6110 }, { - "grad_norm": 0.4431177079677582, - "learning_rate": 8.31337542672773e-05, - "loss": 0.0195, + "grad_norm": 0.5263776183128357, + "learning_rate": 9.926302794349306e-05, + "loss": 0.0239, "step": 6120 }, { - "grad_norm": 0.4641300141811371, - "learning_rate": 8.307179424422158e-05, - "loss": 0.0164, + "grad_norm": 0.5497742891311646, + "learning_rate": 9.925830641021594e-05, + "loss": 0.0242, "step": 6130 }, { - "grad_norm": 0.4123786687850952, - "learning_rate": 8.300974380414327e-05, - "loss": 0.0147, + "grad_norm": 0.4678046405315399, + "learning_rate": 9.925356991356593e-05, + "loss": 0.0244, "step": 6140 }, { - "grad_norm": 0.42310845851898193, - "learning_rate": 8.294760311668586e-05, - "loss": 0.0184, + "grad_norm": 0.5880439877510071, + "learning_rate": 9.924881845498184e-05, + "loss": 0.0263, "step": 6150 }, { - "grad_norm": 0.4000084102153778, - "learning_rate": 8.288537235173961e-05, - "loss": 0.0149, + "grad_norm": 0.47017771005630493, + "learning_rate": 9.924405203590705e-05, + "loss": 0.0254, "step": 6160 }, { - "grad_norm": 0.4384673237800598, - "learning_rate": 8.282305167944108e-05, - "loss": 0.0174, + "grad_norm": 0.4536885619163513, + "learning_rate": 9.923927065778946e-05, + "loss": 0.0301, "step": 6170 }, { - "grad_norm": 0.2722078263759613, - "learning_rate": 8.276064127017262e-05, - "loss": 0.0154, + "grad_norm": 0.3195449411869049, + "learning_rate": 9.923447432208154e-05, + "loss": 0.0256, "step": 6180 }, { - "grad_norm": 0.31681233644485474, - "learning_rate": 8.269814129456189e-05, - "loss": 0.0139, + "grad_norm": 0.34646308422088623, + "learning_rate": 9.922966303024027e-05, + "loss": 0.0216, "step": 6190 }, { - "grad_norm": 0.3676801919937134, - "learning_rate": 8.263555192348143e-05, - "loss": 0.0145, + "grad_norm": 0.4663931131362915, + "learning_rate": 9.922483678372721e-05, + "loss": 0.0211, "step": 6200 }, { - "grad_norm": 0.38583555817604065, - "learning_rate": 8.257287332804819e-05, - "loss": 0.0148, + "grad_norm": 0.35806676745414734, + "learning_rate": 9.921999558400845e-05, + "loss": 0.0208, "step": 6210 }, { - "grad_norm": 0.38549646735191345, - "learning_rate": 8.251010567962307e-05, - "loss": 0.0157, + "grad_norm": 0.39478111267089844, + "learning_rate": 9.92151394325546e-05, + "loss": 0.0213, "step": 6220 }, { - "grad_norm": 0.38993483781814575, - "learning_rate": 8.244724914981041e-05, - "loss": 0.0151, + "grad_norm": 0.49016234278678894, + "learning_rate": 9.921026833084084e-05, + "loss": 0.0202, "step": 6230 }, { - "grad_norm": 0.4005882441997528, - "learning_rate": 8.238430391045757e-05, - "loss": 0.0148, + "grad_norm": 0.37187206745147705, + "learning_rate": 9.920538228034689e-05, + "loss": 0.0191, "step": 6240 }, { - "grad_norm": 0.34108078479766846, - "learning_rate": 8.232127013365445e-05, - "loss": 0.0176, + "grad_norm": 0.3741544187068939, + "learning_rate": 9.920048128255699e-05, + "loss": 0.0226, "step": 6250 }, { - "grad_norm": 0.4081633985042572, - "learning_rate": 8.225814799173295e-05, - "loss": 0.0181, + "grad_norm": 0.4790568947792053, + "learning_rate": 9.919556533895995e-05, + "loss": 0.0224, "step": 6260 }, { - "grad_norm": 0.24599479138851166, - "learning_rate": 8.219493765726663e-05, - "loss": 0.0136, + "grad_norm": 0.347062885761261, + "learning_rate": 9.919063445104907e-05, + "loss": 0.0177, "step": 6270 }, { - "grad_norm": 0.4071696698665619, - "learning_rate": 8.21316393030701e-05, - "loss": 0.0159, + "grad_norm": 0.42166146636009216, + "learning_rate": 9.918568862032227e-05, + "loss": 0.0204, "step": 6280 }, { - "grad_norm": 0.41659772396087646, - "learning_rate": 8.206825310219865e-05, - "loss": 0.0133, + "grad_norm": 0.4601902365684509, + "learning_rate": 9.918072784828194e-05, + "loss": 0.0175, "step": 6290 }, { - "grad_norm": 0.4144860804080963, - "learning_rate": 8.200477922794776e-05, - "loss": 0.0178, + "grad_norm": 0.4728532135486603, + "learning_rate": 9.917575213643501e-05, + "loss": 0.0221, "step": 6300 }, { - "grad_norm": 0.3404088318347931, - "learning_rate": 8.194121785385256e-05, - "loss": 0.016, + "grad_norm": 0.42424672842025757, + "learning_rate": 9.917076148629302e-05, + "loss": 0.0195, "step": 6310 }, { - "grad_norm": 0.4269247055053711, - "learning_rate": 8.187756915368741e-05, - "loss": 0.0144, + "grad_norm": 0.4475298225879669, + "learning_rate": 9.916575589937196e-05, + "loss": 0.0178, "step": 6320 }, { - "grad_norm": 0.34299951791763306, - "learning_rate": 8.181383330146544e-05, - "loss": 0.0161, + "grad_norm": 0.45007506012916565, + "learning_rate": 9.916073537719239e-05, + "loss": 0.0193, "step": 6330 }, { - "grad_norm": 0.27657780051231384, - "learning_rate": 8.175001047143804e-05, - "loss": 0.0172, + "grad_norm": 0.3647649884223938, + "learning_rate": 9.915569992127944e-05, + "loss": 0.0199, "step": 6340 }, { - "grad_norm": 0.4395776391029358, - "learning_rate": 8.168610083809438e-05, - "loss": 0.0173, + "grad_norm": 0.517101526260376, + "learning_rate": 9.915064953316273e-05, + "loss": 0.0237, "step": 6350 }, { - "grad_norm": 0.41414085030555725, - "learning_rate": 8.162210457616095e-05, - "loss": 0.0182, + "grad_norm": 0.46937012672424316, + "learning_rate": 9.914558421437645e-05, + "loss": 0.0259, "step": 6360 }, { - "grad_norm": 0.3814326822757721, - "learning_rate": 8.155802186060109e-05, - "loss": 0.0171, + "grad_norm": 0.4201948046684265, + "learning_rate": 9.914050396645929e-05, + "loss": 0.0227, "step": 6370 }, { - "grad_norm": 0.3817397654056549, - "learning_rate": 8.149385286661453e-05, - "loss": 0.0174, + "grad_norm": 0.4780450463294983, + "learning_rate": 9.913540879095452e-05, + "loss": 0.0237, "step": 6380 }, { - "grad_norm": 0.40936893224716187, - "learning_rate": 8.14295977696368e-05, - "loss": 0.0171, + "grad_norm": 0.4860870838165283, + "learning_rate": 9.913029868940987e-05, + "loss": 0.0239, "step": 6390 }, { - "grad_norm": 0.36244097352027893, - "learning_rate": 8.13652567453389e-05, - "loss": 0.0134, + "grad_norm": 0.4339703917503357, + "learning_rate": 9.912517366337772e-05, + "loss": 0.019, "step": 6400 }, { - "grad_norm": 0.4146624803543091, - "learning_rate": 8.130082996962676e-05, - "loss": 0.0175, + "grad_norm": 0.43234312534332275, + "learning_rate": 9.912003371441487e-05, + "loss": 0.0235, "step": 6410 }, { - "grad_norm": 0.4154786169528961, - "learning_rate": 8.123631761864068e-05, - "loss": 0.0178, + "grad_norm": 0.3167775869369507, + "learning_rate": 9.911487884408271e-05, + "loss": 0.0225, "step": 6420 }, { - "grad_norm": 0.4006327986717224, - "learning_rate": 8.1171719868755e-05, - "loss": 0.0142, + "grad_norm": 0.4893382787704468, + "learning_rate": 9.910970905394719e-05, + "loss": 0.0181, "step": 6430 }, { - "grad_norm": 0.45199131965637207, - "learning_rate": 8.110703689657748e-05, - "loss": 0.0164, + "grad_norm": 0.4834316074848175, + "learning_rate": 9.91045243455787e-05, + "loss": 0.0203, "step": 6440 }, { - "grad_norm": 0.35770922899246216, - "learning_rate": 8.104226887894892e-05, - "loss": 0.0147, + "grad_norm": 0.3625202476978302, + "learning_rate": 9.909932472055225e-05, + "loss": 0.0186, "step": 6450 }, { - "grad_norm": 0.30330339074134827, - "learning_rate": 8.097741599294257e-05, - "loss": 0.0143, + "grad_norm": 0.3890365660190582, + "learning_rate": 9.909411018044734e-05, + "loss": 0.018, "step": 6460 }, { - "grad_norm": 0.32697442173957825, - "learning_rate": 8.091247841586378e-05, - "loss": 0.0143, + "grad_norm": 0.3136729300022125, + "learning_rate": 9.908888072684802e-05, + "loss": 0.0177, "step": 6470 }, { - "grad_norm": 0.3217572867870331, - "learning_rate": 8.084745632524939e-05, - "loss": 0.016, + "grad_norm": 0.34689581394195557, + "learning_rate": 9.908363636134285e-05, + "loss": 0.0194, "step": 6480 }, { - "grad_norm": 0.42020222544670105, - "learning_rate": 8.07823498988673e-05, - "loss": 0.017, + "grad_norm": 0.46506989002227783, + "learning_rate": 9.907837708552493e-05, + "loss": 0.02, "step": 6490 }, { - "grad_norm": 0.3531644642353058, - "learning_rate": 8.071715931471602e-05, - "loss": 0.0144, + "grad_norm": 0.42716407775878906, + "learning_rate": 9.90731029009919e-05, + "loss": 0.0173, "step": 6500 }, { - "grad_norm": 0.33470600843429565, - "learning_rate": 8.06518847510241e-05, - "loss": 0.0135, + "grad_norm": 0.45493918657302856, + "learning_rate": 9.906781380934589e-05, + "loss": 0.017, "step": 6510 }, { - "grad_norm": 0.34689798951148987, - "learning_rate": 8.058652638624971e-05, - "loss": 0.0139, + "grad_norm": 0.35338613390922546, + "learning_rate": 9.906250981219362e-05, + "loss": 0.0169, "step": 6520 }, { - "grad_norm": 0.32701587677001953, - "learning_rate": 8.052108439908013e-05, - "loss": 0.0158, + "grad_norm": 0.38390281796455383, + "learning_rate": 9.905719091114628e-05, + "loss": 0.0179, "step": 6530 }, { - "grad_norm": 0.38156113028526306, - "learning_rate": 8.045555896843125e-05, - "loss": 0.0155, + "grad_norm": 0.37599876523017883, + "learning_rate": 9.905185710781964e-05, + "loss": 0.0184, "step": 6540 }, { - "grad_norm": 0.448223352432251, - "learning_rate": 8.03899502734471e-05, - "loss": 0.0157, + "grad_norm": 0.4138750731945038, + "learning_rate": 9.904650840383392e-05, + "loss": 0.0183, "step": 6550 }, { - "grad_norm": 0.5395865440368652, - "learning_rate": 8.032425849349931e-05, - "loss": 0.0158, + "grad_norm": 0.4354017674922943, + "learning_rate": 9.904114480081397e-05, + "loss": 0.0189, "step": 6560 }, { - "grad_norm": 0.5085408091545105, - "learning_rate": 8.025848380818674e-05, - "loss": 0.0194, + "grad_norm": 0.5025765299797058, + "learning_rate": 9.903576630038906e-05, + "loss": 0.0221, "step": 6570 }, { - "grad_norm": 0.3302910625934601, - "learning_rate": 8.019262639733487e-05, - "loss": 0.0141, + "grad_norm": 0.432136207818985, + "learning_rate": 9.903037290419309e-05, + "loss": 0.0173, "step": 6580 }, { - "grad_norm": 0.5127670764923096, - "learning_rate": 8.012668644099531e-05, - "loss": 0.0159, + "grad_norm": 0.486317902803421, + "learning_rate": 9.902496461386439e-05, + "loss": 0.0185, "step": 6590 }, { - "grad_norm": 0.36805611848831177, - "learning_rate": 8.006066411944542e-05, - "loss": 0.0183, + "grad_norm": 0.5418636798858643, + "learning_rate": 9.901954143104588e-05, + "loss": 0.0222, "step": 6600 }, { - "grad_norm": 0.35506105422973633, - "learning_rate": 7.999455961318769e-05, - "loss": 0.0196, + "grad_norm": 0.36305558681488037, + "learning_rate": 9.901410335738496e-05, + "loss": 0.0228, "step": 6610 }, { - "grad_norm": 0.3373585343360901, - "learning_rate": 7.992837310294932e-05, - "loss": 0.0125, + "grad_norm": 0.42342790961265564, + "learning_rate": 9.900865039453358e-05, + "loss": 0.0155, "step": 6620 }, { - "grad_norm": 0.4130808413028717, - "learning_rate": 7.986210476968167e-05, - "loss": 0.0174, + "grad_norm": 0.4071841537952423, + "learning_rate": 9.900318254414821e-05, + "loss": 0.0213, "step": 6630 }, { - "grad_norm": 0.455818772315979, - "learning_rate": 7.97957547945599e-05, - "loss": 0.0161, + "grad_norm": 0.39720529317855835, + "learning_rate": 9.899769980788985e-05, + "loss": 0.0195, "step": 6640 }, { - "grad_norm": 0.35132884979248047, - "learning_rate": 7.972932335898226e-05, - "loss": 0.0138, + "grad_norm": 0.33522626757621765, + "learning_rate": 9.899220218742398e-05, + "loss": 0.0165, "step": 6650 }, { - "grad_norm": 0.39049220085144043, - "learning_rate": 7.966281064456975e-05, - "loss": 0.0164, + "grad_norm": 0.3815174698829651, + "learning_rate": 9.898668968442066e-05, + "loss": 0.0198, "step": 6660 }, { - "grad_norm": 0.46879708766937256, - "learning_rate": 7.959621683316563e-05, - "loss": 0.0172, + "grad_norm": 0.45704084634780884, + "learning_rate": 9.898116230055443e-05, + "loss": 0.0203, "step": 6670 }, { - "grad_norm": 0.4622355103492737, - "learning_rate": 7.952954210683481e-05, - "loss": 0.0158, + "grad_norm": 0.4906638562679291, + "learning_rate": 9.897562003750437e-05, + "loss": 0.0185, "step": 6680 }, { - "grad_norm": 0.35040467977523804, - "learning_rate": 7.946278664786345e-05, - "loss": 0.015, + "grad_norm": 0.4390787184238434, + "learning_rate": 9.897006289695407e-05, + "loss": 0.0187, "step": 6690 }, { - "grad_norm": 0.36279842257499695, - "learning_rate": 7.939595063875842e-05, - "loss": 0.0144, + "grad_norm": 0.377570241689682, + "learning_rate": 9.896449088059164e-05, + "loss": 0.0179, "step": 6700 }, { - "grad_norm": 0.3879755735397339, - "learning_rate": 7.932903426224683e-05, - "loss": 0.015, + "grad_norm": 0.42955902218818665, + "learning_rate": 9.89589039901097e-05, + "loss": 0.0179, "step": 6710 }, { - "grad_norm": 0.33389654755592346, - "learning_rate": 7.926203770127552e-05, - "loss": 0.0155, + "grad_norm": 0.38894832134246826, + "learning_rate": 9.895330222720542e-05, + "loss": 0.0184, "step": 6720 }, { - "grad_norm": 0.36863645911216736, - "learning_rate": 7.919496113901046e-05, - "loss": 0.0125, + "grad_norm": 0.43698614835739136, + "learning_rate": 9.894768559358047e-05, + "loss": 0.0153, "step": 6730 }, { - "grad_norm": 0.33035746216773987, - "learning_rate": 7.912780475883649e-05, - "loss": 0.0143, + "grad_norm": 0.3228228986263275, + "learning_rate": 9.894205409094101e-05, + "loss": 0.0171, "step": 6740 }, { - "grad_norm": 0.377729594707489, - "learning_rate": 7.906056874435652e-05, - "loss": 0.0125, + "grad_norm": 0.4363706111907959, + "learning_rate": 9.893640772099777e-05, + "loss": 0.0153, "step": 6750 }, { - "grad_norm": 0.311613529920578, - "learning_rate": 7.899325327939131e-05, - "loss": 0.0121, + "grad_norm": 0.3527734875679016, + "learning_rate": 9.893074648546595e-05, + "loss": 0.015, "step": 6760 }, { - "grad_norm": 0.3309403955936432, - "learning_rate": 7.892585854797872e-05, - "loss": 0.0123, + "grad_norm": 0.2969190776348114, + "learning_rate": 9.892507038606528e-05, + "loss": 0.0152, "step": 6770 }, { - "grad_norm": 0.3694356083869934, - "learning_rate": 7.88583847343734e-05, - "loss": 0.0141, + "grad_norm": 0.4266859292984009, + "learning_rate": 9.891937942452003e-05, + "loss": 0.0172, "step": 6780 }, { - "grad_norm": 0.46677365899086, - "learning_rate": 7.879083202304616e-05, - "loss": 0.0134, + "grad_norm": 0.440215528011322, + "learning_rate": 9.891367360255895e-05, + "loss": 0.0165, "step": 6790 }, { - "grad_norm": 0.3483867347240448, - "learning_rate": 7.872320059868355e-05, - "loss": 0.0143, + "grad_norm": 0.41470351815223694, + "learning_rate": 9.890795292191532e-05, + "loss": 0.0177, "step": 6800 }, { - "grad_norm": 0.30982255935668945, - "learning_rate": 7.865549064618729e-05, - "loss": 0.0135, + "grad_norm": 0.3750207722187042, + "learning_rate": 9.890221738432694e-05, + "loss": 0.0168, "step": 6810 }, { - "grad_norm": 0.3763626217842102, - "learning_rate": 7.858770235067381e-05, - "loss": 0.0155, + "grad_norm": 0.41080591082572937, + "learning_rate": 9.88964669915361e-05, + "loss": 0.0189, "step": 6820 }, { - "grad_norm": 0.43991050124168396, - "learning_rate": 7.851983589747374e-05, - "loss": 0.0126, + "grad_norm": 0.40893933176994324, + "learning_rate": 9.889070174528963e-05, + "loss": 0.0153, "step": 6830 }, { - "grad_norm": 0.4573172330856323, - "learning_rate": 7.845189147213133e-05, - "loss": 0.0154, + "grad_norm": 0.4975588023662567, + "learning_rate": 9.888492164733883e-05, + "loss": 0.0181, "step": 6840 }, { - "grad_norm": 0.5023182034492493, - "learning_rate": 7.838386926040407e-05, - "loss": 0.0163, + "grad_norm": 0.4919087290763855, + "learning_rate": 9.88791266994396e-05, + "loss": 0.0197, "step": 6850 }, { - "grad_norm": 0.4336654841899872, - "learning_rate": 7.83157694482621e-05, - "loss": 0.0173, + "grad_norm": 0.49394693970680237, + "learning_rate": 9.887331690335223e-05, + "loss": 0.0212, "step": 6860 }, { - "grad_norm": 0.2743140459060669, - "learning_rate": 7.824759222188768e-05, - "loss": 0.0125, + "grad_norm": 0.3582223653793335, + "learning_rate": 9.886749226084163e-05, + "loss": 0.0163, "step": 6870 }, { - "grad_norm": 0.3396323621273041, - "learning_rate": 7.817933776767478e-05, - "loss": 0.0165, + "grad_norm": 0.41114887595176697, + "learning_rate": 9.886165277367714e-05, + "loss": 0.0197, "step": 6880 }, { - "grad_norm": 0.35954833030700684, - "learning_rate": 7.811100627222842e-05, - "loss": 0.012, + "grad_norm": 0.4360108971595764, + "learning_rate": 9.885579844363265e-05, + "loss": 0.0145, "step": 6890 }, { - "grad_norm": 0.3257628381252289, - "learning_rate": 7.804259792236435e-05, - "loss": 0.0135, + "grad_norm": 0.35609444975852966, + "learning_rate": 9.884992927248656e-05, + "loss": 0.0163, "step": 6900 }, { - "grad_norm": 0.49320679903030396, - "learning_rate": 7.797411290510835e-05, - "loss": 0.0128, + "grad_norm": 0.4820123612880707, + "learning_rate": 9.884404526202178e-05, + "loss": 0.0162, "step": 6910 }, { - "grad_norm": 0.36475473642349243, - "learning_rate": 7.790555140769586e-05, - "loss": 0.0148, + "grad_norm": 0.32953253388404846, + "learning_rate": 9.883814641402568e-05, + "loss": 0.0181, "step": 6920 }, { - "grad_norm": 0.30804768204689026, - "learning_rate": 7.78369136175714e-05, - "loss": 0.0158, + "grad_norm": 0.3404013216495514, + "learning_rate": 9.88322327302902e-05, + "loss": 0.0189, "step": 6930 }, { - "grad_norm": 0.2891436219215393, - "learning_rate": 7.776819972238806e-05, - "loss": 0.0127, + "grad_norm": 0.36602842807769775, + "learning_rate": 9.882630421261176e-05, + "loss": 0.0149, "step": 6940 }, { - "grad_norm": 0.3640880584716797, - "learning_rate": 7.7699409910007e-05, - "loss": 0.0135, + "grad_norm": 0.33201056718826294, + "learning_rate": 9.88203608627913e-05, + "loss": 0.0165, "step": 6950 }, { - "grad_norm": 0.29762402176856995, - "learning_rate": 7.763054436849694e-05, - "loss": 0.0148, + "grad_norm": 0.33390235900878906, + "learning_rate": 9.881440268263422e-05, + "loss": 0.0184, "step": 6960 }, { - "grad_norm": 0.3800078332424164, - "learning_rate": 7.756160328613364e-05, - "loss": 0.0124, + "grad_norm": 0.41959813237190247, + "learning_rate": 9.880842967395048e-05, + "loss": 0.0158, "step": 6970 }, { - "grad_norm": 0.3265116810798645, - "learning_rate": 7.749258685139942e-05, - "loss": 0.0148, + "grad_norm": 0.35842469334602356, + "learning_rate": 9.880244183855452e-05, + "loss": 0.0182, "step": 6980 }, { - "grad_norm": 0.37737640738487244, - "learning_rate": 7.742349525298253e-05, - "loss": 0.0132, + "grad_norm": 0.2936786413192749, + "learning_rate": 9.879643917826527e-05, + "loss": 0.0161, "step": 6990 }, { - "grad_norm": 0.21071003377437592, - "learning_rate": 7.735432867977679e-05, - "loss": 0.0136, + "grad_norm": 0.3380889892578125, + "learning_rate": 9.87904216949062e-05, + "loss": 0.0164, "step": 7000 }, { - "grad_norm": 0.30626237392425537, - "learning_rate": 7.728508732088096e-05, - "loss": 0.0168, + "grad_norm": 0.3669847846031189, + "learning_rate": 9.878438939030526e-05, + "loss": 0.0196, "step": 7010 }, { - "grad_norm": 0.3753643035888672, - "learning_rate": 7.721577136559825e-05, - "loss": 0.0154, + "grad_norm": 0.3744491934776306, + "learning_rate": 9.877834226629489e-05, + "loss": 0.0184, "step": 7020 }, { - "grad_norm": 0.3369854986667633, - "learning_rate": 7.714638100343588e-05, - "loss": 0.0126, + "grad_norm": 0.3672807216644287, + "learning_rate": 9.877228032471206e-05, + "loss": 0.0149, "step": 7030 }, { - "grad_norm": 0.36830559372901917, - "learning_rate": 7.707691642410444e-05, - "loss": 0.014, + "grad_norm": 0.2968101501464844, + "learning_rate": 9.876620356739823e-05, + "loss": 0.0163, "step": 7040 }, { - "grad_norm": 0.4115355908870697, - "learning_rate": 7.70073778175174e-05, - "loss": 0.0148, + "grad_norm": 0.4129272401332855, + "learning_rate": 9.876011199619935e-05, + "loss": 0.0163, "step": 7050 }, { - "grad_norm": 0.2858520448207855, - "learning_rate": 7.69377653737907e-05, - "loss": 0.0139, + "grad_norm": 0.393873393535614, + "learning_rate": 9.875400561296589e-05, + "loss": 0.0163, "step": 7060 }, { - "grad_norm": 0.38406574726104736, - "learning_rate": 7.686807928324209e-05, - "loss": 0.013, + "grad_norm": 0.34575068950653076, + "learning_rate": 9.874788441955278e-05, + "loss": 0.0156, "step": 7070 }, { - "grad_norm": 0.3661177456378937, - "learning_rate": 7.679831973639065e-05, - "loss": 0.0128, + "grad_norm": 0.3829306662082672, + "learning_rate": 9.874174841781951e-05, + "loss": 0.0156, "step": 7080 }, { - "grad_norm": 0.2673167586326599, - "learning_rate": 7.672848692395637e-05, - "loss": 0.0135, + "grad_norm": 0.44204407930374146, + "learning_rate": 9.873559760963003e-05, + "loss": 0.0167, "step": 7090 }, { - "grad_norm": 0.3895997703075409, - "learning_rate": 7.665858103685944e-05, - "loss": 0.0116, + "grad_norm": 0.397582083940506, + "learning_rate": 9.872943199685278e-05, + "loss": 0.0144, "step": 7100 }, { - "grad_norm": 0.3720957934856415, - "learning_rate": 7.658860226621991e-05, - "loss": 0.0144, + "grad_norm": 0.39908432960510254, + "learning_rate": 9.872325158136071e-05, + "loss": 0.017, "step": 7110 }, { - "grad_norm": 0.3137901723384857, - "learning_rate": 7.651855080335708e-05, - "loss": 0.0146, + "grad_norm": 0.43825486302375793, + "learning_rate": 9.871705636503128e-05, + "loss": 0.0171, "step": 7120 }, { - "grad_norm": 0.29563164710998535, - "learning_rate": 7.644842683978896e-05, - "loss": 0.0123, + "grad_norm": 0.34599193930625916, + "learning_rate": 9.871084634974641e-05, + "loss": 0.0152, "step": 7130 }, { - "grad_norm": 0.39437034726142883, - "learning_rate": 7.63782305672318e-05, - "loss": 0.0155, + "grad_norm": 0.43893107771873474, + "learning_rate": 9.870462153739257e-05, + "loss": 0.0182, "step": 7140 }, { - "grad_norm": 0.3885636627674103, - "learning_rate": 7.63079621775995e-05, - "loss": 0.0124, + "grad_norm": 0.3930358290672302, + "learning_rate": 9.869838192986067e-05, + "loss": 0.0157, "step": 7150 }, { - "grad_norm": 0.35217997431755066, - "learning_rate": 7.623762186300319e-05, - "loss": 0.0138, + "grad_norm": 0.4006841480731964, + "learning_rate": 9.869212752904616e-05, + "loss": 0.0161, "step": 7160 }, { - "grad_norm": 0.3057458698749542, - "learning_rate": 7.616720981575057e-05, - "loss": 0.0116, + "grad_norm": 0.3374226689338684, + "learning_rate": 9.868585833684894e-05, + "loss": 0.0141, "step": 7170 }, { - "grad_norm": 0.3529524505138397, - "learning_rate": 7.609672622834552e-05, - "loss": 0.0134, + "grad_norm": 0.37965646386146545, + "learning_rate": 9.867957435517342e-05, + "loss": 0.0163, "step": 7180 }, { - "grad_norm": 0.3752933442592621, - "learning_rate": 7.602617129348747e-05, - "loss": 0.0118, + "grad_norm": 0.4180934429168701, + "learning_rate": 9.867327558592854e-05, + "loss": 0.015, "step": 7190 }, { - "grad_norm": 0.34898602962493896, - "learning_rate": 7.595554520407088e-05, - "loss": 0.0145, + "grad_norm": 0.33238381147384644, + "learning_rate": 9.866696203102766e-05, + "loss": 0.0174, "step": 7200 }, { - "grad_norm": 0.39017370343208313, - "learning_rate": 7.588484815318484e-05, - "loss": 0.0135, + "grad_norm": 0.4209561347961426, + "learning_rate": 9.86606336923887e-05, + "loss": 0.0161, "step": 7210 }, { - "grad_norm": 0.37441468238830566, - "learning_rate": 7.581408033411234e-05, - "loss": 0.0137, + "grad_norm": 0.4806552231311798, + "learning_rate": 9.865429057193403e-05, + "loss": 0.0169, "step": 7220 }, { - "grad_norm": 0.40523502230644226, - "learning_rate": 7.574324194032995e-05, - "loss": 0.0147, + "grad_norm": 0.4034254550933838, + "learning_rate": 9.864793267159053e-05, + "loss": 0.018, "step": 7230 }, { - "grad_norm": 0.2746640145778656, - "learning_rate": 7.567233316550705e-05, - "loss": 0.0118, + "grad_norm": 0.3712209463119507, + "learning_rate": 9.864155999328957e-05, + "loss": 0.0142, "step": 7240 }, { - "grad_norm": 0.3495345115661621, - "learning_rate": 7.560135420350562e-05, - "loss": 0.0117, + "grad_norm": 0.4380336403846741, + "learning_rate": 9.8635172538967e-05, + "loss": 0.0142, "step": 7250 }, { - "grad_norm": 0.4313659965991974, - "learning_rate": 7.553030524837935e-05, - "loss": 0.0128, + "grad_norm": 0.5083938241004944, + "learning_rate": 9.862877031056312e-05, + "loss": 0.0158, "step": 7260 }, { - "grad_norm": 0.4218427538871765, - "learning_rate": 7.545918649437341e-05, - "loss": 0.0121, + "grad_norm": 0.5030562281608582, + "learning_rate": 9.862235331002279e-05, + "loss": 0.0149, "step": 7270 }, { - "grad_norm": 0.37825003266334534, - "learning_rate": 7.538799813592377e-05, - "loss": 0.0107, + "grad_norm": 0.41710764169692993, + "learning_rate": 9.861592153929533e-05, + "loss": 0.0134, "step": 7280 }, { - "grad_norm": 0.36705559492111206, - "learning_rate": 7.531674036765662e-05, - "loss": 0.0134, + "grad_norm": 0.45611754059791565, + "learning_rate": 9.860947500033455e-05, + "loss": 0.0163, "step": 7290 }, { - "grad_norm": 0.3276546001434326, - "learning_rate": 7.524541338438807e-05, - "loss": 0.0145, + "grad_norm": 0.39705395698547363, + "learning_rate": 9.86030136950987e-05, + "loss": 0.0159, "step": 7300 }, { - "grad_norm": 0.45257559418678284, - "learning_rate": 7.517401738112328e-05, - "loss": 0.0138, + "grad_norm": 0.412016361951828, + "learning_rate": 9.85965376255506e-05, + "loss": 0.0148, "step": 7310 }, { - "grad_norm": 0.36200499534606934, - "learning_rate": 7.510255255305628e-05, - "loss": 0.0143, + "grad_norm": 0.44000282883644104, + "learning_rate": 9.859004679365747e-05, + "loss": 0.0157, "step": 7320 }, { - "grad_norm": 0.37191298604011536, - "learning_rate": 7.503101909556911e-05, - "loss": 0.0166, + "grad_norm": 0.3972632884979248, + "learning_rate": 9.858354120139108e-05, + "loss": 0.0181, "step": 7330 }, { - "grad_norm": 0.3438247740268707, - "learning_rate": 7.495941720423154e-05, - "loss": 0.0142, + "grad_norm": 0.3845224380493164, + "learning_rate": 9.857702085072764e-05, + "loss": 0.0161, "step": 7340 }, { - "grad_norm": 0.4581180810928345, - "learning_rate": 7.488774707480042e-05, - "loss": 0.0153, + "grad_norm": 0.5182003974914551, + "learning_rate": 9.857048574364787e-05, + "loss": 0.0177, "step": 7350 }, { - "grad_norm": 0.359670490026474, - "learning_rate": 7.481600890321911e-05, - "loss": 0.0122, + "grad_norm": 0.37718909978866577, + "learning_rate": 9.856393588213698e-05, + "loss": 0.0148, "step": 7360 }, { - "grad_norm": 0.46836525201797485, - "learning_rate": 7.474420288561708e-05, - "loss": 0.0135, + "grad_norm": 0.47215864062309265, + "learning_rate": 9.855737126818458e-05, + "loss": 0.0162, "step": 7370 }, { - "grad_norm": 0.33389872312545776, - "learning_rate": 7.467232921830921e-05, - "loss": 0.0134, + "grad_norm": 0.38670554757118225, + "learning_rate": 9.855079190378491e-05, + "loss": 0.016, "step": 7380 }, { - "grad_norm": 0.33688122034072876, - "learning_rate": 7.460038809779537e-05, - "loss": 0.0136, + "grad_norm": 0.4058266282081604, + "learning_rate": 9.854419779093655e-05, + "loss": 0.0165, "step": 7390 }, { - "grad_norm": 0.41096070408821106, - "learning_rate": 7.452837972075983e-05, - "loss": 0.0127, + "grad_norm": 0.38940101861953735, + "learning_rate": 9.853758893164264e-05, + "loss": 0.0152, "step": 7400 }, { - "grad_norm": 0.41126012802124023, - "learning_rate": 7.445630428407074e-05, - "loss": 0.015, + "grad_norm": 0.41507646441459656, + "learning_rate": 9.853096532791078e-05, + "loss": 0.0178, "step": 7410 }, { - "grad_norm": 0.3702179789543152, - "learning_rate": 7.43841619847796e-05, - "loss": 0.0126, + "grad_norm": 0.3551158905029297, + "learning_rate": 9.852432698175304e-05, + "loss": 0.0151, "step": 7420 }, { - "grad_norm": 0.35743647813796997, - "learning_rate": 7.431195302012072e-05, - "loss": 0.0197, + "grad_norm": 0.4304772615432739, + "learning_rate": 9.851767389518597e-05, + "loss": 0.0224, "step": 7430 }, { - "grad_norm": 0.2985547184944153, - "learning_rate": 7.423967758751061e-05, - "loss": 0.0159, + "grad_norm": 0.39942145347595215, + "learning_rate": 9.85110060702306e-05, + "loss": 0.0183, "step": 7440 }, { - "grad_norm": 0.39072185754776, - "learning_rate": 7.416733588454758e-05, - "loss": 0.0154, + "grad_norm": 0.376549631357193, + "learning_rate": 9.850432350891245e-05, + "loss": 0.0182, "step": 7450 }, { - "grad_norm": 0.4116431474685669, - "learning_rate": 7.409492810901106e-05, - "loss": 0.0163, + "grad_norm": 0.4297804534435272, + "learning_rate": 9.84976262132615e-05, + "loss": 0.0189, "step": 7460 }, { - "grad_norm": 0.4209757149219513, - "learning_rate": 7.402245445886116e-05, - "loss": 0.0197, + "grad_norm": 0.45385608077049255, + "learning_rate": 9.849091418531222e-05, + "loss": 0.0229, "step": 7470 }, { - "grad_norm": 0.348718523979187, - "learning_rate": 7.394991513223806e-05, - "loss": 0.0142, + "grad_norm": 0.35529136657714844, + "learning_rate": 9.848418742710353e-05, + "loss": 0.0175, "step": 7480 }, { - "grad_norm": 0.3451424837112427, - "learning_rate": 7.38773103274615e-05, - "loss": 0.0147, + "grad_norm": 0.397545725107193, + "learning_rate": 9.847744594067885e-05, + "loss": 0.0182, "step": 7490 }, { - "grad_norm": 0.32023975253105164, - "learning_rate": 7.380464024303028e-05, - "loss": 0.0159, + "grad_norm": 0.3288343846797943, + "learning_rate": 9.847068972808607e-05, + "loss": 0.0189, "step": 7500 }, { - "grad_norm": 0.4120250344276428, - "learning_rate": 7.373190507762162e-05, - "loss": 0.0123, + "grad_norm": 0.3809896409511566, + "learning_rate": 9.846391879137756e-05, + "loss": 0.0149, "step": 7510 }, { - "grad_norm": 0.3678809106349945, - "learning_rate": 7.365910503009066e-05, - "loss": 0.0117, + "grad_norm": 0.3377668559551239, + "learning_rate": 9.845713313261012e-05, + "loss": 0.0144, "step": 7520 }, { - "grad_norm": 0.36750277876853943, - "learning_rate": 7.358624029946996e-05, - "loss": 0.0136, + "grad_norm": 0.3758298456668854, + "learning_rate": 9.845033275384505e-05, + "loss": 0.0161, "step": 7530 }, { - "grad_norm": 0.3163510262966156, - "learning_rate": 7.351331108496893e-05, - "loss": 0.014, + "grad_norm": 0.3665297329425812, + "learning_rate": 9.844351765714818e-05, + "loss": 0.0157, "step": 7540 }, { - "grad_norm": 0.4134257435798645, - "learning_rate": 7.344031758597325e-05, - "loss": 0.0133, + "grad_norm": 0.3788328170776367, + "learning_rate": 9.843668784458971e-05, + "loss": 0.0154, "step": 7550 }, { - "grad_norm": 0.392423540353775, - "learning_rate": 7.336726000204435e-05, - "loss": 0.0141, + "grad_norm": 0.5394845008850098, + "learning_rate": 9.842984331824437e-05, + "loss": 0.0169, "step": 7560 }, { - "grad_norm": 0.3356896638870239, - "learning_rate": 7.32941385329189e-05, - "loss": 0.0125, + "grad_norm": 0.31362563371658325, + "learning_rate": 9.842298408019133e-05, + "loss": 0.0146, "step": 7570 }, { - "grad_norm": 0.31061065196990967, - "learning_rate": 7.322095337850816e-05, - "loss": 0.0153, + "grad_norm": 0.32352402806282043, + "learning_rate": 9.841611013251429e-05, + "loss": 0.0178, "step": 7580 }, { - "grad_norm": 0.33278927206993103, - "learning_rate": 7.314770473889758e-05, - "loss": 0.0104, + "grad_norm": 0.35832497477531433, + "learning_rate": 9.840922147730133e-05, + "loss": 0.0127, "step": 7590 }, { - "grad_norm": 0.35223352909088135, - "learning_rate": 7.307439281434615e-05, - "loss": 0.0117, + "grad_norm": 0.41615843772888184, + "learning_rate": 9.840231811664506e-05, + "loss": 0.0139, "step": 7600 }, { - "grad_norm": 0.37879201769828796, - "learning_rate": 7.300101780528585e-05, - "loss": 0.0132, + "grad_norm": 0.3583919405937195, + "learning_rate": 9.839540005264252e-05, + "loss": 0.0155, "step": 7610 }, { - "grad_norm": 0.33135613799095154, - "learning_rate": 7.292757991232117e-05, - "loss": 0.0142, + "grad_norm": 0.45877712965011597, + "learning_rate": 9.838846728739527e-05, + "loss": 0.0174, "step": 7620 }, { - "grad_norm": 0.3964645564556122, - "learning_rate": 7.285407933622848e-05, - "loss": 0.012, + "grad_norm": 0.4497396945953369, + "learning_rate": 9.838151982300927e-05, + "loss": 0.0158, "step": 7630 }, { - "grad_norm": 0.3188134431838989, - "learning_rate": 7.278051627795557e-05, - "loss": 0.0182, + "grad_norm": 0.4297942519187927, + "learning_rate": 9.8374557661595e-05, + "loss": 0.0219, "step": 7640 }, { - "grad_norm": 0.36549797654151917, - "learning_rate": 7.270689093862105e-05, - "loss": 0.0146, + "grad_norm": 0.4039759933948517, + "learning_rate": 9.836758080526735e-05, + "loss": 0.018, "step": 7650 }, { - "grad_norm": 0.3195458948612213, - "learning_rate": 7.263320351951374e-05, - "loss": 0.0105, + "grad_norm": 0.33790767192840576, + "learning_rate": 9.836058925614575e-05, + "loss": 0.0136, "step": 7660 }, { - "grad_norm": 0.3117104768753052, - "learning_rate": 7.255945422209227e-05, - "loss": 0.0127, + "grad_norm": 0.3268868029117584, + "learning_rate": 9.8353583016354e-05, + "loss": 0.0153, "step": 7670 }, { - "grad_norm": 0.31559836864471436, - "learning_rate": 7.248564324798437e-05, - "loss": 0.0139, + "grad_norm": 0.34531736373901367, + "learning_rate": 9.834656208802044e-05, + "loss": 0.0161, "step": 7680 }, { - "grad_norm": 0.2703145742416382, - "learning_rate": 7.241177079898644e-05, - "loss": 0.014, + "grad_norm": 0.2970593571662903, + "learning_rate": 9.833952647327784e-05, + "loss": 0.0167, "step": 7690 }, { - "grad_norm": 0.30897757411003113, - "learning_rate": 7.233783707706295e-05, - "loss": 0.0127, + "grad_norm": 0.3330003619194031, + "learning_rate": 9.833247617426342e-05, + "loss": 0.0153, "step": 7700 }, { - "grad_norm": 0.3173775374889374, - "learning_rate": 7.226384228434586e-05, - "loss": 0.0133, + "grad_norm": 0.3993304669857025, + "learning_rate": 9.832541119311889e-05, + "loss": 0.0165, "step": 7710 }, { - "grad_norm": 0.36799511313438416, - "learning_rate": 7.21897866231341e-05, - "loss": 0.0105, + "grad_norm": 0.43474718928337097, + "learning_rate": 9.83183315319904e-05, + "loss": 0.013, "step": 7720 }, { - "grad_norm": 0.2746727764606476, - "learning_rate": 7.211567029589303e-05, - "loss": 0.0138, + "grad_norm": 0.3155459761619568, + "learning_rate": 9.831123719302855e-05, + "loss": 0.017, "step": 7730 }, { - "grad_norm": 0.3362380266189575, - "learning_rate": 7.204149350525387e-05, - "loss": 0.0123, + "grad_norm": 0.2562251687049866, + "learning_rate": 9.830412817838842e-05, + "loss": 0.0149, "step": 7740 }, { - "grad_norm": 0.33995118737220764, - "learning_rate": 7.196725645401309e-05, - "loss": 0.0127, + "grad_norm": 0.3912941515445709, + "learning_rate": 9.829700449022956e-05, + "loss": 0.0153, "step": 7750 }, { - "grad_norm": 0.332607626914978, - "learning_rate": 7.1892959345132e-05, - "loss": 0.0139, + "grad_norm": 0.3397212624549866, + "learning_rate": 9.828986613071593e-05, + "loss": 0.0161, "step": 7760 }, { - "grad_norm": 0.2877747416496277, - "learning_rate": 7.181860238173605e-05, - "loss": 0.0111, + "grad_norm": 0.33656275272369385, + "learning_rate": 9.828271310201601e-05, + "loss": 0.0133, "step": 7770 }, { - "grad_norm": 0.30644649267196655, - "learning_rate": 7.174418576711432e-05, - "loss": 0.012, + "grad_norm": 0.3815263509750366, + "learning_rate": 9.827554540630268e-05, + "loss": 0.0144, "step": 7780 }, { - "grad_norm": 0.29927390813827515, - "learning_rate": 7.1669709704719e-05, - "loss": 0.0109, + "grad_norm": 0.321825236082077, + "learning_rate": 9.826836304575329e-05, + "loss": 0.0136, "step": 7790 }, { - "grad_norm": 0.3413328528404236, - "learning_rate": 7.159517439816481e-05, - "loss": 0.0143, + "grad_norm": 0.40393421053886414, + "learning_rate": 9.826116602254966e-05, + "loss": 0.0176, "step": 7800 }, { - "grad_norm": 0.24794118106365204, - "learning_rate": 7.152058005122842e-05, - "loss": 0.0148, + "grad_norm": 0.3452446162700653, + "learning_rate": 9.825395433887805e-05, + "loss": 0.0187, "step": 7810 }, { - "grad_norm": 0.30149683356285095, - "learning_rate": 7.144592686784793e-05, - "loss": 0.0118, + "grad_norm": 0.29129523038864136, + "learning_rate": 9.824672799692917e-05, + "loss": 0.0149, "step": 7820 }, { - "grad_norm": 0.33201098442077637, - "learning_rate": 7.137121505212229e-05, - "loss": 0.0111, + "grad_norm": 0.34453850984573364, + "learning_rate": 9.823948699889823e-05, + "loss": 0.0144, "step": 7830 }, { - "grad_norm": 0.40392014384269714, - "learning_rate": 7.129644480831077e-05, - "loss": 0.0126, + "grad_norm": 0.4322754442691803, + "learning_rate": 9.823223134698483e-05, + "loss": 0.016, "step": 7840 }, { - "grad_norm": 0.32815903425216675, - "learning_rate": 7.122161634083234e-05, - "loss": 0.0104, + "grad_norm": 0.36726808547973633, + "learning_rate": 9.822496104339303e-05, + "loss": 0.0141, "step": 7850 }, { - "grad_norm": 0.31929904222488403, - "learning_rate": 7.114672985426516e-05, - "loss": 0.0121, + "grad_norm": 0.4384458661079407, + "learning_rate": 9.821767609033138e-05, + "loss": 0.0153, "step": 7860 }, { - "grad_norm": 0.3497408926486969, - "learning_rate": 7.107178555334606e-05, - "loss": 0.0113, + "grad_norm": 0.4254269003868103, + "learning_rate": 9.821037649001284e-05, + "loss": 0.0142, "step": 7870 }, { - "grad_norm": 0.3377891778945923, - "learning_rate": 7.099678364296989e-05, - "loss": 0.0135, + "grad_norm": 0.349990576505661, + "learning_rate": 9.820306224465486e-05, + "loss": 0.0169, "step": 7880 }, { - "grad_norm": 0.27795737981796265, - "learning_rate": 7.0921724328189e-05, - "loss": 0.0115, + "grad_norm": 0.3649872839450836, + "learning_rate": 9.819573335647928e-05, + "loss": 0.0145, "step": 7890 }, { - "grad_norm": 0.3064858615398407, - "learning_rate": 7.084660781421268e-05, - "loss": 0.0142, + "grad_norm": 0.2825605273246765, + "learning_rate": 9.818838982771246e-05, + "loss": 0.0167, "step": 7900 }, { - "grad_norm": 0.32354864478111267, - "learning_rate": 7.077143430640662e-05, - "loss": 0.0154, + "grad_norm": 0.4013703465461731, + "learning_rate": 9.818103166058514e-05, + "loss": 0.0184, "step": 7910 }, { - "grad_norm": 0.38219061493873596, - "learning_rate": 7.069620401029232e-05, - "loss": 0.0116, + "grad_norm": 0.3298392593860626, + "learning_rate": 9.817365885733254e-05, + "loss": 0.0143, "step": 7920 }, { - "grad_norm": 0.3809864819049835, - "learning_rate": 7.062091713154655e-05, - "loss": 0.0132, + "grad_norm": 0.38088521361351013, + "learning_rate": 9.816627142019434e-05, + "loss": 0.0161, "step": 7930 }, { - "grad_norm": 0.3125777542591095, - "learning_rate": 7.054557387600075e-05, - "loss": 0.0149, + "grad_norm": 0.4659990966320038, + "learning_rate": 9.815886935141463e-05, + "loss": 0.0183, "step": 7940 }, { - "grad_norm": 0.3452930152416229, - "learning_rate": 7.04701744496405e-05, - "loss": 0.0106, + "grad_norm": 0.3853849172592163, + "learning_rate": 9.8151452653242e-05, + "loss": 0.0135, "step": 7950 }, { - "grad_norm": 0.37026354670524597, - "learning_rate": 7.039471905860495e-05, - "loss": 0.0122, + "grad_norm": 0.37548625469207764, + "learning_rate": 9.814402132792939e-05, + "loss": 0.0158, "step": 7960 }, { - "grad_norm": 0.32580164074897766, - "learning_rate": 7.031920790918628e-05, - "loss": 0.013, + "grad_norm": 0.31197187304496765, + "learning_rate": 9.813657537773428e-05, + "loss": 0.0161, "step": 7970 }, { - "grad_norm": 0.288991242647171, - "learning_rate": 7.024364120782906e-05, - "loss": 0.0132, + "grad_norm": 0.32812777161598206, + "learning_rate": 9.812911480491854e-05, + "loss": 0.0165, "step": 7980 }, { - "grad_norm": 0.325033575296402, - "learning_rate": 7.016801916112978e-05, - "loss": 0.0102, + "grad_norm": 0.3862552046775818, + "learning_rate": 9.81216396117485e-05, + "loss": 0.0136, "step": 7990 }, { - "grad_norm": 0.27642202377319336, - "learning_rate": 7.009234197583623e-05, - "loss": 0.0111, + "grad_norm": 0.38808831572532654, + "learning_rate": 9.811414980049491e-05, + "loss": 0.0139, "step": 8000 }, { - "grad_norm": 0.3329485058784485, - "learning_rate": 7.001660985884692e-05, - "loss": 0.0123, + "grad_norm": 0.32167288661003113, + "learning_rate": 9.810664537343301e-05, + "loss": 0.0152, "step": 8010 }, { - "grad_norm": 0.3785879611968994, - "learning_rate": 6.994082301721063e-05, - "loss": 0.0151, + "grad_norm": 0.44605064392089844, + "learning_rate": 9.809912633284243e-05, + "loss": 0.018, "step": 8020 }, { - "grad_norm": 0.34150007367134094, - "learning_rate": 6.986498165812563e-05, - "loss": 0.0117, + "grad_norm": 0.33501434326171875, + "learning_rate": 9.809159268100725e-05, + "loss": 0.0143, "step": 8030 }, { - "grad_norm": 0.439264178276062, - "learning_rate": 6.978908598893932e-05, - "loss": 0.0139, + "grad_norm": 0.4212293028831482, + "learning_rate": 9.808404442021599e-05, + "loss": 0.0168, "step": 8040 }, { - "grad_norm": 0.3500205874443054, - "learning_rate": 6.971313621714756e-05, - "loss": 0.0112, + "grad_norm": 0.3646804392337799, + "learning_rate": 9.807648155276163e-05, + "loss": 0.0136, "step": 8050 }, { - "grad_norm": 0.2777945399284363, - "learning_rate": 6.96371325503941e-05, - "loss": 0.011, + "grad_norm": 0.3829670250415802, + "learning_rate": 9.806890408094156e-05, + "loss": 0.0133, "step": 8060 }, { - "grad_norm": 0.382794052362442, - "learning_rate": 6.956107519647014e-05, - "loss": 0.0125, + "grad_norm": 0.37975287437438965, + "learning_rate": 9.806131200705761e-05, + "loss": 0.0153, "step": 8070 }, { - "grad_norm": 0.30284425616264343, - "learning_rate": 6.94849643633135e-05, - "loss": 0.0133, + "grad_norm": 0.34802359342575073, + "learning_rate": 9.805370533341605e-05, + "loss": 0.016, "step": 8080 }, { - "grad_norm": 0.24802178144454956, - "learning_rate": 6.940880025900834e-05, - "loss": 0.0107, + "grad_norm": 0.34443968534469604, + "learning_rate": 9.804608406232762e-05, + "loss": 0.0134, "step": 8090 }, { - "grad_norm": 0.2981727719306946, - "learning_rate": 6.933258309178438e-05, - "loss": 0.0122, + "grad_norm": 0.34838855266571045, + "learning_rate": 9.803844819610741e-05, + "loss": 0.0143, "step": 8100 }, { - "grad_norm": 0.21943475306034088, - "learning_rate": 6.925631307001646e-05, - "loss": 0.0165, + "grad_norm": 0.3762608766555786, + "learning_rate": 9.803079773707504e-05, + "loss": 0.0194, "step": 8110 }, { - "grad_norm": 0.3769550621509552, - "learning_rate": 6.91799904022239e-05, - "loss": 0.0143, + "grad_norm": 0.3491004407405853, + "learning_rate": 9.802313268755447e-05, + "loss": 0.0168, "step": 8120 }, { - "grad_norm": 0.2876106798648834, - "learning_rate": 6.910361529706997e-05, - "loss": 0.0131, + "grad_norm": 0.3784998953342438, + "learning_rate": 9.801545304987419e-05, + "loss": 0.0159, "step": 8130 }, { - "grad_norm": 0.33470556139945984, - "learning_rate": 6.902718796336131e-05, - "loss": 0.0137, + "grad_norm": 0.4268767237663269, + "learning_rate": 9.800775882636704e-05, + "loss": 0.0174, "step": 8140 }, { - "grad_norm": 0.4283612370491028, - "learning_rate": 6.895070861004729e-05, - "loss": 0.0129, + "grad_norm": 0.4905880093574524, + "learning_rate": 9.800005001937034e-05, + "loss": 0.0162, "step": 8150 }, { - "grad_norm": 0.2765333652496338, - "learning_rate": 6.887417744621956e-05, - "loss": 0.0137, + "grad_norm": 0.36699673533439636, + "learning_rate": 9.79923266312258e-05, + "loss": 0.0164, "step": 8160 }, { - "grad_norm": 0.32283779978752136, - "learning_rate": 6.87975946811114e-05, - "loss": 0.0125, + "grad_norm": 0.39664509892463684, + "learning_rate": 9.79845886642796e-05, + "loss": 0.0156, "step": 8170 }, { - "grad_norm": 0.3863534927368164, - "learning_rate": 6.872096052409718e-05, - "loss": 0.0139, + "grad_norm": 0.44191649556159973, + "learning_rate": 9.797683612088233e-05, + "loss": 0.0167, "step": 8180 }, { - "grad_norm": 0.29665520787239075, - "learning_rate": 6.864427518469174e-05, - "loss": 0.0129, + "grad_norm": 0.346263587474823, + "learning_rate": 9.796906900338898e-05, + "loss": 0.0153, "step": 8190 }, { - "grad_norm": 0.3554975986480713, - "learning_rate": 6.856753887254986e-05, - "loss": 0.0107, + "grad_norm": 0.29614493250846863, + "learning_rate": 9.796128731415903e-05, + "loss": 0.0125, "step": 8200 }, { - "grad_norm": 0.3163433074951172, - "learning_rate": 6.849075179746572e-05, - "loss": 0.0115, + "grad_norm": 0.38059693574905396, + "learning_rate": 9.795349105555634e-05, + "loss": 0.0137, "step": 8210 }, { - "grad_norm": 0.3252999782562256, - "learning_rate": 6.841391416937221e-05, - "loss": 0.01, + "grad_norm": 0.3214937746524811, + "learning_rate": 9.794568022994922e-05, + "loss": 0.0126, "step": 8220 }, { - "grad_norm": 0.38887593150138855, - "learning_rate": 6.833702619834053e-05, - "loss": 0.0107, + "grad_norm": 0.348738431930542, + "learning_rate": 9.793785483971034e-05, + "loss": 0.014, "step": 8230 }, { - "grad_norm": 0.3632328510284424, - "learning_rate": 6.82600880945794e-05, - "loss": 0.0106, + "grad_norm": 0.3860212564468384, + "learning_rate": 9.793001488721691e-05, + "loss": 0.0138, "step": 8240 }, { - "grad_norm": 0.2744136154651642, - "learning_rate": 6.818310006843468e-05, - "loss": 0.0108, + "grad_norm": 0.31617406010627747, + "learning_rate": 9.792216037485047e-05, + "loss": 0.0139, "step": 8250 }, { - "grad_norm": 0.3894250988960266, - "learning_rate": 6.810606233038868e-05, - "loss": 0.0117, + "grad_norm": 0.4328280985355377, + "learning_rate": 9.791429130499704e-05, + "loss": 0.0146, "step": 8260 }, { - "grad_norm": 0.3185424506664276, - "learning_rate": 6.802897509105966e-05, - "loss": 0.0136, + "grad_norm": 0.4389854371547699, + "learning_rate": 9.790640768004698e-05, + "loss": 0.0168, "step": 8270 }, { - "grad_norm": 0.28028440475463867, - "learning_rate": 6.79518385612012e-05, - "loss": 0.0129, + "grad_norm": 0.474202960729599, + "learning_rate": 9.789850950239518e-05, + "loss": 0.0165, "step": 8280 }, { - "grad_norm": 0.25754812359809875, - "learning_rate": 6.787465295170157e-05, - "loss": 0.0121, + "grad_norm": 0.39522701501846313, + "learning_rate": 9.789059677444089e-05, + "loss": 0.0154, "step": 8290 }, { - "grad_norm": 0.2997465431690216, - "learning_rate": 6.779741847358332e-05, - "loss": 0.0108, + "grad_norm": 0.3074120581150055, + "learning_rate": 9.788266949858776e-05, + "loss": 0.0136, "step": 8300 }, { - "grad_norm": 0.3573647141456604, - "learning_rate": 6.772013533800256e-05, - "loss": 0.0133, + "grad_norm": 0.35952499508857727, + "learning_rate": 9.787472767724392e-05, + "loss": 0.016, "step": 8310 }, { - "grad_norm": 0.3089340627193451, - "learning_rate": 6.764280375624843e-05, - "loss": 0.0116, + "grad_norm": 0.3444002568721771, + "learning_rate": 9.786677131282185e-05, + "loss": 0.0143, "step": 8320 }, { - "grad_norm": 0.31227874755859375, - "learning_rate": 6.756542393974252e-05, - "loss": 0.0108, + "grad_norm": 0.32378682494163513, + "learning_rate": 9.785880040773853e-05, + "loss": 0.0138, "step": 8330 }, { - "grad_norm": 0.381428986787796, - "learning_rate": 6.748799610003828e-05, - "loss": 0.0112, + "grad_norm": 0.39044997096061707, + "learning_rate": 9.785081496441527e-05, + "loss": 0.0144, "step": 8340 }, { - "grad_norm": 0.278055340051651, - "learning_rate": 6.741052044882048e-05, - "loss": 0.0117, + "grad_norm": 0.3554976284503937, + "learning_rate": 9.784281498527785e-05, + "loss": 0.0148, "step": 8350 }, { - "grad_norm": 0.2949462831020355, - "learning_rate": 6.73329971979046e-05, - "loss": 0.0108, + "grad_norm": 0.30649831891059875, + "learning_rate": 9.783480047275646e-05, + "loss": 0.0143, "step": 8360 }, { - "grad_norm": 0.3816727101802826, - "learning_rate": 6.725542655923625e-05, - "loss": 0.0116, + "grad_norm": 0.4501311480998993, + "learning_rate": 9.78267714292857e-05, + "loss": 0.0154, "step": 8370 }, { - "grad_norm": 0.2636021673679352, - "learning_rate": 6.717780874489057e-05, - "loss": 0.0128, + "grad_norm": 0.3144357204437256, + "learning_rate": 9.781872785730454e-05, + "loss": 0.0163, "step": 8380 }, { - "grad_norm": 0.38993480801582336, - "learning_rate": 6.710014396707172e-05, - "loss": 0.0119, + "grad_norm": 0.383786141872406, + "learning_rate": 9.781066975925646e-05, + "loss": 0.0144, "step": 8390 }, { - "grad_norm": 0.3024043142795563, - "learning_rate": 6.702243243811221e-05, - "loss": 0.0109, + "grad_norm": 0.34242933988571167, + "learning_rate": 9.780259713758928e-05, + "loss": 0.0135, "step": 8400 }, { - "grad_norm": 0.36139100790023804, - "learning_rate": 6.694467437047244e-05, - "loss": 0.0118, + "grad_norm": 0.43082892894744873, + "learning_rate": 9.779450999475524e-05, + "loss": 0.014, "step": 8410 }, { - "grad_norm": 0.34032484889030457, - "learning_rate": 6.686686997673997e-05, - "loss": 0.0122, + "grad_norm": 0.3656945526599884, + "learning_rate": 9.7786408333211e-05, + "loss": 0.0146, "step": 8420 }, { - "grad_norm": 0.3963174819946289, - "learning_rate": 6.678901946962903e-05, - "loss": 0.0104, + "grad_norm": 0.39802682399749756, + "learning_rate": 9.777829215541764e-05, + "loss": 0.0126, "step": 8430 }, { - "grad_norm": 0.2162044644355774, - "learning_rate": 6.671112306197996e-05, - "loss": 0.0124, + "grad_norm": 0.29535219073295593, + "learning_rate": 9.777016146384064e-05, + "loss": 0.0151, "step": 8440 }, { - "grad_norm": 0.2702636122703552, - "learning_rate": 6.663318096675854e-05, - "loss": 0.0125, + "grad_norm": 0.2743058502674103, + "learning_rate": 9.776201626094988e-05, + "loss": 0.0151, "step": 8450 }, { - "grad_norm": 0.42705151438713074, - "learning_rate": 6.655519339705552e-05, - "loss": 0.0118, + "grad_norm": 0.47285139560699463, + "learning_rate": 9.775385654921965e-05, + "loss": 0.0156, "step": 8460 }, { - "grad_norm": 0.29198169708251953, - "learning_rate": 6.647716056608588e-05, - "loss": 0.0132, + "grad_norm": 0.5085629820823669, + "learning_rate": 9.774568233112868e-05, + "loss": 0.0243, "step": 8470 }, { - "grad_norm": 0.24624188244342804, - "learning_rate": 6.639908268718843e-05, - "loss": 0.011, + "grad_norm": 0.3698887526988983, + "learning_rate": 9.773749360916007e-05, + "loss": 0.0202, "step": 8480 }, { - "grad_norm": 0.29472094774246216, - "learning_rate": 6.632095997382514e-05, - "loss": 0.0103, + "grad_norm": 0.35737812519073486, + "learning_rate": 9.772929038580134e-05, + "loss": 0.0173, "step": 8490 }, { - "grad_norm": 0.28369250893592834, - "learning_rate": 6.624279263958047e-05, - "loss": 0.0121, + "grad_norm": 0.3153495192527771, + "learning_rate": 9.772107266354439e-05, + "loss": 0.0175, "step": 8500 }, { - "grad_norm": 0.3822849690914154, - "learning_rate": 6.616458089816097e-05, - "loss": 0.0121, + "grad_norm": 0.3515191972255707, + "learning_rate": 9.77128404448856e-05, + "loss": 0.0161, "step": 8510 }, { - "grad_norm": 0.3733149766921997, - "learning_rate": 6.608632496339454e-05, - "loss": 0.0127, + "grad_norm": 0.3558012843132019, + "learning_rate": 9.770459373232565e-05, + "loss": 0.0166, "step": 8520 }, { - "grad_norm": 0.25586628913879395, - "learning_rate": 6.600802504922988e-05, - "loss": 0.0149, + "grad_norm": 0.30575254559516907, + "learning_rate": 9.769633252836969e-05, + "loss": 0.0188, "step": 8530 }, { - "grad_norm": 0.36027905344963074, - "learning_rate": 6.592968136973604e-05, - "loss": 0.0132, + "grad_norm": 0.4248638153076172, + "learning_rate": 9.768805683552724e-05, + "loss": 0.0177, "step": 8540 }, { - "grad_norm": 0.30015572905540466, - "learning_rate": 6.585129413910159e-05, - "loss": 0.0119, + "grad_norm": 0.4025588631629944, + "learning_rate": 9.767976665631228e-05, + "loss": 0.0161, "step": 8550 }, { - "grad_norm": 0.3715488016605377, - "learning_rate": 6.577286357163424e-05, - "loss": 0.0123, + "grad_norm": 0.4045548737049103, + "learning_rate": 9.767146199324311e-05, + "loss": 0.0166, "step": 8560 }, { - "grad_norm": 0.39481574296951294, - "learning_rate": 6.569438988176018e-05, - "loss": 0.0103, + "grad_norm": 0.41473588347435, + "learning_rate": 9.766314284884249e-05, + "loss": 0.0133, "step": 8570 }, { - "grad_norm": 0.35289764404296875, - "learning_rate": 6.561587328402347e-05, - "loss": 0.0127, + "grad_norm": 0.352109432220459, + "learning_rate": 9.765480922563752e-05, + "loss": 0.016, "step": 8580 }, { - "grad_norm": 0.34876325726509094, - "learning_rate": 6.553731399308549e-05, - "loss": 0.0127, + "grad_norm": 0.4061742424964905, + "learning_rate": 9.764646112615978e-05, + "loss": 0.0158, "step": 8590 }, { - "grad_norm": 0.26030921936035156, - "learning_rate": 6.545871222372436e-05, - "loss": 0.011, + "grad_norm": 0.3103123903274536, + "learning_rate": 9.763809855294517e-05, + "loss": 0.0144, "step": 8600 }, { - "grad_norm": 0.3516961336135864, - "learning_rate": 6.538006819083426e-05, - "loss": 0.0115, + "grad_norm": 0.3941231369972229, + "learning_rate": 9.762972150853404e-05, + "loss": 0.0155, "step": 8610 }, { - "grad_norm": 0.3771360218524933, - "learning_rate": 6.530138210942505e-05, - "loss": 0.011, + "grad_norm": 0.4432227611541748, + "learning_rate": 9.762132999547111e-05, + "loss": 0.0146, "step": 8620 }, { - "grad_norm": 0.2810113728046417, - "learning_rate": 6.522265419462141e-05, - "loss": 0.0089, + "grad_norm": 0.3516373336315155, + "learning_rate": 9.761292401630549e-05, + "loss": 0.0131, "step": 8630 }, { - "grad_norm": 0.2637802064418793, - "learning_rate": 6.514388466166248e-05, - "loss": 0.0105, + "grad_norm": 0.37528103590011597, + "learning_rate": 9.76045035735907e-05, + "loss": 0.0143, "step": 8640 }, { - "grad_norm": 0.2962155342102051, - "learning_rate": 6.506507372590119e-05, - "loss": 0.0126, + "grad_norm": 0.3847804665565491, + "learning_rate": 9.759606866988464e-05, + "loss": 0.0161, "step": 8650 }, { - "grad_norm": 0.2789287269115448, - "learning_rate": 6.498622160280355e-05, - "loss": 0.0128, + "grad_norm": 0.31890586018562317, + "learning_rate": 9.758761930774963e-05, + "loss": 0.0158, "step": 8660 }, { - "grad_norm": 0.2849268317222595, - "learning_rate": 6.490732850794832e-05, - "loss": 0.0133, + "grad_norm": 0.3136662542819977, + "learning_rate": 9.757915548975235e-05, + "loss": 0.0163, "step": 8670 }, { - "grad_norm": 0.26055100560188293, - "learning_rate": 6.482839465702616e-05, - "loss": 0.0131, + "grad_norm": 0.31251639127731323, + "learning_rate": 9.757067721846389e-05, + "loss": 0.0154, "step": 8680 }, { - "grad_norm": 0.32526013255119324, - "learning_rate": 6.474942026583923e-05, - "loss": 0.0127, + "grad_norm": 0.4165763556957245, + "learning_rate": 9.756218449645971e-05, + "loss": 0.0151, "step": 8690 }, { - "grad_norm": 0.33485621213912964, - "learning_rate": 6.467040555030052e-05, - "loss": 0.0108, + "grad_norm": 0.41660186648368835, + "learning_rate": 9.75536773263197e-05, + "loss": 0.014, "step": 8700 }, { - "grad_norm": 0.2516840994358063, - "learning_rate": 6.459135072643321e-05, - "loss": 0.0115, + "grad_norm": 0.3443741798400879, + "learning_rate": 9.75451557106281e-05, + "loss": 0.015, "step": 8710 }, { - "grad_norm": 0.31756162643432617, - "learning_rate": 6.451225601037019e-05, - "loss": 0.0091, + "grad_norm": 0.3316461741924286, + "learning_rate": 9.753661965197354e-05, + "loss": 0.0123, "step": 8720 }, { - "grad_norm": 0.2860639691352844, - "learning_rate": 6.443312161835338e-05, - "loss": 0.0128, + "grad_norm": 0.4706900417804718, + "learning_rate": 9.752806915294908e-05, + "loss": 0.0167, "step": 8730 }, { - "grad_norm": 0.2940915822982788, - "learning_rate": 6.43539477667332e-05, - "loss": 0.0104, + "grad_norm": 0.32706180214881897, + "learning_rate": 9.75195042161521e-05, + "loss": 0.0131, "step": 8740 }, { - "grad_norm": 0.24016410112380981, - "learning_rate": 6.427473467196793e-05, - "loss": 0.0143, + "grad_norm": 0.3663170039653778, + "learning_rate": 9.751092484418442e-05, + "loss": 0.0176, "step": 8750 }, { - "grad_norm": 0.29369375109672546, - "learning_rate": 6.419548255062315e-05, - "loss": 0.0103, + "grad_norm": 0.3475320041179657, + "learning_rate": 9.750233103965224e-05, + "loss": 0.0131, "step": 8760 }, { - "grad_norm": 0.32075247168540955, - "learning_rate": 6.411619161937112e-05, - "loss": 0.0097, + "grad_norm": 0.3173498511314392, + "learning_rate": 9.749372280516611e-05, + "loss": 0.0123, "step": 8770 }, { - "grad_norm": 0.3262137174606323, - "learning_rate": 6.403686209499022e-05, - "loss": 0.0104, + "grad_norm": 0.3239175081253052, + "learning_rate": 9.748510014334097e-05, + "loss": 0.0131, "step": 8780 }, { - "grad_norm": 0.3471458852291107, - "learning_rate": 6.395749419436437e-05, - "loss": 0.0099, + "grad_norm": 0.39269745349884033, + "learning_rate": 9.747646305679621e-05, + "loss": 0.0122, "step": 8790 }, { - "grad_norm": 0.30499032139778137, - "learning_rate": 6.387808813448234e-05, - "loss": 0.0133, + "grad_norm": 0.3620828092098236, + "learning_rate": 9.74678115481555e-05, + "loss": 0.0162, "step": 8800 }, { - "grad_norm": 0.34709998965263367, - "learning_rate": 6.37986441324373e-05, - "loss": 0.0118, + "grad_norm": 0.42276638746261597, + "learning_rate": 9.745914562004696e-05, + "loss": 0.0148, "step": 8810 }, { - "grad_norm": 0.2934912443161011, - "learning_rate": 6.37191624054261e-05, - "loss": 0.0125, + "grad_norm": 0.40569648146629333, + "learning_rate": 9.745046527510307e-05, + "loss": 0.0157, "step": 8820 }, { - "grad_norm": 0.324332058429718, - "learning_rate": 6.363964317074872e-05, - "loss": 0.0154, + "grad_norm": 0.3451126515865326, + "learning_rate": 9.744177051596068e-05, + "loss": 0.0188, "step": 8830 }, { - "grad_norm": 0.2829475402832031, - "learning_rate": 6.356008664580776e-05, - "loss": 0.0134, + "grad_norm": 0.37246832251548767, + "learning_rate": 9.743306134526105e-05, + "loss": 0.0178, "step": 8840 }, { - "grad_norm": 0.2962392270565033, - "learning_rate": 6.348049304810771e-05, - "loss": 0.0142, + "grad_norm": 0.3463205099105835, + "learning_rate": 9.742433776564977e-05, + "loss": 0.0181, "step": 8850 }, { - "grad_norm": 0.3759327530860901, - "learning_rate": 6.340086259525442e-05, - "loss": 0.0129, + "grad_norm": 0.39877772331237793, + "learning_rate": 9.741559977977683e-05, + "loss": 0.0165, "step": 8860 }, { - "grad_norm": 0.36553898453712463, - "learning_rate": 6.332119550495448e-05, - "loss": 0.0131, + "grad_norm": 0.26505008339881897, + "learning_rate": 9.740684739029661e-05, + "loss": 0.0167, "step": 8870 }, { - "grad_norm": 0.33665499091148376, - "learning_rate": 6.324149199501473e-05, - "loss": 0.0113, + "grad_norm": 0.36500832438468933, + "learning_rate": 9.739808059986789e-05, + "loss": 0.0142, "step": 8880 }, { - "grad_norm": 0.3106566071510315, - "learning_rate": 6.316175228334146e-05, - "loss": 0.0128, + "grad_norm": 0.3650977313518524, + "learning_rate": 9.738929941115373e-05, + "loss": 0.0163, "step": 8890 }, { - "grad_norm": 0.3394717276096344, - "learning_rate": 6.308197658794003e-05, - "loss": 0.0089, + "grad_norm": 0.29584237933158875, + "learning_rate": 9.738050382682167e-05, + "loss": 0.0117, "step": 8900 }, { - "grad_norm": 0.30129188299179077, - "learning_rate": 6.300216512691417e-05, - "loss": 0.0114, + "grad_norm": 0.40130600333213806, + "learning_rate": 9.737169384954355e-05, + "loss": 0.0149, "step": 8910 }, { - "grad_norm": 0.3877800405025482, - "learning_rate": 6.292231811846532e-05, - "loss": 0.0112, + "grad_norm": 0.36694860458374023, + "learning_rate": 9.736286948199562e-05, + "loss": 0.0143, "step": 8920 }, { - "grad_norm": 0.31548699736595154, - "learning_rate": 6.284243578089217e-05, - "loss": 0.0124, + "grad_norm": 0.29328790307044983, + "learning_rate": 9.735403072685848e-05, + "loss": 0.0149, "step": 8930 }, { - "grad_norm": 0.286324143409729, - "learning_rate": 6.276251833258999e-05, - "loss": 0.0104, + "grad_norm": 0.360105961561203, + "learning_rate": 9.734517758681712e-05, + "loss": 0.0133, "step": 8940 }, { - "grad_norm": 0.31045469641685486, - "learning_rate": 6.268256599205003e-05, - "loss": 0.0124, + "grad_norm": 0.3908972442150116, + "learning_rate": 9.733631006456088e-05, + "loss": 0.0157, "step": 8950 }, { - "grad_norm": 0.31697219610214233, - "learning_rate": 6.260257897785892e-05, - "loss": 0.0137, + "grad_norm": 0.32673004269599915, + "learning_rate": 9.732742816278348e-05, + "loss": 0.0168, "step": 8960 }, { - "grad_norm": 0.37782853841781616, - "learning_rate": 6.252255750869811e-05, - "loss": 0.0118, + "grad_norm": 0.389112263917923, + "learning_rate": 9.731853188418302e-05, + "loss": 0.0146, "step": 8970 }, { - "grad_norm": 0.30432772636413574, - "learning_rate": 6.244250180334325e-05, - "loss": 0.0128, + "grad_norm": 0.38246992230415344, + "learning_rate": 9.730962123146194e-05, + "loss": 0.0158, "step": 8980 }, { - "grad_norm": 0.3652268052101135, - "learning_rate": 6.236241208066356e-05, - "loss": 0.0127, + "grad_norm": 0.3654584288597107, + "learning_rate": 9.730069620732709e-05, + "loss": 0.016, "step": 8990 }, { - "grad_norm": 0.2324247658252716, - "learning_rate": 6.228228855962133e-05, - "loss": 0.0162, + "grad_norm": 0.33035150170326233, + "learning_rate": 9.72917568144896e-05, + "loss": 0.0196, "step": 9000 }, { - "grad_norm": 0.31128183007240295, - "learning_rate": 6.220213145927115e-05, - "loss": 0.0135, + "grad_norm": 0.31670403480529785, + "learning_rate": 9.728280305566509e-05, + "loss": 0.0152, "step": 9010 }, { - "grad_norm": 0.32616373896598816, - "learning_rate": 6.212194099875951e-05, - "loss": 0.0101, + "grad_norm": 0.28820958733558655, + "learning_rate": 9.727383493357343e-05, + "loss": 0.0123, "step": 9020 }, { - "grad_norm": 0.25326231122016907, - "learning_rate": 6.204171739732405e-05, - "loss": 0.0133, + "grad_norm": 0.3736238479614258, + "learning_rate": 9.726485245093891e-05, + "loss": 0.0174, "step": 9030 }, { - "grad_norm": 0.3350052237510681, - "learning_rate": 6.196146087429303e-05, - "loss": 0.0134, + "grad_norm": 0.3337687849998474, + "learning_rate": 9.725585561049018e-05, + "loss": 0.0174, "step": 9040 }, { - "grad_norm": 0.40826472640037537, - "learning_rate": 6.188117164908474e-05, - "loss": 0.0112, + "grad_norm": 0.4167155921459198, + "learning_rate": 9.724684441496022e-05, + "loss": 0.0136, "step": 9050 }, { - "grad_norm": 0.2690710127353668, - "learning_rate": 6.180084994120684e-05, - "loss": 0.0116, + "grad_norm": 0.28661665320396423, + "learning_rate": 9.72378188670864e-05, + "loss": 0.0146, "step": 9060 }, { - "grad_norm": 0.20586135983467102, - "learning_rate": 6.17204959702558e-05, - "loss": 0.012, + "grad_norm": 0.3665269911289215, + "learning_rate": 9.722877896961047e-05, + "loss": 0.0155, "step": 9070 }, { - "grad_norm": 0.30860739946365356, - "learning_rate": 6.164010995591635e-05, - "loss": 0.0128, + "grad_norm": 0.2954253852367401, + "learning_rate": 9.721972472527848e-05, + "loss": 0.0156, "step": 9080 }, { - "grad_norm": 0.36940038204193115, - "learning_rate": 6.155969211796076e-05, - "loss": 0.0118, + "grad_norm": 0.3786759078502655, + "learning_rate": 9.721065613684089e-05, + "loss": 0.015, "step": 9090 }, { - "grad_norm": 0.32097283005714417, - "learning_rate": 6.147924267624829e-05, - "loss": 0.0122, + "grad_norm": 0.33991774916648865, + "learning_rate": 9.72015732070525e-05, + "loss": 0.0151, "step": 9100 }, { - "grad_norm": 0.3452341556549072, - "learning_rate": 6.13987618507247e-05, - "loss": 0.0124, + "grad_norm": 0.37635087966918945, + "learning_rate": 9.719247593867244e-05, + "loss": 0.0155, "step": 9110 }, { - "grad_norm": 0.25516965985298157, - "learning_rate": 6.131824986142147e-05, - "loss": 0.0123, + "grad_norm": 0.31850841641426086, + "learning_rate": 9.718336433446423e-05, + "loss": 0.016, "step": 9120 }, { - "grad_norm": 0.3512190878391266, - "learning_rate": 6.123770692845529e-05, - "loss": 0.0108, + "grad_norm": 0.4509117901325226, + "learning_rate": 9.717423839719574e-05, + "loss": 0.013, "step": 9130 }, { - "grad_norm": 0.3161102831363678, - "learning_rate": 6.11571332720275e-05, - "loss": 0.0121, + "grad_norm": 0.26356270909309387, + "learning_rate": 9.71650981296392e-05, + "loss": 0.0154, "step": 9140 }, { - "grad_norm": 0.2909663915634155, - "learning_rate": 6.107652911242336e-05, - "loss": 0.0099, + "grad_norm": 0.3148396909236908, + "learning_rate": 9.715594353457118e-05, + "loss": 0.0131, "step": 9150 }, { - "grad_norm": 0.3107944130897522, - "learning_rate": 6.0995894670011586e-05, - "loss": 0.0108, + "grad_norm": 0.39046502113342285, + "learning_rate": 9.714677461477257e-05, + "loss": 0.0144, "step": 9160 }, { - "grad_norm": 0.2862789034843445, - "learning_rate": 6.091523016524368e-05, - "loss": 0.0121, + "grad_norm": 0.33206799626350403, + "learning_rate": 9.713759137302869e-05, + "loss": 0.0158, "step": 9170 }, { - "grad_norm": 0.3161313235759735, - "learning_rate": 6.083453581865328e-05, - "loss": 0.0105, + "grad_norm": 0.3918338716030121, + "learning_rate": 9.712839381212914e-05, + "loss": 0.0135, "step": 9180 }, { - "grad_norm": 0.2773940861225128, - "learning_rate": 6.075381185085568e-05, - "loss": 0.0108, + "grad_norm": 0.3739299774169922, + "learning_rate": 9.71191819348679e-05, + "loss": 0.0153, "step": 9190 }, { - "grad_norm": 0.2752029299736023, - "learning_rate": 6.067305848254709e-05, - "loss": 0.0115, + "grad_norm": 0.3122602105140686, + "learning_rate": 9.710995574404331e-05, + "loss": 0.016, "step": 9200 }, { - "grad_norm": 0.35372427105903625, - "learning_rate": 6.059227593450418e-05, - "loss": 0.0119, + "grad_norm": 0.3587058186531067, + "learning_rate": 9.710071524245802e-05, + "loss": 0.0163, "step": 9210 }, { - "grad_norm": 0.3142034411430359, - "learning_rate": 6.051146442758333e-05, - "loss": 0.0125, + "grad_norm": 0.27872487902641296, + "learning_rate": 9.709146043291906e-05, + "loss": 0.0172, "step": 9220 }, { - "grad_norm": 0.2529461979866028, - "learning_rate": 6.043062418272012e-05, - "loss": 0.0107, + "grad_norm": 0.3250398635864258, + "learning_rate": 9.70821913182378e-05, + "loss": 0.0144, "step": 9230 }, { - "grad_norm": 0.3262726962566376, - "learning_rate": 6.0349755420928666e-05, - "loss": 0.0127, + "grad_norm": 0.30500882863998413, + "learning_rate": 9.707290790122995e-05, + "loss": 0.0159, "step": 9240 }, { - "grad_norm": 0.26650935411453247, - "learning_rate": 6.0268858363301105e-05, - "loss": 0.0106, + "grad_norm": 0.21951811015605927, + "learning_rate": 9.706361018471557e-05, + "loss": 0.0136, "step": 9250 }, { - "grad_norm": 0.3812367916107178, - "learning_rate": 6.018793323100689e-05, - "loss": 0.0104, + "grad_norm": 0.39244839549064636, + "learning_rate": 9.705429817151906e-05, + "loss": 0.0134, "step": 9260 }, { - "grad_norm": 0.2846297025680542, - "learning_rate": 6.0106980245292255e-05, - "loss": 0.0116, + "grad_norm": 0.29822587966918945, + "learning_rate": 9.704497186446917e-05, + "loss": 0.0137, "step": 9270 }, { - "grad_norm": 0.3379952907562256, - "learning_rate": 6.002599962747957e-05, - "loss": 0.0131, + "grad_norm": 0.40413886308670044, + "learning_rate": 9.703563126639896e-05, + "loss": 0.0148, "step": 9280 }, { - "grad_norm": 0.28340259194374084, - "learning_rate": 5.994499159896673e-05, - "loss": 0.0109, + "grad_norm": 0.3283483684062958, + "learning_rate": 9.70262763801459e-05, + "loss": 0.0138, "step": 9290 }, { - "grad_norm": 0.32905933260917664, - "learning_rate": 5.9863956381226607e-05, - "loss": 0.0127, + "grad_norm": 0.2845418155193329, + "learning_rate": 9.701690720855171e-05, + "loss": 0.0157, "step": 9300 }, { - "grad_norm": 0.34781473875045776, - "learning_rate": 5.9782894195806394e-05, - "loss": 0.0097, + "grad_norm": 0.3722706735134125, + "learning_rate": 9.700752375446253e-05, + "loss": 0.0131, "step": 9310 }, { - "grad_norm": 0.3538030683994293, - "learning_rate": 5.9701805264327004e-05, - "loss": 0.0101, + "grad_norm": 0.26517462730407715, + "learning_rate": 9.69981260207288e-05, + "loss": 0.013, "step": 9320 }, { - "grad_norm": 0.3605461120605469, - "learning_rate": 5.96206898084825e-05, - "loss": 0.0102, + "grad_norm": 0.3415532112121582, + "learning_rate": 9.698871401020529e-05, + "loss": 0.0127, "step": 9330 }, { - "grad_norm": 0.29366081953048706, - "learning_rate": 5.953954805003942e-05, - "loss": 0.0105, + "grad_norm": 0.36040860414505005, + "learning_rate": 9.697928772575112e-05, + "loss": 0.0132, "step": 9340 }, { - "grad_norm": 0.26942163705825806, - "learning_rate": 5.945838021083623e-05, - "loss": 0.0119, + "grad_norm": 0.3559826612472534, + "learning_rate": 9.696984717022976e-05, + "loss": 0.0145, "step": 9350 }, { - "grad_norm": 0.3048942983150482, - "learning_rate": 5.9377186512782714e-05, - "loss": 0.0104, + "grad_norm": 0.3137611746788025, + "learning_rate": 9.6960392346509e-05, + "loss": 0.014, "step": 9360 }, { - "grad_norm": 0.3276144862174988, - "learning_rate": 5.929596717785935e-05, - "loss": 0.0092, + "grad_norm": 0.2644113302230835, + "learning_rate": 9.695092325746097e-05, + "loss": 0.0111, "step": 9370 }, { - "grad_norm": 0.327134370803833, - "learning_rate": 5.921472242811668e-05, - "loss": 0.0106, + "grad_norm": 0.3597872257232666, + "learning_rate": 9.694143990596211e-05, + "loss": 0.0137, "step": 9380 }, { - "grad_norm": 0.2601287066936493, - "learning_rate": 5.913345248567475e-05, - "loss": 0.0095, + "grad_norm": 0.2509511411190033, + "learning_rate": 9.693194229489325e-05, + "loss": 0.0115, "step": 9390 }, { - "grad_norm": 0.3127315044403076, - "learning_rate": 5.905215757272248e-05, - "loss": 0.0113, + "grad_norm": 0.3125064969062805, + "learning_rate": 9.692243042713944e-05, + "loss": 0.0138, "step": 9400 }, { - "grad_norm": 0.2669365406036377, - "learning_rate": 5.897083791151706e-05, - "loss": 0.0082, + "grad_norm": 0.2905729413032532, + "learning_rate": 9.691290430559022e-05, + "loss": 0.011, "step": 9410 }, { - "grad_norm": 0.3623141348361969, - "learning_rate": 5.888949372438336e-05, - "loss": 0.0114, + "grad_norm": 0.32909095287323, + "learning_rate": 9.690336393313932e-05, + "loss": 0.014, "step": 9420 }, { - "grad_norm": 0.28753525018692017, - "learning_rate": 5.8808125233713255e-05, - "loss": 0.0123, + "grad_norm": 0.3342820703983307, + "learning_rate": 9.689380931268487e-05, + "loss": 0.0155, "step": 9430 }, { - "grad_norm": 0.2539946436882019, - "learning_rate": 5.872673266196509e-05, - "loss": 0.0113, + "grad_norm": 0.2983195185661316, + "learning_rate": 9.688424044712932e-05, + "loss": 0.0136, "step": 9440 }, { - "grad_norm": 0.30448299646377563, - "learning_rate": 5.864531623166305e-05, - "loss": 0.0131, + "grad_norm": 0.2790825068950653, + "learning_rate": 9.687465733937942e-05, + "loss": 0.0153, "step": 9450 }, { - "grad_norm": 0.35516154766082764, - "learning_rate": 5.856387616539656e-05, - "loss": 0.0103, + "grad_norm": 0.38151609897613525, + "learning_rate": 9.686505999234627e-05, + "loss": 0.0123, "step": 9460 }, { - "grad_norm": 0.2838006615638733, - "learning_rate": 5.848241268581967e-05, - "loss": 0.0099, + "grad_norm": 0.33219847083091736, + "learning_rate": 9.685544840894529e-05, + "loss": 0.0132, "step": 9470 }, { - "grad_norm": 0.2405579686164856, - "learning_rate": 5.840092601565037e-05, - "loss": 0.0117, + "grad_norm": 0.27068987488746643, + "learning_rate": 9.684582259209624e-05, + "loss": 0.0151, "step": 9480 }, { - "grad_norm": 0.2804074287414551, - "learning_rate": 5.8319416377670144e-05, - "loss": 0.0104, + "grad_norm": 0.32508429884910583, + "learning_rate": 9.683618254472317e-05, + "loss": 0.0137, "step": 9490 }, { - "grad_norm": 0.3132031559944153, - "learning_rate": 5.82378839947232e-05, - "loss": 0.0102, + "grad_norm": 0.4124397039413452, + "learning_rate": 9.682652826975449e-05, + "loss": 0.0134, "step": 9500 }, { - "grad_norm": 0.2594463527202606, - "learning_rate": 5.815632908971599e-05, - "loss": 0.0123, + "grad_norm": 0.30208060145378113, + "learning_rate": 9.681685977012291e-05, + "loss": 0.0154, "step": 9510 }, { - "grad_norm": 0.3764408230781555, - "learning_rate": 5.80747518856165e-05, - "loss": 0.0088, + "grad_norm": 0.31354883313179016, + "learning_rate": 9.680717704876546e-05, + "loss": 0.0118, "step": 9520 }, { - "grad_norm": 0.3109181821346283, - "learning_rate": 5.799315260545367e-05, - "loss": 0.0078, + "grad_norm": 0.318999320268631, + "learning_rate": 9.679748010862349e-05, + "loss": 0.01, "step": 9530 }, { - "grad_norm": 0.26161524653434753, - "learning_rate": 5.791153147231686e-05, - "loss": 0.01, + "grad_norm": 0.284979909658432, + "learning_rate": 9.678776895264267e-05, + "loss": 0.0128, "step": 9540 }, { - "grad_norm": 0.38785234093666077, - "learning_rate": 5.782988870935509e-05, - "loss": 0.0135, + "grad_norm": 0.32563382387161255, + "learning_rate": 9.6778043583773e-05, + "loss": 0.0163, "step": 9550 }, { - "grad_norm": 0.3314606547355652, - "learning_rate": 5.774822453977657e-05, - "loss": 0.0114, + "grad_norm": 0.34569570422172546, + "learning_rate": 9.67683040049688e-05, + "loss": 0.0153, "step": 9560 }, { - "grad_norm": 0.2779804766178131, - "learning_rate": 5.7666539186848036e-05, - "loss": 0.011, + "grad_norm": 0.3157880902290344, + "learning_rate": 9.675855021918869e-05, + "loss": 0.0134, "step": 9570 }, { - "grad_norm": 0.2708481252193451, - "learning_rate": 5.758483287389411e-05, - "loss": 0.0128, + "grad_norm": 0.3187633156776428, + "learning_rate": 9.674878222939561e-05, + "loss": 0.0165, "step": 9580 }, { - "grad_norm": 0.27679356932640076, - "learning_rate": 5.7503105824296735e-05, - "loss": 0.0128, + "grad_norm": 0.33771705627441406, + "learning_rate": 9.673900003855681e-05, + "loss": 0.017, "step": 9590 }, { - "grad_norm": 0.2817783057689667, - "learning_rate": 5.742135826149453e-05, - "loss": 0.0091, + "grad_norm": 0.2888439893722534, + "learning_rate": 9.672920364964389e-05, + "loss": 0.0127, "step": 9600 }, { - "grad_norm": 0.2382742166519165, - "learning_rate": 5.7339590408982223e-05, - "loss": 0.0115, + "grad_norm": 0.3588801622390747, + "learning_rate": 9.671939306563269e-05, + "loss": 0.015, "step": 9610 }, { - "grad_norm": 0.23625558614730835, - "learning_rate": 5.725780249031e-05, - "loss": 0.0086, + "grad_norm": 0.3220963478088379, + "learning_rate": 9.670956828950345e-05, + "loss": 0.0119, "step": 9620 }, { - "grad_norm": 0.30190321803092957, - "learning_rate": 5.717599472908292e-05, - "loss": 0.0097, + "grad_norm": 0.3182002604007721, + "learning_rate": 9.669972932424065e-05, + "loss": 0.0127, "step": 9630 }, { - "grad_norm": 0.31747740507125854, - "learning_rate": 5.7094167348960237e-05, - "loss": 0.0129, + "grad_norm": 0.362656831741333, + "learning_rate": 9.668987617283312e-05, + "loss": 0.0161, "step": 9640 }, { - "grad_norm": 0.27004364132881165, - "learning_rate": 5.7012320573654945e-05, - "loss": 0.0126, + "grad_norm": 0.3145112991333008, + "learning_rate": 9.668000883827397e-05, + "loss": 0.0158, "step": 9650 }, { - "grad_norm": 0.2586357593536377, - "learning_rate": 5.693045462693295e-05, - "loss": 0.0124, + "grad_norm": 0.26826515793800354, + "learning_rate": 9.667012732356067e-05, + "loss": 0.0147, "step": 9660 }, { - "grad_norm": 0.2702503502368927, - "learning_rate": 5.684856973261266e-05, - "loss": 0.009, + "grad_norm": 0.33713433146476746, + "learning_rate": 9.666023163169493e-05, + "loss": 0.0109, "step": 9670 }, { - "grad_norm": 0.33263716101646423, - "learning_rate": 5.6766666114564215e-05, - "loss": 0.0087, + "grad_norm": 0.29669779539108276, + "learning_rate": 9.665032176568281e-05, + "loss": 0.0108, "step": 9680 }, { - "grad_norm": 0.3644093871116638, - "learning_rate": 5.668474399670899e-05, - "loss": 0.0119, + "grad_norm": 0.3480658233165741, + "learning_rate": 9.664039772853469e-05, + "loss": 0.0152, "step": 9690 }, { - "grad_norm": 0.29676637053489685, - "learning_rate": 5.660280360301896e-05, - "loss": 0.0115, + "grad_norm": 0.395332396030426, + "learning_rate": 9.663045952326518e-05, + "loss": 0.0148, "step": 9700 }, { - "grad_norm": 0.2950310707092285, - "learning_rate": 5.652084515751599e-05, - "loss": 0.0098, + "grad_norm": 0.40417957305908203, + "learning_rate": 9.662050715289328e-05, + "loss": 0.0142, "step": 9710 }, { - "grad_norm": 0.30780714750289917, - "learning_rate": 5.643886888427137e-05, - "loss": 0.0105, + "grad_norm": 0.3749469220638275, + "learning_rate": 9.661054062044226e-05, + "loss": 0.0143, "step": 9720 }, { - "grad_norm": 0.31481608748435974, - "learning_rate": 5.6356875007405074e-05, - "loss": 0.0101, + "grad_norm": 0.33351844549179077, + "learning_rate": 9.660055992893968e-05, + "loss": 0.0127, "step": 9730 }, { - "grad_norm": 0.2832598090171814, - "learning_rate": 5.627486375108525e-05, - "loss": 0.0113, + "grad_norm": 0.3418681025505066, + "learning_rate": 9.659056508141739e-05, + "loss": 0.0135, "step": 9740 }, { - "grad_norm": 0.3724987208843231, - "learning_rate": 5.619283533952754e-05, - "loss": 0.0122, + "grad_norm": 0.47097650170326233, + "learning_rate": 9.658055608091161e-05, + "loss": 0.0163, "step": 9750 }, { - "grad_norm": 0.47322213649749756, - "learning_rate": 5.6110789996994474e-05, - "loss": 0.0114, + "grad_norm": 0.4399387240409851, + "learning_rate": 9.657053293046276e-05, + "loss": 0.0149, "step": 9760 }, { - "grad_norm": 0.31387490034103394, - "learning_rate": 5.602872794779491e-05, - "loss": 0.0097, + "grad_norm": 0.39624348282814026, + "learning_rate": 9.656049563311564e-05, + "loss": 0.0126, "step": 9770 }, { - "grad_norm": 0.33728551864624023, - "learning_rate": 5.594664941628334e-05, - "loss": 0.0084, + "grad_norm": 0.3995211720466614, + "learning_rate": 9.655044419191929e-05, + "loss": 0.0117, "step": 9780 }, { - "grad_norm": 0.2974473237991333, - "learning_rate": 5.5864554626859324e-05, - "loss": 0.0096, + "grad_norm": 0.38814017176628113, + "learning_rate": 9.654037860992711e-05, + "loss": 0.0133, "step": 9790 }, { - "grad_norm": 0.2743162214756012, - "learning_rate": 5.578244380396691e-05, - "loss": 0.0102, + "grad_norm": 0.2618493139743805, + "learning_rate": 9.653029889019672e-05, + "loss": 0.0136, "step": 9800 }, { - "grad_norm": 0.2396203577518463, - "learning_rate": 5.570031717209394e-05, - "loss": 0.0172, + "grad_norm": 0.3224608898162842, + "learning_rate": 9.65202050357901e-05, + "loss": 0.0199, "step": 9810 }, { - "grad_norm": 0.3160618245601654, - "learning_rate": 5.561817495577147e-05, - "loss": 0.0117, + "grad_norm": 0.43517857789993286, + "learning_rate": 9.651009704977347e-05, + "loss": 0.0166, "step": 9820 }, { - "grad_norm": 0.31184670329093933, - "learning_rate": 5.5536017379573215e-05, - "loss": 0.0131, + "grad_norm": 0.3874225318431854, + "learning_rate": 9.649997493521738e-05, + "loss": 0.0178, "step": 9830 }, { - "grad_norm": 0.2700205445289612, - "learning_rate": 5.545384466811483e-05, - "loss": 0.0131, + "grad_norm": 0.27843573689460754, + "learning_rate": 9.64898386951967e-05, + "loss": 0.0171, "step": 9840 }, { - "grad_norm": 0.29386454820632935, - "learning_rate": 5.5371657046053384e-05, - "loss": 0.013, + "grad_norm": 0.40193015336990356, + "learning_rate": 9.647968833279049e-05, + "loss": 0.0163, "step": 9850 }, { - "grad_norm": 0.30857086181640625, - "learning_rate": 5.528945473808669e-05, - "loss": 0.0112, + "grad_norm": 0.2597414553165436, + "learning_rate": 9.646952385108218e-05, + "loss": 0.0148, "step": 9860 }, { - "grad_norm": 0.31214991211891174, - "learning_rate": 5.520723796895272e-05, - "loss": 0.0082, + "grad_norm": 0.27404460310935974, + "learning_rate": 9.645934525315951e-05, + "loss": 0.0115, "step": 9870 }, { - "grad_norm": 0.3384791910648346, - "learning_rate": 5.512500696342897e-05, - "loss": 0.0087, + "grad_norm": 0.3651256561279297, + "learning_rate": 9.644915254211442e-05, + "loss": 0.0121, "step": 9880 }, { - "grad_norm": 0.26187098026275635, - "learning_rate": 5.504276194633188e-05, - "loss": 0.0093, + "grad_norm": 0.3848916292190552, + "learning_rate": 9.643894572104321e-05, + "loss": 0.0121, "step": 9890 }, { - "grad_norm": 0.3037872314453125, - "learning_rate": 5.49605031425162e-05, - "loss": 0.0092, + "grad_norm": 0.3190523386001587, + "learning_rate": 9.642872479304644e-05, + "loss": 0.0126, "step": 9900 }, { - "grad_norm": 0.28506723046302795, - "learning_rate": 5.487823077687434e-05, - "loss": 0.0099, + "grad_norm": 0.3862152397632599, + "learning_rate": 9.641848976122895e-05, + "loss": 0.0145, "step": 9910 }, { - "grad_norm": 0.3275642693042755, - "learning_rate": 5.4795945074335806e-05, - "loss": 0.009, + "grad_norm": 0.3423781096935272, + "learning_rate": 9.64082406286999e-05, + "loss": 0.0137, "step": 9920 }, { - "grad_norm": 0.29687318205833435, - "learning_rate": 5.471364625986657e-05, - "loss": 0.0112, + "grad_norm": 0.35022684931755066, + "learning_rate": 9.639797739857269e-05, + "loss": 0.0145, "step": 9930 }, { - "grad_norm": 0.24183917045593262, - "learning_rate": 5.463133455846845e-05, - "loss": 0.0088, + "grad_norm": 0.33175429701805115, + "learning_rate": 9.638770007396498e-05, + "loss": 0.011, "step": 9940 }, { - "grad_norm": 0.32168254256248474, - "learning_rate": 5.4549010195178505e-05, - "loss": 0.0141, + "grad_norm": 0.35739460587501526, + "learning_rate": 9.63774086579988e-05, + "loss": 0.0172, "step": 9950 }, { - "grad_norm": 0.2662130892276764, - "learning_rate": 5.446667339506838e-05, - "loss": 0.0104, + "grad_norm": 0.3348420262336731, + "learning_rate": 9.63671031538004e-05, + "loss": 0.0133, "step": 9960 }, { - "grad_norm": 0.32166045904159546, - "learning_rate": 5.4384324383243756e-05, - "loss": 0.0098, + "grad_norm": 0.327800452709198, + "learning_rate": 9.635678356450031e-05, + "loss": 0.0128, "step": 9970 }, { - "grad_norm": 0.22883684933185577, - "learning_rate": 5.430196338484368e-05, - "loss": 0.0078, + "grad_norm": 0.2548326253890991, + "learning_rate": 9.634644989323336e-05, + "loss": 0.0109, "step": 9980 }, { - "grad_norm": 0.2235536277294159, - "learning_rate": 5.4219590625039975e-05, - "loss": 0.0097, + "grad_norm": 0.36686092615127563, + "learning_rate": 9.633610214313861e-05, + "loss": 0.0131, "step": 9990 }, { - "grad_norm": 0.2653367221355438, - "learning_rate": 5.413720632903664e-05, - "loss": 0.0103, + "grad_norm": 0.37838801741600037, + "learning_rate": 9.632574031735951e-05, + "loss": 0.0137, "step": 10000 }, { - "grad_norm": 0.25344082713127136, - "learning_rate": 5.405481072206917e-05, - "loss": 0.0089, + "grad_norm": 0.26175743341445923, + "learning_rate": 9.631536441904364e-05, + "loss": 0.011, "step": 10010 }, { - "grad_norm": 0.335708886384964, - "learning_rate": 5.397240402940402e-05, - "loss": 0.0102, + "grad_norm": 0.3280612826347351, + "learning_rate": 9.630497445134293e-05, + "loss": 0.0134, "step": 10020 }, { - "grad_norm": 0.2804220914840698, - "learning_rate": 5.388998647633794e-05, - "loss": 0.0119, + "grad_norm": 0.2718747854232788, + "learning_rate": 9.62945704174136e-05, + "loss": 0.0141, "step": 10030 }, { - "grad_norm": 0.3300700783729553, - "learning_rate": 5.380755828819737e-05, - "loss": 0.009, + "grad_norm": 0.22997650504112244, + "learning_rate": 9.628415232041612e-05, + "loss": 0.0121, "step": 10040 }, { - "grad_norm": 0.3184926211833954, - "learning_rate": 5.3725119690337846e-05, - "loss": 0.01, + "grad_norm": 0.2900289297103882, + "learning_rate": 9.627372016351524e-05, + "loss": 0.0126, "step": 10050 }, { - "grad_norm": 0.2888861894607544, - "learning_rate": 5.3642670908143324e-05, - "loss": 0.0085, + "grad_norm": 0.27054861187934875, + "learning_rate": 9.626327394987995e-05, + "loss": 0.0116, "step": 10060 }, { - "grad_norm": 0.2378275841474533, - "learning_rate": 5.356021216702562e-05, - "loss": 0.0083, + "grad_norm": 0.3369966447353363, + "learning_rate": 9.625281368268355e-05, + "loss": 0.0121, "step": 10070 }, { - "grad_norm": 0.30194157361984253, - "learning_rate": 5.347774369242381e-05, - "loss": 0.0144, + "grad_norm": 0.2885831594467163, + "learning_rate": 9.624233936510357e-05, + "loss": 0.0185, "step": 10080 }, { - "grad_norm": 0.30574026703834534, - "learning_rate": 5.3395265709803545e-05, - "loss": 0.0106, + "grad_norm": 0.34082889556884766, + "learning_rate": 9.623185100032187e-05, + "loss": 0.0128, "step": 10090 }, { - "grad_norm": 0.2707749307155609, - "learning_rate": 5.331277844465647e-05, - "loss": 0.009, + "grad_norm": 0.38311946392059326, + "learning_rate": 9.62213485915245e-05, + "loss": 0.0133, "step": 10100 }, { - "grad_norm": 0.2563587427139282, - "learning_rate": 5.323028212249963e-05, - "loss": 0.0079, + "grad_norm": 0.3015669584274292, + "learning_rate": 9.621083214190186e-05, + "loss": 0.0106, "step": 10110 }, { - "grad_norm": 0.23816479742527008, - "learning_rate": 5.314777696887481e-05, - "loss": 0.0092, + "grad_norm": 0.2607632577419281, + "learning_rate": 9.62003016546485e-05, + "loss": 0.0122, "step": 10120 }, { - "grad_norm": 0.2609695494174957, - "learning_rate": 5.306526320934796e-05, - "loss": 0.0115, + "grad_norm": 0.3680003881454468, + "learning_rate": 9.618975713296339e-05, + "loss": 0.0147, "step": 10130 }, { - "grad_norm": 0.32091790437698364, - "learning_rate": 5.298274106950854e-05, - "loss": 0.009, + "grad_norm": 0.4181278347969055, + "learning_rate": 9.61791985800496e-05, + "loss": 0.0115, "step": 10140 }, { - "grad_norm": 0.352911114692688, - "learning_rate": 5.290021077496893e-05, - "loss": 0.0088, + "grad_norm": 0.365619033575058, + "learning_rate": 9.616862599911458e-05, + "loss": 0.0118, "step": 10150 }, { - "grad_norm": 0.2584521174430847, - "learning_rate": 5.2817672551363816e-05, - "loss": 0.0098, + "grad_norm": 0.3310202658176422, + "learning_rate": 9.615803939337e-05, + "loss": 0.013, "step": 10160 }, { - "grad_norm": 0.25784599781036377, - "learning_rate": 5.273512662434952e-05, - "loss": 0.0107, + "grad_norm": 0.31007471680641174, + "learning_rate": 9.614743876603178e-05, + "loss": 0.0142, "step": 10170 }, { - "grad_norm": 0.3719595670700073, - "learning_rate": 5.265257321960349e-05, - "loss": 0.011, + "grad_norm": 0.269034206867218, + "learning_rate": 9.613682412032013e-05, + "loss": 0.0143, "step": 10180 }, { - "grad_norm": 0.30774036049842834, - "learning_rate": 5.257001256282357e-05, - "loss": 0.0096, + "grad_norm": 0.34003397822380066, + "learning_rate": 9.612619545945947e-05, + "loss": 0.0131, "step": 10190 }, { - "grad_norm": 0.3356902599334717, - "learning_rate": 5.248744487972742e-05, - "loss": 0.0094, + "grad_norm": 0.2762850821018219, + "learning_rate": 9.611555278667852e-05, + "loss": 0.0128, "step": 10200 }, { - "grad_norm": 0.3237636685371399, - "learning_rate": 5.240487039605196e-05, - "loss": 0.0113, + "grad_norm": 0.30333468317985535, + "learning_rate": 9.610489610521024e-05, + "loss": 0.016, "step": 10210 }, { - "grad_norm": 0.3064771890640259, - "learning_rate": 5.232228933755267e-05, - "loss": 0.0128, + "grad_norm": 0.3439791202545166, + "learning_rate": 9.609422541829187e-05, + "loss": 0.0168, "step": 10220 }, { - "grad_norm": 0.3693506419658661, - "learning_rate": 5.2239701930003006e-05, - "loss": 0.0099, + "grad_norm": 0.3174102306365967, + "learning_rate": 9.608354072916486e-05, + "loss": 0.0137, "step": 10230 }, { - "grad_norm": 0.2279125452041626, - "learning_rate": 5.215710839919379e-05, - "loss": 0.0092, + "grad_norm": 0.26099175214767456, + "learning_rate": 9.607284204107493e-05, + "loss": 0.0118, "step": 10240 }, { - "grad_norm": 0.2967457175254822, - "learning_rate": 5.207450897093257e-05, - "loss": 0.0082, + "grad_norm": 0.3374234735965729, + "learning_rate": 9.606212935727208e-05, + "loss": 0.0113, "step": 10250 }, { - "grad_norm": 0.2727642059326172, - "learning_rate": 5.1991903871043046e-05, - "loss": 0.0125, + "grad_norm": 0.3489407002925873, + "learning_rate": 9.605140268101052e-05, + "loss": 0.0162, "step": 10260 }, { - "grad_norm": 0.3166670501232147, - "learning_rate": 5.190929332536439e-05, - "loss": 0.0109, + "grad_norm": 0.3476567566394806, + "learning_rate": 9.604066201554875e-05, + "loss": 0.0149, "step": 10270 }, { - "grad_norm": 0.3417263329029083, - "learning_rate": 5.182667755975071e-05, - "loss": 0.0121, + "grad_norm": 0.3224858343601227, + "learning_rate": 9.60299073641495e-05, + "loss": 0.0156, "step": 10280 }, { - "grad_norm": 0.24792048335075378, - "learning_rate": 5.1744056800070315e-05, - "loss": 0.0089, + "grad_norm": 0.2694963216781616, + "learning_rate": 9.601913873007974e-05, + "loss": 0.0126, "step": 10290 }, { - "grad_norm": 0.28874701261520386, - "learning_rate": 5.166143127220524e-05, - "loss": 0.0099, + "grad_norm": 0.3785710036754608, + "learning_rate": 9.60083561166107e-05, + "loss": 0.0129, "step": 10300 }, { - "grad_norm": 0.33919915556907654, - "learning_rate": 5.1578801202050485e-05, - "loss": 0.0092, + "grad_norm": 0.44782793521881104, + "learning_rate": 9.599755952701783e-05, + "loss": 0.0125, "step": 10310 }, { - "grad_norm": 0.24734066426753998, - "learning_rate": 5.149616681551355e-05, - "loss": 0.0084, + "grad_norm": 0.2835095226764679, + "learning_rate": 9.598674896458089e-05, + "loss": 0.012, "step": 10320 }, { - "grad_norm": 0.29600366950035095, - "learning_rate": 5.141352833851367e-05, - "loss": 0.011, + "grad_norm": 0.2620943784713745, + "learning_rate": 9.597592443258383e-05, + "loss": 0.0139, "step": 10330 }, { - "grad_norm": 0.26344043016433716, - "learning_rate": 5.1330885996981285e-05, - "loss": 0.0084, + "grad_norm": 0.22041644155979156, + "learning_rate": 9.596508593431483e-05, + "loss": 0.0116, "step": 10340 }, { - "grad_norm": 0.29486021399497986, - "learning_rate": 5.124824001685741e-05, - "loss": 0.0128, + "grad_norm": 0.4375811815261841, + "learning_rate": 9.59542334730664e-05, + "loss": 0.0167, "step": 10350 }, { - "grad_norm": 0.3321808874607086, - "learning_rate": 5.116559062409298e-05, - "loss": 0.0108, + "grad_norm": 0.4061785936355591, + "learning_rate": 9.594336705213516e-05, + "loss": 0.0155, "step": 10360 }, { - "grad_norm": 0.3323744535446167, - "learning_rate": 5.10829380446483e-05, - "loss": 0.0102, + "grad_norm": 0.26017096638679504, + "learning_rate": 9.593248667482208e-05, + "loss": 0.0136, "step": 10370 }, { - "grad_norm": 0.2918391227722168, - "learning_rate": 5.100028250449235e-05, - "loss": 0.0086, + "grad_norm": 0.3989296853542328, + "learning_rate": 9.592159234443233e-05, + "loss": 0.0123, "step": 10380 }, { - "grad_norm": 0.2917460501194, - "learning_rate": 5.0917624229602234e-05, - "loss": 0.0092, + "grad_norm": 0.35283365845680237, + "learning_rate": 9.59106840642753e-05, + "loss": 0.0124, "step": 10390 }, { - "grad_norm": 0.24862314760684967, - "learning_rate": 5.0834963445962524e-05, - "loss": 0.0092, + "grad_norm": 0.32362040877342224, + "learning_rate": 9.589976183766467e-05, + "loss": 0.0125, "step": 10400 }, { - "grad_norm": 0.3336447775363922, - "learning_rate": 5.075230037956461e-05, - "loss": 0.0089, + "grad_norm": 0.33966749906539917, + "learning_rate": 9.58888256679183e-05, + "loss": 0.0124, "step": 10410 }, { - "grad_norm": 0.2436840683221817, - "learning_rate": 5.0669635256406213e-05, - "loss": 0.0074, + "grad_norm": 0.22488029301166534, + "learning_rate": 9.587787555835832e-05, + "loss": 0.01, "step": 10420 }, { - "grad_norm": 0.23710206151008606, - "learning_rate": 5.058696830249058e-05, - "loss": 0.0109, + "grad_norm": 0.2843591570854187, + "learning_rate": 9.586691151231107e-05, + "loss": 0.0132, "step": 10430 }, { - "grad_norm": 0.2947542071342468, - "learning_rate": 5.050429974382602e-05, - "loss": 0.0099, + "grad_norm": 0.3061405122280121, + "learning_rate": 9.585593353310715e-05, + "loss": 0.0121, "step": 10440 }, { - "grad_norm": 0.28282612562179565, - "learning_rate": 5.042162980642523e-05, - "loss": 0.0083, + "grad_norm": 0.2765599191188812, + "learning_rate": 9.58449416240814e-05, + "loss": 0.0116, "step": 10450 }, { - "grad_norm": 0.2697611451148987, - "learning_rate": 5.033895871630462e-05, - "loss": 0.0096, + "grad_norm": 0.3570658266544342, + "learning_rate": 9.583393578857283e-05, + "loss": 0.0136, "step": 10460 }, { - "grad_norm": 0.28840044140815735, - "learning_rate": 5.025628669948386e-05, - "loss": 0.0099, + "grad_norm": 0.3393988609313965, + "learning_rate": 9.582291602992474e-05, + "loss": 0.0132, "step": 10470 }, { - "grad_norm": 0.26551976799964905, - "learning_rate": 5.017361398198502e-05, - "loss": 0.0095, + "grad_norm": 0.2755260765552521, + "learning_rate": 9.581188235148466e-05, + "loss": 0.0126, "step": 10480 }, { - "grad_norm": 0.344992458820343, - "learning_rate": 5.009094078983221e-05, - "loss": 0.009, + "grad_norm": 0.3000558614730835, + "learning_rate": 9.58008347566043e-05, + "loss": 0.0127, "step": 10490 }, { - "grad_norm": 0.33870258927345276, - "learning_rate": 5.000826734905073e-05, - "loss": 0.0086, + "grad_norm": 0.34652918577194214, + "learning_rate": 9.578977324863965e-05, + "loss": 0.0135, "step": 10500 }, { - "grad_norm": 0.26343870162963867, - "learning_rate": 4.9925593885666645e-05, - "loss": 0.0095, + "grad_norm": 0.30220451951026917, + "learning_rate": 9.577869783095089e-05, + "loss": 0.0141, "step": 10510 }, { - "grad_norm": 0.30816709995269775, - "learning_rate": 4.984292062570602e-05, - "loss": 0.0095, + "grad_norm": 0.38388505578041077, + "learning_rate": 9.576760850690245e-05, + "loss": 0.0148, "step": 10520 }, { - "grad_norm": 0.24630755186080933, - "learning_rate": 4.976024779519442e-05, - "loss": 0.009, + "grad_norm": 0.3771129846572876, + "learning_rate": 9.575650527986298e-05, + "loss": 0.0142, "step": 10530 }, { - "grad_norm": 0.2802034616470337, - "learning_rate": 4.9677575620156194e-05, - "loss": 0.0094, + "grad_norm": 0.3310950696468353, + "learning_rate": 9.574538815320531e-05, + "loss": 0.015, "step": 10540 }, { - "grad_norm": 0.41168075799942017, - "learning_rate": 4.959490432661391e-05, - "loss": 0.012, + "grad_norm": 0.40225401520729065, + "learning_rate": 9.573425713030656e-05, + "loss": 0.0151, "step": 10550 }, { - "grad_norm": 0.2359355241060257, - "learning_rate": 4.9512234140587726e-05, - "loss": 0.0089, + "grad_norm": 0.27139759063720703, + "learning_rate": 9.572311221454806e-05, + "loss": 0.0133, "step": 10560 }, { - "grad_norm": 0.26179102063179016, - "learning_rate": 4.942956528809477e-05, - "loss": 0.011, + "grad_norm": 0.4283144772052765, + "learning_rate": 9.57119534093153e-05, + "loss": 0.0149, "step": 10570 }, { - "grad_norm": 0.21987201273441315, - "learning_rate": 4.934689799514854e-05, - "loss": 0.008, + "grad_norm": 0.275490939617157, + "learning_rate": 9.570078071799806e-05, + "loss": 0.0121, "step": 10580 }, { - "grad_norm": 0.2534409165382385, - "learning_rate": 4.926423248775827e-05, - "loss": 0.0107, + "grad_norm": 0.318310409784317, + "learning_rate": 9.568959414399028e-05, + "loss": 0.0144, "step": 10590 }, { - "grad_norm": 0.32994967699050903, - "learning_rate": 4.918156899192826e-05, - "loss": 0.0106, + "grad_norm": 0.3069532811641693, + "learning_rate": 9.567839369069018e-05, + "loss": 0.0137, "step": 10600 }, { - "grad_norm": 0.19284388422966003, - "learning_rate": 4.909890773365738e-05, - "loss": 0.0117, + "grad_norm": 0.2492484301328659, + "learning_rate": 9.566717936150013e-05, + "loss": 0.0156, "step": 10610 }, { - "grad_norm": 0.2593536078929901, - "learning_rate": 4.9016248938938344e-05, - "loss": 0.0125, + "grad_norm": 0.33000829815864563, + "learning_rate": 9.565595115982678e-05, + "loss": 0.0175, "step": 10620 }, { - "grad_norm": 0.34668925404548645, - "learning_rate": 4.8933592833757156e-05, - "loss": 0.0087, + "grad_norm": 0.34279122948646545, + "learning_rate": 9.564470908908094e-05, + "loss": 0.0132, "step": 10630 }, { - "grad_norm": 0.30534157156944275, - "learning_rate": 4.8850939644092435e-05, - "loss": 0.0119, + "grad_norm": 0.37935033440589905, + "learning_rate": 9.563345315267764e-05, + "loss": 0.0154, "step": 10640 }, { - "grad_norm": 0.2502377927303314, - "learning_rate": 4.876828959591485e-05, - "loss": 0.0087, + "grad_norm": 0.24085286259651184, + "learning_rate": 9.562218335403616e-05, + "loss": 0.012, "step": 10650 }, { - "grad_norm": 0.2559858560562134, - "learning_rate": 4.8685642915186474e-05, - "loss": 0.0098, + "grad_norm": 0.37656792998313904, + "learning_rate": 9.561089969657999e-05, + "loss": 0.0131, "step": 10660 }, { - "grad_norm": 0.17750589549541473, - "learning_rate": 4.860299982786018e-05, - "loss": 0.0084, + "grad_norm": 0.2326240986585617, + "learning_rate": 9.559960218373673e-05, + "loss": 0.0113, "step": 10670 }, { - "grad_norm": 0.29540231823921204, - "learning_rate": 4.852036055987901e-05, - "loss": 0.0069, + "grad_norm": 0.4000895321369171, + "learning_rate": 9.558829081893836e-05, + "loss": 0.01, "step": 10680 }, { - "grad_norm": 0.20562411844730377, - "learning_rate": 4.843772533717558e-05, - "loss": 0.0083, + "grad_norm": 0.35430994629859924, + "learning_rate": 9.55769656056209e-05, + "loss": 0.0121, "step": 10690 }, { - "grad_norm": 0.25039413571357727, - "learning_rate": 4.835509438567142e-05, - "loss": 0.0092, + "grad_norm": 0.30428677797317505, + "learning_rate": 9.556562654722469e-05, + "loss": 0.0133, "step": 10700 }, { - "grad_norm": 0.2514692544937134, - "learning_rate": 4.827246793127639e-05, - "loss": 0.0077, + "grad_norm": 0.25132450461387634, + "learning_rate": 9.555427364719422e-05, + "loss": 0.011, "step": 10710 }, { - "grad_norm": 0.30662453174591064, - "learning_rate": 4.818984619988807e-05, - "loss": 0.0107, + "grad_norm": 0.4662877321243286, + "learning_rate": 9.55429069089782e-05, + "loss": 0.0146, "step": 10720 }, { - "grad_norm": 0.26225242018699646, - "learning_rate": 4.810722941739115e-05, - "loss": 0.0104, + "grad_norm": 0.29919013381004333, + "learning_rate": 9.553152633602956e-05, + "loss": 0.014, "step": 10730 }, { - "grad_norm": 0.3066365122795105, - "learning_rate": 4.8024617809656684e-05, - "loss": 0.0085, + "grad_norm": 0.32089173793792725, + "learning_rate": 9.552013193180543e-05, + "loss": 0.0123, "step": 10740 }, { - "grad_norm": 0.27893194556236267, - "learning_rate": 4.794201160254171e-05, - "loss": 0.0077, + "grad_norm": 0.3637930154800415, + "learning_rate": 9.550872369976707e-05, + "loss": 0.0118, "step": 10750 }, { - "grad_norm": 0.2807646095752716, - "learning_rate": 4.785941102188844e-05, - "loss": 0.0096, + "grad_norm": 0.2593134939670563, + "learning_rate": 9.549730164338007e-05, + "loss": 0.0125, "step": 10760 }, { - "grad_norm": 0.299655944108963, - "learning_rate": 4.7776816293523686e-05, - "loss": 0.0108, + "grad_norm": 0.3442279100418091, + "learning_rate": 9.548586576611408e-05, + "loss": 0.0141, "step": 10770 }, { - "grad_norm": 0.2901594638824463, - "learning_rate": 4.769422764325832e-05, - "loss": 0.0086, + "grad_norm": 0.32184478640556335, + "learning_rate": 9.54744160714431e-05, + "loss": 0.0119, "step": 10780 }, { - "grad_norm": 0.3350774943828583, - "learning_rate": 4.76116452968865e-05, - "loss": 0.0124, + "grad_norm": 0.29462918639183044, + "learning_rate": 9.546295256284516e-05, + "loss": 0.0165, "step": 10790 }, { - "grad_norm": 0.29051363468170166, - "learning_rate": 4.752906948018525e-05, - "loss": 0.0097, + "grad_norm": 0.3887004852294922, + "learning_rate": 9.545147524380265e-05, + "loss": 0.0142, "step": 10800 }, { - "grad_norm": 0.26909443736076355, - "learning_rate": 4.7446500418913684e-05, - "loss": 0.0076, + "grad_norm": 0.3077283501625061, + "learning_rate": 9.543998411780201e-05, + "loss": 0.0107, "step": 10810 }, { - "grad_norm": 0.28262028098106384, - "learning_rate": 4.736393833881247e-05, - "loss": 0.0097, + "grad_norm": 0.2644650340080261, + "learning_rate": 9.542847918833397e-05, + "loss": 0.013, "step": 10820 }, { - "grad_norm": 0.22907830774784088, - "learning_rate": 4.7281383465603194e-05, - "loss": 0.0094, + "grad_norm": 0.2867148220539093, + "learning_rate": 9.541696045889343e-05, + "loss": 0.0137, "step": 10830 }, { - "grad_norm": 0.38901764154434204, - "learning_rate": 4.71988360249877e-05, - "loss": 0.0085, + "grad_norm": 0.3509559631347656, + "learning_rate": 9.540542793297947e-05, + "loss": 0.0129, "step": 10840 }, { - "grad_norm": 0.2833888828754425, - "learning_rate": 4.7116296242647554e-05, - "loss": 0.0076, + "grad_norm": 0.26982375979423523, + "learning_rate": 9.539388161409537e-05, + "loss": 0.0119, "step": 10850 }, { - "grad_norm": 0.2758328318595886, - "learning_rate": 4.703376434424336e-05, - "loss": 0.0106, + "grad_norm": 0.26333704590797424, + "learning_rate": 9.538232150574857e-05, + "loss": 0.0142, "step": 10860 }, { - "grad_norm": 0.26278847455978394, - "learning_rate": 4.695124055541421e-05, - "loss": 0.01, + "grad_norm": 0.3465039134025574, + "learning_rate": 9.537074761145076e-05, + "loss": 0.014, "step": 10870 }, { - "grad_norm": 0.3171476125717163, - "learning_rate": 4.6868725101776934e-05, - "loss": 0.0086, + "grad_norm": 0.27450138330459595, + "learning_rate": 9.535915993471778e-05, + "loss": 0.0119, "step": 10880 }, { - "grad_norm": 0.3477509915828705, - "learning_rate": 4.678621820892567e-05, - "loss": 0.0101, + "grad_norm": 0.3773563504219055, + "learning_rate": 9.534755847906964e-05, + "loss": 0.014, "step": 10890 }, { - "grad_norm": 0.30033016204833984, - "learning_rate": 4.670372010243111e-05, - "loss": 0.0088, + "grad_norm": 0.30463284254074097, + "learning_rate": 9.533594324803057e-05, + "loss": 0.0116, "step": 10900 }, { - "grad_norm": 0.3467949628829956, - "learning_rate": 4.662123100783992e-05, - "loss": 0.0095, + "grad_norm": 0.3722027540206909, + "learning_rate": 9.532431424512895e-05, + "loss": 0.0127, "step": 10910 }, { - "grad_norm": 0.24811244010925293, - "learning_rate": 4.653875115067415e-05, - "loss": 0.0102, + "grad_norm": 0.28425273299217224, + "learning_rate": 9.531267147389741e-05, + "loss": 0.0134, "step": 10920 }, { - "grad_norm": 0.33560365438461304, - "learning_rate": 4.6456280756430545e-05, - "loss": 0.0096, + "grad_norm": 0.30973801016807556, + "learning_rate": 9.530101493787266e-05, + "loss": 0.0132, "step": 10930 }, { - "grad_norm": 0.26739415526390076, - "learning_rate": 4.637382005058004e-05, - "loss": 0.01, + "grad_norm": 0.3881785571575165, + "learning_rate": 9.528934464059571e-05, + "loss": 0.0139, "step": 10940 }, { - "grad_norm": 0.2914515733718872, - "learning_rate": 4.629136925856705e-05, - "loss": 0.0068, + "grad_norm": 0.29282981157302856, + "learning_rate": 9.527766058561163e-05, + "loss": 0.0109, "step": 10950 }, { - "grad_norm": 0.256402850151062, - "learning_rate": 4.6208928605808895e-05, - "loss": 0.0079, + "grad_norm": 0.3170948624610901, + "learning_rate": 9.526596277646976e-05, + "loss": 0.0113, "step": 10960 }, { - "grad_norm": 0.19222067296504974, - "learning_rate": 4.612649831769519e-05, - "loss": 0.0082, + "grad_norm": 0.2517901062965393, + "learning_rate": 9.525425121672358e-05, + "loss": 0.0115, "step": 10970 }, { - "grad_norm": 0.33148807287216187, - "learning_rate": 4.604407861958715e-05, - "loss": 0.0094, + "grad_norm": 0.32891780138015747, + "learning_rate": 9.524252590993074e-05, + "loss": 0.0135, "step": 10980 }, { - "grad_norm": 0.35632526874542236, - "learning_rate": 4.5961669736817114e-05, - "loss": 0.0102, + "grad_norm": 0.2312023788690567, + "learning_rate": 9.523078685965309e-05, + "loss": 0.0136, "step": 10990 }, { - "grad_norm": 0.3180091381072998, - "learning_rate": 4.5879271894687814e-05, - "loss": 0.0096, + "grad_norm": 0.3014102280139923, + "learning_rate": 9.521903406945664e-05, + "loss": 0.0128, "step": 11000 }, { - "grad_norm": 0.3015517592430115, - "learning_rate": 4.5796885318471826e-05, - "loss": 0.0075, + "grad_norm": 0.3667161166667938, + "learning_rate": 9.520726754291158e-05, + "loss": 0.011, "step": 11010 }, { - "grad_norm": 0.24222354590892792, - "learning_rate": 4.571451023341086e-05, - "loss": 0.0083, + "grad_norm": 0.320538729429245, + "learning_rate": 9.519548728359227e-05, + "loss": 0.0111, "step": 11020 }, { - "grad_norm": 0.3168736696243286, - "learning_rate": 4.563214686471527e-05, - "loss": 0.0091, + "grad_norm": 0.31149187684059143, + "learning_rate": 9.518369329507726e-05, + "loss": 0.0122, "step": 11030 }, { - "grad_norm": 0.20201337337493896, - "learning_rate": 4.5549795437563365e-05, - "loss": 0.0082, + "grad_norm": 0.3280191421508789, + "learning_rate": 9.51718855809492e-05, + "loss": 0.0112, "step": 11040 }, { - "grad_norm": 0.3413758873939514, - "learning_rate": 4.546745617710081e-05, - "loss": 0.0091, + "grad_norm": 0.4067302942276001, + "learning_rate": 9.516006414479502e-05, + "loss": 0.0129, "step": 11050 }, { - "grad_norm": 0.19693852961063385, - "learning_rate": 4.5385129308440014e-05, - "loss": 0.0089, + "grad_norm": 0.23629716038703918, + "learning_rate": 9.514822899020572e-05, + "loss": 0.0125, "step": 11060 }, { - "grad_norm": 0.2525441646575928, - "learning_rate": 4.530281505665944e-05, - "loss": 0.0088, + "grad_norm": 0.27616390585899353, + "learning_rate": 9.513638012077654e-05, + "loss": 0.0124, "step": 11070 }, { - "grad_norm": 0.2551596462726593, - "learning_rate": 4.5220513646803134e-05, - "loss": 0.0105, + "grad_norm": 0.3262661099433899, + "learning_rate": 9.512451754010683e-05, + "loss": 0.0133, "step": 11080 }, { - "grad_norm": 0.1865827739238739, - "learning_rate": 4.513822530388003e-05, - "loss": 0.0077, + "grad_norm": 0.28653451800346375, + "learning_rate": 9.511264125180013e-05, + "loss": 0.0121, "step": 11090 }, { - "grad_norm": 0.25236818194389343, - "learning_rate": 4.5055950252863296e-05, - "loss": 0.0078, + "grad_norm": 0.31544029712677, + "learning_rate": 9.510075125946414e-05, + "loss": 0.0119, "step": 11100 }, { - "grad_norm": 0.2237762063741684, - "learning_rate": 4.4973688718689803e-05, - "loss": 0.0081, + "grad_norm": 0.24397487938404083, + "learning_rate": 9.508884756671075e-05, + "loss": 0.011, "step": 11110 }, { - "grad_norm": 0.22772948443889618, - "learning_rate": 4.4891440926259406e-05, - "loss": 0.0076, + "grad_norm": 0.27550360560417175, + "learning_rate": 9.507693017715596e-05, + "loss": 0.0124, "step": 11120 }, { - "grad_norm": 0.23078931868076324, - "learning_rate": 4.480920710043443e-05, - "loss": 0.0088, + "grad_norm": 0.2947140038013458, + "learning_rate": 9.506499909441997e-05, + "loss": 0.012, "step": 11130 }, { - "grad_norm": 0.19781197607517242, - "learning_rate": 4.4726987466039044e-05, - "loss": 0.0098, + "grad_norm": 0.33004024624824524, + "learning_rate": 9.505305432212713e-05, + "loss": 0.0141, "step": 11140 }, { - "grad_norm": 0.24051713943481445, - "learning_rate": 4.46447822478586e-05, - "loss": 0.0118, + "grad_norm": 0.3657475709915161, + "learning_rate": 9.504109586390595e-05, + "loss": 0.0151, "step": 11150 }, { - "grad_norm": 0.23845115303993225, - "learning_rate": 4.4562591670638974e-05, - "loss": 0.0066, + "grad_norm": 0.2962097227573395, + "learning_rate": 9.502912372338908e-05, + "loss": 0.0104, "step": 11160 }, { - "grad_norm": 0.3238190710544586, - "learning_rate": 4.4480415959086105e-05, - "loss": 0.0073, + "grad_norm": 0.2767464816570282, + "learning_rate": 9.501713790421335e-05, + "loss": 0.0105, "step": 11170 }, { - "grad_norm": 0.2884441912174225, - "learning_rate": 4.439825533786522e-05, - "loss": 0.0105, + "grad_norm": 0.3294580578804016, + "learning_rate": 9.500513841001974e-05, + "loss": 0.0148, "step": 11180 }, { - "grad_norm": 0.3803757429122925, - "learning_rate": 4.431611003160035e-05, - "loss": 0.0086, + "grad_norm": 0.29677534103393555, + "learning_rate": 9.499312524445336e-05, + "loss": 0.0125, "step": 11190 }, { - "grad_norm": 0.17523354291915894, - "learning_rate": 4.4233980264873636e-05, - "loss": 0.0093, + "grad_norm": 0.2728707194328308, + "learning_rate": 9.498109841116351e-05, + "loss": 0.0125, "step": 11200 }, { - "grad_norm": 0.2824532687664032, - "learning_rate": 4.4151866262224684e-05, - "loss": 0.0108, + "grad_norm": 0.3406853973865509, + "learning_rate": 9.496905791380363e-05, + "loss": 0.0145, "step": 11210 }, { - "grad_norm": 0.3265339732170105, - "learning_rate": 4.406976824815006e-05, - "loss": 0.0079, + "grad_norm": 0.3211401402950287, + "learning_rate": 9.495700375603129e-05, + "loss": 0.0117, "step": 11220 }, { - "grad_norm": 0.24177534878253937, - "learning_rate": 4.3987686447102595e-05, - "loss": 0.0085, + "grad_norm": 0.2678249180316925, + "learning_rate": 9.494493594150822e-05, + "loss": 0.0126, "step": 11230 }, { - "grad_norm": 0.21874769032001495, - "learning_rate": 4.3905621083490804e-05, - "loss": 0.0076, + "grad_norm": 0.24417801201343536, + "learning_rate": 9.493285447390032e-05, + "loss": 0.011, "step": 11240 }, { - "grad_norm": 0.2425031214952469, - "learning_rate": 4.3823572381678286e-05, - "loss": 0.0089, + "grad_norm": 0.24807077646255493, + "learning_rate": 9.492075935687761e-05, + "loss": 0.0124, "step": 11250 }, { - "grad_norm": 0.2610780894756317, - "learning_rate": 4.374154056598301e-05, - "loss": 0.0118, + "grad_norm": 0.37638401985168457, + "learning_rate": 9.490865059411427e-05, + "loss": 0.0149, "step": 11260 }, { - "grad_norm": 0.36238113045692444, - "learning_rate": 4.3659525860676845e-05, - "loss": 0.0131, + "grad_norm": 0.41031622886657715, + "learning_rate": 9.489652818928863e-05, + "loss": 0.018, "step": 11270 }, { - "grad_norm": 0.25657838582992554, - "learning_rate": 4.3577528489984854e-05, - "loss": 0.0081, + "grad_norm": 0.26087990403175354, + "learning_rate": 9.488439214608315e-05, + "loss": 0.0128, "step": 11280 }, { - "grad_norm": 0.27643704414367676, - "learning_rate": 4.349554867808476e-05, - "loss": 0.008, + "grad_norm": 0.3037140965461731, + "learning_rate": 9.487224246818444e-05, + "loss": 0.0106, "step": 11290 }, { - "grad_norm": 0.2851879596710205, - "learning_rate": 4.34135866491062e-05, - "loss": 0.0084, + "grad_norm": 0.2749677896499634, + "learning_rate": 9.486007915928325e-05, + "loss": 0.0112, "step": 11300 }, { - "grad_norm": 0.3266191780567169, - "learning_rate": 4.333164262713022e-05, - "loss": 0.0119, + "grad_norm": 0.3401210904121399, + "learning_rate": 9.484790222307448e-05, + "loss": 0.0142, "step": 11310 }, { - "grad_norm": 0.3193758428096771, - "learning_rate": 4.324971683618868e-05, - "loss": 0.0075, + "grad_norm": 0.3041849732398987, + "learning_rate": 9.483571166325716e-05, + "loss": 0.0101, "step": 11320 }, { - "grad_norm": 0.32918694615364075, - "learning_rate": 4.316780950026354e-05, - "loss": 0.0098, + "grad_norm": 0.35622477531433105, + "learning_rate": 9.482350748353444e-05, + "loss": 0.0131, "step": 11330 }, { - "grad_norm": 0.2517104148864746, - "learning_rate": 4.308592084328637e-05, - "loss": 0.0098, + "grad_norm": 0.3240088224411011, + "learning_rate": 9.481128968761363e-05, + "loss": 0.0143, "step": 11340 }, { - "grad_norm": 0.31300199031829834, - "learning_rate": 4.3004051089137576e-05, - "loss": 0.0095, + "grad_norm": 0.2930913269519806, + "learning_rate": 9.479905827920621e-05, + "loss": 0.0145, "step": 11350 }, { - "grad_norm": 0.2529290020465851, - "learning_rate": 4.292220046164597e-05, - "loss": 0.0073, + "grad_norm": 0.31744933128356934, + "learning_rate": 9.478681326202773e-05, + "loss": 0.0111, "step": 11360 }, { - "grad_norm": 0.23040375113487244, - "learning_rate": 4.2840369184588035e-05, - "loss": 0.0077, + "grad_norm": 0.31036609411239624, + "learning_rate": 9.477455463979791e-05, + "loss": 0.0117, "step": 11370 }, { - "grad_norm": 0.30496788024902344, - "learning_rate": 4.2758557481687345e-05, - "loss": 0.0096, + "grad_norm": 0.3229518532752991, + "learning_rate": 9.476228241624059e-05, + "loss": 0.0142, "step": 11380 }, { - "grad_norm": 0.2186974138021469, - "learning_rate": 4.267676557661403e-05, - "loss": 0.0084, + "grad_norm": 0.2793971300125122, + "learning_rate": 9.474999659508374e-05, + "loss": 0.0131, "step": 11390 }, { - "grad_norm": 0.251447856426239, - "learning_rate": 4.2594993692983955e-05, - "loss": 0.0082, + "grad_norm": 0.2960239350795746, + "learning_rate": 9.47376971800595e-05, + "loss": 0.0125, "step": 11400 }, { - "grad_norm": 0.29543429613113403, - "learning_rate": 4.251324205435837e-05, - "loss": 0.0093, + "grad_norm": 0.3085460364818573, + "learning_rate": 9.472538417490409e-05, + "loss": 0.0139, "step": 11410 }, { - "grad_norm": 0.3692440688610077, - "learning_rate": 4.243151088424312e-05, - "loss": 0.0108, + "grad_norm": 0.3967900276184082, + "learning_rate": 9.471305758335784e-05, + "loss": 0.0148, "step": 11420 }, { - "grad_norm": 0.27703699469566345, - "learning_rate": 4.234980040608813e-05, - "loss": 0.0118, + "grad_norm": 0.3073781430721283, + "learning_rate": 9.47007174091653e-05, + "loss": 0.0147, "step": 11430 }, { - "grad_norm": 0.2971097528934479, - "learning_rate": 4.22681108432867e-05, - "loss": 0.0097, + "grad_norm": 0.40549206733703613, + "learning_rate": 9.468836365607507e-05, + "loss": 0.0143, "step": 11440 }, { - "grad_norm": 0.2624056041240692, - "learning_rate": 4.2186442419174984e-05, - "loss": 0.0085, + "grad_norm": 0.3226219713687897, + "learning_rate": 9.467599632783988e-05, + "loss": 0.0124, "step": 11450 }, { - "grad_norm": 0.2369023561477661, - "learning_rate": 4.210479535703133e-05, - "loss": 0.0081, + "grad_norm": 0.3155043125152588, + "learning_rate": 9.466361542821662e-05, + "loss": 0.0126, "step": 11460 }, { - "grad_norm": 0.18464338779449463, - "learning_rate": 4.202316988007567e-05, - "loss": 0.0095, + "grad_norm": 0.22298204898834229, + "learning_rate": 9.465122096096625e-05, + "loss": 0.0137, "step": 11470 }, { - "grad_norm": 0.21752451360225677, - "learning_rate": 4.194156621146901e-05, - "loss": 0.008, + "grad_norm": 0.27996012568473816, + "learning_rate": 9.463881292985391e-05, + "loss": 0.0126, "step": 11480 }, { - "grad_norm": 0.20453378558158875, - "learning_rate": 4.1859984574312596e-05, - "loss": 0.0069, + "grad_norm": 0.27538031339645386, + "learning_rate": 9.462639133864881e-05, + "loss": 0.0115, "step": 11490 }, { - "grad_norm": 0.21684104204177856, - "learning_rate": 4.177842519164752e-05, - "loss": 0.0117, + "grad_norm": 0.26578009128570557, + "learning_rate": 9.461395619112432e-05, + "loss": 0.015, "step": 11500 }, { - "grad_norm": 0.22559356689453125, - "learning_rate": 4.169688828645404e-05, - "loss": 0.0082, + "grad_norm": 0.3194080591201782, + "learning_rate": 9.460150749105791e-05, + "loss": 0.0126, "step": 11510 }, { - "grad_norm": 0.31731340289115906, - "learning_rate": 4.161537408165092e-05, - "loss": 0.0098, + "grad_norm": 0.3216540813446045, + "learning_rate": 9.458904524223116e-05, + "loss": 0.0136, "step": 11520 }, { - "grad_norm": 0.2910335063934326, - "learning_rate": 4.1533882800094924e-05, - "loss": 0.0092, + "grad_norm": 0.3181364834308624, + "learning_rate": 9.457656944842976e-05, + "loss": 0.0131, "step": 11530 }, { - "grad_norm": 0.2638309597969055, - "learning_rate": 4.145241466458005e-05, - "loss": 0.0097, + "grad_norm": 0.3125889301300049, + "learning_rate": 9.456408011344353e-05, + "loss": 0.0144, "step": 11540 }, { - "grad_norm": 0.2704751193523407, - "learning_rate": 4.13709698978371e-05, - "loss": 0.0066, + "grad_norm": 0.35436922311782837, + "learning_rate": 9.455157724106643e-05, + "loss": 0.0105, "step": 11550 }, { - "grad_norm": 0.24327205121517181, - "learning_rate": 4.1289548722532944e-05, - "loss": 0.0101, + "grad_norm": 0.3434251546859741, + "learning_rate": 9.453906083509647e-05, + "loss": 0.0144, "step": 11560 }, { - "grad_norm": 0.2812352180480957, - "learning_rate": 4.120815136126999e-05, - "loss": 0.0096, + "grad_norm": 0.3820211887359619, + "learning_rate": 9.45265308993358e-05, + "loss": 0.0142, "step": 11570 }, { - "grad_norm": 0.1935969889163971, - "learning_rate": 4.112677803658548e-05, - "loss": 0.0073, + "grad_norm": 0.2705078721046448, + "learning_rate": 9.451398743759071e-05, + "loss": 0.0113, "step": 11580 }, { - "grad_norm": 0.2166028618812561, - "learning_rate": 4.1045428970951e-05, - "loss": 0.0073, + "grad_norm": 0.2262042909860611, + "learning_rate": 9.450143045367156e-05, + "loss": 0.0101, "step": 11590 }, { - "grad_norm": 0.20732693374156952, - "learning_rate": 4.0964104386771785e-05, - "loss": 0.0068, + "grad_norm": 0.21458709239959717, + "learning_rate": 9.448885995139283e-05, + "loss": 0.0094, "step": 11600 }, { - "grad_norm": 0.36443421244621277, - "learning_rate": 4.0882804506386144e-05, - "loss": 0.009, + "grad_norm": 0.31339216232299805, + "learning_rate": 9.44762759345731e-05, + "loss": 0.0118, "step": 11610 }, { - "grad_norm": 0.2749943733215332, - "learning_rate": 4.080152955206485e-05, - "loss": 0.011, + "grad_norm": 0.27186083793640137, + "learning_rate": 9.446367840703509e-05, + "loss": 0.014, "step": 11620 }, { - "grad_norm": 0.34515616297721863, - "learning_rate": 4.0720279746010505e-05, - "loss": 0.0071, + "grad_norm": 0.23883971571922302, + "learning_rate": 9.445106737260556e-05, + "loss": 0.0108, "step": 11630 }, { - "grad_norm": 0.35089951753616333, - "learning_rate": 4.063905531035699e-05, - "loss": 0.0085, + "grad_norm": 0.30917060375213623, + "learning_rate": 9.443844283511543e-05, + "loss": 0.0116, "step": 11640 }, { - "grad_norm": 0.271034300327301, - "learning_rate": 4.055785646716882e-05, - "loss": 0.0083, + "grad_norm": 0.3272685408592224, + "learning_rate": 9.442580479839968e-05, + "loss": 0.0127, "step": 11650 }, { - "grad_norm": 0.3269774913787842, - "learning_rate": 4.047668343844051e-05, - "loss": 0.0079, + "grad_norm": 0.2757956385612488, + "learning_rate": 9.441315326629745e-05, + "loss": 0.0122, "step": 11660 }, { - "grad_norm": 0.2104874551296234, - "learning_rate": 4.039553644609604e-05, - "loss": 0.0066, + "grad_norm": 0.2818768322467804, + "learning_rate": 9.44004882426519e-05, + "loss": 0.01, "step": 11670 }, { - "grad_norm": 0.23121942579746246, - "learning_rate": 4.0314415711988176e-05, - "loss": 0.0067, + "grad_norm": 0.32806840538978577, + "learning_rate": 9.438780973131037e-05, + "loss": 0.0095, "step": 11680 }, { - "grad_norm": 0.3726987838745117, - "learning_rate": 4.023332145789792e-05, - "loss": 0.0108, + "grad_norm": 0.33475878834724426, + "learning_rate": 9.437511773612423e-05, + "loss": 0.0148, "step": 11690 }, { - "grad_norm": 0.28051552176475525, - "learning_rate": 4.015225390553385e-05, - "loss": 0.0099, + "grad_norm": 0.3011960983276367, + "learning_rate": 9.436241226094896e-05, + "loss": 0.0133, "step": 11700 }, { - "grad_norm": 0.30161479115486145, - "learning_rate": 4.007121327653158e-05, - "loss": 0.0085, + "grad_norm": 0.3173636794090271, + "learning_rate": 9.434969330964418e-05, + "loss": 0.0116, "step": 11710 }, { - "grad_norm": 0.2895340919494629, - "learning_rate": 3.9990199792453064e-05, - "loss": 0.0101, + "grad_norm": 0.3196856379508972, + "learning_rate": 9.433696088607356e-05, + "loss": 0.0133, "step": 11720 }, { - "grad_norm": 0.20414268970489502, - "learning_rate": 3.9909213674786103e-05, - "loss": 0.0067, + "grad_norm": 0.29220321774482727, + "learning_rate": 9.432421499410486e-05, + "loss": 0.0092, "step": 11730 }, { - "grad_norm": 0.17530548572540283, - "learning_rate": 3.982825514494363e-05, - "loss": 0.0077, + "grad_norm": 0.32554659247398376, + "learning_rate": 9.431145563760998e-05, + "loss": 0.0108, "step": 11740 }, { - "grad_norm": 0.2952740490436554, - "learning_rate": 3.974732442426319e-05, - "loss": 0.0076, + "grad_norm": 0.31929922103881836, + "learning_rate": 9.429868282046484e-05, + "loss": 0.0114, "step": 11750 }, { - "grad_norm": 0.2763718068599701, - "learning_rate": 3.966642173400629e-05, - "loss": 0.0081, + "grad_norm": 0.2933848798274994, + "learning_rate": 9.428589654654951e-05, + "loss": 0.0125, "step": 11760 }, { - "grad_norm": 0.24036982655525208, - "learning_rate": 3.9585547295357764e-05, - "loss": 0.0078, + "grad_norm": 0.3677593469619751, + "learning_rate": 9.42730968197481e-05, + "loss": 0.0127, "step": 11770 }, { - "grad_norm": 0.1980796903371811, - "learning_rate": 3.950470132942526e-05, - "loss": 0.0107, + "grad_norm": 0.24362000823020935, + "learning_rate": 9.426028364394883e-05, + "loss": 0.0148, "step": 11780 }, { - "grad_norm": 0.22392122447490692, - "learning_rate": 3.942388405723856e-05, - "loss": 0.0072, + "grad_norm": 0.26775676012039185, + "learning_rate": 9.424745702304402e-05, + "loss": 0.0112, "step": 11790 }, { - "grad_norm": 0.3153569996356964, - "learning_rate": 3.9343095699749e-05, - "loss": 0.0118, + "grad_norm": 0.3327542841434479, + "learning_rate": 9.423461696093006e-05, + "loss": 0.0152, "step": 11800 }, { - "grad_norm": 0.2480199635028839, - "learning_rate": 3.9262336477828874e-05, - "loss": 0.008, + "grad_norm": 0.26998060941696167, + "learning_rate": 9.422176346150741e-05, + "loss": 0.0115, "step": 11810 }, { - "grad_norm": 0.33051490783691406, - "learning_rate": 3.9181606612270794e-05, - "loss": 0.0103, + "grad_norm": 0.34707435965538025, + "learning_rate": 9.420889652868063e-05, + "loss": 0.014, "step": 11820 }, { - "grad_norm": 0.2385939210653305, - "learning_rate": 3.910090632378713e-05, - "loss": 0.0092, + "grad_norm": 0.38913461565971375, + "learning_rate": 9.419601616635836e-05, + "loss": 0.012, "step": 11830 }, { - "grad_norm": 0.23588930070400238, - "learning_rate": 3.90202358330094e-05, - "loss": 0.0078, + "grad_norm": 0.3392130434513092, + "learning_rate": 9.418312237845331e-05, + "loss": 0.0114, "step": 11840 }, { - "grad_norm": 0.24540403485298157, - "learning_rate": 3.8939595360487656e-05, - "loss": 0.0098, + "grad_norm": 0.21947836875915527, + "learning_rate": 9.417021516888225e-05, + "loss": 0.0137, "step": 11850 }, { - "grad_norm": 0.2659599184989929, - "learning_rate": 3.885898512668984e-05, - "loss": 0.0099, + "grad_norm": 0.2975698709487915, + "learning_rate": 9.415729454156608e-05, + "loss": 0.014, "step": 11860 }, { - "grad_norm": 0.2260056883096695, - "learning_rate": 3.877840535200127e-05, - "loss": 0.0111, + "grad_norm": 0.2511023283004761, + "learning_rate": 9.414436050042973e-05, + "loss": 0.0148, "step": 11870 }, { - "grad_norm": 0.2496788501739502, - "learning_rate": 3.869785625672397e-05, - "loss": 0.0078, + "grad_norm": 0.41581928730010986, + "learning_rate": 9.413141304940223e-05, + "loss": 0.0119, "step": 11880 }, { - "grad_norm": 0.24494805932044983, - "learning_rate": 3.8617338061076094e-05, - "loss": 0.0092, + "grad_norm": 0.22800247371196747, + "learning_rate": 9.411845219241666e-05, + "loss": 0.0119, "step": 11890 }, { - "grad_norm": 0.27489352226257324, - "learning_rate": 3.853685098519132e-05, - "loss": 0.0143, + "grad_norm": 0.3093186020851135, + "learning_rate": 9.410547793341021e-05, + "loss": 0.0178, "step": 11900 }, { - "grad_norm": 0.1963207870721817, - "learning_rate": 3.845639524911823e-05, - "loss": 0.0076, + "grad_norm": 0.35192447900772095, + "learning_rate": 9.409249027632408e-05, + "loss": 0.0115, "step": 11910 }, { - "grad_norm": 0.26988306641578674, - "learning_rate": 3.837597107281974e-05, - "loss": 0.0084, + "grad_norm": 0.30905112624168396, + "learning_rate": 9.407948922510362e-05, + "loss": 0.0117, "step": 11920 }, { - "grad_norm": 0.2814962565898895, - "learning_rate": 3.829557867617247e-05, - "loss": 0.0087, + "grad_norm": 0.3030983507633209, + "learning_rate": 9.406647478369817e-05, + "loss": 0.0123, "step": 11930 }, { - "grad_norm": 0.2576271891593933, - "learning_rate": 3.821521827896618e-05, - "loss": 0.0074, + "grad_norm": 0.3390377461910248, + "learning_rate": 9.405344695606118e-05, + "loss": 0.0111, "step": 11940 }, { - "grad_norm": 0.35358384251594543, - "learning_rate": 3.81348901009031e-05, - "loss": 0.0116, + "grad_norm": 0.3951464593410492, + "learning_rate": 9.404040574615018e-05, + "loss": 0.0158, "step": 11950 }, { - "grad_norm": 0.23834174871444702, - "learning_rate": 3.805459436159741e-05, - "loss": 0.0092, + "grad_norm": 0.2875381112098694, + "learning_rate": 9.402735115792674e-05, + "loss": 0.0141, "step": 11960 }, { - "grad_norm": 0.198640838265419, - "learning_rate": 3.797433128057461e-05, - "loss": 0.0081, + "grad_norm": 0.26205626130104065, + "learning_rate": 9.401428319535649e-05, + "loss": 0.0116, "step": 11970 }, { - "grad_norm": 0.2608760595321655, - "learning_rate": 3.789410107727089e-05, - "loss": 0.0118, + "grad_norm": 0.28623369336128235, + "learning_rate": 9.400120186240912e-05, + "loss": 0.0157, "step": 11980 }, { - "grad_norm": 0.3514273762702942, - "learning_rate": 3.781390397103257e-05, - "loss": 0.0086, + "grad_norm": 0.3012058436870575, + "learning_rate": 9.398810716305844e-05, + "loss": 0.0127, "step": 11990 }, { - "grad_norm": 0.2969302833080292, - "learning_rate": 3.7733740181115455e-05, - "loss": 0.0094, + "grad_norm": 0.27840206027030945, + "learning_rate": 9.397499910128222e-05, + "loss": 0.0137, "step": 12000 }, { - "grad_norm": 0.2919679880142212, - "learning_rate": 3.7653609926684306e-05, - "loss": 0.0109, + "grad_norm": 0.2904963493347168, + "learning_rate": 9.396187768106237e-05, + "loss": 0.0162, "step": 12010 }, { - "grad_norm": 0.25600647926330566, - "learning_rate": 3.757351342681217e-05, - "loss": 0.0099, + "grad_norm": 0.36414679884910583, + "learning_rate": 9.394874290638482e-05, + "loss": 0.0151, "step": 12020 }, { - "grad_norm": 0.3260495066642761, - "learning_rate": 3.749345090047982e-05, - "loss": 0.01, + "grad_norm": 0.29719632863998413, + "learning_rate": 9.393559478123959e-05, + "loss": 0.0152, "step": 12030 }, { - "grad_norm": 0.28559738397598267, - "learning_rate": 3.741342256657515e-05, - "loss": 0.0103, + "grad_norm": 0.29279398918151855, + "learning_rate": 9.39224333096207e-05, + "loss": 0.0147, "step": 12040 }, { - "grad_norm": 0.2582564055919647, - "learning_rate": 3.7333428643892567e-05, - "loss": 0.0075, + "grad_norm": 0.2706153988838196, + "learning_rate": 9.390925849552629e-05, + "loss": 0.0109, "step": 12050 }, { - "grad_norm": 0.21928203105926514, - "learning_rate": 3.725346935113239e-05, - "loss": 0.0077, + "grad_norm": 0.25963762402534485, + "learning_rate": 9.389607034295849e-05, + "loss": 0.0115, "step": 12060 }, { - "grad_norm": 0.2669205963611603, - "learning_rate": 3.717354490690029e-05, - "loss": 0.0084, + "grad_norm": 0.266626238822937, + "learning_rate": 9.388286885592355e-05, + "loss": 0.0124, "step": 12070 }, { - "grad_norm": 0.21352745592594147, - "learning_rate": 3.709365552970664e-05, - "loss": 0.0086, + "grad_norm": 0.2981499135494232, + "learning_rate": 9.386965403843168e-05, + "loss": 0.0117, "step": 12080 }, { - "grad_norm": 0.25914400815963745, - "learning_rate": 3.7013801437965945e-05, - "loss": 0.0072, + "grad_norm": 0.37225407361984253, + "learning_rate": 9.385642589449726e-05, + "loss": 0.011, "step": 12090 }, { - "grad_norm": 0.27507898211479187, - "learning_rate": 3.693398284999623e-05, - "loss": 0.0085, + "grad_norm": 0.2745845317840576, + "learning_rate": 9.38431844281386e-05, + "loss": 0.0129, "step": 12100 }, { - "grad_norm": 0.2519773840904236, - "learning_rate": 3.6854199984018484e-05, - "loss": 0.0083, + "grad_norm": 0.32825857400894165, + "learning_rate": 9.38299296433781e-05, + "loss": 0.0113, "step": 12110 }, { - "grad_norm": 0.2433474063873291, - "learning_rate": 3.677445305815601e-05, - "loss": 0.0098, + "grad_norm": 0.31450754404067993, + "learning_rate": 9.381666154424226e-05, + "loss": 0.0136, "step": 12120 }, { - "grad_norm": 0.22368106245994568, - "learning_rate": 3.669474229043387e-05, - "loss": 0.0079, + "grad_norm": 0.3225879669189453, + "learning_rate": 9.380338013476157e-05, + "loss": 0.0116, "step": 12130 }, { - "grad_norm": 0.22219599783420563, - "learning_rate": 3.6615067898778235e-05, - "loss": 0.0061, + "grad_norm": 0.2857224643230438, + "learning_rate": 9.379008541897054e-05, + "loss": 0.0098, "step": 12140 }, { - "grad_norm": 0.2246261090040207, - "learning_rate": 3.6535430101015866e-05, - "loss": 0.0076, + "grad_norm": 0.31626152992248535, + "learning_rate": 9.377677740090777e-05, + "loss": 0.0114, "step": 12150 }, { - "grad_norm": 0.24473929405212402, - "learning_rate": 3.645582911487345e-05, - "loss": 0.0088, + "grad_norm": 0.30775129795074463, + "learning_rate": 9.376345608461588e-05, + "loss": 0.0132, "step": 12160 }, { - "grad_norm": 0.28539374470710754, - "learning_rate": 3.637626515797706e-05, - "loss": 0.0088, + "grad_norm": 0.3363003134727478, + "learning_rate": 9.375012147414155e-05, + "loss": 0.0129, "step": 12170 }, { - "grad_norm": 0.25773289799690247, - "learning_rate": 3.629673844785152e-05, - "loss": 0.0067, + "grad_norm": 0.29895225167274475, + "learning_rate": 9.373677357353545e-05, + "loss": 0.0103, "step": 12180 }, { - "grad_norm": 0.21043069660663605, - "learning_rate": 3.621724920191979e-05, - "loss": 0.0074, + "grad_norm": 0.28134334087371826, + "learning_rate": 9.372341238685237e-05, + "loss": 0.0107, "step": 12190 }, { - "grad_norm": 0.2802955210208893, - "learning_rate": 3.6137797637502444e-05, - "loss": 0.0099, + "grad_norm": 0.3287898004055023, + "learning_rate": 9.371003791815102e-05, + "loss": 0.0141, "step": 12200 }, { - "grad_norm": 0.22656969726085663, - "learning_rate": 3.6058383971817035e-05, - "loss": 0.0068, + "grad_norm": 0.28676196932792664, + "learning_rate": 9.369665017149429e-05, + "loss": 0.0105, "step": 12210 }, { - "grad_norm": 0.33790233731269836, - "learning_rate": 3.59790084219775e-05, - "loss": 0.0084, + "grad_norm": 0.335525780916214, + "learning_rate": 9.368324915094895e-05, + "loss": 0.0126, "step": 12220 }, { - "grad_norm": 0.29881054162979126, - "learning_rate": 3.589967120499353e-05, - "loss": 0.0069, + "grad_norm": 0.31452497839927673, + "learning_rate": 9.366983486058591e-05, + "loss": 0.0112, "step": 12230 }, { - "grad_norm": 0.3114658296108246, - "learning_rate": 3.5820372537770075e-05, - "loss": 0.0073, + "grad_norm": 0.4237510561943054, + "learning_rate": 9.365640730448009e-05, + "loss": 0.0124, "step": 12240 }, { - "grad_norm": 0.25627395510673523, - "learning_rate": 3.5741112637106655e-05, - "loss": 0.0089, + "grad_norm": 0.2449006289243698, + "learning_rate": 9.36429664867104e-05, + "loss": 0.0139, "step": 12250 }, { - "grad_norm": 0.20619019865989685, - "learning_rate": 3.5661891719696804e-05, - "loss": 0.008, + "grad_norm": 0.31632760167121887, + "learning_rate": 9.362951241135982e-05, + "loss": 0.0129, "step": 12260 }, { - "grad_norm": 0.24497458338737488, - "learning_rate": 3.5582710002127504e-05, - "loss": 0.011, + "grad_norm": 0.29561933875083923, + "learning_rate": 9.361604508251534e-05, + "loss": 0.0153, "step": 12270 }, { - "grad_norm": 0.23106102645397186, - "learning_rate": 3.550356770087853e-05, - "loss": 0.0086, + "grad_norm": 0.3216542601585388, + "learning_rate": 9.360256450426799e-05, + "loss": 0.0125, "step": 12280 }, { - "grad_norm": 0.2801083028316498, - "learning_rate": 3.5424465032321914e-05, - "loss": 0.0081, + "grad_norm": 0.26470035314559937, + "learning_rate": 9.358907068071279e-05, + "loss": 0.0115, "step": 12290 }, { - "grad_norm": 0.27025705575942993, - "learning_rate": 3.5345402212721335e-05, - "loss": 0.0082, + "grad_norm": 0.2767382264137268, + "learning_rate": 9.357556361594882e-05, + "loss": 0.0123, "step": 12300 }, { - "grad_norm": 0.2505914270877838, - "learning_rate": 3.526637945823152e-05, - "loss": 0.009, + "grad_norm": 0.3290117084980011, + "learning_rate": 9.356204331407917e-05, + "loss": 0.0124, "step": 12310 }, { - "grad_norm": 0.2746097445487976, - "learning_rate": 3.518739698489767e-05, - "loss": 0.0064, + "grad_norm": 0.3501572906970978, + "learning_rate": 9.354850977921094e-05, + "loss": 0.0102, "step": 12320 }, { - "grad_norm": 0.2840472459793091, - "learning_rate": 3.510845500865485e-05, - "loss": 0.0073, + "grad_norm": 0.2946989834308624, + "learning_rate": 9.353496301545529e-05, + "loss": 0.012, "step": 12330 }, { - "grad_norm": 0.27252042293548584, - "learning_rate": 3.502955374532739e-05, - "loss": 0.011, + "grad_norm": 0.24019166827201843, + "learning_rate": 9.352140302692733e-05, + "loss": 0.015, "step": 12340 }, { - "grad_norm": 0.21478882431983948, - "learning_rate": 3.495069341062836e-05, - "loss": 0.0082, + "grad_norm": 0.23021461069583893, + "learning_rate": 9.350782981774627e-05, + "loss": 0.0116, "step": 12350 }, { - "grad_norm": 0.30031028389930725, - "learning_rate": 3.4871874220158896e-05, - "loss": 0.0115, + "grad_norm": 0.344180703163147, + "learning_rate": 9.349424339203526e-05, + "loss": 0.0155, "step": 12360 }, { - "grad_norm": 0.29065871238708496, - "learning_rate": 3.479309638940762e-05, - "loss": 0.009, + "grad_norm": 0.30660107731819153, + "learning_rate": 9.34806437539215e-05, + "loss": 0.013, "step": 12370 }, { - "grad_norm": 0.3161035478115082, - "learning_rate": 3.4714360133750146e-05, - "loss": 0.0085, + "grad_norm": 0.3166130483150482, + "learning_rate": 9.346703090753622e-05, + "loss": 0.0129, "step": 12380 }, { - "grad_norm": 0.29907113313674927, - "learning_rate": 3.463566566844839e-05, - "loss": 0.0078, + "grad_norm": 0.2784366309642792, + "learning_rate": 9.345340485701461e-05, + "loss": 0.0112, "step": 12390 }, { - "grad_norm": 0.2610228359699249, - "learning_rate": 3.4557013208650016e-05, - "loss": 0.0093, + "grad_norm": 0.3697854280471802, + "learning_rate": 9.343976560649595e-05, + "loss": 0.0146, "step": 12400 }, { - "grad_norm": 0.19939985871315002, - "learning_rate": 3.4478402969387857e-05, - "loss": 0.0075, + "grad_norm": 0.28481167554855347, + "learning_rate": 9.342611316012344e-05, + "loss": 0.0116, "step": 12410 }, { - "grad_norm": 0.22081556916236877, - "learning_rate": 3.4399835165579266e-05, - "loss": 0.0075, + "grad_norm": 0.3195859491825104, + "learning_rate": 9.341244752204437e-05, + "loss": 0.0108, "step": 12420 }, { - "grad_norm": 0.20169289410114288, - "learning_rate": 3.4321310012025645e-05, - "loss": 0.0063, + "grad_norm": 0.2889546751976013, + "learning_rate": 9.339876869640995e-05, + "loss": 0.0098, "step": 12430 }, { - "grad_norm": 0.2489660531282425, - "learning_rate": 3.424282772341176e-05, - "loss": 0.0074, + "grad_norm": 0.3227902054786682, + "learning_rate": 9.33850766873755e-05, + "loss": 0.0116, "step": 12440 }, { - "grad_norm": 0.28628960251808167, - "learning_rate": 3.416438851430519e-05, - "loss": 0.013, + "grad_norm": 0.3246162235736847, + "learning_rate": 9.337137149910028e-05, + "loss": 0.0179, "step": 12450 }, { - "grad_norm": 0.29382604360580444, - "learning_rate": 3.408599259915577e-05, - "loss": 0.0077, + "grad_norm": 0.350818395614624, + "learning_rate": 9.335765313574753e-05, + "loss": 0.0116, "step": 12460 }, { - "grad_norm": 0.3110322952270508, - "learning_rate": 3.400764019229487e-05, - "loss": 0.0108, + "grad_norm": 0.29043349623680115, + "learning_rate": 9.334392160148457e-05, + "loss": 0.0154, "step": 12470 }, { - "grad_norm": 0.23523154854774475, - "learning_rate": 3.3929331507935035e-05, - "loss": 0.0069, + "grad_norm": 0.31587493419647217, + "learning_rate": 9.333017690048264e-05, + "loss": 0.0109, "step": 12480 }, { - "grad_norm": 0.273687481880188, - "learning_rate": 3.3851066760169196e-05, - "loss": 0.0075, + "grad_norm": 0.29283198714256287, + "learning_rate": 9.331641903691706e-05, + "loss": 0.0107, "step": 12490 }, { - "grad_norm": 0.19932150840759277, - "learning_rate": 3.377284616297021e-05, - "loss": 0.007, + "grad_norm": 0.24044612050056458, + "learning_rate": 9.330264801496707e-05, + "loss": 0.0102, "step": 12500 }, { - "grad_norm": 0.24185536801815033, - "learning_rate": 3.3694669930190166e-05, - "loss": 0.0068, + "grad_norm": 0.2525232434272766, + "learning_rate": 9.328886383881594e-05, + "loss": 0.0103, "step": 12510 }, { - "grad_norm": 0.2853364646434784, - "learning_rate": 3.36165382755599e-05, - "loss": 0.0075, + "grad_norm": 0.28715670108795166, + "learning_rate": 9.327506651265095e-05, + "loss": 0.0116, "step": 12520 }, { - "grad_norm": 0.23251241445541382, - "learning_rate": 3.35384514126884e-05, - "loss": 0.009, + "grad_norm": 0.314951092004776, + "learning_rate": 9.326125604066338e-05, + "loss": 0.0137, "step": 12530 }, { - "grad_norm": 0.2282250076532364, - "learning_rate": 3.3460409555062154e-05, - "loss": 0.007, + "grad_norm": 0.3211721181869507, + "learning_rate": 9.324743242704847e-05, + "loss": 0.0109, "step": 12540 }, { - "grad_norm": 0.2507518231868744, - "learning_rate": 3.3382412916044645e-05, - "loss": 0.0073, + "grad_norm": 0.2860586941242218, + "learning_rate": 9.323359567600546e-05, + "loss": 0.0112, "step": 12550 }, { - "grad_norm": 0.17678532004356384, - "learning_rate": 3.330446170887566e-05, - "loss": 0.006, + "grad_norm": 0.24017620086669922, + "learning_rate": 9.321974579173761e-05, + "loss": 0.0099, "step": 12560 }, { - "grad_norm": 0.22255797684192657, - "learning_rate": 3.3226556146670834e-05, - "loss": 0.0069, + "grad_norm": 0.3212856948375702, + "learning_rate": 9.320588277845213e-05, + "loss": 0.0108, "step": 12570 }, { - "grad_norm": 0.22277423739433289, - "learning_rate": 3.314869644242102e-05, - "loss": 0.0067, + "grad_norm": 0.30985161662101746, + "learning_rate": 9.319200664036026e-05, + "loss": 0.01, "step": 12580 }, { - "grad_norm": 0.24665354192256927, - "learning_rate": 3.3070882808991674e-05, - "loss": 0.0077, + "grad_norm": 0.3210581839084625, + "learning_rate": 9.31781173816772e-05, + "loss": 0.0101, "step": 12590 }, { - "grad_norm": 0.2221645563840866, - "learning_rate": 3.2993115459122305e-05, - "loss": 0.008, + "grad_norm": 0.2976618707180023, + "learning_rate": 9.316421500662212e-05, + "loss": 0.0114, "step": 12600 }, { - "grad_norm": 0.25327399373054504, - "learning_rate": 3.2915394605425835e-05, - "loss": 0.007, + "grad_norm": 0.26481467485427856, + "learning_rate": 9.31502995194182e-05, + "loss": 0.0108, "step": 12610 }, { - "grad_norm": 0.2379884123802185, - "learning_rate": 3.283772046038816e-05, - "loss": 0.0084, + "grad_norm": 0.2732757031917572, + "learning_rate": 9.31363709242926e-05, + "loss": 0.0123, "step": 12620 }, { - "grad_norm": 0.2786004841327667, - "learning_rate": 3.276009323636739e-05, - "loss": 0.007, + "grad_norm": 0.44056597352027893, + "learning_rate": 9.312242922547647e-05, + "loss": 0.0108, "step": 12630 }, { - "grad_norm": 0.23235200345516205, - "learning_rate": 3.268251314559344e-05, - "loss": 0.0071, + "grad_norm": 0.33957093954086304, + "learning_rate": 9.310847442720492e-05, + "loss": 0.0125, "step": 12640 }, { - "grad_norm": 0.2734534740447998, - "learning_rate": 3.2604980400167254e-05, - "loss": 0.0065, + "grad_norm": 0.30947208404541016, + "learning_rate": 9.309450653371706e-05, + "loss": 0.0112, "step": 12650 }, { - "grad_norm": 0.20516322553157806, - "learning_rate": 3.252749521206042e-05, - "loss": 0.0057, + "grad_norm": 0.3137624263763428, + "learning_rate": 9.308052554925595e-05, + "loss": 0.0091, "step": 12660 }, { - "grad_norm": 0.2804936170578003, - "learning_rate": 3.2450057793114494e-05, - "loss": 0.0058, + "grad_norm": 0.2852679491043091, + "learning_rate": 9.306653147806867e-05, + "loss": 0.0094, "step": 12670 }, { - "grad_norm": 0.23371683061122894, - "learning_rate": 3.2372668355040435e-05, - "loss": 0.0072, + "grad_norm": 0.3253926634788513, + "learning_rate": 9.305252432440622e-05, + "loss": 0.0117, "step": 12680 }, { - "grad_norm": 0.24273928999900818, - "learning_rate": 3.2295327109418005e-05, - "loss": 0.0076, + "grad_norm": 0.30271226167678833, + "learning_rate": 9.303850409252361e-05, + "loss": 0.0121, "step": 12690 }, { - "grad_norm": 0.2837400734424591, - "learning_rate": 3.221803426769518e-05, - "loss": 0.0059, + "grad_norm": 0.277969628572464, + "learning_rate": 9.302447078667985e-05, + "loss": 0.0103, "step": 12700 }, { - "grad_norm": 0.1821887344121933, - "learning_rate": 3.214079004118768e-05, - "loss": 0.0062, + "grad_norm": 0.2837568521499634, + "learning_rate": 9.301042441113783e-05, + "loss": 0.0096, "step": 12710 }, { - "grad_norm": 0.27825450897216797, - "learning_rate": 3.2063594641078234e-05, - "loss": 0.0075, + "grad_norm": 0.2578808665275574, + "learning_rate": 9.299636497016451e-05, + "loss": 0.0113, "step": 12720 }, { - "grad_norm": 0.30339550971984863, - "learning_rate": 3.198644827841616e-05, - "loss": 0.0066, + "grad_norm": 0.25955691933631897, + "learning_rate": 9.298229246803076e-05, + "loss": 0.01, "step": 12730 }, { - "grad_norm": 0.23971441388130188, - "learning_rate": 3.1909351164116654e-05, - "loss": 0.0066, + "grad_norm": 0.26568764448165894, + "learning_rate": 9.296820690901144e-05, + "loss": 0.0107, "step": 12740 }, { - "grad_norm": 0.18594853579998016, - "learning_rate": 3.183230350896026e-05, - "loss": 0.0085, + "grad_norm": 0.2533382475376129, + "learning_rate": 9.295410829738539e-05, + "loss": 0.0112, "step": 12750 }, { - "grad_norm": 0.2139957696199417, - "learning_rate": 3.1755305523592337e-05, - "loss": 0.0084, + "grad_norm": 0.24403776228427887, + "learning_rate": 9.293999663743535e-05, + "loss": 0.0108, "step": 12760 }, { - "grad_norm": 0.282361775636673, - "learning_rate": 3.167835741852245e-05, - "loss": 0.0078, + "grad_norm": 0.26153847575187683, + "learning_rate": 9.292587193344813e-05, + "loss": 0.0113, "step": 12770 }, { - "grad_norm": 0.23739394545555115, - "learning_rate": 3.160145940412378e-05, - "loss": 0.0095, + "grad_norm": 0.24276554584503174, + "learning_rate": 9.291173418971437e-05, + "loss": 0.0131, "step": 12780 }, { - "grad_norm": 0.22370891273021698, - "learning_rate": 3.1524611690632545e-05, - "loss": 0.0059, + "grad_norm": 0.23707716166973114, + "learning_rate": 9.28975834105288e-05, + "loss": 0.0093, "step": 12790 }, { - "grad_norm": 0.2226971536874771, - "learning_rate": 3.144781448814746e-05, - "loss": 0.0074, + "grad_norm": 0.2535558044910431, + "learning_rate": 9.288341960019004e-05, + "loss": 0.0112, "step": 12800 }, { - "grad_norm": 0.27098506689071655, - "learning_rate": 3.1371068006629145e-05, - "loss": 0.006, + "grad_norm": 0.3209000527858734, + "learning_rate": 9.286924276300067e-05, + "loss": 0.0094, "step": 12810 }, { - "grad_norm": 0.23244546353816986, - "learning_rate": 3.129437245589956e-05, - "loss": 0.0071, + "grad_norm": 0.30806979537010193, + "learning_rate": 9.285505290326726e-05, + "loss": 0.0118, "step": 12820 }, { - "grad_norm": 0.23745259642601013, - "learning_rate": 3.121772804564143e-05, - "loss": 0.0066, + "grad_norm": 0.2941007614135742, + "learning_rate": 9.284085002530027e-05, + "loss": 0.0118, "step": 12830 }, { - "grad_norm": 0.25362375378608704, - "learning_rate": 3.11411349853976e-05, - "loss": 0.0064, + "grad_norm": 0.2706282138824463, + "learning_rate": 9.282663413341422e-05, + "loss": 0.0109, "step": 12840 }, { - "grad_norm": 0.2746460437774658, - "learning_rate": 3.10645934845706e-05, - "loss": 0.0064, + "grad_norm": 0.2256108820438385, + "learning_rate": 9.281240523192747e-05, + "loss": 0.0104, "step": 12850 }, { - "grad_norm": 0.2910458445549011, - "learning_rate": 3.098810375242196e-05, - "loss": 0.008, + "grad_norm": 0.3792514204978943, + "learning_rate": 9.279816332516242e-05, + "loss": 0.012, "step": 12860 }, { - "grad_norm": 0.1930484175682068, - "learning_rate": 3.0911665998071704e-05, - "loss": 0.0058, + "grad_norm": 0.23446276783943176, + "learning_rate": 9.278390841744536e-05, + "loss": 0.0101, "step": 12870 }, { - "grad_norm": 0.26174429059028625, - "learning_rate": 3.083528043049774e-05, - "loss": 0.0061, + "grad_norm": 0.25093555450439453, + "learning_rate": 9.276964051310658e-05, + "loss": 0.0097, "step": 12880 }, { - "grad_norm": 0.237471342086792, - "learning_rate": 3.0758947258535255e-05, - "loss": 0.0082, + "grad_norm": 0.23392321169376373, + "learning_rate": 9.275535961648027e-05, + "loss": 0.011, "step": 12890 }, { - "grad_norm": 0.2530456483364105, - "learning_rate": 3.068266669087625e-05, - "loss": 0.0078, + "grad_norm": 0.2726021707057953, + "learning_rate": 9.274106573190459e-05, + "loss": 0.0112, "step": 12900 }, { - "grad_norm": 0.21137702465057373, - "learning_rate": 3.060643893606887e-05, - "loss": 0.006, + "grad_norm": 0.3146917521953583, + "learning_rate": 9.272675886372168e-05, + "loss": 0.009, "step": 12910 }, { - "grad_norm": 0.1681884080171585, - "learning_rate": 3.053026420251693e-05, - "loss": 0.0068, + "grad_norm": 0.29945164918899536, + "learning_rate": 9.271243901627754e-05, + "loss": 0.0101, "step": 12920 }, { - "grad_norm": 0.25592848658561707, - "learning_rate": 3.0454142698479183e-05, - "loss": 0.0069, + "grad_norm": 0.3533570468425751, + "learning_rate": 9.269810619392219e-05, + "loss": 0.0115, "step": 12930 }, { - "grad_norm": 0.2088608294725418, - "learning_rate": 3.0378074632068954e-05, - "loss": 0.0087, + "grad_norm": 0.25372767448425293, + "learning_rate": 9.268376040100955e-05, + "loss": 0.0119, "step": 12940 }, { - "grad_norm": 0.22707495093345642, - "learning_rate": 3.0302060211253408e-05, - "loss": 0.0066, + "grad_norm": 0.21627280116081238, + "learning_rate": 9.266940164189752e-05, + "loss": 0.01, "step": 12950 }, { - "grad_norm": 0.2611274719238281, - "learning_rate": 3.0226099643853073e-05, - "loss": 0.0075, + "grad_norm": 0.2865859866142273, + "learning_rate": 9.265502992094787e-05, + "loss": 0.0113, "step": 12960 }, { - "grad_norm": 0.20526066422462463, - "learning_rate": 3.0150193137541283e-05, - "loss": 0.006, + "grad_norm": 0.26772934198379517, + "learning_rate": 9.264064524252638e-05, + "loss": 0.0099, "step": 12970 }, { - "grad_norm": 0.26528921723365784, - "learning_rate": 3.0074340899843467e-05, - "loss": 0.0074, + "grad_norm": 0.27988114953041077, + "learning_rate": 9.262624761100271e-05, + "loss": 0.0111, "step": 12980 }, { - "grad_norm": 0.294217973947525, - "learning_rate": 2.999854313813677e-05, - "loss": 0.0112, + "grad_norm": 0.3677988350391388, + "learning_rate": 9.261183703075051e-05, + "loss": 0.0157, "step": 12990 }, { - "grad_norm": 0.2798357307910919, - "learning_rate": 2.9922800059649382e-05, - "loss": 0.0074, + "grad_norm": 0.2572009861469269, + "learning_rate": 9.259741350614733e-05, + "loss": 0.0104, "step": 13000 }, { - "grad_norm": 0.25348061323165894, - "learning_rate": 2.9847111871459976e-05, - "loss": 0.008, + "grad_norm": 0.31377890706062317, + "learning_rate": 9.258297704157464e-05, + "loss": 0.0123, "step": 13010 }, { - "grad_norm": 0.25376802682876587, - "learning_rate": 2.977147878049721e-05, - "loss": 0.0072, + "grad_norm": 0.2724866569042206, + "learning_rate": 9.256852764141786e-05, + "loss": 0.0112, "step": 13020 }, { - "grad_norm": 0.23868699371814728, - "learning_rate": 2.9695900993539006e-05, - "loss": 0.0069, + "grad_norm": 0.3324366509914398, + "learning_rate": 9.255406531006634e-05, + "loss": 0.0115, "step": 13030 }, { - "grad_norm": 0.24584409594535828, - "learning_rate": 2.9620378717212183e-05, - "loss": 0.0113, + "grad_norm": 0.3288393020629883, + "learning_rate": 9.253959005191335e-05, + "loss": 0.0158, "step": 13040 }, { - "grad_norm": 0.2317948341369629, - "learning_rate": 2.9544912157991745e-05, - "loss": 0.0087, + "grad_norm": 0.35578760504722595, + "learning_rate": 9.25251018713561e-05, + "loss": 0.0132, "step": 13050 }, { - "grad_norm": 0.3443603217601776, - "learning_rate": 2.9469501522200405e-05, - "loss": 0.006, + "grad_norm": 0.3423280417919159, + "learning_rate": 9.251060077279571e-05, + "loss": 0.0108, "step": 13060 }, { - "grad_norm": 0.1847532093524933, - "learning_rate": 2.9394147016007946e-05, - "loss": 0.0064, + "grad_norm": 0.3130364418029785, + "learning_rate": 9.249608676063724e-05, + "loss": 0.0103, "step": 13070 }, { - "grad_norm": 0.2874523401260376, - "learning_rate": 2.9318848845430702e-05, - "loss": 0.0058, + "grad_norm": 0.3021275997161865, + "learning_rate": 9.248155983928964e-05, + "loss": 0.0101, "step": 13080 }, { - "grad_norm": 0.20815210044384003, - "learning_rate": 2.9243607216331013e-05, - "loss": 0.0061, + "grad_norm": 0.2814575731754303, + "learning_rate": 9.246702001316583e-05, + "loss": 0.011, "step": 13090 }, { - "grad_norm": 0.1735462099313736, - "learning_rate": 2.916842233441661e-05, - "loss": 0.0049, + "grad_norm": 0.25663796067237854, + "learning_rate": 9.245246728668262e-05, + "loss": 0.0086, "step": 13100 }, { - "grad_norm": 0.2947513163089752, - "learning_rate": 2.90932944052401e-05, - "loss": 0.0078, + "grad_norm": 0.2873521149158478, + "learning_rate": 9.243790166426073e-05, + "loss": 0.0126, "step": 13110 }, { - "grad_norm": 0.185662180185318, - "learning_rate": 2.9018223634198354e-05, - "loss": 0.006, + "grad_norm": 0.21159593760967255, + "learning_rate": 9.242332315032484e-05, + "loss": 0.0092, "step": 13120 }, { - "grad_norm": 0.159962996840477, - "learning_rate": 2.8943210226532025e-05, - "loss": 0.0085, + "grad_norm": 0.3208015263080597, + "learning_rate": 9.240873174930349e-05, + "loss": 0.0132, "step": 13130 }, { - "grad_norm": 0.23721252381801605, - "learning_rate": 2.8868254387324857e-05, - "loss": 0.0082, + "grad_norm": 0.32282617688179016, + "learning_rate": 9.239412746562917e-05, + "loss": 0.0122, "step": 13140 }, { - "grad_norm": 0.2839058041572571, - "learning_rate": 2.8793356321503306e-05, - "loss": 0.0084, + "grad_norm": 0.35617560148239136, + "learning_rate": 9.237951030373828e-05, + "loss": 0.0131, "step": 13150 }, { - "grad_norm": 0.3259228765964508, - "learning_rate": 2.87185162338358e-05, - "loss": 0.0067, + "grad_norm": 0.26838645339012146, + "learning_rate": 9.236488026807113e-05, + "loss": 0.011, "step": 13160 }, { - "grad_norm": 0.2640034258365631, - "learning_rate": 2.8643734328932253e-05, - "loss": 0.0087, + "grad_norm": 0.30789637565612793, + "learning_rate": 9.235023736307193e-05, + "loss": 0.0134, "step": 13170 }, { - "grad_norm": 0.2673685848712921, - "learning_rate": 2.856901081124359e-05, - "loss": 0.0068, + "grad_norm": 0.2878201901912689, + "learning_rate": 9.233558159318881e-05, + "loss": 0.0114, "step": 13180 }, { - "grad_norm": 0.2533598840236664, - "learning_rate": 2.8494345885061002e-05, - "loss": 0.0084, + "grad_norm": 0.27371957898139954, + "learning_rate": 9.232091296287382e-05, + "loss": 0.0127, "step": 13190 }, { - "grad_norm": 0.27771884202957153, - "learning_rate": 2.8419739754515616e-05, - "loss": 0.008, + "grad_norm": 0.26467692852020264, + "learning_rate": 9.230623147658288e-05, + "loss": 0.0119, "step": 13200 }, { - "grad_norm": 0.18158374726772308, - "learning_rate": 2.8345192623577666e-05, - "loss": 0.0066, + "grad_norm": 0.2501579821109772, + "learning_rate": 9.229153713877586e-05, + "loss": 0.0106, "step": 13210 }, { - "grad_norm": 0.20402629673480988, - "learning_rate": 2.8270704696056193e-05, - "loss": 0.0085, + "grad_norm": 0.25455471873283386, + "learning_rate": 9.227682995391649e-05, + "loss": 0.014, "step": 13220 }, { - "grad_norm": 0.22918453812599182, - "learning_rate": 2.8196276175598367e-05, - "loss": 0.0089, + "grad_norm": 0.3140377998352051, + "learning_rate": 9.226210992647243e-05, + "loss": 0.0139, "step": 13230 }, { - "grad_norm": 0.26338842511177063, - "learning_rate": 2.8121907265688884e-05, - "loss": 0.0097, + "grad_norm": 0.29039689898490906, + "learning_rate": 9.224737706091525e-05, + "loss": 0.0145, "step": 13240 }, { - "grad_norm": 0.2552448511123657, - "learning_rate": 2.804759816964957e-05, - "loss": 0.007, + "grad_norm": 0.2891133725643158, + "learning_rate": 9.223263136172039e-05, + "loss": 0.0112, "step": 13250 }, { - "grad_norm": 0.2544209957122803, - "learning_rate": 2.797334909063857e-05, - "loss": 0.0068, + "grad_norm": 0.3550475537776947, + "learning_rate": 9.22178728333672e-05, + "loss": 0.0106, "step": 13260 }, { - "grad_norm": 0.20611265301704407, - "learning_rate": 2.7899160231650056e-05, - "loss": 0.0057, + "grad_norm": 0.24973979592323303, + "learning_rate": 9.220310148033897e-05, + "loss": 0.0097, "step": 13270 }, { - "grad_norm": 0.2455790638923645, - "learning_rate": 2.7825031795513585e-05, - "loss": 0.0064, + "grad_norm": 0.34137365221977234, + "learning_rate": 9.21883173071228e-05, + "loss": 0.011, "step": 13280 }, { - "grad_norm": 0.16692101955413818, - "learning_rate": 2.775096398489341e-05, - "loss": 0.0054, + "grad_norm": 0.3092743456363678, + "learning_rate": 9.217352031820976e-05, + "loss": 0.0089, "step": 13290 }, { - "grad_norm": 0.19417795538902283, - "learning_rate": 2.7676957002288163e-05, - "loss": 0.0069, + "grad_norm": 0.3239731192588806, + "learning_rate": 9.215871051809477e-05, + "loss": 0.0118, "step": 13300 }, { - "grad_norm": 0.246934175491333, - "learning_rate": 2.760301105003003e-05, - "loss": 0.0067, + "grad_norm": 0.30375275015830994, + "learning_rate": 9.214388791127666e-05, + "loss": 0.0107, "step": 13310 }, { - "grad_norm": 0.18580402433872223, - "learning_rate": 2.752912633028446e-05, - "loss": 0.0065, + "grad_norm": 0.22032277286052704, + "learning_rate": 9.212905250225814e-05, + "loss": 0.0113, "step": 13320 }, { - "grad_norm": 0.18205900490283966, - "learning_rate": 2.7455303045049474e-05, - "loss": 0.0065, + "grad_norm": 0.25254762172698975, + "learning_rate": 9.211420429554583e-05, + "loss": 0.0116, "step": 13330 }, { - "grad_norm": 0.24550791084766388, - "learning_rate": 2.7381541396155098e-05, - "loss": 0.0069, + "grad_norm": 0.23423565924167633, + "learning_rate": 9.209934329565022e-05, + "loss": 0.0101, "step": 13340 }, { - "grad_norm": 0.2719643712043762, - "learning_rate": 2.730784158526286e-05, - "loss": 0.0059, + "grad_norm": 0.23686887323856354, + "learning_rate": 9.208446950708568e-05, + "loss": 0.0099, "step": 13350 }, { - "grad_norm": 0.19565212726593018, - "learning_rate": 2.723420381386521e-05, - "loss": 0.0072, + "grad_norm": 0.21654759347438812, + "learning_rate": 9.20695829343705e-05, + "loss": 0.0101, "step": 13360 }, { - "grad_norm": 0.2644125521183014, - "learning_rate": 2.7160628283285018e-05, - "loss": 0.0074, + "grad_norm": 0.3012162446975708, + "learning_rate": 9.205468358202678e-05, + "loss": 0.0107, "step": 13370 }, { - "grad_norm": 0.20331737399101257, - "learning_rate": 2.7087115194675007e-05, - "loss": 0.0059, + "grad_norm": 0.3501008450984955, + "learning_rate": 9.203977145458059e-05, + "loss": 0.0099, "step": 13380 }, { - "grad_norm": 0.18478238582611084, - "learning_rate": 2.701366474901712e-05, - "loss": 0.005, + "grad_norm": 0.24046842753887177, + "learning_rate": 9.202484655656182e-05, + "loss": 0.0085, "step": 13390 }, { - "grad_norm": 0.2070227712392807, - "learning_rate": 2.6940277147122085e-05, - "loss": 0.0063, + "grad_norm": 0.28026625514030457, + "learning_rate": 9.200990889250427e-05, + "loss": 0.01, "step": 13400 }, { - "grad_norm": 0.24267740547657013, - "learning_rate": 2.686695258962878e-05, - "loss": 0.0061, + "grad_norm": 0.30334681272506714, + "learning_rate": 9.19949584669456e-05, + "loss": 0.0101, "step": 13410 }, { - "grad_norm": 0.1909894049167633, - "learning_rate": 2.679369127700375e-05, - "loss": 0.0056, + "grad_norm": 0.27469584345817566, + "learning_rate": 9.197999528442738e-05, + "loss": 0.0093, "step": 13420 }, { - "grad_norm": 0.19610315561294556, - "learning_rate": 2.672049340954067e-05, - "loss": 0.007, + "grad_norm": 0.2630292475223541, + "learning_rate": 9.196501934949499e-05, + "loss": 0.012, "step": 13430 }, { - "grad_norm": 0.2082044929265976, - "learning_rate": 2.6647359187359676e-05, - "loss": 0.0086, + "grad_norm": 0.3277732729911804, + "learning_rate": 9.195003066669776e-05, + "loss": 0.0137, "step": 13440 }, { - "grad_norm": 0.22661364078521729, - "learning_rate": 2.6574288810406946e-05, - "loss": 0.0072, + "grad_norm": 0.2155776172876358, + "learning_rate": 9.193502924058884e-05, + "loss": 0.0123, "step": 13450 }, { - "grad_norm": 0.2202974408864975, - "learning_rate": 2.6501282478454083e-05, - "loss": 0.0077, + "grad_norm": 0.33467185497283936, + "learning_rate": 9.192001507572526e-05, + "loss": 0.0126, "step": 13460 }, { - "grad_norm": 0.2299632430076599, - "learning_rate": 2.6428340391097618e-05, - "loss": 0.0109, + "grad_norm": 0.33216392993927, + "learning_rate": 9.190498817666793e-05, + "loss": 0.0158, "step": 13470 }, { - "grad_norm": 0.3141759932041168, - "learning_rate": 2.6355462747758485e-05, - "loss": 0.0065, + "grad_norm": 0.30800798535346985, + "learning_rate": 9.188994854798163e-05, + "loss": 0.0108, "step": 13480 }, { - "grad_norm": 0.23412247002124786, - "learning_rate": 2.6282649747681304e-05, - "loss": 0.0059, + "grad_norm": 0.2823765277862549, + "learning_rate": 9.187489619423499e-05, + "loss": 0.0097, "step": 13490 }, { - "grad_norm": 0.2105311006307602, - "learning_rate": 2.620990158993406e-05, - "loss": 0.0054, + "grad_norm": 0.2572925388813019, + "learning_rate": 9.185983112000056e-05, + "loss": 0.0095, "step": 13500 }, { - "grad_norm": 0.21451091766357422, - "learning_rate": 2.6137218473407477e-05, - "loss": 0.007, + "grad_norm": 0.2934921085834503, + "learning_rate": 9.184475332985464e-05, + "loss": 0.0119, "step": 13510 }, { - "grad_norm": 0.18160901963710785, - "learning_rate": 2.606460059681436e-05, - "loss": 0.0051, + "grad_norm": 0.2932271957397461, + "learning_rate": 9.182966282837754e-05, + "loss": 0.0098, "step": 13520 }, { - "grad_norm": 0.2187230885028839, - "learning_rate": 2.599204815868928e-05, - "loss": 0.0053, + "grad_norm": 0.3143003582954407, + "learning_rate": 9.18145596201533e-05, + "loss": 0.0099, "step": 13530 }, { - "grad_norm": 0.20273251831531525, - "learning_rate": 2.5919561357387756e-05, - "loss": 0.0054, + "grad_norm": 0.2762475907802582, + "learning_rate": 9.179944370976991e-05, + "loss": 0.009, "step": 13540 }, { - "grad_norm": 0.23314417898654938, - "learning_rate": 2.5847140391085972e-05, - "loss": 0.006, + "grad_norm": 0.3207020163536072, + "learning_rate": 9.178431510181918e-05, + "loss": 0.0106, "step": 13550 }, { - "grad_norm": 0.2788357436656952, - "learning_rate": 2.5774785457780103e-05, - "loss": 0.0093, + "grad_norm": 0.3313432037830353, + "learning_rate": 9.176917380089675e-05, + "loss": 0.014, "step": 13560 }, { - "grad_norm": 0.20386862754821777, - "learning_rate": 2.5702496755285753e-05, - "loss": 0.0064, + "grad_norm": 0.33052942156791687, + "learning_rate": 9.175401981160219e-05, + "loss": 0.011, "step": 13570 }, { - "grad_norm": 0.21573026478290558, - "learning_rate": 2.5630274481237483e-05, - "loss": 0.0057, + "grad_norm": 0.20004190504550934, + "learning_rate": 9.173885313853885e-05, + "loss": 0.0095, "step": 13580 }, { - "grad_norm": 0.1928456872701645, - "learning_rate": 2.5558118833088197e-05, - "loss": 0.0065, + "grad_norm": 0.30103132128715515, + "learning_rate": 9.172367378631398e-05, + "loss": 0.0102, "step": 13590 }, { - "grad_norm": 0.23650315403938293, - "learning_rate": 2.548603000810872e-05, - "loss": 0.0049, + "grad_norm": 0.27073389291763306, + "learning_rate": 9.170848175953866e-05, + "loss": 0.0085, "step": 13600 }, { - "grad_norm": 0.19839148223400116, - "learning_rate": 2.5414008203387152e-05, - "loss": 0.0068, + "grad_norm": 0.23978844285011292, + "learning_rate": 9.169327706282784e-05, + "loss": 0.0109, "step": 13610 }, { - "grad_norm": 0.24050778150558472, - "learning_rate": 2.534205361582834e-05, - "loss": 0.006, + "grad_norm": 0.27707400918006897, + "learning_rate": 9.167805970080029e-05, + "loss": 0.0095, "step": 13620 }, { - "grad_norm": 0.18059442937374115, - "learning_rate": 2.527016644215338e-05, - "loss": 0.0067, + "grad_norm": 0.24676603078842163, + "learning_rate": 9.166282967807864e-05, + "loss": 0.0109, "step": 13630 }, { - "grad_norm": 0.18381164968013763, - "learning_rate": 2.519834687889905e-05, - "loss": 0.007, + "grad_norm": 0.28476017713546753, + "learning_rate": 9.16475869992894e-05, + "loss": 0.0104, "step": 13640 }, { - "grad_norm": 0.2277163565158844, - "learning_rate": 2.5126595122417295e-05, - "loss": 0.0074, + "grad_norm": 0.3044326901435852, + "learning_rate": 9.163233166906284e-05, + "loss": 0.0109, "step": 13650 }, { - "grad_norm": 0.1679578423500061, - "learning_rate": 2.5054911368874713e-05, - "loss": 0.0052, + "grad_norm": 0.2360885739326477, + "learning_rate": 9.161706369203317e-05, + "loss": 0.0093, "step": 13660 }, { - "grad_norm": 0.19412367045879364, - "learning_rate": 2.4983295814251916e-05, - "loss": 0.0077, + "grad_norm": 0.26497066020965576, + "learning_rate": 9.16017830728384e-05, + "loss": 0.0114, "step": 13670 }, { - "grad_norm": 0.22263267636299133, - "learning_rate": 2.4911748654343105e-05, - "loss": 0.006, + "grad_norm": 0.28854620456695557, + "learning_rate": 9.158648981612035e-05, + "loss": 0.0097, "step": 13680 }, { - "grad_norm": 0.23625227808952332, - "learning_rate": 2.4840270084755463e-05, - "loss": 0.0062, + "grad_norm": 0.3069143295288086, + "learning_rate": 9.157118392652472e-05, + "loss": 0.011, "step": 13690 }, { - "grad_norm": 0.22303491830825806, - "learning_rate": 2.4768860300908685e-05, - "loss": 0.0059, + "grad_norm": 0.29258447885513306, + "learning_rate": 9.155586540870104e-05, + "loss": 0.0106, "step": 13700 }, { - "grad_norm": 0.20615343749523163, - "learning_rate": 2.469751949803443e-05, - "loss": 0.0069, + "grad_norm": 0.3380049765110016, + "learning_rate": 9.154053426730267e-05, + "loss": 0.0117, "step": 13710 }, { - "grad_norm": 0.22624443471431732, - "learning_rate": 2.4626247871175666e-05, - "loss": 0.008, + "grad_norm": 0.2883357107639313, + "learning_rate": 9.15251905069868e-05, + "loss": 0.0126, "step": 13720 }, { - "grad_norm": 0.25985249876976013, - "learning_rate": 2.4555045615186346e-05, - "loss": 0.0065, + "grad_norm": 0.26026731729507446, + "learning_rate": 9.150983413241446e-05, + "loss": 0.0101, "step": 13730 }, { - "grad_norm": 0.2458672672510147, - "learning_rate": 2.4483912924730677e-05, - "loss": 0.0072, + "grad_norm": 0.34788039326667786, + "learning_rate": 9.149446514825051e-05, + "loss": 0.012, "step": 13740 }, { - "grad_norm": 0.21012453734874725, - "learning_rate": 2.4412849994282742e-05, - "loss": 0.0048, + "grad_norm": 0.31428471207618713, + "learning_rate": 9.147908355916365e-05, + "loss": 0.0091, "step": 13750 }, { - "grad_norm": 0.14065012335777283, - "learning_rate": 2.434185701812592e-05, - "loss": 0.0052, + "grad_norm": 0.2580438256263733, + "learning_rate": 9.146368936982642e-05, + "loss": 0.009, "step": 13760 }, { - "grad_norm": 0.19618520140647888, - "learning_rate": 2.4270934190352218e-05, - "loss": 0.0071, + "grad_norm": 0.28266477584838867, + "learning_rate": 9.144828258491511e-05, + "loss": 0.0115, "step": 13770 }, { - "grad_norm": 0.1482352614402771, - "learning_rate": 2.4200081704861998e-05, - "loss": 0.0074, + "grad_norm": 0.281694233417511, + "learning_rate": 9.143286320910996e-05, + "loss": 0.012, "step": 13780 }, { - "grad_norm": 0.20872756838798523, - "learning_rate": 2.412929975536321e-05, - "loss": 0.0069, + "grad_norm": 0.23622789978981018, + "learning_rate": 9.141743124709491e-05, + "loss": 0.0098, "step": 13790 }, { - "grad_norm": 0.20830373466014862, - "learning_rate": 2.4058588535371017e-05, - "loss": 0.0062, + "grad_norm": 0.30679628252983093, + "learning_rate": 9.140198670355784e-05, + "loss": 0.0106, "step": 13800 }, { - "grad_norm": 0.2667636275291443, - "learning_rate": 2.3987948238207243e-05, - "loss": 0.0057, + "grad_norm": 0.2917974591255188, + "learning_rate": 9.138652958319034e-05, + "loss": 0.0093, "step": 13810 }, { - "grad_norm": 0.24201737344264984, - "learning_rate": 2.3917379056999678e-05, - "loss": 0.0067, + "grad_norm": 0.2997232973575592, + "learning_rate": 9.137105989068791e-05, + "loss": 0.0109, "step": 13820 }, { - "grad_norm": 0.17350678145885468, - "learning_rate": 2.3846881184681824e-05, - "loss": 0.0046, + "grad_norm": 0.3329172730445862, + "learning_rate": 9.135557763074983e-05, + "loss": 0.0089, "step": 13830 }, { - "grad_norm": 0.17855527997016907, - "learning_rate": 2.377645481399214e-05, - "loss": 0.0052, + "grad_norm": 0.22885291278362274, + "learning_rate": 9.13400828080792e-05, + "loss": 0.0091, "step": 13840 }, { - "grad_norm": 0.24410289525985718, - "learning_rate": 2.3706100137473667e-05, - "loss": 0.0069, + "grad_norm": 0.2628379464149475, + "learning_rate": 9.132457542738292e-05, + "loss": 0.0108, "step": 13850 }, { - "grad_norm": 0.17104095220565796, - "learning_rate": 2.3635817347473394e-05, - "loss": 0.0078, + "grad_norm": 0.34049901366233826, + "learning_rate": 9.130905549337174e-05, + "loss": 0.012, "step": 13860 }, { - "grad_norm": 0.20315547287464142, - "learning_rate": 2.3565606636141757e-05, - "loss": 0.007, + "grad_norm": 0.33799418807029724, + "learning_rate": 9.129352301076021e-05, + "loss": 0.0114, "step": 13870 }, { - "grad_norm": 0.24569718539714813, - "learning_rate": 2.3495468195432203e-05, - "loss": 0.0071, + "grad_norm": 0.29019302129745483, + "learning_rate": 9.127797798426668e-05, + "loss": 0.011, "step": 13880 }, { - "grad_norm": 0.1952163428068161, - "learning_rate": 2.3425402217100507e-05, - "loss": 0.0052, + "grad_norm": 0.22545388340950012, + "learning_rate": 9.126242041861333e-05, + "loss": 0.0089, "step": 13890 }, { - "grad_norm": 0.18253405392169952, - "learning_rate": 2.3355408892704424e-05, - "loss": 0.0068, + "grad_norm": 0.2698425352573395, + "learning_rate": 9.124685031852611e-05, + "loss": 0.0109, "step": 13900 }, { - "grad_norm": 0.16422688961029053, - "learning_rate": 2.3285488413603003e-05, - "loss": 0.0046, + "grad_norm": 0.2501087486743927, + "learning_rate": 9.123126768873482e-05, + "loss": 0.0087, "step": 13910 }, { - "grad_norm": 0.2229805588722229, - "learning_rate": 2.321564097095615e-05, - "loss": 0.0048, + "grad_norm": 0.2929692566394806, + "learning_rate": 9.121567253397308e-05, + "loss": 0.0088, "step": 13920 }, { - "grad_norm": 0.3146781921386719, - "learning_rate": 2.3145866755724142e-05, - "loss": 0.0062, + "grad_norm": 0.305990606546402, + "learning_rate": 9.120006485897824e-05, + "loss": 0.0097, "step": 13930 }, { - "grad_norm": 0.24816946685314178, - "learning_rate": 2.307616595866699e-05, - "loss": 0.0063, + "grad_norm": 0.22306327521800995, + "learning_rate": 9.118444466849152e-05, + "loss": 0.0101, "step": 13940 }, { - "grad_norm": 0.20742735266685486, - "learning_rate": 2.3006538770344032e-05, - "loss": 0.0067, + "grad_norm": 0.3023201823234558, + "learning_rate": 9.116881196725793e-05, + "loss": 0.0113, "step": 13950 }, { - "grad_norm": 0.17962752282619476, - "learning_rate": 2.293698538111334e-05, - "loss": 0.0056, + "grad_norm": 0.33773621916770935, + "learning_rate": 9.115316676002627e-05, + "loss": 0.01, "step": 13960 }, { - "grad_norm": 0.2705751359462738, - "learning_rate": 2.28675059811312e-05, - "loss": 0.0065, + "grad_norm": 0.2963753640651703, + "learning_rate": 9.113750905154911e-05, + "loss": 0.0107, "step": 13970 }, { - "grad_norm": 0.23835845291614532, - "learning_rate": 2.279810076035167e-05, - "loss": 0.0071, + "grad_norm": 0.2869108021259308, + "learning_rate": 9.112183884658289e-05, + "loss": 0.012, "step": 13980 }, { - "grad_norm": 0.24078679084777832, - "learning_rate": 2.272876990852596e-05, - "loss": 0.0076, + "grad_norm": 0.24208708107471466, + "learning_rate": 9.11061561498878e-05, + "loss": 0.0124, "step": 13990 }, { - "grad_norm": 0.22596397995948792, - "learning_rate": 2.265951361520195e-05, - "loss": 0.0101, + "grad_norm": 0.3181329667568207, + "learning_rate": 9.109046096622779e-05, + "loss": 0.0149, "step": 14000 }, { - "grad_norm": 0.19356085360050201, - "learning_rate": 2.2590332069723748e-05, - "loss": 0.0048, + "grad_norm": 0.2435329407453537, + "learning_rate": 9.107475330037069e-05, + "loss": 0.0088, "step": 14010 }, { - "grad_norm": 0.16043022274971008, - "learning_rate": 2.2521225461231004e-05, - "loss": 0.0067, + "grad_norm": 0.26075267791748047, + "learning_rate": 9.105903315708806e-05, + "loss": 0.0101, "step": 14020 }, { - "grad_norm": 0.2451379895210266, - "learning_rate": 2.2452193978658597e-05, - "loss": 0.0049, + "grad_norm": 0.31523725390434265, + "learning_rate": 9.104330054115524e-05, + "loss": 0.0088, "step": 14030 }, { - "grad_norm": 0.19134803116321564, - "learning_rate": 2.238323781073594e-05, - "loss": 0.0071, + "grad_norm": 0.2785132825374603, + "learning_rate": 9.102755545735141e-05, + "loss": 0.0116, "step": 14040 }, { - "grad_norm": 0.1770501583814621, - "learning_rate": 2.2314357145986552e-05, - "loss": 0.007, + "grad_norm": 0.22342707216739655, + "learning_rate": 9.10117979104595e-05, + "loss": 0.0112, "step": 14050 }, { - "grad_norm": 0.19176732003688812, - "learning_rate": 2.224555217272757e-05, - "loss": 0.007, + "grad_norm": 0.3659726083278656, + "learning_rate": 9.099602790526624e-05, + "loss": 0.0105, "step": 14060 }, { - "grad_norm": 0.2289755940437317, - "learning_rate": 2.2176823079069127e-05, - "loss": 0.0065, + "grad_norm": 0.2550984025001526, + "learning_rate": 9.098024544656212e-05, + "loss": 0.0109, "step": 14070 }, { - "grad_norm": 0.27378588914871216, - "learning_rate": 2.210817005291398e-05, - "loss": 0.0058, + "grad_norm": 0.3669590651988983, + "learning_rate": 9.096445053914148e-05, + "loss": 0.0106, "step": 14080 }, { - "grad_norm": 0.2435673326253891, - "learning_rate": 2.203959328195686e-05, - "loss": 0.0069, + "grad_norm": 0.3143695890903473, + "learning_rate": 9.094864318780236e-05, + "loss": 0.0112, "step": 14090 }, { - "grad_norm": 0.22978395223617554, - "learning_rate": 2.1971092953684026e-05, - "loss": 0.006, + "grad_norm": 0.25870826840400696, + "learning_rate": 9.093282339734663e-05, + "loss": 0.0106, "step": 14100 }, { - "grad_norm": 0.19038143754005432, - "learning_rate": 2.1902669255372788e-05, - "loss": 0.0054, + "grad_norm": 0.22017966210842133, + "learning_rate": 9.091699117257992e-05, + "loss": 0.0095, "step": 14110 }, { - "grad_norm": 0.16993741691112518, - "learning_rate": 2.1834322374090897e-05, - "loss": 0.0064, + "grad_norm": 0.2652941048145294, + "learning_rate": 9.090114651831163e-05, + "loss": 0.0106, "step": 14120 }, { - "grad_norm": 0.22536249458789825, - "learning_rate": 2.1766052496696153e-05, - "loss": 0.0061, + "grad_norm": 0.2580643594264984, + "learning_rate": 9.088528943935497e-05, + "loss": 0.0096, "step": 14130 }, { - "grad_norm": 0.1695421189069748, - "learning_rate": 2.169785980983577e-05, - "loss": 0.0053, + "grad_norm": 0.23172371089458466, + "learning_rate": 9.086941994052689e-05, + "loss": 0.009, "step": 14140 }, { - "grad_norm": 0.20765681564807892, - "learning_rate": 2.162974449994593e-05, - "loss": 0.0055, + "grad_norm": 0.2915630042552948, + "learning_rate": 9.085353802664813e-05, + "loss": 0.0098, "step": 14150 }, { - "grad_norm": 0.20967355370521545, - "learning_rate": 2.1561706753251337e-05, - "loss": 0.0071, + "grad_norm": 0.2782306969165802, + "learning_rate": 9.08376437025432e-05, + "loss": 0.0111, "step": 14160 }, { - "grad_norm": 0.24757470190525055, - "learning_rate": 2.1493746755764544e-05, - "loss": 0.006, + "grad_norm": 0.3387666940689087, + "learning_rate": 9.082173697304035e-05, + "loss": 0.0099, "step": 14170 }, { - "grad_norm": 0.17080722749233246, - "learning_rate": 2.1425864693285635e-05, - "loss": 0.0087, + "grad_norm": 0.3233988881111145, + "learning_rate": 9.080581784297166e-05, + "loss": 0.0131, "step": 14180 }, { - "grad_norm": 0.1654953807592392, - "learning_rate": 2.1358060751401547e-05, - "loss": 0.0062, + "grad_norm": 0.26535287499427795, + "learning_rate": 9.078988631717291e-05, + "loss": 0.0103, "step": 14190 }, { - "grad_norm": 0.23503398895263672, - "learning_rate": 2.129033511548566e-05, - "loss": 0.0078, + "grad_norm": 0.32505011558532715, + "learning_rate": 9.077394240048369e-05, + "loss": 0.0134, "step": 14200 }, { - "grad_norm": 0.19658979773521423, - "learning_rate": 2.1222687970697315e-05, - "loss": 0.0043, + "grad_norm": 0.3045431077480316, + "learning_rate": 9.075798609774736e-05, + "loss": 0.0082, "step": 14210 }, { - "grad_norm": 0.24506372213363647, - "learning_rate": 2.1155119501981173e-05, - "loss": 0.0062, + "grad_norm": 0.3463459610939026, + "learning_rate": 9.0742017413811e-05, + "loss": 0.0109, "step": 14220 }, { - "grad_norm": 0.23663446307182312, - "learning_rate": 2.1087629894066895e-05, - "loss": 0.0053, + "grad_norm": 0.31739935278892517, + "learning_rate": 9.072603635352548e-05, + "loss": 0.0103, "step": 14230 }, { - "grad_norm": 0.20866233110427856, - "learning_rate": 2.1020219331468473e-05, - "loss": 0.0076, + "grad_norm": 0.3275075852870941, + "learning_rate": 9.071004292174541e-05, + "loss": 0.0123, "step": 14240 }, { - "grad_norm": 0.2166251242160797, - "learning_rate": 2.095288799848379e-05, - "loss": 0.0069, + "grad_norm": 0.29646533727645874, + "learning_rate": 9.06940371233292e-05, + "loss": 0.0119, "step": 14250 }, { - "grad_norm": 0.17026379704475403, - "learning_rate": 2.088563607919417e-05, - "loss": 0.0052, + "grad_norm": 0.2616863250732422, + "learning_rate": 9.067801896313898e-05, + "loss": 0.0102, "step": 14260 }, { - "grad_norm": 0.26690876483917236, - "learning_rate": 2.0818463757463786e-05, - "loss": 0.0055, + "grad_norm": 0.3164909780025482, + "learning_rate": 9.066198844604064e-05, + "loss": 0.0112, "step": 14270 }, { - "grad_norm": 0.21364854276180267, - "learning_rate": 2.0751371216939175e-05, - "loss": 0.0062, + "grad_norm": 0.2876752018928528, + "learning_rate": 9.06459455769038e-05, + "loss": 0.0116, "step": 14280 }, { - "grad_norm": 0.25960573554039, - "learning_rate": 2.068435864104882e-05, - "loss": 0.0057, + "grad_norm": 0.3506101071834564, + "learning_rate": 9.062989036060193e-05, + "loss": 0.0109, "step": 14290 }, { - "grad_norm": 0.2138037234544754, - "learning_rate": 2.0617426213002506e-05, - "loss": 0.0089, + "grad_norm": 0.32811760902404785, + "learning_rate": 9.061382280201212e-05, + "loss": 0.0134, "step": 14300 }, { - "grad_norm": 0.20547112822532654, - "learning_rate": 2.055057411579097e-05, - "loss": 0.0059, + "grad_norm": 0.30635514855384827, + "learning_rate": 9.059774290601528e-05, + "loss": 0.0103, "step": 14310 }, { - "grad_norm": 0.32553014159202576, - "learning_rate": 2.0483802532185286e-05, - "loss": 0.0066, + "grad_norm": 0.2746696174144745, + "learning_rate": 9.058165067749606e-05, + "loss": 0.0102, "step": 14320 }, { - "grad_norm": 0.22871890664100647, - "learning_rate": 2.041711164473638e-05, - "loss": 0.0072, + "grad_norm": 0.30828356742858887, + "learning_rate": 9.056554612134288e-05, + "loss": 0.0113, "step": 14330 }, { - "grad_norm": 0.20716866850852966, - "learning_rate": 2.0350501635774637e-05, - "loss": 0.0055, + "grad_norm": 0.2606695294380188, + "learning_rate": 9.054942924244785e-05, + "loss": 0.0091, "step": 14340 }, { - "grad_norm": 0.17182990908622742, - "learning_rate": 2.0283972687409247e-05, - "loss": 0.0063, + "grad_norm": 0.24746853113174438, + "learning_rate": 9.053330004570686e-05, + "loss": 0.0107, "step": 14350 }, { - "grad_norm": 0.16304871439933777, - "learning_rate": 2.021752498152784e-05, - "loss": 0.0061, + "grad_norm": 0.28739649057388306, + "learning_rate": 9.051715853601955e-05, + "loss": 0.0104, "step": 14360 }, { - "grad_norm": 0.22496426105499268, - "learning_rate": 2.015115869979589e-05, - "loss": 0.0046, + "grad_norm": 0.2748439610004425, + "learning_rate": 9.050100471828926e-05, + "loss": 0.0084, "step": 14370 }, { - "grad_norm": 0.1372428685426712, - "learning_rate": 2.0084874023656265e-05, - "loss": 0.0049, + "grad_norm": 0.22228635847568512, + "learning_rate": 9.048483859742311e-05, + "loss": 0.0087, "step": 14380 }, { - "grad_norm": 0.22879809141159058, - "learning_rate": 2.001867113432877e-05, - "loss": 0.0075, + "grad_norm": 0.2524581849575043, + "learning_rate": 9.046866017833193e-05, + "loss": 0.0122, "step": 14390 }, { - "grad_norm": 0.18383388221263885, - "learning_rate": 1.995255021280954e-05, - "loss": 0.0059, + "grad_norm": 0.2601880431175232, + "learning_rate": 9.045246946593029e-05, + "loss": 0.0112, "step": 14400 }, { - "grad_norm": 0.1956668198108673, - "learning_rate": 1.9886511439870688e-05, - "loss": 0.007, + "grad_norm": 0.2924646735191345, + "learning_rate": 9.043626646513652e-05, + "loss": 0.0115, "step": 14410 }, { - "grad_norm": 0.2605588436126709, - "learning_rate": 1.9820554996059675e-05, - "loss": 0.0064, + "grad_norm": 0.25464484095573425, + "learning_rate": 9.042005118087267e-05, + "loss": 0.0107, "step": 14420 }, { - "grad_norm": 0.21275591850280762, - "learning_rate": 1.9754681061698893e-05, - "loss": 0.0047, + "grad_norm": 0.2606675922870636, + "learning_rate": 9.040382361806448e-05, + "loss": 0.0095, "step": 14430 }, { - "grad_norm": 0.15767215192317963, - "learning_rate": 1.9688889816885185e-05, - "loss": 0.0051, + "grad_norm": 0.30493468046188354, + "learning_rate": 9.038758378164148e-05, + "loss": 0.0096, "step": 14440 }, { - "grad_norm": 0.206788569688797, - "learning_rate": 1.962318144148928e-05, - "loss": 0.0053, + "grad_norm": 0.3285244107246399, + "learning_rate": 9.037133167653691e-05, + "loss": 0.0099, "step": 14450 }, { - "grad_norm": 0.21391050517559052, - "learning_rate": 1.955755611515539e-05, - "loss": 0.0054, + "grad_norm": 0.26329296827316284, + "learning_rate": 9.035506730768771e-05, + "loss": 0.0106, "step": 14460 }, { - "grad_norm": 0.15545658767223358, - "learning_rate": 1.9492014017300642e-05, - "loss": 0.0068, + "grad_norm": 0.335332989692688, + "learning_rate": 9.033879068003458e-05, + "loss": 0.0119, "step": 14470 }, { - "grad_norm": 0.1992073655128479, - "learning_rate": 1.942655532711461e-05, - "loss": 0.005, + "grad_norm": 0.21017469465732574, + "learning_rate": 9.032250179852193e-05, + "loss": 0.0104, "step": 14480 }, { - "grad_norm": 0.17155469954013824, - "learning_rate": 1.9361180223558882e-05, - "loss": 0.0068, + "grad_norm": 0.30278006196022034, + "learning_rate": 9.030620066809787e-05, + "loss": 0.0112, "step": 14490 }, { - "grad_norm": 0.1756027191877365, - "learning_rate": 1.929588888536647e-05, - "loss": 0.0063, + "grad_norm": 0.27095702290534973, + "learning_rate": 9.028988729371428e-05, + "loss": 0.0104, "step": 14500 }, { - "grad_norm": 0.21194574236869812, - "learning_rate": 1.9230681491041425e-05, - "loss": 0.0057, + "grad_norm": 0.2386666089296341, + "learning_rate": 9.027356168032673e-05, + "loss": 0.011, "step": 14510 }, { - "grad_norm": 0.18223026394844055, - "learning_rate": 1.9165558218858264e-05, - "loss": 0.0052, + "grad_norm": 0.22701570391654968, + "learning_rate": 9.02572238328945e-05, + "loss": 0.0097, "step": 14520 }, { - "grad_norm": 0.250331848859787, - "learning_rate": 1.9100519246861505e-05, - "loss": 0.0053, + "grad_norm": 0.24391967058181763, + "learning_rate": 9.02408737563806e-05, + "loss": 0.0093, "step": 14530 }, { - "grad_norm": 0.2652829885482788, - "learning_rate": 1.9035564752865248e-05, - "loss": 0.0076, + "grad_norm": 0.30137214064598083, + "learning_rate": 9.022451145575174e-05, + "loss": 0.0117, "step": 14540 }, { - "grad_norm": 0.23563922941684723, - "learning_rate": 1.897069491445258e-05, - "loss": 0.0057, + "grad_norm": 0.3091307282447815, + "learning_rate": 9.02081369359784e-05, + "loss": 0.0101, "step": 14550 }, { - "grad_norm": 0.21190163493156433, - "learning_rate": 1.890590990897515e-05, - "loss": 0.0057, + "grad_norm": 0.2756590247154236, + "learning_rate": 9.019175020203465e-05, + "loss": 0.01, "step": 14560 }, { - "grad_norm": 0.19106173515319824, - "learning_rate": 1.884120991355272e-05, - "loss": 0.0059, + "grad_norm": 0.2108382284641266, + "learning_rate": 9.017535125889842e-05, + "loss": 0.0109, "step": 14570 }, { - "grad_norm": 0.19955740869045258, - "learning_rate": 1.8776595105072576e-05, - "loss": 0.0054, + "grad_norm": 0.2355968952178955, + "learning_rate": 9.015894011155124e-05, + "loss": 0.0103, "step": 14580 }, { - "grad_norm": 0.1958319991827011, - "learning_rate": 1.8712065660189166e-05, - "loss": 0.0054, + "grad_norm": 0.25008463859558105, + "learning_rate": 9.014251676497838e-05, + "loss": 0.0096, "step": 14590 }, { - "grad_norm": 0.20281006395816803, - "learning_rate": 1.8647621755323513e-05, - "loss": 0.0054, + "grad_norm": 0.2834985852241516, + "learning_rate": 9.012608122416884e-05, + "loss": 0.0094, "step": 14600 }, { - "grad_norm": 0.2089354693889618, - "learning_rate": 1.858326356666278e-05, - "loss": 0.0055, + "grad_norm": 0.301725834608078, + "learning_rate": 9.010963349411529e-05, + "loss": 0.0101, "step": 14610 }, { - "grad_norm": 0.3210635781288147, - "learning_rate": 1.851899127015983e-05, - "loss": 0.0061, + "grad_norm": 0.25175178050994873, + "learning_rate": 9.00931735798141e-05, + "loss": 0.0103, "step": 14620 }, { - "grad_norm": 0.2990008592605591, - "learning_rate": 1.8454805041532626e-05, - "loss": 0.0067, + "grad_norm": 0.24005499482154846, + "learning_rate": 9.00767014862654e-05, + "loss": 0.0106, "step": 14630 }, { - "grad_norm": 0.2500159740447998, - "learning_rate": 1.8390705056263906e-05, - "loss": 0.0056, + "grad_norm": 0.29720914363861084, + "learning_rate": 9.006021721847295e-05, + "loss": 0.0094, "step": 14640 }, { - "grad_norm": 0.2781897485256195, - "learning_rate": 1.832669148960057e-05, - "loss": 0.0049, + "grad_norm": 0.34712278842926025, + "learning_rate": 9.004372078144423e-05, + "loss": 0.0091, "step": 14650 }, { - "grad_norm": 0.17028845846652985, - "learning_rate": 1.8262764516553233e-05, - "loss": 0.0085, + "grad_norm": 0.22825026512145996, + "learning_rate": 9.002721218019043e-05, + "loss": 0.0132, "step": 14660 }, { - "grad_norm": 0.2419242262840271, - "learning_rate": 1.8198924311895843e-05, - "loss": 0.0053, + "grad_norm": 0.3266952633857727, + "learning_rate": 9.001069141972642e-05, + "loss": 0.0098, "step": 14670 }, { - "grad_norm": 0.13097338378429413, - "learning_rate": 1.813517105016505e-05, - "loss": 0.0059, + "grad_norm": 0.26162558794021606, + "learning_rate": 8.99941585050708e-05, + "loss": 0.0104, "step": 14680 }, { - "grad_norm": 0.20473894476890564, - "learning_rate": 1.8071504905659888e-05, - "loss": 0.0055, + "grad_norm": 0.26830780506134033, + "learning_rate": 8.997761344124578e-05, + "loss": 0.0106, "step": 14690 }, { - "grad_norm": 0.1552768498659134, - "learning_rate": 1.800792605244109e-05, - "loss": 0.0053, + "grad_norm": 0.3264312148094177, + "learning_rate": 8.996105623327737e-05, + "loss": 0.0107, "step": 14700 }, { - "grad_norm": 0.14139609038829803, - "learning_rate": 1.7944434664330844e-05, - "loss": 0.0042, + "grad_norm": 0.2410210818052292, + "learning_rate": 8.994448688619517e-05, + "loss": 0.0082, "step": 14710 }, { - "grad_norm": 0.14989487826824188, - "learning_rate": 1.7881030914912212e-05, - "loss": 0.0051, + "grad_norm": 0.284929484128952, + "learning_rate": 8.992790540503253e-05, + "loss": 0.01, "step": 14720 }, { - "grad_norm": 0.2136131227016449, - "learning_rate": 1.7817714977528577e-05, - "loss": 0.0056, + "grad_norm": 0.2587823271751404, + "learning_rate": 8.991131179482648e-05, + "loss": 0.0094, "step": 14730 }, { - "grad_norm": 0.22079499065876007, - "learning_rate": 1.7754487025283332e-05, - "loss": 0.0078, + "grad_norm": 0.28239545226097107, + "learning_rate": 8.989470606061768e-05, + "loss": 0.0116, "step": 14740 }, { - "grad_norm": 0.18154866993427277, - "learning_rate": 1.7691347231039275e-05, - "loss": 0.0057, + "grad_norm": 0.30279162526130676, + "learning_rate": 8.987808820745056e-05, + "loss": 0.0103, "step": 14750 }, { - "grad_norm": 0.17814518511295319, - "learning_rate": 1.7628295767418164e-05, - "loss": 0.0047, + "grad_norm": 0.29264846444129944, + "learning_rate": 8.986145824037315e-05, + "loss": 0.0088, "step": 14760 }, { - "grad_norm": 0.17428627610206604, - "learning_rate": 1.7565332806800333e-05, - "loss": 0.0055, + "grad_norm": 0.19506973028182983, + "learning_rate": 8.984481616443721e-05, + "loss": 0.01, "step": 14770 }, { - "grad_norm": 0.2982683777809143, - "learning_rate": 1.750245852132408e-05, - "loss": 0.0052, + "grad_norm": 0.2171848863363266, + "learning_rate": 8.982816198469815e-05, + "loss": 0.01, "step": 14780 }, { - "grad_norm": 0.2404620200395584, - "learning_rate": 1.7439673082885323e-05, - "loss": 0.0057, + "grad_norm": 0.3335197865962982, + "learning_rate": 8.98114957062151e-05, + "loss": 0.0101, "step": 14790 }, { - "grad_norm": 0.14670974016189575, - "learning_rate": 1.7376976663137047e-05, - "loss": 0.0056, + "grad_norm": 0.2951051890850067, + "learning_rate": 8.97948173340508e-05, + "loss": 0.0103, "step": 14800 }, { - "grad_norm": 0.2664196193218231, - "learning_rate": 1.7314369433488853e-05, - "loss": 0.0051, + "grad_norm": 0.30811333656311035, + "learning_rate": 8.977812687327172e-05, + "loss": 0.01, "step": 14810 }, { - "grad_norm": 0.19232824444770813, - "learning_rate": 1.7251851565106548e-05, - "loss": 0.0066, + "grad_norm": 0.2803950309753418, + "learning_rate": 8.976142432894798e-05, + "loss": 0.0109, "step": 14820 }, { - "grad_norm": 0.1771378368139267, - "learning_rate": 1.7189423228911574e-05, - "loss": 0.0058, + "grad_norm": 0.3032907545566559, + "learning_rate": 8.974470970615336e-05, + "loss": 0.0113, "step": 14830 }, { - "grad_norm": 0.17615680396556854, - "learning_rate": 1.7127084595580606e-05, - "loss": 0.0061, + "grad_norm": 0.2634539306163788, + "learning_rate": 8.972798300996534e-05, + "loss": 0.0107, "step": 14840 }, { - "grad_norm": 0.20852655172348022, - "learning_rate": 1.706483583554513e-05, - "loss": 0.0051, + "grad_norm": 0.2380373328924179, + "learning_rate": 8.971124424546504e-05, + "loss": 0.0095, "step": 14850 }, { - "grad_norm": 0.16545358300209045, - "learning_rate": 1.700267711899083e-05, - "loss": 0.0042, + "grad_norm": 0.22044654190540314, + "learning_rate": 8.969449341773724e-05, + "loss": 0.0084, "step": 14860 }, { - "grad_norm": 0.20421037077903748, - "learning_rate": 1.69406086158573e-05, - "loss": 0.0044, + "grad_norm": 0.171296626329422, + "learning_rate": 8.967773053187042e-05, + "loss": 0.0082, "step": 14870 }, { - "grad_norm": 0.19449642300605774, - "learning_rate": 1.6878630495837455e-05, - "loss": 0.0061, + "grad_norm": 0.30781790614128113, + "learning_rate": 8.966095559295668e-05, + "loss": 0.0105, "step": 14880 }, { - "grad_norm": 0.23133471608161926, - "learning_rate": 1.681674292837707e-05, - "loss": 0.0064, + "grad_norm": 0.2670572102069855, + "learning_rate": 8.964416860609184e-05, + "loss": 0.0111, "step": 14890 }, { - "grad_norm": 0.14292384684085846, - "learning_rate": 1.6754946082674444e-05, - "loss": 0.0058, + "grad_norm": 0.23065350949764252, + "learning_rate": 8.962736957637532e-05, + "loss": 0.0095, "step": 14900 }, { - "grad_norm": 0.12971968948841095, - "learning_rate": 1.6693240127679748e-05, - "loss": 0.0044, + "grad_norm": 0.2330165058374405, + "learning_rate": 8.96105585089102e-05, + "loss": 0.0085, "step": 14910 }, { - "grad_norm": 0.1865580976009369, - "learning_rate": 1.663162523209475e-05, - "loss": 0.0071, + "grad_norm": 0.2678815424442291, + "learning_rate": 8.959373540880329e-05, + "loss": 0.0122, "step": 14920 }, { - "grad_norm": 0.18029199540615082, - "learning_rate": 1.6570101564372193e-05, - "loss": 0.007, + "grad_norm": 0.30811116099357605, + "learning_rate": 8.957690028116495e-05, + "loss": 0.0123, "step": 14930 }, { - "grad_norm": 0.24925662577152252, - "learning_rate": 1.650866929271543e-05, - "loss": 0.0062, + "grad_norm": 0.2776390612125397, + "learning_rate": 8.956005313110928e-05, + "loss": 0.0113, "step": 14940 }, { - "grad_norm": 0.17494648694992065, - "learning_rate": 1.644732858507797e-05, - "loss": 0.0055, + "grad_norm": 0.2533762454986572, + "learning_rate": 8.9543193963754e-05, + "loss": 0.0091, "step": 14950 }, { - "grad_norm": 0.15730054676532745, - "learning_rate": 1.6386079609162943e-05, - "loss": 0.0047, + "grad_norm": 0.24179822206497192, + "learning_rate": 8.952632278422048e-05, + "loss": 0.0083, "step": 14960 }, { - "grad_norm": 0.2508449852466583, - "learning_rate": 1.6324922532422742e-05, - "loss": 0.0098, + "grad_norm": 0.30182337760925293, + "learning_rate": 8.95094395976337e-05, + "loss": 0.0148, "step": 14970 }, { - "grad_norm": 0.190720796585083, - "learning_rate": 1.6263857522058434e-05, - "loss": 0.0054, + "grad_norm": 0.21201108396053314, + "learning_rate": 8.949254440912239e-05, + "loss": 0.0092, "step": 14980 }, { - "grad_norm": 0.1414359211921692, - "learning_rate": 1.6202884745019443e-05, - "loss": 0.0046, + "grad_norm": 0.2630045711994171, + "learning_rate": 8.94756372238188e-05, + "loss": 0.0085, "step": 14990 }, { - "grad_norm": 0.14060048758983612, - "learning_rate": 1.614200436800304e-05, - "loss": 0.0061, + "grad_norm": 0.2681374251842499, + "learning_rate": 8.945871804685892e-05, + "loss": 0.0106, "step": 15000 }, { - "grad_norm": 0.19313286244869232, - "learning_rate": 1.6081216557453814e-05, - "loss": 0.0049, + "grad_norm": 0.2599766254425049, + "learning_rate": 8.944178688338236e-05, + "loss": 0.0094, "step": 15010 }, { - "grad_norm": 0.17659884691238403, - "learning_rate": 1.6020521479563367e-05, - "loss": 0.0054, + "grad_norm": 0.23193500936031342, + "learning_rate": 8.942484373853233e-05, + "loss": 0.0101, "step": 15020 }, { - "grad_norm": 0.14806364476680756, - "learning_rate": 1.5959919300269654e-05, - "loss": 0.0041, + "grad_norm": 0.2852785289287567, + "learning_rate": 8.940788861745572e-05, + "loss": 0.0086, "step": 15030 }, { - "grad_norm": 0.15692979097366333, - "learning_rate": 1.5899410185256764e-05, - "loss": 0.0037, + "grad_norm": 0.2425108402967453, + "learning_rate": 8.939092152530308e-05, + "loss": 0.0079, "step": 15040 }, { - "grad_norm": 0.14825497567653656, - "learning_rate": 1.583899429995431e-05, - "loss": 0.0046, + "grad_norm": 0.2836150825023651, + "learning_rate": 8.937394246722853e-05, + "loss": 0.0094, "step": 15050 }, { - "grad_norm": 0.1588299572467804, - "learning_rate": 1.5778671809536993e-05, - "loss": 0.0043, + "grad_norm": 0.2823534309864044, + "learning_rate": 8.935695144838984e-05, + "loss": 0.0081, "step": 15060 }, { - "grad_norm": 0.13866159319877625, - "learning_rate": 1.5718442878924246e-05, - "loss": 0.0044, + "grad_norm": 0.292885422706604, + "learning_rate": 8.933994847394849e-05, + "loss": 0.0083, "step": 15070 }, { - "grad_norm": 0.2178276777267456, - "learning_rate": 1.5658307672779593e-05, - "loss": 0.0055, + "grad_norm": 0.3051721751689911, + "learning_rate": 8.932293354906949e-05, + "loss": 0.0103, "step": 15080 }, { - "grad_norm": 0.1646864116191864, - "learning_rate": 1.5598266355510427e-05, - "loss": 0.0045, + "grad_norm": 0.2722909450531006, + "learning_rate": 8.930590667892153e-05, + "loss": 0.0089, "step": 15090 }, { - "grad_norm": 0.21160311996936798, - "learning_rate": 1.553831909126744e-05, - "loss": 0.0049, + "grad_norm": 0.29498255252838135, + "learning_rate": 8.928886786867696e-05, + "loss": 0.0096, "step": 15100 }, { - "grad_norm": 0.23382562398910522, - "learning_rate": 1.5478466043944135e-05, - "loss": 0.0066, + "grad_norm": 0.27731165289878845, + "learning_rate": 8.927181712351168e-05, + "loss": 0.0111, "step": 15110 }, { - "grad_norm": 0.18930543959140778, - "learning_rate": 1.5418707377176468e-05, - "loss": 0.0053, + "grad_norm": 0.2584938406944275, + "learning_rate": 8.925475444860527e-05, + "loss": 0.0096, "step": 15120 }, { - "grad_norm": 0.1883104145526886, - "learning_rate": 1.535904325434233e-05, - "loss": 0.0051, + "grad_norm": 0.2471330314874649, + "learning_rate": 8.923767984914092e-05, + "loss": 0.0092, "step": 15130 }, { - "grad_norm": 0.21340157091617584, - "learning_rate": 1.529947383856118e-05, - "loss": 0.0073, + "grad_norm": 0.292123019695282, + "learning_rate": 8.922059333030545e-05, + "loss": 0.0126, "step": 15140 }, { - "grad_norm": 0.23821984231472015, - "learning_rate": 1.5239999292693524e-05, - "loss": 0.0067, + "grad_norm": 0.3022165894508362, + "learning_rate": 8.920349489728928e-05, + "loss": 0.0112, "step": 15150 }, { - "grad_norm": 0.14447997510433197, - "learning_rate": 1.5180619779340505e-05, - "loss": 0.0067, + "grad_norm": 0.27089715003967285, + "learning_rate": 8.918638455528646e-05, + "loss": 0.011, "step": 15160 }, { - "grad_norm": 0.14597798883914948, - "learning_rate": 1.5121335460843428e-05, - "loss": 0.0051, + "grad_norm": 0.2372835874557495, + "learning_rate": 8.916926230949468e-05, + "loss": 0.0105, "step": 15170 }, { - "grad_norm": 0.19506734609603882, - "learning_rate": 1.5062146499283347e-05, - "loss": 0.0062, + "grad_norm": 0.32664525508880615, + "learning_rate": 8.915212816511522e-05, + "loss": 0.0104, "step": 15180 }, { - "grad_norm": 0.18002669513225555, - "learning_rate": 1.5003053056480643e-05, - "loss": 0.007, + "grad_norm": 0.3257906436920166, + "learning_rate": 8.913498212735296e-05, + "loss": 0.0119, "step": 15190 }, { - "grad_norm": 0.2782810926437378, - "learning_rate": 1.4944055293994551e-05, - "loss": 0.0051, + "grad_norm": 0.29897263646125793, + "learning_rate": 8.911782420141643e-05, + "loss": 0.0107, "step": 15200 }, { - "grad_norm": 0.2039763480424881, - "learning_rate": 1.4885153373122656e-05, - "loss": 0.006, + "grad_norm": 0.21231438219547272, + "learning_rate": 8.910065439251775e-05, + "loss": 0.0106, "step": 15210 }, { - "grad_norm": 0.1816200613975525, - "learning_rate": 1.482634745490059e-05, - "loss": 0.005, + "grad_norm": 0.28170809149742126, + "learning_rate": 8.908347270587268e-05, + "loss": 0.0097, "step": 15220 }, { - "grad_norm": 0.15879617631435394, - "learning_rate": 1.4767637700101466e-05, - "loss": 0.0047, + "grad_norm": 0.26248350739479065, + "learning_rate": 8.906627914670054e-05, + "loss": 0.0095, "step": 15230 }, { - "grad_norm": 0.19043923914432526, - "learning_rate": 1.4709024269235528e-05, - "loss": 0.0051, + "grad_norm": 0.30315592885017395, + "learning_rate": 8.904907372022427e-05, + "loss": 0.0104, "step": 15240 }, { - "grad_norm": 0.2104177474975586, - "learning_rate": 1.4650507322549684e-05, - "loss": 0.0047, + "grad_norm": 0.2650705575942993, + "learning_rate": 8.903185643167042e-05, + "loss": 0.0098, "step": 15250 }, { - "grad_norm": 0.22558549046516418, - "learning_rate": 1.4592087020026972e-05, - "loss": 0.0055, + "grad_norm": 0.2631804347038269, + "learning_rate": 8.901462728626919e-05, + "loss": 0.011, "step": 15260 }, { - "grad_norm": 0.1903344690799713, - "learning_rate": 1.4533763521386318e-05, - "loss": 0.0051, + "grad_norm": 0.24776969850063324, + "learning_rate": 8.899738628925429e-05, + "loss": 0.0096, "step": 15270 }, { - "grad_norm": 0.21042105555534363, - "learning_rate": 1.44755369860819e-05, - "loss": 0.0059, + "grad_norm": 0.29048454761505127, + "learning_rate": 8.898013344586312e-05, + "loss": 0.0106, "step": 15280 }, { - "grad_norm": 0.2088128626346588, - "learning_rate": 1.441740757330287e-05, - "loss": 0.0048, + "grad_norm": 0.2221129685640335, + "learning_rate": 8.896286876133661e-05, + "loss": 0.009, "step": 15290 }, { - "grad_norm": 0.2233678549528122, - "learning_rate": 1.4359375441972844e-05, - "loss": 0.0043, + "grad_norm": 0.2735884487628937, + "learning_rate": 8.894559224091933e-05, + "loss": 0.009, "step": 15300 }, { - "grad_norm": 0.13520531356334686, - "learning_rate": 1.4301440750749395e-05, - "loss": 0.0038, + "grad_norm": 0.23851007223129272, + "learning_rate": 8.892830388985942e-05, + "loss": 0.009, "step": 15310 }, { - "grad_norm": 0.21610109508037567, - "learning_rate": 1.4243603658023808e-05, - "loss": 0.0056, + "grad_norm": 0.28519192337989807, + "learning_rate": 8.891100371340864e-05, + "loss": 0.0104, "step": 15320 }, { - "grad_norm": 0.16219377517700195, - "learning_rate": 1.4185864321920444e-05, - "loss": 0.0043, + "grad_norm": 0.21761135756969452, + "learning_rate": 8.889369171682231e-05, + "loss": 0.0085, "step": 15330 }, { - "grad_norm": 0.2393598109483719, - "learning_rate": 1.4128222900296485e-05, - "loss": 0.0057, + "grad_norm": 0.2853968143463135, + "learning_rate": 8.887636790535936e-05, + "loss": 0.0105, "step": 15340 }, { - "grad_norm": 0.16220656037330627, - "learning_rate": 1.407067955074135e-05, - "loss": 0.0056, + "grad_norm": 0.2600754499435425, + "learning_rate": 8.885903228428231e-05, + "loss": 0.0099, "step": 15350 }, { - "grad_norm": 0.22726091742515564, - "learning_rate": 1.4013234430576356e-05, - "loss": 0.0054, + "grad_norm": 0.22754167020320892, + "learning_rate": 8.884168485885727e-05, + "loss": 0.0099, "step": 15360 }, { - "grad_norm": 0.1691230684518814, - "learning_rate": 1.3955887696854286e-05, - "loss": 0.0066, + "grad_norm": 0.2530911862850189, + "learning_rate": 8.882432563435393e-05, + "loss": 0.0121, "step": 15370 }, { - "grad_norm": 0.24777837097644806, - "learning_rate": 1.38986395063589e-05, - "loss": 0.0073, + "grad_norm": 0.3016035258769989, + "learning_rate": 8.880695461604556e-05, + "loss": 0.0123, "step": 15380 }, { - "grad_norm": 0.2199578583240509, - "learning_rate": 1.3841490015604597e-05, - "loss": 0.0059, + "grad_norm": 0.30656224489212036, + "learning_rate": 8.878957180920901e-05, + "loss": 0.0108, "step": 15390 }, { - "grad_norm": 0.22531665861606598, - "learning_rate": 1.3784439380835879e-05, - "loss": 0.0068, + "grad_norm": 0.33010727167129517, + "learning_rate": 8.877217721912473e-05, + "loss": 0.012, "step": 15400 }, { - "grad_norm": 0.16674892604351044, - "learning_rate": 1.3727487758026986e-05, - "loss": 0.0055, + "grad_norm": 0.2794025242328644, + "learning_rate": 8.875477085107673e-05, + "loss": 0.011, "step": 15410 }, { - "grad_norm": 0.19342897832393646, - "learning_rate": 1.3670635302881525e-05, - "loss": 0.0056, + "grad_norm": 0.32219621539115906, + "learning_rate": 8.87373527103526e-05, + "loss": 0.0107, "step": 15420 }, { - "grad_norm": 0.12783855199813843, - "learning_rate": 1.3613882170831888e-05, - "loss": 0.0066, + "grad_norm": 0.27440646290779114, + "learning_rate": 8.871992280224353e-05, + "loss": 0.0103, "step": 15430 }, { - "grad_norm": 0.17184005677700043, - "learning_rate": 1.355722851703901e-05, - "loss": 0.0069, + "grad_norm": 0.27229413390159607, + "learning_rate": 8.870248113204422e-05, + "loss": 0.0115, "step": 15440 }, { - "grad_norm": 0.20452845096588135, - "learning_rate": 1.3500674496391814e-05, - "loss": 0.0047, + "grad_norm": 0.22909151017665863, + "learning_rate": 8.868502770505306e-05, + "loss": 0.0092, "step": 15450 }, { - "grad_norm": 0.13389019668102264, - "learning_rate": 1.3444220263506795e-05, - "loss": 0.0066, + "grad_norm": 0.24661816656589508, + "learning_rate": 8.86675625265719e-05, + "loss": 0.0112, "step": 15460 }, { - "grad_norm": 0.24266892671585083, - "learning_rate": 1.3387865972727714e-05, - "loss": 0.0067, + "grad_norm": 0.2945097088813782, + "learning_rate": 8.865008560190618e-05, + "loss": 0.0113, "step": 15470 }, { - "grad_norm": 0.1648254692554474, - "learning_rate": 1.3331611778125036e-05, - "loss": 0.0065, + "grad_norm": 0.2632730007171631, + "learning_rate": 8.863259693636496e-05, + "loss": 0.012, "step": 15480 }, { - "grad_norm": 0.20072923600673676, - "learning_rate": 1.3275457833495564e-05, - "loss": 0.0102, + "grad_norm": 0.24616363644599915, + "learning_rate": 8.861509653526083e-05, + "loss": 0.0157, "step": 15490 }, { - "grad_norm": 0.18705320358276367, - "learning_rate": 1.3219404292362065e-05, - "loss": 0.0077, + "grad_norm": 0.32043564319610596, + "learning_rate": 8.859758440390993e-05, + "loss": 0.0135, "step": 15500 }, { - "grad_norm": 0.1857132464647293, - "learning_rate": 1.3163451307972751e-05, - "loss": 0.0067, + "grad_norm": 0.28351980447769165, + "learning_rate": 8.858006054763202e-05, + "loss": 0.012, "step": 15510 }, { - "grad_norm": 0.2010364979505539, - "learning_rate": 1.3107599033300977e-05, - "loss": 0.006, + "grad_norm": 0.27427908778190613, + "learning_rate": 8.856252497175035e-05, + "loss": 0.011, "step": 15520 }, { - "grad_norm": 0.1669953614473343, - "learning_rate": 1.305184762104471e-05, - "loss": 0.006, + "grad_norm": 0.22871659696102142, + "learning_rate": 8.854497768159178e-05, + "loss": 0.0104, "step": 15530 }, { - "grad_norm": 0.16274145245552063, - "learning_rate": 1.2996197223626178e-05, - "loss": 0.0041, + "grad_norm": 0.2508438229560852, + "learning_rate": 8.852741868248671e-05, + "loss": 0.0096, "step": 15540 }, { - "grad_norm": 0.18755671381950378, - "learning_rate": 1.2940647993191457e-05, - "loss": 0.0053, + "grad_norm": 0.27534446120262146, + "learning_rate": 8.85098479797691e-05, + "loss": 0.0114, "step": 15550 }, { - "grad_norm": 0.20959579944610596, - "learning_rate": 1.2885200081610005e-05, - "loss": 0.005, + "grad_norm": 0.2810802757740021, + "learning_rate": 8.849226557877646e-05, + "loss": 0.0092, "step": 15560 }, { - "grad_norm": 0.14389164745807648, - "learning_rate": 1.2829853640474316e-05, - "loss": 0.0045, + "grad_norm": 0.27636170387268066, + "learning_rate": 8.84746714848499e-05, + "loss": 0.0095, "step": 15570 }, { - "grad_norm": 0.15106278657913208, - "learning_rate": 1.2774608821099438e-05, - "loss": 0.0044, + "grad_norm": 0.2919359505176544, + "learning_rate": 8.845706570333397e-05, + "loss": 0.0085, "step": 15580 }, { - "grad_norm": 0.1607772409915924, - "learning_rate": 1.2719465774522577e-05, - "loss": 0.0055, + "grad_norm": 0.27650538086891174, + "learning_rate": 8.84394482395769e-05, + "loss": 0.0112, "step": 15590 }, { - "grad_norm": 0.1950947344303131, - "learning_rate": 1.2664424651502755e-05, - "loss": 0.0064, + "grad_norm": 0.21103999018669128, + "learning_rate": 8.842181909893038e-05, + "loss": 0.0111, "step": 15600 }, { - "grad_norm": 0.20571626722812653, - "learning_rate": 1.260948560252026e-05, - "loss": 0.0054, + "grad_norm": 0.27054905891418457, + "learning_rate": 8.840417828674969e-05, + "loss": 0.0105, "step": 15610 }, { - "grad_norm": 0.1459115594625473, - "learning_rate": 1.2554648777776396e-05, - "loss": 0.0048, + "grad_norm": 0.24890156090259552, + "learning_rate": 8.838652580839364e-05, + "loss": 0.0087, "step": 15620 }, { - "grad_norm": 0.1424199789762497, - "learning_rate": 1.2499914327192919e-05, - "loss": 0.0056, + "grad_norm": 0.3368520736694336, + "learning_rate": 8.836886166922458e-05, + "loss": 0.0094, "step": 15630 }, { - "grad_norm": 0.13897298276424408, - "learning_rate": 1.2445282400411722e-05, - "loss": 0.0055, + "grad_norm": 0.27556174993515015, + "learning_rate": 8.835118587460844e-05, + "loss": 0.0106, "step": 15640 }, { - "grad_norm": 0.13337059319019318, - "learning_rate": 1.2390753146794437e-05, - "loss": 0.0052, + "grad_norm": 0.2553868889808655, + "learning_rate": 8.83334984299146e-05, + "loss": 0.0101, "step": 15650 }, { - "grad_norm": 0.22809644043445587, - "learning_rate": 1.2336326715421925e-05, - "loss": 0.0055, + "grad_norm": 0.2568078637123108, + "learning_rate": 8.83157993405161e-05, + "loss": 0.0111, "step": 15660 }, { - "grad_norm": 0.13422560691833496, - "learning_rate": 1.2282003255094005e-05, - "loss": 0.0051, + "grad_norm": 0.28967228531837463, + "learning_rate": 8.829808861178943e-05, + "loss": 0.0091, "step": 15670 }, { - "grad_norm": 0.18627096712589264, - "learning_rate": 1.2227782914328928e-05, - "loss": 0.0052, + "grad_norm": 0.32905226945877075, + "learning_rate": 8.828036624911464e-05, + "loss": 0.0104, "step": 15680 }, { - "grad_norm": 0.18976987898349762, - "learning_rate": 1.2173665841363018e-05, - "loss": 0.0044, + "grad_norm": 0.30181950330734253, + "learning_rate": 8.826263225787532e-05, + "loss": 0.0097, "step": 15690 }, { - "grad_norm": 0.16836141049861908, - "learning_rate": 1.211965218415032e-05, - "loss": 0.0041, + "grad_norm": 0.28712335228919983, + "learning_rate": 8.824488664345858e-05, + "loss": 0.0085, "step": 15700 }, { - "grad_norm": 0.14203710854053497, - "learning_rate": 1.2065742090362082e-05, - "loss": 0.0055, + "grad_norm": 0.23175868391990662, + "learning_rate": 8.822712941125508e-05, + "loss": 0.0096, "step": 15710 }, { - "grad_norm": 0.16482926905155182, - "learning_rate": 1.2011935707386457e-05, - "loss": 0.0055, + "grad_norm": 0.21660515666007996, + "learning_rate": 8.820936056665898e-05, + "loss": 0.0091, "step": 15720 }, { - "grad_norm": 0.23353098332881927, - "learning_rate": 1.1958233182328044e-05, - "loss": 0.0052, + "grad_norm": 0.3418399691581726, + "learning_rate": 8.819158011506801e-05, + "loss": 0.0107, "step": 15730 }, { - "grad_norm": 0.2196880280971527, - "learning_rate": 1.1904634662007474e-05, - "loss": 0.0052, + "grad_norm": 0.32794296741485596, + "learning_rate": 8.81737880618834e-05, + "loss": 0.0107, "step": 15740 }, { - "grad_norm": 0.20127354562282562, - "learning_rate": 1.1851140292961088e-05, - "loss": 0.0052, + "grad_norm": 0.2936713397502899, + "learning_rate": 8.815598441250987e-05, + "loss": 0.0093, "step": 15750 }, { - "grad_norm": 0.19632349908351898, - "learning_rate": 1.1797750221440424e-05, - "loss": 0.006, + "grad_norm": 0.2799868583679199, + "learning_rate": 8.813816917235576e-05, + "loss": 0.0113, "step": 15760 }, { - "grad_norm": 0.22671037912368774, - "learning_rate": 1.1744464593411897e-05, - "loss": 0.0061, + "grad_norm": 0.3164301812648773, + "learning_rate": 8.812034234683282e-05, + "loss": 0.0115, "step": 15770 }, { - "grad_norm": 0.1899847388267517, - "learning_rate": 1.1691283554556399e-05, - "loss": 0.005, + "grad_norm": 0.3140501081943512, + "learning_rate": 8.810250394135637e-05, + "loss": 0.0098, "step": 15780 }, { - "grad_norm": 0.2320375293493271, - "learning_rate": 1.1638207250268834e-05, - "loss": 0.0048, + "grad_norm": 0.2496100664138794, + "learning_rate": 8.808465396134529e-05, + "loss": 0.0097, "step": 15790 }, { - "grad_norm": 0.1803273856639862, - "learning_rate": 1.158523582565782e-05, - "loss": 0.0052, + "grad_norm": 0.2808854877948761, + "learning_rate": 8.806679241222189e-05, + "loss": 0.0101, "step": 15800 }, { - "grad_norm": 0.12426243722438812, - "learning_rate": 1.1532369425545192e-05, - "loss": 0.0037, + "grad_norm": 0.2741963565349579, + "learning_rate": 8.804891929941203e-05, + "loss": 0.0086, "step": 15810 }, { - "grad_norm": 0.15763874351978302, - "learning_rate": 1.1479608194465662e-05, - "loss": 0.0092, + "grad_norm": 0.2546410858631134, + "learning_rate": 8.803103462834514e-05, + "loss": 0.0137, "step": 15820 }, { - "grad_norm": 0.16793958842754364, - "learning_rate": 1.1426952276666442e-05, - "loss": 0.0044, + "grad_norm": 0.2717438042163849, + "learning_rate": 8.801313840445408e-05, + "loss": 0.0085, "step": 15830 }, { - "grad_norm": 0.18338541686534882, - "learning_rate": 1.1374401816106778e-05, - "loss": 0.0073, + "grad_norm": 0.3066229820251465, + "learning_rate": 8.799523063317524e-05, + "loss": 0.012, "step": 15840 }, { - "grad_norm": 0.16605933010578156, - "learning_rate": 1.1321956956457646e-05, - "loss": 0.005, + "grad_norm": 0.26512670516967773, + "learning_rate": 8.797731131994854e-05, + "loss": 0.01, "step": 15850 }, { - "grad_norm": 0.23071573674678802, - "learning_rate": 1.1269617841101277e-05, - "loss": 0.0058, + "grad_norm": 0.35922300815582275, + "learning_rate": 8.795938047021739e-05, + "loss": 0.0105, "step": 15860 }, { - "grad_norm": 0.16531772911548615, - "learning_rate": 1.1217384613130804e-05, - "loss": 0.0046, + "grad_norm": 0.2575354278087616, + "learning_rate": 8.794143808942872e-05, + "loss": 0.009, "step": 15870 }, { - "grad_norm": 0.17097795009613037, - "learning_rate": 1.11652574153499e-05, - "loss": 0.0056, + "grad_norm": 0.2672472596168518, + "learning_rate": 8.792348418303296e-05, + "loss": 0.0105, "step": 15880 }, { - "grad_norm": 0.14396753907203674, - "learning_rate": 1.1113236390272303e-05, - "loss": 0.0052, + "grad_norm": 0.2731088697910309, + "learning_rate": 8.790551875648398e-05, + "loss": 0.0107, "step": 15890 }, { - "grad_norm": 0.19578398764133453, - "learning_rate": 1.106132168012155e-05, - "loss": 0.0074, + "grad_norm": 0.3232043385505676, + "learning_rate": 8.788754181523926e-05, + "loss": 0.0123, "step": 15900 }, { - "grad_norm": 0.15604498982429504, - "learning_rate": 1.1009513426830448e-05, - "loss": 0.0057, + "grad_norm": 0.23219452798366547, + "learning_rate": 8.78695533647597e-05, + "loss": 0.0104, "step": 15910 }, { - "grad_norm": 0.1655254364013672, - "learning_rate": 1.0957811772040777e-05, - "loss": 0.0081, + "grad_norm": 0.24780035018920898, + "learning_rate": 8.785155341050972e-05, + "loss": 0.0136, "step": 15920 }, { - "grad_norm": 0.20122294127941132, - "learning_rate": 1.0906216857102913e-05, - "loss": 0.005, + "grad_norm": 0.29535433650016785, + "learning_rate": 8.783354195795721e-05, + "loss": 0.0099, "step": 15930 }, { - "grad_norm": 0.1938413679599762, - "learning_rate": 1.0854728823075355e-05, - "loss": 0.0055, + "grad_norm": 0.28305551409721375, + "learning_rate": 8.78155190125736e-05, + "loss": 0.0112, "step": 15940 }, { - "grad_norm": 0.14376655220985413, - "learning_rate": 1.0803347810724452e-05, - "loss": 0.0068, + "grad_norm": 0.2864108383655548, + "learning_rate": 8.779748457983378e-05, + "loss": 0.0126, "step": 15950 }, { - "grad_norm": 0.14233747124671936, - "learning_rate": 1.0752073960523911e-05, - "loss": 0.0051, + "grad_norm": 0.25819188356399536, + "learning_rate": 8.777943866521612e-05, + "loss": 0.0112, "step": 15960 }, { - "grad_norm": 0.14280758798122406, - "learning_rate": 1.070090741265447e-05, - "loss": 0.0047, + "grad_norm": 0.25041109323501587, + "learning_rate": 8.77613812742025e-05, + "loss": 0.0106, "step": 15970 }, { - "grad_norm": 0.14656981825828552, - "learning_rate": 1.0649848307003547e-05, - "loss": 0.0051, + "grad_norm": 0.2609696090221405, + "learning_rate": 8.774331241227829e-05, + "loss": 0.0106, "step": 15980 }, { - "grad_norm": 0.13522081077098846, - "learning_rate": 1.0598896783164757e-05, - "loss": 0.0049, + "grad_norm": 0.2358810156583786, + "learning_rate": 8.772523208493232e-05, + "loss": 0.0094, "step": 15990 }, { - "grad_norm": 0.12241239100694656, - "learning_rate": 1.0548052980437645e-05, - "loss": 0.0073, + "grad_norm": 0.25929760932922363, + "learning_rate": 8.770714029765692e-05, + "loss": 0.0125, "step": 16000 }, { - "grad_norm": 0.1648959517478943, - "learning_rate": 1.049731703782722e-05, - "loss": 0.0055, + "grad_norm": 0.3153972029685974, + "learning_rate": 8.768903705594789e-05, + "loss": 0.0108, "step": 16010 }, { - "grad_norm": 0.13855500519275665, - "learning_rate": 1.0446689094043587e-05, - "loss": 0.0073, + "grad_norm": 0.2483811378479004, + "learning_rate": 8.767092236530453e-05, + "loss": 0.0122, "step": 16020 }, { - "grad_norm": 0.1769500970840454, - "learning_rate": 1.039616928750165e-05, - "loss": 0.0049, + "grad_norm": 0.17887499928474426, + "learning_rate": 8.76527962312296e-05, + "loss": 0.0099, "step": 16030 }, { - "grad_norm": 0.15365423262119293, - "learning_rate": 1.0345757756320612e-05, - "loss": 0.0043, + "grad_norm": 0.22678126394748688, + "learning_rate": 8.763465865922934e-05, + "loss": 0.0093, "step": 16040 }, { - "grad_norm": 0.1871398538351059, - "learning_rate": 1.0295454638323666e-05, - "loss": 0.0055, + "grad_norm": 0.2637060284614563, + "learning_rate": 8.761650965481347e-05, + "loss": 0.0097, "step": 16050 }, { - "grad_norm": 0.22193288803100586, - "learning_rate": 1.0245260071037632e-05, - "loss": 0.0053, + "grad_norm": 0.27175331115722656, + "learning_rate": 8.759834922349516e-05, + "loss": 0.0102, "step": 16060 }, { - "grad_norm": 0.18220891058444977, - "learning_rate": 1.0195174191692518e-05, - "loss": 0.0043, + "grad_norm": 0.28748819231987, + "learning_rate": 8.758017737079108e-05, + "loss": 0.0083, "step": 16070 }, { - "grad_norm": 0.15648461878299713, - "learning_rate": 1.014519713722124e-05, - "loss": 0.0052, + "grad_norm": 0.3020229935646057, + "learning_rate": 8.756199410222137e-05, + "loss": 0.0105, "step": 16080 }, { - "grad_norm": 0.18102756142616272, - "learning_rate": 1.0095329044259132e-05, - "loss": 0.0043, + "grad_norm": 0.2474466860294342, + "learning_rate": 8.754379942330963e-05, + "loss": 0.0091, "step": 16090 }, { - "grad_norm": 0.14919497072696686, - "learning_rate": 1.004557004914365e-05, - "loss": 0.005, + "grad_norm": 0.3583427965641022, + "learning_rate": 8.75255933395829e-05, + "loss": 0.0094, "step": 16100 }, { - "grad_norm": 0.17376764118671417, - "learning_rate": 9.995920287914007e-06, - "loss": 0.0042, + "grad_norm": 0.3202165365219116, + "learning_rate": 8.750737585657171e-05, + "loss": 0.01, "step": 16110 }, { - "grad_norm": 0.13198114931583405, - "learning_rate": 9.946379896310737e-06, - "loss": 0.004, + "grad_norm": 0.28818318247795105, + "learning_rate": 8.748914697981008e-05, + "loss": 0.0089, "step": 16120 }, { - "grad_norm": 0.16041940450668335, - "learning_rate": 9.896949009775396e-06, - "loss": 0.0048, + "grad_norm": 0.306518018245697, + "learning_rate": 8.747090671483542e-05, + "loss": 0.0102, "step": 16130 }, { - "grad_norm": 0.20756278932094574, - "learning_rate": 9.847627763450134e-06, - "loss": 0.0049, + "grad_norm": 0.2634550929069519, + "learning_rate": 8.745265506718869e-05, + "loss": 0.0101, "step": 16140 }, { - "grad_norm": 0.22414100170135498, - "learning_rate": 9.798416292177337e-06, - "loss": 0.0045, + "grad_norm": 0.2795845568180084, + "learning_rate": 8.74343920424142e-05, + "loss": 0.0101, "step": 16150 }, { - "grad_norm": 0.22440825402736664, - "learning_rate": 9.74931473049932e-06, - "loss": 0.0054, + "grad_norm": 0.25344061851501465, + "learning_rate": 8.741611764605982e-05, + "loss": 0.0107, "step": 16160 }, { - "grad_norm": 0.1703796684741974, - "learning_rate": 9.700323212657847e-06, - "loss": 0.0042, + "grad_norm": 0.2501373291015625, + "learning_rate": 8.739783188367682e-05, + "loss": 0.0088, "step": 16170 }, { - "grad_norm": 0.15423569083213806, - "learning_rate": 9.65144187259388e-06, - "loss": 0.0053, + "grad_norm": 0.2246922254562378, + "learning_rate": 8.737953476081991e-05, + "loss": 0.0099, "step": 16180 }, { - "grad_norm": 0.16278226673603058, - "learning_rate": 9.602670843947132e-06, - "loss": 0.0049, + "grad_norm": 0.30105143785476685, + "learning_rate": 8.73612262830473e-05, + "loss": 0.0096, "step": 16190 }, { - "grad_norm": 0.1522931307554245, - "learning_rate": 9.554010260055713e-06, - "loss": 0.0046, + "grad_norm": 0.22231817245483398, + "learning_rate": 8.734290645592061e-05, + "loss": 0.0089, "step": 16200 }, { - "grad_norm": 0.1324155330657959, - "learning_rate": 9.505460253955834e-06, - "loss": 0.0044, + "grad_norm": 0.3406302034854889, + "learning_rate": 8.732457528500493e-05, + "loss": 0.0096, "step": 16210 }, { - "grad_norm": 0.12835343182086945, - "learning_rate": 9.457020958381324e-06, - "loss": 0.0055, + "grad_norm": 0.2609446048736572, + "learning_rate": 8.730623277586875e-05, + "loss": 0.0104, "step": 16220 }, { - "grad_norm": 0.13485650718212128, - "learning_rate": 9.408692505763395e-06, - "loss": 0.0048, + "grad_norm": 0.33388611674308777, + "learning_rate": 8.72878789340841e-05, + "loss": 0.0094, "step": 16230 }, { - "grad_norm": 0.17823177576065063, - "learning_rate": 9.360475028230181e-06, - "loss": 0.0051, + "grad_norm": 0.25460466742515564, + "learning_rate": 8.726951376522635e-05, + "loss": 0.0096, "step": 16240 }, { - "grad_norm": 0.13854046165943146, - "learning_rate": 9.312368657606412e-06, - "loss": 0.0038, + "grad_norm": 0.21380110085010529, + "learning_rate": 8.725113727487435e-05, + "loss": 0.0079, "step": 16250 }, { - "grad_norm": 0.20396800339221954, - "learning_rate": 9.264373525413096e-06, - "loss": 0.0072, + "grad_norm": 0.3228558599948883, + "learning_rate": 8.723274946861042e-05, + "loss": 0.0125, "step": 16260 }, { - "grad_norm": 0.13611064851284027, - "learning_rate": 9.216489762867058e-06, - "loss": 0.0042, + "grad_norm": 0.27122434973716736, + "learning_rate": 8.721435035202026e-05, + "loss": 0.0091, "step": 16270 }, { - "grad_norm": 0.23162850737571716, - "learning_rate": 9.168717500880708e-06, - "loss": 0.0068, + "grad_norm": 0.4003918468952179, + "learning_rate": 8.719593993069306e-05, + "loss": 0.0121, "step": 16280 }, { - "grad_norm": 0.14248082041740417, - "learning_rate": 9.121056870061574e-06, - "loss": 0.0035, + "grad_norm": 0.22706764936447144, + "learning_rate": 8.717751821022139e-05, + "loss": 0.0079, "step": 16290 }, { - "grad_norm": 0.20380237698554993, - "learning_rate": 9.073508000711983e-06, - "loss": 0.0048, + "grad_norm": 0.2985023856163025, + "learning_rate": 8.715908519620134e-05, + "loss": 0.009, "step": 16300 }, { - "grad_norm": 0.18490472435951233, - "learning_rate": 9.026071022828758e-06, - "loss": 0.0062, + "grad_norm": 0.285762757062912, + "learning_rate": 8.71406408942323e-05, + "loss": 0.011, "step": 16310 }, { - "grad_norm": 0.22372393310070038, - "learning_rate": 8.978746066102771e-06, - "loss": 0.0049, + "grad_norm": 0.3188495934009552, + "learning_rate": 8.712218530991723e-05, + "loss": 0.0107, "step": 16320 }, { - "grad_norm": 0.15033039450645447, - "learning_rate": 8.931533259918634e-06, - "loss": 0.0037, + "grad_norm": 0.20482321083545685, + "learning_rate": 8.710371844886241e-05, + "loss": 0.0083, "step": 16330 }, { - "grad_norm": 0.13687333464622498, - "learning_rate": 8.884432733354382e-06, - "loss": 0.004, + "grad_norm": 0.23664060235023499, + "learning_rate": 8.708524031667758e-05, + "loss": 0.0085, "step": 16340 }, { - "grad_norm": 0.15738479793071747, - "learning_rate": 8.837444615181029e-06, - "loss": 0.005, + "grad_norm": 0.40683987736701965, + "learning_rate": 8.706675091897592e-05, + "loss": 0.0106, "step": 16350 }, { - "grad_norm": 0.2072679102420807, - "learning_rate": 8.790569033862323e-06, - "loss": 0.0062, + "grad_norm": 0.22715559601783752, + "learning_rate": 8.704825026137404e-05, + "loss": 0.0114, "step": 16360 }, { - "grad_norm": 0.19508348405361176, - "learning_rate": 8.7438061175543e-06, - "loss": 0.0039, + "grad_norm": 0.21525990962982178, + "learning_rate": 8.702973834949192e-05, + "loss": 0.008, "step": 16370 }, { - "grad_norm": 0.18191103637218475, - "learning_rate": 8.697155994104978e-06, - "loss": 0.0037, + "grad_norm": 0.25327223539352417, + "learning_rate": 8.701121518895301e-05, + "loss": 0.0075, "step": 16380 }, { - "grad_norm": 0.16201141476631165, - "learning_rate": 8.650618791054033e-06, - "loss": 0.0046, + "grad_norm": 0.30771520733833313, + "learning_rate": 8.699268078538414e-05, + "loss": 0.0083, "step": 16390 }, { - "grad_norm": 0.1699906885623932, - "learning_rate": 8.604194635632373e-06, - "loss": 0.0056, + "grad_norm": 0.28778135776519775, + "learning_rate": 8.69741351444156e-05, + "loss": 0.0109, "step": 16400 }, { - "grad_norm": 0.14311489462852478, - "learning_rate": 8.557883654761906e-06, - "loss": 0.0051, + "grad_norm": 0.3142705261707306, + "learning_rate": 8.695557827168101e-05, + "loss": 0.0101, "step": 16410 }, { - "grad_norm": 0.17263031005859375, - "learning_rate": 8.511685975055061e-06, - "loss": 0.0052, + "grad_norm": 0.26974770426750183, + "learning_rate": 8.693701017281753e-05, + "loss": 0.0104, "step": 16420 }, { - "grad_norm": 0.15772181749343872, - "learning_rate": 8.46560172281452e-06, - "loss": 0.0038, + "grad_norm": 0.2427842766046524, + "learning_rate": 8.691843085346563e-05, + "loss": 0.0084, "step": 16430 }, { - "grad_norm": 0.1426163911819458, - "learning_rate": 8.419631024032893e-06, - "loss": 0.004, + "grad_norm": 0.23732158541679382, + "learning_rate": 8.689984031926919e-05, + "loss": 0.0081, "step": 16440 }, { - "grad_norm": 0.15554185211658478, - "learning_rate": 8.373774004392293e-06, - "loss": 0.0048, + "grad_norm": 0.27599337697029114, + "learning_rate": 8.688123857587555e-05, + "loss": 0.0093, "step": 16450 }, { - "grad_norm": 0.16801729798316956, - "learning_rate": 8.32803078926409e-06, - "loss": 0.0058, + "grad_norm": 0.26685118675231934, + "learning_rate": 8.686262562893544e-05, + "loss": 0.0098, "step": 16460 }, { - "grad_norm": 0.15965138375759125, - "learning_rate": 8.282401503708454e-06, - "loss": 0.0044, + "grad_norm": 0.2499106228351593, + "learning_rate": 8.684400148410294e-05, + "loss": 0.0078, "step": 16470 }, { - "grad_norm": 0.1287917047739029, - "learning_rate": 8.23688627247412e-06, - "loss": 0.0051, + "grad_norm": 0.21405087411403656, + "learning_rate": 8.682536614703562e-05, + "loss": 0.0094, "step": 16480 }, { - "grad_norm": 0.16028323769569397, - "learning_rate": 8.191485219998007e-06, - "loss": 0.0042, + "grad_norm": 0.24212081730365753, + "learning_rate": 8.680671962339437e-05, + "loss": 0.0086, "step": 16490 }, { - "grad_norm": 0.11672808229923248, - "learning_rate": 8.146198470404843e-06, - "loss": 0.0046, + "grad_norm": 0.26402127742767334, + "learning_rate": 8.678806191884352e-05, + "loss": 0.0091, "step": 16500 }, { - "grad_norm": 0.13132800161838531, - "learning_rate": 8.101026147506897e-06, - "loss": 0.003, + "grad_norm": 0.23224511742591858, + "learning_rate": 8.67693930390508e-05, + "loss": 0.0064, "step": 16510 }, { - "grad_norm": 0.1448666900396347, - "learning_rate": 8.05596837480353e-06, - "loss": 0.0053, + "grad_norm": 0.25720784068107605, + "learning_rate": 8.67507129896873e-05, + "loss": 0.0095, "step": 16520 }, { - "grad_norm": 0.13154615461826324, - "learning_rate": 8.011025275480998e-06, - "loss": 0.0051, + "grad_norm": 0.2127181887626648, + "learning_rate": 8.673202177642757e-05, + "loss": 0.0094, "step": 16530 }, { - "grad_norm": 0.1266883760690689, - "learning_rate": 7.966196972412027e-06, - "loss": 0.0037, + "grad_norm": 0.21197263896465302, + "learning_rate": 8.671331940494945e-05, + "loss": 0.0081, "step": 16540 }, { - "grad_norm": 0.1132773905992508, - "learning_rate": 7.92148358815547e-06, - "loss": 0.0034, + "grad_norm": 0.2022964358329773, + "learning_rate": 8.669460588093427e-05, + "loss": 0.0065, "step": 16550 }, { - "grad_norm": 0.16018636524677277, - "learning_rate": 7.87688524495604e-06, - "loss": 0.0049, + "grad_norm": 0.24868224561214447, + "learning_rate": 8.667588121006667e-05, + "loss": 0.0097, "step": 16560 }, { - "grad_norm": 0.1342909187078476, - "learning_rate": 7.83240206474386e-06, - "loss": 0.0054, + "grad_norm": 0.27824923396110535, + "learning_rate": 8.665714539803475e-05, + "loss": 0.0103, "step": 16570 }, { - "grad_norm": 0.12503987550735474, - "learning_rate": 7.788034169134272e-06, - "loss": 0.0043, + "grad_norm": 0.2965802550315857, + "learning_rate": 8.663839845052993e-05, + "loss": 0.0084, "step": 16580 }, { - "grad_norm": 0.11737470328807831, - "learning_rate": 7.743781679427414e-06, - "loss": 0.0043, + "grad_norm": 0.21202483773231506, + "learning_rate": 8.661964037324703e-05, + "loss": 0.009, "step": 16590 }, { - "grad_norm": 0.11028105020523071, - "learning_rate": 7.699644716607895e-06, - "loss": 0.0038, + "grad_norm": 0.20447774231433868, + "learning_rate": 8.660087117188427e-05, + "loss": 0.0085, "step": 16600 }, { - "grad_norm": 0.09312699735164642, - "learning_rate": 7.655623401344486e-06, - "loss": 0.0036, + "grad_norm": 0.2292976975440979, + "learning_rate": 8.658209085214325e-05, + "loss": 0.0074, "step": 16610 }, { - "grad_norm": 0.1259354054927826, - "learning_rate": 7.611717853989775e-06, - "loss": 0.0045, + "grad_norm": 0.3335595428943634, + "learning_rate": 8.656329941972891e-05, + "loss": 0.0083, "step": 16620 }, { - "grad_norm": 0.14075908064842224, - "learning_rate": 7.567928194579854e-06, - "loss": 0.0058, + "grad_norm": 0.2836003303527832, + "learning_rate": 8.654449688034963e-05, + "loss": 0.0102, "step": 16630 }, { - "grad_norm": 0.16171559691429138, - "learning_rate": 7.524254542833997e-06, - "loss": 0.0049, + "grad_norm": 0.3065835237503052, + "learning_rate": 8.652568323971706e-05, + "loss": 0.0092, "step": 16640 }, { - "grad_norm": 0.1568921059370041, - "learning_rate": 7.480697018154286e-06, - "loss": 0.009, + "grad_norm": 0.24776628613471985, + "learning_rate": 8.650685850354636e-05, + "loss": 0.0153, "step": 16650 }, { - "grad_norm": 0.14268898963928223, - "learning_rate": 7.437255739625332e-06, - "loss": 0.0036, + "grad_norm": 0.3175615072250366, + "learning_rate": 8.648802267755593e-05, + "loss": 0.0096, "step": 16660 }, { - "grad_norm": 0.2208922803401947, - "learning_rate": 7.393930826013923e-06, - "loss": 0.0064, + "grad_norm": 0.23557539284229279, + "learning_rate": 8.646917576746764e-05, + "loss": 0.0143, "step": 16670 }, { - "grad_norm": 0.1548379510641098, - "learning_rate": 7.350722395768722e-06, - "loss": 0.0043, + "grad_norm": 0.272316038608551, + "learning_rate": 8.645031777900666e-05, + "loss": 0.01, "step": 16680 }, { - "grad_norm": 0.1384071558713913, - "learning_rate": 7.307630567019963e-06, - "loss": 0.0036, + "grad_norm": 0.34998440742492676, + "learning_rate": 8.643144871790154e-05, + "loss": 0.0106, "step": 16690 }, { - "grad_norm": 0.10652858018875122, - "learning_rate": 7.264655457579e-06, - "loss": 0.0041, + "grad_norm": 0.3412337899208069, + "learning_rate": 8.641256858988424e-05, + "loss": 0.0109, "step": 16700 }, { - "grad_norm": 0.10340376943349838, - "learning_rate": 7.221797184938184e-06, - "loss": 0.0059, + "grad_norm": 0.1802694946527481, + "learning_rate": 8.639367740069e-05, + "loss": 0.0111, "step": 16710 }, { - "grad_norm": 0.11764318495988846, - "learning_rate": 7.179055866270373e-06, - "loss": 0.0036, + "grad_norm": 0.2797043025493622, + "learning_rate": 8.63747751560575e-05, + "loss": 0.0089, "step": 16720 }, { - "grad_norm": 0.21147853136062622, - "learning_rate": 7.136431618428707e-06, - "loss": 0.0045, + "grad_norm": 0.32260340452194214, + "learning_rate": 8.635586186172871e-05, + "loss": 0.0099, "step": 16730 }, { - "grad_norm": 0.19434772431850433, - "learning_rate": 7.09392455794628e-06, - "loss": 0.0062, + "grad_norm": 0.278450071811676, + "learning_rate": 8.633693752344902e-05, + "loss": 0.0119, "step": 16740 }, { - "grad_norm": 0.14892810583114624, - "learning_rate": 7.051534801035725e-06, - "loss": 0.0053, + "grad_norm": 0.25058600306510925, + "learning_rate": 8.631800214696713e-05, + "loss": 0.0115, "step": 16750 }, { - "grad_norm": 0.17711889743804932, - "learning_rate": 7.00926246358905e-06, - "loss": 0.0055, + "grad_norm": 0.2945828139781952, + "learning_rate": 8.629905573803511e-05, + "loss": 0.0107, "step": 16760 }, { - "grad_norm": 0.13662099838256836, - "learning_rate": 6.967107661177191e-06, - "loss": 0.0056, + "grad_norm": 0.28727853298187256, + "learning_rate": 8.628009830240839e-05, + "loss": 0.0113, "step": 16770 }, { - "grad_norm": 0.13581331074237823, - "learning_rate": 6.925070509049786e-06, - "loss": 0.0053, + "grad_norm": 0.2422596514225006, + "learning_rate": 8.626112984584571e-05, + "loss": 0.0112, "step": 16780 }, { - "grad_norm": 0.09662957489490509, - "learning_rate": 6.883151122134812e-06, - "loss": 0.0037, + "grad_norm": 0.286075234413147, + "learning_rate": 8.62421503741092e-05, + "loss": 0.009, "step": 16790 }, { - "grad_norm": 0.14342249929904938, - "learning_rate": 6.8413496150382394e-06, - "loss": 0.0047, + "grad_norm": 0.2758397161960602, + "learning_rate": 8.622315989296432e-05, + "loss": 0.0102, "step": 16800 }, { - "grad_norm": 0.07608351111412048, - "learning_rate": 6.7996661020438165e-06, - "loss": 0.0029, + "grad_norm": 0.23855245113372803, + "learning_rate": 8.62041584081799e-05, + "loss": 0.0066, "step": 16810 }, { - "grad_norm": 0.14846809208393097, - "learning_rate": 6.758100697112662e-06, - "loss": 0.0038, + "grad_norm": 0.21355026960372925, + "learning_rate": 8.618514592552807e-05, + "loss": 0.0085, "step": 16820 }, { - "grad_norm": 0.18177205324172974, - "learning_rate": 6.716653513883026e-06, - "loss": 0.005, + "grad_norm": 0.29938891530036926, + "learning_rate": 8.616612245078431e-05, + "loss": 0.0095, "step": 16830 }, { - "grad_norm": 0.15042972564697266, - "learning_rate": 6.675324665669913e-06, - "loss": 0.0039, + "grad_norm": 0.2483750730752945, + "learning_rate": 8.614708798972746e-05, + "loss": 0.0086, "step": 16840 }, { - "grad_norm": 0.17908424139022827, - "learning_rate": 6.634114265464803e-06, - "loss": 0.0056, + "grad_norm": 0.25764375925064087, + "learning_rate": 8.61280425481397e-05, + "loss": 0.0103, "step": 16850 }, { - "grad_norm": 0.1749325394630432, - "learning_rate": 6.59302242593538e-06, - "loss": 0.0041, + "grad_norm": 0.20467785000801086, + "learning_rate": 8.61089861318065e-05, + "loss": 0.0078, "step": 16860 }, { - "grad_norm": 0.11305996030569077, - "learning_rate": 6.552049259425141e-06, - "loss": 0.0042, + "grad_norm": 0.18834789097309113, + "learning_rate": 8.608991874651673e-05, + "loss": 0.0085, "step": 16870 }, { - "grad_norm": 0.11477645486593246, - "learning_rate": 6.511194877953181e-06, - "loss": 0.0046, + "grad_norm": 0.20003071427345276, + "learning_rate": 8.607084039806255e-05, + "loss": 0.0084, "step": 16880 }, { - "grad_norm": 0.1085256040096283, - "learning_rate": 6.470459393213813e-06, - "loss": 0.0046, + "grad_norm": 0.19845867156982422, + "learning_rate": 8.605175109223944e-05, + "loss": 0.0082, "step": 16890 }, { - "grad_norm": 0.16111595928668976, - "learning_rate": 6.429842916576279e-06, - "loss": 0.0049, + "grad_norm": 0.27606236934661865, + "learning_rate": 8.603265083484624e-05, + "loss": 0.0097, "step": 16900 }, { - "grad_norm": 0.1468374878168106, - "learning_rate": 6.389345559084503e-06, - "loss": 0.0036, + "grad_norm": 0.28678980469703674, + "learning_rate": 8.60135396316851e-05, + "loss": 0.0078, "step": 16910 }, { - "grad_norm": 0.17139410972595215, - "learning_rate": 6.348967431456682e-06, - "loss": 0.0064, + "grad_norm": 0.28309038281440735, + "learning_rate": 8.599441748856152e-05, + "loss": 0.0125, "step": 16920 }, { - "grad_norm": 0.17251119017601013, - "learning_rate": 6.30870864408511e-06, - "loss": 0.0036, + "grad_norm": 0.3022243082523346, + "learning_rate": 8.597528441128427e-05, + "loss": 0.0089, "step": 16930 }, { - "grad_norm": 0.14736191928386688, - "learning_rate": 6.268569307035754e-06, - "loss": 0.0041, + "grad_norm": 0.23140695691108704, + "learning_rate": 8.595614040566549e-05, + "loss": 0.009, "step": 16940 }, { - "grad_norm": 0.1500120311975479, - "learning_rate": 6.228549530048022e-06, - "loss": 0.0058, + "grad_norm": 0.2016071230173111, + "learning_rate": 8.593698547752063e-05, + "loss": 0.0106, "step": 16950 }, { - "grad_norm": 0.14884720742702484, - "learning_rate": 6.1886494225344814e-06, - "loss": 0.0064, + "grad_norm": 0.2225610762834549, + "learning_rate": 8.591781963266843e-05, + "loss": 0.0121, "step": 16960 }, { - "grad_norm": 0.15160712599754333, - "learning_rate": 6.148869093580479e-06, - "loss": 0.0056, + "grad_norm": 0.22754250466823578, + "learning_rate": 8.5898642876931e-05, + "loss": 0.0113, "step": 16970 }, { - "grad_norm": 0.1517210155725479, - "learning_rate": 6.109208651943921e-06, - "loss": 0.0047, + "grad_norm": 0.22588352859020233, + "learning_rate": 8.587945521613369e-05, + "loss": 0.0087, "step": 16980 }, { - "grad_norm": 0.155134379863739, - "learning_rate": 6.069668206054946e-06, - "loss": 0.0038, + "grad_norm": 0.2298949807882309, + "learning_rate": 8.586025665610524e-05, + "loss": 0.0078, "step": 16990 }, { - "grad_norm": 0.14380772411823273, - "learning_rate": 6.0302478640156145e-06, - "loss": 0.0053, + "grad_norm": 0.2997662425041199, + "learning_rate": 8.584104720267765e-05, + "loss": 0.0099, "step": 17000 }, { - "grad_norm": 0.16657549142837524, - "learning_rate": 5.990947733599644e-06, - "loss": 0.0068, + "grad_norm": 0.2864239513874054, + "learning_rate": 8.582182686168625e-05, + "loss": 0.0127, "step": 17010 }, { - "grad_norm": 0.16626840829849243, - "learning_rate": 5.951767922252105e-06, - "loss": 0.0059, + "grad_norm": 0.29647591710090637, + "learning_rate": 8.580259563896967e-05, + "loss": 0.0111, "step": 17020 }, { - "grad_norm": 0.14325283467769623, - "learning_rate": 5.912708537089068e-06, - "loss": 0.0039, + "grad_norm": 0.21788856387138367, + "learning_rate": 8.578335354036983e-05, + "loss": 0.0087, "step": 17030 }, { - "grad_norm": 0.15190692245960236, - "learning_rate": 5.873769684897434e-06, - "loss": 0.005, + "grad_norm": 0.23925253748893738, + "learning_rate": 8.576410057173201e-05, + "loss": 0.0097, "step": 17040 }, { - "grad_norm": 0.13617077469825745, - "learning_rate": 5.834951472134514e-06, - "loss": 0.0059, + "grad_norm": 0.4090352952480316, + "learning_rate": 8.574483673890474e-05, + "loss": 0.0105, "step": 17050 }, { - "grad_norm": 0.16800183057785034, - "learning_rate": 5.796254004927832e-06, - "loss": 0.0052, + "grad_norm": 0.2814415693283081, + "learning_rate": 8.572556204773983e-05, + "loss": 0.0105, "step": 17060 }, { - "grad_norm": 0.1636773645877838, - "learning_rate": 5.757677389074806e-06, - "loss": 0.0047, + "grad_norm": 0.24559694528579712, + "learning_rate": 8.570627650409246e-05, + "loss": 0.0093, "step": 17070 }, { - "grad_norm": 0.11355911195278168, - "learning_rate": 5.719221730042385e-06, - "loss": 0.0033, + "grad_norm": 0.3441891372203827, + "learning_rate": 8.568698011382107e-05, + "loss": 0.0075, "step": 17080 }, { - "grad_norm": 0.1289564073085785, - "learning_rate": 5.680887132966911e-06, - "loss": 0.0057, + "grad_norm": 0.32313841581344604, + "learning_rate": 8.566767288278738e-05, + "loss": 0.011, "step": 17090 }, { - "grad_norm": 0.14525209367275238, - "learning_rate": 5.642673702653683e-06, - "loss": 0.004, + "grad_norm": 0.2577977180480957, + "learning_rate": 8.56483548168564e-05, + "loss": 0.0084, "step": 17100 }, { - "grad_norm": 0.18314318358898163, - "learning_rate": 5.604581543576781e-06, - "loss": 0.0041, + "grad_norm": 0.2673834264278412, + "learning_rate": 8.562902592189648e-05, + "loss": 0.0084, "step": 17110 }, { - "grad_norm": 0.17528216540813446, - "learning_rate": 5.566610759878704e-06, - "loss": 0.0052, + "grad_norm": 0.2455013245344162, + "learning_rate": 8.560968620377921e-05, + "loss": 0.0097, "step": 17120 }, { - "grad_norm": 0.1443951427936554, - "learning_rate": 5.528761455370119e-06, - "loss": 0.0041, + "grad_norm": 0.17626313865184784, + "learning_rate": 8.559033566837951e-05, + "loss": 0.0091, "step": 17130 }, { - "grad_norm": 0.20061659812927246, - "learning_rate": 5.491033733529594e-06, - "loss": 0.0051, + "grad_norm": 0.306132048368454, + "learning_rate": 8.557097432157551e-05, + "loss": 0.0102, "step": 17140 }, { - "grad_norm": 0.12190601229667664, - "learning_rate": 5.453427697503255e-06, - "loss": 0.0049, + "grad_norm": 0.2101263701915741, + "learning_rate": 8.555160216924872e-05, + "loss": 0.0094, "step": 17150 }, { - "grad_norm": 0.13548479974269867, - "learning_rate": 5.415943450104599e-06, - "loss": 0.0047, + "grad_norm": 0.27102577686309814, + "learning_rate": 8.55322192172839e-05, + "loss": 0.0102, "step": 17160 }, { - "grad_norm": 0.15832222998142242, - "learning_rate": 5.378581093814111e-06, - "loss": 0.0037, + "grad_norm": 0.22342561185359955, + "learning_rate": 8.551282547156902e-05, + "loss": 0.0081, "step": 17170 }, { - "grad_norm": 0.10824442654848099, - "learning_rate": 5.3413407307790375e-06, - "loss": 0.0041, + "grad_norm": 0.19892767071723938, + "learning_rate": 8.549342093799544e-05, + "loss": 0.0086, "step": 17180 }, { - "grad_norm": 0.18595543503761292, - "learning_rate": 5.30422246281313e-06, - "loss": 0.0063, + "grad_norm": 0.24145124852657318, + "learning_rate": 8.547400562245773e-05, + "loss": 0.0115, "step": 17190 }, { - "grad_norm": 0.13622011244297028, - "learning_rate": 5.267226391396296e-06, - "loss": 0.0032, + "grad_norm": 0.27737998962402344, + "learning_rate": 8.545457953085374e-05, + "loss": 0.0078, "step": 17200 }, { - "grad_norm": 0.10188230872154236, - "learning_rate": 5.2303526176744e-06, - "loss": 0.003, + "grad_norm": 0.22603948414325714, + "learning_rate": 8.543514266908463e-05, + "loss": 0.0071, "step": 17210 }, { - "grad_norm": 0.13925378024578094, - "learning_rate": 5.193601242458929e-06, - "loss": 0.0043, + "grad_norm": 0.2690403461456299, + "learning_rate": 8.541569504305478e-05, + "loss": 0.0087, "step": 17220 }, { - "grad_norm": 0.11982742697000504, - "learning_rate": 5.156972366226714e-06, - "loss": 0.0058, + "grad_norm": 0.2842450737953186, + "learning_rate": 8.539623665867187e-05, + "loss": 0.011, "step": 17230 }, { - "grad_norm": 0.1453980952501297, - "learning_rate": 5.120466089119735e-06, - "loss": 0.0052, + "grad_norm": 0.39218488335609436, + "learning_rate": 8.537676752184685e-05, + "loss": 0.0109, "step": 17240 }, { - "grad_norm": 0.12326829135417938, - "learning_rate": 5.084082510944749e-06, - "loss": 0.0039, + "grad_norm": 0.2956484258174896, + "learning_rate": 8.53572876384939e-05, + "loss": 0.0086, "step": 17250 }, { - "grad_norm": 0.17726564407348633, - "learning_rate": 5.047821731173058e-06, - "loss": 0.0047, + "grad_norm": 0.23740901052951813, + "learning_rate": 8.533779701453056e-05, + "loss": 0.0105, "step": 17260 }, { - "grad_norm": 0.15553253889083862, - "learning_rate": 5.011683848940274e-06, - "loss": 0.0042, + "grad_norm": 0.3055787980556488, + "learning_rate": 8.53182956558775e-05, + "loss": 0.0095, "step": 17270 }, { - "grad_norm": 0.11963417381048203, - "learning_rate": 4.975668963045954e-06, - "loss": 0.004, + "grad_norm": 0.20200638473033905, + "learning_rate": 8.529878356845877e-05, + "loss": 0.0087, "step": 17280 }, { - "grad_norm": 0.12937189638614655, - "learning_rate": 4.9397771719534525e-06, - "loss": 0.0032, + "grad_norm": 0.24597354233264923, + "learning_rate": 8.527926075820158e-05, + "loss": 0.0079, "step": 17290 }, { - "grad_norm": 0.12897104024887085, - "learning_rate": 4.904008573789548e-06, - "loss": 0.0045, + "grad_norm": 0.22677156329154968, + "learning_rate": 8.525972723103648e-05, + "loss": 0.0094, "step": 17300 }, { - "grad_norm": 0.09982036799192429, - "learning_rate": 4.8683632663442005e-06, - "loss": 0.0035, + "grad_norm": 0.29029905796051025, + "learning_rate": 8.524018299289722e-05, + "loss": 0.0083, "step": 17310 }, { - "grad_norm": 0.15100431442260742, - "learning_rate": 4.832841347070343e-06, - "loss": 0.0047, + "grad_norm": 0.2062239795923233, + "learning_rate": 8.522062804972083e-05, + "loss": 0.009, "step": 17320 }, { - "grad_norm": 0.175477996468544, - "learning_rate": 4.797442913083539e-06, - "loss": 0.0059, + "grad_norm": 0.29048898816108704, + "learning_rate": 8.520106240744759e-05, + "loss": 0.0106, "step": 17330 }, { - "grad_norm": 0.14300963282585144, - "learning_rate": 4.7621680611617596e-06, - "loss": 0.0038, + "grad_norm": 0.1675039380788803, + "learning_rate": 8.518148607202102e-05, + "loss": 0.0079, "step": 17340 }, { - "grad_norm": 0.2541049122810364, - "learning_rate": 4.727016887745095e-06, - "loss": 0.0053, + "grad_norm": 0.309238463640213, + "learning_rate": 8.51618990493879e-05, + "loss": 0.0106, "step": 17350 }, { - "grad_norm": 0.12595532834529877, - "learning_rate": 4.691989488935511e-06, - "loss": 0.0041, + "grad_norm": 0.20029228925704956, + "learning_rate": 8.514230134549823e-05, + "loss": 0.009, "step": 17360 }, { - "grad_norm": 0.11065539717674255, - "learning_rate": 4.657085960496588e-06, - "loss": 0.0036, + "grad_norm": 0.24057692289352417, + "learning_rate": 8.51226929663053e-05, + "loss": 0.0077, "step": 17370 }, { - "grad_norm": 0.09352098405361176, - "learning_rate": 4.6223063978532265e-06, - "loss": 0.0052, + "grad_norm": 0.2233814150094986, + "learning_rate": 8.51030739177656e-05, + "loss": 0.0095, "step": 17380 }, { - "grad_norm": 0.09857888519763947, - "learning_rate": 4.587650896091439e-06, - "loss": 0.0046, + "grad_norm": 0.21940739452838898, + "learning_rate": 8.508344420583889e-05, + "loss": 0.0097, "step": 17390 }, { - "grad_norm": 0.1510113924741745, - "learning_rate": 4.553119549958035e-06, - "loss": 0.0059, + "grad_norm": 0.21646778285503387, + "learning_rate": 8.506380383648816e-05, + "loss": 0.0106, "step": 17400 }, { - "grad_norm": 0.14610646665096283, - "learning_rate": 4.518712453860385e-06, - "loss": 0.0047, + "grad_norm": 0.2844802439212799, + "learning_rate": 8.504415281567963e-05, + "loss": 0.0093, "step": 17410 }, { - "grad_norm": 0.10412902384996414, - "learning_rate": 4.484429701866205e-06, - "loss": 0.0038, + "grad_norm": 0.28693559765815735, + "learning_rate": 8.502449114938275e-05, + "loss": 0.0091, "step": 17420 }, { - "grad_norm": 0.1347024291753769, - "learning_rate": 4.4502713877031975e-06, - "loss": 0.004, + "grad_norm": 0.21802087128162384, + "learning_rate": 8.500481884357025e-05, + "loss": 0.0095, "step": 17430 }, { - "grad_norm": 0.11108066141605377, - "learning_rate": 4.416237604758911e-06, - "loss": 0.0044, + "grad_norm": 0.3267443776130676, + "learning_rate": 8.498513590421801e-05, + "loss": 0.0098, "step": 17440 }, { - "grad_norm": 0.17151008546352386, - "learning_rate": 4.3823284460804025e-06, - "loss": 0.0054, + "grad_norm": 0.3616046905517578, + "learning_rate": 8.496544233730522e-05, + "loss": 0.0106, "step": 17450 }, { - "grad_norm": 0.1379305124282837, - "learning_rate": 4.348544004374011e-06, - "loss": 0.0033, + "grad_norm": 0.31874823570251465, + "learning_rate": 8.494573814881426e-05, + "loss": 0.0083, "step": 17460 }, { - "grad_norm": 0.1281946748495102, - "learning_rate": 4.314884372005123e-06, - "loss": 0.0049, + "grad_norm": 0.2720226049423218, + "learning_rate": 8.492602334473074e-05, + "loss": 0.0102, "step": 17470 }, { - "grad_norm": 0.15030136704444885, - "learning_rate": 4.281349640997867e-06, - "loss": 0.0044, + "grad_norm": 0.22996309399604797, + "learning_rate": 8.49062979310435e-05, + "loss": 0.0094, "step": 17480 }, { - "grad_norm": 0.11368031054735184, - "learning_rate": 4.247939903034942e-06, - "loss": 0.0045, + "grad_norm": 0.23823527991771698, + "learning_rate": 8.488656191374458e-05, + "loss": 0.0086, "step": 17490 }, { - "grad_norm": 0.10988582670688629, - "learning_rate": 4.214655249457284e-06, - "loss": 0.0043, + "grad_norm": 0.2258954793214798, + "learning_rate": 8.48668152988293e-05, + "loss": 0.0088, "step": 17500 }, { - "grad_norm": 0.13686129450798035, - "learning_rate": 4.181495771263855e-06, - "loss": 0.0047, + "grad_norm": 0.23459313809871674, + "learning_rate": 8.484705809229612e-05, + "loss": 0.009, "step": 17510 }, { - "grad_norm": 0.08308565616607666, - "learning_rate": 4.148461559111427e-06, - "loss": 0.003, + "grad_norm": 0.213528111577034, + "learning_rate": 8.482729030014677e-05, + "loss": 0.0067, "step": 17520 }, { - "grad_norm": 0.10827736556529999, - "learning_rate": 4.115552703314252e-06, - "loss": 0.0044, + "grad_norm": 0.2736048400402069, + "learning_rate": 8.48075119283862e-05, + "loss": 0.0091, "step": 17530 }, { - "grad_norm": 0.10387372225522995, - "learning_rate": 4.082769293843886e-06, - "loss": 0.0036, + "grad_norm": 0.22485287487506866, + "learning_rate": 8.478772298302254e-05, + "loss": 0.0083, "step": 17540 }, { - "grad_norm": 0.06729380041360855, - "learning_rate": 4.050111420328939e-06, - "loss": 0.0034, + "grad_norm": 0.26154983043670654, + "learning_rate": 8.476792347006716e-05, + "loss": 0.0072, "step": 17550 }, { - "grad_norm": 0.1021871566772461, - "learning_rate": 4.017579172054764e-06, - "loss": 0.0039, + "grad_norm": 0.2836498022079468, + "learning_rate": 8.474811339553462e-05, + "loss": 0.0084, "step": 17560 }, { - "grad_norm": 0.1491686850786209, - "learning_rate": 3.985172637963308e-06, - "loss": 0.0043, + "grad_norm": 0.2520183324813843, + "learning_rate": 8.47282927654427e-05, + "loss": 0.009, "step": 17570 }, { - "grad_norm": 0.14715075492858887, - "learning_rate": 3.952891906652784e-06, - "loss": 0.0036, + "grad_norm": 0.31152215600013733, + "learning_rate": 8.470846158581238e-05, + "loss": 0.0089, "step": 17580 }, { - "grad_norm": 0.10194284468889236, - "learning_rate": 3.920737066377478e-06, - "loss": 0.0051, + "grad_norm": 0.21570228040218353, + "learning_rate": 8.468861986266787e-05, + "loss": 0.0092, "step": 17590 }, { - "grad_norm": 0.09443909674882889, - "learning_rate": 3.888708205047509e-06, - "loss": 0.003, + "grad_norm": 0.21828025579452515, + "learning_rate": 8.466876760203654e-05, + "loss": 0.0071, "step": 17600 }, { - "grad_norm": 0.13160809874534607, - "learning_rate": 3.856805410228542e-06, - "loss": 0.0045, + "grad_norm": 0.25473111867904663, + "learning_rate": 8.464890480994898e-05, + "loss": 0.0085, "step": 17610 }, { - "grad_norm": 0.10030721873044968, - "learning_rate": 3.82502876914162e-06, - "loss": 0.0038, + "grad_norm": 0.2922549843788147, + "learning_rate": 8.462903149243899e-05, + "loss": 0.0079, "step": 17620 }, { - "grad_norm": 0.11953095346689224, - "learning_rate": 3.7933783686628586e-06, - "loss": 0.005, + "grad_norm": 0.28819790482521057, + "learning_rate": 8.460914765554357e-05, + "loss": 0.0095, "step": 17630 }, { - "grad_norm": 0.09695965051651001, - "learning_rate": 3.7618542953232306e-06, - "loss": 0.0046, + "grad_norm": 0.2571139335632324, + "learning_rate": 8.458925330530288e-05, + "loss": 0.0086, "step": 17640 }, { - "grad_norm": 0.07293650507926941, - "learning_rate": 3.7304566353083658e-06, - "loss": 0.0029, + "grad_norm": 0.2384311407804489, + "learning_rate": 8.456934844776032e-05, + "loss": 0.0071, "step": 17650 }, { - "grad_norm": 0.09447503089904785, - "learning_rate": 3.6991854744582555e-06, - "loss": 0.003, + "grad_norm": 0.19701038300991058, + "learning_rate": 8.454943308896246e-05, + "loss": 0.0062, "step": 17660 }, { - "grad_norm": 0.127981036901474, - "learning_rate": 3.6680408982670777e-06, - "loss": 0.0039, + "grad_norm": 0.22720931470394135, + "learning_rate": 8.452950723495905e-05, + "loss": 0.0086, "step": 17670 }, { - "grad_norm": 0.10943655669689178, - "learning_rate": 3.637022991882899e-06, - "loss": 0.0054, + "grad_norm": 0.24483327567577362, + "learning_rate": 8.450957089180303e-05, + "loss": 0.0109, "step": 17680 }, { - "grad_norm": 0.1618594229221344, - "learning_rate": 3.606131840107485e-06, - "loss": 0.0046, + "grad_norm": 0.2452455312013626, + "learning_rate": 8.448962406555055e-05, + "loss": 0.0095, "step": 17690 }, { - "grad_norm": 0.15760213136672974, - "learning_rate": 3.575367527396084e-06, - "loss": 0.0045, + "grad_norm": 0.36983492970466614, + "learning_rate": 8.446966676226093e-05, + "loss": 0.0101, "step": 17700 }, { - "grad_norm": 0.11885190010070801, - "learning_rate": 3.5447301378571386e-06, - "loss": 0.0034, + "grad_norm": 0.2618146538734436, + "learning_rate": 8.444969898799667e-05, + "loss": 0.0088, "step": 17710 }, { - "grad_norm": 0.10916793346405029, - "learning_rate": 3.514219755252113e-06, - "loss": 0.0054, + "grad_norm": 0.3063323199748993, + "learning_rate": 8.442972074882343e-05, + "loss": 0.0104, "step": 17720 }, { - "grad_norm": 0.11321078985929489, - "learning_rate": 3.4838364629952213e-06, - "loss": 0.0045, + "grad_norm": 0.34352055191993713, + "learning_rate": 8.44097320508101e-05, + "loss": 0.0089, "step": 17730 }, { - "grad_norm": 0.08594832569360733, - "learning_rate": 3.4535803441532123e-06, - "loss": 0.0066, + "grad_norm": 0.24822193384170532, + "learning_rate": 8.43897329000287e-05, + "loss": 0.0112, "step": 17740 }, { - "grad_norm": 0.13212276995182037, - "learning_rate": 3.4234514814451836e-06, - "loss": 0.0037, + "grad_norm": 0.2098926603794098, + "learning_rate": 8.436972330255448e-05, + "loss": 0.007, "step": 17750 }, { - "grad_norm": 0.1763709932565689, - "learning_rate": 3.393449957242273e-06, - "loss": 0.0076, + "grad_norm": 0.2527866065502167, + "learning_rate": 8.434970326446579e-05, + "loss": 0.0129, "step": 17760 }, { - "grad_norm": 0.13847731053829193, - "learning_rate": 3.363575853567524e-06, - "loss": 0.0043, + "grad_norm": 0.23664584755897522, + "learning_rate": 8.432967279184418e-05, + "loss": 0.0093, "step": 17770 }, { - "grad_norm": 0.1049138754606247, - "learning_rate": 3.3338292520955826e-06, - "loss": 0.0049, + "grad_norm": 0.2556487023830414, + "learning_rate": 8.430963189077441e-05, + "loss": 0.0093, "step": 17780 }, { - "grad_norm": 0.11970741301774979, - "learning_rate": 3.304210234152516e-06, - "loss": 0.0046, + "grad_norm": 0.21195469796657562, + "learning_rate": 8.428958056734437e-05, + "loss": 0.0092, "step": 17790 }, { - "grad_norm": 0.1296226680278778, - "learning_rate": 3.2747188807155993e-06, - "loss": 0.0038, + "grad_norm": 0.3128160238265991, + "learning_rate": 8.426951882764513e-05, + "loss": 0.009, "step": 17800 }, { - "grad_norm": 0.09861905127763748, - "learning_rate": 3.2453552724130643e-06, - "loss": 0.0053, + "grad_norm": 0.31145206093788147, + "learning_rate": 8.424944667777089e-05, + "loss": 0.0111, "step": 17810 }, { - "grad_norm": 0.12149041891098022, - "learning_rate": 3.216119489523889e-06, - "loss": 0.0047, + "grad_norm": 0.25404298305511475, + "learning_rate": 8.422936412381905e-05, + "loss": 0.0095, "step": 17820 }, { - "grad_norm": 0.12130892276763916, - "learning_rate": 3.1870116119775917e-06, - "loss": 0.0036, + "grad_norm": 0.2762012779712677, + "learning_rate": 8.420927117189017e-05, + "loss": 0.0079, "step": 17830 }, { - "grad_norm": 0.08184216916561127, - "learning_rate": 3.158031719353999e-06, - "loss": 0.0047, + "grad_norm": 0.2649284899234772, + "learning_rate": 8.418916782808795e-05, + "loss": 0.0091, "step": 17840 }, { - "grad_norm": 0.12849783897399902, - "learning_rate": 3.1291798908830273e-06, - "loss": 0.0037, + "grad_norm": 0.21584245562553406, + "learning_rate": 8.416905409851926e-05, + "loss": 0.0078, "step": 17850 }, { - "grad_norm": 0.0973939523100853, - "learning_rate": 3.1004562054444853e-06, - "loss": 0.004, + "grad_norm": 0.19599056243896484, + "learning_rate": 8.41489299892941e-05, + "loss": 0.0082, "step": 17860 }, { - "grad_norm": 0.10297653824090958, - "learning_rate": 3.071860741567806e-06, - "loss": 0.0056, + "grad_norm": 0.3028583228588104, + "learning_rate": 8.412879550652566e-05, + "loss": 0.0095, "step": 17870 }, { - "grad_norm": 0.1703691929578781, - "learning_rate": 3.04339357743193e-06, - "loss": 0.0036, + "grad_norm": 0.3360619843006134, + "learning_rate": 8.410865065633029e-05, + "loss": 0.0075, "step": 17880 }, { - "grad_norm": 0.09887855499982834, - "learning_rate": 3.0150547908649628e-06, - "loss": 0.0048, + "grad_norm": 0.26718395948410034, + "learning_rate": 8.408849544482742e-05, + "loss": 0.0099, "step": 17890 }, { - "grad_norm": 0.17256423830986023, - "learning_rate": 2.9868444593440957e-06, - "loss": 0.0044, + "grad_norm": 0.2591950297355652, + "learning_rate": 8.406832987813968e-05, + "loss": 0.009, "step": 17900 }, { - "grad_norm": 0.14721833169460297, - "learning_rate": 2.9587626599952846e-06, - "loss": 0.0052, + "grad_norm": 0.31409353017807007, + "learning_rate": 8.404815396239286e-05, + "loss": 0.0107, "step": 17910 }, { - "grad_norm": 0.10446737706661224, - "learning_rate": 2.930809469593082e-06, - "loss": 0.0058, + "grad_norm": 0.3196844756603241, + "learning_rate": 8.402796770371587e-05, + "loss": 0.011, "step": 17920 }, { - "grad_norm": 0.1434236317873001, - "learning_rate": 2.9029849645604733e-06, - "loss": 0.0038, + "grad_norm": 0.24811924993991852, + "learning_rate": 8.400777110824071e-05, + "loss": 0.0082, "step": 17930 }, { - "grad_norm": 0.1383984535932541, - "learning_rate": 2.8752892209685632e-06, - "loss": 0.0054, + "grad_norm": 0.2636547088623047, + "learning_rate": 8.398756418210263e-05, + "loss": 0.0106, "step": 17940 }, { - "grad_norm": 0.10035323351621628, - "learning_rate": 2.847722314536483e-06, - "loss": 0.0033, + "grad_norm": 0.20335087180137634, + "learning_rate": 8.396734693143993e-05, + "loss": 0.0075, "step": 17950 }, { - "grad_norm": 0.11949007958173752, - "learning_rate": 2.820284320631078e-06, - "loss": 0.0063, + "grad_norm": 0.2664903402328491, + "learning_rate": 8.39471193623941e-05, + "loss": 0.0115, "step": 17960 }, { - "grad_norm": 0.1266811639070511, - "learning_rate": 2.792975314266788e-06, - "loss": 0.0041, + "grad_norm": 0.21218733489513397, + "learning_rate": 8.392688148110974e-05, + "loss": 0.0092, "step": 17970 }, { - "grad_norm": 0.12894940376281738, - "learning_rate": 2.7657953701054007e-06, - "loss": 0.0037, + "grad_norm": 0.28692907094955444, + "learning_rate": 8.390663329373456e-05, + "loss": 0.0087, "step": 17980 }, { - "grad_norm": 0.08531206846237183, - "learning_rate": 2.7387445624558306e-06, - "loss": 0.0056, + "grad_norm": 0.28370440006256104, + "learning_rate": 8.388637480641944e-05, + "loss": 0.0101, "step": 17990 }, { - "grad_norm": 0.09096170216798782, - "learning_rate": 2.7118229652739747e-06, - "loss": 0.0036, + "grad_norm": 0.29673174023628235, + "learning_rate": 8.386610602531837e-05, + "loss": 0.0079, "step": 18000 }, { - "grad_norm": 0.13481144607067108, - "learning_rate": 2.6850306521624236e-06, - "loss": 0.0047, + "grad_norm": 0.22297346591949463, + "learning_rate": 8.384582695658847e-05, + "loss": 0.0092, "step": 18010 }, { - "grad_norm": 0.0847296267747879, - "learning_rate": 2.6583676963703507e-06, - "loss": 0.0034, + "grad_norm": 0.3078713119029999, + "learning_rate": 8.382553760638999e-05, + "loss": 0.0077, "step": 18020 }, { - "grad_norm": 0.07553955912590027, - "learning_rate": 2.631834170793268e-06, - "loss": 0.0039, + "grad_norm": 0.22283561527729034, + "learning_rate": 8.380523798088631e-05, + "loss": 0.0079, "step": 18030 }, { - "grad_norm": 0.1235010102391243, - "learning_rate": 2.6054301479728036e-06, - "loss": 0.0031, + "grad_norm": 0.2426093965768814, + "learning_rate": 8.378492808624389e-05, + "loss": 0.0083, "step": 18040 }, { - "grad_norm": 0.07297364622354507, - "learning_rate": 2.579155700096575e-06, - "loss": 0.0046, + "grad_norm": 0.2425232082605362, + "learning_rate": 8.376460792863237e-05, + "loss": 0.0092, "step": 18050 }, { - "grad_norm": 0.10736757516860962, - "learning_rate": 2.5530108989978873e-06, - "loss": 0.006, + "grad_norm": 0.2113635241985321, + "learning_rate": 8.374427751422444e-05, + "loss": 0.0106, "step": 18060 }, { - "grad_norm": 0.11740373820066452, - "learning_rate": 2.5269958161556416e-06, - "loss": 0.0064, + "grad_norm": 0.2851988673210144, + "learning_rate": 8.3723936849196e-05, + "loss": 0.0106, "step": 18070 }, { - "grad_norm": 0.0979749783873558, - "learning_rate": 2.5011105226940888e-06, - "loss": 0.0034, + "grad_norm": 0.316919207572937, + "learning_rate": 8.370358593972595e-05, + "loss": 0.0079, "step": 18080 }, { - "grad_norm": 0.13769516348838806, - "learning_rate": 2.4753550893826248e-06, - "loss": 0.0048, + "grad_norm": 0.2205875962972641, + "learning_rate": 8.36832247919964e-05, + "loss": 0.0094, "step": 18090 }, { - "grad_norm": 0.08744233101606369, - "learning_rate": 2.4497295866356296e-06, - "loss": 0.004, + "grad_norm": 0.21108581125736237, + "learning_rate": 8.36628534121925e-05, + "loss": 0.0091, "step": 18100 }, { - "grad_norm": 0.09951108694076538, - "learning_rate": 2.424234084512228e-06, - "loss": 0.0036, + "grad_norm": 0.21722665429115295, + "learning_rate": 8.364247180650254e-05, + "loss": 0.0081, "step": 18110 }, { - "grad_norm": 0.1635640263557434, - "learning_rate": 2.3988686527161687e-06, - "loss": 0.0074, + "grad_norm": 0.2520598769187927, + "learning_rate": 8.362207998111794e-05, + "loss": 0.0124, "step": 18120 }, { - "grad_norm": 0.14396807551383972, - "learning_rate": 2.373633360595573e-06, - "loss": 0.005, + "grad_norm": 0.26642319560050964, + "learning_rate": 8.360167794223318e-05, + "loss": 0.0095, "step": 18130 }, { - "grad_norm": 0.14187557995319366, - "learning_rate": 2.3485282771427585e-06, - "loss": 0.0028, + "grad_norm": 0.21216467022895813, + "learning_rate": 8.358126569604586e-05, + "loss": 0.0068, "step": 18140 }, { - "grad_norm": 0.08442353457212448, - "learning_rate": 2.3235534709940665e-06, - "loss": 0.0031, + "grad_norm": 0.23055925965309143, + "learning_rate": 8.356084324875668e-05, + "loss": 0.0071, "step": 18150 }, { - "grad_norm": 0.1230187937617302, - "learning_rate": 2.2987090104296617e-06, - "loss": 0.004, + "grad_norm": 0.2907140552997589, + "learning_rate": 8.354041060656945e-05, + "loss": 0.0085, "step": 18160 }, { - "grad_norm": 0.10204652696847916, - "learning_rate": 2.273994963373355e-06, - "loss": 0.0038, + "grad_norm": 0.26798176765441895, + "learning_rate": 8.351996777569106e-05, + "loss": 0.0079, "step": 18170 }, { - "grad_norm": 0.07479498535394669, - "learning_rate": 2.249411397392409e-06, - "loss": 0.0045, + "grad_norm": 0.2686067521572113, + "learning_rate": 8.349951476233148e-05, + "loss": 0.0091, "step": 18180 }, { - "grad_norm": 0.10246019065380096, - "learning_rate": 2.2249583796973506e-06, - "loss": 0.0048, + "grad_norm": 0.2189292162656784, + "learning_rate": 8.347905157270386e-05, + "loss": 0.0093, "step": 18190 }, { - "grad_norm": 0.16304340958595276, - "learning_rate": 2.200635977141796e-06, - "loss": 0.0047, + "grad_norm": 0.28690341114997864, + "learning_rate": 8.345857821302432e-05, + "loss": 0.0097, "step": 18200 }, { - "grad_norm": 0.1268751174211502, - "learning_rate": 2.17644425622226e-06, - "loss": 0.0042, + "grad_norm": 0.26576632261276245, + "learning_rate": 8.343809468951213e-05, + "loss": 0.0091, "step": 18210 }, { - "grad_norm": 0.07579465210437775, - "learning_rate": 2.152383283077991e-06, - "loss": 0.0042, + "grad_norm": 0.25923672318458557, + "learning_rate": 8.341760100838965e-05, + "loss": 0.0085, "step": 18220 }, { - "grad_norm": 0.11345085501670837, - "learning_rate": 2.128453123490781e-06, - "loss": 0.0049, + "grad_norm": 0.24061141908168793, + "learning_rate": 8.339709717588233e-05, + "loss": 0.0097, "step": 18230 }, { - "grad_norm": 0.1875070482492447, - "learning_rate": 2.1046538428847462e-06, - "loss": 0.0053, + "grad_norm": 0.2769416272640228, + "learning_rate": 8.33765831982187e-05, + "loss": 0.0104, "step": 18240 }, { - "grad_norm": 0.09485717862844467, - "learning_rate": 2.0809855063262273e-06, - "loss": 0.0042, + "grad_norm": 0.2558574080467224, + "learning_rate": 8.335605908163035e-05, + "loss": 0.0099, "step": 18250 }, { - "grad_norm": 0.07932358235120773, - "learning_rate": 2.057448178523558e-06, - "loss": 0.0043, + "grad_norm": 0.2816294729709625, + "learning_rate": 8.333552483235196e-05, + "loss": 0.0093, "step": 18260 }, { - "grad_norm": 0.11418958753347397, - "learning_rate": 2.034041923826885e-06, - "loss": 0.0051, + "grad_norm": 0.3014129400253296, + "learning_rate": 8.33149804566213e-05, + "loss": 0.0096, "step": 18270 }, { - "grad_norm": 0.08102735131978989, - "learning_rate": 2.0107668062280204e-06, - "loss": 0.0048, + "grad_norm": 0.223966583609581, + "learning_rate": 8.329442596067921e-05, + "loss": 0.0097, "step": 18280 }, { - "grad_norm": 0.08639351278543472, - "learning_rate": 1.9876228893602357e-06, - "loss": 0.0037, + "grad_norm": 0.22248557209968567, + "learning_rate": 8.32738613507696e-05, + "loss": 0.0081, "step": 18290 }, { - "grad_norm": 0.09763824194669724, - "learning_rate": 1.9646102364981266e-06, - "loss": 0.004, + "grad_norm": 0.20519164204597473, + "learning_rate": 8.325328663313946e-05, + "loss": 0.008, "step": 18300 }, { - "grad_norm": 0.08055881410837173, - "learning_rate": 1.9417289105574053e-06, - "loss": 0.0072, + "grad_norm": 0.3648359179496765, + "learning_rate": 8.323270181403884e-05, + "loss": 0.0125, "step": 18310 }, { - "grad_norm": 0.09894809126853943, - "learning_rate": 1.9189789740947427e-06, - "loss": 0.0036, + "grad_norm": 0.23096631467342377, + "learning_rate": 8.321210689972086e-05, + "loss": 0.0083, "step": 18320 }, { - "grad_norm": 0.10058688372373581, - "learning_rate": 1.896360489307597e-06, - "loss": 0.0063, + "grad_norm": 0.2694459557533264, + "learning_rate": 8.319150189644174e-05, + "loss": 0.0119, "step": 18330 }, { - "grad_norm": 0.1028638482093811, - "learning_rate": 1.8738735180340362e-06, - "loss": 0.004, + "grad_norm": 0.23120594024658203, + "learning_rate": 8.31708868104607e-05, + "loss": 0.0088, "step": 18340 }, { - "grad_norm": 0.0817807987332344, - "learning_rate": 1.8515181217525824e-06, - "loss": 0.0038, + "grad_norm": 0.2610611319541931, + "learning_rate": 8.315026164804007e-05, + "loss": 0.009, "step": 18350 }, { - "grad_norm": 0.07613148540258408, - "learning_rate": 1.8292943615820457e-06, - "loss": 0.0036, + "grad_norm": 0.30400434136390686, + "learning_rate": 8.312962641544524e-05, + "loss": 0.0082, "step": 18360 }, { - "grad_norm": 0.05594291910529137, - "learning_rate": 1.8072022982813296e-06, - "loss": 0.0034, + "grad_norm": 0.2520270347595215, + "learning_rate": 8.310898111894465e-05, + "loss": 0.0077, "step": 18370 }, { - "grad_norm": 0.10103534907102585, - "learning_rate": 1.7852419922492925e-06, - "loss": 0.005, + "grad_norm": 0.25413277745246887, + "learning_rate": 8.308832576480977e-05, + "loss": 0.0093, "step": 18380 }, { - "grad_norm": 0.06014038249850273, - "learning_rate": 1.763413503524569e-06, - "loss": 0.0031, + "grad_norm": 0.21032068133354187, + "learning_rate": 8.306766035931519e-05, + "loss": 0.007, "step": 18390 }, { - "grad_norm": 0.11110295355319977, - "learning_rate": 1.7417168917854165e-06, - "loss": 0.0032, + "grad_norm": 0.24193617701530457, + "learning_rate": 8.304698490873847e-05, + "loss": 0.0078, "step": 18400 }, { - "grad_norm": 0.08939255774021149, - "learning_rate": 1.720152216349552e-06, - "loss": 0.0034, + "grad_norm": 0.33122915029525757, + "learning_rate": 8.30262994193603e-05, + "loss": 0.0082, "step": 18410 }, { - "grad_norm": 0.08942482620477676, - "learning_rate": 1.6987195361739595e-06, - "loss": 0.0036, + "grad_norm": 0.22571302950382233, + "learning_rate": 8.300560389746438e-05, + "loss": 0.0082, "step": 18420 }, { - "grad_norm": 0.1863851249217987, - "learning_rate": 1.6774189098547832e-06, - "loss": 0.0051, + "grad_norm": 0.21970698237419128, + "learning_rate": 8.298489834933745e-05, + "loss": 0.0094, "step": 18430 }, { - "grad_norm": 0.12386361509561539, - "learning_rate": 1.6562503956271069e-06, - "loss": 0.0047, + "grad_norm": 0.23111294209957123, + "learning_rate": 8.296418278126934e-05, + "loss": 0.01, "step": 18440 }, { - "grad_norm": 0.12683850526809692, - "learning_rate": 1.6352140513648417e-06, - "loss": 0.0043, + "grad_norm": 0.2237575501203537, + "learning_rate": 8.294345719955284e-05, + "loss": 0.0098, "step": 18450 }, { - "grad_norm": 0.09401161968708038, - "learning_rate": 1.6143099345805712e-06, - "loss": 0.0029, + "grad_norm": 0.18173182010650635, + "learning_rate": 8.29227216104839e-05, + "loss": 0.0074, "step": 18460 }, { - "grad_norm": 0.08010539412498474, - "learning_rate": 1.5935381024253293e-06, - "loss": 0.0035, + "grad_norm": 0.223158597946167, + "learning_rate": 8.290197602036137e-05, + "loss": 0.0078, "step": 18470 }, { - "grad_norm": 0.05355304852128029, - "learning_rate": 1.572898611688517e-06, - "loss": 0.0034, + "grad_norm": 0.21327970921993256, + "learning_rate": 8.288122043548725e-05, + "loss": 0.0073, "step": 18480 }, { - "grad_norm": 0.12306501716375351, - "learning_rate": 1.5523915187977133e-06, - "loss": 0.0056, + "grad_norm": 0.2590601146221161, + "learning_rate": 8.286045486216657e-05, + "loss": 0.0099, "step": 18490 }, { - "grad_norm": 0.10398924350738525, - "learning_rate": 1.532016879818532e-06, - "loss": 0.0039, + "grad_norm": 0.33284223079681396, + "learning_rate": 8.283967930670733e-05, + "loss": 0.0084, "step": 18500 }, { - "grad_norm": 0.07574418187141418, - "learning_rate": 1.51177475045447e-06, - "loss": 0.0037, + "grad_norm": 0.2591264843940735, + "learning_rate": 8.281889377542058e-05, + "loss": 0.0087, "step": 18510 }, { - "grad_norm": 0.0989949107170105, - "learning_rate": 1.4916651860467035e-06, - "loss": 0.0038, + "grad_norm": 0.2607881426811218, + "learning_rate": 8.279809827462045e-05, + "loss": 0.008, "step": 18520 }, { - "grad_norm": 0.06689750403165817, - "learning_rate": 1.471688241574043e-06, - "loss": 0.0031, + "grad_norm": 0.25031501054763794, + "learning_rate": 8.277729281062402e-05, + "loss": 0.0075, "step": 18530 }, { - "grad_norm": 0.09178074449300766, - "learning_rate": 1.451843971652672e-06, - "loss": 0.0041, + "grad_norm": 0.28451529145240784, + "learning_rate": 8.27564773897515e-05, + "loss": 0.0086, "step": 18540 }, { - "grad_norm": 0.12429051101207733, - "learning_rate": 1.432132430536076e-06, - "loss": 0.0042, + "grad_norm": 0.27966487407684326, + "learning_rate": 8.273565201832602e-05, + "loss": 0.0094, "step": 18550 }, { - "grad_norm": 0.08976944535970688, - "learning_rate": 1.412553672114869e-06, - "loss": 0.0035, + "grad_norm": 0.3083644211292267, + "learning_rate": 8.27148167026738e-05, + "loss": 0.0084, "step": 18560 }, { - "grad_norm": 0.07835699617862701, - "learning_rate": 1.3931077499166056e-06, - "loss": 0.0036, + "grad_norm": 0.3163197338581085, + "learning_rate": 8.269397144912405e-05, + "loss": 0.0083, "step": 18570 }, { - "grad_norm": 0.07831160724163055, - "learning_rate": 1.3737947171057085e-06, - "loss": 0.0035, + "grad_norm": 0.21488220989704132, + "learning_rate": 8.267311626400899e-05, + "loss": 0.0069, "step": 18580 }, { - "grad_norm": 0.08674316853284836, - "learning_rate": 1.3546146264832582e-06, - "loss": 0.0038, + "grad_norm": 0.26235413551330566, + "learning_rate": 8.26522511536639e-05, + "loss": 0.0082, "step": 18590 }, { - "grad_norm": 0.06477545201778412, - "learning_rate": 1.3355675304869086e-06, - "loss": 0.0037, + "grad_norm": 0.2428647130727768, + "learning_rate": 8.263137612442706e-05, + "loss": 0.0077, "step": 18600 }, { - "grad_norm": 0.11489281058311462, - "learning_rate": 1.3166534811906827e-06, - "loss": 0.005, + "grad_norm": 0.26525843143463135, + "learning_rate": 8.261049118263971e-05, + "loss": 0.0098, "step": 18610 }, { - "grad_norm": 0.1530413031578064, - "learning_rate": 1.2978725303048666e-06, - "loss": 0.0049, + "grad_norm": 0.22844119369983673, + "learning_rate": 8.258959633464619e-05, + "loss": 0.0097, "step": 18620 }, { - "grad_norm": 0.12522374093532562, - "learning_rate": 1.2792247291758762e-06, - "loss": 0.0045, + "grad_norm": 0.2765507996082306, + "learning_rate": 8.256869158679377e-05, + "loss": 0.0096, "step": 18630 }, { - "grad_norm": 0.06969575583934784, - "learning_rate": 1.2607101287860635e-06, - "loss": 0.0036, + "grad_norm": 0.19514387845993042, + "learning_rate": 8.254777694543278e-05, + "loss": 0.0071, "step": 18640 }, { - "grad_norm": 0.08598778396844864, - "learning_rate": 1.2423287797536654e-06, - "loss": 0.0033, + "grad_norm": 0.2237156480550766, + "learning_rate": 8.252685241691651e-05, + "loss": 0.0076, "step": 18650 }, { - "grad_norm": 0.06391648203134537, - "learning_rate": 1.2240807323325776e-06, - "loss": 0.0037, + "grad_norm": 0.18528668582439423, + "learning_rate": 8.250591800760133e-05, + "loss": 0.0075, "step": 18660 }, { - "grad_norm": 0.10801202058792114, - "learning_rate": 1.205966036412254e-06, - "loss": 0.0033, + "grad_norm": 0.2788935899734497, + "learning_rate": 8.248497372384649e-05, + "loss": 0.0077, "step": 18670 }, { - "grad_norm": 0.11867705732584, - "learning_rate": 1.1879847415175949e-06, - "loss": 0.0051, + "grad_norm": 0.30191946029663086, + "learning_rate": 8.246401957201437e-05, + "loss": 0.0105, "step": 18680 }, { - "grad_norm": 0.11671072244644165, - "learning_rate": 1.1701368968087712e-06, - "loss": 0.0046, + "grad_norm": 0.2995465099811554, + "learning_rate": 8.244305555847027e-05, + "loss": 0.0089, "step": 18690 }, { - "grad_norm": 0.05049500986933708, - "learning_rate": 1.1524225510811116e-06, - "loss": 0.0036, + "grad_norm": 0.2559933662414551, + "learning_rate": 8.24220816895825e-05, + "loss": 0.0084, "step": 18700 }, { - "grad_norm": 0.07130085676908493, - "learning_rate": 1.1348417527649535e-06, - "loss": 0.0034, + "grad_norm": 0.30890005826950073, + "learning_rate": 8.240109797172237e-05, + "loss": 0.0077, "step": 18710 }, { - "grad_norm": 0.054281286895275116, - "learning_rate": 1.1173945499255268e-06, - "loss": 0.0031, + "grad_norm": 0.1988600343465805, + "learning_rate": 8.238010441126416e-05, + "loss": 0.0075, "step": 18720 }, { - "grad_norm": 0.07449881732463837, - "learning_rate": 1.1000809902628307e-06, - "loss": 0.0052, + "grad_norm": 0.19257709383964539, + "learning_rate": 8.23591010145852e-05, + "loss": 0.01, "step": 18730 }, { - "grad_norm": 0.09705372154712677, - "learning_rate": 1.082901121111468e-06, - "loss": 0.0034, + "grad_norm": 0.24763116240501404, + "learning_rate": 8.233808778806571e-05, + "loss": 0.0079, "step": 18740 }, { - "grad_norm": 0.10779692232608795, - "learning_rate": 1.0658549894405456e-06, - "loss": 0.0032, + "grad_norm": 0.29485538601875305, + "learning_rate": 8.231706473808903e-05, + "loss": 0.0073, "step": 18750 }, { - "grad_norm": 0.050938431173563004, - "learning_rate": 1.0489426418535342e-06, - "loss": 0.0026, + "grad_norm": 0.24101503193378448, + "learning_rate": 8.229603187104133e-05, + "loss": 0.0057, "step": 18760 }, { - "grad_norm": 0.07344609498977661, - "learning_rate": 1.0321641245881474e-06, - "loss": 0.003, + "grad_norm": 0.20755937695503235, + "learning_rate": 8.22749891933119e-05, + "loss": 0.0067, "step": 18770 }, { - "grad_norm": 0.0742596834897995, - "learning_rate": 1.015519483516214e-06, - "loss": 0.0037, + "grad_norm": 0.25986117124557495, + "learning_rate": 8.225393671129291e-05, + "loss": 0.0079, "step": 18780 }, { - "grad_norm": 0.06837273389101028, - "learning_rate": 9.990087641435443e-07, - "loss": 0.003, + "grad_norm": 0.2040674239397049, + "learning_rate": 8.223287443137957e-05, + "loss": 0.0068, "step": 18790 }, { - "grad_norm": 0.08643893897533417, - "learning_rate": 9.826320116098132e-07, - "loss": 0.0044, + "grad_norm": 0.24269916117191315, + "learning_rate": 8.221180235997004e-05, + "loss": 0.0082, "step": 18800 }, { - "grad_norm": 0.04812360182404518, - "learning_rate": 9.663892706884447e-07, - "loss": 0.0031, + "grad_norm": 0.1868121176958084, + "learning_rate": 8.219072050346544e-05, + "loss": 0.0075, "step": 18810 }, { - "grad_norm": 0.08176058530807495, - "learning_rate": 9.502805857864616e-07, - "loss": 0.0061, + "grad_norm": 0.27327829599380493, + "learning_rate": 8.216962886826992e-05, + "loss": 0.0103, "step": 18820 }, { - "grad_norm": 0.08198662847280502, - "learning_rate": 9.34306000944396e-07, - "loss": 0.0039, + "grad_norm": 0.2937210500240326, + "learning_rate": 8.214852746079054e-05, + "loss": 0.0094, "step": 18830 }, { - "grad_norm": 0.07465307414531708, - "learning_rate": 9.184655598361624e-07, - "loss": 0.0034, + "grad_norm": 0.27595123648643494, + "learning_rate": 8.212741628743732e-05, + "loss": 0.0085, "step": 18840 }, { - "grad_norm": 0.07182762771844864, - "learning_rate": 9.027593057689076e-07, - "loss": 0.0038, + "grad_norm": 0.20546695590019226, + "learning_rate": 8.210629535462333e-05, + "loss": 0.0083, "step": 18850 }, { - "grad_norm": 0.0734962522983551, - "learning_rate": 8.871872816829441e-07, - "loss": 0.0031, + "grad_norm": 0.32897746562957764, + "learning_rate": 8.208516466876453e-05, + "loss": 0.0075, "step": 18860 }, { - "grad_norm": 0.06601841002702713, - "learning_rate": 8.717495301515777e-07, - "loss": 0.0028, + "grad_norm": 0.2310360074043274, + "learning_rate": 8.206402423627986e-05, + "loss": 0.0073, "step": 18870 }, { - "grad_norm": 0.10713382810354233, - "learning_rate": 8.564460933810415e-07, - "loss": 0.0044, + "grad_norm": 0.23503150045871735, + "learning_rate": 8.204287406359124e-05, + "loss": 0.0093, "step": 18880 }, { - "grad_norm": 0.09342646598815918, - "learning_rate": 8.412770132103453e-07, - "loss": 0.006, + "grad_norm": 0.30840203166007996, + "learning_rate": 8.20217141571235e-05, + "loss": 0.0113, "step": 18890 }, { - "grad_norm": 0.0634833425283432, - "learning_rate": 8.262423311111711e-07, - "loss": 0.0034, + "grad_norm": 0.25985071063041687, + "learning_rate": 8.200054452330449e-05, + "loss": 0.0076, "step": 18900 }, { - "grad_norm": 0.07692574709653854, - "learning_rate": 8.113420881877665e-07, - "loss": 0.0037, + "grad_norm": 0.24178855121135712, + "learning_rate": 8.197936516856499e-05, + "loss": 0.008, "step": 18910 }, { - "grad_norm": 0.07447660714387894, - "learning_rate": 7.965763251768288e-07, - "loss": 0.004, + "grad_norm": 0.28449511528015137, + "learning_rate": 8.195817609933871e-05, + "loss": 0.0082, "step": 18920 }, { - "grad_norm": 0.1114739403128624, - "learning_rate": 7.819450824473995e-07, - "loss": 0.0046, + "grad_norm": 0.2819555401802063, + "learning_rate": 8.193697732206233e-05, + "loss": 0.0088, "step": 18930 }, { - "grad_norm": 0.07135684788227081, - "learning_rate": 7.674484000007198e-07, - "loss": 0.004, + "grad_norm": 0.30575448274612427, + "learning_rate": 8.19157688431755e-05, + "loss": 0.0082, "step": 18940 }, { - "grad_norm": 0.08017963916063309, - "learning_rate": 7.530863174701752e-07, - "loss": 0.0042, + "grad_norm": 0.23637647926807404, + "learning_rate": 8.189455066912077e-05, + "loss": 0.0096, "step": 18950 }, { - "grad_norm": 0.08928892016410828, - "learning_rate": 7.38858874121151e-07, - "loss": 0.0052, + "grad_norm": 0.1974545419216156, + "learning_rate": 8.187332280634369e-05, + "loss": 0.0099, "step": 18960 }, { - "grad_norm": 0.1515638679265976, - "learning_rate": 7.247661088509328e-07, - "loss": 0.0039, + "grad_norm": 0.33112841844558716, + "learning_rate": 8.18520852612927e-05, + "loss": 0.0085, "step": 18970 }, { - "grad_norm": 0.09839137643575668, - "learning_rate": 7.108080601886002e-07, - "loss": 0.0042, + "grad_norm": 0.25162628293037415, + "learning_rate": 8.183083804041921e-05, + "loss": 0.0087, "step": 18980 }, { - "grad_norm": 0.060578133910894394, - "learning_rate": 6.969847662949336e-07, - "loss": 0.0043, + "grad_norm": 0.22467279434204102, + "learning_rate": 8.180958115017757e-05, + "loss": 0.0089, "step": 18990 }, { - "grad_norm": 0.07045385986566544, - "learning_rate": 6.832962649622798e-07, - "loss": 0.0043, + "grad_norm": 0.2644553780555725, + "learning_rate": 8.178831459702505e-05, + "loss": 0.0094, "step": 19000 }, { - "grad_norm": 0.11689253896474838, - "learning_rate": 6.697425936144863e-07, - "loss": 0.0033, + "grad_norm": 0.2908259332180023, + "learning_rate": 8.17670383874219e-05, + "loss": 0.0078, "step": 19010 }, { - "grad_norm": 0.051963455975055695, - "learning_rate": 6.563237893067731e-07, - "loss": 0.0035, + "grad_norm": 0.22885648906230927, + "learning_rate": 8.174575252783124e-05, + "loss": 0.0075, "step": 19020 }, { - "grad_norm": 0.08776947855949402, - "learning_rate": 6.430398887256328e-07, - "loss": 0.0041, + "grad_norm": 0.26460787653923035, + "learning_rate": 8.172445702471914e-05, + "loss": 0.0094, "step": 19030 }, { - "grad_norm": 0.053857170045375824, - "learning_rate": 6.298909281887478e-07, - "loss": 0.0029, + "grad_norm": 0.23526042699813843, + "learning_rate": 8.170315188455466e-05, + "loss": 0.007, "step": 19040 }, { - "grad_norm": 0.1286555677652359, - "learning_rate": 6.168769436448673e-07, - "loss": 0.0033, + "grad_norm": 0.24561813473701477, + "learning_rate": 8.168183711380969e-05, + "loss": 0.008, "step": 19050 }, { - "grad_norm": 0.08176587522029877, - "learning_rate": 6.03997970673742e-07, - "loss": 0.003, + "grad_norm": 0.239788219332695, + "learning_rate": 8.166051271895913e-05, + "loss": 0.0072, "step": 19060 }, { - "grad_norm": 0.09932407736778259, - "learning_rate": 5.912540444859782e-07, - "loss": 0.0044, + "grad_norm": 0.22616542875766754, + "learning_rate": 8.163917870648075e-05, + "loss": 0.0089, "step": 19070 }, { - "grad_norm": 0.0911400094628334, - "learning_rate": 5.786451999229837e-07, - "loss": 0.007, + "grad_norm": 0.2480158805847168, + "learning_rate": 8.161783508285526e-05, + "loss": 0.0112, "step": 19080 }, { - "grad_norm": 0.0852760598063469, - "learning_rate": 5.661714714568722e-07, - "loss": 0.004, + "grad_norm": 0.18843553960323334, + "learning_rate": 8.159648185456628e-05, + "loss": 0.0083, "step": 19090 }, { - "grad_norm": 0.08842559158802032, - "learning_rate": 5.538328931903259e-07, - "loss": 0.0036, + "grad_norm": 0.2588503360748291, + "learning_rate": 8.157511902810038e-05, + "loss": 0.0085, "step": 19100 }, { - "grad_norm": 0.07818997651338577, - "learning_rate": 5.416294988565551e-07, - "loss": 0.0033, + "grad_norm": 0.24212302267551422, + "learning_rate": 8.155374660994701e-05, + "loss": 0.0078, "step": 19110 }, { - "grad_norm": 0.059740930795669556, - "learning_rate": 5.29561321819172e-07, - "loss": 0.004, + "grad_norm": 0.25631436705589294, + "learning_rate": 8.153236460659857e-05, + "loss": 0.0087, "step": 19120 }, { - "grad_norm": 0.07137617468833923, - "learning_rate": 5.176283950721061e-07, - "loss": 0.0034, + "grad_norm": 0.22890721261501312, + "learning_rate": 8.151097302455031e-05, + "loss": 0.0069, "step": 19130 }, { - "grad_norm": 0.08400587737560272, - "learning_rate": 5.058307512395332e-07, - "loss": 0.0037, + "grad_norm": 0.22997477650642395, + "learning_rate": 8.148957187030044e-05, + "loss": 0.0076, "step": 19140 }, { - "grad_norm": 0.19625791907310486, - "learning_rate": 4.941684225757526e-07, - "loss": 0.0035, + "grad_norm": 0.293473482131958, + "learning_rate": 8.146816115035006e-05, + "loss": 0.0076, "step": 19150 }, { - "grad_norm": 0.03737051039934158, - "learning_rate": 4.826414409651314e-07, - "loss": 0.0031, + "grad_norm": 0.23699471354484558, + "learning_rate": 8.14467408712032e-05, + "loss": 0.0066, "step": 19160 }, { - "grad_norm": 0.1145489290356636, - "learning_rate": 4.712498379219943e-07, - "loss": 0.0033, + "grad_norm": 0.1930815726518631, + "learning_rate": 8.142531103936678e-05, + "loss": 0.0075, "step": 19170 }, { - "grad_norm": 0.10778439044952393, - "learning_rate": 4.599936445905506e-07, - "loss": 0.0047, + "grad_norm": 0.25631794333457947, + "learning_rate": 8.14038716613506e-05, + "loss": 0.0096, "step": 19180 }, { - "grad_norm": 0.056398406624794006, - "learning_rate": 4.4887289174480594e-07, - "loss": 0.0035, + "grad_norm": 0.2409273236989975, + "learning_rate": 8.138242274366736e-05, + "loss": 0.0092, "step": 19190 }, { - "grad_norm": 0.05212453380227089, - "learning_rate": 4.378876097884621e-07, - "loss": 0.0036, + "grad_norm": 0.22028669714927673, + "learning_rate": 8.136096429283271e-05, + "loss": 0.0077, "step": 19200 }, { - "grad_norm": 0.0517825148999691, - "learning_rate": 4.2703782875487264e-07, - "loss": 0.0027, + "grad_norm": 0.20971792936325073, + "learning_rate": 8.133949631536515e-05, + "loss": 0.0066, "step": 19210 }, { - "grad_norm": 0.04213884100317955, - "learning_rate": 4.163235783069208e-07, - "loss": 0.0025, + "grad_norm": 0.18486718833446503, + "learning_rate": 8.131801881778607e-05, + "loss": 0.0059, "step": 19220 }, { - "grad_norm": 0.0969507098197937, - "learning_rate": 4.057448877369585e-07, - "loss": 0.0041, + "grad_norm": 0.25684654712677, + "learning_rate": 8.129653180661978e-05, + "loss": 0.008, "step": 19230 }, { - "grad_norm": 0.05608559772372246, - "learning_rate": 3.9530178596672295e-07, - "loss": 0.0037, + "grad_norm": 0.21516060829162598, + "learning_rate": 8.127503528839346e-05, + "loss": 0.0074, "step": 19240 }, { - "grad_norm": 0.056393902748823166, - "learning_rate": 3.849943015472479e-07, - "loss": 0.0047, + "grad_norm": 0.30441102385520935, + "learning_rate": 8.125352926963721e-05, + "loss": 0.01, "step": 19250 }, { - "grad_norm": 0.10194101184606552, - "learning_rate": 3.748224626588137e-07, - "loss": 0.0031, + "grad_norm": 0.2278432697057724, + "learning_rate": 8.123201375688395e-05, + "loss": 0.0076, "step": 19260 }, { - "grad_norm": 0.0955602377653122, - "learning_rate": 3.647862971108307e-07, - "loss": 0.0035, + "grad_norm": 0.20663349330425262, + "learning_rate": 8.121048875666954e-05, + "loss": 0.0075, "step": 19270 }, { - "grad_norm": 0.06831927597522736, - "learning_rate": 3.5488583234179473e-07, - "loss": 0.0046, + "grad_norm": 0.30783334374427795, + "learning_rate": 8.118895427553274e-05, + "loss": 0.01, "step": 19280 }, { - "grad_norm": 0.044883936643600464, - "learning_rate": 3.4512109541920413e-07, - "loss": 0.0041, + "grad_norm": 0.23236003518104553, + "learning_rate": 8.116741032001511e-05, + "loss": 0.0086, "step": 19290 }, { - "grad_norm": 0.046192340552806854, - "learning_rate": 3.354921130394706e-07, - "loss": 0.0026, + "grad_norm": 0.2601412534713745, + "learning_rate": 8.114585689666114e-05, + "loss": 0.0064, "step": 19300 }, { - "grad_norm": 0.04406873881816864, - "learning_rate": 3.259989115278639e-07, - "loss": 0.0029, + "grad_norm": 0.18265371024608612, + "learning_rate": 8.112429401201821e-05, + "loss": 0.0068, "step": 19310 }, { - "grad_norm": 0.07073112577199936, - "learning_rate": 3.1664151683843403e-07, - "loss": 0.0042, + "grad_norm": 0.24811069667339325, + "learning_rate": 8.110272167263656e-05, + "loss": 0.0083, "step": 19320 }, { - "grad_norm": 0.05474907159805298, - "learning_rate": 3.074199545539447e-07, - "loss": 0.0052, + "grad_norm": 0.23625928163528442, + "learning_rate": 8.108113988506929e-05, + "loss": 0.0102, "step": 19330 }, { - "grad_norm": 0.047355033457279205, - "learning_rate": 2.983342498857955e-07, - "loss": 0.0043, + "grad_norm": 0.23172149062156677, + "learning_rate": 8.105954865587235e-05, + "loss": 0.0095, "step": 19340 }, { - "grad_norm": 0.10388905555009842, - "learning_rate": 2.893844276739499e-07, - "loss": 0.003, + "grad_norm": 0.23441234230995178, + "learning_rate": 8.103794799160463e-05, + "loss": 0.0077, "step": 19350 }, { - "grad_norm": 0.10018634796142578, - "learning_rate": 2.8057051238688514e-07, - "loss": 0.0038, + "grad_norm": 0.2271808683872223, + "learning_rate": 8.101633789882781e-05, + "loss": 0.0085, "step": 19360 }, { - "grad_norm": 0.03913424164056778, - "learning_rate": 2.71892528121509e-07, - "loss": 0.0032, + "grad_norm": 0.2427125722169876, + "learning_rate": 8.099471838410648e-05, + "loss": 0.0074, "step": 19370 }, { - "grad_norm": 0.07924854755401611, - "learning_rate": 2.633504986030988e-07, - "loss": 0.0043, + "grad_norm": 0.2313302904367447, + "learning_rate": 8.097308945400806e-05, + "loss": 0.0084, "step": 19380 }, { - "grad_norm": 0.09704583883285522, - "learning_rate": 2.549444471852347e-07, - "loss": 0.0055, + "grad_norm": 0.2358667105436325, + "learning_rate": 8.095145111510288e-05, + "loss": 0.0093, "step": 19390 }, { - "grad_norm": 0.05486253276467323, - "learning_rate": 2.4667439684974423e-07, - "loss": 0.0038, + "grad_norm": 0.20756182074546814, + "learning_rate": 8.092980337396406e-05, + "loss": 0.0086, "step": 19400 }, { - "grad_norm": 0.12703447043895721, - "learning_rate": 2.3854037020662467e-07, - "loss": 0.0056, + "grad_norm": 0.2865169048309326, + "learning_rate": 8.090814623716763e-05, + "loss": 0.0109, "step": 19410 }, { - "grad_norm": 0.06610122323036194, - "learning_rate": 2.3054238949399288e-07, - "loss": 0.0027, + "grad_norm": 0.27583447098731995, + "learning_rate": 8.088647971129246e-05, + "loss": 0.007, "step": 19420 }, { - "grad_norm": 0.08677727729082108, - "learning_rate": 2.2268047657802993e-07, - "loss": 0.0037, + "grad_norm": 0.2980212867259979, + "learning_rate": 8.086480380292026e-05, + "loss": 0.0082, "step": 19430 }, { - "grad_norm": 0.06518864631652832, - "learning_rate": 2.149546529529034e-07, - "loss": 0.0039, + "grad_norm": 0.21550731360912323, + "learning_rate": 8.084311851863562e-05, + "loss": 0.0076, "step": 19440 }, { - "grad_norm": 0.11082996428012848, - "learning_rate": 2.0736493974071736e-07, - "loss": 0.004, + "grad_norm": 0.2747206389904022, + "learning_rate": 8.082142386502591e-05, + "loss": 0.0076, "step": 19450 }, { - "grad_norm": 0.036419276148080826, - "learning_rate": 1.9991135769145686e-07, - "loss": 0.003, + "grad_norm": 0.3016121983528137, + "learning_rate": 8.079971984868145e-05, + "loss": 0.0071, "step": 19460 }, { - "grad_norm": 0.0653953105211258, - "learning_rate": 1.9259392718293245e-07, - "loss": 0.0042, + "grad_norm": 0.3915133774280548, + "learning_rate": 8.077800647619532e-05, + "loss": 0.0082, "step": 19470 }, { - "grad_norm": 0.09075991064310074, - "learning_rate": 1.8541266822072467e-07, - "loss": 0.0056, + "grad_norm": 0.234879732131958, + "learning_rate": 8.075628375416345e-05, + "loss": 0.0097, "step": 19480 }, { - "grad_norm": 0.06219499185681343, - "learning_rate": 1.7836760043811184e-07, - "loss": 0.0033, + "grad_norm": 0.23224681615829468, + "learning_rate": 8.073455168918464e-05, + "loss": 0.007, "step": 19490 }, { - "grad_norm": 0.11633436381816864, - "learning_rate": 1.7145874309604792e-07, - "loss": 0.0031, + "grad_norm": 0.23529082536697388, + "learning_rate": 8.071281028786055e-05, + "loss": 0.0065, "step": 19500 }, { - "grad_norm": 0.11897324025630951, - "learning_rate": 1.6468611508308474e-07, - "loss": 0.0054, + "grad_norm": 0.24942827224731445, + "learning_rate": 8.069105955679562e-05, + "loss": 0.0095, "step": 19510 }, { - "grad_norm": 0.055197566747665405, - "learning_rate": 1.5804973491532204e-07, - "loss": 0.0032, + "grad_norm": 0.20556573569774628, + "learning_rate": 8.066929950259713e-05, + "loss": 0.0067, "step": 19520 }, { - "grad_norm": 0.07045387476682663, - "learning_rate": 1.5154962073637424e-07, - "loss": 0.0043, + "grad_norm": 0.25004634261131287, + "learning_rate": 8.064753013187522e-05, + "loss": 0.0081, "step": 19530 }, { - "grad_norm": 0.06476310640573502, - "learning_rate": 1.4518579031730372e-07, - "loss": 0.0034, + "grad_norm": 0.24985705316066742, + "learning_rate": 8.062575145124289e-05, + "loss": 0.0071, "step": 19540 }, { - "grad_norm": 0.10052386671304703, - "learning_rate": 1.389582610565876e-07, - "loss": 0.0041, + "grad_norm": 0.28442367911338806, + "learning_rate": 8.060396346731587e-05, + "loss": 0.0086, "step": 19550 }, { - "grad_norm": 0.12584728002548218, - "learning_rate": 1.3286704998003995e-07, - "loss": 0.0034, + "grad_norm": 0.2917954623699188, + "learning_rate": 8.058216618671281e-05, + "loss": 0.0079, "step": 19560 }, { - "grad_norm": 0.04220246151089668, - "learning_rate": 1.2691217374080632e-07, - "loss": 0.005, + "grad_norm": 0.2235489934682846, + "learning_rate": 8.056035961605514e-05, + "loss": 0.0089, "step": 19570 }, { - "grad_norm": 0.0714455097913742, - "learning_rate": 1.2109364861929705e-07, - "loss": 0.0024, + "grad_norm": 0.2291390597820282, + "learning_rate": 8.05385437619671e-05, + "loss": 0.0058, "step": 19580 }, { - "grad_norm": 0.07059883326292038, - "learning_rate": 1.1541149052312628e-07, - "loss": 0.0045, + "grad_norm": 0.26902198791503906, + "learning_rate": 8.05167186310758e-05, + "loss": 0.0095, "step": 19590 }, { - "grad_norm": 0.05011598765850067, - "learning_rate": 1.0986571498710074e-07, - "loss": 0.0046, + "grad_norm": 0.2794935703277588, + "learning_rate": 8.049488423001113e-05, + "loss": 0.009, "step": 19600 }, { - "grad_norm": 0.03681618720293045, - "learning_rate": 1.0445633717316438e-07, - "loss": 0.003, + "grad_norm": 0.24435113370418549, + "learning_rate": 8.047304056540581e-05, + "loss": 0.0074, "step": 19610 }, { - "grad_norm": 0.13256072998046875, - "learning_rate": 9.918337187034277e-08, - "loss": 0.0038, + "grad_norm": 0.2055334597826004, + "learning_rate": 8.045118764389534e-05, + "loss": 0.0085, "step": 19620 }, { - "grad_norm": 0.08195763826370239, - "learning_rate": 9.404683349472643e-08, - "loss": 0.0029, + "grad_norm": 0.17402079701423645, + "learning_rate": 8.042932547211809e-05, + "loss": 0.0065, "step": 19630 }, { - "grad_norm": 0.06682359427213669, - "learning_rate": 8.904673608940983e-08, - "loss": 0.005, + "grad_norm": 0.22251446545124054, + "learning_rate": 8.04074540567152e-05, + "loss": 0.0092, "step": 19640 }, { - "grad_norm": 0.11099401861429214, - "learning_rate": 8.418309332447471e-08, - "loss": 0.0044, + "grad_norm": 0.25850194692611694, + "learning_rate": 8.038557340433063e-05, + "loss": 0.0074, "step": 19650 }, { - "grad_norm": 0.06852664053440094, - "learning_rate": 7.945591849692902e-08, - "loss": 0.0033, + "grad_norm": 0.2445290982723236, + "learning_rate": 8.036368352161115e-05, + "loss": 0.0075, "step": 19660 }, { - "grad_norm": 0.10303331166505814, - "learning_rate": 7.486522453069578e-08, - "loss": 0.0044, + "grad_norm": 0.21069346368312836, + "learning_rate": 8.034178441520633e-05, + "loss": 0.0077, "step": 19670 }, { - "grad_norm": 0.11328206211328506, - "learning_rate": 7.041102397655208e-08, - "loss": 0.0048, + "grad_norm": 0.2779119908809662, + "learning_rate": 8.031987609176852e-05, + "loss": 0.0094, "step": 19680 }, { - "grad_norm": 0.053033504635095596, - "learning_rate": 6.609332901210685e-08, - "loss": 0.0031, + "grad_norm": 0.20738078653812408, + "learning_rate": 8.02979585579529e-05, + "loss": 0.0071, "step": 19690 }, { - "grad_norm": 0.06228000670671463, - "learning_rate": 6.191215144178419e-08, - "loss": 0.0028, + "grad_norm": 0.2439897060394287, + "learning_rate": 8.027603182041745e-05, + "loss": 0.0065, "step": 19700 }, { - "grad_norm": 0.06171086058020592, - "learning_rate": 5.786750269675678e-08, - "loss": 0.0041, + "grad_norm": 0.2629689872264862, + "learning_rate": 8.025409588582292e-05, + "loss": 0.0089, "step": 19710 }, { - "grad_norm": 0.11801750957965851, - "learning_rate": 5.395939383494031e-08, - "loss": 0.0035, + "grad_norm": 0.2789883315563202, + "learning_rate": 8.023215076083288e-05, + "loss": 0.0082, "step": 19720 }, { - "grad_norm": 0.0650201365351677, - "learning_rate": 5.018783554095463e-08, - "loss": 0.0061, + "grad_norm": 0.22468744218349457, + "learning_rate": 8.021019645211367e-05, + "loss": 0.0107, "step": 19730 }, { - "grad_norm": 0.06774918735027313, - "learning_rate": 4.655283812610156e-08, - "loss": 0.0037, + "grad_norm": 0.22162243723869324, + "learning_rate": 8.018823296633441e-05, + "loss": 0.0081, "step": 19740 }, { - "grad_norm": 0.05833545699715614, - "learning_rate": 4.305441152831491e-08, - "loss": 0.0036, + "grad_norm": 0.255220502614975, + "learning_rate": 8.016626031016708e-05, + "loss": 0.0073, "step": 19750 }, { - "grad_norm": 0.061846163123846054, - "learning_rate": 3.9692565312171584e-08, - "loss": 0.0026, + "grad_norm": 0.1920865923166275, + "learning_rate": 8.014427849028636e-05, + "loss": 0.0062, "step": 19760 }, { - "grad_norm": 0.11921104788780212, - "learning_rate": 3.6467308668824975e-08, - "loss": 0.0042, + "grad_norm": 0.3471209406852722, + "learning_rate": 8.012228751336974e-05, + "loss": 0.0097, "step": 19770 }, { - "grad_norm": 0.0699174776673317, - "learning_rate": 3.3378650416004964e-08, - "loss": 0.0039, + "grad_norm": 0.2116909772157669, + "learning_rate": 8.01002873860975e-05, + "loss": 0.0079, "step": 19780 }, { - "grad_norm": 0.05499502271413803, - "learning_rate": 3.042659899797906e-08, - "loss": 0.0035, + "grad_norm": 0.285729318857193, + "learning_rate": 8.00782781151527e-05, + "loss": 0.0078, "step": 19790 }, { - "grad_norm": 0.050804853439331055, - "learning_rate": 2.76111624855524e-08, - "loss": 0.003, + "grad_norm": 0.18220259249210358, + "learning_rate": 8.005625970722119e-05, + "loss": 0.007, "step": 19800 }, { - "grad_norm": 0.08485697954893112, - "learning_rate": 2.4932348576017784e-08, - "loss": 0.0034, + "grad_norm": 0.2763148248195648, + "learning_rate": 8.003423216899158e-05, + "loss": 0.0074, "step": 19810 }, { - "grad_norm": 0.11028487235307693, - "learning_rate": 2.239016459314458e-08, - "loss": 0.0039, + "grad_norm": 0.29228675365448, + "learning_rate": 8.001219550715522e-05, + "loss": 0.0083, "step": 19820 }, { - "grad_norm": 0.11637432128190994, - "learning_rate": 1.9984617487173174e-08, - "loss": 0.004, + "grad_norm": 0.23517456650733948, + "learning_rate": 7.999014972840632e-05, + "loss": 0.0087, "step": 19830 }, { - "grad_norm": 0.07264947891235352, - "learning_rate": 1.7715713834776105e-08, - "loss": 0.0038, + "grad_norm": 0.3168605864048004, + "learning_rate": 7.996809483944174e-05, + "loss": 0.0084, "step": 19840 }, { - "grad_norm": 0.0735667273402214, - "learning_rate": 1.5583459839046964e-08, - "loss": 0.0037, + "grad_norm": 0.26795217394828796, + "learning_rate": 7.994603084696124e-05, + "loss": 0.0075, "step": 19850 }, { - "grad_norm": 0.09691016376018524, - "learning_rate": 1.3587861329489304e-08, - "loss": 0.0048, + "grad_norm": 0.20814606547355652, + "learning_rate": 7.992395775766724e-05, + "loss": 0.0085, "step": 19860 }, { - "grad_norm": 0.08705077320337296, - "learning_rate": 1.1728923761994415e-08, - "loss": 0.0046, + "grad_norm": 0.24424220621585846, + "learning_rate": 7.990187557826497e-05, + "loss": 0.009, "step": 19870 }, { - "grad_norm": 0.06617732346057892, - "learning_rate": 1.0006652218819135e-08, - "loss": 0.0049, + "grad_norm": 0.2510085701942444, + "learning_rate": 7.987978431546242e-05, + "loss": 0.0089, "step": 19880 }, { - "grad_norm": 0.060623060911893845, - "learning_rate": 8.421051408596947e-09, - "loss": 0.0052, + "grad_norm": 0.25798335671424866, + "learning_rate": 7.985768397597031e-05, + "loss": 0.0088, "step": 19890 }, { - "grad_norm": 0.10678309947252274, - "learning_rate": 6.972125666299123e-09, - "loss": 0.0043, + "grad_norm": 0.24623669683933258, + "learning_rate": 7.983557456650216e-05, + "loss": 0.0089, "step": 19900 }, { - "grad_norm": 0.11491920053958893, - "learning_rate": 5.659878953229169e-09, - "loss": 0.0058, + "grad_norm": 0.3170531094074249, + "learning_rate": 7.981345609377422e-05, + "loss": 0.0109, "step": 19910 }, { - "grad_norm": 0.0382302887737751, - "learning_rate": 4.48431485701728e-09, - "loss": 0.0035, + "grad_norm": 0.2596462666988373, + "learning_rate": 7.97913285645055e-05, + "loss": 0.0077, "step": 19920 }, { - "grad_norm": 0.10758187621831894, - "learning_rate": 3.4454365916203322e-09, - "loss": 0.0036, + "grad_norm": 0.25415122509002686, + "learning_rate": 7.976919198541776e-05, + "loss": 0.007, "step": 19930 }, { - "grad_norm": 0.11459647119045258, - "learning_rate": 2.5432469972830332e-09, - "loss": 0.0041, + "grad_norm": 0.21104435622692108, + "learning_rate": 7.974704636323548e-05, + "loss": 0.0081, "step": 19940 }, { - "grad_norm": 0.10030755400657654, - "learning_rate": 1.7777485405601203e-09, - "loss": 0.0031, + "grad_norm": 0.2091863602399826, + "learning_rate": 7.972489170468597e-05, + "loss": 0.0067, "step": 19950 }, { - "grad_norm": 0.0468946136534214, - "learning_rate": 1.1489433142941597e-09, - "loss": 0.004, + "grad_norm": 0.31201303005218506, + "learning_rate": 7.970272801649918e-05, + "loss": 0.008, "step": 19960 }, { - "grad_norm": 0.12162579596042633, - "learning_rate": 6.568330376210963e-10, - "loss": 0.0043, + "grad_norm": 0.2933281362056732, + "learning_rate": 7.96805553054079e-05, + "loss": 0.0096, "step": 19970 }, { - "grad_norm": 0.051232777535915375, - "learning_rate": 3.0141905594249787e-10, - "loss": 0.0031, + "grad_norm": 0.2672351896762848, + "learning_rate": 7.965837357814756e-05, + "loss": 0.008, "step": 19980 }, { - "grad_norm": 0.05374916270375252, - "learning_rate": 8.270234094776008e-11, - "loss": 0.0028, + "grad_norm": 0.22798676788806915, + "learning_rate": 7.963618284145643e-05, + "loss": 0.0073, "step": 19990 }, { - "grad_norm": 0.07233867049217224, - "learning_rate": 6.834906085551041e-13, - "loss": 0.0039, + "grad_norm": 0.2415589839220047, + "learning_rate": 7.961398310207544e-05, + "loss": 0.0081, "step": 20000 + }, + { + "grad_norm": 0.2610226273536682, + "learning_rate": 7.95917743667483e-05, + "loss": 0.011, + "step": 20010 + }, + { + "grad_norm": 0.2124694585800171, + "learning_rate": 7.956955664222144e-05, + "loss": 0.0085, + "step": 20020 + }, + { + "grad_norm": 0.16482801735401154, + "learning_rate": 7.954732993524399e-05, + "loss": 0.0079, + "step": 20030 + }, + { + "grad_norm": 0.25093182921409607, + "learning_rate": 7.952509425256786e-05, + "loss": 0.0067, + "step": 20040 + }, + { + "grad_norm": 0.1770930290222168, + "learning_rate": 7.950284960094767e-05, + "loss": 0.0065, + "step": 20050 + }, + { + "grad_norm": 0.18716323375701904, + "learning_rate": 7.948059598714076e-05, + "loss": 0.008, + "step": 20060 + }, + { + "grad_norm": 0.3074689209461212, + "learning_rate": 7.945833341790717e-05, + "loss": 0.0089, + "step": 20070 + }, + { + "grad_norm": 0.20486070215702057, + "learning_rate": 7.94360619000097e-05, + "loss": 0.0077, + "step": 20080 + }, + { + "grad_norm": 0.26866084337234497, + "learning_rate": 7.941378144021381e-05, + "loss": 0.0099, + "step": 20090 + }, + { + "grad_norm": 0.2770794928073883, + "learning_rate": 7.939149204528777e-05, + "loss": 0.0119, + "step": 20100 + }, + { + "grad_norm": 0.22791442275047302, + "learning_rate": 7.936919372200246e-05, + "loss": 0.0101, + "step": 20110 + }, + { + "grad_norm": 0.31353890895843506, + "learning_rate": 7.934688647713158e-05, + "loss": 0.0106, + "step": 20120 + }, + { + "grad_norm": 0.28576356172561646, + "learning_rate": 7.932457031745143e-05, + "loss": 0.0095, + "step": 20130 + }, + { + "grad_norm": 0.3792745769023895, + "learning_rate": 7.930224524974108e-05, + "loss": 0.0099, + "step": 20140 + }, + { + "grad_norm": 0.27687785029411316, + "learning_rate": 7.927991128078232e-05, + "loss": 0.0101, + "step": 20150 + }, + { + "grad_norm": 0.30250903964042664, + "learning_rate": 7.925756841735958e-05, + "loss": 0.0082, + "step": 20160 + }, + { + "grad_norm": 0.24277298152446747, + "learning_rate": 7.923521666626008e-05, + "loss": 0.0074, + "step": 20170 + }, + { + "grad_norm": 0.23286497592926025, + "learning_rate": 7.921285603427366e-05, + "loss": 0.0076, + "step": 20180 + }, + { + "grad_norm": 0.20044639706611633, + "learning_rate": 7.91904865281929e-05, + "loss": 0.0082, + "step": 20190 + }, + { + "grad_norm": 0.22045907378196716, + "learning_rate": 7.916810815481307e-05, + "loss": 0.0073, + "step": 20200 + }, + { + "grad_norm": 0.21045024693012238, + "learning_rate": 7.914572092093211e-05, + "loss": 0.0078, + "step": 20210 + }, + { + "grad_norm": 0.2371012568473816, + "learning_rate": 7.912332483335068e-05, + "loss": 0.0098, + "step": 20220 + }, + { + "grad_norm": 0.21180227398872375, + "learning_rate": 7.910091989887213e-05, + "loss": 0.0072, + "step": 20230 + }, + { + "grad_norm": 0.2546599507331848, + "learning_rate": 7.907850612430248e-05, + "loss": 0.007, + "step": 20240 + }, + { + "grad_norm": 0.28301095962524414, + "learning_rate": 7.905608351645044e-05, + "loss": 0.0085, + "step": 20250 + }, + { + "grad_norm": 0.24904896318912506, + "learning_rate": 7.90336520821274e-05, + "loss": 0.0103, + "step": 20260 + }, + { + "grad_norm": 0.25038453936576843, + "learning_rate": 7.901121182814746e-05, + "loss": 0.0078, + "step": 20270 + }, + { + "grad_norm": 0.26120755076408386, + "learning_rate": 7.898876276132736e-05, + "loss": 0.0076, + "step": 20280 + }, + { + "grad_norm": 0.3288372755050659, + "learning_rate": 7.896630488848654e-05, + "loss": 0.0074, + "step": 20290 + }, + { + "grad_norm": 0.23615920543670654, + "learning_rate": 7.89438382164471e-05, + "loss": 0.0071, + "step": 20300 + }, + { + "grad_norm": 0.3450459837913513, + "learning_rate": 7.892136275203383e-05, + "loss": 0.0108, + "step": 20310 + }, + { + "grad_norm": 0.23624712228775024, + "learning_rate": 7.889887850207418e-05, + "loss": 0.0082, + "step": 20320 + }, + { + "grad_norm": 0.22624339163303375, + "learning_rate": 7.887638547339827e-05, + "loss": 0.0088, + "step": 20330 + }, + { + "grad_norm": 0.21127496659755707, + "learning_rate": 7.885388367283891e-05, + "loss": 0.0083, + "step": 20340 + }, + { + "grad_norm": 0.24286557734012604, + "learning_rate": 7.88313731072315e-05, + "loss": 0.0101, + "step": 20350 + }, + { + "grad_norm": 0.2506529986858368, + "learning_rate": 7.88088537834142e-05, + "loss": 0.0093, + "step": 20360 + }, + { + "grad_norm": 0.238164484500885, + "learning_rate": 7.878632570822778e-05, + "loss": 0.0092, + "step": 20370 + }, + { + "grad_norm": 0.23400647938251495, + "learning_rate": 7.876378888851567e-05, + "loss": 0.0091, + "step": 20380 + }, + { + "grad_norm": 0.21203789114952087, + "learning_rate": 7.874124333112396e-05, + "loss": 0.0104, + "step": 20390 + }, + { + "grad_norm": 0.24512746930122375, + "learning_rate": 7.871868904290138e-05, + "loss": 0.0091, + "step": 20400 + }, + { + "grad_norm": 0.28832200169563293, + "learning_rate": 7.869612603069935e-05, + "loss": 0.009, + "step": 20410 + }, + { + "grad_norm": 0.24954824149608612, + "learning_rate": 7.867355430137192e-05, + "loss": 0.0059, + "step": 20420 + }, + { + "grad_norm": 0.17843236029148102, + "learning_rate": 7.865097386177577e-05, + "loss": 0.0057, + "step": 20430 + }, + { + "grad_norm": 0.21828846633434296, + "learning_rate": 7.862838471877023e-05, + "loss": 0.008, + "step": 20440 + }, + { + "grad_norm": 0.2021966427564621, + "learning_rate": 7.860578687921731e-05, + "loss": 0.0065, + "step": 20450 + }, + { + "grad_norm": 0.17848709225654602, + "learning_rate": 7.858318034998164e-05, + "loss": 0.0077, + "step": 20460 + }, + { + "grad_norm": 0.2623528242111206, + "learning_rate": 7.856056513793046e-05, + "loss": 0.0076, + "step": 20470 + }, + { + "grad_norm": 0.1966804414987564, + "learning_rate": 7.85379412499337e-05, + "loss": 0.0088, + "step": 20480 + }, + { + "grad_norm": 0.2642080783843994, + "learning_rate": 7.851530869286389e-05, + "loss": 0.0093, + "step": 20490 + }, + { + "grad_norm": 0.2431679666042328, + "learning_rate": 7.849266747359619e-05, + "loss": 0.0082, + "step": 20500 + }, + { + "grad_norm": 0.23049116134643555, + "learning_rate": 7.847001759900843e-05, + "loss": 0.0068, + "step": 20510 + }, + { + "grad_norm": 0.172549307346344, + "learning_rate": 7.844735907598102e-05, + "loss": 0.0112, + "step": 20520 + }, + { + "grad_norm": 0.22782915830612183, + "learning_rate": 7.842469191139703e-05, + "loss": 0.0105, + "step": 20530 + }, + { + "grad_norm": 0.22819611430168152, + "learning_rate": 7.840201611214215e-05, + "loss": 0.0064, + "step": 20540 + }, + { + "grad_norm": 0.2272072434425354, + "learning_rate": 7.837933168510469e-05, + "loss": 0.0088, + "step": 20550 + }, + { + "grad_norm": 0.21519364416599274, + "learning_rate": 7.835663863717559e-05, + "loss": 0.0097, + "step": 20560 + }, + { + "grad_norm": 0.3380639851093292, + "learning_rate": 7.833393697524838e-05, + "loss": 0.0079, + "step": 20570 + }, + { + "grad_norm": 0.32779932022094727, + "learning_rate": 7.831122670621922e-05, + "loss": 0.0101, + "step": 20580 + }, + { + "grad_norm": 0.26737362146377563, + "learning_rate": 7.82885078369869e-05, + "loss": 0.0077, + "step": 20590 + }, + { + "grad_norm": 0.1945435255765915, + "learning_rate": 7.826578037445283e-05, + "loss": 0.0068, + "step": 20600 + }, + { + "grad_norm": 0.22408734261989594, + "learning_rate": 7.824304432552097e-05, + "loss": 0.0074, + "step": 20610 + }, + { + "grad_norm": 0.17780053615570068, + "learning_rate": 7.822029969709798e-05, + "loss": 0.006, + "step": 20620 + }, + { + "grad_norm": 0.2417307198047638, + "learning_rate": 7.819754649609306e-05, + "loss": 0.0094, + "step": 20630 + }, + { + "grad_norm": 0.28193891048431396, + "learning_rate": 7.817478472941802e-05, + "loss": 0.0079, + "step": 20640 + }, + { + "grad_norm": 0.3269716501235962, + "learning_rate": 7.815201440398727e-05, + "loss": 0.0093, + "step": 20650 + }, + { + "grad_norm": 0.24986889958381653, + "learning_rate": 7.812923552671789e-05, + "loss": 0.0098, + "step": 20660 + }, + { + "grad_norm": 0.22790886461734772, + "learning_rate": 7.810644810452945e-05, + "loss": 0.0086, + "step": 20670 + }, + { + "grad_norm": 0.21844862401485443, + "learning_rate": 7.808365214434417e-05, + "loss": 0.0079, + "step": 20680 + }, + { + "grad_norm": 0.29999908804893494, + "learning_rate": 7.80608476530869e-05, + "loss": 0.0076, + "step": 20690 + }, + { + "grad_norm": 0.2469121366739273, + "learning_rate": 7.8038034637685e-05, + "loss": 0.009, + "step": 20700 + }, + { + "grad_norm": 0.22816407680511475, + "learning_rate": 7.801521310506848e-05, + "loss": 0.0076, + "step": 20710 + }, + { + "grad_norm": 0.23810826241970062, + "learning_rate": 7.799238306216994e-05, + "loss": 0.0079, + "step": 20720 + }, + { + "grad_norm": 0.27262425422668457, + "learning_rate": 7.796954451592448e-05, + "loss": 0.0095, + "step": 20730 + }, + { + "grad_norm": 0.2820528745651245, + "learning_rate": 7.794669747326992e-05, + "loss": 0.0107, + "step": 20740 + }, + { + "grad_norm": 0.2760917842388153, + "learning_rate": 7.792384194114654e-05, + "loss": 0.0085, + "step": 20750 + }, + { + "grad_norm": 0.21847929060459137, + "learning_rate": 7.790097792649729e-05, + "loss": 0.008, + "step": 20760 + }, + { + "grad_norm": 0.2442878782749176, + "learning_rate": 7.787810543626762e-05, + "loss": 0.0106, + "step": 20770 + }, + { + "grad_norm": 0.18228621780872345, + "learning_rate": 7.785522447740558e-05, + "loss": 0.0062, + "step": 20780 + }, + { + "grad_norm": 0.2158331573009491, + "learning_rate": 7.783233505686182e-05, + "loss": 0.007, + "step": 20790 + }, + { + "grad_norm": 0.2960889935493469, + "learning_rate": 7.780943718158955e-05, + "loss": 0.0078, + "step": 20800 + }, + { + "grad_norm": 0.1813492625951767, + "learning_rate": 7.778653085854453e-05, + "loss": 0.0064, + "step": 20810 + }, + { + "grad_norm": 0.21400456130504608, + "learning_rate": 7.77636160946851e-05, + "loss": 0.007, + "step": 20820 + }, + { + "grad_norm": 0.2015584409236908, + "learning_rate": 7.774069289697215e-05, + "loss": 0.0077, + "step": 20830 + }, + { + "grad_norm": 0.30142104625701904, + "learning_rate": 7.771776127236913e-05, + "loss": 0.0064, + "step": 20840 + }, + { + "grad_norm": 0.24986571073532104, + "learning_rate": 7.769482122784212e-05, + "loss": 0.007, + "step": 20850 + }, + { + "grad_norm": 0.2626262903213501, + "learning_rate": 7.767187277035963e-05, + "loss": 0.0077, + "step": 20860 + }, + { + "grad_norm": 0.28270116448402405, + "learning_rate": 7.764891590689285e-05, + "loss": 0.0074, + "step": 20870 + }, + { + "grad_norm": 0.22431358695030212, + "learning_rate": 7.762595064441542e-05, + "loss": 0.0077, + "step": 20880 + }, + { + "grad_norm": 0.22290213406085968, + "learning_rate": 7.760297698990362e-05, + "loss": 0.0071, + "step": 20890 + }, + { + "grad_norm": 0.19826994836330414, + "learning_rate": 7.757999495033623e-05, + "loss": 0.0064, + "step": 20900 + }, + { + "grad_norm": 0.2533501982688904, + "learning_rate": 7.755700453269456e-05, + "loss": 0.0081, + "step": 20910 + }, + { + "grad_norm": 0.21321673691272736, + "learning_rate": 7.753400574396254e-05, + "loss": 0.007, + "step": 20920 + }, + { + "grad_norm": 0.21878372132778168, + "learning_rate": 7.751099859112655e-05, + "loss": 0.0063, + "step": 20930 + }, + { + "grad_norm": 0.28217369318008423, + "learning_rate": 7.748798308117557e-05, + "loss": 0.01, + "step": 20940 + }, + { + "grad_norm": 0.26973801851272583, + "learning_rate": 7.746495922110112e-05, + "loss": 0.007, + "step": 20950 + }, + { + "grad_norm": 0.2012580782175064, + "learning_rate": 7.744192701789723e-05, + "loss": 0.0067, + "step": 20960 + }, + { + "grad_norm": 0.25700366497039795, + "learning_rate": 7.741888647856046e-05, + "loss": 0.0076, + "step": 20970 + }, + { + "grad_norm": 0.21343129873275757, + "learning_rate": 7.739583761008994e-05, + "loss": 0.0081, + "step": 20980 + }, + { + "grad_norm": 0.2504539489746094, + "learning_rate": 7.73727804194873e-05, + "loss": 0.0064, + "step": 20990 + }, + { + "grad_norm": 0.22741132974624634, + "learning_rate": 7.734971491375671e-05, + "loss": 0.0104, + "step": 21000 + }, + { + "grad_norm": 0.2709076702594757, + "learning_rate": 7.732664109990485e-05, + "loss": 0.0074, + "step": 21010 + }, + { + "grad_norm": 0.22615349292755127, + "learning_rate": 7.730355898494095e-05, + "loss": 0.0093, + "step": 21020 + }, + { + "grad_norm": 0.22514010965824127, + "learning_rate": 7.728046857587673e-05, + "loss": 0.0076, + "step": 21030 + }, + { + "grad_norm": 0.23027095198631287, + "learning_rate": 7.725736987972647e-05, + "loss": 0.0081, + "step": 21040 + }, + { + "grad_norm": 0.2386668622493744, + "learning_rate": 7.723426290350691e-05, + "loss": 0.0096, + "step": 21050 + }, + { + "grad_norm": 0.22972188889980316, + "learning_rate": 7.721114765423736e-05, + "loss": 0.0067, + "step": 21060 + }, + { + "grad_norm": 0.2910795211791992, + "learning_rate": 7.718802413893963e-05, + "loss": 0.01, + "step": 21070 + }, + { + "grad_norm": 0.2245515137910843, + "learning_rate": 7.716489236463802e-05, + "loss": 0.0092, + "step": 21080 + }, + { + "grad_norm": 0.20972386002540588, + "learning_rate": 7.714175233835936e-05, + "loss": 0.0077, + "step": 21090 + }, + { + "grad_norm": 0.2070750743150711, + "learning_rate": 7.711860406713299e-05, + "loss": 0.0076, + "step": 21100 + }, + { + "grad_norm": 0.28407394886016846, + "learning_rate": 7.70954475579907e-05, + "loss": 0.0093, + "step": 21110 + }, + { + "grad_norm": 0.23277485370635986, + "learning_rate": 7.707228281796688e-05, + "loss": 0.007, + "step": 21120 + }, + { + "grad_norm": 0.24423246085643768, + "learning_rate": 7.704910985409833e-05, + "loss": 0.0087, + "step": 21130 + }, + { + "grad_norm": 0.23047204315662384, + "learning_rate": 7.702592867342439e-05, + "loss": 0.0086, + "step": 21140 + }, + { + "grad_norm": 0.23932595551013947, + "learning_rate": 7.700273928298691e-05, + "loss": 0.0085, + "step": 21150 + }, + { + "grad_norm": 0.23834581673145294, + "learning_rate": 7.697954168983021e-05, + "loss": 0.0075, + "step": 21160 + }, + { + "grad_norm": 0.20849624276161194, + "learning_rate": 7.695633590100109e-05, + "loss": 0.0076, + "step": 21170 + }, + { + "grad_norm": 0.26982787251472473, + "learning_rate": 7.693312192354886e-05, + "loss": 0.0084, + "step": 21180 + }, + { + "grad_norm": 0.23794211447238922, + "learning_rate": 7.690989976452532e-05, + "loss": 0.0091, + "step": 21190 + }, + { + "grad_norm": 0.2747703194618225, + "learning_rate": 7.688666943098475e-05, + "loss": 0.008, + "step": 21200 + }, + { + "grad_norm": 0.2960923910140991, + "learning_rate": 7.686343092998389e-05, + "loss": 0.0075, + "step": 21210 + }, + { + "grad_norm": 0.2536032199859619, + "learning_rate": 7.684018426858202e-05, + "loss": 0.0077, + "step": 21220 + }, + { + "grad_norm": 0.258856862783432, + "learning_rate": 7.681692945384084e-05, + "loss": 0.0079, + "step": 21230 + }, + { + "grad_norm": 0.20317433774471283, + "learning_rate": 7.679366649282456e-05, + "loss": 0.0076, + "step": 21240 + }, + { + "grad_norm": 0.2373640388250351, + "learning_rate": 7.677039539259983e-05, + "loss": 0.0069, + "step": 21250 + }, + { + "grad_norm": 0.2372748702764511, + "learning_rate": 7.674711616023581e-05, + "loss": 0.0072, + "step": 21260 + }, + { + "grad_norm": 0.2482282519340515, + "learning_rate": 7.672382880280413e-05, + "loss": 0.0093, + "step": 21270 + }, + { + "grad_norm": 0.25534459948539734, + "learning_rate": 7.670053332737885e-05, + "loss": 0.0076, + "step": 21280 + }, + { + "grad_norm": 0.27775219082832336, + "learning_rate": 7.667722974103654e-05, + "loss": 0.0088, + "step": 21290 + }, + { + "grad_norm": 0.2930542528629303, + "learning_rate": 7.66539180508562e-05, + "loss": 0.0072, + "step": 21300 + }, + { + "grad_norm": 0.21638818085193634, + "learning_rate": 7.663059826391932e-05, + "loss": 0.0097, + "step": 21310 + }, + { + "grad_norm": 0.23934610188007355, + "learning_rate": 7.660727038730981e-05, + "loss": 0.0085, + "step": 21320 + }, + { + "grad_norm": 0.16057957708835602, + "learning_rate": 7.65839344281141e-05, + "loss": 0.0056, + "step": 21330 + }, + { + "grad_norm": 0.22868148982524872, + "learning_rate": 7.656059039342101e-05, + "loss": 0.0086, + "step": 21340 + }, + { + "grad_norm": 0.34593066573143005, + "learning_rate": 7.653723829032187e-05, + "loss": 0.0084, + "step": 21350 + }, + { + "grad_norm": 0.3157254457473755, + "learning_rate": 7.65138781259104e-05, + "loss": 0.0072, + "step": 21360 + }, + { + "grad_norm": 0.2019934505224228, + "learning_rate": 7.649050990728279e-05, + "loss": 0.0068, + "step": 21370 + }, + { + "grad_norm": 0.17256787419319153, + "learning_rate": 7.646713364153774e-05, + "loss": 0.0065, + "step": 21380 + }, + { + "grad_norm": 0.18384099006652832, + "learning_rate": 7.64437493357763e-05, + "loss": 0.0087, + "step": 21390 + }, + { + "grad_norm": 0.2063303291797638, + "learning_rate": 7.642035699710202e-05, + "loss": 0.0069, + "step": 21400 + }, + { + "grad_norm": 0.21035006642341614, + "learning_rate": 7.639695663262089e-05, + "loss": 0.0077, + "step": 21410 + }, + { + "grad_norm": 0.2264682501554489, + "learning_rate": 7.637354824944128e-05, + "loss": 0.008, + "step": 21420 + }, + { + "grad_norm": 0.2054148018360138, + "learning_rate": 7.635013185467408e-05, + "loss": 0.0068, + "step": 21430 + }, + { + "grad_norm": 0.2782699763774872, + "learning_rate": 7.632670745543256e-05, + "loss": 0.0086, + "step": 21440 + }, + { + "grad_norm": 0.26364076137542725, + "learning_rate": 7.630327505883242e-05, + "loss": 0.0106, + "step": 21450 + }, + { + "grad_norm": 0.18610930442810059, + "learning_rate": 7.627983467199182e-05, + "loss": 0.0065, + "step": 21460 + }, + { + "grad_norm": 0.2130100131034851, + "learning_rate": 7.625638630203132e-05, + "loss": 0.0062, + "step": 21470 + }, + { + "grad_norm": 0.2408430278301239, + "learning_rate": 7.623292995607394e-05, + "loss": 0.0073, + "step": 21480 + }, + { + "grad_norm": 0.21088026463985443, + "learning_rate": 7.620946564124507e-05, + "loss": 0.0075, + "step": 21490 + }, + { + "grad_norm": 0.1982228010892868, + "learning_rate": 7.618599336467256e-05, + "loss": 0.009, + "step": 21500 + }, + { + "grad_norm": 0.20202748477458954, + "learning_rate": 7.616251313348666e-05, + "loss": 0.0062, + "step": 21510 + }, + { + "grad_norm": 0.30146968364715576, + "learning_rate": 7.613902495482005e-05, + "loss": 0.0073, + "step": 21520 + }, + { + "grad_norm": 0.21921776235103607, + "learning_rate": 7.611552883580784e-05, + "loss": 0.0081, + "step": 21530 + }, + { + "grad_norm": 0.1917729377746582, + "learning_rate": 7.609202478358748e-05, + "loss": 0.0084, + "step": 21540 + }, + { + "grad_norm": 0.2060953825712204, + "learning_rate": 7.606851280529895e-05, + "loss": 0.0069, + "step": 21550 + }, + { + "grad_norm": 0.2721261978149414, + "learning_rate": 7.604499290808449e-05, + "loss": 0.007, + "step": 21560 + }, + { + "grad_norm": 0.18917472660541534, + "learning_rate": 7.602146509908888e-05, + "loss": 0.0064, + "step": 21570 + }, + { + "grad_norm": 0.25411972403526306, + "learning_rate": 7.599792938545921e-05, + "loss": 0.0095, + "step": 21580 + }, + { + "grad_norm": 0.23172226548194885, + "learning_rate": 7.597438577434506e-05, + "loss": 0.0088, + "step": 21590 + }, + { + "grad_norm": 0.21325945854187012, + "learning_rate": 7.595083427289831e-05, + "loss": 0.0092, + "step": 21600 + }, + { + "grad_norm": 0.20827534794807434, + "learning_rate": 7.59272748882733e-05, + "loss": 0.0077, + "step": 21610 + }, + { + "grad_norm": 0.19146892428398132, + "learning_rate": 7.590370762762675e-05, + "loss": 0.0077, + "step": 21620 + }, + { + "grad_norm": 0.20657746493816376, + "learning_rate": 7.588013249811777e-05, + "loss": 0.0081, + "step": 21630 + }, + { + "grad_norm": 0.24334716796875, + "learning_rate": 7.585654950690786e-05, + "loss": 0.0076, + "step": 21640 + }, + { + "grad_norm": 0.23545455932617188, + "learning_rate": 7.583295866116091e-05, + "loss": 0.0061, + "step": 21650 + }, + { + "grad_norm": 0.23728017508983612, + "learning_rate": 7.580935996804321e-05, + "loss": 0.0074, + "step": 21660 + }, + { + "grad_norm": 0.22799402475357056, + "learning_rate": 7.57857534347234e-05, + "loss": 0.0076, + "step": 21670 + }, + { + "grad_norm": 0.28084051609039307, + "learning_rate": 7.576213906837254e-05, + "loss": 0.0084, + "step": 21680 + }, + { + "grad_norm": 0.2438400685787201, + "learning_rate": 7.573851687616403e-05, + "loss": 0.0078, + "step": 21690 + }, + { + "grad_norm": 0.17609958350658417, + "learning_rate": 7.571488686527368e-05, + "loss": 0.0087, + "step": 21700 + }, + { + "grad_norm": 0.2439611256122589, + "learning_rate": 7.569124904287968e-05, + "loss": 0.0088, + "step": 21710 + }, + { + "grad_norm": 0.22961488366127014, + "learning_rate": 7.566760341616254e-05, + "loss": 0.0118, + "step": 21720 + }, + { + "grad_norm": 0.23086677491664886, + "learning_rate": 7.564394999230519e-05, + "loss": 0.0101, + "step": 21730 + }, + { + "grad_norm": 0.27812400460243225, + "learning_rate": 7.562028877849294e-05, + "loss": 0.0068, + "step": 21740 + }, + { + "grad_norm": 0.26642799377441406, + "learning_rate": 7.559661978191341e-05, + "loss": 0.0075, + "step": 21750 + }, + { + "grad_norm": 0.18844999372959137, + "learning_rate": 7.557294300975664e-05, + "loss": 0.0083, + "step": 21760 + }, + { + "grad_norm": 0.23796232044696808, + "learning_rate": 7.554925846921499e-05, + "loss": 0.009, + "step": 21770 + }, + { + "grad_norm": 0.27837589383125305, + "learning_rate": 7.552556616748321e-05, + "loss": 0.0071, + "step": 21780 + }, + { + "grad_norm": 0.24125047028064728, + "learning_rate": 7.550186611175838e-05, + "loss": 0.0066, + "step": 21790 + }, + { + "grad_norm": 0.17244909703731537, + "learning_rate": 7.547815830923998e-05, + "loss": 0.007, + "step": 21800 + }, + { + "grad_norm": 0.27542996406555176, + "learning_rate": 7.54544427671298e-05, + "loss": 0.0081, + "step": 21810 + }, + { + "grad_norm": 0.22847610712051392, + "learning_rate": 7.543071949263198e-05, + "loss": 0.006, + "step": 21820 + }, + { + "grad_norm": 0.2737162113189697, + "learning_rate": 7.540698849295305e-05, + "loss": 0.0078, + "step": 21830 + }, + { + "grad_norm": 0.2757057547569275, + "learning_rate": 7.538324977530183e-05, + "loss": 0.007, + "step": 21840 + }, + { + "grad_norm": 0.26891055703163147, + "learning_rate": 7.535950334688955e-05, + "loss": 0.0094, + "step": 21850 + }, + { + "grad_norm": 0.14977210760116577, + "learning_rate": 7.533574921492972e-05, + "loss": 0.0077, + "step": 21860 + }, + { + "grad_norm": 0.32933393120765686, + "learning_rate": 7.531198738663824e-05, + "loss": 0.0085, + "step": 21870 + }, + { + "grad_norm": 0.24526238441467285, + "learning_rate": 7.528821786923333e-05, + "loss": 0.0069, + "step": 21880 + }, + { + "grad_norm": 0.2873361110687256, + "learning_rate": 7.52644406699355e-05, + "loss": 0.0074, + "step": 21890 + }, + { + "grad_norm": 0.25182682275772095, + "learning_rate": 7.524065579596766e-05, + "loss": 0.0091, + "step": 21900 + }, + { + "grad_norm": 0.2680301368236542, + "learning_rate": 7.521686325455506e-05, + "loss": 0.0076, + "step": 21910 + }, + { + "grad_norm": 0.24315355718135834, + "learning_rate": 7.51930630529252e-05, + "loss": 0.0071, + "step": 21920 + }, + { + "grad_norm": 0.22277019917964935, + "learning_rate": 7.516925519830797e-05, + "loss": 0.0081, + "step": 21930 + }, + { + "grad_norm": 0.231625497341156, + "learning_rate": 7.514543969793557e-05, + "loss": 0.0078, + "step": 21940 + }, + { + "grad_norm": 0.2680265009403229, + "learning_rate": 7.512161655904251e-05, + "loss": 0.0058, + "step": 21950 + }, + { + "grad_norm": 0.19067281484603882, + "learning_rate": 7.509778578886563e-05, + "loss": 0.0067, + "step": 21960 + }, + { + "grad_norm": 0.26889294385910034, + "learning_rate": 7.507394739464412e-05, + "loss": 0.0092, + "step": 21970 + }, + { + "grad_norm": 0.2563711702823639, + "learning_rate": 7.50501013836194e-05, + "loss": 0.009, + "step": 21980 + }, + { + "grad_norm": 0.2342793047428131, + "learning_rate": 7.50262477630353e-05, + "loss": 0.0075, + "step": 21990 + }, + { + "grad_norm": 0.28217384219169617, + "learning_rate": 7.500238654013794e-05, + "loss": 0.0061, + "step": 22000 + }, + { + "grad_norm": 0.22224083542823792, + "learning_rate": 7.497851772217566e-05, + "loss": 0.0074, + "step": 22010 + }, + { + "grad_norm": 0.15639682114124298, + "learning_rate": 7.495464131639924e-05, + "loss": 0.0072, + "step": 22020 + }, + { + "grad_norm": 0.21873731911182404, + "learning_rate": 7.493075733006166e-05, + "loss": 0.0063, + "step": 22030 + }, + { + "grad_norm": 0.2252448946237564, + "learning_rate": 7.490686577041828e-05, + "loss": 0.0076, + "step": 22040 + }, + { + "grad_norm": 0.21436147391796112, + "learning_rate": 7.488296664472668e-05, + "loss": 0.0064, + "step": 22050 + }, + { + "grad_norm": 0.24536632001399994, + "learning_rate": 7.485905996024682e-05, + "loss": 0.006, + "step": 22060 + }, + { + "grad_norm": 0.30665984749794006, + "learning_rate": 7.483514572424093e-05, + "loss": 0.0075, + "step": 22070 + }, + { + "grad_norm": 0.1688169538974762, + "learning_rate": 7.481122394397349e-05, + "loss": 0.0065, + "step": 22080 + }, + { + "grad_norm": 0.27143412828445435, + "learning_rate": 7.478729462671131e-05, + "loss": 0.0065, + "step": 22090 + }, + { + "grad_norm": 0.2068139612674713, + "learning_rate": 7.47633577797235e-05, + "loss": 0.0105, + "step": 22100 + }, + { + "grad_norm": 0.2206212729215622, + "learning_rate": 7.473941341028144e-05, + "loss": 0.0072, + "step": 22110 + }, + { + "grad_norm": 0.21949850022792816, + "learning_rate": 7.471546152565879e-05, + "loss": 0.0077, + "step": 22120 + }, + { + "grad_norm": 0.2864134609699249, + "learning_rate": 7.46915021331315e-05, + "loss": 0.0106, + "step": 22130 + }, + { + "grad_norm": 0.2223970741033554, + "learning_rate": 7.466753523997778e-05, + "loss": 0.008, + "step": 22140 + }, + { + "grad_norm": 0.1761765331029892, + "learning_rate": 7.464356085347819e-05, + "loss": 0.0073, + "step": 22150 + }, + { + "grad_norm": 0.23056161403656006, + "learning_rate": 7.461957898091548e-05, + "loss": 0.0068, + "step": 22160 + }, + { + "grad_norm": 0.22583626210689545, + "learning_rate": 7.459558962957473e-05, + "loss": 0.0073, + "step": 22170 + }, + { + "grad_norm": 0.20231269299983978, + "learning_rate": 7.457159280674326e-05, + "loss": 0.0081, + "step": 22180 + }, + { + "grad_norm": 0.2062925100326538, + "learning_rate": 7.454758851971066e-05, + "loss": 0.0068, + "step": 22190 + }, + { + "grad_norm": 0.24801717698574066, + "learning_rate": 7.45235767757688e-05, + "loss": 0.0076, + "step": 22200 + }, + { + "grad_norm": 0.22194969654083252, + "learning_rate": 7.449955758221183e-05, + "loss": 0.0079, + "step": 22210 + }, + { + "grad_norm": 0.27872583270072937, + "learning_rate": 7.447553094633615e-05, + "loss": 0.009, + "step": 22220 + }, + { + "grad_norm": 0.26207977533340454, + "learning_rate": 7.445149687544039e-05, + "loss": 0.0071, + "step": 22230 + }, + { + "grad_norm": 0.17819800972938538, + "learning_rate": 7.44274553768255e-05, + "loss": 0.0076, + "step": 22240 + }, + { + "grad_norm": 0.1804770529270172, + "learning_rate": 7.440340645779464e-05, + "loss": 0.0117, + "step": 22250 + }, + { + "grad_norm": 0.23417186737060547, + "learning_rate": 7.437935012565322e-05, + "loss": 0.0088, + "step": 22260 + }, + { + "grad_norm": 0.26907482743263245, + "learning_rate": 7.435528638770893e-05, + "loss": 0.0067, + "step": 22270 + }, + { + "grad_norm": 0.23220942914485931, + "learning_rate": 7.433121525127171e-05, + "loss": 0.0074, + "step": 22280 + }, + { + "grad_norm": 0.18864183127880096, + "learning_rate": 7.430713672365371e-05, + "loss": 0.0074, + "step": 22290 + }, + { + "grad_norm": 0.21367451548576355, + "learning_rate": 7.428305081216938e-05, + "loss": 0.0064, + "step": 22300 + }, + { + "grad_norm": 0.24732628464698792, + "learning_rate": 7.425895752413536e-05, + "loss": 0.0076, + "step": 22310 + }, + { + "grad_norm": 0.23585957288742065, + "learning_rate": 7.423485686687057e-05, + "loss": 0.0069, + "step": 22320 + }, + { + "grad_norm": 0.25474631786346436, + "learning_rate": 7.421074884769616e-05, + "loss": 0.0104, + "step": 22330 + }, + { + "grad_norm": 0.22462978959083557, + "learning_rate": 7.418663347393548e-05, + "loss": 0.0078, + "step": 22340 + }, + { + "grad_norm": 0.269584059715271, + "learning_rate": 7.416251075291418e-05, + "loss": 0.0072, + "step": 22350 + }, + { + "grad_norm": 0.2947555482387543, + "learning_rate": 7.413838069196007e-05, + "loss": 0.0071, + "step": 22360 + }, + { + "grad_norm": 0.2860237658023834, + "learning_rate": 7.411424329840324e-05, + "loss": 0.0074, + "step": 22370 + }, + { + "grad_norm": 0.18522831797599792, + "learning_rate": 7.409009857957601e-05, + "loss": 0.0065, + "step": 22380 + }, + { + "grad_norm": 0.22943225502967834, + "learning_rate": 7.40659465428129e-05, + "loss": 0.0091, + "step": 22390 + }, + { + "grad_norm": 0.25011688470840454, + "learning_rate": 7.404178719545063e-05, + "loss": 0.009, + "step": 22400 + }, + { + "grad_norm": 0.2632109820842743, + "learning_rate": 7.401762054482822e-05, + "loss": 0.0072, + "step": 22410 + }, + { + "grad_norm": 0.24603308737277985, + "learning_rate": 7.39934465982868e-05, + "loss": 0.0089, + "step": 22420 + }, + { + "grad_norm": 0.20079408586025238, + "learning_rate": 7.396926536316984e-05, + "loss": 0.0087, + "step": 22430 + }, + { + "grad_norm": 0.24733145534992218, + "learning_rate": 7.394507684682293e-05, + "loss": 0.0089, + "step": 22440 + }, + { + "grad_norm": 0.24166692793369293, + "learning_rate": 7.392088105659393e-05, + "loss": 0.0071, + "step": 22450 + }, + { + "grad_norm": 0.2011084109544754, + "learning_rate": 7.389667799983284e-05, + "loss": 0.0069, + "step": 22460 + }, + { + "grad_norm": 0.2692710757255554, + "learning_rate": 7.387246768389193e-05, + "loss": 0.0068, + "step": 22470 + }, + { + "grad_norm": 0.20514404773712158, + "learning_rate": 7.384825011612563e-05, + "loss": 0.0061, + "step": 22480 + }, + { + "grad_norm": 0.17967849969863892, + "learning_rate": 7.382402530389066e-05, + "loss": 0.0066, + "step": 22490 + }, + { + "grad_norm": 0.20874296128749847, + "learning_rate": 7.379979325454582e-05, + "loss": 0.0079, + "step": 22500 + }, + { + "grad_norm": 0.2606438398361206, + "learning_rate": 7.37755539754522e-05, + "loss": 0.007, + "step": 22510 + }, + { + "grad_norm": 0.25000566244125366, + "learning_rate": 7.375130747397302e-05, + "loss": 0.0069, + "step": 22520 + }, + { + "grad_norm": 0.2225477546453476, + "learning_rate": 7.372705375747377e-05, + "loss": 0.0078, + "step": 22530 + }, + { + "grad_norm": 0.22811853885650635, + "learning_rate": 7.370279283332205e-05, + "loss": 0.0084, + "step": 22540 + }, + { + "grad_norm": 0.1927633285522461, + "learning_rate": 7.36785247088877e-05, + "loss": 0.0066, + "step": 22550 + }, + { + "grad_norm": 0.2659134566783905, + "learning_rate": 7.365424939154275e-05, + "loss": 0.0077, + "step": 22560 + }, + { + "grad_norm": 0.1849370151758194, + "learning_rate": 7.362996688866138e-05, + "loss": 0.0073, + "step": 22570 + }, + { + "grad_norm": 0.22637829184532166, + "learning_rate": 7.360567720761999e-05, + "loss": 0.0061, + "step": 22580 + }, + { + "grad_norm": 0.22359804809093475, + "learning_rate": 7.358138035579711e-05, + "loss": 0.0059, + "step": 22590 + }, + { + "grad_norm": 0.21463163197040558, + "learning_rate": 7.355707634057354e-05, + "loss": 0.0069, + "step": 22600 + }, + { + "grad_norm": 0.21173028647899628, + "learning_rate": 7.353276516933215e-05, + "loss": 0.0071, + "step": 22610 + }, + { + "grad_norm": 0.166578009724617, + "learning_rate": 7.350844684945806e-05, + "loss": 0.0056, + "step": 22620 + }, + { + "grad_norm": 0.23739005625247955, + "learning_rate": 7.348412138833851e-05, + "loss": 0.006, + "step": 22630 + }, + { + "grad_norm": 0.24811996519565582, + "learning_rate": 7.345978879336295e-05, + "loss": 0.0077, + "step": 22640 + }, + { + "grad_norm": 0.2003709226846695, + "learning_rate": 7.343544907192296e-05, + "loss": 0.0096, + "step": 22650 + }, + { + "grad_norm": 0.2116670310497284, + "learning_rate": 7.341110223141235e-05, + "loss": 0.0063, + "step": 22660 + }, + { + "grad_norm": 0.2476043403148651, + "learning_rate": 7.3386748279227e-05, + "loss": 0.0101, + "step": 22670 + }, + { + "grad_norm": 0.2581193447113037, + "learning_rate": 7.336238722276501e-05, + "loss": 0.008, + "step": 22680 + }, + { + "grad_norm": 0.22614417970180511, + "learning_rate": 7.333801906942663e-05, + "loss": 0.0078, + "step": 22690 + }, + { + "grad_norm": 0.20376884937286377, + "learning_rate": 7.331364382661428e-05, + "loss": 0.0077, + "step": 22700 + }, + { + "grad_norm": 0.2353021800518036, + "learning_rate": 7.328926150173248e-05, + "loss": 0.0084, + "step": 22710 + }, + { + "grad_norm": 0.2068241536617279, + "learning_rate": 7.326487210218795e-05, + "loss": 0.0069, + "step": 22720 + }, + { + "grad_norm": 0.14977042376995087, + "learning_rate": 7.324047563538955e-05, + "loss": 0.0062, + "step": 22730 + }, + { + "grad_norm": 0.21605587005615234, + "learning_rate": 7.321607210874828e-05, + "loss": 0.0082, + "step": 22740 + }, + { + "grad_norm": 0.21768251061439514, + "learning_rate": 7.31916615296773e-05, + "loss": 0.0075, + "step": 22750 + }, + { + "grad_norm": 0.22607772052288055, + "learning_rate": 7.316724390559188e-05, + "loss": 0.0067, + "step": 22760 + }, + { + "grad_norm": 0.20838750898838043, + "learning_rate": 7.314281924390946e-05, + "loss": 0.0082, + "step": 22770 + }, + { + "grad_norm": 0.23569554090499878, + "learning_rate": 7.311838755204959e-05, + "loss": 0.0068, + "step": 22780 + }, + { + "grad_norm": 0.25966617465019226, + "learning_rate": 7.3093948837434e-05, + "loss": 0.0073, + "step": 22790 + }, + { + "grad_norm": 0.2056553214788437, + "learning_rate": 7.306950310748651e-05, + "loss": 0.0072, + "step": 22800 + }, + { + "grad_norm": 0.26063889265060425, + "learning_rate": 7.304505036963311e-05, + "loss": 0.01, + "step": 22810 + }, + { + "grad_norm": 0.2838689088821411, + "learning_rate": 7.302059063130186e-05, + "loss": 0.0061, + "step": 22820 + }, + { + "grad_norm": 0.20723405480384827, + "learning_rate": 7.2996123899923e-05, + "loss": 0.0078, + "step": 22830 + }, + { + "grad_norm": 0.3118131160736084, + "learning_rate": 7.297165018292886e-05, + "loss": 0.0069, + "step": 22840 + }, + { + "grad_norm": 0.24340879917144775, + "learning_rate": 7.294716948775396e-05, + "loss": 0.0083, + "step": 22850 + }, + { + "grad_norm": 0.34460562467575073, + "learning_rate": 7.292268182183484e-05, + "loss": 0.0074, + "step": 22860 + }, + { + "grad_norm": 0.2676179111003876, + "learning_rate": 7.28981871926102e-05, + "loss": 0.0072, + "step": 22870 + }, + { + "grad_norm": 0.23573555052280426, + "learning_rate": 7.28736856075209e-05, + "loss": 0.0079, + "step": 22880 + }, + { + "grad_norm": 0.20188726484775543, + "learning_rate": 7.284917707400985e-05, + "loss": 0.007, + "step": 22890 + }, + { + "grad_norm": 0.22615191340446472, + "learning_rate": 7.282466159952212e-05, + "loss": 0.0084, + "step": 22900 + }, + { + "grad_norm": 0.2418040782213211, + "learning_rate": 7.280013919150483e-05, + "loss": 0.0073, + "step": 22910 + }, + { + "grad_norm": 0.20078390836715698, + "learning_rate": 7.277560985740728e-05, + "loss": 0.008, + "step": 22920 + }, + { + "grad_norm": 0.2556045353412628, + "learning_rate": 7.275107360468079e-05, + "loss": 0.0066, + "step": 22930 + }, + { + "grad_norm": 0.23398616909980774, + "learning_rate": 7.272653044077885e-05, + "loss": 0.0079, + "step": 22940 + }, + { + "grad_norm": 0.2393350750207901, + "learning_rate": 7.270198037315703e-05, + "loss": 0.0077, + "step": 22950 + }, + { + "grad_norm": 0.26894935965538025, + "learning_rate": 7.267742340927297e-05, + "loss": 0.009, + "step": 22960 + }, + { + "grad_norm": 0.3006245195865631, + "learning_rate": 7.265285955658645e-05, + "loss": 0.0073, + "step": 22970 + }, + { + "grad_norm": 0.2342676818370819, + "learning_rate": 7.26282888225593e-05, + "loss": 0.006, + "step": 22980 + }, + { + "grad_norm": 0.21574576199054718, + "learning_rate": 7.260371121465548e-05, + "loss": 0.006, + "step": 22990 + }, + { + "grad_norm": 0.2520429790019989, + "learning_rate": 7.2579126740341e-05, + "loss": 0.0079, + "step": 23000 + }, + { + "grad_norm": 0.23564472794532776, + "learning_rate": 7.2554535407084e-05, + "loss": 0.0066, + "step": 23010 + }, + { + "grad_norm": 0.19552455842494965, + "learning_rate": 7.252993722235464e-05, + "loss": 0.0068, + "step": 23020 + }, + { + "grad_norm": 0.207731693983078, + "learning_rate": 7.250533219362523e-05, + "loss": 0.006, + "step": 23030 + }, + { + "grad_norm": 0.22337056696414948, + "learning_rate": 7.248072032837012e-05, + "loss": 0.0072, + "step": 23040 + }, + { + "grad_norm": 0.19904747605323792, + "learning_rate": 7.245610163406575e-05, + "loss": 0.0058, + "step": 23050 + }, + { + "grad_norm": 0.26901039481163025, + "learning_rate": 7.243147611819061e-05, + "loss": 0.0075, + "step": 23060 + }, + { + "grad_norm": 0.17573612928390503, + "learning_rate": 7.240684378822531e-05, + "loss": 0.0075, + "step": 23070 + }, + { + "grad_norm": 0.23055042326450348, + "learning_rate": 7.238220465165248e-05, + "loss": 0.0087, + "step": 23080 + }, + { + "grad_norm": 0.21898643672466278, + "learning_rate": 7.235755871595684e-05, + "loss": 0.0074, + "step": 23090 + }, + { + "grad_norm": 0.20287759602069855, + "learning_rate": 7.233290598862517e-05, + "loss": 0.0074, + "step": 23100 + }, + { + "grad_norm": 0.19096684455871582, + "learning_rate": 7.230824647714635e-05, + "loss": 0.0064, + "step": 23110 + }, + { + "grad_norm": 0.14748680591583252, + "learning_rate": 7.228358018901124e-05, + "loss": 0.0057, + "step": 23120 + }, + { + "grad_norm": 0.20211993157863617, + "learning_rate": 7.225890713171286e-05, + "loss": 0.0065, + "step": 23130 + }, + { + "grad_norm": 0.17459122836589813, + "learning_rate": 7.223422731274618e-05, + "loss": 0.0059, + "step": 23140 + }, + { + "grad_norm": 0.21621543169021606, + "learning_rate": 7.220954073960832e-05, + "loss": 0.0073, + "step": 23150 + }, + { + "grad_norm": 0.2575087249279022, + "learning_rate": 7.218484741979838e-05, + "loss": 0.0067, + "step": 23160 + }, + { + "grad_norm": 0.22043457627296448, + "learning_rate": 7.216014736081756e-05, + "loss": 0.0068, + "step": 23170 + }, + { + "grad_norm": 0.21308642625808716, + "learning_rate": 7.213544057016906e-05, + "loss": 0.0063, + "step": 23180 + }, + { + "grad_norm": 0.31472212076187134, + "learning_rate": 7.211072705535819e-05, + "loss": 0.0098, + "step": 23190 + }, + { + "grad_norm": 0.22318722307682037, + "learning_rate": 7.208600682389224e-05, + "loss": 0.0111, + "step": 23200 + }, + { + "grad_norm": 0.23881292343139648, + "learning_rate": 7.206127988328055e-05, + "loss": 0.0062, + "step": 23210 + }, + { + "grad_norm": 0.23514747619628906, + "learning_rate": 7.203654624103453e-05, + "loss": 0.0058, + "step": 23220 + }, + { + "grad_norm": 0.24831639230251312, + "learning_rate": 7.201180590466761e-05, + "loss": 0.0069, + "step": 23230 + }, + { + "grad_norm": 0.20711083710193634, + "learning_rate": 7.198705888169523e-05, + "loss": 0.0079, + "step": 23240 + }, + { + "grad_norm": 0.20910988748073578, + "learning_rate": 7.196230517963491e-05, + "loss": 0.0054, + "step": 23250 + }, + { + "grad_norm": 0.2589011490345001, + "learning_rate": 7.193754480600615e-05, + "loss": 0.0069, + "step": 23260 + }, + { + "grad_norm": 0.22907955944538116, + "learning_rate": 7.19127777683305e-05, + "loss": 0.0069, + "step": 23270 + }, + { + "grad_norm": 0.2453407645225525, + "learning_rate": 7.188800407413156e-05, + "loss": 0.0079, + "step": 23280 + }, + { + "grad_norm": 0.25670960545539856, + "learning_rate": 7.186322373093489e-05, + "loss": 0.0071, + "step": 23290 + }, + { + "grad_norm": 0.19476912915706635, + "learning_rate": 7.18384367462681e-05, + "loss": 0.007, + "step": 23300 + }, + { + "grad_norm": 0.2318083643913269, + "learning_rate": 7.181364312766085e-05, + "loss": 0.007, + "step": 23310 + }, + { + "grad_norm": 0.21937629580497742, + "learning_rate": 7.178884288264477e-05, + "loss": 0.0084, + "step": 23320 + }, + { + "grad_norm": 0.23231841623783112, + "learning_rate": 7.176403601875353e-05, + "loss": 0.0061, + "step": 23330 + }, + { + "grad_norm": 0.2002200335264206, + "learning_rate": 7.173922254352279e-05, + "loss": 0.0053, + "step": 23340 + }, + { + "grad_norm": 0.25092506408691406, + "learning_rate": 7.171440246449024e-05, + "loss": 0.0069, + "step": 23350 + }, + { + "grad_norm": 0.23472483456134796, + "learning_rate": 7.168957578919555e-05, + "loss": 0.0081, + "step": 23360 + }, + { + "grad_norm": 0.23603397607803345, + "learning_rate": 7.16647425251804e-05, + "loss": 0.0073, + "step": 23370 + }, + { + "grad_norm": 0.2890200614929199, + "learning_rate": 7.163990267998852e-05, + "loss": 0.0094, + "step": 23380 + }, + { + "grad_norm": 0.18937300145626068, + "learning_rate": 7.161505626116556e-05, + "loss": 0.0069, + "step": 23390 + }, + { + "grad_norm": 0.19407562911510468, + "learning_rate": 7.159020327625923e-05, + "loss": 0.0073, + "step": 23400 + }, + { + "grad_norm": 0.3111112415790558, + "learning_rate": 7.15653437328192e-05, + "loss": 0.0071, + "step": 23410 + }, + { + "grad_norm": 0.24332955479621887, + "learning_rate": 7.154047763839713e-05, + "loss": 0.0096, + "step": 23420 + }, + { + "grad_norm": 0.259840190410614, + "learning_rate": 7.15156050005467e-05, + "loss": 0.0097, + "step": 23430 + }, + { + "grad_norm": 0.20390920341014862, + "learning_rate": 7.149072582682357e-05, + "loss": 0.0064, + "step": 23440 + }, + { + "grad_norm": 0.2323862463235855, + "learning_rate": 7.146584012478535e-05, + "loss": 0.0084, + "step": 23450 + }, + { + "grad_norm": 0.2258659154176712, + "learning_rate": 7.144094790199169e-05, + "loss": 0.0085, + "step": 23460 + }, + { + "grad_norm": 0.18803656101226807, + "learning_rate": 7.141604916600415e-05, + "loss": 0.0076, + "step": 23470 + }, + { + "grad_norm": 0.2015160620212555, + "learning_rate": 7.139114392438635e-05, + "loss": 0.0062, + "step": 23480 + }, + { + "grad_norm": 0.2064451277256012, + "learning_rate": 7.136623218470382e-05, + "loss": 0.005, + "step": 23490 + }, + { + "grad_norm": 0.28598785400390625, + "learning_rate": 7.13413139545241e-05, + "loss": 0.0067, + "step": 23500 + }, + { + "grad_norm": 0.2777937352657318, + "learning_rate": 7.131638924141668e-05, + "loss": 0.008, + "step": 23510 + }, + { + "grad_norm": 0.3024005889892578, + "learning_rate": 7.129145805295304e-05, + "loss": 0.0081, + "step": 23520 + }, + { + "grad_norm": 0.20812225341796875, + "learning_rate": 7.126652039670661e-05, + "loss": 0.0082, + "step": 23530 + }, + { + "grad_norm": 0.30336740612983704, + "learning_rate": 7.124157628025278e-05, + "loss": 0.0088, + "step": 23540 + }, + { + "grad_norm": 0.26776477694511414, + "learning_rate": 7.121662571116894e-05, + "loss": 0.0071, + "step": 23550 + }, + { + "grad_norm": 0.21023081243038177, + "learning_rate": 7.119166869703441e-05, + "loss": 0.0061, + "step": 23560 + }, + { + "grad_norm": 0.2137153446674347, + "learning_rate": 7.116670524543044e-05, + "loss": 0.0068, + "step": 23570 + }, + { + "grad_norm": 0.2735556364059448, + "learning_rate": 7.114173536394032e-05, + "loss": 0.0098, + "step": 23580 + }, + { + "grad_norm": 0.16731102764606476, + "learning_rate": 7.111675906014917e-05, + "loss": 0.0077, + "step": 23590 + }, + { + "grad_norm": 0.23282283544540405, + "learning_rate": 7.109177634164421e-05, + "loss": 0.0064, + "step": 23600 + }, + { + "grad_norm": 0.14320172369480133, + "learning_rate": 7.106678721601449e-05, + "loss": 0.0057, + "step": 23610 + }, + { + "grad_norm": 0.19735263288021088, + "learning_rate": 7.104179169085103e-05, + "loss": 0.006, + "step": 23620 + }, + { + "grad_norm": 0.2534981667995453, + "learning_rate": 7.101678977374683e-05, + "loss": 0.0065, + "step": 23630 + }, + { + "grad_norm": 0.20142443478107452, + "learning_rate": 7.099178147229685e-05, + "loss": 0.0062, + "step": 23640 + }, + { + "grad_norm": 0.19346113502979279, + "learning_rate": 7.096676679409789e-05, + "loss": 0.006, + "step": 23650 + }, + { + "grad_norm": 0.24220320582389832, + "learning_rate": 7.094174574674877e-05, + "loss": 0.0062, + "step": 23660 + }, + { + "grad_norm": 0.23936009407043457, + "learning_rate": 7.091671833785025e-05, + "loss": 0.0067, + "step": 23670 + }, + { + "grad_norm": 0.25479021668434143, + "learning_rate": 7.089168457500493e-05, + "loss": 0.0058, + "step": 23680 + }, + { + "grad_norm": 0.22986072301864624, + "learning_rate": 7.086664446581747e-05, + "loss": 0.0063, + "step": 23690 + }, + { + "grad_norm": 0.21007581055164337, + "learning_rate": 7.084159801789438e-05, + "loss": 0.0063, + "step": 23700 + }, + { + "grad_norm": 0.19588488340377808, + "learning_rate": 7.081654523884411e-05, + "loss": 0.0056, + "step": 23710 + }, + { + "grad_norm": 0.23593758046627045, + "learning_rate": 7.0791486136277e-05, + "loss": 0.0075, + "step": 23720 + }, + { + "grad_norm": 0.2843397557735443, + "learning_rate": 7.07664207178054e-05, + "loss": 0.0065, + "step": 23730 + }, + { + "grad_norm": 0.2020701915025711, + "learning_rate": 7.074134899104345e-05, + "loss": 0.0085, + "step": 23740 + }, + { + "grad_norm": 0.20319929718971252, + "learning_rate": 7.071627096360735e-05, + "loss": 0.0066, + "step": 23750 + }, + { + "grad_norm": 0.2622431218624115, + "learning_rate": 7.069118664311511e-05, + "loss": 0.0077, + "step": 23760 + }, + { + "grad_norm": 0.2573256492614746, + "learning_rate": 7.06660960371867e-05, + "loss": 0.0062, + "step": 23770 + }, + { + "grad_norm": 0.21369421482086182, + "learning_rate": 7.064099915344396e-05, + "loss": 0.0068, + "step": 23780 + }, + { + "grad_norm": 0.2916682958602905, + "learning_rate": 7.061589599951066e-05, + "loss": 0.0071, + "step": 23790 + }, + { + "grad_norm": 0.30225372314453125, + "learning_rate": 7.05907865830125e-05, + "loss": 0.0071, + "step": 23800 + }, + { + "grad_norm": 0.27047479152679443, + "learning_rate": 7.056567091157703e-05, + "loss": 0.007, + "step": 23810 + }, + { + "grad_norm": 0.22229182720184326, + "learning_rate": 7.054054899283375e-05, + "loss": 0.0082, + "step": 23820 + }, + { + "grad_norm": 0.2269914299249649, + "learning_rate": 7.051542083441403e-05, + "loss": 0.006, + "step": 23830 + }, + { + "grad_norm": 0.2352418452501297, + "learning_rate": 7.049028644395113e-05, + "loss": 0.0074, + "step": 23840 + }, + { + "grad_norm": 0.24701836705207825, + "learning_rate": 7.046514582908024e-05, + "loss": 0.0069, + "step": 23850 + }, + { + "grad_norm": 0.26358070969581604, + "learning_rate": 7.043999899743838e-05, + "loss": 0.0057, + "step": 23860 + }, + { + "grad_norm": 0.2301684021949768, + "learning_rate": 7.041484595666451e-05, + "loss": 0.0071, + "step": 23870 + }, + { + "grad_norm": 0.21277542412281036, + "learning_rate": 7.038968671439948e-05, + "loss": 0.006, + "step": 23880 + }, + { + "grad_norm": 0.23770518600940704, + "learning_rate": 7.036452127828596e-05, + "loss": 0.0072, + "step": 23890 + }, + { + "grad_norm": 0.23463623225688934, + "learning_rate": 7.033934965596859e-05, + "loss": 0.0072, + "step": 23900 + }, + { + "grad_norm": 0.2334730327129364, + "learning_rate": 7.031417185509381e-05, + "loss": 0.0066, + "step": 23910 + }, + { + "grad_norm": 0.15661562979221344, + "learning_rate": 7.028898788331e-05, + "loss": 0.0063, + "step": 23920 + }, + { + "grad_norm": 0.21705128252506256, + "learning_rate": 7.026379774826736e-05, + "loss": 0.0058, + "step": 23930 + }, + { + "grad_norm": 0.19458024203777313, + "learning_rate": 7.0238601457618e-05, + "loss": 0.0059, + "step": 23940 + }, + { + "grad_norm": 0.2701191008090973, + "learning_rate": 7.02133990190159e-05, + "loss": 0.0067, + "step": 23950 + }, + { + "grad_norm": 0.23577691614627838, + "learning_rate": 7.018819044011687e-05, + "loss": 0.0078, + "step": 23960 + }, + { + "grad_norm": 0.15558412671089172, + "learning_rate": 7.016297572857863e-05, + "loss": 0.0053, + "step": 23970 + }, + { + "grad_norm": 0.18498459458351135, + "learning_rate": 7.013775489206072e-05, + "loss": 0.0051, + "step": 23980 + }, + { + "grad_norm": 0.2610701322555542, + "learning_rate": 7.01125279382246e-05, + "loss": 0.0057, + "step": 23990 + }, + { + "grad_norm": 0.2301194816827774, + "learning_rate": 7.008729487473351e-05, + "loss": 0.0066, + "step": 24000 + }, + { + "grad_norm": 0.1587142050266266, + "learning_rate": 7.006205570925263e-05, + "loss": 0.0061, + "step": 24010 + }, + { + "grad_norm": 0.2386852502822876, + "learning_rate": 7.003681044944892e-05, + "loss": 0.0081, + "step": 24020 + }, + { + "grad_norm": 0.18989433348178864, + "learning_rate": 7.001155910299126e-05, + "loss": 0.0061, + "step": 24030 + }, + { + "grad_norm": 0.2027624249458313, + "learning_rate": 6.99863016775503e-05, + "loss": 0.0068, + "step": 24040 + }, + { + "grad_norm": 0.253236323595047, + "learning_rate": 6.996103818079859e-05, + "loss": 0.0079, + "step": 24050 + }, + { + "grad_norm": 0.27774831652641296, + "learning_rate": 6.993576862041054e-05, + "loss": 0.0084, + "step": 24060 + }, + { + "grad_norm": 0.21400724351406097, + "learning_rate": 6.991049300406235e-05, + "loss": 0.0068, + "step": 24070 + }, + { + "grad_norm": 0.19312065839767456, + "learning_rate": 6.988521133943209e-05, + "loss": 0.0047, + "step": 24080 + }, + { + "grad_norm": 0.2008533626794815, + "learning_rate": 6.985992363419966e-05, + "loss": 0.0073, + "step": 24090 + }, + { + "grad_norm": 0.1768791675567627, + "learning_rate": 6.983462989604682e-05, + "loss": 0.0063, + "step": 24100 + }, + { + "grad_norm": 0.3028063178062439, + "learning_rate": 6.980933013265709e-05, + "loss": 0.0069, + "step": 24110 + }, + { + "grad_norm": 0.2480514645576477, + "learning_rate": 6.978402435171592e-05, + "loss": 0.0073, + "step": 24120 + }, + { + "grad_norm": 0.17038322985172272, + "learning_rate": 6.975871256091052e-05, + "loss": 0.0062, + "step": 24130 + }, + { + "grad_norm": 0.2071041464805603, + "learning_rate": 6.973339476792995e-05, + "loss": 0.0065, + "step": 24140 + }, + { + "grad_norm": 0.21248899400234222, + "learning_rate": 6.970807098046505e-05, + "loss": 0.006, + "step": 24150 + }, + { + "grad_norm": 0.2238602340221405, + "learning_rate": 6.968274120620858e-05, + "loss": 0.0074, + "step": 24160 + }, + { + "grad_norm": 0.22012795507907867, + "learning_rate": 6.965740545285499e-05, + "loss": 0.0054, + "step": 24170 + }, + { + "grad_norm": 0.29043570160865784, + "learning_rate": 6.963206372810068e-05, + "loss": 0.0083, + "step": 24180 + }, + { + "grad_norm": 0.2822639048099518, + "learning_rate": 6.960671603964375e-05, + "loss": 0.0087, + "step": 24190 + }, + { + "grad_norm": 0.19819259643554688, + "learning_rate": 6.958136239518418e-05, + "loss": 0.006, + "step": 24200 + }, + { + "grad_norm": 0.18711893260478973, + "learning_rate": 6.955600280242371e-05, + "loss": 0.0069, + "step": 24210 + }, + { + "grad_norm": 0.1911400556564331, + "learning_rate": 6.953063726906596e-05, + "loss": 0.0088, + "step": 24220 + }, + { + "grad_norm": 0.195345938205719, + "learning_rate": 6.950526580281626e-05, + "loss": 0.0085, + "step": 24230 + }, + { + "grad_norm": 0.18867981433868408, + "learning_rate": 6.947988841138184e-05, + "loss": 0.0063, + "step": 24240 + }, + { + "grad_norm": 0.21644961833953857, + "learning_rate": 6.945450510247165e-05, + "loss": 0.0065, + "step": 24250 + }, + { + "grad_norm": 0.2197287678718567, + "learning_rate": 6.942911588379647e-05, + "loss": 0.0064, + "step": 24260 + }, + { + "grad_norm": 0.24320431053638458, + "learning_rate": 6.940372076306888e-05, + "loss": 0.0058, + "step": 24270 + }, + { + "grad_norm": 0.1552775651216507, + "learning_rate": 6.937831974800326e-05, + "loss": 0.0066, + "step": 24280 + }, + { + "grad_norm": 0.2661644518375397, + "learning_rate": 6.935291284631574e-05, + "loss": 0.0087, + "step": 24290 + }, + { + "grad_norm": 0.2091807872056961, + "learning_rate": 6.932750006572428e-05, + "loss": 0.0072, + "step": 24300 + }, + { + "grad_norm": 0.2535192668437958, + "learning_rate": 6.930208141394863e-05, + "loss": 0.0078, + "step": 24310 + }, + { + "grad_norm": 0.23283711075782776, + "learning_rate": 6.927665689871026e-05, + "loss": 0.0062, + "step": 24320 + }, + { + "grad_norm": 0.24870415031909943, + "learning_rate": 6.925122652773253e-05, + "loss": 0.0064, + "step": 24330 + }, + { + "grad_norm": 0.3209875822067261, + "learning_rate": 6.922579030874046e-05, + "loss": 0.0073, + "step": 24340 + }, + { + "grad_norm": 0.25944098830223083, + "learning_rate": 6.920034824946093e-05, + "loss": 0.0059, + "step": 24350 + }, + { + "grad_norm": 0.2029070258140564, + "learning_rate": 6.917490035762255e-05, + "loss": 0.0077, + "step": 24360 + }, + { + "grad_norm": 0.16941291093826294, + "learning_rate": 6.914944664095573e-05, + "loss": 0.0079, + "step": 24370 + }, + { + "grad_norm": 0.20521265268325806, + "learning_rate": 6.912398710719264e-05, + "loss": 0.0069, + "step": 24380 + }, + { + "grad_norm": 0.25230786204338074, + "learning_rate": 6.90985217640672e-05, + "loss": 0.007, + "step": 24390 + }, + { + "grad_norm": 0.21439579129219055, + "learning_rate": 6.90730506193151e-05, + "loss": 0.0068, + "step": 24400 + }, + { + "grad_norm": 0.21030278503894806, + "learning_rate": 6.904757368067384e-05, + "loss": 0.0076, + "step": 24410 + }, + { + "grad_norm": 0.17052730917930603, + "learning_rate": 6.90220909558826e-05, + "loss": 0.0061, + "step": 24420 + }, + { + "grad_norm": 0.2733839750289917, + "learning_rate": 6.899660245268237e-05, + "loss": 0.0075, + "step": 24430 + }, + { + "grad_norm": 0.2304183542728424, + "learning_rate": 6.897110817881592e-05, + "loss": 0.0089, + "step": 24440 + }, + { + "grad_norm": 0.2562944293022156, + "learning_rate": 6.894560814202769e-05, + "loss": 0.0077, + "step": 24450 + }, + { + "grad_norm": 0.28432559967041016, + "learning_rate": 6.892010235006394e-05, + "loss": 0.0061, + "step": 24460 + }, + { + "grad_norm": 0.22223351895809174, + "learning_rate": 6.889459081067264e-05, + "loss": 0.0069, + "step": 24470 + }, + { + "grad_norm": 0.24769578874111176, + "learning_rate": 6.886907353160356e-05, + "loss": 0.0069, + "step": 24480 + }, + { + "grad_norm": 0.24927619099617004, + "learning_rate": 6.884355052060814e-05, + "loss": 0.0075, + "step": 24490 + }, + { + "grad_norm": 0.2924117147922516, + "learning_rate": 6.88180217854396e-05, + "loss": 0.0087, + "step": 24500 + }, + { + "grad_norm": 0.2673661708831787, + "learning_rate": 6.87924873338529e-05, + "loss": 0.0069, + "step": 24510 + }, + { + "grad_norm": 0.2333325296640396, + "learning_rate": 6.876694717360475e-05, + "loss": 0.0078, + "step": 24520 + }, + { + "grad_norm": 0.18996460735797882, + "learning_rate": 6.874140131245355e-05, + "loss": 0.0072, + "step": 24530 + }, + { + "grad_norm": 0.23210026323795319, + "learning_rate": 6.871584975815948e-05, + "loss": 0.0104, + "step": 24540 + }, + { + "grad_norm": 0.14410477876663208, + "learning_rate": 6.86902925184844e-05, + "loss": 0.0072, + "step": 24550 + }, + { + "grad_norm": 0.23743757605552673, + "learning_rate": 6.866472960119195e-05, + "loss": 0.0072, + "step": 24560 + }, + { + "grad_norm": 0.2170960158109665, + "learning_rate": 6.863916101404748e-05, + "loss": 0.0091, + "step": 24570 + }, + { + "grad_norm": 0.22801494598388672, + "learning_rate": 6.8613586764818e-05, + "loss": 0.006, + "step": 24580 + }, + { + "grad_norm": 0.24854330718517303, + "learning_rate": 6.858800686127233e-05, + "loss": 0.0062, + "step": 24590 + }, + { + "grad_norm": 0.2309127151966095, + "learning_rate": 6.856242131118097e-05, + "loss": 0.0088, + "step": 24600 + }, + { + "grad_norm": 0.20720410346984863, + "learning_rate": 6.853683012231614e-05, + "loss": 0.0065, + "step": 24610 + }, + { + "grad_norm": 0.2361244559288025, + "learning_rate": 6.851123330245173e-05, + "loss": 0.0083, + "step": 24620 + }, + { + "grad_norm": 0.20022918283939362, + "learning_rate": 6.848563085936343e-05, + "loss": 0.0062, + "step": 24630 + }, + { + "grad_norm": 0.15003590285778046, + "learning_rate": 6.846002280082853e-05, + "loss": 0.0055, + "step": 24640 + }, + { + "grad_norm": 0.16061264276504517, + "learning_rate": 6.843440913462614e-05, + "loss": 0.0064, + "step": 24650 + }, + { + "grad_norm": 0.2030072808265686, + "learning_rate": 6.840878986853698e-05, + "loss": 0.007, + "step": 24660 + }, + { + "grad_norm": 0.236202672123909, + "learning_rate": 6.838316501034352e-05, + "loss": 0.0066, + "step": 24670 + }, + { + "grad_norm": 0.22119833528995514, + "learning_rate": 6.83575345678299e-05, + "loss": 0.0077, + "step": 24680 + }, + { + "grad_norm": 0.2633749842643738, + "learning_rate": 6.833189854878196e-05, + "loss": 0.008, + "step": 24690 + }, + { + "grad_norm": 0.21641162037849426, + "learning_rate": 6.83062569609873e-05, + "loss": 0.0065, + "step": 24700 + }, + { + "grad_norm": 0.23379985988140106, + "learning_rate": 6.828060981223512e-05, + "loss": 0.006, + "step": 24710 + }, + { + "grad_norm": 0.20191489160060883, + "learning_rate": 6.825495711031634e-05, + "loss": 0.0091, + "step": 24720 + }, + { + "grad_norm": 0.21407240629196167, + "learning_rate": 6.822929886302359e-05, + "loss": 0.0063, + "step": 24730 + }, + { + "grad_norm": 0.25733667612075806, + "learning_rate": 6.820363507815116e-05, + "loss": 0.0075, + "step": 24740 + }, + { + "grad_norm": 0.22397202253341675, + "learning_rate": 6.817796576349501e-05, + "loss": 0.0069, + "step": 24750 + }, + { + "grad_norm": 0.2072632610797882, + "learning_rate": 6.815229092685285e-05, + "loss": 0.0092, + "step": 24760 + }, + { + "grad_norm": 0.1787474900484085, + "learning_rate": 6.812661057602399e-05, + "loss": 0.0056, + "step": 24770 + }, + { + "grad_norm": 0.20838628709316254, + "learning_rate": 6.810092471880943e-05, + "loss": 0.0063, + "step": 24780 + }, + { + "grad_norm": 0.20440411567687988, + "learning_rate": 6.807523336301187e-05, + "loss": 0.0075, + "step": 24790 + }, + { + "grad_norm": 0.20928245782852173, + "learning_rate": 6.804953651643566e-05, + "loss": 0.006, + "step": 24800 + }, + { + "grad_norm": 0.21724002063274384, + "learning_rate": 6.802383418688685e-05, + "loss": 0.0085, + "step": 24810 + }, + { + "grad_norm": 0.1721886396408081, + "learning_rate": 6.799812638217309e-05, + "loss": 0.0057, + "step": 24820 + }, + { + "grad_norm": 0.2006726711988449, + "learning_rate": 6.797241311010373e-05, + "loss": 0.0065, + "step": 24830 + }, + { + "grad_norm": 0.21389009058475494, + "learning_rate": 6.794669437848982e-05, + "loss": 0.0057, + "step": 24840 + }, + { + "grad_norm": 0.2328222244977951, + "learning_rate": 6.792097019514402e-05, + "loss": 0.0064, + "step": 24850 + }, + { + "grad_norm": 0.2139717936515808, + "learning_rate": 6.789524056788064e-05, + "loss": 0.0051, + "step": 24860 + }, + { + "grad_norm": 0.18213902413845062, + "learning_rate": 6.786950550451567e-05, + "loss": 0.0056, + "step": 24870 + }, + { + "grad_norm": 0.1845434010028839, + "learning_rate": 6.784376501286676e-05, + "loss": 0.0062, + "step": 24880 + }, + { + "grad_norm": 0.18680481612682343, + "learning_rate": 6.781801910075316e-05, + "loss": 0.0051, + "step": 24890 + }, + { + "grad_norm": 0.1739758849143982, + "learning_rate": 6.779226777599581e-05, + "loss": 0.0057, + "step": 24900 + }, + { + "grad_norm": 0.22760528326034546, + "learning_rate": 6.776651104641729e-05, + "loss": 0.0059, + "step": 24910 + }, + { + "grad_norm": 0.21769170463085175, + "learning_rate": 6.774074891984183e-05, + "loss": 0.0077, + "step": 24920 + }, + { + "grad_norm": 0.22711685299873352, + "learning_rate": 6.771498140409526e-05, + "loss": 0.0063, + "step": 24930 + }, + { + "grad_norm": 0.2919391095638275, + "learning_rate": 6.768920850700506e-05, + "loss": 0.0062, + "step": 24940 + }, + { + "grad_norm": 0.21893933415412903, + "learning_rate": 6.766343023640039e-05, + "loss": 0.0065, + "step": 24950 + }, + { + "grad_norm": 0.18296101689338684, + "learning_rate": 6.763764660011198e-05, + "loss": 0.007, + "step": 24960 + }, + { + "grad_norm": 0.19271834194660187, + "learning_rate": 6.761185760597223e-05, + "loss": 0.0058, + "step": 24970 + }, + { + "grad_norm": 0.33522406220436096, + "learning_rate": 6.758606326181515e-05, + "loss": 0.0087, + "step": 24980 + }, + { + "grad_norm": 0.23418518900871277, + "learning_rate": 6.75602635754764e-05, + "loss": 0.0072, + "step": 24990 + }, + { + "grad_norm": 0.23341238498687744, + "learning_rate": 6.75344585547932e-05, + "loss": 0.0073, + "step": 25000 + }, + { + "grad_norm": 0.22817090153694153, + "learning_rate": 6.750864820760449e-05, + "loss": 0.0059, + "step": 25010 + }, + { + "grad_norm": 0.20839819312095642, + "learning_rate": 6.748283254175072e-05, + "loss": 0.0056, + "step": 25020 + }, + { + "grad_norm": 0.18569742143154144, + "learning_rate": 6.745701156507404e-05, + "loss": 0.007, + "step": 25030 + }, + { + "grad_norm": 0.19279655814170837, + "learning_rate": 6.743118528541818e-05, + "loss": 0.0053, + "step": 25040 + }, + { + "grad_norm": 0.3393954336643219, + "learning_rate": 6.740535371062846e-05, + "loss": 0.0073, + "step": 25050 + }, + { + "grad_norm": 0.24617277085781097, + "learning_rate": 6.737951684855185e-05, + "loss": 0.0067, + "step": 25060 + }, + { + "grad_norm": 0.2199849784374237, + "learning_rate": 6.735367470703691e-05, + "loss": 0.0085, + "step": 25070 + }, + { + "grad_norm": 0.2632351517677307, + "learning_rate": 6.732782729393379e-05, + "loss": 0.0068, + "step": 25080 + }, + { + "grad_norm": 0.20831426978111267, + "learning_rate": 6.730197461709425e-05, + "loss": 0.006, + "step": 25090 + }, + { + "grad_norm": 0.2053185999393463, + "learning_rate": 6.727611668437164e-05, + "loss": 0.0072, + "step": 25100 + }, + { + "grad_norm": 0.19841746985912323, + "learning_rate": 6.725025350362094e-05, + "loss": 0.0063, + "step": 25110 + }, + { + "grad_norm": 0.2458447515964508, + "learning_rate": 6.72243850826987e-05, + "loss": 0.0074, + "step": 25120 + }, + { + "grad_norm": 0.19741925597190857, + "learning_rate": 6.719851142946305e-05, + "loss": 0.0067, + "step": 25130 + }, + { + "grad_norm": 0.21869440376758575, + "learning_rate": 6.717263255177372e-05, + "loss": 0.0081, + "step": 25140 + }, + { + "grad_norm": 0.250583291053772, + "learning_rate": 6.714674845749205e-05, + "loss": 0.0064, + "step": 25150 + }, + { + "grad_norm": 0.20444005727767944, + "learning_rate": 6.712085915448092e-05, + "loss": 0.0058, + "step": 25160 + }, + { + "grad_norm": 0.21273529529571533, + "learning_rate": 6.709496465060486e-05, + "loss": 0.0059, + "step": 25170 + }, + { + "grad_norm": 0.18499837815761566, + "learning_rate": 6.706906495372987e-05, + "loss": 0.0071, + "step": 25180 + }, + { + "grad_norm": 0.2700726389884949, + "learning_rate": 6.704316007172365e-05, + "loss": 0.0069, + "step": 25190 + }, + { + "grad_norm": 0.22663068771362305, + "learning_rate": 6.701725001245539e-05, + "loss": 0.0052, + "step": 25200 + }, + { + "grad_norm": 0.18079528212547302, + "learning_rate": 6.699133478379588e-05, + "loss": 0.0051, + "step": 25210 + }, + { + "grad_norm": 0.19316238164901733, + "learning_rate": 6.69654143936175e-05, + "loss": 0.0079, + "step": 25220 + }, + { + "grad_norm": 0.21459133923053741, + "learning_rate": 6.693948884979419e-05, + "loss": 0.0072, + "step": 25230 + }, + { + "grad_norm": 0.23129896819591522, + "learning_rate": 6.691355816020142e-05, + "loss": 0.0066, + "step": 25240 + }, + { + "grad_norm": 0.1841215342283249, + "learning_rate": 6.688762233271624e-05, + "loss": 0.0057, + "step": 25250 + }, + { + "grad_norm": 0.16104280948638916, + "learning_rate": 6.68616813752173e-05, + "loss": 0.0062, + "step": 25260 + }, + { + "grad_norm": 0.24268488585948944, + "learning_rate": 6.683573529558477e-05, + "loss": 0.0072, + "step": 25270 + }, + { + "grad_norm": 0.259103000164032, + "learning_rate": 6.680978410170037e-05, + "loss": 0.0064, + "step": 25280 + }, + { + "grad_norm": 0.2177886962890625, + "learning_rate": 6.678382780144741e-05, + "loss": 0.0076, + "step": 25290 + }, + { + "grad_norm": 0.1991564780473709, + "learning_rate": 6.675786640271071e-05, + "loss": 0.0072, + "step": 25300 + }, + { + "grad_norm": 0.22543810307979584, + "learning_rate": 6.673189991337665e-05, + "loss": 0.0073, + "step": 25310 + }, + { + "grad_norm": 0.24554254114627838, + "learning_rate": 6.670592834133317e-05, + "loss": 0.0078, + "step": 25320 + }, + { + "grad_norm": 0.15265288949012756, + "learning_rate": 6.667995169446979e-05, + "loss": 0.0063, + "step": 25330 + }, + { + "grad_norm": 0.16831746697425842, + "learning_rate": 6.665396998067747e-05, + "loss": 0.0057, + "step": 25340 + }, + { + "grad_norm": 0.19175328314304352, + "learning_rate": 6.66279832078488e-05, + "loss": 0.0048, + "step": 25350 + }, + { + "grad_norm": 0.18119966983795166, + "learning_rate": 6.660199138387786e-05, + "loss": 0.0055, + "step": 25360 + }, + { + "grad_norm": 0.24650025367736816, + "learning_rate": 6.65759945166603e-05, + "loss": 0.007, + "step": 25370 + }, + { + "grad_norm": 0.21147359907627106, + "learning_rate": 6.654999261409326e-05, + "loss": 0.0071, + "step": 25380 + }, + { + "grad_norm": 0.22879630327224731, + "learning_rate": 6.652398568407544e-05, + "loss": 0.0074, + "step": 25390 + }, + { + "grad_norm": 0.22692294418811798, + "learning_rate": 6.649797373450707e-05, + "loss": 0.0081, + "step": 25400 + }, + { + "grad_norm": 0.2574363946914673, + "learning_rate": 6.647195677328988e-05, + "loss": 0.0094, + "step": 25410 + }, + { + "grad_norm": 0.2013295441865921, + "learning_rate": 6.644593480832712e-05, + "loss": 0.0086, + "step": 25420 + }, + { + "grad_norm": 0.15698054432868958, + "learning_rate": 6.641990784752363e-05, + "loss": 0.0059, + "step": 25430 + }, + { + "grad_norm": 0.1815074235200882, + "learning_rate": 6.639387589878566e-05, + "loss": 0.0072, + "step": 25440 + }, + { + "grad_norm": 0.16183070838451385, + "learning_rate": 6.636783897002103e-05, + "loss": 0.0049, + "step": 25450 + }, + { + "grad_norm": 0.20442305505275726, + "learning_rate": 6.63417970691391e-05, + "loss": 0.0082, + "step": 25460 + }, + { + "grad_norm": 0.2078857719898224, + "learning_rate": 6.63157502040507e-05, + "loss": 0.0101, + "step": 25470 + }, + { + "grad_norm": 0.2201431542634964, + "learning_rate": 6.628969838266819e-05, + "loss": 0.0088, + "step": 25480 + }, + { + "grad_norm": 0.24755345284938812, + "learning_rate": 6.626364161290541e-05, + "loss": 0.0076, + "step": 25490 + }, + { + "grad_norm": 0.22055433690547943, + "learning_rate": 6.623757990267774e-05, + "loss": 0.0063, + "step": 25500 + }, + { + "grad_norm": 0.18375305831432343, + "learning_rate": 6.621151325990201e-05, + "loss": 0.007, + "step": 25510 + }, + { + "grad_norm": 0.2530211806297302, + "learning_rate": 6.618544169249657e-05, + "loss": 0.0088, + "step": 25520 + }, + { + "grad_norm": 0.26939213275909424, + "learning_rate": 6.615936520838133e-05, + "loss": 0.0082, + "step": 25530 + }, + { + "grad_norm": 0.3377273678779602, + "learning_rate": 6.613328381547759e-05, + "loss": 0.0075, + "step": 25540 + }, + { + "grad_norm": 0.26195868849754333, + "learning_rate": 6.610719752170821e-05, + "loss": 0.0049, + "step": 25550 + }, + { + "grad_norm": 0.20831884443759918, + "learning_rate": 6.60811063349975e-05, + "loss": 0.0054, + "step": 25560 + }, + { + "grad_norm": 0.23061248660087585, + "learning_rate": 6.605501026327127e-05, + "loss": 0.0067, + "step": 25570 + }, + { + "grad_norm": 0.19042521715164185, + "learning_rate": 6.602890931445685e-05, + "loss": 0.0071, + "step": 25580 + }, + { + "grad_norm": 0.2882711887359619, + "learning_rate": 6.6002803496483e-05, + "loss": 0.0083, + "step": 25590 + }, + { + "grad_norm": 0.2180003523826599, + "learning_rate": 6.597669281727997e-05, + "loss": 0.0075, + "step": 25600 + }, + { + "grad_norm": 0.20100033283233643, + "learning_rate": 6.595057728477949e-05, + "loss": 0.0062, + "step": 25610 + }, + { + "grad_norm": 0.20027615129947662, + "learning_rate": 6.59244569069148e-05, + "loss": 0.0066, + "step": 25620 + }, + { + "grad_norm": 0.26127099990844727, + "learning_rate": 6.589833169162054e-05, + "loss": 0.0077, + "step": 25630 + }, + { + "grad_norm": 0.24925601482391357, + "learning_rate": 6.587220164683291e-05, + "loss": 0.0059, + "step": 25640 + }, + { + "grad_norm": 0.271277517080307, + "learning_rate": 6.58460667804895e-05, + "loss": 0.0066, + "step": 25650 + }, + { + "grad_norm": 0.20398743450641632, + "learning_rate": 6.581992710052938e-05, + "loss": 0.0067, + "step": 25660 + }, + { + "grad_norm": 0.20166295766830444, + "learning_rate": 6.579378261489311e-05, + "loss": 0.0096, + "step": 25670 + }, + { + "grad_norm": 0.23275814950466156, + "learning_rate": 6.576763333152268e-05, + "loss": 0.0073, + "step": 25680 + }, + { + "grad_norm": 0.135862797498703, + "learning_rate": 6.574147925836159e-05, + "loss": 0.0049, + "step": 25690 + }, + { + "grad_norm": 0.1966680884361267, + "learning_rate": 6.571532040335472e-05, + "loss": 0.0072, + "step": 25700 + }, + { + "grad_norm": 0.2585192024707794, + "learning_rate": 6.568915677444845e-05, + "loss": 0.0079, + "step": 25710 + }, + { + "grad_norm": 0.16709916293621063, + "learning_rate": 6.56629883795906e-05, + "loss": 0.0053, + "step": 25720 + }, + { + "grad_norm": 0.1999332308769226, + "learning_rate": 6.563681522673043e-05, + "loss": 0.0073, + "step": 25730 + }, + { + "grad_norm": 0.22824805974960327, + "learning_rate": 6.561063732381867e-05, + "loss": 0.0068, + "step": 25740 + }, + { + "grad_norm": 0.21648965775966644, + "learning_rate": 6.558445467880745e-05, + "loss": 0.0071, + "step": 25750 + }, + { + "grad_norm": 0.2542528510093689, + "learning_rate": 6.55582672996504e-05, + "loss": 0.0073, + "step": 25760 + }, + { + "grad_norm": 0.2426459789276123, + "learning_rate": 6.553207519430253e-05, + "loss": 0.0056, + "step": 25770 + }, + { + "grad_norm": 0.2039594054222107, + "learning_rate": 6.550587837072032e-05, + "loss": 0.0062, + "step": 25780 + }, + { + "grad_norm": 0.23912934958934784, + "learning_rate": 6.547967683686166e-05, + "loss": 0.0063, + "step": 25790 + }, + { + "grad_norm": 0.18202634155750275, + "learning_rate": 6.545347060068591e-05, + "loss": 0.005, + "step": 25800 + }, + { + "grad_norm": 0.22227893769741058, + "learning_rate": 6.542725967015382e-05, + "loss": 0.006, + "step": 25810 + }, + { + "grad_norm": 0.21977637708187103, + "learning_rate": 6.540104405322757e-05, + "loss": 0.0072, + "step": 25820 + }, + { + "grad_norm": 0.2197333425283432, + "learning_rate": 6.537482375787077e-05, + "loss": 0.009, + "step": 25830 + }, + { + "grad_norm": 0.15945011377334595, + "learning_rate": 6.534859879204845e-05, + "loss": 0.0072, + "step": 25840 + }, + { + "grad_norm": 0.1592593640089035, + "learning_rate": 6.532236916372709e-05, + "loss": 0.007, + "step": 25850 + }, + { + "grad_norm": 0.2686917185783386, + "learning_rate": 6.529613488087454e-05, + "loss": 0.0097, + "step": 25860 + }, + { + "grad_norm": 0.24457047879695892, + "learning_rate": 6.526989595146009e-05, + "loss": 0.0056, + "step": 25870 + }, + { + "grad_norm": 0.21197229623794556, + "learning_rate": 6.524365238345441e-05, + "loss": 0.009, + "step": 25880 + }, + { + "grad_norm": 0.22684945166110992, + "learning_rate": 6.521740418482964e-05, + "loss": 0.006, + "step": 25890 + }, + { + "grad_norm": 0.20982791483402252, + "learning_rate": 6.519115136355925e-05, + "loss": 0.0058, + "step": 25900 + }, + { + "grad_norm": 0.23438315093517303, + "learning_rate": 6.51648939276182e-05, + "loss": 0.0081, + "step": 25910 + }, + { + "grad_norm": 0.21843135356903076, + "learning_rate": 6.513863188498277e-05, + "loss": 0.007, + "step": 25920 + }, + { + "grad_norm": 0.20278333127498627, + "learning_rate": 6.511236524363068e-05, + "loss": 0.0048, + "step": 25930 + }, + { + "grad_norm": 0.20683413743972778, + "learning_rate": 6.508609401154104e-05, + "loss": 0.0051, + "step": 25940 + }, + { + "grad_norm": 0.2596301734447479, + "learning_rate": 6.505981819669439e-05, + "loss": 0.0062, + "step": 25950 + }, + { + "grad_norm": 0.23002870380878448, + "learning_rate": 6.503353780707258e-05, + "loss": 0.0058, + "step": 25960 + }, + { + "grad_norm": 0.20470206439495087, + "learning_rate": 6.500725285065895e-05, + "loss": 0.0049, + "step": 25970 + }, + { + "grad_norm": 0.19926327466964722, + "learning_rate": 6.498096333543813e-05, + "loss": 0.0055, + "step": 25980 + }, + { + "grad_norm": 0.19966010749340057, + "learning_rate": 6.49546692693962e-05, + "loss": 0.0054, + "step": 25990 + }, + { + "grad_norm": 0.19689211249351501, + "learning_rate": 6.492837066052059e-05, + "loss": 0.0048, + "step": 26000 + }, + { + "grad_norm": 0.2969319224357605, + "learning_rate": 6.490206751680014e-05, + "loss": 0.009, + "step": 26010 + }, + { + "grad_norm": 0.22720490396022797, + "learning_rate": 6.487575984622505e-05, + "loss": 0.0053, + "step": 26020 + }, + { + "grad_norm": 0.2233182191848755, + "learning_rate": 6.484944765678689e-05, + "loss": 0.0079, + "step": 26030 + }, + { + "grad_norm": 0.1987072080373764, + "learning_rate": 6.482313095647861e-05, + "loss": 0.0054, + "step": 26040 + }, + { + "grad_norm": 0.1756778359413147, + "learning_rate": 6.479680975329451e-05, + "loss": 0.0054, + "step": 26050 + }, + { + "grad_norm": 0.15614423155784607, + "learning_rate": 6.477048405523031e-05, + "loss": 0.0055, + "step": 26060 + }, + { + "grad_norm": 0.21151113510131836, + "learning_rate": 6.474415387028304e-05, + "loss": 0.0067, + "step": 26070 + }, + { + "grad_norm": 0.19343069195747375, + "learning_rate": 6.471781920645114e-05, + "loss": 0.0066, + "step": 26080 + }, + { + "grad_norm": 0.1986183077096939, + "learning_rate": 6.469148007173434e-05, + "loss": 0.0074, + "step": 26090 + }, + { + "grad_norm": 0.22761370241641998, + "learning_rate": 6.466513647413381e-05, + "loss": 0.0065, + "step": 26100 + }, + { + "grad_norm": 0.18695175647735596, + "learning_rate": 6.463878842165203e-05, + "loss": 0.006, + "step": 26110 + }, + { + "grad_norm": 0.15956836938858032, + "learning_rate": 6.461243592229286e-05, + "loss": 0.0051, + "step": 26120 + }, + { + "grad_norm": 0.1976875364780426, + "learning_rate": 6.458607898406146e-05, + "loss": 0.0061, + "step": 26130 + }, + { + "grad_norm": 0.19613993167877197, + "learning_rate": 6.455971761496439e-05, + "loss": 0.0057, + "step": 26140 + }, + { + "grad_norm": 0.2289198637008667, + "learning_rate": 6.453335182300953e-05, + "loss": 0.0073, + "step": 26150 + }, + { + "grad_norm": 0.2191610187292099, + "learning_rate": 6.450698161620612e-05, + "loss": 0.0054, + "step": 26160 + }, + { + "grad_norm": 0.19832280278205872, + "learning_rate": 6.448060700256473e-05, + "loss": 0.0082, + "step": 26170 + }, + { + "grad_norm": 0.20640571415424347, + "learning_rate": 6.445422799009726e-05, + "loss": 0.005, + "step": 26180 + }, + { + "grad_norm": 0.22177433967590332, + "learning_rate": 6.442784458681699e-05, + "loss": 0.0056, + "step": 26190 + }, + { + "grad_norm": 0.19210496544837952, + "learning_rate": 6.440145680073847e-05, + "loss": 0.0079, + "step": 26200 + }, + { + "grad_norm": 0.26254889369010925, + "learning_rate": 6.437506463987762e-05, + "loss": 0.0084, + "step": 26210 + }, + { + "grad_norm": 0.209589883685112, + "learning_rate": 6.434866811225168e-05, + "loss": 0.0065, + "step": 26220 + }, + { + "grad_norm": 0.22509878873825073, + "learning_rate": 6.432226722587923e-05, + "loss": 0.0085, + "step": 26230 + }, + { + "grad_norm": 0.21649697422981262, + "learning_rate": 6.429586198878015e-05, + "loss": 0.0059, + "step": 26240 + }, + { + "grad_norm": 0.201228067278862, + "learning_rate": 6.426945240897566e-05, + "loss": 0.0053, + "step": 26250 + }, + { + "grad_norm": 0.1925959438085556, + "learning_rate": 6.424303849448829e-05, + "loss": 0.0063, + "step": 26260 + }, + { + "grad_norm": 0.2186427116394043, + "learning_rate": 6.42166202533419e-05, + "loss": 0.0062, + "step": 26270 + }, + { + "grad_norm": 0.23884597420692444, + "learning_rate": 6.419019769356164e-05, + "loss": 0.0048, + "step": 26280 + }, + { + "grad_norm": 0.21245859563350677, + "learning_rate": 6.416377082317398e-05, + "loss": 0.0052, + "step": 26290 + }, + { + "grad_norm": 0.2216310203075409, + "learning_rate": 6.413733965020674e-05, + "loss": 0.006, + "step": 26300 + }, + { + "grad_norm": 0.15961231291294098, + "learning_rate": 6.411090418268896e-05, + "loss": 0.0057, + "step": 26310 + }, + { + "grad_norm": 0.22585490345954895, + "learning_rate": 6.408446442865109e-05, + "loss": 0.0066, + "step": 26320 + }, + { + "grad_norm": 0.20640522241592407, + "learning_rate": 6.405802039612479e-05, + "loss": 0.0075, + "step": 26330 + }, + { + "grad_norm": 0.23104189336299896, + "learning_rate": 6.403157209314308e-05, + "loss": 0.0075, + "step": 26340 + }, + { + "grad_norm": 0.19122198224067688, + "learning_rate": 6.400511952774024e-05, + "loss": 0.0094, + "step": 26350 + }, + { + "grad_norm": 0.2364351451396942, + "learning_rate": 6.397866270795187e-05, + "loss": 0.0056, + "step": 26360 + }, + { + "grad_norm": 0.22780410945415497, + "learning_rate": 6.395220164181489e-05, + "loss": 0.0055, + "step": 26370 + }, + { + "grad_norm": 0.2700885832309723, + "learning_rate": 6.39257363373674e-05, + "loss": 0.0067, + "step": 26380 + }, + { + "grad_norm": 0.26172471046447754, + "learning_rate": 6.389926680264892e-05, + "loss": 0.0074, + "step": 26390 + }, + { + "grad_norm": 0.22210976481437683, + "learning_rate": 6.387279304570017e-05, + "loss": 0.0057, + "step": 26400 + }, + { + "grad_norm": 0.2141304612159729, + "learning_rate": 6.384631507456319e-05, + "loss": 0.0065, + "step": 26410 + }, + { + "grad_norm": 0.22563396394252777, + "learning_rate": 6.381983289728126e-05, + "loss": 0.0067, + "step": 26420 + }, + { + "grad_norm": 0.19412635266780853, + "learning_rate": 6.3793346521899e-05, + "loss": 0.0058, + "step": 26430 + }, + { + "grad_norm": 0.16654404997825623, + "learning_rate": 6.376685595646226e-05, + "loss": 0.0072, + "step": 26440 + }, + { + "grad_norm": 0.16080507636070251, + "learning_rate": 6.374036120901816e-05, + "loss": 0.0056, + "step": 26450 + }, + { + "grad_norm": 0.32743343710899353, + "learning_rate": 6.371386228761514e-05, + "loss": 0.0065, + "step": 26460 + }, + { + "grad_norm": 0.28099605441093445, + "learning_rate": 6.368735920030283e-05, + "loss": 0.0064, + "step": 26470 + }, + { + "grad_norm": 0.19512446224689484, + "learning_rate": 6.366085195513218e-05, + "loss": 0.0068, + "step": 26480 + }, + { + "grad_norm": 0.2368878275156021, + "learning_rate": 6.363434056015543e-05, + "loss": 0.0073, + "step": 26490 + }, + { + "grad_norm": 0.18388403952121735, + "learning_rate": 6.360782502342599e-05, + "loss": 0.0055, + "step": 26500 + }, + { + "grad_norm": 0.1994674950838089, + "learning_rate": 6.358130535299862e-05, + "loss": 0.0064, + "step": 26510 + }, + { + "grad_norm": 0.26185137033462524, + "learning_rate": 6.355478155692926e-05, + "loss": 0.0065, + "step": 26520 + }, + { + "grad_norm": 0.19742263853549957, + "learning_rate": 6.352825364327517e-05, + "loss": 0.0058, + "step": 26530 + }, + { + "grad_norm": 0.24318774044513702, + "learning_rate": 6.350172162009482e-05, + "loss": 0.0078, + "step": 26540 + }, + { + "grad_norm": 0.2802782952785492, + "learning_rate": 6.347518549544793e-05, + "loss": 0.0075, + "step": 26550 + }, + { + "grad_norm": 0.1994744837284088, + "learning_rate": 6.344864527739547e-05, + "loss": 0.0062, + "step": 26560 + }, + { + "grad_norm": 0.2588227093219757, + "learning_rate": 6.342210097399966e-05, + "loss": 0.0079, + "step": 26570 + }, + { + "grad_norm": 0.20334990322589874, + "learning_rate": 6.339555259332398e-05, + "loss": 0.0076, + "step": 26580 + }, + { + "grad_norm": 0.2466357797384262, + "learning_rate": 6.33690001434331e-05, + "loss": 0.0055, + "step": 26590 + }, + { + "grad_norm": 0.1516738384962082, + "learning_rate": 6.334244363239296e-05, + "loss": 0.0059, + "step": 26600 + }, + { + "grad_norm": 0.25755223631858826, + "learning_rate": 6.331588306827073e-05, + "loss": 0.0082, + "step": 26610 + }, + { + "grad_norm": 0.20359931886196136, + "learning_rate": 6.328931845913483e-05, + "loss": 0.0078, + "step": 26620 + }, + { + "grad_norm": 0.26368001103401184, + "learning_rate": 6.326274981305484e-05, + "loss": 0.0061, + "step": 26630 + }, + { + "grad_norm": 0.20655317604541779, + "learning_rate": 6.323617713810166e-05, + "loss": 0.0056, + "step": 26640 + }, + { + "grad_norm": 0.23046205937862396, + "learning_rate": 6.320960044234734e-05, + "loss": 0.0055, + "step": 26650 + }, + { + "grad_norm": 0.2450997680425644, + "learning_rate": 6.318301973386518e-05, + "loss": 0.0066, + "step": 26660 + }, + { + "grad_norm": 0.21167534589767456, + "learning_rate": 6.315643502072971e-05, + "loss": 0.0056, + "step": 26670 + }, + { + "grad_norm": 0.17250072956085205, + "learning_rate": 6.312984631101667e-05, + "loss": 0.0055, + "step": 26680 + }, + { + "grad_norm": 0.24523146450519562, + "learning_rate": 6.310325361280297e-05, + "loss": 0.0059, + "step": 26690 + }, + { + "grad_norm": 0.24308966100215912, + "learning_rate": 6.30766569341668e-05, + "loss": 0.006, + "step": 26700 + }, + { + "grad_norm": 0.18970511853694916, + "learning_rate": 6.305005628318753e-05, + "loss": 0.0067, + "step": 26710 + }, + { + "grad_norm": 0.20905692875385284, + "learning_rate": 6.302345166794572e-05, + "loss": 0.0055, + "step": 26720 + }, + { + "grad_norm": 0.18546536564826965, + "learning_rate": 6.299684309652316e-05, + "loss": 0.0053, + "step": 26730 + }, + { + "grad_norm": 0.12105439603328705, + "learning_rate": 6.297023057700283e-05, + "loss": 0.0044, + "step": 26740 + }, + { + "grad_norm": 0.20489977300167084, + "learning_rate": 6.294361411746891e-05, + "loss": 0.0062, + "step": 26750 + }, + { + "grad_norm": 0.24438318610191345, + "learning_rate": 6.291699372600677e-05, + "loss": 0.0064, + "step": 26760 + }, + { + "grad_norm": 0.22675564885139465, + "learning_rate": 6.2890369410703e-05, + "loss": 0.005, + "step": 26770 + }, + { + "grad_norm": 0.2818538248538971, + "learning_rate": 6.286374117964534e-05, + "loss": 0.0074, + "step": 26780 + }, + { + "grad_norm": 0.18035154044628143, + "learning_rate": 6.283710904092277e-05, + "loss": 0.0043, + "step": 26790 + }, + { + "grad_norm": 0.1388295739889145, + "learning_rate": 6.281047300262542e-05, + "loss": 0.0059, + "step": 26800 + }, + { + "grad_norm": 0.1769358366727829, + "learning_rate": 6.278383307284461e-05, + "loss": 0.0061, + "step": 26810 + }, + { + "grad_norm": 0.16022364795207977, + "learning_rate": 6.275718925967284e-05, + "loss": 0.0056, + "step": 26820 + }, + { + "grad_norm": 0.22695006430149078, + "learning_rate": 6.273054157120382e-05, + "loss": 0.0068, + "step": 26830 + }, + { + "grad_norm": 0.21520768105983734, + "learning_rate": 6.270389001553238e-05, + "loss": 0.0058, + "step": 26840 + }, + { + "grad_norm": 0.19172999262809753, + "learning_rate": 6.26772346007546e-05, + "loss": 0.0056, + "step": 26850 + }, + { + "grad_norm": 0.2523164451122284, + "learning_rate": 6.265057533496767e-05, + "loss": 0.0049, + "step": 26860 + }, + { + "grad_norm": 0.26460522413253784, + "learning_rate": 6.262391222626997e-05, + "loss": 0.0069, + "step": 26870 + }, + { + "grad_norm": 0.25140002369880676, + "learning_rate": 6.259724528276106e-05, + "loss": 0.0064, + "step": 26880 + }, + { + "grad_norm": 0.1940133422613144, + "learning_rate": 6.257057451254162e-05, + "loss": 0.008, + "step": 26890 + }, + { + "grad_norm": 0.22486016154289246, + "learning_rate": 6.254389992371357e-05, + "loss": 0.0068, + "step": 26900 + }, + { + "grad_norm": 0.2173072248697281, + "learning_rate": 6.25172215243799e-05, + "loss": 0.006, + "step": 26910 + }, + { + "grad_norm": 0.2645227909088135, + "learning_rate": 6.249053932264486e-05, + "loss": 0.0064, + "step": 26920 + }, + { + "grad_norm": 0.1636214554309845, + "learning_rate": 6.246385332661376e-05, + "loss": 0.0066, + "step": 26930 + }, + { + "grad_norm": 0.23108375072479248, + "learning_rate": 6.24371635443931e-05, + "loss": 0.0062, + "step": 26940 + }, + { + "grad_norm": 0.20354636013507843, + "learning_rate": 6.241046998409054e-05, + "loss": 0.0074, + "step": 26950 + }, + { + "grad_norm": 0.23577523231506348, + "learning_rate": 6.238377265381489e-05, + "loss": 0.0063, + "step": 26960 + }, + { + "grad_norm": 0.3131958544254303, + "learning_rate": 6.235707156167607e-05, + "loss": 0.0058, + "step": 26970 + }, + { + "grad_norm": 0.25694388151168823, + "learning_rate": 6.233036671578519e-05, + "loss": 0.0065, + "step": 26980 + }, + { + "grad_norm": 0.19088886678218842, + "learning_rate": 6.230365812425445e-05, + "loss": 0.0069, + "step": 26990 + }, + { + "grad_norm": 0.247942715883255, + "learning_rate": 6.227694579519724e-05, + "loss": 0.0055, + "step": 27000 + }, + { + "grad_norm": 0.23688679933547974, + "learning_rate": 6.225022973672805e-05, + "loss": 0.0061, + "step": 27010 + }, + { + "grad_norm": 0.2073313444852829, + "learning_rate": 6.222350995696253e-05, + "loss": 0.0058, + "step": 27020 + }, + { + "grad_norm": 0.24426770210266113, + "learning_rate": 6.21967864640174e-05, + "loss": 0.0057, + "step": 27030 + }, + { + "grad_norm": 0.1389980912208557, + "learning_rate": 6.217005926601059e-05, + "loss": 0.0052, + "step": 27040 + }, + { + "grad_norm": 0.18472421169281006, + "learning_rate": 6.214332837106111e-05, + "loss": 0.0052, + "step": 27050 + }, + { + "grad_norm": 0.2815677523612976, + "learning_rate": 6.21165937872891e-05, + "loss": 0.0071, + "step": 27060 + }, + { + "grad_norm": 0.2159937471151352, + "learning_rate": 6.208985552281582e-05, + "loss": 0.0063, + "step": 27070 + }, + { + "grad_norm": 0.19135768711566925, + "learning_rate": 6.206311358576364e-05, + "loss": 0.0064, + "step": 27080 + }, + { + "grad_norm": 0.1525515466928482, + "learning_rate": 6.203636798425608e-05, + "loss": 0.0047, + "step": 27090 + }, + { + "grad_norm": 0.15612834692001343, + "learning_rate": 6.20096187264177e-05, + "loss": 0.0053, + "step": 27100 + }, + { + "grad_norm": 0.1883126199245453, + "learning_rate": 6.198286582037425e-05, + "loss": 0.0051, + "step": 27110 + }, + { + "grad_norm": 0.21747751533985138, + "learning_rate": 6.195610927425256e-05, + "loss": 0.0057, + "step": 27120 + }, + { + "grad_norm": 0.20136936008930206, + "learning_rate": 6.192934909618056e-05, + "loss": 0.0054, + "step": 27130 + }, + { + "grad_norm": 0.18872247636318207, + "learning_rate": 6.190258529428728e-05, + "loss": 0.0049, + "step": 27140 + }, + { + "grad_norm": 0.28938767313957214, + "learning_rate": 6.187581787670285e-05, + "loss": 0.0072, + "step": 27150 + }, + { + "grad_norm": 0.2653864324092865, + "learning_rate": 6.184904685155852e-05, + "loss": 0.0077, + "step": 27160 + }, + { + "grad_norm": 0.2535940706729889, + "learning_rate": 6.18222722269866e-05, + "loss": 0.0076, + "step": 27170 + }, + { + "grad_norm": 0.1849503070116043, + "learning_rate": 6.179549401112053e-05, + "loss": 0.0067, + "step": 27180 + }, + { + "grad_norm": 0.2222738415002823, + "learning_rate": 6.176871221209482e-05, + "loss": 0.0079, + "step": 27190 + }, + { + "grad_norm": 0.29542919993400574, + "learning_rate": 6.174192683804508e-05, + "loss": 0.0059, + "step": 27200 + }, + { + "grad_norm": 0.22121202945709229, + "learning_rate": 6.1715137897108e-05, + "loss": 0.0067, + "step": 27210 + }, + { + "grad_norm": 0.21276335418224335, + "learning_rate": 6.168834539742134e-05, + "loss": 0.0055, + "step": 27220 + }, + { + "grad_norm": 0.29806381464004517, + "learning_rate": 6.166154934712397e-05, + "loss": 0.0052, + "step": 27230 + }, + { + "grad_norm": 0.2035537213087082, + "learning_rate": 6.163474975435581e-05, + "loss": 0.0053, + "step": 27240 + }, + { + "grad_norm": 0.2204233556985855, + "learning_rate": 6.160794662725787e-05, + "loss": 0.006, + "step": 27250 + }, + { + "grad_norm": 0.23584558069705963, + "learning_rate": 6.158113997397222e-05, + "loss": 0.0063, + "step": 27260 + }, + { + "grad_norm": 0.19361534714698792, + "learning_rate": 6.155432980264205e-05, + "loss": 0.006, + "step": 27270 + }, + { + "grad_norm": 0.18379713594913483, + "learning_rate": 6.152751612141156e-05, + "loss": 0.0061, + "step": 27280 + }, + { + "grad_norm": 0.18036022782325745, + "learning_rate": 6.150069893842602e-05, + "loss": 0.006, + "step": 27290 + }, + { + "grad_norm": 0.13554903864860535, + "learning_rate": 6.147387826183182e-05, + "loss": 0.0085, + "step": 27300 + }, + { + "grad_norm": 0.23212015628814697, + "learning_rate": 6.144705409977635e-05, + "loss": 0.0061, + "step": 27310 + }, + { + "grad_norm": 0.21660080552101135, + "learning_rate": 6.142022646040808e-05, + "loss": 0.0058, + "step": 27320 + }, + { + "grad_norm": 0.18117165565490723, + "learning_rate": 6.139339535187653e-05, + "loss": 0.0049, + "step": 27330 + }, + { + "grad_norm": 0.16411201655864716, + "learning_rate": 6.136656078233232e-05, + "loss": 0.0051, + "step": 27340 + }, + { + "grad_norm": 0.1901129186153412, + "learning_rate": 6.133972275992707e-05, + "loss": 0.0046, + "step": 27350 + }, + { + "grad_norm": 0.1666228026151657, + "learning_rate": 6.131288129281342e-05, + "loss": 0.0045, + "step": 27360 + }, + { + "grad_norm": 0.18389329314231873, + "learning_rate": 6.128603638914516e-05, + "loss": 0.0066, + "step": 27370 + }, + { + "grad_norm": 0.1898213028907776, + "learning_rate": 6.125918805707704e-05, + "loss": 0.006, + "step": 27380 + }, + { + "grad_norm": 0.19179952144622803, + "learning_rate": 6.123233630476485e-05, + "loss": 0.0059, + "step": 27390 + }, + { + "grad_norm": 0.22159911692142487, + "learning_rate": 6.120548114036547e-05, + "loss": 0.0056, + "step": 27400 + }, + { + "grad_norm": 0.2556234896183014, + "learning_rate": 6.117862257203679e-05, + "loss": 0.006, + "step": 27410 + }, + { + "grad_norm": 0.2184114009141922, + "learning_rate": 6.115176060793771e-05, + "loss": 0.006, + "step": 27420 + }, + { + "grad_norm": 0.2209833711385727, + "learning_rate": 6.112489525622822e-05, + "loss": 0.0074, + "step": 27430 + }, + { + "grad_norm": 0.2665866017341614, + "learning_rate": 6.109802652506928e-05, + "loss": 0.0065, + "step": 27440 + }, + { + "grad_norm": 0.21960771083831787, + "learning_rate": 6.107115442262291e-05, + "loss": 0.0075, + "step": 27450 + }, + { + "grad_norm": 0.15529780089855194, + "learning_rate": 6.104427895705214e-05, + "loss": 0.0048, + "step": 27460 + }, + { + "grad_norm": 0.17456737160682678, + "learning_rate": 6.101740013652103e-05, + "loss": 0.0067, + "step": 27470 + }, + { + "grad_norm": 0.17030955851078033, + "learning_rate": 6.099051796919465e-05, + "loss": 0.0059, + "step": 27480 + }, + { + "grad_norm": 0.14730389416217804, + "learning_rate": 6.096363246323911e-05, + "loss": 0.0047, + "step": 27490 + }, + { + "grad_norm": 0.19984987378120422, + "learning_rate": 6.0936743626821504e-05, + "loss": 0.0058, + "step": 27500 + }, + { + "grad_norm": 0.24629443883895874, + "learning_rate": 6.090985146810996e-05, + "loss": 0.0056, + "step": 27510 + }, + { + "grad_norm": 0.19523531198501587, + "learning_rate": 6.088295599527357e-05, + "loss": 0.0055, + "step": 27520 + }, + { + "grad_norm": 0.19997794926166534, + "learning_rate": 6.085605721648252e-05, + "loss": 0.0044, + "step": 27530 + }, + { + "grad_norm": 0.2269504815340042, + "learning_rate": 6.082915513990792e-05, + "loss": 0.0061, + "step": 27540 + }, + { + "grad_norm": 0.19974826276302338, + "learning_rate": 6.080224977372192e-05, + "loss": 0.0059, + "step": 27550 + }, + { + "grad_norm": 0.14757752418518066, + "learning_rate": 6.0775341126097666e-05, + "loss": 0.0049, + "step": 27560 + }, + { + "grad_norm": 0.18435977399349213, + "learning_rate": 6.074842920520926e-05, + "loss": 0.0065, + "step": 27570 + }, + { + "grad_norm": 0.2411748319864273, + "learning_rate": 6.072151401923186e-05, + "loss": 0.0068, + "step": 27580 + }, + { + "grad_norm": 0.20606940984725952, + "learning_rate": 6.069459557634159e-05, + "loss": 0.008, + "step": 27590 + }, + { + "grad_norm": 0.2300151288509369, + "learning_rate": 6.066767388471557e-05, + "loss": 0.0064, + "step": 27600 + }, + { + "grad_norm": 0.19813773036003113, + "learning_rate": 6.064074895253188e-05, + "loss": 0.007, + "step": 27610 + }, + { + "grad_norm": 0.20168276131153107, + "learning_rate": 6.061382078796961e-05, + "loss": 0.01, + "step": 27620 + }, + { + "grad_norm": 0.20673015713691711, + "learning_rate": 6.0586889399208814e-05, + "loss": 0.0075, + "step": 27630 + }, + { + "grad_norm": 0.25116589665412903, + "learning_rate": 6.0559954794430565e-05, + "loss": 0.0058, + "step": 27640 + }, + { + "grad_norm": 0.22388678789138794, + "learning_rate": 6.053301698181687e-05, + "loss": 0.006, + "step": 27650 + }, + { + "grad_norm": 0.1884300410747528, + "learning_rate": 6.0506075969550725e-05, + "loss": 0.0052, + "step": 27660 + }, + { + "grad_norm": 0.16081227362155914, + "learning_rate": 6.047913176581609e-05, + "loss": 0.0055, + "step": 27670 + }, + { + "grad_norm": 0.1714400202035904, + "learning_rate": 6.0452184378797904e-05, + "loss": 0.0044, + "step": 27680 + }, + { + "grad_norm": 0.20271414518356323, + "learning_rate": 6.042523381668209e-05, + "loss": 0.0063, + "step": 27690 + }, + { + "grad_norm": 0.25421056151390076, + "learning_rate": 6.03982800876555e-05, + "loss": 0.0064, + "step": 27700 + }, + { + "grad_norm": 0.20130525529384613, + "learning_rate": 6.0371323199905975e-05, + "loss": 0.0058, + "step": 27710 + }, + { + "grad_norm": 0.19146931171417236, + "learning_rate": 6.03443631616223e-05, + "loss": 0.0047, + "step": 27720 + }, + { + "grad_norm": 0.1841273307800293, + "learning_rate": 6.031739998099421e-05, + "loss": 0.0044, + "step": 27730 + }, + { + "grad_norm": 0.19991715252399445, + "learning_rate": 6.029043366621243e-05, + "loss": 0.0064, + "step": 27740 + }, + { + "grad_norm": 0.2284337431192398, + "learning_rate": 6.0263464225468615e-05, + "loss": 0.0065, + "step": 27750 + }, + { + "grad_norm": 0.29523906111717224, + "learning_rate": 6.023649166695534e-05, + "loss": 0.0061, + "step": 27760 + }, + { + "grad_norm": 0.20799440145492554, + "learning_rate": 6.0209515998866186e-05, + "loss": 0.0079, + "step": 27770 + }, + { + "grad_norm": 0.1822252869606018, + "learning_rate": 6.018253722939563e-05, + "loss": 0.0051, + "step": 27780 + }, + { + "grad_norm": 0.2678062617778778, + "learning_rate": 6.015555536673914e-05, + "loss": 0.0069, + "step": 27790 + }, + { + "grad_norm": 0.23746149241924286, + "learning_rate": 6.0128570419093054e-05, + "loss": 0.006, + "step": 27800 + }, + { + "grad_norm": 0.22552046179771423, + "learning_rate": 6.010158239465471e-05, + "loss": 0.0061, + "step": 27810 + }, + { + "grad_norm": 0.30230170488357544, + "learning_rate": 6.007459130162235e-05, + "loss": 0.0072, + "step": 27820 + }, + { + "grad_norm": 0.22161483764648438, + "learning_rate": 6.004759714819516e-05, + "loss": 0.0055, + "step": 27830 + }, + { + "grad_norm": 0.19161266088485718, + "learning_rate": 6.002059994257323e-05, + "loss": 0.0051, + "step": 27840 + }, + { + "grad_norm": 0.18854838609695435, + "learning_rate": 5.999359969295764e-05, + "loss": 0.0053, + "step": 27850 + }, + { + "grad_norm": 0.1774907410144806, + "learning_rate": 5.9966596407550314e-05, + "loss": 0.0055, + "step": 27860 + }, + { + "grad_norm": 0.19992147386074066, + "learning_rate": 5.993959009455416e-05, + "loss": 0.0061, + "step": 27870 + }, + { + "grad_norm": 0.23237383365631104, + "learning_rate": 5.991258076217298e-05, + "loss": 0.0062, + "step": 27880 + }, + { + "grad_norm": 0.18414562940597534, + "learning_rate": 5.988556841861147e-05, + "loss": 0.0072, + "step": 27890 + }, + { + "grad_norm": 0.22399203479290009, + "learning_rate": 5.985855307207531e-05, + "loss": 0.0053, + "step": 27900 + }, + { + "grad_norm": 0.19938924908638, + "learning_rate": 5.9831534730771e-05, + "loss": 0.0079, + "step": 27910 + }, + { + "grad_norm": 0.18949943780899048, + "learning_rate": 5.980451340290605e-05, + "loss": 0.0049, + "step": 27920 + }, + { + "grad_norm": 0.17120018601417542, + "learning_rate": 5.97774890966888e-05, + "loss": 0.0073, + "step": 27930 + }, + { + "grad_norm": 0.3061679005622864, + "learning_rate": 5.975046182032851e-05, + "loss": 0.0049, + "step": 27940 + }, + { + "grad_norm": 0.25692275166511536, + "learning_rate": 5.972343158203537e-05, + "loss": 0.0083, + "step": 27950 + }, + { + "grad_norm": 0.25221413373947144, + "learning_rate": 5.969639839002045e-05, + "loss": 0.0058, + "step": 27960 + }, + { + "grad_norm": 0.23802326619625092, + "learning_rate": 5.966936225249572e-05, + "loss": 0.0064, + "step": 27970 + }, + { + "grad_norm": 0.2714694142341614, + "learning_rate": 5.9642323177674044e-05, + "loss": 0.0062, + "step": 27980 + }, + { + "grad_norm": 0.16080555319786072, + "learning_rate": 5.9615281173769154e-05, + "loss": 0.007, + "step": 27990 + }, + { + "grad_norm": 0.18601246178150177, + "learning_rate": 5.958823624899574e-05, + "loss": 0.0051, + "step": 28000 + }, + { + "grad_norm": 0.1622346192598343, + "learning_rate": 5.956118841156933e-05, + "loss": 0.0064, + "step": 28010 + }, + { + "grad_norm": 0.2637445330619812, + "learning_rate": 5.953413766970631e-05, + "loss": 0.0059, + "step": 28020 + }, + { + "grad_norm": 0.2083210051059723, + "learning_rate": 5.9507084031624e-05, + "loss": 0.0066, + "step": 28030 + }, + { + "grad_norm": 0.23868678510189056, + "learning_rate": 5.948002750554058e-05, + "loss": 0.0056, + "step": 28040 + }, + { + "grad_norm": 0.20230062305927277, + "learning_rate": 5.9452968099675124e-05, + "loss": 0.008, + "step": 28050 + }, + { + "grad_norm": 0.21888235211372375, + "learning_rate": 5.9425905822247527e-05, + "loss": 0.0059, + "step": 28060 + }, + { + "grad_norm": 0.1834922730922699, + "learning_rate": 5.939884068147864e-05, + "loss": 0.005, + "step": 28070 + }, + { + "grad_norm": 0.2151925414800644, + "learning_rate": 5.937177268559011e-05, + "loss": 0.005, + "step": 28080 + }, + { + "grad_norm": 0.15771934390068054, + "learning_rate": 5.934470184280448e-05, + "loss": 0.0042, + "step": 28090 + }, + { + "grad_norm": 0.1876213699579239, + "learning_rate": 5.931762816134516e-05, + "loss": 0.0071, + "step": 28100 + }, + { + "grad_norm": 0.11851833760738373, + "learning_rate": 5.9290551649436434e-05, + "loss": 0.0049, + "step": 28110 + }, + { + "grad_norm": 0.1731957197189331, + "learning_rate": 5.9263472315303416e-05, + "loss": 0.0047, + "step": 28120 + }, + { + "grad_norm": 0.18775789439678192, + "learning_rate": 5.9236390167172096e-05, + "loss": 0.0045, + "step": 28130 + }, + { + "grad_norm": 0.24825909733772278, + "learning_rate": 5.920930521326932e-05, + "loss": 0.0058, + "step": 28140 + }, + { + "grad_norm": 0.1368711143732071, + "learning_rate": 5.918221746182276e-05, + "loss": 0.0044, + "step": 28150 + }, + { + "grad_norm": 0.21826130151748657, + "learning_rate": 5.9155126921061e-05, + "loss": 0.0063, + "step": 28160 + }, + { + "grad_norm": 0.27194494009017944, + "learning_rate": 5.91280335992134e-05, + "loss": 0.0068, + "step": 28170 + }, + { + "grad_norm": 0.16480326652526855, + "learning_rate": 5.91009375045102e-05, + "loss": 0.0053, + "step": 28180 + }, + { + "grad_norm": 0.2562893331050873, + "learning_rate": 5.9073838645182476e-05, + "loss": 0.0067, + "step": 28190 + }, + { + "grad_norm": 0.28722527623176575, + "learning_rate": 5.904673702946217e-05, + "loss": 0.0057, + "step": 28200 + }, + { + "grad_norm": 0.2712259292602539, + "learning_rate": 5.9019632665582004e-05, + "loss": 0.0063, + "step": 28210 + }, + { + "grad_norm": 0.23969018459320068, + "learning_rate": 5.899252556177559e-05, + "loss": 0.0054, + "step": 28220 + }, + { + "grad_norm": 0.14926593005657196, + "learning_rate": 5.896541572627735e-05, + "loss": 0.0046, + "step": 28230 + }, + { + "grad_norm": 0.21841435134410858, + "learning_rate": 5.893830316732253e-05, + "loss": 0.0059, + "step": 28240 + }, + { + "grad_norm": 0.17031942307949066, + "learning_rate": 5.8911187893147214e-05, + "loss": 0.0052, + "step": 28250 + }, + { + "grad_norm": 0.19262637197971344, + "learning_rate": 5.888406991198828e-05, + "loss": 0.0058, + "step": 28260 + }, + { + "grad_norm": 0.2473677545785904, + "learning_rate": 5.885694923208349e-05, + "loss": 0.005, + "step": 28270 + }, + { + "grad_norm": 0.1710784137248993, + "learning_rate": 5.882982586167138e-05, + "loss": 0.0056, + "step": 28280 + }, + { + "grad_norm": 0.16016383469104767, + "learning_rate": 5.880269980899131e-05, + "loss": 0.0047, + "step": 28290 + }, + { + "grad_norm": 0.21569634974002838, + "learning_rate": 5.8775571082283465e-05, + "loss": 0.0054, + "step": 28300 + }, + { + "grad_norm": 0.2260180413722992, + "learning_rate": 5.8748439689788824e-05, + "loss": 0.0063, + "step": 28310 + }, + { + "grad_norm": 0.25348004698753357, + "learning_rate": 5.87213056397492e-05, + "loss": 0.0071, + "step": 28320 + }, + { + "grad_norm": 0.20415283739566803, + "learning_rate": 5.869416894040719e-05, + "loss": 0.0052, + "step": 28330 + }, + { + "grad_norm": 0.1627565622329712, + "learning_rate": 5.866702960000621e-05, + "loss": 0.0046, + "step": 28340 + }, + { + "grad_norm": 0.18913191556930542, + "learning_rate": 5.863988762679048e-05, + "loss": 0.0052, + "step": 28350 + }, + { + "grad_norm": 0.19791871309280396, + "learning_rate": 5.8612743029005e-05, + "loss": 0.005, + "step": 28360 + }, + { + "grad_norm": 0.17180344462394714, + "learning_rate": 5.858559581489561e-05, + "loss": 0.0044, + "step": 28370 + }, + { + "grad_norm": 0.1873185634613037, + "learning_rate": 5.85584459927089e-05, + "loss": 0.0059, + "step": 28380 + }, + { + "grad_norm": 0.18828752636909485, + "learning_rate": 5.853129357069227e-05, + "loss": 0.0059, + "step": 28390 + }, + { + "grad_norm": 0.1895962655544281, + "learning_rate": 5.8504138557093913e-05, + "loss": 0.0042, + "step": 28400 + }, + { + "grad_norm": 0.17760205268859863, + "learning_rate": 5.8476980960162784e-05, + "loss": 0.0074, + "step": 28410 + }, + { + "grad_norm": 0.20683133602142334, + "learning_rate": 5.844982078814868e-05, + "loss": 0.0054, + "step": 28420 + }, + { + "grad_norm": 0.1892419159412384, + "learning_rate": 5.842265804930211e-05, + "loss": 0.006, + "step": 28430 + }, + { + "grad_norm": 0.18328820168972015, + "learning_rate": 5.839549275187444e-05, + "loss": 0.0056, + "step": 28440 + }, + { + "grad_norm": 0.21130099892616272, + "learning_rate": 5.836832490411771e-05, + "loss": 0.0055, + "step": 28450 + }, + { + "grad_norm": 0.242754265666008, + "learning_rate": 5.834115451428485e-05, + "loss": 0.0055, + "step": 28460 + }, + { + "grad_norm": 0.2330566793680191, + "learning_rate": 5.831398159062946e-05, + "loss": 0.0075, + "step": 28470 + }, + { + "grad_norm": 0.2693495452404022, + "learning_rate": 5.828680614140599e-05, + "loss": 0.0069, + "step": 28480 + }, + { + "grad_norm": 0.22436171770095825, + "learning_rate": 5.825962817486962e-05, + "loss": 0.0061, + "step": 28490 + }, + { + "grad_norm": 0.20326124131679535, + "learning_rate": 5.823244769927629e-05, + "loss": 0.0062, + "step": 28500 + }, + { + "grad_norm": 0.2075069695711136, + "learning_rate": 5.8205264722882716e-05, + "loss": 0.0043, + "step": 28510 + }, + { + "grad_norm": 0.16633442044258118, + "learning_rate": 5.817807925394636e-05, + "loss": 0.0053, + "step": 28520 + }, + { + "grad_norm": 0.14915831387043, + "learning_rate": 5.815089130072546e-05, + "loss": 0.0061, + "step": 28530 + }, + { + "grad_norm": 0.175053209066391, + "learning_rate": 5.8123700871479e-05, + "loss": 0.0056, + "step": 28540 + }, + { + "grad_norm": 0.23188957571983337, + "learning_rate": 5.809650797446671e-05, + "loss": 0.0058, + "step": 28550 + }, + { + "grad_norm": 0.2453858107328415, + "learning_rate": 5.806931261794907e-05, + "loss": 0.0068, + "step": 28560 + }, + { + "grad_norm": 0.20580248534679413, + "learning_rate": 5.804211481018731e-05, + "loss": 0.0059, + "step": 28570 + }, + { + "grad_norm": 0.2458379715681076, + "learning_rate": 5.801491455944341e-05, + "loss": 0.0058, + "step": 28580 + }, + { + "grad_norm": 0.2027692049741745, + "learning_rate": 5.79877118739801e-05, + "loss": 0.0059, + "step": 28590 + }, + { + "grad_norm": 0.22771303355693817, + "learning_rate": 5.7960506762060816e-05, + "loss": 0.0047, + "step": 28600 + }, + { + "grad_norm": 0.21013477444648743, + "learning_rate": 5.793329923194977e-05, + "loss": 0.0052, + "step": 28610 + }, + { + "grad_norm": 0.19989684224128723, + "learning_rate": 5.790608929191187e-05, + "loss": 0.0056, + "step": 28620 + }, + { + "grad_norm": 0.16074591875076294, + "learning_rate": 5.78788769502128e-05, + "loss": 0.0072, + "step": 28630 + }, + { + "grad_norm": 0.17087243497371674, + "learning_rate": 5.785166221511894e-05, + "loss": 0.0047, + "step": 28640 + }, + { + "grad_norm": 0.22725346684455872, + "learning_rate": 5.7824445094897415e-05, + "loss": 0.0055, + "step": 28650 + }, + { + "grad_norm": 0.18366172909736633, + "learning_rate": 5.7797225597816065e-05, + "loss": 0.0043, + "step": 28660 + }, + { + "grad_norm": 0.1574377864599228, + "learning_rate": 5.777000373214345e-05, + "loss": 0.0045, + "step": 28670 + }, + { + "grad_norm": 0.2128681093454361, + "learning_rate": 5.774277950614885e-05, + "loss": 0.006, + "step": 28680 + }, + { + "grad_norm": 0.2590664327144623, + "learning_rate": 5.771555292810227e-05, + "loss": 0.0052, + "step": 28690 + }, + { + "grad_norm": 0.23427312076091766, + "learning_rate": 5.768832400627444e-05, + "loss": 0.0064, + "step": 28700 + }, + { + "grad_norm": 0.22861208021640778, + "learning_rate": 5.7661092748936775e-05, + "loss": 0.0062, + "step": 28710 + }, + { + "grad_norm": 0.20201127231121063, + "learning_rate": 5.76338591643614e-05, + "loss": 0.0046, + "step": 28720 + }, + { + "grad_norm": 0.2151762843132019, + "learning_rate": 5.760662326082118e-05, + "loss": 0.0059, + "step": 28730 + }, + { + "grad_norm": 0.1839437037706375, + "learning_rate": 5.757938504658965e-05, + "loss": 0.0056, + "step": 28740 + }, + { + "grad_norm": 0.1831098049879074, + "learning_rate": 5.755214452994107e-05, + "loss": 0.0057, + "step": 28750 + }, + { + "grad_norm": 0.140682652592659, + "learning_rate": 5.752490171915039e-05, + "loss": 0.0046, + "step": 28760 + }, + { + "grad_norm": 0.16588278114795685, + "learning_rate": 5.749765662249324e-05, + "loss": 0.0065, + "step": 28770 + }, + { + "grad_norm": 0.19379691779613495, + "learning_rate": 5.747040924824596e-05, + "loss": 0.0059, + "step": 28780 + }, + { + "grad_norm": 0.20092302560806274, + "learning_rate": 5.7443159604685613e-05, + "loss": 0.0052, + "step": 28790 + }, + { + "grad_norm": 0.21887630224227905, + "learning_rate": 5.74159077000899e-05, + "loss": 0.0049, + "step": 28800 + }, + { + "grad_norm": 0.2055760771036148, + "learning_rate": 5.7388653542737235e-05, + "loss": 0.0051, + "step": 28810 + }, + { + "grad_norm": 0.22898320853710175, + "learning_rate": 5.736139714090672e-05, + "loss": 0.008, + "step": 28820 + }, + { + "grad_norm": 0.2068106085062027, + "learning_rate": 5.73341385028781e-05, + "loss": 0.005, + "step": 28830 + }, + { + "grad_norm": 0.1926884949207306, + "learning_rate": 5.7306877636931855e-05, + "loss": 0.0058, + "step": 28840 + }, + { + "grad_norm": 0.18023042380809784, + "learning_rate": 5.7279614551349125e-05, + "loss": 0.0051, + "step": 28850 + }, + { + "grad_norm": 0.21538949012756348, + "learning_rate": 5.725234925441169e-05, + "loss": 0.0048, + "step": 28860 + }, + { + "grad_norm": 0.1922045350074768, + "learning_rate": 5.7225081754402044e-05, + "loss": 0.0054, + "step": 28870 + }, + { + "grad_norm": 0.19482499361038208, + "learning_rate": 5.7197812059603326e-05, + "loss": 0.0054, + "step": 28880 + }, + { + "grad_norm": 0.18533092737197876, + "learning_rate": 5.717054017829934e-05, + "loss": 0.0051, + "step": 28890 + }, + { + "grad_norm": 0.2453930377960205, + "learning_rate": 5.7143266118774584e-05, + "loss": 0.0062, + "step": 28900 + }, + { + "grad_norm": 0.24753063917160034, + "learning_rate": 5.711598988931418e-05, + "loss": 0.005, + "step": 28910 + }, + { + "grad_norm": 0.21387207508087158, + "learning_rate": 5.7088711498203954e-05, + "loss": 0.0046, + "step": 28920 + }, + { + "grad_norm": 0.2242448627948761, + "learning_rate": 5.706143095373033e-05, + "loss": 0.0067, + "step": 28930 + }, + { + "grad_norm": 0.17051264643669128, + "learning_rate": 5.703414826418042e-05, + "loss": 0.0046, + "step": 28940 + }, + { + "grad_norm": 0.1759202480316162, + "learning_rate": 5.7006863437842007e-05, + "loss": 0.0054, + "step": 28950 + }, + { + "grad_norm": 0.151921346783638, + "learning_rate": 5.697957648300348e-05, + "loss": 0.0076, + "step": 28960 + }, + { + "grad_norm": 0.22174721956253052, + "learning_rate": 5.695228740795391e-05, + "loss": 0.0056, + "step": 28970 + }, + { + "grad_norm": 0.19715653359889984, + "learning_rate": 5.6924996220982985e-05, + "loss": 0.0051, + "step": 28980 + }, + { + "grad_norm": 0.21052099764347076, + "learning_rate": 5.6897702930381045e-05, + "loss": 0.0064, + "step": 28990 + }, + { + "grad_norm": 0.19617202877998352, + "learning_rate": 5.687040754443908e-05, + "loss": 0.0056, + "step": 29000 + }, + { + "grad_norm": 0.20572853088378906, + "learning_rate": 5.6843110071448725e-05, + "loss": 0.0057, + "step": 29010 + }, + { + "grad_norm": 0.16405658423900604, + "learning_rate": 5.6815810519702194e-05, + "loss": 0.0057, + "step": 29020 + }, + { + "grad_norm": 0.1357373595237732, + "learning_rate": 5.6788508897492396e-05, + "loss": 0.0044, + "step": 29030 + }, + { + "grad_norm": 0.30207711458206177, + "learning_rate": 5.676120521311282e-05, + "loss": 0.0087, + "step": 29040 + }, + { + "grad_norm": 0.18085119128227234, + "learning_rate": 5.6733899474857634e-05, + "loss": 0.004, + "step": 29050 + }, + { + "grad_norm": 0.2029605656862259, + "learning_rate": 5.670659169102157e-05, + "loss": 0.0051, + "step": 29060 + }, + { + "grad_norm": 0.2264328896999359, + "learning_rate": 5.6679281869900044e-05, + "loss": 0.0063, + "step": 29070 + }, + { + "grad_norm": 0.16823899745941162, + "learning_rate": 5.6651970019789045e-05, + "loss": 0.0055, + "step": 29080 + }, + { + "grad_norm": 0.14622972905635834, + "learning_rate": 5.662465614898519e-05, + "loss": 0.0056, + "step": 29090 + }, + { + "grad_norm": 0.2224721759557724, + "learning_rate": 5.6597340265785695e-05, + "loss": 0.0056, + "step": 29100 + }, + { + "grad_norm": 0.18313968181610107, + "learning_rate": 5.657002237848843e-05, + "loss": 0.0044, + "step": 29110 + }, + { + "grad_norm": 0.2091536968946457, + "learning_rate": 5.654270249539183e-05, + "loss": 0.005, + "step": 29120 + }, + { + "grad_norm": 0.15692025423049927, + "learning_rate": 5.651538062479498e-05, + "loss": 0.0058, + "step": 29130 + }, + { + "grad_norm": 0.15955621004104614, + "learning_rate": 5.648805677499751e-05, + "loss": 0.0067, + "step": 29140 + }, + { + "grad_norm": 0.23243653774261475, + "learning_rate": 5.646073095429969e-05, + "loss": 0.0059, + "step": 29150 + }, + { + "grad_norm": 0.25523442029953003, + "learning_rate": 5.643340317100241e-05, + "loss": 0.0059, + "step": 29160 + }, + { + "grad_norm": 0.22557803988456726, + "learning_rate": 5.64060734334071e-05, + "loss": 0.0052, + "step": 29170 + }, + { + "grad_norm": 0.18731147050857544, + "learning_rate": 5.637874174981583e-05, + "loss": 0.0053, + "step": 29180 + }, + { + "grad_norm": 0.20181021094322205, + "learning_rate": 5.635140812853124e-05, + "loss": 0.005, + "step": 29190 + }, + { + "grad_norm": 0.17075319588184357, + "learning_rate": 5.6324072577856544e-05, + "loss": 0.0043, + "step": 29200 + }, + { + "grad_norm": 0.18515950441360474, + "learning_rate": 5.629673510609559e-05, + "loss": 0.0045, + "step": 29210 + }, + { + "grad_norm": 0.2035454660654068, + "learning_rate": 5.626939572155276e-05, + "loss": 0.0064, + "step": 29220 + }, + { + "grad_norm": 0.19095073640346527, + "learning_rate": 5.6242054432533054e-05, + "loss": 0.0054, + "step": 29230 + }, + { + "grad_norm": 0.16964705288410187, + "learning_rate": 5.621471124734201e-05, + "loss": 0.0069, + "step": 29240 + }, + { + "grad_norm": 0.22279876470565796, + "learning_rate": 5.6187366174285794e-05, + "loss": 0.006, + "step": 29250 + }, + { + "grad_norm": 0.18500900268554688, + "learning_rate": 5.616001922167109e-05, + "loss": 0.0046, + "step": 29260 + }, + { + "grad_norm": 0.16835862398147583, + "learning_rate": 5.61326703978052e-05, + "loss": 0.0048, + "step": 29270 + }, + { + "grad_norm": 0.22742272913455963, + "learning_rate": 5.6105319710995964e-05, + "loss": 0.0052, + "step": 29280 + }, + { + "grad_norm": 0.1859036237001419, + "learning_rate": 5.60779671695518e-05, + "loss": 0.0059, + "step": 29290 + }, + { + "grad_norm": 0.20393681526184082, + "learning_rate": 5.6050612781781684e-05, + "loss": 0.0067, + "step": 29300 + }, + { + "grad_norm": 0.19883304834365845, + "learning_rate": 5.602325655599516e-05, + "loss": 0.0052, + "step": 29310 + }, + { + "grad_norm": 0.15367934107780457, + "learning_rate": 5.599589850050234e-05, + "loss": 0.0056, + "step": 29320 + }, + { + "grad_norm": 0.15047653019428253, + "learning_rate": 5.5968538623613874e-05, + "loss": 0.006, + "step": 29330 + }, + { + "grad_norm": 0.18690043687820435, + "learning_rate": 5.594117693364095e-05, + "loss": 0.0049, + "step": 29340 + }, + { + "grad_norm": 0.21890491247177124, + "learning_rate": 5.591381343889535e-05, + "loss": 0.0062, + "step": 29350 + }, + { + "grad_norm": 0.16709180176258087, + "learning_rate": 5.5886448147689355e-05, + "loss": 0.0058, + "step": 29360 + }, + { + "grad_norm": 0.20897939801216125, + "learning_rate": 5.585908106833585e-05, + "loss": 0.0057, + "step": 29370 + }, + { + "grad_norm": 0.21187157928943634, + "learning_rate": 5.5831712209148226e-05, + "loss": 0.0052, + "step": 29380 + }, + { + "grad_norm": 0.16506585478782654, + "learning_rate": 5.58043415784404e-05, + "loss": 0.0048, + "step": 29390 + }, + { + "grad_norm": 0.2711997330188751, + "learning_rate": 5.577696918452686e-05, + "loss": 0.0056, + "step": 29400 + }, + { + "grad_norm": 0.2071024626493454, + "learning_rate": 5.5749595035722604e-05, + "loss": 0.0041, + "step": 29410 + }, + { + "grad_norm": 0.1898985505104065, + "learning_rate": 5.5722219140343193e-05, + "loss": 0.004, + "step": 29420 + }, + { + "grad_norm": 0.2062532603740692, + "learning_rate": 5.56948415067047e-05, + "loss": 0.006, + "step": 29430 + }, + { + "grad_norm": 0.19252778589725494, + "learning_rate": 5.5667462143123704e-05, + "loss": 0.0053, + "step": 29440 + }, + { + "grad_norm": 0.2357863485813141, + "learning_rate": 5.564008105791737e-05, + "loss": 0.0056, + "step": 29450 + }, + { + "grad_norm": 0.2055368274450302, + "learning_rate": 5.5612698259403316e-05, + "loss": 0.0059, + "step": 29460 + }, + { + "grad_norm": 0.2414434850215912, + "learning_rate": 5.5585313755899724e-05, + "loss": 0.0044, + "step": 29470 + }, + { + "grad_norm": 0.20956923067569733, + "learning_rate": 5.5557927555725285e-05, + "loss": 0.0057, + "step": 29480 + }, + { + "grad_norm": 0.14181269705295563, + "learning_rate": 5.55305396671992e-05, + "loss": 0.0042, + "step": 29490 + }, + { + "grad_norm": 0.19984854757785797, + "learning_rate": 5.55031500986412e-05, + "loss": 0.0056, + "step": 29500 + }, + { + "grad_norm": 0.1677868664264679, + "learning_rate": 5.547575885837149e-05, + "loss": 0.0062, + "step": 29510 + }, + { + "grad_norm": 0.2121850699186325, + "learning_rate": 5.5448365954710825e-05, + "loss": 0.0062, + "step": 29520 + }, + { + "grad_norm": 0.16570019721984863, + "learning_rate": 5.5420971395980446e-05, + "loss": 0.0045, + "step": 29530 + }, + { + "grad_norm": 0.30435338616371155, + "learning_rate": 5.539357519050209e-05, + "loss": 0.0079, + "step": 29540 + }, + { + "grad_norm": 0.1941782832145691, + "learning_rate": 5.536617734659799e-05, + "loss": 0.0055, + "step": 29550 + }, + { + "grad_norm": 0.20635464787483215, + "learning_rate": 5.533877787259091e-05, + "loss": 0.0066, + "step": 29560 + }, + { + "grad_norm": 0.22745567560195923, + "learning_rate": 5.5311376776804044e-05, + "loss": 0.0055, + "step": 29570 + }, + { + "grad_norm": 0.1521776020526886, + "learning_rate": 5.528397406756118e-05, + "loss": 0.005, + "step": 29580 + }, + { + "grad_norm": 0.14877501130104065, + "learning_rate": 5.525656975318652e-05, + "loss": 0.0051, + "step": 29590 + }, + { + "grad_norm": 0.18450523912906647, + "learning_rate": 5.522916384200474e-05, + "loss": 0.0068, + "step": 29600 + }, + { + "grad_norm": 0.30232444405555725, + "learning_rate": 5.520175634234106e-05, + "loss": 0.0086, + "step": 29610 + }, + { + "grad_norm": 0.24299657344818115, + "learning_rate": 5.517434726252113e-05, + "loss": 0.0068, + "step": 29620 + }, + { + "grad_norm": 0.22698938846588135, + "learning_rate": 5.514693661087113e-05, + "loss": 0.0049, + "step": 29630 + }, + { + "grad_norm": 0.24346503615379333, + "learning_rate": 5.511952439571769e-05, + "loss": 0.0058, + "step": 29640 + }, + { + "grad_norm": 0.19434130191802979, + "learning_rate": 5.509211062538791e-05, + "loss": 0.0071, + "step": 29650 + }, + { + "grad_norm": 0.39745065569877625, + "learning_rate": 5.506469530820939e-05, + "loss": 0.0065, + "step": 29660 + }, + { + "grad_norm": 0.19955550134181976, + "learning_rate": 5.503727845251014e-05, + "loss": 0.0072, + "step": 29670 + }, + { + "grad_norm": 0.251282662153244, + "learning_rate": 5.50098600666187e-05, + "loss": 0.0055, + "step": 29680 + }, + { + "grad_norm": 0.18520720303058624, + "learning_rate": 5.498244015886406e-05, + "loss": 0.0072, + "step": 29690 + }, + { + "grad_norm": 0.22318704426288605, + "learning_rate": 5.495501873757565e-05, + "loss": 0.0066, + "step": 29700 + }, + { + "grad_norm": 0.19224198162555695, + "learning_rate": 5.492759581108336e-05, + "loss": 0.0064, + "step": 29710 + }, + { + "grad_norm": 0.20975066721439362, + "learning_rate": 5.490017138771759e-05, + "loss": 0.0067, + "step": 29720 + }, + { + "grad_norm": 0.18289147317409515, + "learning_rate": 5.487274547580912e-05, + "loss": 0.0096, + "step": 29730 + }, + { + "grad_norm": 0.15758202970027924, + "learning_rate": 5.484531808368923e-05, + "loss": 0.0053, + "step": 29740 + }, + { + "grad_norm": 0.17166924476623535, + "learning_rate": 5.4817889219689656e-05, + "loss": 0.0066, + "step": 29750 + }, + { + "grad_norm": 0.2286929488182068, + "learning_rate": 5.4790458892142536e-05, + "loss": 0.0061, + "step": 29760 + }, + { + "grad_norm": 0.20322489738464355, + "learning_rate": 5.476302710938048e-05, + "loss": 0.0073, + "step": 29770 + }, + { + "grad_norm": 0.15503044426441193, + "learning_rate": 5.473559387973657e-05, + "loss": 0.0059, + "step": 29780 + }, + { + "grad_norm": 0.21335212886333466, + "learning_rate": 5.470815921154425e-05, + "loss": 0.0057, + "step": 29790 + }, + { + "grad_norm": 0.25170767307281494, + "learning_rate": 5.468072311313749e-05, + "loss": 0.0074, + "step": 29800 + }, + { + "grad_norm": 0.2249269336462021, + "learning_rate": 5.465328559285063e-05, + "loss": 0.0043, + "step": 29810 + }, + { + "grad_norm": 0.15145115554332733, + "learning_rate": 5.462584665901849e-05, + "loss": 0.0055, + "step": 29820 + }, + { + "grad_norm": 0.1966078132390976, + "learning_rate": 5.4598406319976235e-05, + "loss": 0.0056, + "step": 29830 + }, + { + "grad_norm": 0.16924650967121124, + "learning_rate": 5.457096458405958e-05, + "loss": 0.0052, + "step": 29840 + }, + { + "grad_norm": 0.21797440946102142, + "learning_rate": 5.454352145960457e-05, + "loss": 0.0063, + "step": 29850 + }, + { + "grad_norm": 0.2023417055606842, + "learning_rate": 5.4516076954947715e-05, + "loss": 0.0065, + "step": 29860 + }, + { + "grad_norm": 0.23208950459957123, + "learning_rate": 5.448863107842591e-05, + "loss": 0.0061, + "step": 29870 + }, + { + "grad_norm": 0.15935717523097992, + "learning_rate": 5.446118383837651e-05, + "loss": 0.0068, + "step": 29880 + }, + { + "grad_norm": 0.2505706250667572, + "learning_rate": 5.443373524313722e-05, + "loss": 0.0072, + "step": 29890 + }, + { + "grad_norm": 0.27601441740989685, + "learning_rate": 5.440628530104626e-05, + "loss": 0.0061, + "step": 29900 + }, + { + "grad_norm": 0.19104118645191193, + "learning_rate": 5.4378834020442146e-05, + "loss": 0.0074, + "step": 29910 + }, + { + "grad_norm": 0.1839321255683899, + "learning_rate": 5.4351381409663884e-05, + "loss": 0.0052, + "step": 29920 + }, + { + "grad_norm": 0.23367559909820557, + "learning_rate": 5.432392747705084e-05, + "loss": 0.0071, + "step": 29930 + }, + { + "grad_norm": 0.17363567650318146, + "learning_rate": 5.429647223094278e-05, + "loss": 0.0064, + "step": 29940 + }, + { + "grad_norm": 0.21096286177635193, + "learning_rate": 5.4269015679679924e-05, + "loss": 0.0061, + "step": 29950 + }, + { + "grad_norm": 0.15616314113140106, + "learning_rate": 5.424155783160281e-05, + "loss": 0.0043, + "step": 29960 + }, + { + "grad_norm": 0.17585046589374542, + "learning_rate": 5.4214098695052415e-05, + "loss": 0.0046, + "step": 29970 + }, + { + "grad_norm": 0.17712654173374176, + "learning_rate": 5.418663827837012e-05, + "loss": 0.0039, + "step": 29980 + }, + { + "grad_norm": 0.22880172729492188, + "learning_rate": 5.415917658989763e-05, + "loss": 0.0047, + "step": 29990 + }, + { + "grad_norm": 0.17595677077770233, + "learning_rate": 5.413171363797713e-05, + "loss": 0.0052, + "step": 30000 + }, + { + "grad_norm": 0.20251259207725525, + "learning_rate": 5.4104249430951116e-05, + "loss": 0.0047, + "step": 30010 + }, + { + "grad_norm": 0.14635731279850006, + "learning_rate": 5.4076783977162494e-05, + "loss": 0.0077, + "step": 30020 + }, + { + "grad_norm": 0.19002458453178406, + "learning_rate": 5.4049317284954525e-05, + "loss": 0.0057, + "step": 30030 + }, + { + "grad_norm": 0.15300147235393524, + "learning_rate": 5.4021849362670884e-05, + "loss": 0.0044, + "step": 30040 + }, + { + "grad_norm": 0.1581823080778122, + "learning_rate": 5.3994380218655604e-05, + "loss": 0.0061, + "step": 30050 + }, + { + "grad_norm": 0.3429616689682007, + "learning_rate": 5.396690986125309e-05, + "loss": 0.0071, + "step": 30060 + }, + { + "grad_norm": 0.27367961406707764, + "learning_rate": 5.3939438298808075e-05, + "loss": 0.0053, + "step": 30070 + }, + { + "grad_norm": 0.16606593132019043, + "learning_rate": 5.3911965539665744e-05, + "loss": 0.0049, + "step": 30080 + }, + { + "grad_norm": 0.1713843047618866, + "learning_rate": 5.388449159217156e-05, + "loss": 0.0063, + "step": 30090 + }, + { + "grad_norm": 0.1901925504207611, + "learning_rate": 5.3857016464671385e-05, + "loss": 0.0055, + "step": 30100 + }, + { + "grad_norm": 0.20465657114982605, + "learning_rate": 5.382954016551146e-05, + "loss": 0.0063, + "step": 30110 + }, + { + "grad_norm": 0.16736078262329102, + "learning_rate": 5.380206270303835e-05, + "loss": 0.0049, + "step": 30120 + }, + { + "grad_norm": 0.23199190199375153, + "learning_rate": 5.377458408559897e-05, + "loss": 0.005, + "step": 30130 + }, + { + "grad_norm": 0.17988093197345734, + "learning_rate": 5.374710432154061e-05, + "loss": 0.0089, + "step": 30140 + }, + { + "grad_norm": 0.2259960025548935, + "learning_rate": 5.3719623419210886e-05, + "loss": 0.0057, + "step": 30150 + }, + { + "grad_norm": 0.17799469828605652, + "learning_rate": 5.3692141386957786e-05, + "loss": 0.0057, + "step": 30160 + }, + { + "grad_norm": 0.17998215556144714, + "learning_rate": 5.3664658233129616e-05, + "loss": 0.0064, + "step": 30170 + }, + { + "grad_norm": 0.2078496515750885, + "learning_rate": 5.363717396607504e-05, + "loss": 0.005, + "step": 30180 + }, + { + "grad_norm": 0.19409272074699402, + "learning_rate": 5.360968859414305e-05, + "loss": 0.0051, + "step": 30190 + }, + { + "grad_norm": 0.20358821749687195, + "learning_rate": 5.358220212568295e-05, + "loss": 0.006, + "step": 30200 + }, + { + "grad_norm": 0.176359623670578, + "learning_rate": 5.355471456904444e-05, + "loss": 0.0058, + "step": 30210 + }, + { + "grad_norm": 0.2315361052751541, + "learning_rate": 5.3527225932577495e-05, + "loss": 0.0077, + "step": 30220 + }, + { + "grad_norm": 0.22574542462825775, + "learning_rate": 5.349973622463246e-05, + "loss": 0.0047, + "step": 30230 + }, + { + "grad_norm": 0.25905847549438477, + "learning_rate": 5.3472245453559956e-05, + "loss": 0.0054, + "step": 30240 + }, + { + "grad_norm": 0.24247246980667114, + "learning_rate": 5.3444753627710955e-05, + "loss": 0.0078, + "step": 30250 + }, + { + "grad_norm": 0.1722092181444168, + "learning_rate": 5.341726075543676e-05, + "loss": 0.0048, + "step": 30260 + }, + { + "grad_norm": 0.17815691232681274, + "learning_rate": 5.338976684508898e-05, + "loss": 0.0059, + "step": 30270 + }, + { + "grad_norm": 0.23875965178012848, + "learning_rate": 5.336227190501953e-05, + "loss": 0.0075, + "step": 30280 + }, + { + "grad_norm": 0.20787908136844635, + "learning_rate": 5.3334775943580664e-05, + "loss": 0.0057, + "step": 30290 + }, + { + "grad_norm": 0.15233765542507172, + "learning_rate": 5.330727896912491e-05, + "loss": 0.0049, + "step": 30300 + }, + { + "grad_norm": 0.16355381906032562, + "learning_rate": 5.327978099000511e-05, + "loss": 0.0044, + "step": 30310 + }, + { + "grad_norm": 0.18125870823860168, + "learning_rate": 5.3252282014574465e-05, + "loss": 0.0059, + "step": 30320 + }, + { + "grad_norm": 0.19926011562347412, + "learning_rate": 5.322478205118641e-05, + "loss": 0.0052, + "step": 30330 + }, + { + "grad_norm": 0.26833435893058777, + "learning_rate": 5.3197281108194704e-05, + "loss": 0.0061, + "step": 30340 + }, + { + "grad_norm": 0.2054644078016281, + "learning_rate": 5.316977919395342e-05, + "loss": 0.0059, + "step": 30350 + }, + { + "grad_norm": 0.20752538740634918, + "learning_rate": 5.314227631681691e-05, + "loss": 0.006, + "step": 30360 + }, + { + "grad_norm": 0.1510171741247177, + "learning_rate": 5.311477248513982e-05, + "loss": 0.0065, + "step": 30370 + }, + { + "grad_norm": 0.18885274231433868, + "learning_rate": 5.30872677072771e-05, + "loss": 0.0046, + "step": 30380 + }, + { + "grad_norm": 0.16170471906661987, + "learning_rate": 5.3059761991583954e-05, + "loss": 0.0051, + "step": 30390 + }, + { + "grad_norm": 0.14873333275318146, + "learning_rate": 5.303225534641592e-05, + "loss": 0.0043, + "step": 30400 + }, + { + "grad_norm": 0.16585133969783783, + "learning_rate": 5.300474778012875e-05, + "loss": 0.0051, + "step": 30410 + }, + { + "grad_norm": 0.18362580239772797, + "learning_rate": 5.297723930107855e-05, + "loss": 0.005, + "step": 30420 + }, + { + "grad_norm": 0.13625064492225647, + "learning_rate": 5.294972991762167e-05, + "loss": 0.0031, + "step": 30430 + }, + { + "grad_norm": 0.17825248837471008, + "learning_rate": 5.292221963811472e-05, + "loss": 0.004, + "step": 30440 + }, + { + "grad_norm": 0.25440919399261475, + "learning_rate": 5.28947084709146e-05, + "loss": 0.0045, + "step": 30450 + }, + { + "grad_norm": 0.1803164780139923, + "learning_rate": 5.2867196424378465e-05, + "loss": 0.0041, + "step": 30460 + }, + { + "grad_norm": 0.2411777675151825, + "learning_rate": 5.2839683506863765e-05, + "loss": 0.005, + "step": 30470 + }, + { + "grad_norm": 0.17569312453269958, + "learning_rate": 5.281216972672821e-05, + "loss": 0.0043, + "step": 30480 + }, + { + "grad_norm": 0.2172713726758957, + "learning_rate": 5.278465509232973e-05, + "loss": 0.005, + "step": 30490 + }, + { + "grad_norm": 0.23277314007282257, + "learning_rate": 5.275713961202655e-05, + "loss": 0.0056, + "step": 30500 + }, + { + "grad_norm": 0.25411856174468994, + "learning_rate": 5.2729623294177165e-05, + "loss": 0.0051, + "step": 30510 + }, + { + "grad_norm": 0.16799281537532806, + "learning_rate": 5.270210614714028e-05, + "loss": 0.0046, + "step": 30520 + }, + { + "grad_norm": 0.21863852441310883, + "learning_rate": 5.267458817927491e-05, + "loss": 0.0058, + "step": 30530 + }, + { + "grad_norm": 0.16732220351696014, + "learning_rate": 5.264706939894026e-05, + "loss": 0.0052, + "step": 30540 + }, + { + "grad_norm": 0.15635059773921967, + "learning_rate": 5.261954981449584e-05, + "loss": 0.0042, + "step": 30550 + }, + { + "grad_norm": 0.19723793864250183, + "learning_rate": 5.2592029434301324e-05, + "loss": 0.0058, + "step": 30560 + }, + { + "grad_norm": 0.17818093299865723, + "learning_rate": 5.256450826671672e-05, + "loss": 0.0083, + "step": 30570 + }, + { + "grad_norm": 0.20513881742954254, + "learning_rate": 5.253698632010221e-05, + "loss": 0.0062, + "step": 30580 + }, + { + "grad_norm": 0.18832646310329437, + "learning_rate": 5.2509463602818246e-05, + "loss": 0.0041, + "step": 30590 + }, + { + "grad_norm": 0.24967971444129944, + "learning_rate": 5.248194012322549e-05, + "loss": 0.0049, + "step": 30600 + }, + { + "grad_norm": 0.22181031107902527, + "learning_rate": 5.245441588968486e-05, + "loss": 0.0062, + "step": 30610 + }, + { + "grad_norm": 0.19122512638568878, + "learning_rate": 5.242689091055748e-05, + "loss": 0.004, + "step": 30620 + }, + { + "grad_norm": 0.1803915798664093, + "learning_rate": 5.239936519420473e-05, + "loss": 0.0058, + "step": 30630 + }, + { + "grad_norm": 0.22564992308616638, + "learning_rate": 5.2371838748988175e-05, + "loss": 0.0056, + "step": 30640 + }, + { + "grad_norm": 0.184902623295784, + "learning_rate": 5.234431158326965e-05, + "loss": 0.0047, + "step": 30650 + }, + { + "grad_norm": 0.23818105459213257, + "learning_rate": 5.231678370541115e-05, + "loss": 0.0056, + "step": 30660 + }, + { + "grad_norm": 0.1931838095188141, + "learning_rate": 5.228925512377495e-05, + "loss": 0.0042, + "step": 30670 + }, + { + "grad_norm": 0.25122156739234924, + "learning_rate": 5.2261725846723465e-05, + "loss": 0.0047, + "step": 30680 + }, + { + "grad_norm": 0.17127318680286407, + "learning_rate": 5.22341958826194e-05, + "loss": 0.0053, + "step": 30690 + }, + { + "grad_norm": 0.19694863259792328, + "learning_rate": 5.22066652398256e-05, + "loss": 0.0044, + "step": 30700 + }, + { + "grad_norm": 0.17200535535812378, + "learning_rate": 5.2179133926705185e-05, + "loss": 0.0055, + "step": 30710 + }, + { + "grad_norm": 0.20394374430179596, + "learning_rate": 5.215160195162141e-05, + "loss": 0.0046, + "step": 30720 + }, + { + "grad_norm": 0.20772460103034973, + "learning_rate": 5.212406932293776e-05, + "loss": 0.0055, + "step": 30730 + }, + { + "grad_norm": 0.17080987989902496, + "learning_rate": 5.209653604901795e-05, + "loss": 0.0057, + "step": 30740 + }, + { + "grad_norm": 0.16024409234523773, + "learning_rate": 5.206900213822584e-05, + "loss": 0.0041, + "step": 30750 + }, + { + "grad_norm": 0.14029310643672943, + "learning_rate": 5.204146759892551e-05, + "loss": 0.005, + "step": 30760 + }, + { + "grad_norm": 0.20849694311618805, + "learning_rate": 5.2013932439481216e-05, + "loss": 0.0052, + "step": 30770 + }, + { + "grad_norm": 0.18949127197265625, + "learning_rate": 5.198639666825743e-05, + "loss": 0.0065, + "step": 30780 + }, + { + "grad_norm": 0.22011834383010864, + "learning_rate": 5.195886029361877e-05, + "loss": 0.0043, + "step": 30790 + }, + { + "grad_norm": 0.2479807585477829, + "learning_rate": 5.193132332393009e-05, + "loss": 0.0065, + "step": 30800 + }, + { + "grad_norm": 0.23610201478004456, + "learning_rate": 5.1903785767556376e-05, + "loss": 0.0058, + "step": 30810 + }, + { + "grad_norm": 0.2370217740535736, + "learning_rate": 5.187624763286282e-05, + "loss": 0.0061, + "step": 30820 + }, + { + "grad_norm": 0.19059418141841888, + "learning_rate": 5.184870892821475e-05, + "loss": 0.0065, + "step": 30830 + }, + { + "grad_norm": 0.20750829577445984, + "learning_rate": 5.182116966197773e-05, + "loss": 0.0042, + "step": 30840 + }, + { + "grad_norm": 0.1982104480266571, + "learning_rate": 5.1793629842517466e-05, + "loss": 0.0045, + "step": 30850 + }, + { + "grad_norm": 0.2042876034975052, + "learning_rate": 5.17660894781998e-05, + "loss": 0.0061, + "step": 30860 + }, + { + "grad_norm": 0.2952621579170227, + "learning_rate": 5.173854857739079e-05, + "loss": 0.0076, + "step": 30870 + }, + { + "grad_norm": 0.18933899700641632, + "learning_rate": 5.171100714845661e-05, + "loss": 0.0057, + "step": 30880 + }, + { + "grad_norm": 0.1879114955663681, + "learning_rate": 5.1683465199763646e-05, + "loss": 0.0049, + "step": 30890 + }, + { + "grad_norm": 0.23395439982414246, + "learning_rate": 5.16559227396784e-05, + "loss": 0.0048, + "step": 30900 + }, + { + "grad_norm": 0.2327568233013153, + "learning_rate": 5.1628379776567556e-05, + "loss": 0.0053, + "step": 30910 + }, + { + "grad_norm": 0.17762382328510284, + "learning_rate": 5.160083631879792e-05, + "loss": 0.0052, + "step": 30920 + }, + { + "grad_norm": 0.23868829011917114, + "learning_rate": 5.1573292374736484e-05, + "loss": 0.0052, + "step": 30930 + }, + { + "grad_norm": 0.16216345131397247, + "learning_rate": 5.1545747952750356e-05, + "loss": 0.0042, + "step": 30940 + }, + { + "grad_norm": 0.13944071531295776, + "learning_rate": 5.151820306120682e-05, + "loss": 0.0043, + "step": 30950 + }, + { + "grad_norm": 0.14964541792869568, + "learning_rate": 5.149065770847328e-05, + "loss": 0.0054, + "step": 30960 + }, + { + "grad_norm": 0.1824244260787964, + "learning_rate": 5.1463111902917297e-05, + "loss": 0.0046, + "step": 30970 + }, + { + "grad_norm": 0.1795882135629654, + "learning_rate": 5.143556565290654e-05, + "loss": 0.0044, + "step": 30980 + }, + { + "grad_norm": 0.1381848007440567, + "learning_rate": 5.140801896680882e-05, + "loss": 0.0064, + "step": 30990 + }, + { + "grad_norm": 0.15474601089954376, + "learning_rate": 5.1380471852992144e-05, + "loss": 0.0041, + "step": 31000 + }, + { + "grad_norm": 0.1834709495306015, + "learning_rate": 5.135292431982457e-05, + "loss": 0.0049, + "step": 31010 + }, + { + "grad_norm": 0.24421149492263794, + "learning_rate": 5.1325376375674294e-05, + "loss": 0.0056, + "step": 31020 + }, + { + "grad_norm": 0.1933891922235489, + "learning_rate": 5.129782802890968e-05, + "loss": 0.0045, + "step": 31030 + }, + { + "grad_norm": 0.2036387324333191, + "learning_rate": 5.127027928789916e-05, + "loss": 0.007, + "step": 31040 + }, + { + "grad_norm": 0.2099962830543518, + "learning_rate": 5.124273016101135e-05, + "loss": 0.0049, + "step": 31050 + }, + { + "grad_norm": 0.19814443588256836, + "learning_rate": 5.121518065661492e-05, + "loss": 0.0047, + "step": 31060 + }, + { + "grad_norm": 0.2068185955286026, + "learning_rate": 5.11876307830787e-05, + "loss": 0.0057, + "step": 31070 + }, + { + "grad_norm": 0.16235095262527466, + "learning_rate": 5.1160080548771596e-05, + "loss": 0.0057, + "step": 31080 + }, + { + "grad_norm": 0.12750132381916046, + "learning_rate": 5.1132529962062656e-05, + "loss": 0.0044, + "step": 31090 + }, + { + "grad_norm": 0.2374095767736435, + "learning_rate": 5.110497903132101e-05, + "loss": 0.0057, + "step": 31100 + }, + { + "grad_norm": 0.21322056651115417, + "learning_rate": 5.107742776491592e-05, + "loss": 0.0047, + "step": 31110 + }, + { + "grad_norm": 0.24200071394443512, + "learning_rate": 5.104987617121673e-05, + "loss": 0.0061, + "step": 31120 + }, + { + "grad_norm": 0.15461426973342896, + "learning_rate": 5.102232425859287e-05, + "loss": 0.0046, + "step": 31130 + }, + { + "grad_norm": 0.20958973467350006, + "learning_rate": 5.09947720354139e-05, + "loss": 0.0043, + "step": 31140 + }, + { + "grad_norm": 0.17817848920822144, + "learning_rate": 5.096721951004942e-05, + "loss": 0.0047, + "step": 31150 + }, + { + "grad_norm": 0.18468694388866425, + "learning_rate": 5.0939666690869227e-05, + "loss": 0.0059, + "step": 31160 + }, + { + "grad_norm": 0.17305415868759155, + "learning_rate": 5.0912113586243096e-05, + "loss": 0.0043, + "step": 31170 + }, + { + "grad_norm": 0.2267351597547531, + "learning_rate": 5.0884560204540935e-05, + "loss": 0.005, + "step": 31180 + }, + { + "grad_norm": 0.16885721683502197, + "learning_rate": 5.0857006554132736e-05, + "loss": 0.0039, + "step": 31190 + }, + { + "grad_norm": 0.24405206739902496, + "learning_rate": 5.0829452643388575e-05, + "loss": 0.004, + "step": 31200 + }, + { + "grad_norm": 0.20062245428562164, + "learning_rate": 5.08018984806786e-05, + "loss": 0.0049, + "step": 31210 + }, + { + "grad_norm": 0.23667308688163757, + "learning_rate": 5.0774344074373036e-05, + "loss": 0.0062, + "step": 31220 + }, + { + "grad_norm": 0.23339824378490448, + "learning_rate": 5.07467894328422e-05, + "loss": 0.0055, + "step": 31230 + }, + { + "grad_norm": 0.1691211313009262, + "learning_rate": 5.0719234564456454e-05, + "loss": 0.0071, + "step": 31240 + }, + { + "grad_norm": 0.2070344090461731, + "learning_rate": 5.0691679477586216e-05, + "loss": 0.0052, + "step": 31250 + }, + { + "grad_norm": 0.17927901446819305, + "learning_rate": 5.0664124180602035e-05, + "loss": 0.0045, + "step": 31260 + }, + { + "grad_norm": 0.15509898960590363, + "learning_rate": 5.063656868187447e-05, + "loss": 0.0056, + "step": 31270 + }, + { + "grad_norm": 0.14657287299633026, + "learning_rate": 5.060901298977413e-05, + "loss": 0.0075, + "step": 31280 + }, + { + "grad_norm": 0.24408428370952606, + "learning_rate": 5.0581457112671725e-05, + "loss": 0.0065, + "step": 31290 + }, + { + "grad_norm": 0.20375415682792664, + "learning_rate": 5.0553901058938016e-05, + "loss": 0.007, + "step": 31300 + }, + { + "grad_norm": 0.20089338719844818, + "learning_rate": 5.052634483694377e-05, + "loss": 0.0046, + "step": 31310 + }, + { + "grad_norm": 0.18726156651973724, + "learning_rate": 5.049878845505988e-05, + "loss": 0.0046, + "step": 31320 + }, + { + "grad_norm": 0.2967512309551239, + "learning_rate": 5.047123192165721e-05, + "loss": 0.0052, + "step": 31330 + }, + { + "grad_norm": 0.20723484456539154, + "learning_rate": 5.0443675245106735e-05, + "loss": 0.0055, + "step": 31340 + }, + { + "grad_norm": 0.2124812752008438, + "learning_rate": 5.0416118433779426e-05, + "loss": 0.006, + "step": 31350 + }, + { + "grad_norm": 0.19948455691337585, + "learning_rate": 5.038856149604633e-05, + "loss": 0.0054, + "step": 31360 + }, + { + "grad_norm": 0.22487370669841766, + "learning_rate": 5.03610044402785e-05, + "loss": 0.0059, + "step": 31370 + }, + { + "grad_norm": 0.21467886865139008, + "learning_rate": 5.033344727484707e-05, + "loss": 0.0055, + "step": 31380 + }, + { + "grad_norm": 0.19325312972068787, + "learning_rate": 5.030589000812315e-05, + "loss": 0.0075, + "step": 31390 + }, + { + "grad_norm": 0.18458259105682373, + "learning_rate": 5.027833264847793e-05, + "loss": 0.0055, + "step": 31400 + }, + { + "grad_norm": 0.17840997874736786, + "learning_rate": 5.025077520428258e-05, + "loss": 0.0065, + "step": 31410 + }, + { + "grad_norm": 0.21097692847251892, + "learning_rate": 5.022321768390837e-05, + "loss": 0.0067, + "step": 31420 + }, + { + "grad_norm": 0.18814191222190857, + "learning_rate": 5.0195660095726516e-05, + "loss": 0.0047, + "step": 31430 + }, + { + "grad_norm": 0.21292294561862946, + "learning_rate": 5.016810244810829e-05, + "loss": 0.0052, + "step": 31440 + }, + { + "grad_norm": 0.26881682872772217, + "learning_rate": 5.0140544749424976e-05, + "loss": 0.0048, + "step": 31450 + }, + { + "grad_norm": 0.2497543841600418, + "learning_rate": 5.0112987008047874e-05, + "loss": 0.0063, + "step": 31460 + }, + { + "grad_norm": 0.18842750787734985, + "learning_rate": 5.008542923234831e-05, + "loss": 0.0051, + "step": 31470 + }, + { + "grad_norm": 0.17844758927822113, + "learning_rate": 5.00578714306976e-05, + "loss": 0.0055, + "step": 31480 + }, + { + "grad_norm": 0.15989556908607483, + "learning_rate": 5.0030313611467084e-05, + "loss": 0.0054, + "step": 31490 + }, + { + "grad_norm": 0.190887913107872, + "learning_rate": 5.0002755783028074e-05, + "loss": 0.0055, + "step": 31500 + }, + { + "grad_norm": 0.19800716638565063, + "learning_rate": 4.997519795375194e-05, + "loss": 0.0045, + "step": 31510 + }, + { + "grad_norm": 0.16440793871879578, + "learning_rate": 4.9947640132010016e-05, + "loss": 0.0047, + "step": 31520 + }, + { + "grad_norm": 0.18869194388389587, + "learning_rate": 4.9920082326173625e-05, + "loss": 0.0046, + "step": 31530 + }, + { + "grad_norm": 0.20321868360042572, + "learning_rate": 4.9892524544614114e-05, + "loss": 0.0065, + "step": 31540 + }, + { + "grad_norm": 0.27668923139572144, + "learning_rate": 4.986496679570283e-05, + "loss": 0.0082, + "step": 31550 + }, + { + "grad_norm": 0.24034807085990906, + "learning_rate": 4.983740908781105e-05, + "loss": 0.0044, + "step": 31560 + }, + { + "grad_norm": 0.15371045470237732, + "learning_rate": 4.9809851429310116e-05, + "loss": 0.004, + "step": 31570 + }, + { + "grad_norm": 0.20635126531124115, + "learning_rate": 4.9782293828571275e-05, + "loss": 0.0041, + "step": 31580 + }, + { + "grad_norm": 0.1869991570711136, + "learning_rate": 4.9754736293965846e-05, + "loss": 0.0045, + "step": 31590 + }, + { + "grad_norm": 0.17415070533752441, + "learning_rate": 4.972717883386502e-05, + "loss": 0.0046, + "step": 31600 + }, + { + "grad_norm": 0.16830730438232422, + "learning_rate": 4.9699621456640075e-05, + "loss": 0.0044, + "step": 31610 + }, + { + "grad_norm": 0.21755871176719666, + "learning_rate": 4.9672064170662214e-05, + "loss": 0.007, + "step": 31620 + }, + { + "grad_norm": 0.21022994816303253, + "learning_rate": 4.9644506984302583e-05, + "loss": 0.0036, + "step": 31630 + }, + { + "grad_norm": 0.2332187443971634, + "learning_rate": 4.9616949905932356e-05, + "loss": 0.0059, + "step": 31640 + }, + { + "grad_norm": 0.1786932349205017, + "learning_rate": 4.9589392943922615e-05, + "loss": 0.005, + "step": 31650 + }, + { + "grad_norm": 0.27448326349258423, + "learning_rate": 4.956183610664447e-05, + "loss": 0.0059, + "step": 31660 + }, + { + "grad_norm": 0.24794825911521912, + "learning_rate": 4.9534279402468945e-05, + "loss": 0.0067, + "step": 31670 + }, + { + "grad_norm": 0.17630399763584137, + "learning_rate": 4.9506722839767036e-05, + "loss": 0.0052, + "step": 31680 + }, + { + "grad_norm": 0.14124633371829987, + "learning_rate": 4.947916642690972e-05, + "loss": 0.0042, + "step": 31690 + }, + { + "grad_norm": 0.22503885626792908, + "learning_rate": 4.9451610172267874e-05, + "loss": 0.0054, + "step": 31700 + }, + { + "grad_norm": 0.23687101900577545, + "learning_rate": 4.9424054084212376e-05, + "loss": 0.0044, + "step": 31710 + }, + { + "grad_norm": 0.15688809752464294, + "learning_rate": 4.939649817111407e-05, + "loss": 0.004, + "step": 31720 + }, + { + "grad_norm": 0.27571186423301697, + "learning_rate": 4.936894244134365e-05, + "loss": 0.0043, + "step": 31730 + }, + { + "grad_norm": 0.2929707467556, + "learning_rate": 4.9341386903271886e-05, + "loss": 0.005, + "step": 31740 + }, + { + "grad_norm": 0.17888648808002472, + "learning_rate": 4.931383156526936e-05, + "loss": 0.0058, + "step": 31750 + }, + { + "grad_norm": 0.2074800282716751, + "learning_rate": 4.92862764357067e-05, + "loss": 0.0053, + "step": 31760 + }, + { + "grad_norm": 0.24055776000022888, + "learning_rate": 4.925872152295443e-05, + "loss": 0.0054, + "step": 31770 + }, + { + "grad_norm": 0.19162774085998535, + "learning_rate": 4.923116683538296e-05, + "loss": 0.0045, + "step": 31780 + }, + { + "grad_norm": 0.24251529574394226, + "learning_rate": 4.920361238136273e-05, + "loss": 0.0054, + "step": 31790 + }, + { + "grad_norm": 0.2408338189125061, + "learning_rate": 4.9176058169264014e-05, + "loss": 0.005, + "step": 31800 + }, + { + "grad_norm": 0.19583497941493988, + "learning_rate": 4.9148504207457074e-05, + "loss": 0.0049, + "step": 31810 + }, + { + "grad_norm": 0.1445378214120865, + "learning_rate": 4.912095050431208e-05, + "loss": 0.0043, + "step": 31820 + }, + { + "grad_norm": 0.13216830790042877, + "learning_rate": 4.909339706819911e-05, + "loss": 0.0045, + "step": 31830 + }, + { + "grad_norm": 0.18678554892539978, + "learning_rate": 4.906584390748819e-05, + "loss": 0.0045, + "step": 31840 + }, + { + "grad_norm": 0.19915135204792023, + "learning_rate": 4.9038291030549195e-05, + "loss": 0.0047, + "step": 31850 + }, + { + "grad_norm": 0.2730449438095093, + "learning_rate": 4.9010738445751995e-05, + "loss": 0.0051, + "step": 31860 + }, + { + "grad_norm": 0.16995364427566528, + "learning_rate": 4.8983186161466364e-05, + "loss": 0.0044, + "step": 31870 + }, + { + "grad_norm": 0.15751521289348602, + "learning_rate": 4.89556341860619e-05, + "loss": 0.0044, + "step": 31880 + }, + { + "grad_norm": 0.17200878262519836, + "learning_rate": 4.892808252790822e-05, + "loss": 0.0035, + "step": 31890 + }, + { + "grad_norm": 0.2017274796962738, + "learning_rate": 4.890053119537475e-05, + "loss": 0.0041, + "step": 31900 + }, + { + "grad_norm": 0.1692100614309311, + "learning_rate": 4.887298019683087e-05, + "loss": 0.0048, + "step": 31910 + }, + { + "grad_norm": 0.22506344318389893, + "learning_rate": 4.884542954064587e-05, + "loss": 0.0045, + "step": 31920 + }, + { + "grad_norm": 0.24472633004188538, + "learning_rate": 4.881787923518887e-05, + "loss": 0.0048, + "step": 31930 + }, + { + "grad_norm": 0.19130097329616547, + "learning_rate": 4.879032928882896e-05, + "loss": 0.007, + "step": 31940 + }, + { + "grad_norm": 0.2052450329065323, + "learning_rate": 4.876277970993505e-05, + "loss": 0.0049, + "step": 31950 + }, + { + "grad_norm": 0.1908828765153885, + "learning_rate": 4.873523050687602e-05, + "loss": 0.0082, + "step": 31960 + }, + { + "grad_norm": 0.16499602794647217, + "learning_rate": 4.870768168802056e-05, + "loss": 0.0046, + "step": 31970 + }, + { + "grad_norm": 0.15493258833885193, + "learning_rate": 4.868013326173728e-05, + "loss": 0.0058, + "step": 31980 + }, + { + "grad_norm": 0.12490259110927582, + "learning_rate": 4.865258523639468e-05, + "loss": 0.0045, + "step": 31990 + }, + { + "grad_norm": 0.15310634672641754, + "learning_rate": 4.862503762036109e-05, + "loss": 0.0047, + "step": 32000 + }, + { + "grad_norm": 0.19791896641254425, + "learning_rate": 4.859749042200478e-05, + "loss": 0.0063, + "step": 32010 + }, + { + "grad_norm": 0.2279396653175354, + "learning_rate": 4.856994364969384e-05, + "loss": 0.0047, + "step": 32020 + }, + { + "grad_norm": 0.1784597784280777, + "learning_rate": 4.854239731179625e-05, + "loss": 0.0052, + "step": 32030 + }, + { + "grad_norm": 0.14760909974575043, + "learning_rate": 4.85148514166799e-05, + "loss": 0.0042, + "step": 32040 + }, + { + "grad_norm": 0.19168606400489807, + "learning_rate": 4.8487305972712456e-05, + "loss": 0.004, + "step": 32050 + }, + { + "grad_norm": 0.23768550157546997, + "learning_rate": 4.8459760988261526e-05, + "loss": 0.005, + "step": 32060 + }, + { + "grad_norm": 0.15142159163951874, + "learning_rate": 4.843221647169453e-05, + "loss": 0.0034, + "step": 32070 + }, + { + "grad_norm": 0.21645669639110565, + "learning_rate": 4.840467243137878e-05, + "loss": 0.0053, + "step": 32080 + }, + { + "grad_norm": 0.20354746282100677, + "learning_rate": 4.837712887568143e-05, + "loss": 0.0053, + "step": 32090 + }, + { + "grad_norm": 0.2716869115829468, + "learning_rate": 4.8349585812969464e-05, + "loss": 0.0041, + "step": 32100 + }, + { + "grad_norm": 0.20231391489505768, + "learning_rate": 4.8322043251609775e-05, + "loss": 0.0047, + "step": 32110 + }, + { + "grad_norm": 0.14112946391105652, + "learning_rate": 4.8294501199969015e-05, + "loss": 0.005, + "step": 32120 + }, + { + "grad_norm": 0.28037208318710327, + "learning_rate": 4.826695966641376e-05, + "loss": 0.0048, + "step": 32130 + }, + { + "grad_norm": 0.16090203821659088, + "learning_rate": 4.823941865931043e-05, + "loss": 0.006, + "step": 32140 + }, + { + "grad_norm": 0.15256863832473755, + "learning_rate": 4.82118781870252e-05, + "loss": 0.0066, + "step": 32150 + }, + { + "grad_norm": 0.1753431260585785, + "learning_rate": 4.8184338257924185e-05, + "loss": 0.0041, + "step": 32160 + }, + { + "grad_norm": 0.19041666388511658, + "learning_rate": 4.815679888037324e-05, + "loss": 0.0039, + "step": 32170 + }, + { + "grad_norm": 0.22406138479709625, + "learning_rate": 4.8129260062738135e-05, + "loss": 0.0047, + "step": 32180 + }, + { + "grad_norm": 0.1523721069097519, + "learning_rate": 4.810172181338445e-05, + "loss": 0.0037, + "step": 32190 + }, + { + "grad_norm": 0.20484371483325958, + "learning_rate": 4.807418414067753e-05, + "loss": 0.0041, + "step": 32200 + }, + { + "grad_norm": 0.2010655701160431, + "learning_rate": 4.804664705298264e-05, + "loss": 0.0078, + "step": 32210 + }, + { + "grad_norm": 0.23369023203849792, + "learning_rate": 4.80191105586648e-05, + "loss": 0.0048, + "step": 32220 + }, + { + "grad_norm": 0.23071376979351044, + "learning_rate": 4.799157466608886e-05, + "loss": 0.0047, + "step": 32230 + }, + { + "grad_norm": 0.14918065071105957, + "learning_rate": 4.796403938361951e-05, + "loss": 0.0049, + "step": 32240 + }, + { + "grad_norm": 0.18900185823440552, + "learning_rate": 4.793650471962123e-05, + "loss": 0.0042, + "step": 32250 + }, + { + "grad_norm": 0.21083198487758636, + "learning_rate": 4.790897068245835e-05, + "loss": 0.0064, + "step": 32260 + }, + { + "grad_norm": 0.25141823291778564, + "learning_rate": 4.7881437280494954e-05, + "loss": 0.0061, + "step": 32270 + }, + { + "grad_norm": 0.21428143978118896, + "learning_rate": 4.7853904522094965e-05, + "loss": 0.006, + "step": 32280 + }, + { + "grad_norm": 0.1968785971403122, + "learning_rate": 4.782637241562215e-05, + "loss": 0.0061, + "step": 32290 + }, + { + "grad_norm": 0.15142029523849487, + "learning_rate": 4.779884096943997e-05, + "loss": 0.005, + "step": 32300 + }, + { + "grad_norm": 0.18293695151805878, + "learning_rate": 4.777131019191182e-05, + "loss": 0.0042, + "step": 32310 + }, + { + "grad_norm": 0.13914795219898224, + "learning_rate": 4.774378009140076e-05, + "loss": 0.0033, + "step": 32320 + }, + { + "grad_norm": 0.1742292195558548, + "learning_rate": 4.7716250676269735e-05, + "loss": 0.0049, + "step": 32330 + }, + { + "grad_norm": 0.18444855511188507, + "learning_rate": 4.7688721954881485e-05, + "loss": 0.0048, + "step": 32340 + }, + { + "grad_norm": 0.1803952008485794, + "learning_rate": 4.7661193935598446e-05, + "loss": 0.0044, + "step": 32350 + }, + { + "grad_norm": 0.15445967018604279, + "learning_rate": 4.763366662678296e-05, + "loss": 0.0052, + "step": 32360 + }, + { + "grad_norm": 0.18442274630069733, + "learning_rate": 4.7606140036797064e-05, + "loss": 0.0051, + "step": 32370 + }, + { + "grad_norm": 0.17230643332004547, + "learning_rate": 4.7578614174002614e-05, + "loss": 0.0042, + "step": 32380 + }, + { + "grad_norm": 0.20352445542812347, + "learning_rate": 4.755108904676125e-05, + "loss": 0.0072, + "step": 32390 + }, + { + "grad_norm": 0.26936498284339905, + "learning_rate": 4.752356466343436e-05, + "loss": 0.0053, + "step": 32400 + }, + { + "grad_norm": 0.2861011326313019, + "learning_rate": 4.7496041032383174e-05, + "loss": 0.007, + "step": 32410 + }, + { + "grad_norm": 0.190623477101326, + "learning_rate": 4.746851816196858e-05, + "loss": 0.0056, + "step": 32420 + }, + { + "grad_norm": 0.1682569533586502, + "learning_rate": 4.744099606055135e-05, + "loss": 0.0047, + "step": 32430 + }, + { + "grad_norm": 0.22984115779399872, + "learning_rate": 4.741347473649193e-05, + "loss": 0.0046, + "step": 32440 + }, + { + "grad_norm": 0.19816507399082184, + "learning_rate": 4.738595419815058e-05, + "loss": 0.005, + "step": 32450 + }, + { + "grad_norm": 0.15016452968120575, + "learning_rate": 4.7358434453887365e-05, + "loss": 0.0055, + "step": 32460 + }, + { + "grad_norm": 0.23969532549381256, + "learning_rate": 4.7330915512061976e-05, + "loss": 0.0064, + "step": 32470 + }, + { + "grad_norm": 0.17962460219860077, + "learning_rate": 4.730339738103402e-05, + "loss": 0.0059, + "step": 32480 + }, + { + "grad_norm": 0.16040557622909546, + "learning_rate": 4.727588006916271e-05, + "loss": 0.0045, + "step": 32490 + }, + { + "grad_norm": 0.20202873647212982, + "learning_rate": 4.724836358480711e-05, + "loss": 0.0039, + "step": 32500 + }, + { + "grad_norm": 0.12208746373653412, + "learning_rate": 4.722084793632601e-05, + "loss": 0.004, + "step": 32510 + }, + { + "grad_norm": 0.15922726690769196, + "learning_rate": 4.719333313207792e-05, + "loss": 0.0049, + "step": 32520 + }, + { + "grad_norm": 0.1514817327260971, + "learning_rate": 4.716581918042114e-05, + "loss": 0.0043, + "step": 32530 + }, + { + "grad_norm": 0.26068586111068726, + "learning_rate": 4.7138306089713636e-05, + "loss": 0.0086, + "step": 32540 + }, + { + "grad_norm": 0.19129633903503418, + "learning_rate": 4.7110793868313183e-05, + "loss": 0.0048, + "step": 32550 + }, + { + "grad_norm": 0.2191700041294098, + "learning_rate": 4.708328252457729e-05, + "loss": 0.0043, + "step": 32560 + }, + { + "grad_norm": 0.24066494405269623, + "learning_rate": 4.7055772066863135e-05, + "loss": 0.0048, + "step": 32570 + }, + { + "grad_norm": 0.24875414371490479, + "learning_rate": 4.702826250352771e-05, + "loss": 0.0062, + "step": 32580 + }, + { + "grad_norm": 0.24485665559768677, + "learning_rate": 4.7000753842927653e-05, + "loss": 0.007, + "step": 32590 + }, + { + "grad_norm": 0.18917761743068695, + "learning_rate": 4.6973246093419384e-05, + "loss": 0.0068, + "step": 32600 + }, + { + "grad_norm": 0.22483518719673157, + "learning_rate": 4.694573926335906e-05, + "loss": 0.0036, + "step": 32610 + }, + { + "grad_norm": 0.1390780657529831, + "learning_rate": 4.6918233361102476e-05, + "loss": 0.0032, + "step": 32620 + }, + { + "grad_norm": 0.1989414244890213, + "learning_rate": 4.689072839500525e-05, + "loss": 0.0054, + "step": 32630 + }, + { + "grad_norm": 0.18637628853321075, + "learning_rate": 4.6863224373422635e-05, + "loss": 0.004, + "step": 32640 + }, + { + "grad_norm": 0.171858012676239, + "learning_rate": 4.683572130470962e-05, + "loss": 0.0033, + "step": 32650 + }, + { + "grad_norm": 0.16169320046901703, + "learning_rate": 4.680821919722094e-05, + "loss": 0.005, + "step": 32660 + }, + { + "grad_norm": 0.14494270086288452, + "learning_rate": 4.6780718059310975e-05, + "loss": 0.0044, + "step": 32670 + }, + { + "grad_norm": 0.14536269009113312, + "learning_rate": 4.675321789933389e-05, + "loss": 0.0067, + "step": 32680 + }, + { + "grad_norm": 0.199925497174263, + "learning_rate": 4.6725718725643464e-05, + "loss": 0.0049, + "step": 32690 + }, + { + "grad_norm": 0.19077004492282867, + "learning_rate": 4.669822054659323e-05, + "loss": 0.0049, + "step": 32700 + }, + { + "grad_norm": 0.19654829800128937, + "learning_rate": 4.667072337053644e-05, + "loss": 0.0038, + "step": 32710 + }, + { + "grad_norm": 0.20727397501468658, + "learning_rate": 4.6643227205825965e-05, + "loss": 0.0043, + "step": 32720 + }, + { + "grad_norm": 0.21547819674015045, + "learning_rate": 4.6615732060814454e-05, + "loss": 0.0059, + "step": 32730 + }, + { + "grad_norm": 0.1741621047258377, + "learning_rate": 4.658823794385417e-05, + "loss": 0.0034, + "step": 32740 + }, + { + "grad_norm": 0.16072964668273926, + "learning_rate": 4.6560744863297115e-05, + "loss": 0.0053, + "step": 32750 + }, + { + "grad_norm": 0.1594333052635193, + "learning_rate": 4.653325282749498e-05, + "loss": 0.0044, + "step": 32760 + }, + { + "grad_norm": 0.2023230791091919, + "learning_rate": 4.6505761844799075e-05, + "loss": 0.0053, + "step": 32770 + }, + { + "grad_norm": 0.1880958080291748, + "learning_rate": 4.647827192356048e-05, + "loss": 0.005, + "step": 32780 + }, + { + "grad_norm": 0.19728344678878784, + "learning_rate": 4.645078307212989e-05, + "loss": 0.0039, + "step": 32790 + }, + { + "grad_norm": 0.1750764399766922, + "learning_rate": 4.642329529885768e-05, + "loss": 0.0045, + "step": 32800 + }, + { + "grad_norm": 0.14193753898143768, + "learning_rate": 4.639580861209393e-05, + "loss": 0.0038, + "step": 32810 + }, + { + "grad_norm": 0.14580261707305908, + "learning_rate": 4.636832302018835e-05, + "loss": 0.0039, + "step": 32820 + }, + { + "grad_norm": 0.22704246640205383, + "learning_rate": 4.6340838531490365e-05, + "loss": 0.0066, + "step": 32830 + }, + { + "grad_norm": 0.16191266477108002, + "learning_rate": 4.6313355154349e-05, + "loss": 0.004, + "step": 32840 + }, + { + "grad_norm": 0.15272437036037445, + "learning_rate": 4.6285872897113025e-05, + "loss": 0.0042, + "step": 32850 + }, + { + "grad_norm": 0.18530550599098206, + "learning_rate": 4.625839176813077e-05, + "loss": 0.004, + "step": 32860 + }, + { + "grad_norm": 0.14753226935863495, + "learning_rate": 4.623091177575031e-05, + "loss": 0.0045, + "step": 32870 + }, + { + "grad_norm": 0.15381473302841187, + "learning_rate": 4.620343292831936e-05, + "loss": 0.0064, + "step": 32880 + }, + { + "grad_norm": 0.1351233720779419, + "learning_rate": 4.6175955234185206e-05, + "loss": 0.0042, + "step": 32890 + }, + { + "grad_norm": 0.17321446537971497, + "learning_rate": 4.614847870169492e-05, + "loss": 0.0044, + "step": 32900 + }, + { + "grad_norm": 0.17199234664440155, + "learning_rate": 4.612100333919509e-05, + "loss": 0.0032, + "step": 32910 + }, + { + "grad_norm": 0.18708087503910065, + "learning_rate": 4.609352915503202e-05, + "loss": 0.0046, + "step": 32920 + }, + { + "grad_norm": 0.19583621621131897, + "learning_rate": 4.606605615755166e-05, + "loss": 0.0044, + "step": 32930 + }, + { + "grad_norm": 0.25135812163352966, + "learning_rate": 4.6038584355099576e-05, + "loss": 0.0062, + "step": 32940 + }, + { + "grad_norm": 0.2475704401731491, + "learning_rate": 4.6011113756020964e-05, + "loss": 0.0053, + "step": 32950 + }, + { + "grad_norm": 0.13889190554618835, + "learning_rate": 4.598364436866066e-05, + "loss": 0.0041, + "step": 32960 + }, + { + "grad_norm": 0.19973912835121155, + "learning_rate": 4.595617620136316e-05, + "loss": 0.0036, + "step": 32970 + }, + { + "grad_norm": 0.18490223586559296, + "learning_rate": 4.592870926247257e-05, + "loss": 0.0048, + "step": 32980 + }, + { + "grad_norm": 0.1921015828847885, + "learning_rate": 4.5901243560332594e-05, + "loss": 0.0052, + "step": 32990 + }, + { + "grad_norm": 0.23081588745117188, + "learning_rate": 4.587377910328662e-05, + "loss": 0.0053, + "step": 33000 + }, + { + "grad_norm": 0.16712728142738342, + "learning_rate": 4.5846315899677586e-05, + "loss": 0.0045, + "step": 33010 + }, + { + "grad_norm": 0.18717841804027557, + "learning_rate": 4.5818853957848114e-05, + "loss": 0.0045, + "step": 33020 + }, + { + "grad_norm": 0.20282988250255585, + "learning_rate": 4.579139328614043e-05, + "loss": 0.0064, + "step": 33030 + }, + { + "grad_norm": 0.22016045451164246, + "learning_rate": 4.576393389289633e-05, + "loss": 0.0059, + "step": 33040 + }, + { + "grad_norm": 0.17904025316238403, + "learning_rate": 4.573647578645728e-05, + "loss": 0.006, + "step": 33050 + }, + { + "grad_norm": 0.19048282504081726, + "learning_rate": 4.57090189751643e-05, + "loss": 0.0052, + "step": 33060 + }, + { + "grad_norm": 0.1914157122373581, + "learning_rate": 4.568156346735806e-05, + "loss": 0.0045, + "step": 33070 + }, + { + "grad_norm": 0.20221026241779327, + "learning_rate": 4.565410927137882e-05, + "loss": 0.0065, + "step": 33080 + }, + { + "grad_norm": 0.21152468025684357, + "learning_rate": 4.562665639556644e-05, + "loss": 0.0066, + "step": 33090 + }, + { + "grad_norm": 0.17515084147453308, + "learning_rate": 4.559920484826037e-05, + "loss": 0.0055, + "step": 33100 + }, + { + "grad_norm": 0.13399997353553772, + "learning_rate": 4.5571754637799665e-05, + "loss": 0.0044, + "step": 33110 + }, + { + "grad_norm": 0.15194153785705566, + "learning_rate": 4.554430577252298e-05, + "loss": 0.0043, + "step": 33120 + }, + { + "grad_norm": 0.16189908981323242, + "learning_rate": 4.551685826076858e-05, + "loss": 0.0046, + "step": 33130 + }, + { + "grad_norm": 0.22422078251838684, + "learning_rate": 4.5489412110874246e-05, + "loss": 0.007, + "step": 33140 + }, + { + "grad_norm": 0.2511410415172577, + "learning_rate": 4.5461967331177444e-05, + "loss": 0.0053, + "step": 33150 + }, + { + "grad_norm": 0.21357578039169312, + "learning_rate": 4.5434523930015115e-05, + "loss": 0.0048, + "step": 33160 + }, + { + "grad_norm": 0.19267761707305908, + "learning_rate": 4.540708191572388e-05, + "loss": 0.0062, + "step": 33170 + }, + { + "grad_norm": 0.208268940448761, + "learning_rate": 4.537964129663991e-05, + "loss": 0.0058, + "step": 33180 + }, + { + "grad_norm": 0.22184109687805176, + "learning_rate": 4.535220208109889e-05, + "loss": 0.0041, + "step": 33190 + }, + { + "grad_norm": 0.2716061472892761, + "learning_rate": 4.5324764277436194e-05, + "loss": 0.0042, + "step": 33200 + }, + { + "grad_norm": 0.23763802647590637, + "learning_rate": 4.529732789398664e-05, + "loss": 0.0069, + "step": 33210 + }, + { + "grad_norm": 0.19782312214374542, + "learning_rate": 4.526989293908472e-05, + "loss": 0.0049, + "step": 33220 + }, + { + "grad_norm": 0.18328730762004852, + "learning_rate": 4.524245942106442e-05, + "loss": 0.005, + "step": 33230 + }, + { + "grad_norm": 0.18177179992198944, + "learning_rate": 4.5215027348259345e-05, + "loss": 0.0049, + "step": 33240 + }, + { + "grad_norm": 0.15582804381847382, + "learning_rate": 4.5187596729002616e-05, + "loss": 0.0046, + "step": 33250 + }, + { + "grad_norm": 0.1172068864107132, + "learning_rate": 4.516016757162693e-05, + "loss": 0.0072, + "step": 33260 + }, + { + "grad_norm": 0.16688767075538635, + "learning_rate": 4.513273988446457e-05, + "loss": 0.0051, + "step": 33270 + }, + { + "grad_norm": 0.1660337597131729, + "learning_rate": 4.5105313675847296e-05, + "loss": 0.0051, + "step": 33280 + }, + { + "grad_norm": 0.15883983671665192, + "learning_rate": 4.5077888954106495e-05, + "loss": 0.0047, + "step": 33290 + }, + { + "grad_norm": 0.19077563285827637, + "learning_rate": 4.505046572757309e-05, + "loss": 0.0044, + "step": 33300 + }, + { + "grad_norm": 0.18063879013061523, + "learning_rate": 4.502304400457749e-05, + "loss": 0.0048, + "step": 33310 + }, + { + "grad_norm": 0.1593950390815735, + "learning_rate": 4.499562379344973e-05, + "loss": 0.0038, + "step": 33320 + }, + { + "grad_norm": 0.20816084742546082, + "learning_rate": 4.4968205102519306e-05, + "loss": 0.0079, + "step": 33330 + }, + { + "grad_norm": 0.17260810732841492, + "learning_rate": 4.494078794011532e-05, + "loss": 0.0049, + "step": 33340 + }, + { + "grad_norm": 0.1723584085702896, + "learning_rate": 4.491337231456639e-05, + "loss": 0.0039, + "step": 33350 + }, + { + "grad_norm": 0.15594543516635895, + "learning_rate": 4.4885958234200634e-05, + "loss": 0.0045, + "step": 33360 + }, + { + "grad_norm": 0.18546228110790253, + "learning_rate": 4.485854570734575e-05, + "loss": 0.0044, + "step": 33370 + }, + { + "grad_norm": 0.14639759063720703, + "learning_rate": 4.483113474232891e-05, + "loss": 0.0054, + "step": 33380 + }, + { + "grad_norm": 0.21834929287433624, + "learning_rate": 4.480372534747688e-05, + "loss": 0.0048, + "step": 33390 + }, + { + "grad_norm": 0.2139056921005249, + "learning_rate": 4.477631753111588e-05, + "loss": 0.0051, + "step": 33400 + }, + { + "grad_norm": 0.22326582670211792, + "learning_rate": 4.4748911301571686e-05, + "loss": 0.0049, + "step": 33410 + }, + { + "grad_norm": 0.15286721289157867, + "learning_rate": 4.472150666716961e-05, + "loss": 0.0038, + "step": 33420 + }, + { + "grad_norm": 0.1834818422794342, + "learning_rate": 4.469410363623442e-05, + "loss": 0.0046, + "step": 33430 + }, + { + "grad_norm": 0.1544836163520813, + "learning_rate": 4.466670221709044e-05, + "loss": 0.0082, + "step": 33440 + }, + { + "grad_norm": 0.18582512438297272, + "learning_rate": 4.463930241806154e-05, + "loss": 0.0045, + "step": 33450 + }, + { + "grad_norm": 0.1381050944328308, + "learning_rate": 4.4611904247471006e-05, + "loss": 0.0038, + "step": 33460 + }, + { + "grad_norm": 0.19770090281963348, + "learning_rate": 4.458450771364171e-05, + "loss": 0.0057, + "step": 33470 + }, + { + "grad_norm": 0.20572297275066376, + "learning_rate": 4.4557112824895965e-05, + "loss": 0.0062, + "step": 33480 + }, + { + "grad_norm": 0.15222179889678955, + "learning_rate": 4.452971958955563e-05, + "loss": 0.0041, + "step": 33490 + }, + { + "grad_norm": 0.18206976354122162, + "learning_rate": 4.450232801594208e-05, + "loss": 0.0038, + "step": 33500 + }, + { + "grad_norm": 0.22142469882965088, + "learning_rate": 4.447493811237609e-05, + "loss": 0.007, + "step": 33510 + }, + { + "grad_norm": 0.18259239196777344, + "learning_rate": 4.444754988717804e-05, + "loss": 0.0037, + "step": 33520 + }, + { + "grad_norm": 0.1644694209098816, + "learning_rate": 4.442016334866771e-05, + "loss": 0.0057, + "step": 33530 + }, + { + "grad_norm": 0.12861062586307526, + "learning_rate": 4.4392778505164445e-05, + "loss": 0.0039, + "step": 33540 + }, + { + "grad_norm": 0.130414217710495, + "learning_rate": 4.436539536498702e-05, + "loss": 0.0047, + "step": 33550 + }, + { + "grad_norm": 0.1650528758764267, + "learning_rate": 4.433801393645369e-05, + "loss": 0.0046, + "step": 33560 + }, + { + "grad_norm": 0.1523527055978775, + "learning_rate": 4.431063422788226e-05, + "loss": 0.0041, + "step": 33570 + }, + { + "grad_norm": 0.17007499933242798, + "learning_rate": 4.428325624758991e-05, + "loss": 0.0049, + "step": 33580 + }, + { + "grad_norm": 0.20401349663734436, + "learning_rate": 4.4255880003893366e-05, + "loss": 0.0044, + "step": 33590 + }, + { + "grad_norm": 0.14803670346736908, + "learning_rate": 4.422850550510884e-05, + "loss": 0.0039, + "step": 33600 + }, + { + "grad_norm": 0.20467841625213623, + "learning_rate": 4.4201132759551934e-05, + "loss": 0.0061, + "step": 33610 + }, + { + "grad_norm": 0.18328902125358582, + "learning_rate": 4.4173761775537804e-05, + "loss": 0.0042, + "step": 33620 + }, + { + "grad_norm": 0.17253699898719788, + "learning_rate": 4.414639256138099e-05, + "loss": 0.0053, + "step": 33630 + }, + { + "grad_norm": 0.20801103115081787, + "learning_rate": 4.411902512539557e-05, + "loss": 0.0046, + "step": 33640 + }, + { + "grad_norm": 0.1973229944705963, + "learning_rate": 4.4091659475895044e-05, + "loss": 0.0059, + "step": 33650 + }, + { + "grad_norm": 0.16329631209373474, + "learning_rate": 4.406429562119235e-05, + "loss": 0.0044, + "step": 33660 + }, + { + "grad_norm": 0.18277055025100708, + "learning_rate": 4.4036933569599945e-05, + "loss": 0.0042, + "step": 33670 + }, + { + "grad_norm": 0.16051676869392395, + "learning_rate": 4.400957332942965e-05, + "loss": 0.0038, + "step": 33680 + }, + { + "grad_norm": 0.16573944687843323, + "learning_rate": 4.3982214908992844e-05, + "loss": 0.0057, + "step": 33690 + }, + { + "grad_norm": 0.19477687776088715, + "learning_rate": 4.3954858316600235e-05, + "loss": 0.0035, + "step": 33700 + }, + { + "grad_norm": 0.21553660929203033, + "learning_rate": 4.392750356056205e-05, + "loss": 0.0049, + "step": 33710 + }, + { + "grad_norm": 0.21639776229858398, + "learning_rate": 4.390015064918798e-05, + "loss": 0.0051, + "step": 33720 + }, + { + "grad_norm": 0.1857653558254242, + "learning_rate": 4.387279959078705e-05, + "loss": 0.0044, + "step": 33730 + }, + { + "grad_norm": 0.16132286190986633, + "learning_rate": 4.384545039366786e-05, + "loss": 0.0035, + "step": 33740 + }, + { + "grad_norm": 0.1379638910293579, + "learning_rate": 4.381810306613831e-05, + "loss": 0.0047, + "step": 33750 + }, + { + "grad_norm": 0.13853634893894196, + "learning_rate": 4.3790757616505826e-05, + "loss": 0.004, + "step": 33760 + }, + { + "grad_norm": 0.13563930988311768, + "learning_rate": 4.376341405307725e-05, + "loss": 0.0041, + "step": 33770 + }, + { + "grad_norm": 0.21742835640907288, + "learning_rate": 4.37360723841588e-05, + "loss": 0.004, + "step": 33780 + }, + { + "grad_norm": 0.1971443146467209, + "learning_rate": 4.370873261805619e-05, + "loss": 0.005, + "step": 33790 + }, + { + "grad_norm": 0.2042785882949829, + "learning_rate": 4.368139476307449e-05, + "loss": 0.0053, + "step": 33800 + }, + { + "grad_norm": 0.22481749951839447, + "learning_rate": 4.365405882751822e-05, + "loss": 0.005, + "step": 33810 + }, + { + "grad_norm": 0.18183766305446625, + "learning_rate": 4.3626724819691326e-05, + "loss": 0.0038, + "step": 33820 + }, + { + "grad_norm": 0.26173731684684753, + "learning_rate": 4.359939274789715e-05, + "loss": 0.0053, + "step": 33830 + }, + { + "grad_norm": 0.14960384368896484, + "learning_rate": 4.357206262043848e-05, + "loss": 0.0042, + "step": 33840 + }, + { + "grad_norm": 0.23865333199501038, + "learning_rate": 4.354473444561745e-05, + "loss": 0.0053, + "step": 33850 + }, + { + "grad_norm": 0.13766765594482422, + "learning_rate": 4.3517408231735644e-05, + "loss": 0.0035, + "step": 33860 + }, + { + "grad_norm": 0.1455104500055313, + "learning_rate": 4.3490083987094086e-05, + "loss": 0.0044, + "step": 33870 + }, + { + "grad_norm": 0.2013978511095047, + "learning_rate": 4.34627617199931e-05, + "loss": 0.0047, + "step": 33880 + }, + { + "grad_norm": 0.22904083132743835, + "learning_rate": 4.3435441438732526e-05, + "loss": 0.0048, + "step": 33890 + }, + { + "grad_norm": 0.15855951607227325, + "learning_rate": 4.340812315161149e-05, + "loss": 0.0058, + "step": 33900 + }, + { + "grad_norm": 0.14015594124794006, + "learning_rate": 4.338080686692859e-05, + "loss": 0.0035, + "step": 33910 + }, + { + "grad_norm": 0.14604465663433075, + "learning_rate": 4.3353492592981816e-05, + "loss": 0.0045, + "step": 33920 + }, + { + "grad_norm": 0.1883566826581955, + "learning_rate": 4.3326180338068485e-05, + "loss": 0.0061, + "step": 33930 + }, + { + "grad_norm": 0.12513059377670288, + "learning_rate": 4.3298870110485356e-05, + "loss": 0.004, + "step": 33940 + }, + { + "grad_norm": 0.17817538976669312, + "learning_rate": 4.3271561918528567e-05, + "loss": 0.0037, + "step": 33950 + }, + { + "grad_norm": 0.15345628559589386, + "learning_rate": 4.324425577049359e-05, + "loss": 0.0047, + "step": 33960 + }, + { + "grad_norm": 0.1819748431444168, + "learning_rate": 4.321695167467535e-05, + "loss": 0.0051, + "step": 33970 + }, + { + "grad_norm": 0.20217286050319672, + "learning_rate": 4.3189649639368093e-05, + "loss": 0.0061, + "step": 33980 + }, + { + "grad_norm": 0.18917502462863922, + "learning_rate": 4.316234967286547e-05, + "loss": 0.0045, + "step": 33990 + }, + { + "grad_norm": 0.19039934873580933, + "learning_rate": 4.313505178346046e-05, + "loss": 0.0058, + "step": 34000 + }, + { + "grad_norm": 0.18818224966526031, + "learning_rate": 4.3107755979445465e-05, + "loss": 0.0039, + "step": 34010 + }, + { + "grad_norm": 0.18497240543365479, + "learning_rate": 4.308046226911224e-05, + "loss": 0.0049, + "step": 34020 + }, + { + "grad_norm": 0.19601446390151978, + "learning_rate": 4.305317066075185e-05, + "loss": 0.007, + "step": 34030 + }, + { + "grad_norm": 0.19892337918281555, + "learning_rate": 4.302588116265482e-05, + "loss": 0.005, + "step": 34040 + }, + { + "grad_norm": 0.19490763545036316, + "learning_rate": 4.299859378311094e-05, + "loss": 0.0042, + "step": 34050 + }, + { + "grad_norm": 0.14752547442913055, + "learning_rate": 4.2971308530409424e-05, + "loss": 0.0069, + "step": 34060 + }, + { + "grad_norm": 0.18503214418888092, + "learning_rate": 4.2944025412838765e-05, + "loss": 0.0046, + "step": 34070 + }, + { + "grad_norm": 0.1781253069639206, + "learning_rate": 4.291674443868689e-05, + "loss": 0.0044, + "step": 34080 + }, + { + "grad_norm": 0.21360643208026886, + "learning_rate": 4.288946561624104e-05, + "loss": 0.0051, + "step": 34090 + }, + { + "grad_norm": 0.20713235437870026, + "learning_rate": 4.2862188953787794e-05, + "loss": 0.0039, + "step": 34100 + }, + { + "grad_norm": 0.2380281239748001, + "learning_rate": 4.283491445961308e-05, + "loss": 0.0053, + "step": 34110 + }, + { + "grad_norm": 0.1647782176733017, + "learning_rate": 4.2807642142002155e-05, + "loss": 0.0058, + "step": 34120 + }, + { + "grad_norm": 0.20737305283546448, + "learning_rate": 4.278037200923966e-05, + "loss": 0.0054, + "step": 34130 + }, + { + "grad_norm": 0.23220381140708923, + "learning_rate": 4.275310406960953e-05, + "loss": 0.0035, + "step": 34140 + }, + { + "grad_norm": 0.18818068504333496, + "learning_rate": 4.272583833139502e-05, + "loss": 0.0053, + "step": 34150 + }, + { + "grad_norm": 0.17668244242668152, + "learning_rate": 4.2698574802878794e-05, + "loss": 0.004, + "step": 34160 + }, + { + "grad_norm": 0.15770000219345093, + "learning_rate": 4.2671313492342734e-05, + "loss": 0.004, + "step": 34170 + }, + { + "grad_norm": 0.12426867336034775, + "learning_rate": 4.264405440806813e-05, + "loss": 0.0051, + "step": 34180 + }, + { + "grad_norm": 0.20602105557918549, + "learning_rate": 4.26167975583356e-05, + "loss": 0.0039, + "step": 34190 + }, + { + "grad_norm": 0.18737304210662842, + "learning_rate": 4.2589542951425e-05, + "loss": 0.0036, + "step": 34200 + }, + { + "grad_norm": 0.15843060612678528, + "learning_rate": 4.2562290595615615e-05, + "loss": 0.0032, + "step": 34210 + }, + { + "grad_norm": 0.14224781095981598, + "learning_rate": 4.2535040499185946e-05, + "loss": 0.0036, + "step": 34220 + }, + { + "grad_norm": 0.18161417543888092, + "learning_rate": 4.250779267041387e-05, + "loss": 0.0044, + "step": 34230 + }, + { + "grad_norm": 0.16561119258403778, + "learning_rate": 4.248054711757657e-05, + "loss": 0.0055, + "step": 34240 + }, + { + "grad_norm": 0.135898157954216, + "learning_rate": 4.245330384895052e-05, + "loss": 0.0044, + "step": 34250 + }, + { + "grad_norm": 0.1834179162979126, + "learning_rate": 4.242606287281151e-05, + "loss": 0.0049, + "step": 34260 + }, + { + "grad_norm": 0.18382470309734344, + "learning_rate": 4.2398824197434595e-05, + "loss": 0.0044, + "step": 34270 + }, + { + "grad_norm": 0.21192659437656403, + "learning_rate": 4.23715878310942e-05, + "loss": 0.0057, + "step": 34280 + }, + { + "grad_norm": 0.16714639961719513, + "learning_rate": 4.234435378206402e-05, + "loss": 0.0042, + "step": 34290 + }, + { + "grad_norm": 0.16182488203048706, + "learning_rate": 4.2317122058617006e-05, + "loss": 0.0034, + "step": 34300 + }, + { + "grad_norm": 0.16742929816246033, + "learning_rate": 4.2289892669025485e-05, + "loss": 0.0034, + "step": 34310 + }, + { + "grad_norm": 0.18080438673496246, + "learning_rate": 4.226266562156097e-05, + "loss": 0.0056, + "step": 34320 + }, + { + "grad_norm": 0.17973585426807404, + "learning_rate": 4.223544092449435e-05, + "loss": 0.0046, + "step": 34330 + }, + { + "grad_norm": 0.1538865566253662, + "learning_rate": 4.2208218586095784e-05, + "loss": 0.0034, + "step": 34340 + }, + { + "grad_norm": 0.1607210338115692, + "learning_rate": 4.218099861463466e-05, + "loss": 0.0046, + "step": 34350 + }, + { + "grad_norm": 0.17730094492435455, + "learning_rate": 4.215378101837972e-05, + "loss": 0.0047, + "step": 34360 + }, + { + "grad_norm": 0.1639341413974762, + "learning_rate": 4.2126565805598937e-05, + "loss": 0.0039, + "step": 34370 + }, + { + "grad_norm": 0.19921983778476715, + "learning_rate": 4.209935298455957e-05, + "loss": 0.0046, + "step": 34380 + }, + { + "grad_norm": 0.19436313211917877, + "learning_rate": 4.207214256352817e-05, + "loss": 0.0075, + "step": 34390 + }, + { + "grad_norm": 0.191510409116745, + "learning_rate": 4.2044934550770524e-05, + "loss": 0.0038, + "step": 34400 + }, + { + "grad_norm": 0.1624172478914261, + "learning_rate": 4.201772895455174e-05, + "loss": 0.0038, + "step": 34410 + }, + { + "grad_norm": 0.15403953194618225, + "learning_rate": 4.199052578313613e-05, + "loss": 0.0036, + "step": 34420 + }, + { + "grad_norm": 0.15793345868587494, + "learning_rate": 4.1963325044787294e-05, + "loss": 0.0042, + "step": 34430 + }, + { + "grad_norm": 0.13288035988807678, + "learning_rate": 4.193612674776814e-05, + "loss": 0.0036, + "step": 34440 + }, + { + "grad_norm": 0.20503857731819153, + "learning_rate": 4.1908930900340745e-05, + "loss": 0.0045, + "step": 34450 + }, + { + "grad_norm": 0.1907370239496231, + "learning_rate": 4.1881737510766536e-05, + "loss": 0.0041, + "step": 34460 + }, + { + "grad_norm": 0.2307111769914627, + "learning_rate": 4.185454658730609e-05, + "loss": 0.0038, + "step": 34470 + }, + { + "grad_norm": 0.22476862370967865, + "learning_rate": 4.1827358138219355e-05, + "loss": 0.0047, + "step": 34480 + }, + { + "grad_norm": 0.22284075617790222, + "learning_rate": 4.1800172171765404e-05, + "loss": 0.0038, + "step": 34490 + }, + { + "grad_norm": 0.14311498403549194, + "learning_rate": 4.177298869620264e-05, + "loss": 0.0041, + "step": 34500 + }, + { + "grad_norm": 0.1221272423863411, + "learning_rate": 4.1745807719788705e-05, + "loss": 0.0045, + "step": 34510 + }, + { + "grad_norm": 0.16835187375545502, + "learning_rate": 4.1718629250780445e-05, + "loss": 0.0052, + "step": 34520 + }, + { + "grad_norm": 0.23141171038150787, + "learning_rate": 4.1691453297433956e-05, + "loss": 0.0048, + "step": 34530 + }, + { + "grad_norm": 0.13610680401325226, + "learning_rate": 4.166427986800457e-05, + "loss": 0.0041, + "step": 34540 + }, + { + "grad_norm": 0.16812001168727875, + "learning_rate": 4.163710897074688e-05, + "loss": 0.0044, + "step": 34550 + }, + { + "grad_norm": 0.26125040650367737, + "learning_rate": 4.1609940613914686e-05, + "loss": 0.0041, + "step": 34560 + }, + { + "grad_norm": 0.23603585362434387, + "learning_rate": 4.1582774805760996e-05, + "loss": 0.0047, + "step": 34570 + }, + { + "grad_norm": 0.18174543976783752, + "learning_rate": 4.155561155453809e-05, + "loss": 0.0062, + "step": 34580 + }, + { + "grad_norm": 0.1725202053785324, + "learning_rate": 4.15284508684974e-05, + "loss": 0.004, + "step": 34590 + }, + { + "grad_norm": 0.22396813333034515, + "learning_rate": 4.1501292755889675e-05, + "loss": 0.0037, + "step": 34600 + }, + { + "grad_norm": 0.14621104300022125, + "learning_rate": 4.1474137224964833e-05, + "loss": 0.0052, + "step": 34610 + }, + { + "grad_norm": 0.21945036947727203, + "learning_rate": 4.144698428397197e-05, + "loss": 0.0045, + "step": 34620 + }, + { + "grad_norm": 0.19067789614200592, + "learning_rate": 4.1419833941159466e-05, + "loss": 0.0061, + "step": 34630 + }, + { + "grad_norm": 0.17680232226848602, + "learning_rate": 4.1392686204774846e-05, + "loss": 0.0058, + "step": 34640 + }, + { + "grad_norm": 0.17204707860946655, + "learning_rate": 4.13655410830649e-05, + "loss": 0.0073, + "step": 34650 + }, + { + "grad_norm": 0.16198855638504028, + "learning_rate": 4.1338398584275594e-05, + "loss": 0.004, + "step": 34660 + }, + { + "grad_norm": 0.1765734851360321, + "learning_rate": 4.1311258716652104e-05, + "loss": 0.004, + "step": 34670 + }, + { + "grad_norm": 0.18409477174282074, + "learning_rate": 4.128412148843881e-05, + "loss": 0.0056, + "step": 34680 + }, + { + "grad_norm": 0.16727179288864136, + "learning_rate": 4.125698690787926e-05, + "loss": 0.0045, + "step": 34690 + }, + { + "grad_norm": 0.20261996984481812, + "learning_rate": 4.1229854983216245e-05, + "loss": 0.0055, + "step": 34700 + }, + { + "grad_norm": 0.15564371645450592, + "learning_rate": 4.120272572269175e-05, + "loss": 0.005, + "step": 34710 + }, + { + "grad_norm": 0.21736681461334229, + "learning_rate": 4.117559913454687e-05, + "loss": 0.0044, + "step": 34720 + }, + { + "grad_norm": 0.1425514668226242, + "learning_rate": 4.114847522702201e-05, + "loss": 0.0038, + "step": 34730 + }, + { + "grad_norm": 0.17848694324493408, + "learning_rate": 4.112135400835664e-05, + "loss": 0.0058, + "step": 34740 + }, + { + "grad_norm": 0.14816541969776154, + "learning_rate": 4.109423548678949e-05, + "loss": 0.0049, + "step": 34750 + }, + { + "grad_norm": 0.16022250056266785, + "learning_rate": 4.106711967055848e-05, + "loss": 0.0038, + "step": 34760 + }, + { + "grad_norm": 0.20128020644187927, + "learning_rate": 4.1040006567900636e-05, + "loss": 0.004, + "step": 34770 + }, + { + "grad_norm": 0.1692381352186203, + "learning_rate": 4.101289618705224e-05, + "loss": 0.0034, + "step": 34780 + }, + { + "grad_norm": 0.1550602912902832, + "learning_rate": 4.0985788536248675e-05, + "loss": 0.0037, + "step": 34790 + }, + { + "grad_norm": 0.15244947373867035, + "learning_rate": 4.095868362372454e-05, + "loss": 0.0038, + "step": 34800 + }, + { + "grad_norm": 0.1967269331216812, + "learning_rate": 4.0931581457713614e-05, + "loss": 0.0046, + "step": 34810 + }, + { + "grad_norm": 0.17493416368961334, + "learning_rate": 4.09044820464488e-05, + "loss": 0.0037, + "step": 34820 + }, + { + "grad_norm": 0.17064078152179718, + "learning_rate": 4.087738539816219e-05, + "loss": 0.0051, + "step": 34830 + }, + { + "grad_norm": 0.1470010131597519, + "learning_rate": 4.085029152108501e-05, + "loss": 0.0039, + "step": 34840 + }, + { + "grad_norm": 0.19303803145885468, + "learning_rate": 4.0823200423447714e-05, + "loss": 0.0039, + "step": 34850 + }, + { + "grad_norm": 0.21545466780662537, + "learning_rate": 4.079611211347981e-05, + "loss": 0.0038, + "step": 34860 + }, + { + "grad_norm": 0.1646214872598648, + "learning_rate": 4.076902659941002e-05, + "loss": 0.0059, + "step": 34870 + }, + { + "grad_norm": 0.1329984962940216, + "learning_rate": 4.074194388946624e-05, + "loss": 0.0042, + "step": 34880 + }, + { + "grad_norm": 0.19727373123168945, + "learning_rate": 4.071486399187545e-05, + "loss": 0.0049, + "step": 34890 + }, + { + "grad_norm": 0.16062751412391663, + "learning_rate": 4.0687786914863836e-05, + "loss": 0.0038, + "step": 34900 + }, + { + "grad_norm": 0.17265285551548004, + "learning_rate": 4.0660712666656666e-05, + "loss": 0.0033, + "step": 34910 + }, + { + "grad_norm": 0.16694913804531097, + "learning_rate": 4.0633641255478394e-05, + "loss": 0.0034, + "step": 34920 + }, + { + "grad_norm": 0.19818897545337677, + "learning_rate": 4.0606572689552624e-05, + "loss": 0.004, + "step": 34930 + }, + { + "grad_norm": 0.18170298635959625, + "learning_rate": 4.0579506977102036e-05, + "loss": 0.0056, + "step": 34940 + }, + { + "grad_norm": 0.16174526512622833, + "learning_rate": 4.055244412634849e-05, + "loss": 0.0034, + "step": 34950 + }, + { + "grad_norm": 0.1887652575969696, + "learning_rate": 4.052538414551298e-05, + "loss": 0.0038, + "step": 34960 + }, + { + "grad_norm": 0.1330665647983551, + "learning_rate": 4.0498327042815596e-05, + "loss": 0.0034, + "step": 34970 + }, + { + "grad_norm": 0.16771569848060608, + "learning_rate": 4.047127282647559e-05, + "loss": 0.0059, + "step": 34980 + }, + { + "grad_norm": 0.14479343593120575, + "learning_rate": 4.04442215047113e-05, + "loss": 0.0048, + "step": 34990 + }, + { + "grad_norm": 0.18350155651569366, + "learning_rate": 4.041717308574023e-05, + "loss": 0.0041, + "step": 35000 + }, + { + "grad_norm": 0.19537577033042908, + "learning_rate": 4.039012757777893e-05, + "loss": 0.0088, + "step": 35010 + }, + { + "grad_norm": 0.16308967769145966, + "learning_rate": 4.036308498904314e-05, + "loss": 0.004, + "step": 35020 + }, + { + "grad_norm": 0.21453475952148438, + "learning_rate": 4.033604532774771e-05, + "loss": 0.005, + "step": 35030 + }, + { + "grad_norm": 0.1916845738887787, + "learning_rate": 4.030900860210652e-05, + "loss": 0.0049, + "step": 35040 + }, + { + "grad_norm": 0.26032671332359314, + "learning_rate": 4.028197482033266e-05, + "loss": 0.004, + "step": 35050 + }, + { + "grad_norm": 0.1645016223192215, + "learning_rate": 4.0254943990638246e-05, + "loss": 0.0039, + "step": 35060 + }, + { + "grad_norm": 0.17950184643268585, + "learning_rate": 4.022791612123454e-05, + "loss": 0.0048, + "step": 35070 + }, + { + "grad_norm": 0.18055547773838043, + "learning_rate": 4.020089122033192e-05, + "loss": 0.0071, + "step": 35080 + }, + { + "grad_norm": 0.1926897168159485, + "learning_rate": 4.01738692961398e-05, + "loss": 0.0046, + "step": 35090 + }, + { + "grad_norm": 0.19580279290676117, + "learning_rate": 4.014685035686675e-05, + "loss": 0.0043, + "step": 35100 + }, + { + "grad_norm": 0.23293651640415192, + "learning_rate": 4.011983441072039e-05, + "loss": 0.0067, + "step": 35110 + }, + { + "grad_norm": 0.2102246731519699, + "learning_rate": 4.0092821465907485e-05, + "loss": 0.005, + "step": 35120 + }, + { + "grad_norm": 0.14274299144744873, + "learning_rate": 4.006581153063383e-05, + "loss": 0.0044, + "step": 35130 + }, + { + "grad_norm": 0.17626157402992249, + "learning_rate": 4.003880461310432e-05, + "loss": 0.0037, + "step": 35140 + }, + { + "grad_norm": 0.1781103014945984, + "learning_rate": 4.001180072152298e-05, + "loss": 0.0038, + "step": 35150 + }, + { + "grad_norm": 0.169856995344162, + "learning_rate": 3.998479986409285e-05, + "loss": 0.0041, + "step": 35160 + }, + { + "grad_norm": 0.19370587170124054, + "learning_rate": 3.995780204901607e-05, + "loss": 0.0038, + "step": 35170 + }, + { + "grad_norm": 0.182124063372612, + "learning_rate": 3.993080728449391e-05, + "loss": 0.0054, + "step": 35180 + }, + { + "grad_norm": 0.14766483008861542, + "learning_rate": 3.990381557872661e-05, + "loss": 0.0057, + "step": 35190 + }, + { + "grad_norm": 0.2013283669948578, + "learning_rate": 3.987682693991359e-05, + "loss": 0.0069, + "step": 35200 + }, + { + "grad_norm": 0.20138674974441528, + "learning_rate": 3.9849841376253226e-05, + "loss": 0.0042, + "step": 35210 + }, + { + "grad_norm": 0.15418125689029694, + "learning_rate": 3.982285889594306e-05, + "loss": 0.0044, + "step": 35220 + }, + { + "grad_norm": 0.13768728077411652, + "learning_rate": 3.9795879507179665e-05, + "loss": 0.0047, + "step": 35230 + }, + { + "grad_norm": 0.18905511498451233, + "learning_rate": 3.9768903218158634e-05, + "loss": 0.0043, + "step": 35240 + }, + { + "grad_norm": 0.14632026851177216, + "learning_rate": 3.974193003707468e-05, + "loss": 0.0048, + "step": 35250 + }, + { + "grad_norm": 0.14236973226070404, + "learning_rate": 3.971495997212152e-05, + "loss": 0.005, + "step": 35260 + }, + { + "grad_norm": 0.15416570007801056, + "learning_rate": 3.9687993031491985e-05, + "loss": 0.0041, + "step": 35270 + }, + { + "grad_norm": 0.13119575381278992, + "learning_rate": 3.966102922337787e-05, + "loss": 0.0028, + "step": 35280 + }, + { + "grad_norm": 0.20743732154369354, + "learning_rate": 3.963406855597009e-05, + "loss": 0.0046, + "step": 35290 + }, + { + "grad_norm": 0.1820625215768814, + "learning_rate": 3.960711103745861e-05, + "loss": 0.0045, + "step": 35300 + }, + { + "grad_norm": 0.21300853788852692, + "learning_rate": 3.958015667603237e-05, + "loss": 0.004, + "step": 35310 + }, + { + "grad_norm": 0.1771298050880432, + "learning_rate": 3.955320547987943e-05, + "loss": 0.0046, + "step": 35320 + }, + { + "grad_norm": 0.2095687985420227, + "learning_rate": 3.952625745718681e-05, + "loss": 0.005, + "step": 35330 + }, + { + "grad_norm": 0.16684526205062866, + "learning_rate": 3.949931261614064e-05, + "loss": 0.0046, + "step": 35340 + }, + { + "grad_norm": 0.19215929508209229, + "learning_rate": 3.947237096492605e-05, + "loss": 0.0053, + "step": 35350 + }, + { + "grad_norm": 0.1813025027513504, + "learning_rate": 3.944543251172719e-05, + "loss": 0.0038, + "step": 35360 + }, + { + "grad_norm": 0.18504272401332855, + "learning_rate": 3.941849726472725e-05, + "loss": 0.007, + "step": 35370 + }, + { + "grad_norm": 0.12724550068378448, + "learning_rate": 3.939156523210846e-05, + "loss": 0.0033, + "step": 35380 + }, + { + "grad_norm": 0.142240509390831, + "learning_rate": 3.9364636422052046e-05, + "loss": 0.0049, + "step": 35390 + }, + { + "grad_norm": 0.21093280613422394, + "learning_rate": 3.933771084273828e-05, + "loss": 0.0046, + "step": 35400 + }, + { + "grad_norm": 0.17055660486221313, + "learning_rate": 3.931078850234643e-05, + "loss": 0.0038, + "step": 35410 + }, + { + "grad_norm": 0.18784572184085846, + "learning_rate": 3.928386940905483e-05, + "loss": 0.0054, + "step": 35420 + }, + { + "grad_norm": 0.2002873420715332, + "learning_rate": 3.925695357104073e-05, + "loss": 0.0048, + "step": 35430 + }, + { + "grad_norm": 0.21293509006500244, + "learning_rate": 3.923004099648049e-05, + "loss": 0.0068, + "step": 35440 + }, + { + "grad_norm": 0.22828805446624756, + "learning_rate": 3.920313169354944e-05, + "loss": 0.0051, + "step": 35450 + }, + { + "grad_norm": 0.26494100689888, + "learning_rate": 3.9176225670421897e-05, + "loss": 0.0059, + "step": 35460 + }, + { + "grad_norm": 0.18764545023441315, + "learning_rate": 3.9149322935271224e-05, + "loss": 0.0051, + "step": 35470 + }, + { + "grad_norm": 0.1537577360868454, + "learning_rate": 3.9122423496269725e-05, + "loss": 0.0049, + "step": 35480 + }, + { + "grad_norm": 0.13006477057933807, + "learning_rate": 3.909552736158877e-05, + "loss": 0.0028, + "step": 35490 + }, + { + "grad_norm": 0.13274535536766052, + "learning_rate": 3.90686345393987e-05, + "loss": 0.0053, + "step": 35500 + }, + { + "grad_norm": 0.20071713626384735, + "learning_rate": 3.9041745037868816e-05, + "loss": 0.0066, + "step": 35510 + }, + { + "grad_norm": 0.15042737126350403, + "learning_rate": 3.9014858865167465e-05, + "loss": 0.005, + "step": 35520 + }, + { + "grad_norm": 0.18074959516525269, + "learning_rate": 3.8987976029461935e-05, + "loss": 0.0046, + "step": 35530 + }, + { + "grad_norm": 0.1836220771074295, + "learning_rate": 3.896109653891853e-05, + "loss": 0.0047, + "step": 35540 + }, + { + "grad_norm": 0.18061544001102448, + "learning_rate": 3.893422040170254e-05, + "loss": 0.0039, + "step": 35550 + }, + { + "grad_norm": 0.12281706929206848, + "learning_rate": 3.8907347625978207e-05, + "loss": 0.0046, + "step": 35560 + }, + { + "grad_norm": 0.17047938704490662, + "learning_rate": 3.88804782199088e-05, + "loss": 0.0043, + "step": 35570 + }, + { + "grad_norm": 0.1955205798149109, + "learning_rate": 3.8853612191656495e-05, + "loss": 0.0036, + "step": 35580 + }, + { + "grad_norm": 0.2124948799610138, + "learning_rate": 3.88267495493825e-05, + "loss": 0.0043, + "step": 35590 + }, + { + "grad_norm": 0.22286926209926605, + "learning_rate": 3.8799890301247004e-05, + "loss": 0.0044, + "step": 35600 + }, + { + "grad_norm": 0.17063790559768677, + "learning_rate": 3.8773034455409096e-05, + "loss": 0.0049, + "step": 35610 + }, + { + "grad_norm": 0.18034812808036804, + "learning_rate": 3.8746182020026904e-05, + "loss": 0.0055, + "step": 35620 + }, + { + "grad_norm": 0.23504638671875, + "learning_rate": 3.871933300325745e-05, + "loss": 0.0062, + "step": 35630 + }, + { + "grad_norm": 0.2064599096775055, + "learning_rate": 3.869248741325679e-05, + "loss": 0.0059, + "step": 35640 + }, + { + "grad_norm": 0.17354246973991394, + "learning_rate": 3.866564525817992e-05, + "loss": 0.0042, + "step": 35650 + }, + { + "grad_norm": 0.19360032677650452, + "learning_rate": 3.8638806546180725e-05, + "loss": 0.0047, + "step": 35660 + }, + { + "grad_norm": 0.1580348163843155, + "learning_rate": 3.861197128541213e-05, + "loss": 0.0045, + "step": 35670 + }, + { + "grad_norm": 0.14937378466129303, + "learning_rate": 3.858513948402599e-05, + "loss": 0.0046, + "step": 35680 + }, + { + "grad_norm": 0.1953752487897873, + "learning_rate": 3.8558311150173077e-05, + "loss": 0.0063, + "step": 35690 + }, + { + "grad_norm": 0.1868027001619339, + "learning_rate": 3.853148629200312e-05, + "loss": 0.0046, + "step": 35700 + }, + { + "grad_norm": 0.16624651849269867, + "learning_rate": 3.850466491766482e-05, + "loss": 0.0039, + "step": 35710 + }, + { + "grad_norm": 0.15174205601215363, + "learning_rate": 3.847784703530583e-05, + "loss": 0.0071, + "step": 35720 + }, + { + "grad_norm": 0.1895485520362854, + "learning_rate": 3.845103265307266e-05, + "loss": 0.0051, + "step": 35730 + }, + { + "grad_norm": 0.24633023142814636, + "learning_rate": 3.842422177911086e-05, + "loss": 0.004, + "step": 35740 + }, + { + "grad_norm": 0.1753271222114563, + "learning_rate": 3.8397414421564826e-05, + "loss": 0.0043, + "step": 35750 + }, + { + "grad_norm": 0.16444982588291168, + "learning_rate": 3.8370610588577935e-05, + "loss": 0.0039, + "step": 35760 + }, + { + "grad_norm": 0.19640284776687622, + "learning_rate": 3.834381028829251e-05, + "loss": 0.0053, + "step": 35770 + }, + { + "grad_norm": 0.1419774740934372, + "learning_rate": 3.8317013528849745e-05, + "loss": 0.003, + "step": 35780 + }, + { + "grad_norm": 0.15221914649009705, + "learning_rate": 3.8290220318389815e-05, + "loss": 0.0036, + "step": 35790 + }, + { + "grad_norm": 0.17576512694358826, + "learning_rate": 3.8263430665051746e-05, + "loss": 0.0053, + "step": 35800 + }, + { + "grad_norm": 0.18683822453022003, + "learning_rate": 3.8236644576973554e-05, + "loss": 0.006, + "step": 35810 + }, + { + "grad_norm": 0.20936179161071777, + "learning_rate": 3.820986206229217e-05, + "loss": 0.0037, + "step": 35820 + }, + { + "grad_norm": 0.12605148553848267, + "learning_rate": 3.8183083129143384e-05, + "loss": 0.0039, + "step": 35830 + }, + { + "grad_norm": 0.1542615443468094, + "learning_rate": 3.815630778566193e-05, + "loss": 0.0043, + "step": 35840 + }, + { + "grad_norm": 0.14638002216815948, + "learning_rate": 3.812953603998145e-05, + "loss": 0.004, + "step": 35850 + }, + { + "grad_norm": 0.12962792813777924, + "learning_rate": 3.8102767900234504e-05, + "loss": 0.0032, + "step": 35860 + }, + { + "grad_norm": 0.178972065448761, + "learning_rate": 3.807600337455256e-05, + "loss": 0.0064, + "step": 35870 + }, + { + "grad_norm": 0.14596381783485413, + "learning_rate": 3.804924247106593e-05, + "loss": 0.0034, + "step": 35880 + }, + { + "grad_norm": 0.16244550049304962, + "learning_rate": 3.8022485197903925e-05, + "loss": 0.0038, + "step": 35890 + }, + { + "grad_norm": 0.2238420695066452, + "learning_rate": 3.799573156319464e-05, + "loss": 0.0033, + "step": 35900 + }, + { + "grad_norm": 0.16197574138641357, + "learning_rate": 3.796898157506515e-05, + "loss": 0.0055, + "step": 35910 + }, + { + "grad_norm": 0.16722843050956726, + "learning_rate": 3.794223524164143e-05, + "loss": 0.0032, + "step": 35920 + }, + { + "grad_norm": 0.14297403395175934, + "learning_rate": 3.7915492571048245e-05, + "loss": 0.004, + "step": 35930 + }, + { + "grad_norm": 0.16478975117206573, + "learning_rate": 3.788875357140937e-05, + "loss": 0.0033, + "step": 35940 + }, + { + "grad_norm": 0.101627878844738, + "learning_rate": 3.786201825084736e-05, + "loss": 0.0038, + "step": 35950 + }, + { + "grad_norm": 0.14181600511074066, + "learning_rate": 3.783528661748372e-05, + "loss": 0.0031, + "step": 35960 + }, + { + "grad_norm": 0.2099507600069046, + "learning_rate": 3.780855867943882e-05, + "loss": 0.0045, + "step": 35970 + }, + { + "grad_norm": 0.16354510188102722, + "learning_rate": 3.778183444483189e-05, + "loss": 0.0048, + "step": 35980 + }, + { + "grad_norm": 0.20852220058441162, + "learning_rate": 3.775511392178108e-05, + "loss": 0.0047, + "step": 35990 + }, + { + "grad_norm": 0.16216205060482025, + "learning_rate": 3.772839711840332e-05, + "loss": 0.0057, + "step": 36000 + }, + { + "grad_norm": 0.14635902643203735, + "learning_rate": 3.7701684042814515e-05, + "loss": 0.0043, + "step": 36010 + }, + { + "grad_norm": 0.15734851360321045, + "learning_rate": 3.76749747031294e-05, + "loss": 0.0035, + "step": 36020 + }, + { + "grad_norm": 0.19009758532047272, + "learning_rate": 3.764826910746152e-05, + "loss": 0.0062, + "step": 36030 + }, + { + "grad_norm": 0.16363267600536346, + "learning_rate": 3.762156726392338e-05, + "loss": 0.005, + "step": 36040 + }, + { + "grad_norm": 0.17221324145793915, + "learning_rate": 3.759486918062625e-05, + "loss": 0.0038, + "step": 36050 + }, + { + "grad_norm": 0.17455646395683289, + "learning_rate": 3.756817486568033e-05, + "loss": 0.0054, + "step": 36060 + }, + { + "grad_norm": 0.1589495986700058, + "learning_rate": 3.7541484327194654e-05, + "loss": 0.0045, + "step": 36070 + }, + { + "grad_norm": 0.1859559565782547, + "learning_rate": 3.751479757327707e-05, + "loss": 0.0045, + "step": 36080 + }, + { + "grad_norm": 0.16552455723285675, + "learning_rate": 3.7488114612034345e-05, + "loss": 0.0046, + "step": 36090 + }, + { + "grad_norm": 0.17324581742286682, + "learning_rate": 3.7461435451572044e-05, + "loss": 0.0058, + "step": 36100 + }, + { + "grad_norm": 0.1281116008758545, + "learning_rate": 3.743476009999459e-05, + "loss": 0.0044, + "step": 36110 + }, + { + "grad_norm": 0.2255716323852539, + "learning_rate": 3.7408088565405245e-05, + "loss": 0.0046, + "step": 36120 + }, + { + "grad_norm": 0.18869923055171967, + "learning_rate": 3.738142085590612e-05, + "loss": 0.0039, + "step": 36130 + }, + { + "grad_norm": 0.19128362834453583, + "learning_rate": 3.7354756979598194e-05, + "loss": 0.0035, + "step": 36140 + }, + { + "grad_norm": 0.23120026290416718, + "learning_rate": 3.7328096944581187e-05, + "loss": 0.004, + "step": 36150 + }, + { + "grad_norm": 0.23744067549705505, + "learning_rate": 3.730144075895377e-05, + "loss": 0.0033, + "step": 36160 + }, + { + "grad_norm": 0.2170422524213791, + "learning_rate": 3.727478843081335e-05, + "loss": 0.0049, + "step": 36170 + }, + { + "grad_norm": 0.17735883593559265, + "learning_rate": 3.72481399682562e-05, + "loss": 0.0037, + "step": 36180 + }, + { + "grad_norm": 0.20702946186065674, + "learning_rate": 3.722149537937747e-05, + "loss": 0.0042, + "step": 36190 + }, + { + "grad_norm": 0.17318615317344666, + "learning_rate": 3.7194854672271015e-05, + "loss": 0.0037, + "step": 36200 + }, + { + "grad_norm": 0.12289384007453918, + "learning_rate": 3.7168217855029644e-05, + "loss": 0.0033, + "step": 36210 + }, + { + "grad_norm": 0.13628794252872467, + "learning_rate": 3.7141584935744856e-05, + "loss": 0.0035, + "step": 36220 + }, + { + "grad_norm": 0.1993735432624817, + "learning_rate": 3.7114955922507055e-05, + "loss": 0.0056, + "step": 36230 + }, + { + "grad_norm": 0.24133211374282837, + "learning_rate": 3.708833082340545e-05, + "loss": 0.0045, + "step": 36240 + }, + { + "grad_norm": 0.1532207429409027, + "learning_rate": 3.7061709646528034e-05, + "loss": 0.0041, + "step": 36250 + }, + { + "grad_norm": 0.19150801002979279, + "learning_rate": 3.7035092399961604e-05, + "loss": 0.0041, + "step": 36260 + }, + { + "grad_norm": 0.1891794502735138, + "learning_rate": 3.700847909179177e-05, + "loss": 0.004, + "step": 36270 + }, + { + "grad_norm": 0.20939947664737701, + "learning_rate": 3.698186973010297e-05, + "loss": 0.0044, + "step": 36280 + }, + { + "grad_norm": 0.14455635845661163, + "learning_rate": 3.695526432297844e-05, + "loss": 0.0037, + "step": 36290 + }, + { + "grad_norm": 0.1436607986688614, + "learning_rate": 3.692866287850017e-05, + "loss": 0.0036, + "step": 36300 + }, + { + "grad_norm": 0.1685621589422226, + "learning_rate": 3.6902065404749006e-05, + "loss": 0.0036, + "step": 36310 + }, + { + "grad_norm": 0.1561403125524521, + "learning_rate": 3.6875471909804516e-05, + "loss": 0.0033, + "step": 36320 + }, + { + "grad_norm": 0.1796732246875763, + "learning_rate": 3.6848882401745135e-05, + "loss": 0.0046, + "step": 36330 + }, + { + "grad_norm": 0.18659226596355438, + "learning_rate": 3.682229688864806e-05, + "loss": 0.0048, + "step": 36340 + }, + { + "grad_norm": 0.20979976654052734, + "learning_rate": 3.6795715378589235e-05, + "loss": 0.0042, + "step": 36350 + }, + { + "grad_norm": 0.18520836532115936, + "learning_rate": 3.676913787964345e-05, + "loss": 0.0032, + "step": 36360 + }, + { + "grad_norm": 0.15382996201515198, + "learning_rate": 3.674256439988423e-05, + "loss": 0.0045, + "step": 36370 + }, + { + "grad_norm": 0.18960382044315338, + "learning_rate": 3.6715994947383904e-05, + "loss": 0.0052, + "step": 36380 + }, + { + "grad_norm": 0.20157665014266968, + "learning_rate": 3.668942953021357e-05, + "loss": 0.0054, + "step": 36390 + }, + { + "grad_norm": 0.1395639330148697, + "learning_rate": 3.66628681564431e-05, + "loss": 0.0037, + "step": 36400 + }, + { + "grad_norm": 0.16661204397678375, + "learning_rate": 3.663631083414114e-05, + "loss": 0.004, + "step": 36410 + }, + { + "grad_norm": 0.18493399024009705, + "learning_rate": 3.660975757137509e-05, + "loss": 0.0037, + "step": 36420 + }, + { + "grad_norm": 0.15603278577327728, + "learning_rate": 3.658320837621114e-05, + "loss": 0.005, + "step": 36430 + }, + { + "grad_norm": 0.19166652858257294, + "learning_rate": 3.655666325671426e-05, + "loss": 0.0044, + "step": 36440 + }, + { + "grad_norm": 0.19289395213127136, + "learning_rate": 3.65301222209481e-05, + "loss": 0.004, + "step": 36450 + }, + { + "grad_norm": 0.13206928968429565, + "learning_rate": 3.650358527697519e-05, + "loss": 0.0035, + "step": 36460 + }, + { + "grad_norm": 0.15644435584545135, + "learning_rate": 3.64770524328567e-05, + "loss": 0.0038, + "step": 36470 + }, + { + "grad_norm": 0.12828420102596283, + "learning_rate": 3.645052369665265e-05, + "loss": 0.0036, + "step": 36480 + }, + { + "grad_norm": 0.2099485844373703, + "learning_rate": 3.6423999076421724e-05, + "loss": 0.0042, + "step": 36490 + }, + { + "grad_norm": 0.18850077688694, + "learning_rate": 3.639747858022142e-05, + "loss": 0.0036, + "step": 36500 + }, + { + "grad_norm": 0.1435929387807846, + "learning_rate": 3.637096221610799e-05, + "loss": 0.0029, + "step": 36510 + }, + { + "grad_norm": 0.13435561954975128, + "learning_rate": 3.634444999213638e-05, + "loss": 0.0029, + "step": 36520 + }, + { + "grad_norm": 0.16715021431446075, + "learning_rate": 3.6317941916360296e-05, + "loss": 0.003, + "step": 36530 + }, + { + "grad_norm": 0.18665917217731476, + "learning_rate": 3.629143799683221e-05, + "loss": 0.0042, + "step": 36540 + }, + { + "grad_norm": 0.21228618919849396, + "learning_rate": 3.626493824160331e-05, + "loss": 0.0034, + "step": 36550 + }, + { + "grad_norm": 0.1824057251214981, + "learning_rate": 3.623844265872352e-05, + "loss": 0.003, + "step": 36560 + }, + { + "grad_norm": 0.1753162145614624, + "learning_rate": 3.621195125624149e-05, + "loss": 0.0033, + "step": 36570 + }, + { + "grad_norm": 0.22050733864307404, + "learning_rate": 3.618546404220463e-05, + "loss": 0.0057, + "step": 36580 + }, + { + "grad_norm": 0.2052624672651291, + "learning_rate": 3.615898102465903e-05, + "loss": 0.0055, + "step": 36590 + }, + { + "grad_norm": 0.16407272219657898, + "learning_rate": 3.6132502211649544e-05, + "loss": 0.0047, + "step": 36600 + }, + { + "grad_norm": 0.20948605239391327, + "learning_rate": 3.610602761121975e-05, + "loss": 0.0055, + "step": 36610 + }, + { + "grad_norm": 0.16353218257427216, + "learning_rate": 3.6079557231411897e-05, + "loss": 0.0054, + "step": 36620 + }, + { + "grad_norm": 0.14929567277431488, + "learning_rate": 3.6053091080267035e-05, + "loss": 0.0034, + "step": 36630 + }, + { + "grad_norm": 0.18421703577041626, + "learning_rate": 3.602662916582483e-05, + "loss": 0.0036, + "step": 36640 + }, + { + "grad_norm": 0.16220426559448242, + "learning_rate": 3.600017149612375e-05, + "loss": 0.0059, + "step": 36650 + }, + { + "grad_norm": 0.16695435345172882, + "learning_rate": 3.5973718079200935e-05, + "loss": 0.0032, + "step": 36660 + }, + { + "grad_norm": 0.21801723539829254, + "learning_rate": 3.5947268923092216e-05, + "loss": 0.0035, + "step": 36670 + }, + { + "grad_norm": 0.14759376645088196, + "learning_rate": 3.592082403583216e-05, + "loss": 0.0041, + "step": 36680 + }, + { + "grad_norm": 0.21544069051742554, + "learning_rate": 3.5894383425454004e-05, + "loss": 0.0056, + "step": 36690 + }, + { + "grad_norm": 0.12972599267959595, + "learning_rate": 3.586794709998975e-05, + "loss": 0.0044, + "step": 36700 + }, + { + "grad_norm": 0.13849015533924103, + "learning_rate": 3.584151506747002e-05, + "loss": 0.0039, + "step": 36710 + }, + { + "grad_norm": 0.17557428777217865, + "learning_rate": 3.581508733592418e-05, + "loss": 0.0045, + "step": 36720 + }, + { + "grad_norm": 0.17779088020324707, + "learning_rate": 3.5788663913380297e-05, + "loss": 0.0035, + "step": 36730 + }, + { + "grad_norm": 0.16489960253238678, + "learning_rate": 3.576224480786506e-05, + "loss": 0.0036, + "step": 36740 + }, + { + "grad_norm": 0.13422001898288727, + "learning_rate": 3.573583002740393e-05, + "loss": 0.0059, + "step": 36750 + }, + { + "grad_norm": 0.23320846259593964, + "learning_rate": 3.570941958002103e-05, + "loss": 0.0053, + "step": 36760 + }, + { + "grad_norm": 0.153609961271286, + "learning_rate": 3.568301347373912e-05, + "loss": 0.0033, + "step": 36770 + }, + { + "grad_norm": 0.14701387286186218, + "learning_rate": 3.5656611716579726e-05, + "loss": 0.0033, + "step": 36780 + }, + { + "grad_norm": 0.19461458921432495, + "learning_rate": 3.5630214316562946e-05, + "loss": 0.0046, + "step": 36790 + }, + { + "grad_norm": 0.17379187047481537, + "learning_rate": 3.560382128170766e-05, + "loss": 0.0048, + "step": 36800 + }, + { + "grad_norm": 0.21024708449840546, + "learning_rate": 3.5577432620031374e-05, + "loss": 0.0036, + "step": 36810 + }, + { + "grad_norm": 0.20913304388523102, + "learning_rate": 3.5551048339550216e-05, + "loss": 0.0045, + "step": 36820 + }, + { + "grad_norm": 0.19161812961101532, + "learning_rate": 3.55246684482791e-05, + "loss": 0.0055, + "step": 36830 + }, + { + "grad_norm": 0.17264392971992493, + "learning_rate": 3.5498292954231496e-05, + "loss": 0.0042, + "step": 36840 + }, + { + "grad_norm": 0.16684402525424957, + "learning_rate": 3.54719218654196e-05, + "loss": 0.0047, + "step": 36850 + }, + { + "grad_norm": 0.16446812450885773, + "learning_rate": 3.544555518985425e-05, + "loss": 0.0046, + "step": 36860 + }, + { + "grad_norm": 0.12240175902843475, + "learning_rate": 3.541919293554494e-05, + "loss": 0.0042, + "step": 36870 + }, + { + "grad_norm": 0.15868568420410156, + "learning_rate": 3.539283511049985e-05, + "loss": 0.0041, + "step": 36880 + }, + { + "grad_norm": 0.25886955857276917, + "learning_rate": 3.5366481722725755e-05, + "loss": 0.0032, + "step": 36890 + }, + { + "grad_norm": 0.12546958029270172, + "learning_rate": 3.534013278022816e-05, + "loss": 0.0035, + "step": 36900 + }, + { + "grad_norm": 0.15886810421943665, + "learning_rate": 3.531378829101113e-05, + "loss": 0.0065, + "step": 36910 + }, + { + "grad_norm": 0.1262883096933365, + "learning_rate": 3.528744826307746e-05, + "loss": 0.0037, + "step": 36920 + }, + { + "grad_norm": 0.13824865221977234, + "learning_rate": 3.5261112704428554e-05, + "loss": 0.0031, + "step": 36930 + }, + { + "grad_norm": 0.17388708889484406, + "learning_rate": 3.523478162306443e-05, + "loss": 0.0045, + "step": 36940 + }, + { + "grad_norm": 0.17003922164440155, + "learning_rate": 3.520845502698381e-05, + "loss": 0.0066, + "step": 36950 + }, + { + "grad_norm": 0.18128372728824615, + "learning_rate": 3.5182132924184005e-05, + "loss": 0.0062, + "step": 36960 + }, + { + "grad_norm": 0.1506938487291336, + "learning_rate": 3.5155815322660966e-05, + "loss": 0.0044, + "step": 36970 + }, + { + "grad_norm": 0.10939103364944458, + "learning_rate": 3.512950223040931e-05, + "loss": 0.0039, + "step": 36980 + }, + { + "grad_norm": 0.13286690413951874, + "learning_rate": 3.5103193655422216e-05, + "loss": 0.0035, + "step": 36990 + }, + { + "grad_norm": 0.18425068259239197, + "learning_rate": 3.5076889605691596e-05, + "loss": 0.0052, + "step": 37000 + }, + { + "grad_norm": 0.2002500742673874, + "learning_rate": 3.505059008920787e-05, + "loss": 0.0043, + "step": 37010 + }, + { + "grad_norm": 0.17909139394760132, + "learning_rate": 3.502429511396016e-05, + "loss": 0.0033, + "step": 37020 + }, + { + "grad_norm": 0.13977700471878052, + "learning_rate": 3.4998004687936196e-05, + "loss": 0.0039, + "step": 37030 + }, + { + "grad_norm": 0.16155658662319183, + "learning_rate": 3.497171881912229e-05, + "loss": 0.0059, + "step": 37040 + }, + { + "grad_norm": 0.1531539261341095, + "learning_rate": 3.494543751550342e-05, + "loss": 0.0045, + "step": 37050 + }, + { + "grad_norm": 0.16485808789730072, + "learning_rate": 3.491916078506313e-05, + "loss": 0.0058, + "step": 37060 + }, + { + "grad_norm": 0.17877094447612762, + "learning_rate": 3.489288863578361e-05, + "loss": 0.0037, + "step": 37070 + }, + { + "grad_norm": 0.18021926283836365, + "learning_rate": 3.4866621075645646e-05, + "loss": 0.0037, + "step": 37080 + }, + { + "grad_norm": 0.21794147789478302, + "learning_rate": 3.4840358112628614e-05, + "loss": 0.0043, + "step": 37090 + }, + { + "grad_norm": 0.15519224107265472, + "learning_rate": 3.481409975471053e-05, + "loss": 0.0037, + "step": 37100 + }, + { + "grad_norm": 0.17899326980113983, + "learning_rate": 3.4787846009867986e-05, + "loss": 0.0037, + "step": 37110 + }, + { + "grad_norm": 0.17512184381484985, + "learning_rate": 3.476159688607615e-05, + "loss": 0.0032, + "step": 37120 + }, + { + "grad_norm": 0.12973667681217194, + "learning_rate": 3.4735352391308854e-05, + "loss": 0.0034, + "step": 37130 + }, + { + "grad_norm": 0.14092187583446503, + "learning_rate": 3.4709112533538446e-05, + "loss": 0.0032, + "step": 37140 + }, + { + "grad_norm": 0.1747317612171173, + "learning_rate": 3.4682877320735934e-05, + "loss": 0.0042, + "step": 37150 + }, + { + "grad_norm": 0.11954566836357117, + "learning_rate": 3.465664676087085e-05, + "loss": 0.0037, + "step": 37160 + }, + { + "grad_norm": 0.18138955533504486, + "learning_rate": 3.463042086191136e-05, + "loss": 0.0049, + "step": 37170 + }, + { + "grad_norm": 0.1978742778301239, + "learning_rate": 3.460419963182423e-05, + "loss": 0.0055, + "step": 37180 + }, + { + "grad_norm": 0.18544071912765503, + "learning_rate": 3.457798307857473e-05, + "loss": 0.006, + "step": 37190 + }, + { + "grad_norm": 0.17738130688667297, + "learning_rate": 3.455177121012678e-05, + "loss": 0.0042, + "step": 37200 + }, + { + "grad_norm": 0.1703493446111679, + "learning_rate": 3.452556403444285e-05, + "loss": 0.0062, + "step": 37210 + }, + { + "grad_norm": 0.16046540439128876, + "learning_rate": 3.4499361559483975e-05, + "loss": 0.0044, + "step": 37220 + }, + { + "grad_norm": 0.21843072772026062, + "learning_rate": 3.44731637932098e-05, + "loss": 0.0057, + "step": 37230 + }, + { + "grad_norm": 0.1549636721611023, + "learning_rate": 3.44469707435785e-05, + "loss": 0.0035, + "step": 37240 + }, + { + "grad_norm": 0.1514454334974289, + "learning_rate": 3.4420782418546835e-05, + "loss": 0.0038, + "step": 37250 + }, + { + "grad_norm": 0.14715375006198883, + "learning_rate": 3.439459882607012e-05, + "loss": 0.0041, + "step": 37260 + }, + { + "grad_norm": 0.19239705801010132, + "learning_rate": 3.436841997410225e-05, + "loss": 0.0036, + "step": 37270 + }, + { + "grad_norm": 0.23121893405914307, + "learning_rate": 3.434224587059567e-05, + "loss": 0.0054, + "step": 37280 + }, + { + "grad_norm": 0.17497767508029938, + "learning_rate": 3.431607652350136e-05, + "loss": 0.0032, + "step": 37290 + }, + { + "grad_norm": 0.1866466999053955, + "learning_rate": 3.428991194076891e-05, + "loss": 0.0049, + "step": 37300 + }, + { + "grad_norm": 0.17356768250465393, + "learning_rate": 3.4263752130346394e-05, + "loss": 0.0034, + "step": 37310 + }, + { + "grad_norm": 0.12500721216201782, + "learning_rate": 3.4237597100180515e-05, + "loss": 0.0066, + "step": 37320 + }, + { + "grad_norm": 0.25784391164779663, + "learning_rate": 3.4211446858216427e-05, + "loss": 0.0052, + "step": 37330 + }, + { + "grad_norm": 0.19535090029239655, + "learning_rate": 3.4185301412397915e-05, + "loss": 0.0048, + "step": 37340 + }, + { + "grad_norm": 0.1677456945180893, + "learning_rate": 3.415916077066729e-05, + "loss": 0.0045, + "step": 37350 + }, + { + "grad_norm": 0.17326194047927856, + "learning_rate": 3.413302494096535e-05, + "loss": 0.005, + "step": 37360 + }, + { + "grad_norm": 0.15238283574581146, + "learning_rate": 3.410689393123151e-05, + "loss": 0.0033, + "step": 37370 + }, + { + "grad_norm": 0.1412474364042282, + "learning_rate": 3.408076774940364e-05, + "loss": 0.0039, + "step": 37380 + }, + { + "grad_norm": 0.12052754312753677, + "learning_rate": 3.40546464034182e-05, + "loss": 0.005, + "step": 37390 + }, + { + "grad_norm": 0.1928095519542694, + "learning_rate": 3.4028529901210185e-05, + "loss": 0.0042, + "step": 37400 + }, + { + "grad_norm": 0.17098627984523773, + "learning_rate": 3.4002418250713086e-05, + "loss": 0.0055, + "step": 37410 + }, + { + "grad_norm": 0.1517518311738968, + "learning_rate": 3.3976311459858936e-05, + "loss": 0.0035, + "step": 37420 + }, + { + "grad_norm": 0.14871418476104736, + "learning_rate": 3.395020953657826e-05, + "loss": 0.0031, + "step": 37430 + }, + { + "grad_norm": 0.12419089674949646, + "learning_rate": 3.3924112488800165e-05, + "loss": 0.0057, + "step": 37440 + }, + { + "grad_norm": 0.12372568249702454, + "learning_rate": 3.389802032445225e-05, + "loss": 0.0038, + "step": 37450 + }, + { + "grad_norm": 0.19098053872585297, + "learning_rate": 3.38719330514606e-05, + "loss": 0.0052, + "step": 37460 + }, + { + "grad_norm": 0.2306004911661148, + "learning_rate": 3.3845850677749866e-05, + "loss": 0.0052, + "step": 37470 + }, + { + "grad_norm": 0.1845322996377945, + "learning_rate": 3.3819773211243157e-05, + "loss": 0.0038, + "step": 37480 + }, + { + "grad_norm": 0.13190294802188873, + "learning_rate": 3.379370065986213e-05, + "loss": 0.0049, + "step": 37490 + }, + { + "grad_norm": 0.230205237865448, + "learning_rate": 3.3767633031526955e-05, + "loss": 0.0045, + "step": 37500 + }, + { + "grad_norm": 0.15713104605674744, + "learning_rate": 3.374157033415626e-05, + "loss": 0.0039, + "step": 37510 + }, + { + "grad_norm": 0.15131232142448425, + "learning_rate": 3.371551257566723e-05, + "loss": 0.0058, + "step": 37520 + }, + { + "grad_norm": 0.18210501968860626, + "learning_rate": 3.36894597639755e-05, + "loss": 0.0046, + "step": 37530 + }, + { + "grad_norm": 0.14680740237236023, + "learning_rate": 3.366341190699523e-05, + "loss": 0.0039, + "step": 37540 + }, + { + "grad_norm": 0.17157544195652008, + "learning_rate": 3.36373690126391e-05, + "loss": 0.0038, + "step": 37550 + }, + { + "grad_norm": 0.1537700593471527, + "learning_rate": 3.3611331088818234e-05, + "loss": 0.0042, + "step": 37560 + }, + { + "grad_norm": 0.17883045971393585, + "learning_rate": 3.3585298143442265e-05, + "loss": 0.0061, + "step": 37570 + }, + { + "grad_norm": 0.1626027673482895, + "learning_rate": 3.35592701844193e-05, + "loss": 0.0038, + "step": 37580 + }, + { + "grad_norm": 0.20300991833209991, + "learning_rate": 3.353324721965596e-05, + "loss": 0.0035, + "step": 37590 + }, + { + "grad_norm": 0.1403523087501526, + "learning_rate": 3.350722925705736e-05, + "loss": 0.0026, + "step": 37600 + }, + { + "grad_norm": 0.17607346177101135, + "learning_rate": 3.348121630452703e-05, + "loss": 0.0035, + "step": 37610 + }, + { + "grad_norm": 0.1412612348794937, + "learning_rate": 3.3455208369967044e-05, + "loss": 0.0029, + "step": 37620 + }, + { + "grad_norm": 0.12900951504707336, + "learning_rate": 3.34292054612779e-05, + "loss": 0.0031, + "step": 37630 + }, + { + "grad_norm": 0.17048409581184387, + "learning_rate": 3.340320758635861e-05, + "loss": 0.0034, + "step": 37640 + }, + { + "grad_norm": 0.1889721155166626, + "learning_rate": 3.337721475310666e-05, + "loss": 0.0049, + "step": 37650 + }, + { + "grad_norm": 0.16436006128787994, + "learning_rate": 3.335122696941795e-05, + "loss": 0.0046, + "step": 37660 + }, + { + "grad_norm": 0.20078778266906738, + "learning_rate": 3.332524424318692e-05, + "loss": 0.0047, + "step": 37670 + }, + { + "grad_norm": 0.20041373372077942, + "learning_rate": 3.32992665823064e-05, + "loss": 0.006, + "step": 37680 + }, + { + "grad_norm": 0.1866205930709839, + "learning_rate": 3.327329399466774e-05, + "loss": 0.0053, + "step": 37690 + }, + { + "grad_norm": 0.16264699399471283, + "learning_rate": 3.324732648816072e-05, + "loss": 0.0035, + "step": 37700 + }, + { + "grad_norm": 0.15720997750759125, + "learning_rate": 3.322136407067358e-05, + "loss": 0.0042, + "step": 37710 + }, + { + "grad_norm": 0.11677990108728409, + "learning_rate": 3.3195406750093036e-05, + "loss": 0.0034, + "step": 37720 + }, + { + "grad_norm": 0.14262202382087708, + "learning_rate": 3.3169454534304205e-05, + "loss": 0.0032, + "step": 37730 + }, + { + "grad_norm": 0.17104850709438324, + "learning_rate": 3.3143507431190725e-05, + "loss": 0.0039, + "step": 37740 + }, + { + "grad_norm": 0.18489162623882294, + "learning_rate": 3.311756544863459e-05, + "loss": 0.004, + "step": 37750 + }, + { + "grad_norm": 0.17596763372421265, + "learning_rate": 3.309162859451633e-05, + "loss": 0.0036, + "step": 37760 + }, + { + "grad_norm": 0.1680329442024231, + "learning_rate": 3.306569687671487e-05, + "loss": 0.0032, + "step": 37770 + }, + { + "grad_norm": 0.14729122817516327, + "learning_rate": 3.303977030310756e-05, + "loss": 0.0049, + "step": 37780 + }, + { + "grad_norm": 0.15907712280750275, + "learning_rate": 3.3013848881570245e-05, + "loss": 0.0029, + "step": 37790 + }, + { + "grad_norm": 0.19403396546840668, + "learning_rate": 3.298793261997712e-05, + "loss": 0.0037, + "step": 37800 + }, + { + "grad_norm": 0.16667458415031433, + "learning_rate": 3.2962021526200893e-05, + "loss": 0.0044, + "step": 37810 + }, + { + "grad_norm": 0.17531436681747437, + "learning_rate": 3.293611560811268e-05, + "loss": 0.0042, + "step": 37820 + }, + { + "grad_norm": 0.14528824388980865, + "learning_rate": 3.291021487358199e-05, + "loss": 0.0028, + "step": 37830 + }, + { + "grad_norm": 0.178969144821167, + "learning_rate": 3.28843193304768e-05, + "loss": 0.0035, + "step": 37840 + }, + { + "grad_norm": 0.15281599760055542, + "learning_rate": 3.2858428986663456e-05, + "loss": 0.0044, + "step": 37850 + }, + { + "grad_norm": 0.14106035232543945, + "learning_rate": 3.283254385000681e-05, + "loss": 0.0037, + "step": 37860 + }, + { + "grad_norm": 0.13820761442184448, + "learning_rate": 3.2806663928370076e-05, + "loss": 0.0028, + "step": 37870 + }, + { + "grad_norm": 0.192808136343956, + "learning_rate": 3.278078922961485e-05, + "loss": 0.0034, + "step": 37880 + }, + { + "grad_norm": 0.14884616434574127, + "learning_rate": 3.275491976160123e-05, + "loss": 0.0034, + "step": 37890 + }, + { + "grad_norm": 0.11686035245656967, + "learning_rate": 3.2729055532187645e-05, + "loss": 0.0028, + "step": 37900 + }, + { + "grad_norm": 0.141291081905365, + "learning_rate": 3.270319654923097e-05, + "loss": 0.0032, + "step": 37910 + }, + { + "grad_norm": 0.18173593282699585, + "learning_rate": 3.2677342820586506e-05, + "loss": 0.0031, + "step": 37920 + }, + { + "grad_norm": 0.2224915623664856, + "learning_rate": 3.2651494354107905e-05, + "loss": 0.0036, + "step": 37930 + }, + { + "grad_norm": 0.14514945447444916, + "learning_rate": 3.2625651157647266e-05, + "loss": 0.0033, + "step": 37940 + }, + { + "grad_norm": 0.14295557141304016, + "learning_rate": 3.259981323905505e-05, + "loss": 0.0032, + "step": 37950 + }, + { + "grad_norm": 0.13157346844673157, + "learning_rate": 3.257398060618014e-05, + "loss": 0.0038, + "step": 37960 + }, + { + "grad_norm": 0.14938926696777344, + "learning_rate": 3.254815326686983e-05, + "loss": 0.0042, + "step": 37970 + }, + { + "grad_norm": 0.1744973063468933, + "learning_rate": 3.2522331228969774e-05, + "loss": 0.0038, + "step": 37980 + }, + { + "grad_norm": 0.08774067461490631, + "learning_rate": 3.2496514500324006e-05, + "loss": 0.0027, + "step": 37990 + }, + { + "grad_norm": 0.16428858041763306, + "learning_rate": 3.247070308877498e-05, + "loss": 0.0055, + "step": 38000 + }, + { + "grad_norm": 0.26228994131088257, + "learning_rate": 3.2444897002163515e-05, + "loss": 0.0039, + "step": 38010 + }, + { + "grad_norm": 0.16996653378009796, + "learning_rate": 3.241909624832885e-05, + "loss": 0.0034, + "step": 38020 + }, + { + "grad_norm": 0.1171463206410408, + "learning_rate": 3.239330083510852e-05, + "loss": 0.0026, + "step": 38030 + }, + { + "grad_norm": 0.1099950522184372, + "learning_rate": 3.236751077033855e-05, + "loss": 0.0032, + "step": 38040 + }, + { + "grad_norm": 0.15454281866550446, + "learning_rate": 3.234172606185322e-05, + "loss": 0.0043, + "step": 38050 + }, + { + "grad_norm": 0.1872277706861496, + "learning_rate": 3.231594671748528e-05, + "loss": 0.0034, + "step": 38060 + }, + { + "grad_norm": 0.13560956716537476, + "learning_rate": 3.2290172745065815e-05, + "loss": 0.0032, + "step": 38070 + }, + { + "grad_norm": 0.19233420491218567, + "learning_rate": 3.226440415242426e-05, + "loss": 0.0035, + "step": 38080 + }, + { + "grad_norm": 0.13916891813278198, + "learning_rate": 3.223864094738846e-05, + "loss": 0.0052, + "step": 38090 + }, + { + "grad_norm": 0.1733429878950119, + "learning_rate": 3.221288313778456e-05, + "loss": 0.0061, + "step": 38100 + }, + { + "grad_norm": 0.1555292010307312, + "learning_rate": 3.2187130731437125e-05, + "loss": 0.0037, + "step": 38110 + }, + { + "grad_norm": 0.15827955305576324, + "learning_rate": 3.216138373616905e-05, + "loss": 0.0038, + "step": 38120 + }, + { + "grad_norm": 0.17322027683258057, + "learning_rate": 3.21356421598016e-05, + "loss": 0.0058, + "step": 38130 + }, + { + "grad_norm": 0.1480608880519867, + "learning_rate": 3.210990601015438e-05, + "loss": 0.0027, + "step": 38140 + }, + { + "grad_norm": 0.11870847642421722, + "learning_rate": 3.208417529504535e-05, + "loss": 0.0027, + "step": 38150 + }, + { + "grad_norm": 0.1499631553888321, + "learning_rate": 3.205845002229084e-05, + "loss": 0.0036, + "step": 38160 + }, + { + "grad_norm": 0.17069754004478455, + "learning_rate": 3.203273019970547e-05, + "loss": 0.0053, + "step": 38170 + }, + { + "grad_norm": 0.16122272610664368, + "learning_rate": 3.200701583510227e-05, + "loss": 0.0029, + "step": 38180 + }, + { + "grad_norm": 0.14388661086559296, + "learning_rate": 3.198130693629261e-05, + "loss": 0.0041, + "step": 38190 + }, + { + "grad_norm": 0.18408207595348358, + "learning_rate": 3.195560351108612e-05, + "loss": 0.0037, + "step": 38200 + }, + { + "grad_norm": 0.16047623753547668, + "learning_rate": 3.1929905567290865e-05, + "loss": 0.0042, + "step": 38210 + }, + { + "grad_norm": 0.20117397606372833, + "learning_rate": 3.1904213112713164e-05, + "loss": 0.0042, + "step": 38220 + }, + { + "grad_norm": 0.21903356909751892, + "learning_rate": 3.187852615515774e-05, + "loss": 0.0047, + "step": 38230 + }, + { + "grad_norm": 0.12816491723060608, + "learning_rate": 3.1852844702427606e-05, + "loss": 0.0036, + "step": 38240 + }, + { + "grad_norm": 0.19357353448867798, + "learning_rate": 3.18271687623241e-05, + "loss": 0.0044, + "step": 38250 + }, + { + "grad_norm": 0.1963251680135727, + "learning_rate": 3.1801498342646896e-05, + "loss": 0.0039, + "step": 38260 + }, + { + "grad_norm": 0.17671528458595276, + "learning_rate": 3.177583345119398e-05, + "loss": 0.0039, + "step": 38270 + }, + { + "grad_norm": 0.17538240551948547, + "learning_rate": 3.17501740957617e-05, + "loss": 0.0038, + "step": 38280 + }, + { + "grad_norm": 0.1837548166513443, + "learning_rate": 3.172452028414467e-05, + "loss": 0.0038, + "step": 38290 + }, + { + "grad_norm": 0.17383086681365967, + "learning_rate": 3.169887202413583e-05, + "loss": 0.0035, + "step": 38300 + }, + { + "grad_norm": 0.16568560898303986, + "learning_rate": 3.167322932352646e-05, + "loss": 0.0045, + "step": 38310 + }, + { + "grad_norm": 0.20074434578418732, + "learning_rate": 3.164759219010613e-05, + "loss": 0.0037, + "step": 38320 + }, + { + "grad_norm": 0.14772246778011322, + "learning_rate": 3.1621960631662725e-05, + "loss": 0.005, + "step": 38330 + }, + { + "grad_norm": 0.21931256353855133, + "learning_rate": 3.159633465598245e-05, + "loss": 0.0069, + "step": 38340 + }, + { + "grad_norm": 0.21280939877033234, + "learning_rate": 3.1570714270849767e-05, + "loss": 0.0038, + "step": 38350 + }, + { + "grad_norm": 0.1555836945772171, + "learning_rate": 3.1545099484047516e-05, + "loss": 0.0032, + "step": 38360 + }, + { + "grad_norm": 0.22691906988620758, + "learning_rate": 3.151949030335674e-05, + "loss": 0.0042, + "step": 38370 + }, + { + "grad_norm": 0.18548434972763062, + "learning_rate": 3.149388673655687e-05, + "loss": 0.0038, + "step": 38380 + }, + { + "grad_norm": 0.17932584881782532, + "learning_rate": 3.146828879142559e-05, + "loss": 0.0034, + "step": 38390 + }, + { + "grad_norm": 0.13206318020820618, + "learning_rate": 3.1442696475738866e-05, + "loss": 0.0032, + "step": 38400 + }, + { + "grad_norm": 0.19317376613616943, + "learning_rate": 3.141710979727098e-05, + "loss": 0.0055, + "step": 38410 + }, + { + "grad_norm": 0.17274326086044312, + "learning_rate": 3.139152876379447e-05, + "loss": 0.0047, + "step": 38420 + }, + { + "grad_norm": 0.14006538689136505, + "learning_rate": 3.1365953383080214e-05, + "loss": 0.0042, + "step": 38430 + }, + { + "grad_norm": 0.20923461019992828, + "learning_rate": 3.134038366289731e-05, + "loss": 0.0033, + "step": 38440 + }, + { + "grad_norm": 0.1356029212474823, + "learning_rate": 3.131481961101317e-05, + "loss": 0.0035, + "step": 38450 + }, + { + "grad_norm": 0.1677478551864624, + "learning_rate": 3.128926123519349e-05, + "loss": 0.0051, + "step": 38460 + }, + { + "grad_norm": 0.16832514107227325, + "learning_rate": 3.1263708543202194e-05, + "loss": 0.0042, + "step": 38470 + }, + { + "grad_norm": 0.15600119531154633, + "learning_rate": 3.123816154280155e-05, + "loss": 0.0039, + "step": 38480 + }, + { + "grad_norm": 0.19982333481311798, + "learning_rate": 3.121262024175207e-05, + "loss": 0.004, + "step": 38490 + }, + { + "grad_norm": 0.14756712317466736, + "learning_rate": 3.118708464781248e-05, + "loss": 0.0057, + "step": 38500 + }, + { + "grad_norm": 0.16085180640220642, + "learning_rate": 3.116155476873987e-05, + "loss": 0.0051, + "step": 38510 + }, + { + "grad_norm": 0.13968053460121155, + "learning_rate": 3.11360306122895e-05, + "loss": 0.0032, + "step": 38520 + }, + { + "grad_norm": 0.13419944047927856, + "learning_rate": 3.1110512186214975e-05, + "loss": 0.0033, + "step": 38530 + }, + { + "grad_norm": 0.18088099360466003, + "learning_rate": 3.1084999498268095e-05, + "loss": 0.0049, + "step": 38540 + }, + { + "grad_norm": 0.15623392164707184, + "learning_rate": 3.1059492556198934e-05, + "loss": 0.0056, + "step": 38550 + }, + { + "grad_norm": 0.13471689820289612, + "learning_rate": 3.103399136775586e-05, + "loss": 0.0033, + "step": 38560 + }, + { + "grad_norm": 0.16237109899520874, + "learning_rate": 3.100849594068541e-05, + "loss": 0.0037, + "step": 38570 + }, + { + "grad_norm": 0.16273000836372375, + "learning_rate": 3.0983006282732484e-05, + "loss": 0.0036, + "step": 38580 + }, + { + "grad_norm": 0.1692211925983429, + "learning_rate": 3.0957522401640116e-05, + "loss": 0.0045, + "step": 38590 + }, + { + "grad_norm": 0.1641266644001007, + "learning_rate": 3.0932044305149645e-05, + "loss": 0.0037, + "step": 38600 + }, + { + "grad_norm": 0.1974402368068695, + "learning_rate": 3.090657200100068e-05, + "loss": 0.0041, + "step": 38610 + }, + { + "grad_norm": 0.18272998929023743, + "learning_rate": 3.088110549693099e-05, + "loss": 0.0058, + "step": 38620 + }, + { + "grad_norm": 0.13980145752429962, + "learning_rate": 3.085564480067667e-05, + "loss": 0.0038, + "step": 38630 + }, + { + "grad_norm": 0.15106166899204254, + "learning_rate": 3.0830189919971955e-05, + "loss": 0.0037, + "step": 38640 + }, + { + "grad_norm": 0.14273133873939514, + "learning_rate": 3.080474086254939e-05, + "loss": 0.0039, + "step": 38650 + }, + { + "grad_norm": 0.2325126677751541, + "learning_rate": 3.077929763613975e-05, + "loss": 0.004, + "step": 38660 + }, + { + "grad_norm": 0.16288131475448608, + "learning_rate": 3.075386024847198e-05, + "loss": 0.0035, + "step": 38670 + }, + { + "grad_norm": 0.17242486774921417, + "learning_rate": 3.072842870727331e-05, + "loss": 0.0042, + "step": 38680 + }, + { + "grad_norm": 0.16674727201461792, + "learning_rate": 3.070300302026916e-05, + "loss": 0.004, + "step": 38690 + }, + { + "grad_norm": 0.16508355736732483, + "learning_rate": 3.067758319518318e-05, + "loss": 0.0032, + "step": 38700 + }, + { + "grad_norm": 0.18067985773086548, + "learning_rate": 3.065216923973725e-05, + "loss": 0.0037, + "step": 38710 + }, + { + "grad_norm": 0.138757586479187, + "learning_rate": 3.062676116165145e-05, + "loss": 0.0042, + "step": 38720 + }, + { + "grad_norm": 0.1765752136707306, + "learning_rate": 3.06013589686441e-05, + "loss": 0.0048, + "step": 38730 + }, + { + "grad_norm": 0.13081985712051392, + "learning_rate": 3.05759626684317e-05, + "loss": 0.004, + "step": 38740 + }, + { + "grad_norm": 0.15878133475780487, + "learning_rate": 3.055057226872896e-05, + "loss": 0.0054, + "step": 38750 + }, + { + "grad_norm": 0.19448181986808777, + "learning_rate": 3.052518777724887e-05, + "loss": 0.0033, + "step": 38760 + }, + { + "grad_norm": 0.13335123658180237, + "learning_rate": 3.04998092017025e-05, + "loss": 0.0055, + "step": 38770 + }, + { + "grad_norm": 0.14835438132286072, + "learning_rate": 3.0474436549799246e-05, + "loss": 0.0036, + "step": 38780 + }, + { + "grad_norm": 0.14918258786201477, + "learning_rate": 3.044906982924661e-05, + "loss": 0.0035, + "step": 38790 + }, + { + "grad_norm": 0.11435020714998245, + "learning_rate": 3.0423709047750337e-05, + "loss": 0.0035, + "step": 38800 + }, + { + "grad_norm": 0.16265302896499634, + "learning_rate": 3.03983542130144e-05, + "loss": 0.0044, + "step": 38810 + }, + { + "grad_norm": 0.17572930455207825, + "learning_rate": 3.0373005332740877e-05, + "loss": 0.004, + "step": 38820 + }, + { + "grad_norm": 0.20133958756923676, + "learning_rate": 3.034766241463013e-05, + "loss": 0.0041, + "step": 38830 + }, + { + "grad_norm": 0.12758953869342804, + "learning_rate": 3.032232546638064e-05, + "loss": 0.0039, + "step": 38840 + }, + { + "grad_norm": 0.147571861743927, + "learning_rate": 3.0296994495689114e-05, + "loss": 0.0035, + "step": 38850 + }, + { + "grad_norm": 0.16126450896263123, + "learning_rate": 3.0271669510250444e-05, + "loss": 0.0067, + "step": 38860 + }, + { + "grad_norm": 0.2174222469329834, + "learning_rate": 3.024635051775766e-05, + "loss": 0.0037, + "step": 38870 + }, + { + "grad_norm": 0.22608225047588348, + "learning_rate": 3.022103752590205e-05, + "loss": 0.0041, + "step": 38880 + }, + { + "grad_norm": 0.13144844770431519, + "learning_rate": 3.0195730542372992e-05, + "loss": 0.004, + "step": 38890 + }, + { + "grad_norm": 0.14710941910743713, + "learning_rate": 3.0170429574858084e-05, + "loss": 0.0041, + "step": 38900 + }, + { + "grad_norm": 0.15190009772777557, + "learning_rate": 3.0145134631043127e-05, + "loss": 0.004, + "step": 38910 + }, + { + "grad_norm": 0.14443771541118622, + "learning_rate": 3.0119845718612018e-05, + "loss": 0.0037, + "step": 38920 + }, + { + "grad_norm": 0.14645619690418243, + "learning_rate": 3.009456284524688e-05, + "loss": 0.0027, + "step": 38930 + }, + { + "grad_norm": 0.12203684449195862, + "learning_rate": 3.0069286018627967e-05, + "loss": 0.0037, + "step": 38940 + }, + { + "grad_norm": 0.16224557161331177, + "learning_rate": 3.0044015246433743e-05, + "loss": 0.0059, + "step": 38950 + }, + { + "grad_norm": 0.21660684049129486, + "learning_rate": 3.0018750536340755e-05, + "loss": 0.0033, + "step": 38960 + }, + { + "grad_norm": 0.1786169558763504, + "learning_rate": 2.999349189602378e-05, + "loss": 0.0037, + "step": 38970 + }, + { + "grad_norm": 0.206757053732872, + "learning_rate": 2.9968239333155733e-05, + "loss": 0.0037, + "step": 38980 + }, + { + "grad_norm": 0.1491784304380417, + "learning_rate": 2.994299285540767e-05, + "loss": 0.0049, + "step": 38990 + }, + { + "grad_norm": 0.1379040628671646, + "learning_rate": 2.9917752470448813e-05, + "loss": 0.0037, + "step": 39000 + }, + { + "grad_norm": 0.1490938812494278, + "learning_rate": 2.9892518185946495e-05, + "loss": 0.0049, + "step": 39010 + }, + { + "grad_norm": 0.1295795440673828, + "learning_rate": 2.986729000956624e-05, + "loss": 0.0033, + "step": 39020 + }, + { + "grad_norm": 0.10465283691883087, + "learning_rate": 2.9842067948971736e-05, + "loss": 0.0038, + "step": 39030 + }, + { + "grad_norm": 0.20977923274040222, + "learning_rate": 2.9816852011824727e-05, + "loss": 0.0047, + "step": 39040 + }, + { + "grad_norm": 0.18177370727062225, + "learning_rate": 2.979164220578519e-05, + "loss": 0.0026, + "step": 39050 + }, + { + "grad_norm": 0.11753685027360916, + "learning_rate": 2.9766438538511165e-05, + "loss": 0.0044, + "step": 39060 + }, + { + "grad_norm": 0.1519283503293991, + "learning_rate": 2.9741241017658873e-05, + "loss": 0.0042, + "step": 39070 + }, + { + "grad_norm": 0.16709767282009125, + "learning_rate": 2.971604965088267e-05, + "loss": 0.003, + "step": 39080 + }, + { + "grad_norm": 0.25406086444854736, + "learning_rate": 2.9690864445835008e-05, + "loss": 0.0041, + "step": 39090 + }, + { + "grad_norm": 0.18339000642299652, + "learning_rate": 2.966568541016651e-05, + "loss": 0.0044, + "step": 39100 + }, + { + "grad_norm": 0.16771723330020905, + "learning_rate": 2.9640512551525867e-05, + "loss": 0.003, + "step": 39110 + }, + { + "grad_norm": 0.17842097580432892, + "learning_rate": 2.961534587755995e-05, + "loss": 0.0057, + "step": 39120 + }, + { + "grad_norm": 0.14806637167930603, + "learning_rate": 2.959018539591375e-05, + "loss": 0.0046, + "step": 39130 + }, + { + "grad_norm": 0.2111293375492096, + "learning_rate": 2.9565031114230325e-05, + "loss": 0.0042, + "step": 39140 + }, + { + "grad_norm": 0.12681937217712402, + "learning_rate": 2.9539883040150895e-05, + "loss": 0.0027, + "step": 39150 + }, + { + "grad_norm": 0.13751573860645294, + "learning_rate": 2.9514741181314774e-05, + "loss": 0.0031, + "step": 39160 + }, + { + "grad_norm": 0.13514862954616547, + "learning_rate": 2.94896055453594e-05, + "loss": 0.0028, + "step": 39170 + }, + { + "grad_norm": 0.15924686193466187, + "learning_rate": 2.9464476139920332e-05, + "loss": 0.0026, + "step": 39180 + }, + { + "grad_norm": 0.1754576563835144, + "learning_rate": 2.9439352972631186e-05, + "loss": 0.0044, + "step": 39190 + }, + { + "grad_norm": 0.17615650594234467, + "learning_rate": 2.9414236051123757e-05, + "loss": 0.0031, + "step": 39200 + }, + { + "grad_norm": 0.1596420705318451, + "learning_rate": 2.938912538302785e-05, + "loss": 0.0037, + "step": 39210 + }, + { + "grad_norm": 0.18859560787677765, + "learning_rate": 2.9364020975971464e-05, + "loss": 0.0037, + "step": 39220 + }, + { + "grad_norm": 0.14963360130786896, + "learning_rate": 2.9338922837580657e-05, + "loss": 0.0033, + "step": 39230 + }, + { + "grad_norm": 0.25426939129829407, + "learning_rate": 2.931383097547955e-05, + "loss": 0.0063, + "step": 39240 + }, + { + "grad_norm": 0.19684921205043793, + "learning_rate": 2.928874539729043e-05, + "loss": 0.0029, + "step": 39250 + }, + { + "grad_norm": 0.15413810312747955, + "learning_rate": 2.926366611063358e-05, + "loss": 0.0031, + "step": 39260 + }, + { + "grad_norm": 0.1927766352891922, + "learning_rate": 2.9238593123127463e-05, + "loss": 0.0028, + "step": 39270 + }, + { + "grad_norm": 0.17459021508693695, + "learning_rate": 2.9213526442388583e-05, + "loss": 0.0062, + "step": 39280 + }, + { + "grad_norm": 0.18897677958011627, + "learning_rate": 2.9188466076031545e-05, + "loss": 0.0058, + "step": 39290 + }, + { + "grad_norm": 0.2039947509765625, + "learning_rate": 2.9163412031669012e-05, + "loss": 0.0036, + "step": 39300 + }, + { + "grad_norm": 0.21787191927433014, + "learning_rate": 2.913836431691175e-05, + "loss": 0.0044, + "step": 39310 + }, + { + "grad_norm": 0.2503415048122406, + "learning_rate": 2.9113322939368583e-05, + "loss": 0.004, + "step": 39320 + }, + { + "grad_norm": 0.1635635793209076, + "learning_rate": 2.9088287906646427e-05, + "loss": 0.0055, + "step": 39330 + }, + { + "grad_norm": 0.143579363822937, + "learning_rate": 2.906325922635024e-05, + "loss": 0.0028, + "step": 39340 + }, + { + "grad_norm": 0.15177857875823975, + "learning_rate": 2.903823690608313e-05, + "loss": 0.0043, + "step": 39350 + }, + { + "grad_norm": 0.21038132905960083, + "learning_rate": 2.9013220953446174e-05, + "loss": 0.0043, + "step": 39360 + }, + { + "grad_norm": 0.14685139060020447, + "learning_rate": 2.8988211376038564e-05, + "loss": 0.0032, + "step": 39370 + }, + { + "grad_norm": 0.1463918536901474, + "learning_rate": 2.8963208181457564e-05, + "loss": 0.0035, + "step": 39380 + }, + { + "grad_norm": 0.1303980052471161, + "learning_rate": 2.8938211377298453e-05, + "loss": 0.0029, + "step": 39390 + }, + { + "grad_norm": 0.15008729696273804, + "learning_rate": 2.8913220971154652e-05, + "loss": 0.0029, + "step": 39400 + }, + { + "grad_norm": 0.17849008738994598, + "learning_rate": 2.888823697061753e-05, + "loss": 0.0031, + "step": 39410 + }, + { + "grad_norm": 0.16182662546634674, + "learning_rate": 2.8863259383276618e-05, + "loss": 0.0034, + "step": 39420 + }, + { + "grad_norm": 0.19884468615055084, + "learning_rate": 2.8838288216719395e-05, + "loss": 0.0044, + "step": 39430 + }, + { + "grad_norm": 0.1171545535326004, + "learning_rate": 2.8813323478531484e-05, + "loss": 0.0051, + "step": 39440 + }, + { + "grad_norm": 0.15900716185569763, + "learning_rate": 2.8788365176296496e-05, + "loss": 0.0041, + "step": 39450 + }, + { + "grad_norm": 0.18207363784313202, + "learning_rate": 2.876341331759611e-05, + "loss": 0.003, + "step": 39460 + }, + { + "grad_norm": 0.1453072428703308, + "learning_rate": 2.8738467910010036e-05, + "loss": 0.0032, + "step": 39470 + }, + { + "grad_norm": 0.09435635805130005, + "learning_rate": 2.8713528961116032e-05, + "loss": 0.0029, + "step": 39480 + }, + { + "grad_norm": 0.1797606647014618, + "learning_rate": 2.8688596478489875e-05, + "loss": 0.0038, + "step": 39490 + }, + { + "grad_norm": 0.14052733778953552, + "learning_rate": 2.8663670469705434e-05, + "loss": 0.0029, + "step": 39500 + }, + { + "grad_norm": 0.20392489433288574, + "learning_rate": 2.8638750942334546e-05, + "loss": 0.0066, + "step": 39510 + }, + { + "grad_norm": 0.16696028411388397, + "learning_rate": 2.8613837903947115e-05, + "loss": 0.0044, + "step": 39520 + }, + { + "grad_norm": 0.12981997430324554, + "learning_rate": 2.858893136211106e-05, + "loss": 0.0028, + "step": 39530 + }, + { + "grad_norm": 0.15271884202957153, + "learning_rate": 2.8564031324392315e-05, + "loss": 0.0033, + "step": 39540 + }, + { + "grad_norm": 0.12272675335407257, + "learning_rate": 2.85391377983549e-05, + "loss": 0.0031, + "step": 39550 + }, + { + "grad_norm": 0.11483343690633774, + "learning_rate": 2.851425079156075e-05, + "loss": 0.003, + "step": 39560 + }, + { + "grad_norm": 0.13703300058841705, + "learning_rate": 2.848937031156994e-05, + "loss": 0.0031, + "step": 39570 + }, + { + "grad_norm": 0.15563447773456573, + "learning_rate": 2.846449636594044e-05, + "loss": 0.0046, + "step": 39580 + }, + { + "grad_norm": 0.16307058930397034, + "learning_rate": 2.843962896222836e-05, + "loss": 0.0047, + "step": 39590 + }, + { + "grad_norm": 0.2005375623703003, + "learning_rate": 2.8414768107987722e-05, + "loss": 0.0035, + "step": 39600 + }, + { + "grad_norm": 0.166842520236969, + "learning_rate": 2.838991381077061e-05, + "loss": 0.0047, + "step": 39610 + }, + { + "grad_norm": 0.15242613852024078, + "learning_rate": 2.83650660781271e-05, + "loss": 0.003, + "step": 39620 + }, + { + "grad_norm": 0.11560218781232834, + "learning_rate": 2.8340224917605285e-05, + "loss": 0.0035, + "step": 39630 + }, + { + "grad_norm": 0.1581832319498062, + "learning_rate": 2.831539033675122e-05, + "loss": 0.0033, + "step": 39640 + }, + { + "grad_norm": 0.12408707290887833, + "learning_rate": 2.8290562343109038e-05, + "loss": 0.0025, + "step": 39650 + }, + { + "grad_norm": 0.17381906509399414, + "learning_rate": 2.826574094422082e-05, + "loss": 0.0034, + "step": 39660 + }, + { + "grad_norm": 0.17554503679275513, + "learning_rate": 2.8240926147626645e-05, + "loss": 0.0037, + "step": 39670 + }, + { + "grad_norm": 0.14596165716648102, + "learning_rate": 2.8216117960864586e-05, + "loss": 0.0035, + "step": 39680 + }, + { + "grad_norm": 0.15940487384796143, + "learning_rate": 2.8191316391470703e-05, + "loss": 0.0032, + "step": 39690 + }, + { + "grad_norm": 0.12426194548606873, + "learning_rate": 2.816652144697911e-05, + "loss": 0.0039, + "step": 39700 + }, + { + "grad_norm": 0.24336692690849304, + "learning_rate": 2.8141733134921783e-05, + "loss": 0.005, + "step": 39710 + }, + { + "grad_norm": 0.1461370289325714, + "learning_rate": 2.811695146282884e-05, + "loss": 0.0032, + "step": 39720 + }, + { + "grad_norm": 0.1991419494152069, + "learning_rate": 2.8092176438228212e-05, + "loss": 0.0041, + "step": 39730 + }, + { + "grad_norm": 0.18520689010620117, + "learning_rate": 2.806740806864598e-05, + "loss": 0.007, + "step": 39740 + }, + { + "grad_norm": 0.13291534781455994, + "learning_rate": 2.804264636160604e-05, + "loss": 0.0028, + "step": 39750 + }, + { + "grad_norm": 0.14441566169261932, + "learning_rate": 2.8017891324630402e-05, + "loss": 0.0037, + "step": 39760 + }, + { + "grad_norm": 0.2251356840133667, + "learning_rate": 2.7993142965238976e-05, + "loss": 0.0033, + "step": 39770 + }, + { + "grad_norm": 0.14895671606063843, + "learning_rate": 2.7968401290949665e-05, + "loss": 0.0036, + "step": 39780 + }, + { + "grad_norm": 0.1317809373140335, + "learning_rate": 2.7943666309278328e-05, + "loss": 0.0051, + "step": 39790 + }, + { + "grad_norm": 0.1881517618894577, + "learning_rate": 2.7918938027738783e-05, + "loss": 0.0039, + "step": 39800 + }, + { + "grad_norm": 0.1234879419207573, + "learning_rate": 2.789421645384287e-05, + "loss": 0.0037, + "step": 39810 + }, + { + "grad_norm": 0.17134664952754974, + "learning_rate": 2.786950159510032e-05, + "loss": 0.0043, + "step": 39820 + }, + { + "grad_norm": 0.14779159426689148, + "learning_rate": 2.7844793459018876e-05, + "loss": 0.0033, + "step": 39830 + }, + { + "grad_norm": 0.1418488472700119, + "learning_rate": 2.7820092053104195e-05, + "loss": 0.0034, + "step": 39840 + }, + { + "grad_norm": 0.18395917117595673, + "learning_rate": 2.7795397384859933e-05, + "loss": 0.0046, + "step": 39850 + }, + { + "grad_norm": 0.20731322467327118, + "learning_rate": 2.7770709461787638e-05, + "loss": 0.0037, + "step": 39860 + }, + { + "grad_norm": 0.154915913939476, + "learning_rate": 2.7746028291386915e-05, + "loss": 0.0033, + "step": 39870 + }, + { + "grad_norm": 0.1507592797279358, + "learning_rate": 2.772135388115519e-05, + "loss": 0.0034, + "step": 39880 + }, + { + "grad_norm": 0.23889128863811493, + "learning_rate": 2.7696686238587945e-05, + "loss": 0.0049, + "step": 39890 + }, + { + "grad_norm": 0.15484386682510376, + "learning_rate": 2.7672025371178505e-05, + "loss": 0.0049, + "step": 39900 + }, + { + "grad_norm": 0.15138769149780273, + "learning_rate": 2.7647371286418238e-05, + "loss": 0.0036, + "step": 39910 + }, + { + "grad_norm": 0.12590371072292328, + "learning_rate": 2.762272399179639e-05, + "loss": 0.004, + "step": 39920 + }, + { + "grad_norm": 0.1896025687456131, + "learning_rate": 2.7598083494800154e-05, + "loss": 0.0051, + "step": 39930 + }, + { + "grad_norm": 0.19800323247909546, + "learning_rate": 2.7573449802914664e-05, + "loss": 0.0031, + "step": 39940 + }, + { + "grad_norm": 0.13694854080677032, + "learning_rate": 2.7548822923622964e-05, + "loss": 0.0029, + "step": 39950 + }, + { + "grad_norm": 0.1395803689956665, + "learning_rate": 2.752420286440609e-05, + "loss": 0.004, + "step": 39960 + }, + { + "grad_norm": 0.1329280436038971, + "learning_rate": 2.749958963274295e-05, + "loss": 0.0026, + "step": 39970 + }, + { + "grad_norm": 0.10489252954721451, + "learning_rate": 2.747498323611039e-05, + "loss": 0.0036, + "step": 39980 + }, + { + "grad_norm": 0.11536134034395218, + "learning_rate": 2.7450383681983184e-05, + "loss": 0.0027, + "step": 39990 + }, + { + "grad_norm": 0.130319744348526, + "learning_rate": 2.742579097783403e-05, + "loss": 0.0027, + "step": 40000 + }, + { + "grad_norm": 0.16949599981307983, + "learning_rate": 2.7401205131133512e-05, + "loss": 0.0038, + "step": 40010 + }, + { + "grad_norm": 0.19407635927200317, + "learning_rate": 2.7376626149350238e-05, + "loss": 0.0049, + "step": 40020 + }, + { + "grad_norm": 0.12247345596551895, + "learning_rate": 2.735205403995056e-05, + "loss": 0.0037, + "step": 40030 + }, + { + "grad_norm": 0.141652911901474, + "learning_rate": 2.7327488810398917e-05, + "loss": 0.0032, + "step": 40040 + }, + { + "grad_norm": 0.17005136609077454, + "learning_rate": 2.7302930468157507e-05, + "loss": 0.0048, + "step": 40050 + }, + { + "grad_norm": 0.17325459420681, + "learning_rate": 2.727837902068655e-05, + "loss": 0.0035, + "step": 40060 + }, + { + "grad_norm": 0.12086674571037292, + "learning_rate": 2.7253834475444123e-05, + "loss": 0.0031, + "step": 40070 + }, + { + "grad_norm": 0.14297521114349365, + "learning_rate": 2.7229296839886204e-05, + "loss": 0.0034, + "step": 40080 + }, + { + "grad_norm": 0.16402524709701538, + "learning_rate": 2.720476612146668e-05, + "loss": 0.0045, + "step": 40090 + }, + { + "grad_norm": 0.13627465069293976, + "learning_rate": 2.7180242327637317e-05, + "loss": 0.0049, + "step": 40100 + }, + { + "grad_norm": 0.17370890080928802, + "learning_rate": 2.7155725465847826e-05, + "loss": 0.0036, + "step": 40110 + }, + { + "grad_norm": 0.11252837628126144, + "learning_rate": 2.713121554354578e-05, + "loss": 0.0042, + "step": 40120 + }, + { + "grad_norm": 0.0983961671590805, + "learning_rate": 2.7106712568176628e-05, + "loss": 0.0062, + "step": 40130 + }, + { + "grad_norm": 0.16455169022083282, + "learning_rate": 2.708221654718374e-05, + "loss": 0.0058, + "step": 40140 + }, + { + "grad_norm": 0.1063307672739029, + "learning_rate": 2.7057727488008357e-05, + "loss": 0.0029, + "step": 40150 + }, + { + "grad_norm": 0.13380520045757294, + "learning_rate": 2.703324539808961e-05, + "loss": 0.0035, + "step": 40160 + }, + { + "grad_norm": 0.23055113852024078, + "learning_rate": 2.7008770284864505e-05, + "loss": 0.0047, + "step": 40170 + }, + { + "grad_norm": 0.21245726943016052, + "learning_rate": 2.6984302155767916e-05, + "loss": 0.0033, + "step": 40180 + }, + { + "grad_norm": 0.14027787744998932, + "learning_rate": 2.6959841018232683e-05, + "loss": 0.0035, + "step": 40190 + }, + { + "grad_norm": 0.18057382106781006, + "learning_rate": 2.693538687968937e-05, + "loss": 0.0042, + "step": 40200 + }, + { + "grad_norm": 0.13222627341747284, + "learning_rate": 2.6910939747566556e-05, + "loss": 0.0031, + "step": 40210 + }, + { + "grad_norm": 0.1441546380519867, + "learning_rate": 2.6886499629290607e-05, + "loss": 0.0034, + "step": 40220 + }, + { + "grad_norm": 0.15969042479991913, + "learning_rate": 2.6862066532285802e-05, + "loss": 0.0038, + "step": 40230 + }, + { + "grad_norm": 0.11195027083158493, + "learning_rate": 2.6837640463974262e-05, + "loss": 0.0034, + "step": 40240 + }, + { + "grad_norm": 0.09255232661962509, + "learning_rate": 2.681322143177596e-05, + "loss": 0.0021, + "step": 40250 + }, + { + "grad_norm": 0.16341856122016907, + "learning_rate": 2.678880944310882e-05, + "loss": 0.0038, + "step": 40260 + }, + { + "grad_norm": 0.10179568827152252, + "learning_rate": 2.6764404505388474e-05, + "loss": 0.0023, + "step": 40270 + }, + { + "grad_norm": 0.16367743909358978, + "learning_rate": 2.6740006626028558e-05, + "loss": 0.0047, + "step": 40280 + }, + { + "grad_norm": 0.2046918272972107, + "learning_rate": 2.671561581244048e-05, + "loss": 0.0027, + "step": 40290 + }, + { + "grad_norm": 0.20121406018733978, + "learning_rate": 2.6691232072033536e-05, + "loss": 0.003, + "step": 40300 + }, + { + "grad_norm": 0.23787392675876617, + "learning_rate": 2.6666855412214852e-05, + "loss": 0.0042, + "step": 40310 + }, + { + "grad_norm": 0.17558088898658752, + "learning_rate": 2.664248584038942e-05, + "loss": 0.0048, + "step": 40320 + }, + { + "grad_norm": 0.17522291839122772, + "learning_rate": 2.6618123363960047e-05, + "loss": 0.0052, + "step": 40330 + }, + { + "grad_norm": 0.18726029992103577, + "learning_rate": 2.659376799032748e-05, + "loss": 0.0032, + "step": 40340 + }, + { + "grad_norm": 0.1825515180826187, + "learning_rate": 2.6569419726890145e-05, + "loss": 0.0029, + "step": 40350 + }, + { + "grad_norm": 0.14483273029327393, + "learning_rate": 2.654507858104447e-05, + "loss": 0.0062, + "step": 40360 + }, + { + "grad_norm": 0.18235516548156738, + "learning_rate": 2.652074456018463e-05, + "loss": 0.0035, + "step": 40370 + }, + { + "grad_norm": 0.12875507771968842, + "learning_rate": 2.6496417671702646e-05, + "loss": 0.0028, + "step": 40380 + }, + { + "grad_norm": 0.12988851964473724, + "learning_rate": 2.6472097922988427e-05, + "loss": 0.0027, + "step": 40390 + }, + { + "grad_norm": 0.15920481085777283, + "learning_rate": 2.6447785321429607e-05, + "loss": 0.0047, + "step": 40400 + }, + { + "grad_norm": 0.18997079133987427, + "learning_rate": 2.6423479874411784e-05, + "loss": 0.0032, + "step": 40410 + }, + { + "grad_norm": 0.16491113603115082, + "learning_rate": 2.6399181589318234e-05, + "loss": 0.0028, + "step": 40420 + }, + { + "grad_norm": 0.20085182785987854, + "learning_rate": 2.6374890473530188e-05, + "loss": 0.0037, + "step": 40430 + }, + { + "grad_norm": 0.21777567267417908, + "learning_rate": 2.635060653442664e-05, + "loss": 0.0053, + "step": 40440 + }, + { + "grad_norm": 0.12248072773218155, + "learning_rate": 2.6326329779384395e-05, + "loss": 0.0037, + "step": 40450 + }, + { + "grad_norm": 0.14673012495040894, + "learning_rate": 2.63020602157781e-05, + "loss": 0.0056, + "step": 40460 + }, + { + "grad_norm": 0.13401389122009277, + "learning_rate": 2.62777978509802e-05, + "loss": 0.0029, + "step": 40470 + }, + { + "grad_norm": 0.18501581251621246, + "learning_rate": 2.6253542692360954e-05, + "loss": 0.0036, + "step": 40480 + }, + { + "grad_norm": 0.12535792589187622, + "learning_rate": 2.6229294747288458e-05, + "loss": 0.0028, + "step": 40490 + }, + { + "grad_norm": 0.11880595237016678, + "learning_rate": 2.6205054023128596e-05, + "loss": 0.0028, + "step": 40500 + }, + { + "grad_norm": 0.17363210022449493, + "learning_rate": 2.6180820527245043e-05, + "loss": 0.0036, + "step": 40510 + }, + { + "grad_norm": 0.13007038831710815, + "learning_rate": 2.6156594266999313e-05, + "loss": 0.0033, + "step": 40520 + }, + { + "grad_norm": 0.12292372435331345, + "learning_rate": 2.6132375249750672e-05, + "loss": 0.0037, + "step": 40530 + }, + { + "grad_norm": 0.13666892051696777, + "learning_rate": 2.6108163482856286e-05, + "loss": 0.0075, + "step": 40540 + }, + { + "grad_norm": 0.15705054998397827, + "learning_rate": 2.6083958973670964e-05, + "loss": 0.005, + "step": 40550 + }, + { + "grad_norm": 0.1749950349330902, + "learning_rate": 2.6059761729547483e-05, + "loss": 0.0044, + "step": 40560 + }, + { + "grad_norm": 0.1375347375869751, + "learning_rate": 2.603557175783624e-05, + "loss": 0.0027, + "step": 40570 + }, + { + "grad_norm": 0.1371878683567047, + "learning_rate": 2.601138906588559e-05, + "loss": 0.0031, + "step": 40580 + }, + { + "grad_norm": 0.1526304930448532, + "learning_rate": 2.598721366104152e-05, + "loss": 0.0025, + "step": 40590 + }, + { + "grad_norm": 0.11683519184589386, + "learning_rate": 2.5963045550647945e-05, + "loss": 0.0039, + "step": 40600 + }, + { + "grad_norm": 0.15001489222049713, + "learning_rate": 2.5938884742046466e-05, + "loss": 0.0031, + "step": 40610 + }, + { + "grad_norm": 0.13099738955497742, + "learning_rate": 2.5914731242576507e-05, + "loss": 0.0023, + "step": 40620 + }, + { + "grad_norm": 0.15278936922550201, + "learning_rate": 2.5890585059575268e-05, + "loss": 0.003, + "step": 40630 + }, + { + "grad_norm": 0.12717005610466003, + "learning_rate": 2.5866446200377688e-05, + "loss": 0.003, + "step": 40640 + }, + { + "grad_norm": 0.15024015307426453, + "learning_rate": 2.5842314672316566e-05, + "loss": 0.0027, + "step": 40650 + }, + { + "grad_norm": 0.11710398644208908, + "learning_rate": 2.581819048272239e-05, + "loss": 0.0029, + "step": 40660 + }, + { + "grad_norm": 0.10066734254360199, + "learning_rate": 2.5794073638923478e-05, + "loss": 0.0024, + "step": 40670 + }, + { + "grad_norm": 0.13636569678783417, + "learning_rate": 2.576996414824586e-05, + "loss": 0.0028, + "step": 40680 + }, + { + "grad_norm": 0.15097136795520782, + "learning_rate": 2.574586201801339e-05, + "loss": 0.0055, + "step": 40690 + }, + { + "grad_norm": 0.10483056306838989, + "learning_rate": 2.572176725554762e-05, + "loss": 0.0025, + "step": 40700 + }, + { + "grad_norm": 0.19350281357765198, + "learning_rate": 2.5697679868167966e-05, + "loss": 0.0031, + "step": 40710 + }, + { + "grad_norm": 0.16532692313194275, + "learning_rate": 2.5673599863191468e-05, + "loss": 0.0045, + "step": 40720 + }, + { + "grad_norm": 0.11395798623561859, + "learning_rate": 2.564952724793306e-05, + "loss": 0.0028, + "step": 40730 + }, + { + "grad_norm": 0.16589069366455078, + "learning_rate": 2.5625462029705306e-05, + "loss": 0.0042, + "step": 40740 + }, + { + "grad_norm": 0.12468626350164413, + "learning_rate": 2.5601404215818624e-05, + "loss": 0.003, + "step": 40750 + }, + { + "grad_norm": 0.15399008989334106, + "learning_rate": 2.5577353813581144e-05, + "loss": 0.0033, + "step": 40760 + }, + { + "grad_norm": 0.10989420115947723, + "learning_rate": 2.5553310830298733e-05, + "loss": 0.0021, + "step": 40770 + }, + { + "grad_norm": 0.1566045880317688, + "learning_rate": 2.5529275273275012e-05, + "loss": 0.0034, + "step": 40780 + }, + { + "grad_norm": 0.15581819415092468, + "learning_rate": 2.550524714981133e-05, + "loss": 0.0038, + "step": 40790 + }, + { + "grad_norm": 0.23047573864459991, + "learning_rate": 2.5481226467206837e-05, + "loss": 0.0047, + "step": 40800 + }, + { + "grad_norm": 0.18920201063156128, + "learning_rate": 2.5457213232758365e-05, + "loss": 0.0039, + "step": 40810 + }, + { + "grad_norm": 0.14982692897319794, + "learning_rate": 2.5433207453760498e-05, + "loss": 0.0034, + "step": 40820 + }, + { + "grad_norm": 0.17577974498271942, + "learning_rate": 2.5409209137505552e-05, + "loss": 0.0031, + "step": 40830 + }, + { + "grad_norm": 0.12767860293388367, + "learning_rate": 2.5385218291283597e-05, + "loss": 0.0031, + "step": 40840 + }, + { + "grad_norm": 0.23887500166893005, + "learning_rate": 2.5361234922382383e-05, + "loss": 0.0025, + "step": 40850 + }, + { + "grad_norm": 0.13887599110603333, + "learning_rate": 2.533725903808749e-05, + "loss": 0.0023, + "step": 40860 + }, + { + "grad_norm": 0.124891497194767, + "learning_rate": 2.5313290645682085e-05, + "loss": 0.0028, + "step": 40870 + }, + { + "grad_norm": 0.15588347613811493, + "learning_rate": 2.52893297524472e-05, + "loss": 0.0035, + "step": 40880 + }, + { + "grad_norm": 0.20391803979873657, + "learning_rate": 2.526537636566145e-05, + "loss": 0.0036, + "step": 40890 + }, + { + "grad_norm": 0.11534092575311661, + "learning_rate": 2.5241430492601305e-05, + "loss": 0.0042, + "step": 40900 + }, + { + "grad_norm": 0.13671579957008362, + "learning_rate": 2.5217492140540867e-05, + "loss": 0.0042, + "step": 40910 + }, + { + "grad_norm": 0.1937016099691391, + "learning_rate": 2.5193561316751967e-05, + "loss": 0.0027, + "step": 40920 + }, + { + "grad_norm": 0.14325913786888123, + "learning_rate": 2.516963802850416e-05, + "loss": 0.003, + "step": 40930 + }, + { + "grad_norm": 0.13932280242443085, + "learning_rate": 2.5145722283064698e-05, + "loss": 0.0035, + "step": 40940 + }, + { + "grad_norm": 0.11823240667581558, + "learning_rate": 2.5121814087698602e-05, + "loss": 0.0037, + "step": 40950 + }, + { + "grad_norm": 0.12812559306621552, + "learning_rate": 2.509791344966848e-05, + "loss": 0.0028, + "step": 40960 + }, + { + "grad_norm": 0.21148385107517242, + "learning_rate": 2.5074020376234768e-05, + "loss": 0.0051, + "step": 40970 + }, + { + "grad_norm": 0.12552213668823242, + "learning_rate": 2.5050134874655534e-05, + "loss": 0.0027, + "step": 40980 + }, + { + "grad_norm": 0.18470603227615356, + "learning_rate": 2.5026256952186566e-05, + "loss": 0.0044, + "step": 40990 + }, + { + "grad_norm": 0.12672986090183258, + "learning_rate": 2.5002386616081335e-05, + "loss": 0.0037, + "step": 41000 + }, + { + "grad_norm": 0.18359945714473724, + "learning_rate": 2.497852387359103e-05, + "loss": 0.0043, + "step": 41010 + }, + { + "grad_norm": 0.17262490093708038, + "learning_rate": 2.4954668731964496e-05, + "loss": 0.0031, + "step": 41020 + }, + { + "grad_norm": 0.14123176038265228, + "learning_rate": 2.4930821198448364e-05, + "loss": 0.0036, + "step": 41030 + }, + { + "grad_norm": 0.16663289070129395, + "learning_rate": 2.4906981280286796e-05, + "loss": 0.0027, + "step": 41040 + }, + { + "grad_norm": 0.11400409042835236, + "learning_rate": 2.488314898472179e-05, + "loss": 0.0028, + "step": 41050 + }, + { + "grad_norm": 0.09916108846664429, + "learning_rate": 2.485932431899295e-05, + "loss": 0.0023, + "step": 41060 + }, + { + "grad_norm": 0.09815013408660889, + "learning_rate": 2.4835507290337584e-05, + "loss": 0.0029, + "step": 41070 + }, + { + "grad_norm": 0.133224755525589, + "learning_rate": 2.4811697905990672e-05, + "loss": 0.0035, + "step": 41080 + }, + { + "grad_norm": 0.1441335529088974, + "learning_rate": 2.4787896173184854e-05, + "loss": 0.0034, + "step": 41090 + }, + { + "grad_norm": 0.14974042773246765, + "learning_rate": 2.4764102099150534e-05, + "loss": 0.0033, + "step": 41100 + }, + { + "grad_norm": 0.1326851099729538, + "learning_rate": 2.4740315691115644e-05, + "loss": 0.0027, + "step": 41110 + }, + { + "grad_norm": 0.16287770867347717, + "learning_rate": 2.4716536956305918e-05, + "loss": 0.0026, + "step": 41120 + }, + { + "grad_norm": 0.11091754585504532, + "learning_rate": 2.4692765901944697e-05, + "loss": 0.0026, + "step": 41130 + }, + { + "grad_norm": 0.13137423992156982, + "learning_rate": 2.4669002535253e-05, + "loss": 0.0027, + "step": 41140 + }, + { + "grad_norm": 0.09918172657489777, + "learning_rate": 2.46452468634495e-05, + "loss": 0.0027, + "step": 41150 + }, + { + "grad_norm": 0.13535438477993011, + "learning_rate": 2.462149889375055e-05, + "loss": 0.0025, + "step": 41160 + }, + { + "grad_norm": 0.1631234884262085, + "learning_rate": 2.459775863337014e-05, + "loss": 0.0033, + "step": 41170 + }, + { + "grad_norm": 0.1580214500427246, + "learning_rate": 2.4574026089519985e-05, + "loss": 0.0031, + "step": 41180 + }, + { + "grad_norm": 0.21731801331043243, + "learning_rate": 2.4550301269409333e-05, + "loss": 0.0045, + "step": 41190 + }, + { + "grad_norm": 0.19862671196460724, + "learning_rate": 2.4526584180245216e-05, + "loss": 0.006, + "step": 41200 + }, + { + "grad_norm": 0.18760280311107635, + "learning_rate": 2.4502874829232236e-05, + "loss": 0.003, + "step": 41210 + }, + { + "grad_norm": 0.13663454353809357, + "learning_rate": 2.447917322357267e-05, + "loss": 0.003, + "step": 41220 + }, + { + "grad_norm": 0.11657775193452835, + "learning_rate": 2.4455479370466443e-05, + "loss": 0.0026, + "step": 41230 + }, + { + "grad_norm": 0.15587320923805237, + "learning_rate": 2.4431793277111097e-05, + "loss": 0.0033, + "step": 41240 + }, + { + "grad_norm": 0.136717289686203, + "learning_rate": 2.4408114950701905e-05, + "loss": 0.0053, + "step": 41250 + }, + { + "grad_norm": 0.17365571856498718, + "learning_rate": 2.4384444398431634e-05, + "loss": 0.0035, + "step": 41260 + }, + { + "grad_norm": 0.1508485972881317, + "learning_rate": 2.4360781627490837e-05, + "loss": 0.005, + "step": 41270 + }, + { + "grad_norm": 0.15707406401634216, + "learning_rate": 2.433712664506762e-05, + "loss": 0.0048, + "step": 41280 + }, + { + "grad_norm": 0.11897261440753937, + "learning_rate": 2.431347945834774e-05, + "loss": 0.0023, + "step": 41290 + }, + { + "grad_norm": 0.19674824178218842, + "learning_rate": 2.428984007451458e-05, + "loss": 0.003, + "step": 41300 + }, + { + "grad_norm": 0.1379646211862564, + "learning_rate": 2.426620850074917e-05, + "loss": 0.0029, + "step": 41310 + }, + { + "grad_norm": 0.16234001517295837, + "learning_rate": 2.424258474423014e-05, + "loss": 0.0037, + "step": 41320 + }, + { + "grad_norm": 0.16070736944675446, + "learning_rate": 2.421896881213382e-05, + "loss": 0.0033, + "step": 41330 + }, + { + "grad_norm": 0.1205456554889679, + "learning_rate": 2.419536071163402e-05, + "loss": 0.0037, + "step": 41340 + }, + { + "grad_norm": 0.15602603554725647, + "learning_rate": 2.417176044990233e-05, + "loss": 0.0035, + "step": 41350 + }, + { + "grad_norm": 0.17379288375377655, + "learning_rate": 2.4148168034107855e-05, + "loss": 0.0034, + "step": 41360 + }, + { + "grad_norm": 0.12310892343521118, + "learning_rate": 2.4124583471417355e-05, + "loss": 0.0028, + "step": 41370 + }, + { + "grad_norm": 0.14370737969875336, + "learning_rate": 2.41010067689952e-05, + "loss": 0.0034, + "step": 41380 + }, + { + "grad_norm": 0.12322472780942917, + "learning_rate": 2.4077437934003338e-05, + "loss": 0.005, + "step": 41390 + }, + { + "grad_norm": 0.11277265101671219, + "learning_rate": 2.405387697360143e-05, + "loss": 0.0024, + "step": 41400 + }, + { + "grad_norm": 0.13990013301372528, + "learning_rate": 2.4030323894946595e-05, + "loss": 0.0042, + "step": 41410 + }, + { + "grad_norm": 0.12131306529045105, + "learning_rate": 2.40067787051937e-05, + "loss": 0.0035, + "step": 41420 + }, + { + "grad_norm": 0.14810140430927277, + "learning_rate": 2.3983241411495087e-05, + "loss": 0.0029, + "step": 41430 + }, + { + "grad_norm": 0.16509346663951874, + "learning_rate": 2.3959712021000823e-05, + "loss": 0.0041, + "step": 41440 + }, + { + "grad_norm": 0.1655333787202835, + "learning_rate": 2.3936190540858495e-05, + "loss": 0.003, + "step": 41450 + }, + { + "grad_norm": 0.127410888671875, + "learning_rate": 2.39126769782133e-05, + "loss": 0.0029, + "step": 41460 + }, + { + "grad_norm": 0.15603305399417877, + "learning_rate": 2.388917134020805e-05, + "loss": 0.0042, + "step": 41470 + }, + { + "grad_norm": 0.13136667013168335, + "learning_rate": 2.3865673633983128e-05, + "loss": 0.0033, + "step": 41480 + }, + { + "grad_norm": 0.15359167754650116, + "learning_rate": 2.3842183866676492e-05, + "loss": 0.0039, + "step": 41490 + }, + { + "grad_norm": 0.16049325466156006, + "learning_rate": 2.381870204542377e-05, + "loss": 0.0033, + "step": 41500 + }, + { + "grad_norm": 0.14517591893672943, + "learning_rate": 2.379522817735808e-05, + "loss": 0.0034, + "step": 41510 + }, + { + "grad_norm": 0.09945697337388992, + "learning_rate": 2.377176226961018e-05, + "loss": 0.0028, + "step": 41520 + }, + { + "grad_norm": 0.14102882146835327, + "learning_rate": 2.3748304329308384e-05, + "loss": 0.0036, + "step": 41530 + }, + { + "grad_norm": 0.14894405007362366, + "learning_rate": 2.372485436357858e-05, + "loss": 0.0038, + "step": 41540 + }, + { + "grad_norm": 0.18140621483325958, + "learning_rate": 2.3701412379544296e-05, + "loss": 0.0037, + "step": 41550 + }, + { + "grad_norm": 0.14919565618038177, + "learning_rate": 2.367797838432653e-05, + "loss": 0.0026, + "step": 41560 + }, + { + "grad_norm": 0.14559563994407654, + "learning_rate": 2.3654552385043967e-05, + "loss": 0.0044, + "step": 41570 + }, + { + "grad_norm": 0.15985830128192902, + "learning_rate": 2.3631134388812742e-05, + "loss": 0.0032, + "step": 41580 + }, + { + "grad_norm": 0.13166458904743195, + "learning_rate": 2.3607724402746684e-05, + "loss": 0.0051, + "step": 41590 + }, + { + "grad_norm": 0.13355065882205963, + "learning_rate": 2.35843224339571e-05, + "loss": 0.0029, + "step": 41600 + }, + { + "grad_norm": 0.1258516162633896, + "learning_rate": 2.3560928489552897e-05, + "loss": 0.0028, + "step": 41610 + }, + { + "grad_norm": 0.19604630768299103, + "learning_rate": 2.353754257664053e-05, + "loss": 0.0051, + "step": 41620 + }, + { + "grad_norm": 0.19981656968593597, + "learning_rate": 2.3514164702324037e-05, + "loss": 0.0027, + "step": 41630 + }, + { + "grad_norm": 0.14043785631656647, + "learning_rate": 2.3490794873704963e-05, + "loss": 0.0027, + "step": 41640 + }, + { + "grad_norm": 0.15612654387950897, + "learning_rate": 2.3467433097882496e-05, + "loss": 0.0021, + "step": 41650 + }, + { + "grad_norm": 0.10130954533815384, + "learning_rate": 2.34440793819533e-05, + "loss": 0.0021, + "step": 41660 + }, + { + "grad_norm": 0.15227554738521576, + "learning_rate": 2.3420733733011617e-05, + "loss": 0.0051, + "step": 41670 + }, + { + "grad_norm": 0.17008958756923676, + "learning_rate": 2.3397396158149243e-05, + "loss": 0.0025, + "step": 41680 + }, + { + "grad_norm": 0.16074174642562866, + "learning_rate": 2.3374066664455498e-05, + "loss": 0.0042, + "step": 41690 + }, + { + "grad_norm": 0.1633201241493225, + "learning_rate": 2.3350745259017315e-05, + "loss": 0.0032, + "step": 41700 + }, + { + "grad_norm": 0.13288123905658722, + "learning_rate": 2.332743194891906e-05, + "loss": 0.0035, + "step": 41710 + }, + { + "grad_norm": 0.1581203043460846, + "learning_rate": 2.330412674124276e-05, + "loss": 0.0036, + "step": 41720 + }, + { + "grad_norm": 0.10304956883192062, + "learning_rate": 2.328082964306786e-05, + "loss": 0.0028, + "step": 41730 + }, + { + "grad_norm": 0.1855410635471344, + "learning_rate": 2.325754066147145e-05, + "loss": 0.0037, + "step": 41740 + }, + { + "grad_norm": 0.1351453959941864, + "learning_rate": 2.32342598035281e-05, + "loss": 0.003, + "step": 41750 + }, + { + "grad_norm": 0.14050233364105225, + "learning_rate": 2.321098707630991e-05, + "loss": 0.0024, + "step": 41760 + }, + { + "grad_norm": 0.12904061377048492, + "learning_rate": 2.318772248688652e-05, + "loss": 0.0061, + "step": 41770 + }, + { + "grad_norm": 0.14126089215278625, + "learning_rate": 2.3164466042325107e-05, + "loss": 0.0032, + "step": 41780 + }, + { + "grad_norm": 0.15881861746311188, + "learning_rate": 2.3141217749690353e-05, + "loss": 0.0038, + "step": 41790 + }, + { + "grad_norm": 0.12823353707790375, + "learning_rate": 2.3117977616044466e-05, + "loss": 0.0031, + "step": 41800 + }, + { + "grad_norm": 0.124204620718956, + "learning_rate": 2.309474564844722e-05, + "loss": 0.0029, + "step": 41810 + }, + { + "grad_norm": 0.1603548377752304, + "learning_rate": 2.307152185395585e-05, + "loss": 0.0032, + "step": 41820 + }, + { + "grad_norm": 0.1899862140417099, + "learning_rate": 2.3048306239625144e-05, + "loss": 0.004, + "step": 41830 + }, + { + "grad_norm": 0.10973954200744629, + "learning_rate": 2.3025098812507378e-05, + "loss": 0.0021, + "step": 41840 + }, + { + "grad_norm": 0.13352397084236145, + "learning_rate": 2.3001899579652366e-05, + "loss": 0.0046, + "step": 41850 + }, + { + "grad_norm": 0.1227138563990593, + "learning_rate": 2.2978708548107393e-05, + "loss": 0.0033, + "step": 41860 + }, + { + "grad_norm": 0.11199397593736649, + "learning_rate": 2.2955525724917348e-05, + "loss": 0.0026, + "step": 41870 + }, + { + "grad_norm": 0.23392455279827118, + "learning_rate": 2.2932351117124477e-05, + "loss": 0.0055, + "step": 41880 + }, + { + "grad_norm": 0.16378571093082428, + "learning_rate": 2.29091847317687e-05, + "loss": 0.0027, + "step": 41890 + }, + { + "grad_norm": 0.11797003448009491, + "learning_rate": 2.2886026575887277e-05, + "loss": 0.0029, + "step": 41900 + }, + { + "grad_norm": 0.12120306491851807, + "learning_rate": 2.2862876656515094e-05, + "loss": 0.0042, + "step": 41910 + }, + { + "grad_norm": 0.12483644485473633, + "learning_rate": 2.2839734980684464e-05, + "loss": 0.0028, + "step": 41920 + }, + { + "grad_norm": 0.11894072592258453, + "learning_rate": 2.281660155542522e-05, + "loss": 0.0029, + "step": 41930 + }, + { + "grad_norm": 0.16940362751483917, + "learning_rate": 2.279347638776469e-05, + "loss": 0.0024, + "step": 41940 + }, + { + "grad_norm": 0.16561780869960785, + "learning_rate": 2.2770359484727665e-05, + "loss": 0.0029, + "step": 41950 + }, + { + "grad_norm": 0.13039658963680267, + "learning_rate": 2.27472508533365e-05, + "loss": 0.0032, + "step": 41960 + }, + { + "grad_norm": 0.16566585004329681, + "learning_rate": 2.2724150500610948e-05, + "loss": 0.0033, + "step": 41970 + }, + { + "grad_norm": 0.10417769849300385, + "learning_rate": 2.2701058433568302e-05, + "loss": 0.0031, + "step": 41980 + }, + { + "grad_norm": 0.15752018988132477, + "learning_rate": 2.2677974659223318e-05, + "loss": 0.0035, + "step": 41990 + }, + { + "grad_norm": 0.09713495522737503, + "learning_rate": 2.2654899184588235e-05, + "loss": 0.0025, + "step": 42000 + }, + { + "grad_norm": 0.18814073503017426, + "learning_rate": 2.2631832016672756e-05, + "loss": 0.0031, + "step": 42010 + }, + { + "grad_norm": 0.15588198602199554, + "learning_rate": 2.2608773162484127e-05, + "loss": 0.0028, + "step": 42020 + }, + { + "grad_norm": 0.17948514223098755, + "learning_rate": 2.2585722629026958e-05, + "loss": 0.0045, + "step": 42030 + }, + { + "grad_norm": 0.12258731573820114, + "learning_rate": 2.2562680423303457e-05, + "loss": 0.0033, + "step": 42040 + }, + { + "grad_norm": 0.11625772714614868, + "learning_rate": 2.2539646552313165e-05, + "loss": 0.0024, + "step": 42050 + }, + { + "grad_norm": 0.1443842649459839, + "learning_rate": 2.251662102305322e-05, + "loss": 0.0043, + "step": 42060 + }, + { + "grad_norm": 0.1871681809425354, + "learning_rate": 2.2493603842518152e-05, + "loss": 0.0029, + "step": 42070 + }, + { + "grad_norm": 0.12985116243362427, + "learning_rate": 2.2470595017699974e-05, + "loss": 0.0029, + "step": 42080 + }, + { + "grad_norm": 0.16299666464328766, + "learning_rate": 2.244759455558816e-05, + "loss": 0.0028, + "step": 42090 + }, + { + "grad_norm": 0.13300037384033203, + "learning_rate": 2.2424602463169614e-05, + "loss": 0.0037, + "step": 42100 + }, + { + "grad_norm": 0.1265360414981842, + "learning_rate": 2.2401618747428776e-05, + "loss": 0.0034, + "step": 42110 + }, + { + "grad_norm": 0.10668852925300598, + "learning_rate": 2.237864341534747e-05, + "loss": 0.0026, + "step": 42120 + }, + { + "grad_norm": 0.16422423720359802, + "learning_rate": 2.2355676473904998e-05, + "loss": 0.0029, + "step": 42130 + }, + { + "grad_norm": 0.08642151206731796, + "learning_rate": 2.2332717930078108e-05, + "loss": 0.0025, + "step": 42140 + }, + { + "grad_norm": 0.1471499502658844, + "learning_rate": 2.2309767790840992e-05, + "loss": 0.0031, + "step": 42150 + }, + { + "grad_norm": 0.13878025114536285, + "learning_rate": 2.228682606316529e-05, + "loss": 0.0035, + "step": 42160 + }, + { + "grad_norm": 0.195353701710701, + "learning_rate": 2.2263892754020138e-05, + "loss": 0.0029, + "step": 42170 + }, + { + "grad_norm": 0.13064812123775482, + "learning_rate": 2.2240967870372004e-05, + "loss": 0.0036, + "step": 42180 + }, + { + "grad_norm": 0.15243151783943176, + "learning_rate": 2.2218051419184933e-05, + "loss": 0.0034, + "step": 42190 + }, + { + "grad_norm": 0.19037245213985443, + "learning_rate": 2.219514340742026e-05, + "loss": 0.0033, + "step": 42200 + }, + { + "grad_norm": 0.13021117448806763, + "learning_rate": 2.2172243842036898e-05, + "loss": 0.0032, + "step": 42210 + }, + { + "grad_norm": 0.13808530569076538, + "learning_rate": 2.2149352729991107e-05, + "loss": 0.0026, + "step": 42220 + }, + { + "grad_norm": 0.24775077402591705, + "learning_rate": 2.2126470078236605e-05, + "loss": 0.0035, + "step": 42230 + }, + { + "grad_norm": 0.1521518975496292, + "learning_rate": 2.2103595893724533e-05, + "loss": 0.0033, + "step": 42240 + }, + { + "grad_norm": 0.1505928635597229, + "learning_rate": 2.208073018340345e-05, + "loss": 0.003, + "step": 42250 + }, + { + "grad_norm": 0.14481805264949799, + "learning_rate": 2.2057872954219405e-05, + "loss": 0.0031, + "step": 42260 + }, + { + "grad_norm": 0.1576819270849228, + "learning_rate": 2.203502421311575e-05, + "loss": 0.0036, + "step": 42270 + }, + { + "grad_norm": 0.1533893644809723, + "learning_rate": 2.2012183967033388e-05, + "loss": 0.0025, + "step": 42280 + }, + { + "grad_norm": 0.1618533879518509, + "learning_rate": 2.198935222291056e-05, + "loss": 0.0026, + "step": 42290 + }, + { + "grad_norm": 0.1628190279006958, + "learning_rate": 2.1966528987682948e-05, + "loss": 0.0042, + "step": 42300 + }, + { + "grad_norm": 0.13917112350463867, + "learning_rate": 2.194371426828365e-05, + "loss": 0.0027, + "step": 42310 + }, + { + "grad_norm": 0.17998965084552765, + "learning_rate": 2.192090807164317e-05, + "loss": 0.0061, + "step": 42320 + }, + { + "grad_norm": 0.1781017780303955, + "learning_rate": 2.1898110404689422e-05, + "loss": 0.0038, + "step": 42330 + }, + { + "grad_norm": 0.17894701659679413, + "learning_rate": 2.1875321274347776e-05, + "loss": 0.0028, + "step": 42340 + }, + { + "grad_norm": 0.1310628354549408, + "learning_rate": 2.18525406875409e-05, + "loss": 0.004, + "step": 42350 + }, + { + "grad_norm": 0.10396751016378403, + "learning_rate": 2.1829768651188997e-05, + "loss": 0.0029, + "step": 42360 + }, + { + "grad_norm": 0.12844139337539673, + "learning_rate": 2.180700517220958e-05, + "loss": 0.004, + "step": 42370 + }, + { + "grad_norm": 0.20522047579288483, + "learning_rate": 2.1784250257517603e-05, + "loss": 0.0055, + "step": 42380 + }, + { + "grad_norm": 0.1703515201807022, + "learning_rate": 2.1761503914025406e-05, + "loss": 0.0044, + "step": 42390 + }, + { + "grad_norm": 0.14236022531986237, + "learning_rate": 2.1738766148642705e-05, + "loss": 0.0024, + "step": 42400 + }, + { + "grad_norm": 0.11090123653411865, + "learning_rate": 2.1716036968276683e-05, + "loss": 0.0031, + "step": 42410 + }, + { + "grad_norm": 0.16948281228542328, + "learning_rate": 2.1693316379831808e-05, + "loss": 0.004, + "step": 42420 + }, + { + "grad_norm": 0.15160688757896423, + "learning_rate": 2.1670604390210037e-05, + "loss": 0.0049, + "step": 42430 + }, + { + "grad_norm": 0.11981331557035446, + "learning_rate": 2.1647901006310656e-05, + "loss": 0.0027, + "step": 42440 + }, + { + "grad_norm": 0.1265793740749359, + "learning_rate": 2.1625206235030353e-05, + "loss": 0.0029, + "step": 42450 + }, + { + "grad_norm": 0.12867692112922668, + "learning_rate": 2.160252008326321e-05, + "loss": 0.0041, + "step": 42460 + }, + { + "grad_norm": 0.1542431265115738, + "learning_rate": 2.157984255790067e-05, + "loss": 0.0038, + "step": 42470 + }, + { + "grad_norm": 0.12827377021312714, + "learning_rate": 2.1557173665831553e-05, + "loss": 0.0041, + "step": 42480 + }, + { + "grad_norm": 0.1588735729455948, + "learning_rate": 2.153451341394212e-05, + "loss": 0.0033, + "step": 42490 + }, + { + "grad_norm": 0.17535175383090973, + "learning_rate": 2.151186180911589e-05, + "loss": 0.0036, + "step": 42500 + }, + { + "grad_norm": 0.11486001312732697, + "learning_rate": 2.1489218858233877e-05, + "loss": 0.003, + "step": 42510 + }, + { + "grad_norm": 0.16072852909564972, + "learning_rate": 2.1466584568174392e-05, + "loss": 0.0052, + "step": 42520 + }, + { + "grad_norm": 0.14080722630023956, + "learning_rate": 2.1443958945813132e-05, + "loss": 0.0034, + "step": 42530 + }, + { + "grad_norm": 0.13305938243865967, + "learning_rate": 2.1421341998023163e-05, + "loss": 0.003, + "step": 42540 + }, + { + "grad_norm": 0.13315212726593018, + "learning_rate": 2.139873373167491e-05, + "loss": 0.0032, + "step": 42550 + }, + { + "grad_norm": 0.09000915288925171, + "learning_rate": 2.13761341536362e-05, + "loss": 0.0021, + "step": 42560 + }, + { + "grad_norm": 0.11708277463912964, + "learning_rate": 2.1353543270772136e-05, + "loss": 0.0025, + "step": 42570 + }, + { + "grad_norm": 0.15893222391605377, + "learning_rate": 2.1330961089945297e-05, + "loss": 0.0036, + "step": 42580 + }, + { + "grad_norm": 0.16108164191246033, + "learning_rate": 2.130838761801548e-05, + "loss": 0.0032, + "step": 42590 + }, + { + "grad_norm": 0.16260536015033722, + "learning_rate": 2.1285822861839966e-05, + "loss": 0.0058, + "step": 42600 + }, + { + "grad_norm": 0.1277707815170288, + "learning_rate": 2.126326682827331e-05, + "loss": 0.0031, + "step": 42610 + }, + { + "grad_norm": 0.1873548924922943, + "learning_rate": 2.124071952416744e-05, + "loss": 0.0031, + "step": 42620 + }, + { + "grad_norm": 0.17107369005680084, + "learning_rate": 2.1218180956371634e-05, + "loss": 0.0045, + "step": 42630 + }, + { + "grad_norm": 0.1146121546626091, + "learning_rate": 2.119565113173252e-05, + "loss": 0.003, + "step": 42640 + }, + { + "grad_norm": 0.0860452950000763, + "learning_rate": 2.1173130057094033e-05, + "loss": 0.0043, + "step": 42650 + }, + { + "grad_norm": 0.15265463292598724, + "learning_rate": 2.115061773929753e-05, + "loss": 0.0022, + "step": 42660 + }, + { + "grad_norm": 0.13698548078536987, + "learning_rate": 2.1128114185181623e-05, + "loss": 0.0027, + "step": 42670 + }, + { + "grad_norm": 0.12606237828731537, + "learning_rate": 2.1105619401582317e-05, + "loss": 0.0032, + "step": 42680 + }, + { + "grad_norm": 0.1708395928144455, + "learning_rate": 2.1083133395332928e-05, + "loss": 0.0038, + "step": 42690 + }, + { + "grad_norm": 0.2103361338376999, + "learning_rate": 2.1060656173264082e-05, + "loss": 0.0044, + "step": 42700 + }, + { + "grad_norm": 0.21404999494552612, + "learning_rate": 2.103818774220383e-05, + "loss": 0.0045, + "step": 42710 + }, + { + "grad_norm": 0.16158972680568695, + "learning_rate": 2.1015728108977412e-05, + "loss": 0.0034, + "step": 42720 + }, + { + "grad_norm": 0.16464819014072418, + "learning_rate": 2.0993277280407548e-05, + "loss": 0.0036, + "step": 42730 + }, + { + "grad_norm": 0.1364492028951645, + "learning_rate": 2.0970835263314132e-05, + "loss": 0.0033, + "step": 42740 + }, + { + "grad_norm": 0.11004994809627533, + "learning_rate": 2.094840206451451e-05, + "loss": 0.0022, + "step": 42750 + }, + { + "grad_norm": 0.12882307171821594, + "learning_rate": 2.0925977690823273e-05, + "loss": 0.0033, + "step": 42760 + }, + { + "grad_norm": 0.11047834157943726, + "learning_rate": 2.0903562149052364e-05, + "loss": 0.0022, + "step": 42770 + }, + { + "grad_norm": 0.10292172431945801, + "learning_rate": 2.0881155446011025e-05, + "loss": 0.0035, + "step": 42780 + }, + { + "grad_norm": 0.12612217664718628, + "learning_rate": 2.0858757588505823e-05, + "loss": 0.0029, + "step": 42790 + }, + { + "grad_norm": 0.21856506168842316, + "learning_rate": 2.0836368583340622e-05, + "loss": 0.0032, + "step": 42800 + }, + { + "grad_norm": 0.13722337782382965, + "learning_rate": 2.081398843731664e-05, + "loss": 0.0022, + "step": 42810 + }, + { + "grad_norm": 0.14211338758468628, + "learning_rate": 2.0791617157232357e-05, + "loss": 0.0056, + "step": 42820 + }, + { + "grad_norm": 0.17739957571029663, + "learning_rate": 2.0769254749883576e-05, + "loss": 0.0031, + "step": 42830 + }, + { + "grad_norm": 0.11499431729316711, + "learning_rate": 2.0746901222063415e-05, + "loss": 0.0037, + "step": 42840 + }, + { + "grad_norm": 0.1767963021993637, + "learning_rate": 2.072455658056226e-05, + "loss": 0.003, + "step": 42850 + }, + { + "grad_norm": 0.19275100529193878, + "learning_rate": 2.0702220832167873e-05, + "loss": 0.0034, + "step": 42860 + }, + { + "grad_norm": 0.13615760207176208, + "learning_rate": 2.0679893983665205e-05, + "loss": 0.0022, + "step": 42870 + }, + { + "grad_norm": 0.12432663887739182, + "learning_rate": 2.0657576041836622e-05, + "loss": 0.0051, + "step": 42880 + }, + { + "grad_norm": 0.1212100014090538, + "learning_rate": 2.0635267013461666e-05, + "loss": 0.0025, + "step": 42890 + }, + { + "grad_norm": 0.13655124604701996, + "learning_rate": 2.061296690531728e-05, + "loss": 0.0032, + "step": 42900 + }, + { + "grad_norm": 0.12499711662530899, + "learning_rate": 2.0590675724177622e-05, + "loss": 0.0026, + "step": 42910 + }, + { + "grad_norm": 0.14739903807640076, + "learning_rate": 2.0568393476814167e-05, + "loss": 0.0026, + "step": 42920 + }, + { + "grad_norm": 0.14726291596889496, + "learning_rate": 2.0546120169995685e-05, + "loss": 0.0036, + "step": 42930 + }, + { + "grad_norm": 0.16694466769695282, + "learning_rate": 2.0523855810488214e-05, + "loss": 0.0024, + "step": 42940 + }, + { + "grad_norm": 0.12402988970279694, + "learning_rate": 2.050160040505505e-05, + "loss": 0.0023, + "step": 42950 + }, + { + "grad_norm": 0.14468993246555328, + "learning_rate": 2.0479353960456843e-05, + "loss": 0.0029, + "step": 42960 + }, + { + "grad_norm": 0.12729193270206451, + "learning_rate": 2.0457116483451456e-05, + "loss": 0.0026, + "step": 42970 + }, + { + "grad_norm": 0.15954506397247314, + "learning_rate": 2.0434887980794043e-05, + "loss": 0.0039, + "step": 42980 + }, + { + "grad_norm": 0.14968369901180267, + "learning_rate": 2.0412668459237043e-05, + "loss": 0.0023, + "step": 42990 + }, + { + "grad_norm": 0.16360320150852203, + "learning_rate": 2.039045792553016e-05, + "loss": 0.0029, + "step": 43000 + }, + { + "grad_norm": 0.138072207570076, + "learning_rate": 2.036825638642036e-05, + "loss": 0.0029, + "step": 43010 + }, + { + "grad_norm": 0.14269419014453888, + "learning_rate": 2.0346063848651868e-05, + "loss": 0.0026, + "step": 43020 + }, + { + "grad_norm": 0.12009543180465698, + "learning_rate": 2.0323880318966254e-05, + "loss": 0.0024, + "step": 43030 + }, + { + "grad_norm": 0.15685294568538666, + "learning_rate": 2.030170580410221e-05, + "loss": 0.0031, + "step": 43040 + }, + { + "grad_norm": 0.11948579549789429, + "learning_rate": 2.0279540310795837e-05, + "loss": 0.0032, + "step": 43050 + }, + { + "grad_norm": 0.13224561512470245, + "learning_rate": 2.0257383845780365e-05, + "loss": 0.0031, + "step": 43060 + }, + { + "grad_norm": 0.19799253344535828, + "learning_rate": 2.0235236415786384e-05, + "loss": 0.0051, + "step": 43070 + }, + { + "grad_norm": 0.15876567363739014, + "learning_rate": 2.021309802754169e-05, + "loss": 0.0033, + "step": 43080 + }, + { + "grad_norm": 0.15996171534061432, + "learning_rate": 2.0190968687771332e-05, + "loss": 0.004, + "step": 43090 + }, + { + "grad_norm": 0.13225407898426056, + "learning_rate": 2.016884840319763e-05, + "loss": 0.0025, + "step": 43100 + }, + { + "grad_norm": 0.10967284440994263, + "learning_rate": 2.0146737180540122e-05, + "loss": 0.0035, + "step": 43110 + }, + { + "grad_norm": 0.1429014503955841, + "learning_rate": 2.012463502651564e-05, + "loss": 0.0037, + "step": 43120 + }, + { + "grad_norm": 0.142154723405838, + "learning_rate": 2.0102541947838228e-05, + "loss": 0.0041, + "step": 43130 + }, + { + "grad_norm": 0.12910042703151703, + "learning_rate": 2.0080457951219173e-05, + "loss": 0.002, + "step": 43140 + }, + { + "grad_norm": 0.14672163128852844, + "learning_rate": 2.0058383043367017e-05, + "loss": 0.0033, + "step": 43150 + }, + { + "grad_norm": 0.10366083681583405, + "learning_rate": 2.0036317230987528e-05, + "loss": 0.0025, + "step": 43160 + }, + { + "grad_norm": 0.1546521633863449, + "learning_rate": 2.0014260520783696e-05, + "loss": 0.0034, + "step": 43170 + }, + { + "grad_norm": 0.1403999626636505, + "learning_rate": 1.9992212919455834e-05, + "loss": 0.0034, + "step": 43180 + }, + { + "grad_norm": 0.1530219465494156, + "learning_rate": 1.9970174433701333e-05, + "loss": 0.0032, + "step": 43190 + }, + { + "grad_norm": 0.15878364443778992, + "learning_rate": 1.9948145070214992e-05, + "loss": 0.0028, + "step": 43200 + }, + { + "grad_norm": 0.1135379821062088, + "learning_rate": 1.9926124835688663e-05, + "loss": 0.0033, + "step": 43210 + }, + { + "grad_norm": 0.1723741739988327, + "learning_rate": 1.9904113736811576e-05, + "loss": 0.0049, + "step": 43220 + }, + { + "grad_norm": 0.13433456420898438, + "learning_rate": 1.9882111780270096e-05, + "loss": 0.0058, + "step": 43230 + }, + { + "grad_norm": 0.1358075588941574, + "learning_rate": 1.986011897274784e-05, + "loss": 0.0036, + "step": 43240 + }, + { + "grad_norm": 0.11093565821647644, + "learning_rate": 1.983813532092565e-05, + "loss": 0.0033, + "step": 43250 + }, + { + "grad_norm": 0.1462140679359436, + "learning_rate": 1.981616083148155e-05, + "loss": 0.0031, + "step": 43260 + }, + { + "grad_norm": 0.17562538385391235, + "learning_rate": 1.9794195511090845e-05, + "loss": 0.0025, + "step": 43270 + }, + { + "grad_norm": 0.1119430661201477, + "learning_rate": 1.977223936642601e-05, + "loss": 0.0027, + "step": 43280 + }, + { + "grad_norm": 0.1822449415922165, + "learning_rate": 1.975029240415674e-05, + "loss": 0.0032, + "step": 43290 + }, + { + "grad_norm": 0.10740764439105988, + "learning_rate": 1.9728354630949936e-05, + "loss": 0.0028, + "step": 43300 + }, + { + "grad_norm": 0.12288554012775421, + "learning_rate": 1.9706426053469716e-05, + "loss": 0.0024, + "step": 43310 + }, + { + "grad_norm": 0.1317206770181656, + "learning_rate": 1.9684506678377396e-05, + "loss": 0.0029, + "step": 43320 + }, + { + "grad_norm": 0.19511902332305908, + "learning_rate": 1.9662596512331544e-05, + "loss": 0.0049, + "step": 43330 + }, + { + "grad_norm": 0.2005932778120041, + "learning_rate": 1.964069556198782e-05, + "loss": 0.0029, + "step": 43340 + }, + { + "grad_norm": 0.16023850440979004, + "learning_rate": 1.9618803833999232e-05, + "loss": 0.0038, + "step": 43350 + }, + { + "grad_norm": 0.1335248202085495, + "learning_rate": 1.9596921335015838e-05, + "loss": 0.0034, + "step": 43360 + }, + { + "grad_norm": 0.13381657004356384, + "learning_rate": 1.957504807168501e-05, + "loss": 0.0039, + "step": 43370 + }, + { + "grad_norm": 0.11346202343702316, + "learning_rate": 1.9553184050651253e-05, + "loss": 0.0033, + "step": 43380 + }, + { + "grad_norm": 0.14977198839187622, + "learning_rate": 1.953132927855628e-05, + "loss": 0.0038, + "step": 43390 + }, + { + "grad_norm": 0.13065074384212494, + "learning_rate": 1.9509483762038995e-05, + "loss": 0.0034, + "step": 43400 + }, + { + "grad_norm": 0.1366412341594696, + "learning_rate": 1.9487647507735467e-05, + "loss": 0.0035, + "step": 43410 + }, + { + "grad_norm": 0.19153328239917755, + "learning_rate": 1.9465820522279032e-05, + "loss": 0.0061, + "step": 43420 + }, + { + "grad_norm": 0.1296260505914688, + "learning_rate": 1.9444002812300078e-05, + "loss": 0.0044, + "step": 43430 + }, + { + "grad_norm": 0.15517450869083405, + "learning_rate": 1.94221943844263e-05, + "loss": 0.0033, + "step": 43440 + }, + { + "grad_norm": 0.10487982630729675, + "learning_rate": 1.9400395245282515e-05, + "loss": 0.0025, + "step": 43450 + }, + { + "grad_norm": 0.20621170103549957, + "learning_rate": 1.937860540149071e-05, + "loss": 0.0027, + "step": 43460 + }, + { + "grad_norm": 0.10109329223632812, + "learning_rate": 1.9356824859670082e-05, + "loss": 0.0027, + "step": 43470 + }, + { + "grad_norm": 0.1692003309726715, + "learning_rate": 1.9335053626436967e-05, + "loss": 0.0033, + "step": 43480 + }, + { + "grad_norm": 0.1333615928888321, + "learning_rate": 1.9313291708404885e-05, + "loss": 0.0034, + "step": 43490 + }, + { + "grad_norm": 0.15854498744010925, + "learning_rate": 1.9291539112184587e-05, + "loss": 0.0029, + "step": 43500 + }, + { + "grad_norm": 0.14657294750213623, + "learning_rate": 1.9269795844383854e-05, + "loss": 0.002, + "step": 43510 + }, + { + "grad_norm": 0.19784286618232727, + "learning_rate": 1.9248061911607777e-05, + "loss": 0.003, + "step": 43520 + }, + { + "grad_norm": 0.16682776808738708, + "learning_rate": 1.9226337320458538e-05, + "loss": 0.003, + "step": 43530 + }, + { + "grad_norm": 0.1091584786772728, + "learning_rate": 1.9204622077535488e-05, + "loss": 0.0027, + "step": 43540 + }, + { + "grad_norm": 0.08509393781423569, + "learning_rate": 1.9182916189435147e-05, + "loss": 0.0028, + "step": 43550 + }, + { + "grad_norm": 0.12800520658493042, + "learning_rate": 1.916121966275117e-05, + "loss": 0.0028, + "step": 43560 + }, + { + "grad_norm": 0.16889889538288116, + "learning_rate": 1.9139532504074443e-05, + "loss": 0.0031, + "step": 43570 + }, + { + "grad_norm": 0.14365145564079285, + "learning_rate": 1.9117854719992885e-05, + "loss": 0.003, + "step": 43580 + }, + { + "grad_norm": 0.12228386849164963, + "learning_rate": 1.9096186317091687e-05, + "loss": 0.0025, + "step": 43590 + }, + { + "grad_norm": 0.14174194633960724, + "learning_rate": 1.9074527301953116e-05, + "loss": 0.0034, + "step": 43600 + }, + { + "grad_norm": 0.14558447897434235, + "learning_rate": 1.9052877681156607e-05, + "loss": 0.0031, + "step": 43610 + }, + { + "grad_norm": 0.16939060389995575, + "learning_rate": 1.903123746127875e-05, + "loss": 0.0029, + "step": 43620 + }, + { + "grad_norm": 0.1154969111084938, + "learning_rate": 1.900960664889327e-05, + "loss": 0.0023, + "step": 43630 + }, + { + "grad_norm": 0.11383353173732758, + "learning_rate": 1.8987985250571015e-05, + "loss": 0.0037, + "step": 43640 + }, + { + "grad_norm": 0.12167482078075409, + "learning_rate": 1.8966373272880054e-05, + "loss": 0.0037, + "step": 43650 + }, + { + "grad_norm": 0.11626416444778442, + "learning_rate": 1.8944770722385462e-05, + "loss": 0.0027, + "step": 43660 + }, + { + "grad_norm": 0.11196047812700272, + "learning_rate": 1.8923177605649576e-05, + "loss": 0.0025, + "step": 43670 + }, + { + "grad_norm": 0.129922553896904, + "learning_rate": 1.8901593929231802e-05, + "loss": 0.0036, + "step": 43680 + }, + { + "grad_norm": 0.11380775272846222, + "learning_rate": 1.8880019699688684e-05, + "loss": 0.0036, + "step": 43690 + }, + { + "grad_norm": 0.16569480299949646, + "learning_rate": 1.8858454923573904e-05, + "loss": 0.0037, + "step": 43700 + }, + { + "grad_norm": 0.15776319801807404, + "learning_rate": 1.8836899607438253e-05, + "loss": 0.0034, + "step": 43710 + }, + { + "grad_norm": 0.20473143458366394, + "learning_rate": 1.8815353757829723e-05, + "loss": 0.0044, + "step": 43720 + }, + { + "grad_norm": 0.09540721029043198, + "learning_rate": 1.879381738129331e-05, + "loss": 0.0022, + "step": 43730 + }, + { + "grad_norm": 0.1435556560754776, + "learning_rate": 1.8772290484371236e-05, + "loss": 0.0029, + "step": 43740 + }, + { + "grad_norm": 0.1187094897031784, + "learning_rate": 1.8750773073602795e-05, + "loss": 0.0029, + "step": 43750 + }, + { + "grad_norm": 0.16787496209144592, + "learning_rate": 1.8729265155524405e-05, + "loss": 0.0025, + "step": 43760 + }, + { + "grad_norm": 0.12616588175296783, + "learning_rate": 1.8707766736669607e-05, + "loss": 0.0021, + "step": 43770 + }, + { + "grad_norm": 0.14102788269519806, + "learning_rate": 1.8686277823569055e-05, + "loss": 0.0032, + "step": 43780 + }, + { + "grad_norm": 0.1212286725640297, + "learning_rate": 1.8664798422750484e-05, + "loss": 0.0033, + "step": 43790 + }, + { + "grad_norm": 0.1220666915178299, + "learning_rate": 1.8643328540738832e-05, + "loss": 0.005, + "step": 43800 + }, + { + "grad_norm": 0.11089291423559189, + "learning_rate": 1.862186818405601e-05, + "loss": 0.0018, + "step": 43810 + }, + { + "grad_norm": 0.1053742915391922, + "learning_rate": 1.8600417359221156e-05, + "loss": 0.0037, + "step": 43820 + }, + { + "grad_norm": 0.109604611992836, + "learning_rate": 1.8578976072750454e-05, + "loss": 0.0033, + "step": 43830 + }, + { + "grad_norm": 0.18026070296764374, + "learning_rate": 1.8557544331157194e-05, + "loss": 0.0026, + "step": 43840 + }, + { + "grad_norm": 0.21883097290992737, + "learning_rate": 1.8536122140951785e-05, + "loss": 0.0039, + "step": 43850 + }, + { + "grad_norm": 0.15479585528373718, + "learning_rate": 1.8514709508641688e-05, + "loss": 0.0044, + "step": 43860 + }, + { + "grad_norm": 0.16708962619304657, + "learning_rate": 1.8493306440731555e-05, + "loss": 0.004, + "step": 43870 + }, + { + "grad_norm": 0.15324416756629944, + "learning_rate": 1.8471912943723013e-05, + "loss": 0.0037, + "step": 43880 + }, + { + "grad_norm": 0.09840118885040283, + "learning_rate": 1.8450529024114894e-05, + "loss": 0.0021, + "step": 43890 + }, + { + "grad_norm": 0.10716372728347778, + "learning_rate": 1.842915468840301e-05, + "loss": 0.0032, + "step": 43900 + }, + { + "grad_norm": 0.12303082644939423, + "learning_rate": 1.840778994308037e-05, + "loss": 0.0033, + "step": 43910 + }, + { + "grad_norm": 0.13007058203220367, + "learning_rate": 1.8386434794637004e-05, + "loss": 0.0023, + "step": 43920 + }, + { + "grad_norm": 0.15132826566696167, + "learning_rate": 1.8365089249560034e-05, + "loss": 0.0023, + "step": 43930 + }, + { + "grad_norm": 0.09762345999479294, + "learning_rate": 1.8343753314333683e-05, + "loss": 0.0041, + "step": 43940 + }, + { + "grad_norm": 0.19953911006450653, + "learning_rate": 1.8322426995439236e-05, + "loss": 0.0048, + "step": 43950 + }, + { + "grad_norm": 0.10680358111858368, + "learning_rate": 1.8301110299355058e-05, + "loss": 0.0029, + "step": 43960 + }, + { + "grad_norm": 0.12017536908388138, + "learning_rate": 1.8279803232556625e-05, + "loss": 0.002, + "step": 43970 + }, + { + "grad_norm": 0.12636029720306396, + "learning_rate": 1.8258505801516444e-05, + "loss": 0.0036, + "step": 43980 + }, + { + "grad_norm": 0.1203877180814743, + "learning_rate": 1.8237218012704117e-05, + "loss": 0.0034, + "step": 43990 + }, + { + "grad_norm": 0.18636785447597504, + "learning_rate": 1.821593987258631e-05, + "loss": 0.0027, + "step": 44000 + }, + { + "grad_norm": 0.15933936834335327, + "learning_rate": 1.8194671387626744e-05, + "loss": 0.0025, + "step": 44010 + }, + { + "grad_norm": 0.14290201663970947, + "learning_rate": 1.8173412564286276e-05, + "loss": 0.0028, + "step": 44020 + }, + { + "grad_norm": 0.11299658566713333, + "learning_rate": 1.8152163409022697e-05, + "loss": 0.0026, + "step": 44030 + }, + { + "grad_norm": 0.10744449496269226, + "learning_rate": 1.8130923928291023e-05, + "loss": 0.0024, + "step": 44040 + }, + { + "grad_norm": 0.12385797500610352, + "learning_rate": 1.8109694128543163e-05, + "loss": 0.0019, + "step": 44050 + }, + { + "grad_norm": 0.1592237651348114, + "learning_rate": 1.8088474016228237e-05, + "loss": 0.0032, + "step": 44060 + }, + { + "grad_norm": 0.14553458988666534, + "learning_rate": 1.8067263597792328e-05, + "loss": 0.0032, + "step": 44070 + }, + { + "grad_norm": 0.15482890605926514, + "learning_rate": 1.80460628796786e-05, + "loss": 0.0027, + "step": 44080 + }, + { + "grad_norm": 0.09575201570987701, + "learning_rate": 1.8024871868327276e-05, + "loss": 0.0019, + "step": 44090 + }, + { + "grad_norm": 0.11266981065273285, + "learning_rate": 1.8003690570175608e-05, + "loss": 0.0025, + "step": 44100 + }, + { + "grad_norm": 0.1318572610616684, + "learning_rate": 1.7982518991657943e-05, + "loss": 0.0029, + "step": 44110 + }, + { + "grad_norm": 0.11195012927055359, + "learning_rate": 1.7961357139205643e-05, + "loss": 0.0027, + "step": 44120 + }, + { + "grad_norm": 0.09446470439434052, + "learning_rate": 1.7940205019247108e-05, + "loss": 0.0026, + "step": 44130 + }, + { + "grad_norm": 0.1283102184534073, + "learning_rate": 1.79190626382078e-05, + "loss": 0.003, + "step": 44140 + }, + { + "grad_norm": 0.14767558872699738, + "learning_rate": 1.7897930002510215e-05, + "loss": 0.0026, + "step": 44150 + }, + { + "grad_norm": 0.12476933747529984, + "learning_rate": 1.787680711857387e-05, + "loss": 0.0024, + "step": 44160 + }, + { + "grad_norm": 0.1404978185892105, + "learning_rate": 1.7855693992815398e-05, + "loss": 0.0038, + "step": 44170 + }, + { + "grad_norm": 0.1445080041885376, + "learning_rate": 1.7834590631648328e-05, + "loss": 0.0023, + "step": 44180 + }, + { + "grad_norm": 0.13157252967357635, + "learning_rate": 1.7813497041483384e-05, + "loss": 0.0023, + "step": 44190 + }, + { + "grad_norm": 0.1158943697810173, + "learning_rate": 1.779241322872817e-05, + "loss": 0.0037, + "step": 44200 + }, + { + "grad_norm": 0.09384501725435257, + "learning_rate": 1.777133919978744e-05, + "loss": 0.0024, + "step": 44210 + }, + { + "grad_norm": 0.09926478564739227, + "learning_rate": 1.7750274961062912e-05, + "loss": 0.0023, + "step": 44220 + }, + { + "grad_norm": 0.13296474516391754, + "learning_rate": 1.772922051895335e-05, + "loss": 0.0043, + "step": 44230 + }, + { + "grad_norm": 0.16705164313316345, + "learning_rate": 1.770817587985453e-05, + "loss": 0.0022, + "step": 44240 + }, + { + "grad_norm": 0.1257438212633133, + "learning_rate": 1.7687141050159246e-05, + "loss": 0.0033, + "step": 44250 + }, + { + "grad_norm": 0.1263522207736969, + "learning_rate": 1.7666116036257375e-05, + "loss": 0.0033, + "step": 44260 + }, + { + "grad_norm": 0.13838563859462738, + "learning_rate": 1.764510084453569e-05, + "loss": 0.0034, + "step": 44270 + }, + { + "grad_norm": 0.11271598190069199, + "learning_rate": 1.76240954813781e-05, + "loss": 0.0038, + "step": 44280 + }, + { + "grad_norm": 0.10200243443250656, + "learning_rate": 1.7603099953165476e-05, + "loss": 0.0037, + "step": 44290 + }, + { + "grad_norm": 0.1016218364238739, + "learning_rate": 1.7582114266275683e-05, + "loss": 0.0023, + "step": 44300 + }, + { + "grad_norm": 0.10926016420125961, + "learning_rate": 1.756113842708364e-05, + "loss": 0.0025, + "step": 44310 + }, + { + "grad_norm": 0.19211642444133759, + "learning_rate": 1.7540172441961245e-05, + "loss": 0.0039, + "step": 44320 + }, + { + "grad_norm": 0.17272014915943146, + "learning_rate": 1.7519216317277387e-05, + "loss": 0.0038, + "step": 44330 + }, + { + "grad_norm": 0.13535098731517792, + "learning_rate": 1.7498270059398046e-05, + "loss": 0.003, + "step": 44340 + }, + { + "grad_norm": 0.17224624752998352, + "learning_rate": 1.7477333674686062e-05, + "loss": 0.0031, + "step": 44350 + }, + { + "grad_norm": 0.10299841314554214, + "learning_rate": 1.745640716950142e-05, + "loss": 0.0026, + "step": 44360 + }, + { + "grad_norm": 0.07292947918176651, + "learning_rate": 1.7435490550201017e-05, + "loss": 0.0018, + "step": 44370 + }, + { + "grad_norm": 0.12707625329494476, + "learning_rate": 1.7414583823138762e-05, + "loss": 0.0031, + "step": 44380 + }, + { + "grad_norm": 0.18536223471164703, + "learning_rate": 1.739368699466558e-05, + "loss": 0.005, + "step": 44390 + }, + { + "grad_norm": 0.1214776486158371, + "learning_rate": 1.737280007112935e-05, + "loss": 0.002, + "step": 44400 + }, + { + "grad_norm": 0.14398516714572906, + "learning_rate": 1.735192305887502e-05, + "loss": 0.0033, + "step": 44410 + }, + { + "grad_norm": 0.1109752506017685, + "learning_rate": 1.733105596424441e-05, + "loss": 0.0025, + "step": 44420 + }, + { + "grad_norm": 0.1155310571193695, + "learning_rate": 1.7310198793576437e-05, + "loss": 0.0031, + "step": 44430 + }, + { + "grad_norm": 0.11506570130586624, + "learning_rate": 1.7289351553206952e-05, + "loss": 0.0033, + "step": 44440 + }, + { + "grad_norm": 0.10263778269290924, + "learning_rate": 1.7268514249468788e-05, + "loss": 0.0032, + "step": 44450 + }, + { + "grad_norm": 0.12946584820747375, + "learning_rate": 1.7247686888691765e-05, + "loss": 0.0027, + "step": 44460 + }, + { + "grad_norm": 0.12770754098892212, + "learning_rate": 1.7226869477202694e-05, + "loss": 0.004, + "step": 44470 + }, + { + "grad_norm": 0.16242554783821106, + "learning_rate": 1.7206062021325336e-05, + "loss": 0.0036, + "step": 44480 + }, + { + "grad_norm": 0.09878756105899811, + "learning_rate": 1.7185264527380502e-05, + "loss": 0.0018, + "step": 44490 + }, + { + "grad_norm": 0.18068929016590118, + "learning_rate": 1.716447700168584e-05, + "loss": 0.0026, + "step": 44500 + }, + { + "grad_norm": 0.11795713007450104, + "learning_rate": 1.714369945055611e-05, + "loss": 0.0022, + "step": 44510 + }, + { + "grad_norm": 0.12547971308231354, + "learning_rate": 1.7122931880302968e-05, + "loss": 0.0022, + "step": 44520 + }, + { + "grad_norm": 0.10362352430820465, + "learning_rate": 1.710217429723505e-05, + "loss": 0.0022, + "step": 44530 + }, + { + "grad_norm": 0.12436658889055252, + "learning_rate": 1.7081426707657972e-05, + "loss": 0.0029, + "step": 44540 + }, + { + "grad_norm": 0.09812696278095245, + "learning_rate": 1.7060689117874275e-05, + "loss": 0.0025, + "step": 44550 + }, + { + "grad_norm": 0.11503946036100388, + "learning_rate": 1.703996153418354e-05, + "loss": 0.0027, + "step": 44560 + }, + { + "grad_norm": 0.13398377597332, + "learning_rate": 1.7019243962882205e-05, + "loss": 0.0035, + "step": 44570 + }, + { + "grad_norm": 0.17513889074325562, + "learning_rate": 1.6998536410263754e-05, + "loss": 0.0041, + "step": 44580 + }, + { + "grad_norm": 0.15084140002727509, + "learning_rate": 1.6977838882618596e-05, + "loss": 0.0035, + "step": 44590 + }, + { + "grad_norm": 0.11744046211242676, + "learning_rate": 1.6957151386234088e-05, + "loss": 0.0029, + "step": 44600 + }, + { + "grad_norm": 0.14456751942634583, + "learning_rate": 1.6936473927394536e-05, + "loss": 0.0034, + "step": 44610 + }, + { + "grad_norm": 0.15840619802474976, + "learning_rate": 1.6915806512381222e-05, + "loss": 0.0036, + "step": 44620 + }, + { + "grad_norm": 0.1667514145374298, + "learning_rate": 1.6895149147472344e-05, + "loss": 0.0029, + "step": 44630 + }, + { + "grad_norm": 0.14782719314098358, + "learning_rate": 1.6874501838943073e-05, + "loss": 0.003, + "step": 44640 + }, + { + "grad_norm": 0.1551159769296646, + "learning_rate": 1.6853864593065506e-05, + "loss": 0.0039, + "step": 44650 + }, + { + "grad_norm": 0.11422882974147797, + "learning_rate": 1.683323741610871e-05, + "loss": 0.0028, + "step": 44660 + }, + { + "grad_norm": 0.120932936668396, + "learning_rate": 1.6812620314338674e-05, + "loss": 0.0029, + "step": 44670 + }, + { + "grad_norm": 0.15635259449481964, + "learning_rate": 1.6792013294018326e-05, + "loss": 0.0035, + "step": 44680 + }, + { + "grad_norm": 0.12792742252349854, + "learning_rate": 1.6771416361407526e-05, + "loss": 0.0028, + "step": 44690 + }, + { + "grad_norm": 0.12818844616413116, + "learning_rate": 1.675082952276308e-05, + "loss": 0.0033, + "step": 44700 + }, + { + "grad_norm": 0.13887102901935577, + "learning_rate": 1.6730252784338757e-05, + "loss": 0.0031, + "step": 44710 + }, + { + "grad_norm": 0.1204068586230278, + "learning_rate": 1.6709686152385166e-05, + "loss": 0.0028, + "step": 44720 + }, + { + "grad_norm": 0.1587284654378891, + "learning_rate": 1.668912963314998e-05, + "loss": 0.0025, + "step": 44730 + }, + { + "grad_norm": 0.10827206075191498, + "learning_rate": 1.6668583232877653e-05, + "loss": 0.0036, + "step": 44740 + }, + { + "grad_norm": 0.16987235844135284, + "learning_rate": 1.6648046957809698e-05, + "loss": 0.0034, + "step": 44750 + }, + { + "grad_norm": 0.10383376479148865, + "learning_rate": 1.6627520814184462e-05, + "loss": 0.0023, + "step": 44760 + }, + { + "grad_norm": 0.10195975005626678, + "learning_rate": 1.660700480823726e-05, + "loss": 0.0023, + "step": 44770 + }, + { + "grad_norm": 0.15911747515201569, + "learning_rate": 1.65864989462003e-05, + "loss": 0.004, + "step": 44780 + }, + { + "grad_norm": 0.12350042164325714, + "learning_rate": 1.656600323430273e-05, + "loss": 0.0021, + "step": 44790 + }, + { + "grad_norm": 0.10617683082818985, + "learning_rate": 1.654551767877059e-05, + "loss": 0.0033, + "step": 44800 + }, + { + "grad_norm": 0.11906342208385468, + "learning_rate": 1.6525042285826874e-05, + "loss": 0.002, + "step": 44810 + }, + { + "grad_norm": 0.09882324188947678, + "learning_rate": 1.6504577061691468e-05, + "loss": 0.0035, + "step": 44820 + }, + { + "grad_norm": 0.1142059937119484, + "learning_rate": 1.6484122012581143e-05, + "loss": 0.0056, + "step": 44830 + }, + { + "grad_norm": 0.10681496560573578, + "learning_rate": 1.6463677144709623e-05, + "loss": 0.0035, + "step": 44840 + }, + { + "grad_norm": 0.1290084272623062, + "learning_rate": 1.6443242464287493e-05, + "loss": 0.0025, + "step": 44850 + }, + { + "grad_norm": 0.1337740272283554, + "learning_rate": 1.642281797752232e-05, + "loss": 0.0032, + "step": 44860 + }, + { + "grad_norm": 0.11962473392486572, + "learning_rate": 1.6402403690618456e-05, + "loss": 0.0032, + "step": 44870 + }, + { + "grad_norm": 0.15387757122516632, + "learning_rate": 1.6381999609777295e-05, + "loss": 0.0028, + "step": 44880 + }, + { + "grad_norm": 0.10399586707353592, + "learning_rate": 1.6361605741196983e-05, + "loss": 0.002, + "step": 44890 + }, + { + "grad_norm": 0.1211291179060936, + "learning_rate": 1.63412220910727e-05, + "loss": 0.0029, + "step": 44900 + }, + { + "grad_norm": 0.1669653207063675, + "learning_rate": 1.6320848665596433e-05, + "loss": 0.003, + "step": 44910 + }, + { + "grad_norm": 0.09013701975345612, + "learning_rate": 1.6300485470957095e-05, + "loss": 0.0026, + "step": 44920 + }, + { + "grad_norm": 0.15025624632835388, + "learning_rate": 1.6280132513340483e-05, + "loss": 0.0064, + "step": 44930 + }, + { + "grad_norm": 0.141093909740448, + "learning_rate": 1.62597897989293e-05, + "loss": 0.0029, + "step": 44940 + }, + { + "grad_norm": 0.15742972493171692, + "learning_rate": 1.623945733390309e-05, + "loss": 0.0026, + "step": 44950 + }, + { + "grad_norm": 0.1692814975976944, + "learning_rate": 1.6219135124438374e-05, + "loss": 0.0025, + "step": 44960 + }, + { + "grad_norm": 0.09857185184955597, + "learning_rate": 1.6198823176708465e-05, + "loss": 0.0019, + "step": 44970 + }, + { + "grad_norm": 0.14189870655536652, + "learning_rate": 1.6178521496883613e-05, + "loss": 0.0028, + "step": 44980 + }, + { + "grad_norm": 0.1352986991405487, + "learning_rate": 1.6158230091130926e-05, + "loss": 0.0052, + "step": 44990 + }, + { + "grad_norm": 0.13977892696857452, + "learning_rate": 1.613794896561438e-05, + "loss": 0.0024, + "step": 45000 + }, + { + "grad_norm": 0.12037627398967743, + "learning_rate": 1.6117678126494894e-05, + "loss": 0.0027, + "step": 45010 + }, + { + "grad_norm": 0.11894038319587708, + "learning_rate": 1.6097417579930153e-05, + "loss": 0.003, + "step": 45020 + }, + { + "grad_norm": 0.11856943368911743, + "learning_rate": 1.6077167332074834e-05, + "loss": 0.0031, + "step": 45030 + }, + { + "grad_norm": 0.1582639366388321, + "learning_rate": 1.605692738908037e-05, + "loss": 0.0033, + "step": 45040 + }, + { + "grad_norm": 0.1551557034254074, + "learning_rate": 1.6036697757095176e-05, + "loss": 0.0039, + "step": 45050 + }, + { + "grad_norm": 0.1763678640127182, + "learning_rate": 1.6016478442264428e-05, + "loss": 0.0033, + "step": 45060 + }, + { + "grad_norm": 0.11375752091407776, + "learning_rate": 1.599626945073026e-05, + "loss": 0.0023, + "step": 45070 + }, + { + "grad_norm": 0.14403843879699707, + "learning_rate": 1.597607078863162e-05, + "loss": 0.0031, + "step": 45080 + }, + { + "grad_norm": 0.13409040868282318, + "learning_rate": 1.595588246210432e-05, + "loss": 0.002, + "step": 45090 + }, + { + "grad_norm": 0.12490037828683853, + "learning_rate": 1.5935704477281048e-05, + "loss": 0.0029, + "step": 45100 + }, + { + "grad_norm": 0.11826036125421524, + "learning_rate": 1.5915536840291323e-05, + "loss": 0.0035, + "step": 45110 + }, + { + "grad_norm": 0.11063234508037567, + "learning_rate": 1.5895379557261576e-05, + "loss": 0.0033, + "step": 45120 + }, + { + "grad_norm": 0.1400458812713623, + "learning_rate": 1.5875232634315033e-05, + "loss": 0.0026, + "step": 45130 + }, + { + "grad_norm": 0.1644551157951355, + "learning_rate": 1.5855096077571812e-05, + "loss": 0.0038, + "step": 45140 + }, + { + "grad_norm": 0.14464645087718964, + "learning_rate": 1.5834969893148855e-05, + "loss": 0.0031, + "step": 45150 + }, + { + "grad_norm": 0.14953793585300446, + "learning_rate": 1.581485408715997e-05, + "loss": 0.0033, + "step": 45160 + }, + { + "grad_norm": 0.16871076822280884, + "learning_rate": 1.5794748665715785e-05, + "loss": 0.0031, + "step": 45170 + }, + { + "grad_norm": 0.11060309410095215, + "learning_rate": 1.5774653634923857e-05, + "loss": 0.0042, + "step": 45180 + }, + { + "grad_norm": 0.11836708337068558, + "learning_rate": 1.575456900088845e-05, + "loss": 0.0026, + "step": 45190 + }, + { + "grad_norm": 0.15805479884147644, + "learning_rate": 1.5734494769710816e-05, + "loss": 0.0041, + "step": 45200 + }, + { + "grad_norm": 0.1307588368654251, + "learning_rate": 1.5714430947488912e-05, + "loss": 0.0025, + "step": 45210 + }, + { + "grad_norm": 0.14632175862789154, + "learning_rate": 1.5694377540317645e-05, + "loss": 0.002, + "step": 45220 + }, + { + "grad_norm": 0.11879955977201462, + "learning_rate": 1.5674334554288694e-05, + "loss": 0.0024, + "step": 45230 + }, + { + "grad_norm": 0.1097344234585762, + "learning_rate": 1.5654301995490582e-05, + "loss": 0.003, + "step": 45240 + }, + { + "grad_norm": 0.13567927479743958, + "learning_rate": 1.5634279870008685e-05, + "loss": 0.0049, + "step": 45250 + }, + { + "grad_norm": 0.1696065068244934, + "learning_rate": 1.5614268183925174e-05, + "loss": 0.0029, + "step": 45260 + }, + { + "grad_norm": 0.142481729388237, + "learning_rate": 1.5594266943319097e-05, + "loss": 0.0034, + "step": 45270 + }, + { + "grad_norm": 0.12296462804079056, + "learning_rate": 1.5574276154266294e-05, + "loss": 0.0032, + "step": 45280 + }, + { + "grad_norm": 0.13821595907211304, + "learning_rate": 1.5554295822839437e-05, + "loss": 0.0032, + "step": 45290 + }, + { + "grad_norm": 0.18884176015853882, + "learning_rate": 1.5534325955108025e-05, + "loss": 0.0037, + "step": 45300 + }, + { + "grad_norm": 0.1347174197435379, + "learning_rate": 1.5514366557138373e-05, + "loss": 0.0022, + "step": 45310 + }, + { + "grad_norm": 0.1188732236623764, + "learning_rate": 1.5494417634993602e-05, + "loss": 0.0022, + "step": 45320 + }, + { + "grad_norm": 0.12445495277643204, + "learning_rate": 1.547447919473372e-05, + "loss": 0.0026, + "step": 45330 + }, + { + "grad_norm": 0.12494923919439316, + "learning_rate": 1.5454551242415434e-05, + "loss": 0.0026, + "step": 45340 + }, + { + "grad_norm": 0.14090991020202637, + "learning_rate": 1.543463378409239e-05, + "loss": 0.0024, + "step": 45350 + }, + { + "grad_norm": 0.11229303479194641, + "learning_rate": 1.541472682581493e-05, + "loss": 0.0025, + "step": 45360 + }, + { + "grad_norm": 0.12597060203552246, + "learning_rate": 1.5394830373630298e-05, + "loss": 0.0033, + "step": 45370 + }, + { + "grad_norm": 0.19662624597549438, + "learning_rate": 1.5374944433582506e-05, + "loss": 0.003, + "step": 45380 + }, + { + "grad_norm": 0.17440171539783478, + "learning_rate": 1.5355069011712375e-05, + "loss": 0.0031, + "step": 45390 + }, + { + "grad_norm": 0.18217696249485016, + "learning_rate": 1.5335204114057526e-05, + "loss": 0.0047, + "step": 45400 + }, + { + "grad_norm": 0.13746672868728638, + "learning_rate": 1.5315349746652387e-05, + "loss": 0.0028, + "step": 45410 + }, + { + "grad_norm": 0.16245613992214203, + "learning_rate": 1.5295505915528212e-05, + "loss": 0.0022, + "step": 45420 + }, + { + "grad_norm": 0.08981003612279892, + "learning_rate": 1.5275672626713024e-05, + "loss": 0.002, + "step": 45430 + }, + { + "grad_norm": 0.13700507581233978, + "learning_rate": 1.5255849886231643e-05, + "loss": 0.0028, + "step": 45440 + }, + { + "grad_norm": 0.12492302060127258, + "learning_rate": 1.523603770010571e-05, + "loss": 0.0047, + "step": 45450 + }, + { + "grad_norm": 0.13143570721149445, + "learning_rate": 1.521623607435363e-05, + "loss": 0.0022, + "step": 45460 + }, + { + "grad_norm": 0.12899400293827057, + "learning_rate": 1.5196445014990612e-05, + "loss": 0.0032, + "step": 45470 + }, + { + "grad_norm": 0.08515872806310654, + "learning_rate": 1.5176664528028672e-05, + "loss": 0.0027, + "step": 45480 + }, + { + "grad_norm": 0.11894240975379944, + "learning_rate": 1.5156894619476574e-05, + "loss": 0.0029, + "step": 45490 + }, + { + "grad_norm": 0.1269761621952057, + "learning_rate": 1.5137135295339938e-05, + "loss": 0.0022, + "step": 45500 + }, + { + "grad_norm": 0.15310552716255188, + "learning_rate": 1.5117386561621073e-05, + "loss": 0.0026, + "step": 45510 + }, + { + "grad_norm": 0.17156606912612915, + "learning_rate": 1.5097648424319167e-05, + "loss": 0.004, + "step": 45520 + }, + { + "grad_norm": 0.12597014009952545, + "learning_rate": 1.5077920889430119e-05, + "loss": 0.0024, + "step": 45530 + }, + { + "grad_norm": 0.09204896539449692, + "learning_rate": 1.5058203962946644e-05, + "loss": 0.0029, + "step": 45540 + }, + { + "grad_norm": 0.09744611382484436, + "learning_rate": 1.503849765085822e-05, + "loss": 0.0021, + "step": 45550 + }, + { + "grad_norm": 0.14112231135368347, + "learning_rate": 1.501880195915109e-05, + "loss": 0.0029, + "step": 45560 + }, + { + "grad_norm": 0.09753841161727905, + "learning_rate": 1.499911689380833e-05, + "loss": 0.0032, + "step": 45570 + }, + { + "grad_norm": 0.12400601804256439, + "learning_rate": 1.4979442460809683e-05, + "loss": 0.0026, + "step": 45580 + }, + { + "grad_norm": 0.09993378818035126, + "learning_rate": 1.4959778666131763e-05, + "loss": 0.0033, + "step": 45590 + }, + { + "grad_norm": 0.12758682668209076, + "learning_rate": 1.4940125515747905e-05, + "loss": 0.0048, + "step": 45600 + }, + { + "grad_norm": 0.09667088836431503, + "learning_rate": 1.4920483015628211e-05, + "loss": 0.0028, + "step": 45610 + }, + { + "grad_norm": 0.103914774954319, + "learning_rate": 1.490085117173956e-05, + "loss": 0.0027, + "step": 45620 + }, + { + "grad_norm": 0.08405863493680954, + "learning_rate": 1.488122999004558e-05, + "loss": 0.0018, + "step": 45630 + }, + { + "grad_norm": 0.13954687118530273, + "learning_rate": 1.486161947650666e-05, + "loss": 0.0037, + "step": 45640 + }, + { + "grad_norm": 0.11454354226589203, + "learning_rate": 1.4842019637079995e-05, + "loss": 0.0024, + "step": 45650 + }, + { + "grad_norm": 0.12007882446050644, + "learning_rate": 1.482243047771944e-05, + "loss": 0.002, + "step": 45660 + }, + { + "grad_norm": 0.0748705193400383, + "learning_rate": 1.4802852004375712e-05, + "loss": 0.0023, + "step": 45670 + }, + { + "grad_norm": 0.06864310055971146, + "learning_rate": 1.4783284222996218e-05, + "loss": 0.0028, + "step": 45680 + }, + { + "grad_norm": 0.16428780555725098, + "learning_rate": 1.4763727139525135e-05, + "loss": 0.0031, + "step": 45690 + }, + { + "grad_norm": 0.1233481839299202, + "learning_rate": 1.4744180759903392e-05, + "loss": 0.0022, + "step": 45700 + }, + { + "grad_norm": 0.13660785555839539, + "learning_rate": 1.4724645090068635e-05, + "loss": 0.0038, + "step": 45710 + }, + { + "grad_norm": 0.2045556902885437, + "learning_rate": 1.4705120135955341e-05, + "loss": 0.0024, + "step": 45720 + }, + { + "grad_norm": 0.1555570662021637, + "learning_rate": 1.4685605903494614e-05, + "loss": 0.0043, + "step": 45730 + }, + { + "grad_norm": 0.11550207436084747, + "learning_rate": 1.46661023986144e-05, + "loss": 0.0031, + "step": 45740 + }, + { + "grad_norm": 0.13337044417858124, + "learning_rate": 1.4646609627239344e-05, + "loss": 0.0049, + "step": 45750 + }, + { + "grad_norm": 0.17174091935157776, + "learning_rate": 1.4627127595290835e-05, + "loss": 0.0032, + "step": 45760 + }, + { + "grad_norm": 0.17372578382492065, + "learning_rate": 1.460765630868699e-05, + "loss": 0.004, + "step": 45770 + }, + { + "grad_norm": 0.1773354560136795, + "learning_rate": 1.4588195773342678e-05, + "loss": 0.0038, + "step": 45780 + }, + { + "grad_norm": 0.14584480226039886, + "learning_rate": 1.4568745995169485e-05, + "loss": 0.003, + "step": 45790 + }, + { + "grad_norm": 0.18555361032485962, + "learning_rate": 1.4549306980075778e-05, + "loss": 0.0056, + "step": 45800 + }, + { + "grad_norm": 0.1179942637681961, + "learning_rate": 1.4529878733966557e-05, + "loss": 0.0024, + "step": 45810 + }, + { + "grad_norm": 0.12227556109428406, + "learning_rate": 1.4510461262743658e-05, + "loss": 0.0025, + "step": 45820 + }, + { + "grad_norm": 0.10669393837451935, + "learning_rate": 1.4491054572305585e-05, + "loss": 0.0028, + "step": 45830 + }, + { + "grad_norm": 0.1617264449596405, + "learning_rate": 1.4471658668547566e-05, + "loss": 0.0034, + "step": 45840 + }, + { + "grad_norm": 0.1140698790550232, + "learning_rate": 1.4452273557361579e-05, + "loss": 0.0034, + "step": 45850 + }, + { + "grad_norm": 0.17291259765625, + "learning_rate": 1.4432899244636282e-05, + "loss": 0.003, + "step": 45860 + }, + { + "grad_norm": 0.1903328001499176, + "learning_rate": 1.4413535736257134e-05, + "loss": 0.0044, + "step": 45870 + }, + { + "grad_norm": 0.23108699917793274, + "learning_rate": 1.439418303810619e-05, + "loss": 0.0036, + "step": 45880 + }, + { + "grad_norm": 0.17733709514141083, + "learning_rate": 1.4374841156062352e-05, + "loss": 0.0025, + "step": 45890 + }, + { + "grad_norm": 0.10132022202014923, + "learning_rate": 1.4355510096001112e-05, + "loss": 0.0023, + "step": 45900 + }, + { + "grad_norm": 0.09983125329017639, + "learning_rate": 1.4336189863794786e-05, + "loss": 0.0026, + "step": 45910 + }, + { + "grad_norm": 0.141797736287117, + "learning_rate": 1.4316880465312327e-05, + "loss": 0.0036, + "step": 45920 + }, + { + "grad_norm": 0.10112389177083969, + "learning_rate": 1.4297581906419426e-05, + "loss": 0.0023, + "step": 45930 + }, + { + "grad_norm": 0.13054825365543365, + "learning_rate": 1.4278294192978475e-05, + "loss": 0.002, + "step": 45940 + }, + { + "grad_norm": 0.16906709969043732, + "learning_rate": 1.4259017330848574e-05, + "loss": 0.0022, + "step": 45950 + }, + { + "grad_norm": 0.12141149491071701, + "learning_rate": 1.4239751325885498e-05, + "loss": 0.0024, + "step": 45960 + }, + { + "grad_norm": 0.09307458996772766, + "learning_rate": 1.4220496183941795e-05, + "loss": 0.0027, + "step": 45970 + }, + { + "grad_norm": 0.1287001073360443, + "learning_rate": 1.4201251910866648e-05, + "loss": 0.0023, + "step": 45980 + }, + { + "grad_norm": 0.1382073163986206, + "learning_rate": 1.4182018512505957e-05, + "loss": 0.0027, + "step": 45990 + }, + { + "grad_norm": 0.14326943457126617, + "learning_rate": 1.4162795994702327e-05, + "loss": 0.0041, + "step": 46000 + }, + { + "grad_norm": 0.1298130750656128, + "learning_rate": 1.4143584363295032e-05, + "loss": 0.0026, + "step": 46010 + }, + { + "grad_norm": 0.10548557341098785, + "learning_rate": 1.4124383624120101e-05, + "loss": 0.0023, + "step": 46020 + }, + { + "grad_norm": 0.10755957663059235, + "learning_rate": 1.4105193783010151e-05, + "loss": 0.0025, + "step": 46030 + }, + { + "grad_norm": 0.15985740721225739, + "learning_rate": 1.4086014845794621e-05, + "loss": 0.0033, + "step": 46040 + }, + { + "grad_norm": 0.13729777932167053, + "learning_rate": 1.4066846818299489e-05, + "loss": 0.0041, + "step": 46050 + }, + { + "grad_norm": 0.13783517479896545, + "learning_rate": 1.4047689706347555e-05, + "loss": 0.0037, + "step": 46060 + }, + { + "grad_norm": 0.14581967890262604, + "learning_rate": 1.402854351575822e-05, + "loss": 0.003, + "step": 46070 + }, + { + "grad_norm": 0.12558075785636902, + "learning_rate": 1.4009408252347588e-05, + "loss": 0.0035, + "step": 46080 + }, + { + "grad_norm": 0.14792941510677338, + "learning_rate": 1.399028392192846e-05, + "loss": 0.0023, + "step": 46090 + }, + { + "grad_norm": 0.09443023055791855, + "learning_rate": 1.397117053031029e-05, + "loss": 0.0024, + "step": 46100 + }, + { + "grad_norm": 0.11221737414598465, + "learning_rate": 1.3952068083299213e-05, + "loss": 0.0027, + "step": 46110 + }, + { + "grad_norm": 0.1332656890153885, + "learning_rate": 1.3932976586698082e-05, + "loss": 0.0036, + "step": 46120 + }, + { + "grad_norm": 0.1910703182220459, + "learning_rate": 1.3913896046306363e-05, + "loss": 0.0036, + "step": 46130 + }, + { + "grad_norm": 0.1155548170208931, + "learning_rate": 1.389482646792023e-05, + "loss": 0.003, + "step": 46140 + }, + { + "grad_norm": 0.10850763320922852, + "learning_rate": 1.387576785733251e-05, + "loss": 0.0021, + "step": 46150 + }, + { + "grad_norm": 0.09422829747200012, + "learning_rate": 1.3856720220332703e-05, + "loss": 0.0024, + "step": 46160 + }, + { + "grad_norm": 0.1184690073132515, + "learning_rate": 1.383768356270701e-05, + "loss": 0.0025, + "step": 46170 + }, + { + "grad_norm": 0.11434689909219742, + "learning_rate": 1.3818657890238207e-05, + "loss": 0.0022, + "step": 46180 + }, + { + "grad_norm": 0.14526799321174622, + "learning_rate": 1.3799643208705859e-05, + "loss": 0.003, + "step": 46190 + }, + { + "grad_norm": 0.14587469398975372, + "learning_rate": 1.3780639523886058e-05, + "loss": 0.003, + "step": 46200 + }, + { + "grad_norm": 0.10978133231401443, + "learning_rate": 1.3761646841551668e-05, + "loss": 0.0025, + "step": 46210 + }, + { + "grad_norm": 0.06338368356227875, + "learning_rate": 1.3742665167472146e-05, + "loss": 0.0018, + "step": 46220 + }, + { + "grad_norm": 0.13494592905044556, + "learning_rate": 1.372369450741363e-05, + "loss": 0.0028, + "step": 46230 + }, + { + "grad_norm": 0.1185169517993927, + "learning_rate": 1.3704734867138901e-05, + "loss": 0.0021, + "step": 46240 + }, + { + "grad_norm": 0.08597885817289352, + "learning_rate": 1.36857862524074e-05, + "loss": 0.0021, + "step": 46250 + }, + { + "grad_norm": 0.09294528514146805, + "learning_rate": 1.3666848668975213e-05, + "loss": 0.0019, + "step": 46260 + }, + { + "grad_norm": 0.10184340924024582, + "learning_rate": 1.3647922122595063e-05, + "loss": 0.002, + "step": 46270 + }, + { + "grad_norm": 0.10174739360809326, + "learning_rate": 1.3629006619016366e-05, + "loss": 0.002, + "step": 46280 + }, + { + "grad_norm": 0.11188867688179016, + "learning_rate": 1.3610102163985139e-05, + "loss": 0.0021, + "step": 46290 + }, + { + "grad_norm": 0.12490741908550262, + "learning_rate": 1.3591208763244057e-05, + "loss": 0.0023, + "step": 46300 + }, + { + "grad_norm": 0.143723726272583, + "learning_rate": 1.3572326422532428e-05, + "loss": 0.0029, + "step": 46310 + }, + { + "grad_norm": 0.1327531784772873, + "learning_rate": 1.355345514758622e-05, + "loss": 0.003, + "step": 46320 + }, + { + "grad_norm": 0.1411413550376892, + "learning_rate": 1.3534594944138007e-05, + "loss": 0.0028, + "step": 46330 + }, + { + "grad_norm": 0.15518443286418915, + "learning_rate": 1.3515745817917069e-05, + "loss": 0.0041, + "step": 46340 + }, + { + "grad_norm": 0.09662064164876938, + "learning_rate": 1.3496907774649208e-05, + "loss": 0.0042, + "step": 46350 + }, + { + "grad_norm": 0.1721542775630951, + "learning_rate": 1.3478080820056987e-05, + "loss": 0.003, + "step": 46360 + }, + { + "grad_norm": 0.11976107209920883, + "learning_rate": 1.3459264959859474e-05, + "loss": 0.0033, + "step": 46370 + }, + { + "grad_norm": 0.19108441472053528, + "learning_rate": 1.3440460199772487e-05, + "loss": 0.0024, + "step": 46380 + }, + { + "grad_norm": 0.12142565846443176, + "learning_rate": 1.3421666545508382e-05, + "loss": 0.0034, + "step": 46390 + }, + { + "grad_norm": 0.1365564912557602, + "learning_rate": 1.3402884002776194e-05, + "loss": 0.0022, + "step": 46400 + }, + { + "grad_norm": 0.100091852247715, + "learning_rate": 1.3384112577281555e-05, + "loss": 0.0037, + "step": 46410 + }, + { + "grad_norm": 0.1566309630870819, + "learning_rate": 1.3365352274726711e-05, + "loss": 0.003, + "step": 46420 + }, + { + "grad_norm": 0.1333821415901184, + "learning_rate": 1.3346603100810578e-05, + "loss": 0.0032, + "step": 46430 + }, + { + "grad_norm": 0.10575415194034576, + "learning_rate": 1.3327865061228645e-05, + "loss": 0.0029, + "step": 46440 + }, + { + "grad_norm": 0.14837437868118286, + "learning_rate": 1.330913816167304e-05, + "loss": 0.0025, + "step": 46450 + }, + { + "grad_norm": 0.12326712161302567, + "learning_rate": 1.3290422407832492e-05, + "loss": 0.0032, + "step": 46460 + }, + { + "grad_norm": 0.10345402359962463, + "learning_rate": 1.3271717805392354e-05, + "loss": 0.0027, + "step": 46470 + }, + { + "grad_norm": 0.12002383917570114, + "learning_rate": 1.3253024360034582e-05, + "loss": 0.004, + "step": 46480 + }, + { + "grad_norm": 0.09263564646244049, + "learning_rate": 1.323434207743779e-05, + "loss": 0.0025, + "step": 46490 + }, + { + "grad_norm": 0.10615521669387817, + "learning_rate": 1.3215670963277105e-05, + "loss": 0.0018, + "step": 46500 + }, + { + "grad_norm": 0.15661312639713287, + "learning_rate": 1.3197011023224376e-05, + "loss": 0.0035, + "step": 46510 + }, + { + "grad_norm": 0.1299227774143219, + "learning_rate": 1.3178362262947941e-05, + "loss": 0.0021, + "step": 46520 + }, + { + "grad_norm": 0.14572419226169586, + "learning_rate": 1.3159724688112845e-05, + "loss": 0.0028, + "step": 46530 + }, + { + "grad_norm": 0.13466009497642517, + "learning_rate": 1.3141098304380683e-05, + "loss": 0.002, + "step": 46540 + }, + { + "grad_norm": 0.1126943975687027, + "learning_rate": 1.3122483117409651e-05, + "loss": 0.0029, + "step": 46550 + }, + { + "grad_norm": 0.14645519852638245, + "learning_rate": 1.3103879132854552e-05, + "loss": 0.0022, + "step": 46560 + }, + { + "grad_norm": 0.10317640006542206, + "learning_rate": 1.3085286356366771e-05, + "loss": 0.0026, + "step": 46570 + }, + { + "grad_norm": 0.1310623437166214, + "learning_rate": 1.3066704793594337e-05, + "loss": 0.0029, + "step": 46580 + }, + { + "grad_norm": 0.12525010108947754, + "learning_rate": 1.3048134450181816e-05, + "loss": 0.0022, + "step": 46590 + }, + { + "grad_norm": 0.126299187541008, + "learning_rate": 1.3029575331770394e-05, + "loss": 0.0022, + "step": 46600 + }, + { + "grad_norm": 0.07357867062091827, + "learning_rate": 1.3011027443997837e-05, + "loss": 0.0026, + "step": 46610 + }, + { + "grad_norm": 0.10681796818971634, + "learning_rate": 1.2992490792498507e-05, + "loss": 0.0034, + "step": 46620 + }, + { + "grad_norm": 0.16927707195281982, + "learning_rate": 1.297396538290333e-05, + "loss": 0.0028, + "step": 46630 + }, + { + "grad_norm": 0.10481976717710495, + "learning_rate": 1.2955451220839888e-05, + "loss": 0.0018, + "step": 46640 + }, + { + "grad_norm": 0.12929436564445496, + "learning_rate": 1.2936948311932223e-05, + "loss": 0.0025, + "step": 46650 + }, + { + "grad_norm": 0.08394632488489151, + "learning_rate": 1.2918456661801104e-05, + "loss": 0.0023, + "step": 46660 + }, + { + "grad_norm": 0.11800488084554672, + "learning_rate": 1.2899976276063736e-05, + "loss": 0.0025, + "step": 46670 + }, + { + "grad_norm": 0.10719099640846252, + "learning_rate": 1.2881507160334022e-05, + "loss": 0.0033, + "step": 46680 + }, + { + "grad_norm": 0.1297837644815445, + "learning_rate": 1.286304932022238e-05, + "loss": 0.0032, + "step": 46690 + }, + { + "grad_norm": 0.14928150177001953, + "learning_rate": 1.2844602761335806e-05, + "loss": 0.0048, + "step": 46700 + }, + { + "grad_norm": 0.10335986316204071, + "learning_rate": 1.2826167489277885e-05, + "loss": 0.0023, + "step": 46710 + }, + { + "grad_norm": 0.12226234376430511, + "learning_rate": 1.2807743509648745e-05, + "loss": 0.0029, + "step": 46720 + }, + { + "grad_norm": 0.1272212266921997, + "learning_rate": 1.2789330828045149e-05, + "loss": 0.0035, + "step": 46730 + }, + { + "grad_norm": 0.1568411886692047, + "learning_rate": 1.2770929450060332e-05, + "loss": 0.0024, + "step": 46740 + }, + { + "grad_norm": 0.12614105641841888, + "learning_rate": 1.2752539381284184e-05, + "loss": 0.0032, + "step": 46750 + }, + { + "grad_norm": 0.13119417428970337, + "learning_rate": 1.273416062730311e-05, + "loss": 0.0023, + "step": 46760 + }, + { + "grad_norm": 0.12374354153871536, + "learning_rate": 1.2715793193700088e-05, + "loss": 0.0025, + "step": 46770 + }, + { + "grad_norm": 0.08981849253177643, + "learning_rate": 1.2697437086054664e-05, + "loss": 0.0021, + "step": 46780 + }, + { + "grad_norm": 0.10945316404104233, + "learning_rate": 1.2679092309942937e-05, + "loss": 0.0027, + "step": 46790 + }, + { + "grad_norm": 0.21132531762123108, + "learning_rate": 1.266075887093755e-05, + "loss": 0.0033, + "step": 46800 + }, + { + "grad_norm": 0.11576346307992935, + "learning_rate": 1.2642436774607757e-05, + "loss": 0.0023, + "step": 46810 + }, + { + "grad_norm": 0.15515871345996857, + "learning_rate": 1.2624126026519278e-05, + "loss": 0.0023, + "step": 46820 + }, + { + "grad_norm": 0.09275900572538376, + "learning_rate": 1.2605826632234474e-05, + "loss": 0.0031, + "step": 46830 + }, + { + "grad_norm": 0.11182555556297302, + "learning_rate": 1.2587538597312198e-05, + "loss": 0.0019, + "step": 46840 + }, + { + "grad_norm": 0.08751679956912994, + "learning_rate": 1.2569261927307884e-05, + "loss": 0.0018, + "step": 46850 + }, + { + "grad_norm": 0.12337523698806763, + "learning_rate": 1.2550996627773493e-05, + "loss": 0.0027, + "step": 46860 + }, + { + "grad_norm": 0.07998523116111755, + "learning_rate": 1.2532742704257527e-05, + "loss": 0.0016, + "step": 46870 + }, + { + "grad_norm": 0.1518530547618866, + "learning_rate": 1.2514500162305087e-05, + "loss": 0.0041, + "step": 46880 + }, + { + "grad_norm": 0.08548180758953094, + "learning_rate": 1.2496269007457728e-05, + "loss": 0.0016, + "step": 46890 + }, + { + "grad_norm": 0.12792423367500305, + "learning_rate": 1.2478049245253625e-05, + "loss": 0.002, + "step": 46900 + }, + { + "grad_norm": 0.11787933856248856, + "learning_rate": 1.2459840881227459e-05, + "loss": 0.0022, + "step": 46910 + }, + { + "grad_norm": 0.0767868235707283, + "learning_rate": 1.2441643920910435e-05, + "loss": 0.0017, + "step": 46920 + }, + { + "grad_norm": 0.07796741276979446, + "learning_rate": 1.2423458369830322e-05, + "loss": 0.0028, + "step": 46930 + }, + { + "grad_norm": 0.11237072199583054, + "learning_rate": 1.2405284233511406e-05, + "loss": 0.0027, + "step": 46940 + }, + { + "grad_norm": 0.1674765795469284, + "learning_rate": 1.2387121517474487e-05, + "loss": 0.0032, + "step": 46950 + }, + { + "grad_norm": 0.13526573777198792, + "learning_rate": 1.2368970227236975e-05, + "loss": 0.0028, + "step": 46960 + }, + { + "grad_norm": 0.09588257968425751, + "learning_rate": 1.2350830368312688e-05, + "loss": 0.0025, + "step": 46970 + }, + { + "grad_norm": 0.16969755291938782, + "learning_rate": 1.2332701946212083e-05, + "loss": 0.004, + "step": 46980 + }, + { + "grad_norm": 0.1334334909915924, + "learning_rate": 1.2314584966442077e-05, + "loss": 0.002, + "step": 46990 + }, + { + "grad_norm": 0.10175280272960663, + "learning_rate": 1.2296479434506136e-05, + "loss": 0.0016, + "step": 47000 + }, + { + "grad_norm": 0.11347930878400803, + "learning_rate": 1.2278385355904232e-05, + "loss": 0.0025, + "step": 47010 + }, + { + "grad_norm": 0.11240120232105255, + "learning_rate": 1.2260302736132867e-05, + "loss": 0.0022, + "step": 47020 + }, + { + "grad_norm": 0.12349376827478409, + "learning_rate": 1.2242231580685098e-05, + "loss": 0.0025, + "step": 47030 + }, + { + "grad_norm": 0.19221635162830353, + "learning_rate": 1.2224171895050413e-05, + "loss": 0.004, + "step": 47040 + }, + { + "grad_norm": 0.1625974476337433, + "learning_rate": 1.2206123684714903e-05, + "loss": 0.0026, + "step": 47050 + }, + { + "grad_norm": 0.13469119369983673, + "learning_rate": 1.2188086955161132e-05, + "loss": 0.0029, + "step": 47060 + }, + { + "grad_norm": 0.08854442089796066, + "learning_rate": 1.2170061711868175e-05, + "loss": 0.0029, + "step": 47070 + }, + { + "grad_norm": 0.08915482461452484, + "learning_rate": 1.215204796031163e-05, + "loss": 0.002, + "step": 47080 + }, + { + "grad_norm": 0.1259669065475464, + "learning_rate": 1.2134045705963599e-05, + "loss": 0.0024, + "step": 47090 + }, + { + "grad_norm": 0.10597129911184311, + "learning_rate": 1.2116054954292689e-05, + "loss": 0.0034, + "step": 47100 + }, + { + "grad_norm": 0.1195705458521843, + "learning_rate": 1.2098075710764011e-05, + "loss": 0.0027, + "step": 47110 + }, + { + "grad_norm": 0.1130211278796196, + "learning_rate": 1.2080107980839183e-05, + "loss": 0.0023, + "step": 47120 + }, + { + "grad_norm": 0.1250617951154709, + "learning_rate": 1.2062151769976343e-05, + "loss": 0.0025, + "step": 47130 + }, + { + "grad_norm": 0.1296345293521881, + "learning_rate": 1.204420708363011e-05, + "loss": 0.0024, + "step": 47140 + }, + { + "grad_norm": 0.1487611085176468, + "learning_rate": 1.2026273927251597e-05, + "loss": 0.003, + "step": 47150 + }, + { + "grad_norm": 0.1340097188949585, + "learning_rate": 1.2008352306288424e-05, + "loss": 0.003, + "step": 47160 + }, + { + "grad_norm": 0.12724849581718445, + "learning_rate": 1.1990442226184695e-05, + "loss": 0.0018, + "step": 47170 + }, + { + "grad_norm": 0.14960534870624542, + "learning_rate": 1.1972543692381066e-05, + "loss": 0.0033, + "step": 47180 + }, + { + "grad_norm": 0.12741266191005707, + "learning_rate": 1.1954656710314576e-05, + "loss": 0.0024, + "step": 47190 + }, + { + "grad_norm": 0.10716425627470016, + "learning_rate": 1.1936781285418875e-05, + "loss": 0.0045, + "step": 47200 + }, + { + "grad_norm": 0.1180964782834053, + "learning_rate": 1.1918917423123993e-05, + "loss": 0.003, + "step": 47210 + }, + { + "grad_norm": 0.1549995392560959, + "learning_rate": 1.1901065128856537e-05, + "loss": 0.0031, + "step": 47220 + }, + { + "grad_norm": 0.1273275911808014, + "learning_rate": 1.1883224408039551e-05, + "loss": 0.0023, + "step": 47230 + }, + { + "grad_norm": 0.08809608221054077, + "learning_rate": 1.1865395266092578e-05, + "loss": 0.0029, + "step": 47240 + }, + { + "grad_norm": 0.12624306976795197, + "learning_rate": 1.1847577708431633e-05, + "loss": 0.0025, + "step": 47250 + }, + { + "grad_norm": 0.10553889721632004, + "learning_rate": 1.1829771740469225e-05, + "loss": 0.0024, + "step": 47260 + }, + { + "grad_norm": 0.12091819196939468, + "learning_rate": 1.1811977367614324e-05, + "loss": 0.0038, + "step": 47270 + }, + { + "grad_norm": 0.116727814078331, + "learning_rate": 1.1794194595272412e-05, + "loss": 0.0024, + "step": 47280 + }, + { + "grad_norm": 0.1217162162065506, + "learning_rate": 1.1776423428845423e-05, + "loss": 0.0031, + "step": 47290 + }, + { + "grad_norm": 0.09838945418596268, + "learning_rate": 1.1758663873731756e-05, + "loss": 0.0016, + "step": 47300 + }, + { + "grad_norm": 0.08561576157808304, + "learning_rate": 1.1740915935326302e-05, + "loss": 0.0022, + "step": 47310 + }, + { + "grad_norm": 0.09892681241035461, + "learning_rate": 1.1723179619020396e-05, + "loss": 0.0027, + "step": 47320 + }, + { + "grad_norm": 0.10146355628967285, + "learning_rate": 1.1705454930201914e-05, + "loss": 0.0018, + "step": 47330 + }, + { + "grad_norm": 0.13942992687225342, + "learning_rate": 1.1687741874255087e-05, + "loss": 0.0022, + "step": 47340 + }, + { + "grad_norm": 0.1388404220342636, + "learning_rate": 1.1670040456560728e-05, + "loss": 0.004, + "step": 47350 + }, + { + "grad_norm": 0.10307937860488892, + "learning_rate": 1.1652350682496005e-05, + "loss": 0.003, + "step": 47360 + }, + { + "grad_norm": 0.12538784742355347, + "learning_rate": 1.163467255743465e-05, + "loss": 0.0023, + "step": 47370 + }, + { + "grad_norm": 0.1075529232621193, + "learning_rate": 1.1617006086746796e-05, + "loss": 0.0027, + "step": 47380 + }, + { + "grad_norm": 0.10925360023975372, + "learning_rate": 1.1599351275799047e-05, + "loss": 0.0027, + "step": 47390 + }, + { + "grad_norm": 0.15213201940059662, + "learning_rate": 1.1581708129954466e-05, + "loss": 0.0022, + "step": 47400 + }, + { + "grad_norm": 0.22908484935760498, + "learning_rate": 1.1564076654572587e-05, + "loss": 0.0032, + "step": 47410 + }, + { + "grad_norm": 0.12508384883403778, + "learning_rate": 1.1546456855009358e-05, + "loss": 0.0027, + "step": 47420 + }, + { + "grad_norm": 0.12114439904689789, + "learning_rate": 1.1528848736617248e-05, + "loss": 0.0025, + "step": 47430 + }, + { + "grad_norm": 0.1250842660665512, + "learning_rate": 1.1511252304745112e-05, + "loss": 0.0024, + "step": 47440 + }, + { + "grad_norm": 0.11026407033205032, + "learning_rate": 1.1493667564738297e-05, + "loss": 0.0024, + "step": 47450 + }, + { + "grad_norm": 0.16509422659873962, + "learning_rate": 1.1476094521938574e-05, + "loss": 0.0038, + "step": 47460 + }, + { + "grad_norm": 0.11542849242687225, + "learning_rate": 1.1458533181684167e-05, + "loss": 0.002, + "step": 47470 + }, + { + "grad_norm": 0.142963707447052, + "learning_rate": 1.1440983549309753e-05, + "loss": 0.0032, + "step": 47480 + }, + { + "grad_norm": 0.10162800550460815, + "learning_rate": 1.1423445630146434e-05, + "loss": 0.0023, + "step": 47490 + }, + { + "grad_norm": 0.11569596081972122, + "learning_rate": 1.1405919429521799e-05, + "loss": 0.0024, + "step": 47500 + }, + { + "grad_norm": 0.12443464249372482, + "learning_rate": 1.1388404952759802e-05, + "loss": 0.0024, + "step": 47510 + }, + { + "grad_norm": 0.08241724967956543, + "learning_rate": 1.1370902205180923e-05, + "loss": 0.0018, + "step": 47520 + }, + { + "grad_norm": 0.09481204301118851, + "learning_rate": 1.1353411192101987e-05, + "loss": 0.0019, + "step": 47530 + }, + { + "grad_norm": 0.1298399120569229, + "learning_rate": 1.133593191883634e-05, + "loss": 0.0028, + "step": 47540 + }, + { + "grad_norm": 0.11450059711933136, + "learning_rate": 1.1318464390693711e-05, + "loss": 0.0022, + "step": 47550 + }, + { + "grad_norm": 0.1073218435049057, + "learning_rate": 1.1301008612980257e-05, + "loss": 0.0028, + "step": 47560 + }, + { + "grad_norm": 0.10504373162984848, + "learning_rate": 1.128356459099863e-05, + "loss": 0.0025, + "step": 47570 + }, + { + "grad_norm": 0.135113924741745, + "learning_rate": 1.1266132330047802e-05, + "loss": 0.0033, + "step": 47580 + }, + { + "grad_norm": 0.12440724670886993, + "learning_rate": 1.1248711835423281e-05, + "loss": 0.0024, + "step": 47590 + }, + { + "grad_norm": 0.08071115612983704, + "learning_rate": 1.123130311241693e-05, + "loss": 0.0015, + "step": 47600 + }, + { + "grad_norm": 0.09911885112524033, + "learning_rate": 1.1213906166317068e-05, + "loss": 0.0049, + "step": 47610 + }, + { + "grad_norm": 0.0713583305478096, + "learning_rate": 1.1196521002408427e-05, + "loss": 0.0039, + "step": 47620 + }, + { + "grad_norm": 0.08284076303243637, + "learning_rate": 1.1179147625972159e-05, + "loss": 0.0018, + "step": 47630 + }, + { + "grad_norm": 0.08442492038011551, + "learning_rate": 1.1161786042285822e-05, + "loss": 0.0022, + "step": 47640 + }, + { + "grad_norm": 0.10332383215427399, + "learning_rate": 1.1144436256623447e-05, + "loss": 0.0029, + "step": 47650 + }, + { + "grad_norm": 0.08446500450372696, + "learning_rate": 1.1127098274255392e-05, + "loss": 0.0019, + "step": 47660 + }, + { + "grad_norm": 0.11181280761957169, + "learning_rate": 1.1109772100448512e-05, + "loss": 0.0022, + "step": 47670 + }, + { + "grad_norm": 0.0828646644949913, + "learning_rate": 1.1092457740466033e-05, + "loss": 0.0021, + "step": 47680 + }, + { + "grad_norm": 0.1020873636007309, + "learning_rate": 1.10751551995676e-05, + "loss": 0.0021, + "step": 47690 + }, + { + "grad_norm": 0.09450958669185638, + "learning_rate": 1.1057864483009262e-05, + "loss": 0.0018, + "step": 47700 + }, + { + "grad_norm": 0.16156481206417084, + "learning_rate": 1.1040585596043473e-05, + "loss": 0.0025, + "step": 47710 + }, + { + "grad_norm": 0.12269661575555801, + "learning_rate": 1.1023318543919148e-05, + "loss": 0.0021, + "step": 47720 + }, + { + "grad_norm": 0.105267733335495, + "learning_rate": 1.10060633318815e-05, + "loss": 0.0025, + "step": 47730 + }, + { + "grad_norm": 0.12487838417291641, + "learning_rate": 1.0988819965172248e-05, + "loss": 0.0032, + "step": 47740 + }, + { + "grad_norm": 0.1382634937763214, + "learning_rate": 1.0971588449029462e-05, + "loss": 0.0035, + "step": 47750 + }, + { + "grad_norm": 0.08704044669866562, + "learning_rate": 1.095436878868762e-05, + "loss": 0.0021, + "step": 47760 + }, + { + "grad_norm": 0.12546859681606293, + "learning_rate": 1.0937160989377598e-05, + "loss": 0.002, + "step": 47770 + }, + { + "grad_norm": 0.09371322393417358, + "learning_rate": 1.0919965056326676e-05, + "loss": 0.0022, + "step": 47780 + }, + { + "grad_norm": 0.13431386649608612, + "learning_rate": 1.0902780994758504e-05, + "loss": 0.0024, + "step": 47790 + }, + { + "grad_norm": 0.10775809735059738, + "learning_rate": 1.0885608809893193e-05, + "loss": 0.0039, + "step": 47800 + }, + { + "grad_norm": 0.1157064437866211, + "learning_rate": 1.0868448506947142e-05, + "loss": 0.0023, + "step": 47810 + }, + { + "grad_norm": 0.10234591364860535, + "learning_rate": 1.0851300091133243e-05, + "loss": 0.0026, + "step": 47820 + }, + { + "grad_norm": 0.14829765260219574, + "learning_rate": 1.083416356766071e-05, + "loss": 0.0023, + "step": 47830 + }, + { + "grad_norm": 0.07441359758377075, + "learning_rate": 1.0817038941735175e-05, + "loss": 0.0018, + "step": 47840 + }, + { + "grad_norm": 0.15241529047489166, + "learning_rate": 1.0799926218558642e-05, + "loss": 0.0024, + "step": 47850 + }, + { + "grad_norm": 0.09379903972148895, + "learning_rate": 1.0782825403329488e-05, + "loss": 0.0019, + "step": 47860 + }, + { + "grad_norm": 0.13571493327617645, + "learning_rate": 1.076573650124254e-05, + "loss": 0.0028, + "step": 47870 + }, + { + "grad_norm": 0.17612747848033905, + "learning_rate": 1.0748659517488891e-05, + "loss": 0.0033, + "step": 47880 + }, + { + "grad_norm": 0.17345304787158966, + "learning_rate": 1.0731594457256138e-05, + "loss": 0.0039, + "step": 47890 + }, + { + "grad_norm": 0.11342008411884308, + "learning_rate": 1.0714541325728139e-05, + "loss": 0.0023, + "step": 47900 + }, + { + "grad_norm": 0.1162128821015358, + "learning_rate": 1.0697500128085231e-05, + "loss": 0.0022, + "step": 47910 + }, + { + "grad_norm": 0.12873190641403198, + "learning_rate": 1.0680470869504055e-05, + "loss": 0.0025, + "step": 47920 + }, + { + "grad_norm": 0.11322584003210068, + "learning_rate": 1.066345355515766e-05, + "loss": 0.0033, + "step": 47930 + }, + { + "grad_norm": 0.11463586241006851, + "learning_rate": 1.0646448190215453e-05, + "loss": 0.002, + "step": 47940 + }, + { + "grad_norm": 0.08118673413991928, + "learning_rate": 1.0629454779843217e-05, + "loss": 0.0029, + "step": 47950 + }, + { + "grad_norm": 0.11862143129110336, + "learning_rate": 1.0612473329203082e-05, + "loss": 0.0029, + "step": 47960 + }, + { + "grad_norm": 0.12997980415821075, + "learning_rate": 1.0595503843453596e-05, + "loss": 0.0038, + "step": 47970 + }, + { + "grad_norm": 0.1317794770002365, + "learning_rate": 1.0578546327749634e-05, + "loss": 0.0024, + "step": 47980 + }, + { + "grad_norm": 0.10178518295288086, + "learning_rate": 1.0561600787242425e-05, + "loss": 0.004, + "step": 47990 + }, + { + "grad_norm": 0.09716015309095383, + "learning_rate": 1.0544667227079591e-05, + "loss": 0.0029, + "step": 48000 + }, + { + "grad_norm": 0.12273216247558594, + "learning_rate": 1.0527745652405085e-05, + "loss": 0.0022, + "step": 48010 + }, + { + "grad_norm": 0.10136248171329498, + "learning_rate": 1.051083606835927e-05, + "loss": 0.0023, + "step": 48020 + }, + { + "grad_norm": 0.084026038646698, + "learning_rate": 1.049393848007878e-05, + "loss": 0.0027, + "step": 48030 + }, + { + "grad_norm": 0.11330205947160721, + "learning_rate": 1.0477052892696709e-05, + "loss": 0.002, + "step": 48040 + }, + { + "grad_norm": 0.08676755428314209, + "learning_rate": 1.0460179311342394e-05, + "loss": 0.0022, + "step": 48050 + }, + { + "grad_norm": 0.09153099358081818, + "learning_rate": 1.0443317741141634e-05, + "loss": 0.0024, + "step": 48060 + }, + { + "grad_norm": 0.11852974444627762, + "learning_rate": 1.0426468187216514e-05, + "loss": 0.0027, + "step": 48070 + }, + { + "grad_norm": 0.1029120534658432, + "learning_rate": 1.0409630654685477e-05, + "loss": 0.0018, + "step": 48080 + }, + { + "grad_norm": 0.11213282495737076, + "learning_rate": 1.039280514866332e-05, + "loss": 0.0033, + "step": 48090 + }, + { + "grad_norm": 0.11281349509954453, + "learning_rate": 1.0375991674261198e-05, + "loss": 0.0041, + "step": 48100 + }, + { + "grad_norm": 0.11131361871957779, + "learning_rate": 1.0359190236586575e-05, + "loss": 0.0024, + "step": 48110 + }, + { + "grad_norm": 0.10350585728883743, + "learning_rate": 1.0342400840743322e-05, + "loss": 0.0025, + "step": 48120 + }, + { + "grad_norm": 0.1123080626130104, + "learning_rate": 1.0325623491831593e-05, + "loss": 0.0024, + "step": 48130 + }, + { + "grad_norm": 0.08169658482074738, + "learning_rate": 1.0308858194947906e-05, + "loss": 0.002, + "step": 48140 + }, + { + "grad_norm": 0.10029498487710953, + "learning_rate": 1.0292104955185111e-05, + "loss": 0.003, + "step": 48150 + }, + { + "grad_norm": 0.08469241112470627, + "learning_rate": 1.0275363777632396e-05, + "loss": 0.0019, + "step": 48160 + }, + { + "grad_norm": 0.10511776059865952, + "learning_rate": 1.0258634667375321e-05, + "loss": 0.002, + "step": 48170 + }, + { + "grad_norm": 0.09187564998865128, + "learning_rate": 1.02419176294957e-05, + "loss": 0.0025, + "step": 48180 + }, + { + "grad_norm": 0.08772830665111542, + "learning_rate": 1.0225212669071782e-05, + "loss": 0.002, + "step": 48190 + }, + { + "grad_norm": 0.10771210491657257, + "learning_rate": 1.0208519791178029e-05, + "loss": 0.0022, + "step": 48200 + }, + { + "grad_norm": 0.08751191943883896, + "learning_rate": 1.019183900088535e-05, + "loss": 0.0022, + "step": 48210 + }, + { + "grad_norm": 0.08429927378892899, + "learning_rate": 1.0175170303260906e-05, + "loss": 0.0019, + "step": 48220 + }, + { + "grad_norm": 0.11151190847158432, + "learning_rate": 1.0158513703368206e-05, + "loss": 0.0026, + "step": 48230 + }, + { + "grad_norm": 0.10996969044208527, + "learning_rate": 1.0141869206267095e-05, + "loss": 0.0033, + "step": 48240 + }, + { + "grad_norm": 0.11298898607492447, + "learning_rate": 1.0125236817013723e-05, + "loss": 0.003, + "step": 48250 + }, + { + "grad_norm": 0.11981301009654999, + "learning_rate": 1.010861654066056e-05, + "loss": 0.0023, + "step": 48260 + }, + { + "grad_norm": 0.11962216347455978, + "learning_rate": 1.0092008382256434e-05, + "loss": 0.0024, + "step": 48270 + }, + { + "grad_norm": 0.10287439078092575, + "learning_rate": 1.0075412346846458e-05, + "loss": 0.0021, + "step": 48280 + }, + { + "grad_norm": 0.08729320019483566, + "learning_rate": 1.0058828439472056e-05, + "loss": 0.002, + "step": 48290 + }, + { + "grad_norm": 0.15193483233451843, + "learning_rate": 1.0042256665170996e-05, + "loss": 0.0026, + "step": 48300 + }, + { + "grad_norm": 0.14248016476631165, + "learning_rate": 1.0025697028977332e-05, + "loss": 0.0034, + "step": 48310 + }, + { + "grad_norm": 0.13116611540317535, + "learning_rate": 1.0009149535921454e-05, + "loss": 0.0033, + "step": 48320 + }, + { + "grad_norm": 0.11773749440908432, + "learning_rate": 9.992614191030031e-06, + "loss": 0.0019, + "step": 48330 + }, + { + "grad_norm": 0.1203927993774414, + "learning_rate": 9.976090999326115e-06, + "loss": 0.0025, + "step": 48340 + }, + { + "grad_norm": 0.09188685566186905, + "learning_rate": 9.959579965828952e-06, + "loss": 0.002, + "step": 48350 + }, + { + "grad_norm": 0.15167517960071564, + "learning_rate": 9.943081095554218e-06, + "loss": 0.0038, + "step": 48360 + }, + { + "grad_norm": 0.1040549948811531, + "learning_rate": 9.926594393513783e-06, + "loss": 0.0034, + "step": 48370 + }, + { + "grad_norm": 0.15255528688430786, + "learning_rate": 9.910119864715906e-06, + "loss": 0.0026, + "step": 48380 + }, + { + "grad_norm": 0.11402716487646103, + "learning_rate": 9.8936575141651e-06, + "loss": 0.0028, + "step": 48390 + }, + { + "grad_norm": 0.11590831726789474, + "learning_rate": 9.877207346862194e-06, + "loss": 0.0026, + "step": 48400 + }, + { + "grad_norm": 0.07933210581541061, + "learning_rate": 9.860769367804312e-06, + "loss": 0.0019, + "step": 48410 + }, + { + "grad_norm": 0.0818268209695816, + "learning_rate": 9.844343581984877e-06, + "loss": 0.0022, + "step": 48420 + }, + { + "grad_norm": 0.09188905358314514, + "learning_rate": 9.82792999439362e-06, + "loss": 0.002, + "step": 48430 + }, + { + "grad_norm": 0.1119978129863739, + "learning_rate": 9.811528610016546e-06, + "loss": 0.0035, + "step": 48440 + }, + { + "grad_norm": 0.10409771651029587, + "learning_rate": 9.79513943383597e-06, + "loss": 0.0019, + "step": 48450 + }, + { + "grad_norm": 0.09029753506183624, + "learning_rate": 9.778762470830489e-06, + "loss": 0.0028, + "step": 48460 + }, + { + "grad_norm": 0.11212477833032608, + "learning_rate": 9.762397725974982e-06, + "loss": 0.002, + "step": 48470 + }, + { + "grad_norm": 0.11352349072694778, + "learning_rate": 9.746045204240622e-06, + "loss": 0.0017, + "step": 48480 + }, + { + "grad_norm": 0.10118387639522552, + "learning_rate": 9.729704910594917e-06, + "loss": 0.0033, + "step": 48490 + }, + { + "grad_norm": 0.0925510972738266, + "learning_rate": 9.713376850001554e-06, + "loss": 0.0028, + "step": 48500 + }, + { + "grad_norm": 0.1536151021718979, + "learning_rate": 9.697061027420622e-06, + "loss": 0.0035, + "step": 48510 + }, + { + "grad_norm": 0.13699395954608917, + "learning_rate": 9.680757447808385e-06, + "loss": 0.005, + "step": 48520 + }, + { + "grad_norm": 0.14559189975261688, + "learning_rate": 9.664466116117488e-06, + "loss": 0.0031, + "step": 48530 + }, + { + "grad_norm": 0.13472463190555573, + "learning_rate": 9.64818703729678e-06, + "loss": 0.0026, + "step": 48540 + }, + { + "grad_norm": 0.10501401126384735, + "learning_rate": 9.631920216291423e-06, + "loss": 0.0022, + "step": 48550 + }, + { + "grad_norm": 0.12361995875835419, + "learning_rate": 9.615665658042849e-06, + "loss": 0.0029, + "step": 48560 + }, + { + "grad_norm": 0.09984498471021652, + "learning_rate": 9.599423367488747e-06, + "loss": 0.0017, + "step": 48570 + }, + { + "grad_norm": 0.12647372484207153, + "learning_rate": 9.583193349563124e-06, + "loss": 0.0023, + "step": 48580 + }, + { + "grad_norm": 0.13131414353847504, + "learning_rate": 9.566975609196216e-06, + "loss": 0.0026, + "step": 48590 + }, + { + "grad_norm": 0.11458467692136765, + "learning_rate": 9.550770151314548e-06, + "loss": 0.0021, + "step": 48600 + }, + { + "grad_norm": 0.12321095168590546, + "learning_rate": 9.53457698084091e-06, + "loss": 0.0037, + "step": 48610 + }, + { + "grad_norm": 0.12712985277175903, + "learning_rate": 9.518396102694355e-06, + "loss": 0.0021, + "step": 48620 + }, + { + "grad_norm": 0.15009577572345734, + "learning_rate": 9.502227521790198e-06, + "loss": 0.0019, + "step": 48630 + }, + { + "grad_norm": 0.11379413306713104, + "learning_rate": 9.486071243040063e-06, + "loss": 0.0024, + "step": 48640 + }, + { + "grad_norm": 0.149822399020195, + "learning_rate": 9.469927271351747e-06, + "loss": 0.0037, + "step": 48650 + }, + { + "grad_norm": 0.09868599474430084, + "learning_rate": 9.453795611629419e-06, + "loss": 0.0021, + "step": 48660 + }, + { + "grad_norm": 0.13636882603168488, + "learning_rate": 9.437676268773399e-06, + "loss": 0.0025, + "step": 48670 + }, + { + "grad_norm": 0.09216433018445969, + "learning_rate": 9.421569247680357e-06, + "loss": 0.0019, + "step": 48680 + }, + { + "grad_norm": 0.07814184576272964, + "learning_rate": 9.40547455324316e-06, + "loss": 0.0018, + "step": 48690 + }, + { + "grad_norm": 0.12194598466157913, + "learning_rate": 9.389392190350965e-06, + "loss": 0.0034, + "step": 48700 + }, + { + "grad_norm": 0.10378128290176392, + "learning_rate": 9.373322163889153e-06, + "loss": 0.0028, + "step": 48710 + }, + { + "grad_norm": 0.0951819121837616, + "learning_rate": 9.357264478739375e-06, + "loss": 0.0019, + "step": 48720 + }, + { + "grad_norm": 0.1617048680782318, + "learning_rate": 9.341219139779567e-06, + "loss": 0.0029, + "step": 48730 + }, + { + "grad_norm": 0.11434527486562729, + "learning_rate": 9.325186151883824e-06, + "loss": 0.0018, + "step": 48740 + }, + { + "grad_norm": 0.10056687146425247, + "learning_rate": 9.30916551992258e-06, + "loss": 0.0021, + "step": 48750 + }, + { + "grad_norm": 0.13249000906944275, + "learning_rate": 9.293157248762479e-06, + "loss": 0.0018, + "step": 48760 + }, + { + "grad_norm": 0.13988246023654938, + "learning_rate": 9.2771613432664e-06, + "loss": 0.0025, + "step": 48770 + }, + { + "grad_norm": 0.1329503208398819, + "learning_rate": 9.261177808293481e-06, + "loss": 0.0029, + "step": 48780 + }, + { + "grad_norm": 0.11927880346775055, + "learning_rate": 9.245206648699096e-06, + "loss": 0.0023, + "step": 48790 + }, + { + "grad_norm": 0.12103567272424698, + "learning_rate": 9.22924786933485e-06, + "loss": 0.0018, + "step": 48800 + }, + { + "grad_norm": 0.10801254957914352, + "learning_rate": 9.213301475048642e-06, + "loss": 0.0018, + "step": 48810 + }, + { + "grad_norm": 0.09657008945941925, + "learning_rate": 9.197367470684504e-06, + "loss": 0.0022, + "step": 48820 + }, + { + "grad_norm": 0.10457472503185272, + "learning_rate": 9.181445861082816e-06, + "loss": 0.0028, + "step": 48830 + }, + { + "grad_norm": 0.14161191880702972, + "learning_rate": 9.16553665108012e-06, + "loss": 0.0048, + "step": 48840 + }, + { + "grad_norm": 0.11951187252998352, + "learning_rate": 9.149639845509223e-06, + "loss": 0.0042, + "step": 48850 + }, + { + "grad_norm": 0.14019037783145905, + "learning_rate": 9.133755449199144e-06, + "loss": 0.003, + "step": 48860 + }, + { + "grad_norm": 0.128916397690773, + "learning_rate": 9.117883466975135e-06, + "loss": 0.0026, + "step": 48870 + }, + { + "grad_norm": 0.1488710343837738, + "learning_rate": 9.10202390365873e-06, + "loss": 0.0025, + "step": 48880 + }, + { + "grad_norm": 0.11417711526155472, + "learning_rate": 9.086176764067583e-06, + "loss": 0.0018, + "step": 48890 + }, + { + "grad_norm": 0.11061181128025055, + "learning_rate": 9.070342053015684e-06, + "loss": 0.0031, + "step": 48900 + }, + { + "grad_norm": 0.08189591765403748, + "learning_rate": 9.054519775313187e-06, + "loss": 0.0018, + "step": 48910 + }, + { + "grad_norm": 0.09690409898757935, + "learning_rate": 9.038709935766476e-06, + "loss": 0.0021, + "step": 48920 + }, + { + "grad_norm": 0.06724182516336441, + "learning_rate": 9.02291253917817e-06, + "loss": 0.0017, + "step": 48930 + }, + { + "grad_norm": 0.12101121991872787, + "learning_rate": 9.007127590347091e-06, + "loss": 0.0039, + "step": 48940 + }, + { + "grad_norm": 0.08156918734312057, + "learning_rate": 8.991355094068288e-06, + "loss": 0.0028, + "step": 48950 + }, + { + "grad_norm": 0.07716041803359985, + "learning_rate": 8.975595055133062e-06, + "loss": 0.0018, + "step": 48960 + }, + { + "grad_norm": 0.07253929227590561, + "learning_rate": 8.959847478328848e-06, + "loss": 0.0023, + "step": 48970 + }, + { + "grad_norm": 0.11566736549139023, + "learning_rate": 8.944112368439378e-06, + "loss": 0.0024, + "step": 48980 + }, + { + "grad_norm": 0.07291276007890701, + "learning_rate": 8.928389730244552e-06, + "loss": 0.002, + "step": 48990 + }, + { + "grad_norm": 0.15503861010074615, + "learning_rate": 8.912679568520494e-06, + "loss": 0.0035, + "step": 49000 + }, + { + "grad_norm": 0.09338108450174332, + "learning_rate": 8.896981888039534e-06, + "loss": 0.002, + "step": 49010 + }, + { + "grad_norm": 0.1186317428946495, + "learning_rate": 8.881296693570201e-06, + "loss": 0.0021, + "step": 49020 + }, + { + "grad_norm": 0.09802825003862381, + "learning_rate": 8.865623989877281e-06, + "loss": 0.003, + "step": 49030 + }, + { + "grad_norm": 0.09408789128065109, + "learning_rate": 8.849963781721681e-06, + "loss": 0.0024, + "step": 49040 + }, + { + "grad_norm": 0.09908238798379898, + "learning_rate": 8.834316073860588e-06, + "loss": 0.0031, + "step": 49050 + }, + { + "grad_norm": 0.07211890816688538, + "learning_rate": 8.818680871047357e-06, + "loss": 0.0016, + "step": 49060 + }, + { + "grad_norm": 0.07321076095104218, + "learning_rate": 8.803058178031549e-06, + "loss": 0.0016, + "step": 49070 + }, + { + "grad_norm": 0.12595193088054657, + "learning_rate": 8.787447999558922e-06, + "loss": 0.0055, + "step": 49080 + }, + { + "grad_norm": 0.11518527567386627, + "learning_rate": 8.77185034037144e-06, + "loss": 0.0016, + "step": 49090 + }, + { + "grad_norm": 0.11184072494506836, + "learning_rate": 8.756265205207259e-06, + "loss": 0.0018, + "step": 49100 + }, + { + "grad_norm": 0.15995003283023834, + "learning_rate": 8.740692598800732e-06, + "loss": 0.0036, + "step": 49110 + }, + { + "grad_norm": 0.13087674975395203, + "learning_rate": 8.72513252588239e-06, + "loss": 0.0015, + "step": 49120 + }, + { + "grad_norm": 0.10785440355539322, + "learning_rate": 8.709584991178998e-06, + "loss": 0.0017, + "step": 49130 + }, + { + "grad_norm": 0.07876037061214447, + "learning_rate": 8.694049999413479e-06, + "loss": 0.0029, + "step": 49140 + }, + { + "grad_norm": 0.1258798986673355, + "learning_rate": 8.678527555304945e-06, + "loss": 0.002, + "step": 49150 + }, + { + "grad_norm": 0.16029977798461914, + "learning_rate": 8.663017663568712e-06, + "loss": 0.0034, + "step": 49160 + }, + { + "grad_norm": 0.09392597526311874, + "learning_rate": 8.647520328916259e-06, + "loss": 0.0022, + "step": 49170 + }, + { + "grad_norm": 0.07043136656284332, + "learning_rate": 8.632035556055307e-06, + "loss": 0.0023, + "step": 49180 + }, + { + "grad_norm": 0.1295660138130188, + "learning_rate": 8.616563349689672e-06, + "loss": 0.0026, + "step": 49190 + }, + { + "grad_norm": 0.11023680865764618, + "learning_rate": 8.601103714519448e-06, + "loss": 0.0043, + "step": 49200 + }, + { + "grad_norm": 0.11860691756010056, + "learning_rate": 8.58565665524082e-06, + "loss": 0.0024, + "step": 49210 + }, + { + "grad_norm": 0.10789217054843903, + "learning_rate": 8.570222176546222e-06, + "loss": 0.0024, + "step": 49220 + }, + { + "grad_norm": 0.09861379116773605, + "learning_rate": 8.554800283124242e-06, + "loss": 0.0023, + "step": 49230 + }, + { + "grad_norm": 0.11307317763566971, + "learning_rate": 8.539390979659639e-06, + "loss": 0.0026, + "step": 49240 + }, + { + "grad_norm": 0.17224572598934174, + "learning_rate": 8.523994270833352e-06, + "loss": 0.0037, + "step": 49250 + }, + { + "grad_norm": 0.11919666826725006, + "learning_rate": 8.5086101613225e-06, + "loss": 0.0019, + "step": 49260 + }, + { + "grad_norm": 0.1128411591053009, + "learning_rate": 8.493238655800346e-06, + "loss": 0.0026, + "step": 49270 + }, + { + "grad_norm": 0.08248217403888702, + "learning_rate": 8.47787975893638e-06, + "loss": 0.0021, + "step": 49280 + }, + { + "grad_norm": 0.07551082223653793, + "learning_rate": 8.462533475396211e-06, + "loss": 0.0019, + "step": 49290 + }, + { + "grad_norm": 0.10485408455133438, + "learning_rate": 8.447199809841643e-06, + "loss": 0.003, + "step": 49300 + }, + { + "grad_norm": 0.10779853910207748, + "learning_rate": 8.431878766930635e-06, + "loss": 0.0017, + "step": 49310 + }, + { + "grad_norm": 0.102345772087574, + "learning_rate": 8.416570351317304e-06, + "loss": 0.0017, + "step": 49320 + }, + { + "grad_norm": 0.10136539489030838, + "learning_rate": 8.401274567651973e-06, + "loss": 0.0021, + "step": 49330 + }, + { + "grad_norm": 0.10500148683786392, + "learning_rate": 8.385991420581058e-06, + "loss": 0.0035, + "step": 49340 + }, + { + "grad_norm": 0.07619517296552658, + "learning_rate": 8.370720914747215e-06, + "loss": 0.0021, + "step": 49350 + }, + { + "grad_norm": 0.1480957418680191, + "learning_rate": 8.355463054789181e-06, + "loss": 0.0019, + "step": 49360 + }, + { + "grad_norm": 0.08972138166427612, + "learning_rate": 8.340217845341919e-06, + "loss": 0.0023, + "step": 49370 + }, + { + "grad_norm": 0.17312607169151306, + "learning_rate": 8.324985291036514e-06, + "loss": 0.0029, + "step": 49380 + }, + { + "grad_norm": 0.1291695237159729, + "learning_rate": 8.309765396500213e-06, + "loss": 0.0035, + "step": 49390 + }, + { + "grad_norm": 0.07369519025087357, + "learning_rate": 8.294558166356419e-06, + "loss": 0.0017, + "step": 49400 + }, + { + "grad_norm": 0.11528974026441574, + "learning_rate": 8.279363605224683e-06, + "loss": 0.0031, + "step": 49410 + }, + { + "grad_norm": 0.07922019809484482, + "learning_rate": 8.264181717720704e-06, + "loss": 0.0036, + "step": 49420 + }, + { + "grad_norm": 0.11961455643177032, + "learning_rate": 8.249012508456361e-06, + "loss": 0.0023, + "step": 49430 + }, + { + "grad_norm": 0.11232568323612213, + "learning_rate": 8.233855982039646e-06, + "loss": 0.002, + "step": 49440 + }, + { + "grad_norm": 0.06828329712152481, + "learning_rate": 8.218712143074708e-06, + "loss": 0.0017, + "step": 49450 + }, + { + "grad_norm": 0.09473037719726562, + "learning_rate": 8.203580996161858e-06, + "loss": 0.0023, + "step": 49460 + }, + { + "grad_norm": 0.07595035433769226, + "learning_rate": 8.188462545897512e-06, + "loss": 0.0017, + "step": 49470 + }, + { + "grad_norm": 0.07802329212427139, + "learning_rate": 8.173356796874304e-06, + "loss": 0.0024, + "step": 49480 + }, + { + "grad_norm": 0.09403416514396667, + "learning_rate": 8.158263753680906e-06, + "loss": 0.003, + "step": 49490 + }, + { + "grad_norm": 0.07932818681001663, + "learning_rate": 8.143183420902239e-06, + "loss": 0.0026, + "step": 49500 + }, + { + "grad_norm": 0.12975916266441345, + "learning_rate": 8.128115803119258e-06, + "loss": 0.0025, + "step": 49510 + }, + { + "grad_norm": 0.09234490245580673, + "learning_rate": 8.11306090490916e-06, + "loss": 0.0018, + "step": 49520 + }, + { + "grad_norm": 0.09094514697790146, + "learning_rate": 8.098018730845169e-06, + "loss": 0.0017, + "step": 49530 + }, + { + "grad_norm": 0.12187396734952927, + "learning_rate": 8.082989285496745e-06, + "loss": 0.0022, + "step": 49540 + }, + { + "grad_norm": 0.1424001008272171, + "learning_rate": 8.067972573429416e-06, + "loss": 0.002, + "step": 49550 + }, + { + "grad_norm": 0.09127826243638992, + "learning_rate": 8.052968599204874e-06, + "loss": 0.0023, + "step": 49560 + }, + { + "grad_norm": 0.1185259222984314, + "learning_rate": 8.037977367380922e-06, + "loss": 0.0027, + "step": 49570 + }, + { + "grad_norm": 0.11775805801153183, + "learning_rate": 8.022998882511495e-06, + "loss": 0.0024, + "step": 49580 + }, + { + "grad_norm": 0.09068689495325089, + "learning_rate": 8.008033149146677e-06, + "loss": 0.0053, + "step": 49590 + }, + { + "grad_norm": 0.08658009022474289, + "learning_rate": 7.993080171832656e-06, + "loss": 0.0027, + "step": 49600 + }, + { + "grad_norm": 0.09561969339847565, + "learning_rate": 7.978139955111752e-06, + "loss": 0.0023, + "step": 49610 + }, + { + "grad_norm": 0.0997595265507698, + "learning_rate": 7.9632125035224e-06, + "loss": 0.0017, + "step": 49620 + }, + { + "grad_norm": 0.11469525843858719, + "learning_rate": 7.948297821599177e-06, + "loss": 0.0022, + "step": 49630 + }, + { + "grad_norm": 0.1325279176235199, + "learning_rate": 7.933395913872755e-06, + "loss": 0.0022, + "step": 49640 + }, + { + "grad_norm": 0.09660835564136505, + "learning_rate": 7.918506784869972e-06, + "loss": 0.0021, + "step": 49650 + }, + { + "grad_norm": 0.10351784527301788, + "learning_rate": 7.903630439113707e-06, + "loss": 0.0017, + "step": 49660 + }, + { + "grad_norm": 0.0911022424697876, + "learning_rate": 7.888766881123044e-06, + "loss": 0.0039, + "step": 49670 + }, + { + "grad_norm": 0.10272479802370071, + "learning_rate": 7.873916115413099e-06, + "loss": 0.0024, + "step": 49680 + }, + { + "grad_norm": 0.09037468582391739, + "learning_rate": 7.85907814649518e-06, + "loss": 0.0041, + "step": 49690 + }, + { + "grad_norm": 0.11527078598737717, + "learning_rate": 7.844252978876649e-06, + "loss": 0.002, + "step": 49700 + }, + { + "grad_norm": 0.15726034343242645, + "learning_rate": 7.829440617061001e-06, + "loss": 0.0025, + "step": 49710 + }, + { + "grad_norm": 0.09577140212059021, + "learning_rate": 7.814641065547851e-06, + "loss": 0.0036, + "step": 49720 + }, + { + "grad_norm": 0.12195814400911331, + "learning_rate": 7.79985432883289e-06, + "loss": 0.0027, + "step": 49730 + }, + { + "grad_norm": 0.11293379217386246, + "learning_rate": 7.78508041140797e-06, + "loss": 0.0036, + "step": 49740 + }, + { + "grad_norm": 0.07693137228488922, + "learning_rate": 7.770319317760993e-06, + "loss": 0.0019, + "step": 49750 + }, + { + "grad_norm": 0.08573081344366074, + "learning_rate": 7.755571052376004e-06, + "loss": 0.0024, + "step": 49760 + }, + { + "grad_norm": 0.11723636835813522, + "learning_rate": 7.740835619733128e-06, + "loss": 0.0028, + "step": 49770 + }, + { + "grad_norm": 0.09361232817173004, + "learning_rate": 7.726113024308601e-06, + "loss": 0.0017, + "step": 49780 + }, + { + "grad_norm": 0.08562222123146057, + "learning_rate": 7.711403270574746e-06, + "loss": 0.0024, + "step": 49790 + }, + { + "grad_norm": 0.08299718052148819, + "learning_rate": 7.696706363000039e-06, + "loss": 0.0024, + "step": 49800 + }, + { + "grad_norm": 0.08675315976142883, + "learning_rate": 7.682022306048959e-06, + "loss": 0.002, + "step": 49810 + }, + { + "grad_norm": 0.12459217756986618, + "learning_rate": 7.667351104182186e-06, + "loss": 0.0022, + "step": 49820 + }, + { + "grad_norm": 0.09263528138399124, + "learning_rate": 7.652692761856395e-06, + "loss": 0.002, + "step": 49830 + }, + { + "grad_norm": 0.08735678344964981, + "learning_rate": 7.63804728352444e-06, + "loss": 0.0024, + "step": 49840 + }, + { + "grad_norm": 0.13695622980594635, + "learning_rate": 7.623414673635215e-06, + "loss": 0.0029, + "step": 49850 + }, + { + "grad_norm": 0.08536655455827713, + "learning_rate": 7.608794936633723e-06, + "loss": 0.0018, + "step": 49860 + }, + { + "grad_norm": 0.13196313381195068, + "learning_rate": 7.594188076961056e-06, + "loss": 0.0024, + "step": 49870 + }, + { + "grad_norm": 0.09964917600154877, + "learning_rate": 7.579594099054382e-06, + "loss": 0.0021, + "step": 49880 + }, + { + "grad_norm": 0.10163655132055283, + "learning_rate": 7.565013007346983e-06, + "loss": 0.0026, + "step": 49890 + }, + { + "grad_norm": 0.07348078489303589, + "learning_rate": 7.5504448062682035e-06, + "loss": 0.0022, + "step": 49900 + }, + { + "grad_norm": 0.0986529067158699, + "learning_rate": 7.53588950024347e-06, + "loss": 0.0017, + "step": 49910 + }, + { + "grad_norm": 0.09526107460260391, + "learning_rate": 7.5213470936943145e-06, + "loss": 0.0025, + "step": 49920 + }, + { + "grad_norm": 0.13952034711837769, + "learning_rate": 7.506817591038323e-06, + "loss": 0.0038, + "step": 49930 + }, + { + "grad_norm": 0.09679154306650162, + "learning_rate": 7.492300996689183e-06, + "loss": 0.0025, + "step": 49940 + }, + { + "grad_norm": 0.10739537328481674, + "learning_rate": 7.477797315056645e-06, + "loss": 0.0018, + "step": 49950 + }, + { + "grad_norm": 0.11501108109951019, + "learning_rate": 7.463306550546539e-06, + "loss": 0.002, + "step": 49960 + }, + { + "grad_norm": 0.09192699939012527, + "learning_rate": 7.448828707560812e-06, + "loss": 0.0018, + "step": 49970 + }, + { + "grad_norm": 0.1040724515914917, + "learning_rate": 7.4343637904974e-06, + "loss": 0.0019, + "step": 49980 + }, + { + "grad_norm": 0.10239234566688538, + "learning_rate": 7.419911803750401e-06, + "loss": 0.0025, + "step": 49990 + }, + { + "grad_norm": 0.10671626031398773, + "learning_rate": 7.405472751709935e-06, + "loss": 0.0025, + "step": 50000 + }, + { + "grad_norm": 0.10485080629587173, + "learning_rate": 7.3910466387622e-06, + "loss": 0.0052, + "step": 50010 + }, + { + "grad_norm": 0.1321965903043747, + "learning_rate": 7.3766334692894735e-06, + "loss": 0.0022, + "step": 50020 + }, + { + "grad_norm": 0.12933506071567535, + "learning_rate": 7.3622332476700865e-06, + "loss": 0.0033, + "step": 50030 + }, + { + "grad_norm": 0.10197607427835464, + "learning_rate": 7.347845978278472e-06, + "loss": 0.0022, + "step": 50040 + }, + { + "grad_norm": 0.16024062037467957, + "learning_rate": 7.333471665485065e-06, + "loss": 0.003, + "step": 50050 + }, + { + "grad_norm": 0.09591895341873169, + "learning_rate": 7.31911031365643e-06, + "loss": 0.0022, + "step": 50060 + }, + { + "grad_norm": 0.09845178574323654, + "learning_rate": 7.304761927155157e-06, + "loss": 0.0015, + "step": 50070 + }, + { + "grad_norm": 0.11886335909366608, + "learning_rate": 7.29042651033991e-06, + "loss": 0.0022, + "step": 50080 + }, + { + "grad_norm": 0.17459557950496674, + "learning_rate": 7.276104067565409e-06, + "loss": 0.0032, + "step": 50090 + }, + { + "grad_norm": 0.10992390662431717, + "learning_rate": 7.261794603182431e-06, + "loss": 0.0023, + "step": 50100 + }, + { + "grad_norm": 0.10051076114177704, + "learning_rate": 7.24749812153781e-06, + "loss": 0.0023, + "step": 50110 + }, + { + "grad_norm": 0.09980118274688721, + "learning_rate": 7.2332146269744605e-06, + "loss": 0.0041, + "step": 50120 + }, + { + "grad_norm": 0.1500774770975113, + "learning_rate": 7.218944123831295e-06, + "loss": 0.0041, + "step": 50130 + }, + { + "grad_norm": 0.11630157381296158, + "learning_rate": 7.204686616443351e-06, + "loss": 0.0019, + "step": 50140 + }, + { + "grad_norm": 0.10871115326881409, + "learning_rate": 7.190442109141665e-06, + "loss": 0.0025, + "step": 50150 + }, + { + "grad_norm": 0.1109437569975853, + "learning_rate": 7.176210606253347e-06, + "loss": 0.002, + "step": 50160 + }, + { + "grad_norm": 0.08186427503824234, + "learning_rate": 7.161992112101551e-06, + "loss": 0.0023, + "step": 50170 + }, + { + "grad_norm": 0.0854170173406601, + "learning_rate": 7.147786631005465e-06, + "loss": 0.002, + "step": 50180 + }, + { + "grad_norm": 0.09184639900922775, + "learning_rate": 7.1335941672803775e-06, + "loss": 0.0022, + "step": 50190 + }, + { + "grad_norm": 0.10300204902887344, + "learning_rate": 7.1194147252375384e-06, + "loss": 0.002, + "step": 50200 + }, + { + "grad_norm": 0.08573468029499054, + "learning_rate": 7.10524830918432e-06, + "loss": 0.0023, + "step": 50210 + }, + { + "grad_norm": 0.0983940064907074, + "learning_rate": 7.091094923424097e-06, + "loss": 0.0018, + "step": 50220 + }, + { + "grad_norm": 0.14008750021457672, + "learning_rate": 7.0769545722562894e-06, + "loss": 0.0017, + "step": 50230 + }, + { + "grad_norm": 0.07016535848379135, + "learning_rate": 7.0628272599763675e-06, + "loss": 0.0019, + "step": 50240 + }, + { + "grad_norm": 0.10257017612457275, + "learning_rate": 7.048712990875828e-06, + "loss": 0.0017, + "step": 50250 + }, + { + "grad_norm": 0.09225355088710785, + "learning_rate": 7.034611769242216e-06, + "loss": 0.0026, + "step": 50260 + }, + { + "grad_norm": 0.1315232515335083, + "learning_rate": 7.02052359935913e-06, + "loss": 0.0031, + "step": 50270 + }, + { + "grad_norm": 0.10144662111997604, + "learning_rate": 7.006448485506145e-06, + "loss": 0.003, + "step": 50280 + }, + { + "grad_norm": 0.14347226917743683, + "learning_rate": 6.992386431958942e-06, + "loss": 0.0018, + "step": 50290 + }, + { + "grad_norm": 0.10548096150159836, + "learning_rate": 6.978337442989197e-06, + "loss": 0.0018, + "step": 50300 + }, + { + "grad_norm": 0.13043780624866486, + "learning_rate": 6.964301522864608e-06, + "loss": 0.0026, + "step": 50310 + }, + { + "grad_norm": 0.108993299305439, + "learning_rate": 6.950278675848926e-06, + "loss": 0.0033, + "step": 50320 + }, + { + "grad_norm": 0.10760018974542618, + "learning_rate": 6.9362689062019145e-06, + "loss": 0.0022, + "step": 50330 + }, + { + "grad_norm": 0.07013796269893646, + "learning_rate": 6.922272218179393e-06, + "loss": 0.0019, + "step": 50340 + }, + { + "grad_norm": 0.0823139101266861, + "learning_rate": 6.908288616033148e-06, + "loss": 0.0019, + "step": 50350 + }, + { + "grad_norm": 0.11102023720741272, + "learning_rate": 6.894318104011077e-06, + "loss": 0.0025, + "step": 50360 + }, + { + "grad_norm": 0.10970284044742584, + "learning_rate": 6.880360686357007e-06, + "loss": 0.0022, + "step": 50370 + }, + { + "grad_norm": 0.0939311683177948, + "learning_rate": 6.8664163673108575e-06, + "loss": 0.0019, + "step": 50380 + }, + { + "grad_norm": 0.09024213254451752, + "learning_rate": 6.85248515110854e-06, + "loss": 0.0022, + "step": 50390 + }, + { + "grad_norm": 0.08812045305967331, + "learning_rate": 6.838567041981992e-06, + "loss": 0.0021, + "step": 50400 + }, + { + "grad_norm": 0.14436450600624084, + "learning_rate": 6.8246620441591634e-06, + "loss": 0.0025, + "step": 50410 + }, + { + "grad_norm": 0.1031867042183876, + "learning_rate": 6.8107701618640275e-06, + "loss": 0.002, + "step": 50420 + }, + { + "grad_norm": 0.06161723658442497, + "learning_rate": 6.796891399316557e-06, + "loss": 0.0017, + "step": 50430 + }, + { + "grad_norm": 0.16692671179771423, + "learning_rate": 6.7830257607327804e-06, + "loss": 0.0043, + "step": 50440 + }, + { + "grad_norm": 0.11882632225751877, + "learning_rate": 6.7691732503247e-06, + "loss": 0.0021, + "step": 50450 + }, + { + "grad_norm": 0.10401830077171326, + "learning_rate": 6.755333872300346e-06, + "loss": 0.0022, + "step": 50460 + }, + { + "grad_norm": 0.0780961662530899, + "learning_rate": 6.741507630863747e-06, + "loss": 0.0037, + "step": 50470 + }, + { + "grad_norm": 0.09252814948558807, + "learning_rate": 6.727694530214945e-06, + "loss": 0.002, + "step": 50480 + }, + { + "grad_norm": 0.09737398475408554, + "learning_rate": 6.713894574550028e-06, + "loss": 0.0023, + "step": 50490 + }, + { + "grad_norm": 0.09577210992574692, + "learning_rate": 6.700107768061015e-06, + "loss": 0.0032, + "step": 50500 + }, + { + "grad_norm": 0.0999421775341034, + "learning_rate": 6.686334114936016e-06, + "loss": 0.0029, + "step": 50510 + }, + { + "grad_norm": 0.08325758576393127, + "learning_rate": 6.672573619359063e-06, + "loss": 0.0015, + "step": 50520 + }, + { + "grad_norm": 0.09199460595846176, + "learning_rate": 6.658826285510256e-06, + "loss": 0.0019, + "step": 50530 + }, + { + "grad_norm": 0.07033105939626694, + "learning_rate": 6.645092117565666e-06, + "loss": 0.0018, + "step": 50540 + }, + { + "grad_norm": 0.09250234812498093, + "learning_rate": 6.631371119697371e-06, + "loss": 0.0022, + "step": 50550 + }, + { + "grad_norm": 0.14956402778625488, + "learning_rate": 6.6176632960734505e-06, + "loss": 0.0028, + "step": 50560 + }, + { + "grad_norm": 0.11683869361877441, + "learning_rate": 6.603968650857978e-06, + "loss": 0.0026, + "step": 50570 + }, + { + "grad_norm": 0.12380315363407135, + "learning_rate": 6.5902871882110085e-06, + "loss": 0.0018, + "step": 50580 + }, + { + "grad_norm": 0.0813598781824112, + "learning_rate": 6.576618912288635e-06, + "loss": 0.0018, + "step": 50590 + }, + { + "grad_norm": 0.09748382866382599, + "learning_rate": 6.562963827242913e-06, + "loss": 0.0022, + "step": 50600 + }, + { + "grad_norm": 0.10563912242650986, + "learning_rate": 6.549321937221886e-06, + "loss": 0.0026, + "step": 50610 + }, + { + "grad_norm": 0.10386590659618378, + "learning_rate": 6.5356932463696064e-06, + "loss": 0.0026, + "step": 50620 + }, + { + "grad_norm": 0.10276325792074203, + "learning_rate": 6.522077758826101e-06, + "loss": 0.0027, + "step": 50630 + }, + { + "grad_norm": 0.07653278112411499, + "learning_rate": 6.5084754787274275e-06, + "loss": 0.0026, + "step": 50640 + }, + { + "grad_norm": 0.07074282318353653, + "learning_rate": 6.494886410205553e-06, + "loss": 0.0015, + "step": 50650 + }, + { + "grad_norm": 0.08966255933046341, + "learning_rate": 6.481310557388521e-06, + "loss": 0.0021, + "step": 50660 + }, + { + "grad_norm": 0.07163316756486893, + "learning_rate": 6.46774792440028e-06, + "loss": 0.0015, + "step": 50670 + }, + { + "grad_norm": 0.09443012624979019, + "learning_rate": 6.4541985153608206e-06, + "loss": 0.0018, + "step": 50680 + }, + { + "grad_norm": 0.10940142720937729, + "learning_rate": 6.4406623343861e-06, + "loss": 0.0016, + "step": 50690 + }, + { + "grad_norm": 0.07139239460229874, + "learning_rate": 6.427139385588038e-06, + "loss": 0.0024, + "step": 50700 + }, + { + "grad_norm": 0.09700915217399597, + "learning_rate": 6.413629673074561e-06, + "loss": 0.002, + "step": 50710 + }, + { + "grad_norm": 0.0829000398516655, + "learning_rate": 6.400133200949554e-06, + "loss": 0.0017, + "step": 50720 + }, + { + "grad_norm": 0.09213865548372269, + "learning_rate": 6.386649973312897e-06, + "loss": 0.0023, + "step": 50730 + }, + { + "grad_norm": 0.06495214253664017, + "learning_rate": 6.37317999426042e-06, + "loss": 0.0019, + "step": 50740 + }, + { + "grad_norm": 0.1298716962337494, + "learning_rate": 6.359723267883977e-06, + "loss": 0.0024, + "step": 50750 + }, + { + "grad_norm": 0.11895918101072311, + "learning_rate": 6.346279798271343e-06, + "loss": 0.0025, + "step": 50760 + }, + { + "grad_norm": 0.10852912068367004, + "learning_rate": 6.332849589506301e-06, + "loss": 0.003, + "step": 50770 + }, + { + "grad_norm": 0.08950640261173248, + "learning_rate": 6.319432645668588e-06, + "loss": 0.002, + "step": 50780 + }, + { + "grad_norm": 0.08060760796070099, + "learning_rate": 6.306028970833922e-06, + "loss": 0.0019, + "step": 50790 + }, + { + "grad_norm": 0.0718727558851242, + "learning_rate": 6.2926385690739665e-06, + "loss": 0.0018, + "step": 50800 + }, + { + "grad_norm": 0.10022716969251633, + "learning_rate": 6.279261444456413e-06, + "loss": 0.0019, + "step": 50810 + }, + { + "grad_norm": 0.06802347302436829, + "learning_rate": 6.265897601044829e-06, + "loss": 0.0022, + "step": 50820 + }, + { + "grad_norm": 0.10718294978141785, + "learning_rate": 6.2525470428988434e-06, + "loss": 0.0032, + "step": 50830 + }, + { + "grad_norm": 0.09632886946201324, + "learning_rate": 6.239209774073962e-06, + "loss": 0.0022, + "step": 50840 + }, + { + "grad_norm": 0.11730077862739563, + "learning_rate": 6.225885798621728e-06, + "loss": 0.0017, + "step": 50850 + }, + { + "grad_norm": 0.1273166835308075, + "learning_rate": 6.2125751205895925e-06, + "loss": 0.0031, + "step": 50860 + }, + { + "grad_norm": 0.1078796461224556, + "learning_rate": 6.199277744020998e-06, + "loss": 0.0022, + "step": 50870 + }, + { + "grad_norm": 0.1520080268383026, + "learning_rate": 6.185993672955337e-06, + "loss": 0.0033, + "step": 50880 + }, + { + "grad_norm": 0.06875384598970413, + "learning_rate": 6.172722911427947e-06, + "loss": 0.002, + "step": 50890 + }, + { + "grad_norm": 0.1419181376695633, + "learning_rate": 6.159465463470149e-06, + "loss": 0.0028, + "step": 50900 + }, + { + "grad_norm": 0.1305476278066635, + "learning_rate": 6.146221333109204e-06, + "loss": 0.0031, + "step": 50910 + }, + { + "grad_norm": 0.10579439252614975, + "learning_rate": 6.132990524368326e-06, + "loss": 0.0021, + "step": 50920 + }, + { + "grad_norm": 0.06911089271306992, + "learning_rate": 6.119773041266685e-06, + "loss": 0.0021, + "step": 50930 + }, + { + "grad_norm": 0.08394245058298111, + "learning_rate": 6.106568887819402e-06, + "loss": 0.003, + "step": 50940 + }, + { + "grad_norm": 0.08618783205747604, + "learning_rate": 6.093378068037548e-06, + "loss": 0.0023, + "step": 50950 + }, + { + "grad_norm": 0.1016187071800232, + "learning_rate": 6.080200585928164e-06, + "loss": 0.0019, + "step": 50960 + }, + { + "grad_norm": 0.0799061506986618, + "learning_rate": 6.06703644549419e-06, + "loss": 0.0019, + "step": 50970 + }, + { + "grad_norm": 0.05634094774723053, + "learning_rate": 6.053885650734576e-06, + "loss": 0.0015, + "step": 50980 + }, + { + "grad_norm": 0.06543285399675369, + "learning_rate": 6.040748205644153e-06, + "loss": 0.0017, + "step": 50990 + }, + { + "grad_norm": 0.11701717972755432, + "learning_rate": 6.0276241142137646e-06, + "loss": 0.0018, + "step": 51000 + }, + { + "grad_norm": 0.07951046526432037, + "learning_rate": 6.014513380430142e-06, + "loss": 0.0017, + "step": 51010 + }, + { + "grad_norm": 0.07820898294448853, + "learning_rate": 6.001416008275984e-06, + "loss": 0.0026, + "step": 51020 + }, + { + "grad_norm": 0.09279879927635193, + "learning_rate": 5.988332001729929e-06, + "loss": 0.0021, + "step": 51030 + }, + { + "grad_norm": 0.08558051288127899, + "learning_rate": 5.975261364766543e-06, + "loss": 0.0018, + "step": 51040 + }, + { + "grad_norm": 0.09042573720216751, + "learning_rate": 5.962204101356356e-06, + "loss": 0.0021, + "step": 51050 + }, + { + "grad_norm": 0.07813333719968796, + "learning_rate": 5.94916021546581e-06, + "loss": 0.0029, + "step": 51060 + }, + { + "grad_norm": 0.10995281487703323, + "learning_rate": 5.936129711057298e-06, + "loss": 0.0027, + "step": 51070 + }, + { + "grad_norm": 0.16042004525661469, + "learning_rate": 5.923112592089142e-06, + "loss": 0.0026, + "step": 51080 + }, + { + "grad_norm": 0.062433984130620956, + "learning_rate": 5.9101088625155954e-06, + "loss": 0.0013, + "step": 51090 + }, + { + "grad_norm": 0.11844075471162796, + "learning_rate": 5.897118526286843e-06, + "loss": 0.0035, + "step": 51100 + }, + { + "grad_norm": 0.10972762107849121, + "learning_rate": 5.884141587349035e-06, + "loss": 0.0026, + "step": 51110 + }, + { + "grad_norm": 0.12986643612384796, + "learning_rate": 5.871178049644177e-06, + "loss": 0.0025, + "step": 51120 + }, + { + "grad_norm": 0.12324147671461105, + "learning_rate": 5.858227917110293e-06, + "loss": 0.0023, + "step": 51130 + }, + { + "grad_norm": 0.056577168405056, + "learning_rate": 5.845291193681252e-06, + "loss": 0.0019, + "step": 51140 + }, + { + "grad_norm": 0.08215837925672531, + "learning_rate": 5.832367883286921e-06, + "loss": 0.002, + "step": 51150 + }, + { + "grad_norm": 0.08361116051673889, + "learning_rate": 5.81945798985305e-06, + "loss": 0.0022, + "step": 51160 + }, + { + "grad_norm": 0.05113065242767334, + "learning_rate": 5.806561517301306e-06, + "loss": 0.0014, + "step": 51170 + }, + { + "grad_norm": 0.14439591765403748, + "learning_rate": 5.793678469549335e-06, + "loss": 0.0022, + "step": 51180 + }, + { + "grad_norm": 0.08118671178817749, + "learning_rate": 5.780808850510627e-06, + "loss": 0.0026, + "step": 51190 + }, + { + "grad_norm": 0.1326126754283905, + "learning_rate": 5.767952664094673e-06, + "loss": 0.002, + "step": 51200 + }, + { + "grad_norm": 0.10221609473228455, + "learning_rate": 5.755109914206791e-06, + "loss": 0.0017, + "step": 51210 + }, + { + "grad_norm": 0.07862391322851181, + "learning_rate": 5.7422806047483125e-06, + "loss": 0.0019, + "step": 51220 + }, + { + "grad_norm": 0.10415737330913544, + "learning_rate": 5.72946473961643e-06, + "loss": 0.0022, + "step": 51230 + }, + { + "grad_norm": 0.09854258596897125, + "learning_rate": 5.716662322704264e-06, + "loss": 0.003, + "step": 51240 + }, + { + "grad_norm": 0.12208380550146103, + "learning_rate": 5.703873357900852e-06, + "loss": 0.0022, + "step": 51250 + }, + { + "grad_norm": 0.12907817959785461, + "learning_rate": 5.691097849091143e-06, + "loss": 0.0024, + "step": 51260 + }, + { + "grad_norm": 0.1858043521642685, + "learning_rate": 5.678335800155982e-06, + "loss": 0.0033, + "step": 51270 + }, + { + "grad_norm": 0.13309243321418762, + "learning_rate": 5.665587214972174e-06, + "loss": 0.0027, + "step": 51280 + }, + { + "grad_norm": 0.08783125132322311, + "learning_rate": 5.652852097412386e-06, + "loss": 0.002, + "step": 51290 + }, + { + "grad_norm": 0.11275880038738251, + "learning_rate": 5.640130451345216e-06, + "loss": 0.0023, + "step": 51300 + }, + { + "grad_norm": 0.062131527811288834, + "learning_rate": 5.627422280635159e-06, + "loss": 0.0025, + "step": 51310 + }, + { + "grad_norm": 0.07860226929187775, + "learning_rate": 5.61472758914261e-06, + "loss": 0.0022, + "step": 51320 + }, + { + "grad_norm": 0.05721055343747139, + "learning_rate": 5.602046380723918e-06, + "loss": 0.0019, + "step": 51330 + }, + { + "grad_norm": 0.05749586969614029, + "learning_rate": 5.5893786592312535e-06, + "loss": 0.0025, + "step": 51340 + }, + { + "grad_norm": 0.09556373953819275, + "learning_rate": 5.576724428512775e-06, + "loss": 0.0023, + "step": 51350 + }, + { + "grad_norm": 0.08765895664691925, + "learning_rate": 5.56408369241247e-06, + "loss": 0.002, + "step": 51360 + }, + { + "grad_norm": 0.07582288235425949, + "learning_rate": 5.5514564547702875e-06, + "loss": 0.0027, + "step": 51370 + }, + { + "grad_norm": 0.08892741799354553, + "learning_rate": 5.538842719422038e-06, + "loss": 0.0018, + "step": 51380 + }, + { + "grad_norm": 0.080121710896492, + "learning_rate": 5.52624249019944e-06, + "loss": 0.0018, + "step": 51390 + }, + { + "grad_norm": 0.09261548519134521, + "learning_rate": 5.51365577093011e-06, + "loss": 0.0018, + "step": 51400 + }, + { + "grad_norm": 0.10181830078363419, + "learning_rate": 5.501082565437565e-06, + "loss": 0.0028, + "step": 51410 + }, + { + "grad_norm": 0.08727855235338211, + "learning_rate": 5.488522877541202e-06, + "loss": 0.0022, + "step": 51420 + }, + { + "grad_norm": 0.11304180324077606, + "learning_rate": 5.475976711056341e-06, + "loss": 0.0025, + "step": 51430 + }, + { + "grad_norm": 0.14425863325595856, + "learning_rate": 5.463444069794166e-06, + "loss": 0.0024, + "step": 51440 + }, + { + "grad_norm": 0.09790679067373276, + "learning_rate": 5.4509249575617594e-06, + "loss": 0.0029, + "step": 51450 + }, + { + "grad_norm": 0.10836488008499146, + "learning_rate": 5.438419378162107e-06, + "loss": 0.0021, + "step": 51460 + }, + { + "grad_norm": 0.1139838844537735, + "learning_rate": 5.425927335394054e-06, + "loss": 0.0026, + "step": 51470 + }, + { + "grad_norm": 0.14746569097042084, + "learning_rate": 5.413448833052387e-06, + "loss": 0.0031, + "step": 51480 + }, + { + "grad_norm": 0.07528004050254822, + "learning_rate": 5.400983874927701e-06, + "loss": 0.0014, + "step": 51490 + }, + { + "grad_norm": 0.07845738530158997, + "learning_rate": 5.388532464806567e-06, + "loss": 0.0023, + "step": 51500 + }, + { + "grad_norm": 0.04815568029880524, + "learning_rate": 5.3760946064713546e-06, + "loss": 0.0014, + "step": 51510 + }, + { + "grad_norm": 0.12126094847917557, + "learning_rate": 5.363670303700386e-06, + "loss": 0.0016, + "step": 51520 + }, + { + "grad_norm": 0.13244712352752686, + "learning_rate": 5.351259560267824e-06, + "loss": 0.0027, + "step": 51530 + }, + { + "grad_norm": 0.0808882862329483, + "learning_rate": 5.338862379943721e-06, + "loss": 0.0021, + "step": 51540 + }, + { + "grad_norm": 0.06990167498588562, + "learning_rate": 5.326478766494025e-06, + "loss": 0.0016, + "step": 51550 + }, + { + "grad_norm": 0.08757064491510391, + "learning_rate": 5.3141087236805385e-06, + "loss": 0.0024, + "step": 51560 + }, + { + "grad_norm": 0.179501473903656, + "learning_rate": 5.3017522552609615e-06, + "loss": 0.0028, + "step": 51570 + }, + { + "grad_norm": 0.09101392328739166, + "learning_rate": 5.289409364988851e-06, + "loss": 0.0017, + "step": 51580 + }, + { + "grad_norm": 0.12195906788110733, + "learning_rate": 5.277080056613671e-06, + "loss": 0.0019, + "step": 51590 + }, + { + "grad_norm": 0.18641230463981628, + "learning_rate": 5.264764333880729e-06, + "loss": 0.0033, + "step": 51600 + }, + { + "grad_norm": 0.07430040091276169, + "learning_rate": 5.252462200531216e-06, + "loss": 0.0013, + "step": 51610 + }, + { + "grad_norm": 0.08553164452314377, + "learning_rate": 5.240173660302194e-06, + "loss": 0.0016, + "step": 51620 + }, + { + "grad_norm": 0.15592463314533234, + "learning_rate": 5.2278987169266044e-06, + "loss": 0.0021, + "step": 51630 + }, + { + "grad_norm": 0.12957355380058289, + "learning_rate": 5.215637374133231e-06, + "loss": 0.0031, + "step": 51640 + }, + { + "grad_norm": 0.09143024682998657, + "learning_rate": 5.203389635646782e-06, + "loss": 0.0021, + "step": 51650 + }, + { + "grad_norm": 0.19930338859558105, + "learning_rate": 5.191155505187756e-06, + "loss": 0.0024, + "step": 51660 + }, + { + "grad_norm": 0.11485683917999268, + "learning_rate": 5.1789349864726e-06, + "loss": 0.0019, + "step": 51670 + }, + { + "grad_norm": 0.08265043050050735, + "learning_rate": 5.16672808321354e-06, + "loss": 0.0041, + "step": 51680 + }, + { + "grad_norm": 0.08319453150033951, + "learning_rate": 5.154534799118749e-06, + "loss": 0.0023, + "step": 51690 + }, + { + "grad_norm": 0.06945076584815979, + "learning_rate": 5.142355137892207e-06, + "loss": 0.0014, + "step": 51700 + }, + { + "grad_norm": 0.05214717611670494, + "learning_rate": 5.130189103233779e-06, + "loss": 0.0016, + "step": 51710 + }, + { + "grad_norm": 0.11276382952928543, + "learning_rate": 5.118036698839179e-06, + "loss": 0.0024, + "step": 51720 + }, + { + "grad_norm": 0.08282233029603958, + "learning_rate": 5.105897928399983e-06, + "loss": 0.0013, + "step": 51730 + }, + { + "grad_norm": 0.0844469889998436, + "learning_rate": 5.0937727956036464e-06, + "loss": 0.0023, + "step": 51740 + }, + { + "grad_norm": 0.07699114084243774, + "learning_rate": 5.081661304133456e-06, + "loss": 0.0025, + "step": 51750 + }, + { + "grad_norm": 0.07751737534999847, + "learning_rate": 5.069563457668558e-06, + "loss": 0.0027, + "step": 51760 + }, + { + "grad_norm": 0.08662010729312897, + "learning_rate": 5.0574792598839624e-06, + "loss": 0.0026, + "step": 51770 + }, + { + "grad_norm": 0.0768296867609024, + "learning_rate": 5.0454087144505276e-06, + "loss": 0.0039, + "step": 51780 + }, + { + "grad_norm": 0.07720673084259033, + "learning_rate": 5.0333518250349655e-06, + "loss": 0.0018, + "step": 51790 + }, + { + "grad_norm": 0.08852304518222809, + "learning_rate": 5.021308595299856e-06, + "loss": 0.004, + "step": 51800 + }, + { + "grad_norm": 0.11592481285333633, + "learning_rate": 5.009279028903585e-06, + "loss": 0.0018, + "step": 51810 + }, + { + "grad_norm": 0.16419260203838348, + "learning_rate": 4.997263129500452e-06, + "loss": 0.0048, + "step": 51820 + }, + { + "grad_norm": 0.09170050919055939, + "learning_rate": 4.985260900740535e-06, + "loss": 0.0017, + "step": 51830 + }, + { + "grad_norm": 0.12243098765611649, + "learning_rate": 4.973272346269814e-06, + "loss": 0.0031, + "step": 51840 + }, + { + "grad_norm": 0.15324784815311432, + "learning_rate": 4.961297469730097e-06, + "loss": 0.0025, + "step": 51850 + }, + { + "grad_norm": 0.1402556151151657, + "learning_rate": 4.949336274759031e-06, + "loss": 0.0027, + "step": 51860 + }, + { + "grad_norm": 0.08906861394643784, + "learning_rate": 4.9373887649901144e-06, + "loss": 0.0018, + "step": 51870 + }, + { + "grad_norm": 0.07055895030498505, + "learning_rate": 4.925454944052666e-06, + "loss": 0.0021, + "step": 51880 + }, + { + "grad_norm": 0.1100592315196991, + "learning_rate": 4.913534815571891e-06, + "loss": 0.0019, + "step": 51890 + }, + { + "grad_norm": 0.13338735699653625, + "learning_rate": 4.901628383168805e-06, + "loss": 0.0022, + "step": 51900 + }, + { + "grad_norm": 0.0836353525519371, + "learning_rate": 4.8897356504602585e-06, + "loss": 0.0025, + "step": 51910 + }, + { + "grad_norm": 0.08108049631118774, + "learning_rate": 4.877856621058957e-06, + "loss": 0.0026, + "step": 51920 + }, + { + "grad_norm": 0.08092668652534485, + "learning_rate": 4.86599129857343e-06, + "loss": 0.0016, + "step": 51930 + }, + { + "grad_norm": 0.07439397275447845, + "learning_rate": 4.85413968660805e-06, + "loss": 0.0016, + "step": 51940 + }, + { + "grad_norm": 0.06807383894920349, + "learning_rate": 4.842301788763031e-06, + "loss": 0.0019, + "step": 51950 + }, + { + "grad_norm": 0.09692571312189102, + "learning_rate": 4.830477608634393e-06, + "loss": 0.0014, + "step": 51960 + }, + { + "grad_norm": 0.07418592274188995, + "learning_rate": 4.818667149814049e-06, + "loss": 0.0016, + "step": 51970 + }, + { + "grad_norm": 0.06503136456012726, + "learning_rate": 4.806870415889664e-06, + "loss": 0.0017, + "step": 51980 + }, + { + "grad_norm": 0.132238507270813, + "learning_rate": 4.795087410444798e-06, + "loss": 0.0036, + "step": 51990 + }, + { + "grad_norm": 0.11996836215257645, + "learning_rate": 4.783318137058807e-06, + "loss": 0.0029, + "step": 52000 + }, + { + "grad_norm": 0.07210696488618851, + "learning_rate": 4.771562599306895e-06, + "loss": 0.0016, + "step": 52010 + }, + { + "grad_norm": 0.0911535993218422, + "learning_rate": 4.759820800760073e-06, + "loss": 0.0027, + "step": 52020 + }, + { + "grad_norm": 0.0797157809138298, + "learning_rate": 4.7480927449851834e-06, + "loss": 0.002, + "step": 52030 + }, + { + "grad_norm": 0.07628306746482849, + "learning_rate": 4.73637843554493e-06, + "loss": 0.0021, + "step": 52040 + }, + { + "grad_norm": 0.07218877226114273, + "learning_rate": 4.724677875997774e-06, + "loss": 0.0016, + "step": 52050 + }, + { + "grad_norm": 0.08089448511600494, + "learning_rate": 4.712991069898065e-06, + "loss": 0.002, + "step": 52060 + }, + { + "grad_norm": 0.09109161049127579, + "learning_rate": 4.7013180207959305e-06, + "loss": 0.0019, + "step": 52070 + }, + { + "grad_norm": 0.05619505047798157, + "learning_rate": 4.6896587322373395e-06, + "loss": 0.0016, + "step": 52080 + }, + { + "grad_norm": 0.0744321346282959, + "learning_rate": 4.678013207764081e-06, + "loss": 0.0019, + "step": 52090 + }, + { + "grad_norm": 0.11107102781534195, + "learning_rate": 4.666381450913748e-06, + "loss": 0.0021, + "step": 52100 + }, + { + "grad_norm": 0.10511229187250137, + "learning_rate": 4.654763465219752e-06, + "loss": 0.0019, + "step": 52110 + }, + { + "grad_norm": 0.04569906368851662, + "learning_rate": 4.643159254211371e-06, + "loss": 0.0015, + "step": 52120 + }, + { + "grad_norm": 0.09290259331464767, + "learning_rate": 4.631568821413606e-06, + "loss": 0.0018, + "step": 52130 + }, + { + "grad_norm": 0.08251085132360458, + "learning_rate": 4.619992170347359e-06, + "loss": 0.0017, + "step": 52140 + }, + { + "grad_norm": 0.09558875858783722, + "learning_rate": 4.608429304529305e-06, + "loss": 0.002, + "step": 52150 + }, + { + "grad_norm": 0.08549585938453674, + "learning_rate": 4.596880227471928e-06, + "loss": 0.0019, + "step": 52160 + }, + { + "grad_norm": 0.09379017353057861, + "learning_rate": 4.585344942683539e-06, + "loss": 0.0016, + "step": 52170 + }, + { + "grad_norm": 0.061152227222919464, + "learning_rate": 4.573823453668241e-06, + "loss": 0.0015, + "step": 52180 + }, + { + "grad_norm": 0.08461473137140274, + "learning_rate": 4.562315763925995e-06, + "loss": 0.0033, + "step": 52190 + }, + { + "grad_norm": 0.06666522473096848, + "learning_rate": 4.5508218769524825e-06, + "loss": 0.0027, + "step": 52200 + }, + { + "grad_norm": 0.0728725865483284, + "learning_rate": 4.539341796239277e-06, + "loss": 0.0016, + "step": 52210 + }, + { + "grad_norm": 0.08726110309362411, + "learning_rate": 4.527875525273717e-06, + "loss": 0.002, + "step": 52220 + }, + { + "grad_norm": 0.08086732774972916, + "learning_rate": 4.51642306753895e-06, + "loss": 0.0025, + "step": 52230 + }, + { + "grad_norm": 0.07932565361261368, + "learning_rate": 4.5049844265139306e-06, + "loss": 0.0025, + "step": 52240 + }, + { + "grad_norm": 0.07438625395298004, + "learning_rate": 4.4935596056734144e-06, + "loss": 0.0019, + "step": 52250 + }, + { + "grad_norm": 0.09787704050540924, + "learning_rate": 4.482148608487957e-06, + "loss": 0.0016, + "step": 52260 + }, + { + "grad_norm": 0.058351270854473114, + "learning_rate": 4.4707514384239365e-06, + "loss": 0.0026, + "step": 52270 + }, + { + "grad_norm": 0.07209163159132004, + "learning_rate": 4.459368098943484e-06, + "loss": 0.0029, + "step": 52280 + }, + { + "grad_norm": 0.05841663107275963, + "learning_rate": 4.447998593504582e-06, + "loss": 0.0017, + "step": 52290 + }, + { + "grad_norm": 0.09702879190444946, + "learning_rate": 4.4366429255609744e-06, + "loss": 0.0041, + "step": 52300 + }, + { + "grad_norm": 0.08758071064949036, + "learning_rate": 4.425301098562212e-06, + "loss": 0.002, + "step": 52310 + }, + { + "grad_norm": 0.1202865019440651, + "learning_rate": 4.413973115953651e-06, + "loss": 0.002, + "step": 52320 + }, + { + "grad_norm": 0.06758531928062439, + "learning_rate": 4.402658981176416e-06, + "loss": 0.0025, + "step": 52330 + }, + { + "grad_norm": 0.11205927282571793, + "learning_rate": 4.391358697667475e-06, + "loss": 0.0021, + "step": 52340 + }, + { + "grad_norm": 0.09588846564292908, + "learning_rate": 4.3800722688595195e-06, + "loss": 0.002, + "step": 52350 + }, + { + "grad_norm": 0.11360388249158859, + "learning_rate": 4.368799698181097e-06, + "loss": 0.003, + "step": 52360 + }, + { + "grad_norm": 0.07467757910490036, + "learning_rate": 4.357540989056486e-06, + "loss": 0.0019, + "step": 52370 + }, + { + "grad_norm": 0.06339841336011887, + "learning_rate": 4.346296144905815e-06, + "loss": 0.0021, + "step": 52380 + }, + { + "grad_norm": 0.12538567185401917, + "learning_rate": 4.335065169144958e-06, + "loss": 0.0016, + "step": 52390 + }, + { + "grad_norm": 0.06279294192790985, + "learning_rate": 4.323848065185593e-06, + "loss": 0.0016, + "step": 52400 + }, + { + "grad_norm": 0.0800292119383812, + "learning_rate": 4.31264483643517e-06, + "loss": 0.0051, + "step": 52410 + }, + { + "grad_norm": 0.10177330672740936, + "learning_rate": 4.301455486296946e-06, + "loss": 0.0032, + "step": 52420 + }, + { + "grad_norm": 0.1045570895075798, + "learning_rate": 4.290280018169935e-06, + "loss": 0.0028, + "step": 52430 + }, + { + "grad_norm": 0.06981983780860901, + "learning_rate": 4.27911843544897e-06, + "loss": 0.0025, + "step": 52440 + }, + { + "grad_norm": 0.07184161990880966, + "learning_rate": 4.2679707415246294e-06, + "loss": 0.0028, + "step": 52450 + }, + { + "grad_norm": 0.11678671091794968, + "learning_rate": 4.256836939783299e-06, + "loss": 0.003, + "step": 52460 + }, + { + "grad_norm": 0.12661907076835632, + "learning_rate": 4.245717033607127e-06, + "loss": 0.0034, + "step": 52470 + }, + { + "grad_norm": 0.10484634339809418, + "learning_rate": 4.234611026374035e-06, + "loss": 0.0026, + "step": 52480 + }, + { + "grad_norm": 0.09484592080116272, + "learning_rate": 4.2235189214577694e-06, + "loss": 0.0032, + "step": 52490 + }, + { + "grad_norm": 0.1650460809469223, + "learning_rate": 4.212440722227779e-06, + "loss": 0.0035, + "step": 52500 + }, + { + "grad_norm": 0.14628027379512787, + "learning_rate": 4.201376432049364e-06, + "loss": 0.0024, + "step": 52510 + }, + { + "grad_norm": 0.13583262264728546, + "learning_rate": 4.1903260542835275e-06, + "loss": 0.0019, + "step": 52520 + }, + { + "grad_norm": 0.13224215805530548, + "learning_rate": 4.1792895922871114e-06, + "loss": 0.0021, + "step": 52530 + }, + { + "grad_norm": 0.10863717645406723, + "learning_rate": 4.168267049412694e-06, + "loss": 0.0017, + "step": 52540 + }, + { + "grad_norm": 0.0944783166050911, + "learning_rate": 4.157258429008626e-06, + "loss": 0.0023, + "step": 52550 + }, + { + "grad_norm": 0.12254945933818817, + "learning_rate": 4.146263734419043e-06, + "loss": 0.0028, + "step": 52560 + }, + { + "grad_norm": 0.0702785775065422, + "learning_rate": 4.135282968983839e-06, + "loss": 0.0018, + "step": 52570 + }, + { + "grad_norm": 0.12036824226379395, + "learning_rate": 4.124316136038675e-06, + "loss": 0.002, + "step": 52580 + }, + { + "grad_norm": 0.06493907421827316, + "learning_rate": 4.1133632389149965e-06, + "loss": 0.0018, + "step": 52590 + }, + { + "grad_norm": 0.07807005941867828, + "learning_rate": 4.102424280939998e-06, + "loss": 0.0026, + "step": 52600 + }, + { + "grad_norm": 0.058739788830280304, + "learning_rate": 4.091499265436649e-06, + "loss": 0.0027, + "step": 52610 + }, + { + "grad_norm": 0.06150120496749878, + "learning_rate": 4.080588195723684e-06, + "loss": 0.0015, + "step": 52620 + }, + { + "grad_norm": 0.08427347987890244, + "learning_rate": 4.069691075115578e-06, + "loss": 0.0015, + "step": 52630 + }, + { + "grad_norm": 0.10936084389686584, + "learning_rate": 4.0588079069226235e-06, + "loss": 0.0014, + "step": 52640 + }, + { + "grad_norm": 0.03771369531750679, + "learning_rate": 4.0479386944508034e-06, + "loss": 0.0019, + "step": 52650 + }, + { + "grad_norm": 0.04970606416463852, + "learning_rate": 4.037083441001932e-06, + "loss": 0.0013, + "step": 52660 + }, + { + "grad_norm": 0.06829822808504105, + "learning_rate": 4.026242149873516e-06, + "loss": 0.0017, + "step": 52670 + }, + { + "grad_norm": 0.04925454035401344, + "learning_rate": 4.015414824358871e-06, + "loss": 0.0022, + "step": 52680 + }, + { + "grad_norm": 0.06825439631938934, + "learning_rate": 4.004601467747054e-06, + "loss": 0.0013, + "step": 52690 + }, + { + "grad_norm": 0.05642141029238701, + "learning_rate": 3.993802083322873e-06, + "loss": 0.0023, + "step": 52700 + }, + { + "grad_norm": 0.09069765359163284, + "learning_rate": 3.9830166743668906e-06, + "loss": 0.0017, + "step": 52710 + }, + { + "grad_norm": 0.08697748184204102, + "learning_rate": 3.9722452441554425e-06, + "loss": 0.0026, + "step": 52720 + }, + { + "grad_norm": 0.05683028697967529, + "learning_rate": 3.961487795960584e-06, + "loss": 0.0015, + "step": 52730 + }, + { + "grad_norm": 0.059424031525850296, + "learning_rate": 3.95074433305016e-06, + "loss": 0.0036, + "step": 52740 + }, + { + "grad_norm": 0.08129140734672546, + "learning_rate": 3.940014858687752e-06, + "loss": 0.0014, + "step": 52750 + }, + { + "grad_norm": 0.11415276676416397, + "learning_rate": 3.929299376132689e-06, + "loss": 0.0047, + "step": 52760 + }, + { + "grad_norm": 0.07907243072986603, + "learning_rate": 3.918597888640047e-06, + "loss": 0.0025, + "step": 52770 + }, + { + "grad_norm": 0.08863958716392517, + "learning_rate": 3.907910399460657e-06, + "loss": 0.0015, + "step": 52780 + }, + { + "grad_norm": 0.05264083668589592, + "learning_rate": 3.8972369118410956e-06, + "loss": 0.0013, + "step": 52790 + }, + { + "grad_norm": 0.1157156378030777, + "learning_rate": 3.88657742902368e-06, + "loss": 0.0027, + "step": 52800 + }, + { + "grad_norm": 0.16505104303359985, + "learning_rate": 3.875931954246504e-06, + "loss": 0.0025, + "step": 52810 + }, + { + "grad_norm": 0.06951690465211868, + "learning_rate": 3.865300490743351e-06, + "loss": 0.0023, + "step": 52820 + }, + { + "grad_norm": 0.0755576565861702, + "learning_rate": 3.854683041743806e-06, + "loss": 0.0032, + "step": 52830 + }, + { + "grad_norm": 0.11203145980834961, + "learning_rate": 3.844079610473139e-06, + "loss": 0.0017, + "step": 52840 + }, + { + "grad_norm": 0.056018002331256866, + "learning_rate": 3.833490200152423e-06, + "loss": 0.0019, + "step": 52850 + }, + { + "grad_norm": 0.12049967050552368, + "learning_rate": 3.822914813998424e-06, + "loss": 0.0018, + "step": 52860 + }, + { + "grad_norm": 0.08500275015830994, + "learning_rate": 3.812353455223666e-06, + "loss": 0.0027, + "step": 52870 + }, + { + "grad_norm": 0.07628992199897766, + "learning_rate": 3.8018061270364225e-06, + "loss": 0.0026, + "step": 52880 + }, + { + "grad_norm": 0.08573714643716812, + "learning_rate": 3.7912728326406688e-06, + "loss": 0.0028, + "step": 52890 + }, + { + "grad_norm": 0.061480812728405, + "learning_rate": 3.7807535752361732e-06, + "loss": 0.0016, + "step": 52900 + }, + { + "grad_norm": 0.09375834465026855, + "learning_rate": 3.7702483580183855e-06, + "loss": 0.0038, + "step": 52910 + }, + { + "grad_norm": 0.09858549386262894, + "learning_rate": 3.759757184178525e-06, + "loss": 0.0046, + "step": 52920 + }, + { + "grad_norm": 0.07815971225500107, + "learning_rate": 3.7492800569035312e-06, + "loss": 0.0029, + "step": 52930 + }, + { + "grad_norm": 0.10285542160272598, + "learning_rate": 3.7388169793760754e-06, + "loss": 0.0017, + "step": 52940 + }, + { + "grad_norm": 0.13717767596244812, + "learning_rate": 3.728367954774553e-06, + "loss": 0.0021, + "step": 52950 + }, + { + "grad_norm": 0.07499244064092636, + "learning_rate": 3.7179329862731317e-06, + "loss": 0.0013, + "step": 52960 + }, + { + "grad_norm": 0.08867670595645905, + "learning_rate": 3.707512077041647e-06, + "loss": 0.0021, + "step": 52970 + }, + { + "grad_norm": 0.09590643644332886, + "learning_rate": 3.6971052302457288e-06, + "loss": 0.0022, + "step": 52980 + }, + { + "grad_norm": 0.11790993809700012, + "learning_rate": 3.6867124490466698e-06, + "loss": 0.0019, + "step": 52990 + }, + { + "grad_norm": 0.06500764936208725, + "learning_rate": 3.6763337366015393e-06, + "loss": 0.0014, + "step": 53000 + }, + { + "grad_norm": 0.07242599129676819, + "learning_rate": 3.665969096063121e-06, + "loss": 0.0021, + "step": 53010 + }, + { + "grad_norm": 0.09093352407217026, + "learning_rate": 3.6556185305799074e-06, + "loss": 0.0022, + "step": 53020 + }, + { + "grad_norm": 0.08295711129903793, + "learning_rate": 3.645282043296133e-06, + "loss": 0.0018, + "step": 53030 + }, + { + "grad_norm": 0.059141773730516434, + "learning_rate": 3.6349596373517427e-06, + "loss": 0.0024, + "step": 53040 + }, + { + "grad_norm": 0.11438831686973572, + "learning_rate": 3.6246513158824215e-06, + "loss": 0.0025, + "step": 53050 + }, + { + "grad_norm": 0.07515961676836014, + "learning_rate": 3.6143570820195593e-06, + "loss": 0.0025, + "step": 53060 + }, + { + "grad_norm": 0.15036208927631378, + "learning_rate": 3.6040769388902773e-06, + "loss": 0.003, + "step": 53070 + }, + { + "grad_norm": 0.07366809993982315, + "learning_rate": 3.593810889617405e-06, + "loss": 0.0016, + "step": 53080 + }, + { + "grad_norm": 0.1306273639202118, + "learning_rate": 3.5835589373194978e-06, + "loss": 0.0057, + "step": 53090 + }, + { + "grad_norm": 0.11764480918645859, + "learning_rate": 3.5733210851108257e-06, + "loss": 0.0019, + "step": 53100 + }, + { + "grad_norm": 0.08813925832509995, + "learning_rate": 3.5630973361014008e-06, + "loss": 0.0023, + "step": 53110 + }, + { + "grad_norm": 0.07190808653831482, + "learning_rate": 3.552887693396889e-06, + "loss": 0.0019, + "step": 53120 + }, + { + "grad_norm": 0.09255789965391159, + "learning_rate": 3.542692160098754e-06, + "loss": 0.0022, + "step": 53130 + }, + { + "grad_norm": 0.13056059181690216, + "learning_rate": 3.5325107393040846e-06, + "loss": 0.0016, + "step": 53140 + }, + { + "grad_norm": 0.08913110196590424, + "learning_rate": 3.522343434105757e-06, + "loss": 0.0016, + "step": 53150 + }, + { + "grad_norm": 0.056747760623693466, + "learning_rate": 3.512190247592323e-06, + "loss": 0.0014, + "step": 53160 + }, + { + "grad_norm": 0.08582012355327606, + "learning_rate": 3.502051182848054e-06, + "loss": 0.0025, + "step": 53170 + }, + { + "grad_norm": 0.09070461243391037, + "learning_rate": 3.4919262429529308e-06, + "loss": 0.0019, + "step": 53180 + }, + { + "grad_norm": 0.07242628186941147, + "learning_rate": 3.4818154309826325e-06, + "loss": 0.0025, + "step": 53190 + }, + { + "grad_norm": 0.0876319631934166, + "learning_rate": 3.4717187500085734e-06, + "loss": 0.0019, + "step": 53200 + }, + { + "grad_norm": 0.07199565321207047, + "learning_rate": 3.46163620309784e-06, + "loss": 0.002, + "step": 53210 + }, + { + "grad_norm": 0.055618077516555786, + "learning_rate": 3.4515677933132595e-06, + "loss": 0.0016, + "step": 53220 + }, + { + "grad_norm": 0.05885442718863487, + "learning_rate": 3.4415135237133466e-06, + "loss": 0.0014, + "step": 53230 + }, + { + "grad_norm": 0.0822029858827591, + "learning_rate": 3.4314733973523196e-06, + "loss": 0.0019, + "step": 53240 + }, + { + "grad_norm": 0.05147421732544899, + "learning_rate": 3.4214474172800993e-06, + "loss": 0.0014, + "step": 53250 + }, + { + "grad_norm": 0.07879266887903214, + "learning_rate": 3.411435586542322e-06, + "loss": 0.0016, + "step": 53260 + }, + { + "grad_norm": 0.061414893716573715, + "learning_rate": 3.4014379081802995e-06, + "loss": 0.0018, + "step": 53270 + }, + { + "grad_norm": 0.07778201997280121, + "learning_rate": 3.391454385231102e-06, + "loss": 0.0019, + "step": 53280 + }, + { + "grad_norm": 0.07013441622257233, + "learning_rate": 3.3814850207274095e-06, + "loss": 0.0021, + "step": 53290 + }, + { + "grad_norm": 0.10824336856603622, + "learning_rate": 3.371529817697694e-06, + "loss": 0.0027, + "step": 53300 + }, + { + "grad_norm": 0.09860820323228836, + "learning_rate": 3.3615887791660585e-06, + "loss": 0.0024, + "step": 53310 + }, + { + "grad_norm": 0.08826929330825806, + "learning_rate": 3.3516619081523426e-06, + "loss": 0.0017, + "step": 53320 + }, + { + "grad_norm": 0.06803655624389648, + "learning_rate": 3.3417492076720567e-06, + "loss": 0.0035, + "step": 53330 + }, + { + "grad_norm": 0.05405272915959358, + "learning_rate": 3.3318506807364147e-06, + "loss": 0.0028, + "step": 53340 + }, + { + "grad_norm": 0.09187202900648117, + "learning_rate": 3.3219663303523553e-06, + "loss": 0.0019, + "step": 53350 + }, + { + "grad_norm": 0.07730714231729507, + "learning_rate": 3.3120961595224374e-06, + "loss": 0.002, + "step": 53360 + }, + { + "grad_norm": 0.10263852030038834, + "learning_rate": 3.302240171245002e-06, + "loss": 0.0029, + "step": 53370 + }, + { + "grad_norm": 0.08190537244081497, + "learning_rate": 3.29239836851401e-06, + "loss": 0.0026, + "step": 53380 + }, + { + "grad_norm": 0.0946035087108612, + "learning_rate": 3.2825707543191588e-06, + "loss": 0.0023, + "step": 53390 + }, + { + "grad_norm": 0.07189280539751053, + "learning_rate": 3.272757331645804e-06, + "loss": 0.002, + "step": 53400 + }, + { + "grad_norm": 0.05237928405404091, + "learning_rate": 3.2629581034750166e-06, + "loss": 0.002, + "step": 53410 + }, + { + "grad_norm": 0.07859513908624649, + "learning_rate": 3.2531730727835218e-06, + "loss": 0.0025, + "step": 53420 + }, + { + "grad_norm": 0.1653396189212799, + "learning_rate": 3.2434022425437914e-06, + "loss": 0.0027, + "step": 53430 + }, + { + "grad_norm": 0.07641898840665817, + "learning_rate": 3.233645615723907e-06, + "loss": 0.0044, + "step": 53440 + }, + { + "grad_norm": 0.0855804979801178, + "learning_rate": 3.2239031952876918e-06, + "loss": 0.0023, + "step": 53450 + }, + { + "grad_norm": 0.12052163481712341, + "learning_rate": 3.21417498419464e-06, + "loss": 0.002, + "step": 53460 + }, + { + "grad_norm": 0.11012811958789825, + "learning_rate": 3.204460985399921e-06, + "loss": 0.0027, + "step": 53470 + }, + { + "grad_norm": 0.06710916757583618, + "learning_rate": 3.1947612018543903e-06, + "loss": 0.0022, + "step": 53480 + }, + { + "grad_norm": 0.14122194051742554, + "learning_rate": 3.1850756365045753e-06, + "loss": 0.0017, + "step": 53490 + }, + { + "grad_norm": 0.07987992465496063, + "learning_rate": 3.175404292292722e-06, + "loss": 0.0023, + "step": 53500 + }, + { + "grad_norm": 0.06671053916215897, + "learning_rate": 3.1657471721566965e-06, + "loss": 0.0019, + "step": 53510 + }, + { + "grad_norm": 0.10298892110586166, + "learning_rate": 3.1561042790300977e-06, + "loss": 0.0017, + "step": 53520 + }, + { + "grad_norm": 0.06646808236837387, + "learning_rate": 3.1464756158421816e-06, + "loss": 0.0022, + "step": 53530 + }, + { + "grad_norm": 0.09134387969970703, + "learning_rate": 3.136861185517875e-06, + "loss": 0.0019, + "step": 53540 + }, + { + "grad_norm": 0.07304519414901733, + "learning_rate": 3.127260990977798e-06, + "loss": 0.0035, + "step": 53550 + }, + { + "grad_norm": 0.08984380960464478, + "learning_rate": 3.1176750351382235e-06, + "loss": 0.0034, + "step": 53560 + }, + { + "grad_norm": 0.06049609184265137, + "learning_rate": 3.1081033209111153e-06, + "loss": 0.0012, + "step": 53570 + }, + { + "grad_norm": 0.1328003853559494, + "learning_rate": 3.0985458512041155e-06, + "loss": 0.0026, + "step": 53580 + }, + { + "grad_norm": 0.06519491970539093, + "learning_rate": 3.089002628920512e-06, + "loss": 0.0022, + "step": 53590 + }, + { + "grad_norm": 0.08595166355371475, + "learning_rate": 3.079473656959303e-06, + "loss": 0.0017, + "step": 53600 + }, + { + "grad_norm": 0.2910189628601074, + "learning_rate": 3.0699589382151393e-06, + "loss": 0.0016, + "step": 53610 + }, + { + "grad_norm": 0.05026281625032425, + "learning_rate": 3.060458475578326e-06, + "loss": 0.0023, + "step": 53620 + }, + { + "grad_norm": 0.08304903656244278, + "learning_rate": 3.0509722719348656e-06, + "loss": 0.0024, + "step": 53630 + }, + { + "grad_norm": 0.05646422505378723, + "learning_rate": 3.041500330166408e-06, + "loss": 0.0014, + "step": 53640 + }, + { + "grad_norm": 0.07882935553789139, + "learning_rate": 3.032042653150291e-06, + "loss": 0.0035, + "step": 53650 + }, + { + "grad_norm": 0.17481666803359985, + "learning_rate": 3.0225992437594887e-06, + "loss": 0.0037, + "step": 53660 + }, + { + "grad_norm": 0.11247922480106354, + "learning_rate": 3.0131701048626783e-06, + "loss": 0.0026, + "step": 53670 + }, + { + "grad_norm": 0.06159326434135437, + "learning_rate": 3.003755239324163e-06, + "loss": 0.0022, + "step": 53680 + }, + { + "grad_norm": 0.07959142327308655, + "learning_rate": 2.9943546500039553e-06, + "loss": 0.0014, + "step": 53690 + }, + { + "grad_norm": 0.06081371009349823, + "learning_rate": 2.9849683397576877e-06, + "loss": 0.0018, + "step": 53700 + }, + { + "grad_norm": 0.07618306577205658, + "learning_rate": 2.975596311436679e-06, + "loss": 0.0016, + "step": 53710 + }, + { + "grad_norm": 0.10597028583288193, + "learning_rate": 2.966238567887902e-06, + "loss": 0.0028, + "step": 53720 + }, + { + "grad_norm": 0.06285658478736877, + "learning_rate": 2.9568951119539943e-06, + "loss": 0.0033, + "step": 53730 + }, + { + "grad_norm": 0.09460027515888214, + "learning_rate": 2.947565946473241e-06, + "loss": 0.0034, + "step": 53740 + }, + { + "grad_norm": 0.06624805927276611, + "learning_rate": 2.9382510742796188e-06, + "loss": 0.0016, + "step": 53750 + }, + { + "grad_norm": 0.09915328025817871, + "learning_rate": 2.9289504982027204e-06, + "loss": 0.0018, + "step": 53760 + }, + { + "grad_norm": 0.06237504631280899, + "learning_rate": 2.9196642210678248e-06, + "loss": 0.0016, + "step": 53770 + }, + { + "grad_norm": 0.07811791449785233, + "learning_rate": 2.910392245695853e-06, + "loss": 0.0026, + "step": 53780 + }, + { + "grad_norm": 0.06234855577349663, + "learning_rate": 2.9011345749033804e-06, + "loss": 0.0015, + "step": 53790 + }, + { + "grad_norm": 0.06481218338012695, + "learning_rate": 2.8918912115026677e-06, + "loss": 0.0019, + "step": 53800 + }, + { + "grad_norm": 0.08792757987976074, + "learning_rate": 2.8826621583015743e-06, + "loss": 0.0023, + "step": 53810 + }, + { + "grad_norm": 0.07245846837759018, + "learning_rate": 2.8734474181036643e-06, + "loss": 0.0024, + "step": 53820 + }, + { + "grad_norm": 0.07888416945934296, + "learning_rate": 2.864246993708114e-06, + "loss": 0.0019, + "step": 53830 + }, + { + "grad_norm": 0.1260496973991394, + "learning_rate": 2.8550608879097884e-06, + "loss": 0.0017, + "step": 53840 + }, + { + "grad_norm": 0.06381430476903915, + "learning_rate": 2.845889103499172e-06, + "loss": 0.0026, + "step": 53850 + }, + { + "grad_norm": 0.06190713122487068, + "learning_rate": 2.8367316432624138e-06, + "loss": 0.0015, + "step": 53860 + }, + { + "grad_norm": 0.09318691492080688, + "learning_rate": 2.8275885099813105e-06, + "loss": 0.0018, + "step": 53870 + }, + { + "grad_norm": 0.08710166066884995, + "learning_rate": 2.8184597064332963e-06, + "loss": 0.0026, + "step": 53880 + }, + { + "grad_norm": 0.05651054531335831, + "learning_rate": 2.809345235391464e-06, + "loss": 0.0017, + "step": 53890 + }, + { + "grad_norm": 0.08247019350528717, + "learning_rate": 2.80024509962456e-06, + "loss": 0.0035, + "step": 53900 + }, + { + "grad_norm": 0.0925832986831665, + "learning_rate": 2.7911593018969563e-06, + "loss": 0.0015, + "step": 53910 + }, + { + "grad_norm": 0.07202859967947006, + "learning_rate": 2.7820878449686838e-06, + "loss": 0.0027, + "step": 53920 + }, + { + "grad_norm": 0.06817338615655899, + "learning_rate": 2.7730307315953995e-06, + "loss": 0.0012, + "step": 53930 + }, + { + "grad_norm": 0.0424310527741909, + "learning_rate": 2.763987964528425e-06, + "loss": 0.0015, + "step": 53940 + }, + { + "grad_norm": 0.08619671314954758, + "learning_rate": 2.754959546514718e-06, + "loss": 0.0021, + "step": 53950 + }, + { + "grad_norm": 0.10410776734352112, + "learning_rate": 2.7459454802968576e-06, + "loss": 0.0022, + "step": 53960 + }, + { + "grad_norm": 0.06249423325061798, + "learning_rate": 2.7369457686131028e-06, + "loss": 0.0017, + "step": 53970 + }, + { + "grad_norm": 0.1610705703496933, + "learning_rate": 2.7279604141973004e-06, + "loss": 0.0025, + "step": 53980 + }, + { + "grad_norm": 0.08082373440265656, + "learning_rate": 2.7189894197789946e-06, + "loss": 0.0018, + "step": 53990 + }, + { + "grad_norm": 0.10258739441633224, + "learning_rate": 2.7100327880833055e-06, + "loss": 0.0015, + "step": 54000 + }, + { + "grad_norm": 0.0672992393374443, + "learning_rate": 2.70109052183104e-06, + "loss": 0.0019, + "step": 54010 + }, + { + "grad_norm": 0.07101277261972427, + "learning_rate": 2.69216262373862e-06, + "loss": 0.002, + "step": 54020 + }, + { + "grad_norm": 0.06556190550327301, + "learning_rate": 2.6832490965181036e-06, + "loss": 0.0018, + "step": 54030 + }, + { + "grad_norm": 0.09636989235877991, + "learning_rate": 2.6743499428771857e-06, + "loss": 0.0013, + "step": 54040 + }, + { + "grad_norm": 0.05531807616353035, + "learning_rate": 2.6654651655191875e-06, + "loss": 0.003, + "step": 54050 + }, + { + "grad_norm": 0.12198468297719955, + "learning_rate": 2.6565947671430836e-06, + "loss": 0.0015, + "step": 54060 + }, + { + "grad_norm": 0.0538400262594223, + "learning_rate": 2.647738750443457e-06, + "loss": 0.0015, + "step": 54070 + }, + { + "grad_norm": 0.10752842575311661, + "learning_rate": 2.6388971181105393e-06, + "loss": 0.0023, + "step": 54080 + }, + { + "grad_norm": 0.0972432792186737, + "learning_rate": 2.630069872830171e-06, + "loss": 0.0021, + "step": 54090 + }, + { + "grad_norm": 0.0944426953792572, + "learning_rate": 2.6212570172838514e-06, + "loss": 0.0018, + "step": 54100 + }, + { + "grad_norm": 0.08618577569723129, + "learning_rate": 2.6124585541486778e-06, + "loss": 0.0028, + "step": 54110 + }, + { + "grad_norm": 0.055841874331235886, + "learning_rate": 2.6036744860974127e-06, + "loss": 0.0015, + "step": 54120 + }, + { + "grad_norm": 0.1259326934814453, + "learning_rate": 2.594904815798399e-06, + "loss": 0.0019, + "step": 54130 + }, + { + "grad_norm": 0.09413226693868637, + "learning_rate": 2.586149545915656e-06, + "loss": 0.002, + "step": 54140 + }, + { + "grad_norm": 0.11932085454463959, + "learning_rate": 2.577408679108778e-06, + "loss": 0.0038, + "step": 54150 + }, + { + "grad_norm": 0.1117199957370758, + "learning_rate": 2.5686822180330306e-06, + "loss": 0.0026, + "step": 54160 + }, + { + "grad_norm": 0.10979679971933365, + "learning_rate": 2.5599701653392703e-06, + "loss": 0.0019, + "step": 54170 + }, + { + "grad_norm": 0.07426301389932632, + "learning_rate": 2.551272523673992e-06, + "loss": 0.0022, + "step": 54180 + }, + { + "grad_norm": 0.0741279125213623, + "learning_rate": 2.542589295679315e-06, + "loss": 0.0014, + "step": 54190 + }, + { + "grad_norm": 0.050029970705509186, + "learning_rate": 2.5339204839929575e-06, + "loss": 0.0014, + "step": 54200 + }, + { + "grad_norm": 0.0576753169298172, + "learning_rate": 2.525266091248296e-06, + "loss": 0.0027, + "step": 54210 + }, + { + "grad_norm": 0.06492278724908829, + "learning_rate": 2.5166261200743e-06, + "loss": 0.0016, + "step": 54220 + }, + { + "grad_norm": 0.096565842628479, + "learning_rate": 2.5080005730955646e-06, + "loss": 0.0018, + "step": 54230 + }, + { + "grad_norm": 0.10656166076660156, + "learning_rate": 2.499389452932299e-06, + "loss": 0.0019, + "step": 54240 + }, + { + "grad_norm": 0.0756705030798912, + "learning_rate": 2.4907927622003336e-06, + "loss": 0.0016, + "step": 54250 + }, + { + "grad_norm": 0.08639020472764969, + "learning_rate": 2.4822105035111177e-06, + "loss": 0.0021, + "step": 54260 + }, + { + "grad_norm": 0.07466516643762589, + "learning_rate": 2.4736426794717273e-06, + "loss": 0.002, + "step": 54270 + }, + { + "grad_norm": 0.10140484571456909, + "learning_rate": 2.4650892926848135e-06, + "loss": 0.0023, + "step": 54280 + }, + { + "grad_norm": 0.09284615516662598, + "learning_rate": 2.456550345748704e-06, + "loss": 0.0024, + "step": 54290 + }, + { + "grad_norm": 0.09955625236034393, + "learning_rate": 2.4480258412572733e-06, + "loss": 0.0027, + "step": 54300 + }, + { + "grad_norm": 0.0661083534359932, + "learning_rate": 2.4395157818000612e-06, + "loss": 0.0034, + "step": 54310 + }, + { + "grad_norm": 0.09930426627397537, + "learning_rate": 2.431020169962189e-06, + "loss": 0.0027, + "step": 54320 + }, + { + "grad_norm": 0.04940907657146454, + "learning_rate": 2.422539008324409e-06, + "loss": 0.0014, + "step": 54330 + }, + { + "grad_norm": 0.08067351579666138, + "learning_rate": 2.414072299463066e-06, + "loss": 0.0021, + "step": 54340 + }, + { + "grad_norm": 0.13216720521450043, + "learning_rate": 2.4056200459501186e-06, + "loss": 0.0032, + "step": 54350 + }, + { + "grad_norm": 0.20564766228199005, + "learning_rate": 2.397182250353147e-06, + "loss": 0.0026, + "step": 54360 + }, + { + "grad_norm": 0.0816926583647728, + "learning_rate": 2.388758915235334e-06, + "loss": 0.0016, + "step": 54370 + }, + { + "grad_norm": 0.0625016987323761, + "learning_rate": 2.380350043155455e-06, + "loss": 0.0018, + "step": 54380 + }, + { + "grad_norm": 0.057611215859651566, + "learning_rate": 2.371955636667911e-06, + "loss": 0.0018, + "step": 54390 + }, + { + "grad_norm": 0.059634532779455185, + "learning_rate": 2.3635756983227008e-06, + "loss": 0.0024, + "step": 54400 + }, + { + "grad_norm": 0.04650024697184563, + "learning_rate": 2.3552102306654278e-06, + "loss": 0.0016, + "step": 54410 + }, + { + "grad_norm": 0.07028137892484665, + "learning_rate": 2.346859236237292e-06, + "loss": 0.0019, + "step": 54420 + }, + { + "grad_norm": 0.09724060446023941, + "learning_rate": 2.3385227175751145e-06, + "loss": 0.0025, + "step": 54430 + }, + { + "grad_norm": 0.09171615540981293, + "learning_rate": 2.330200677211314e-06, + "loss": 0.0022, + "step": 54440 + }, + { + "grad_norm": 0.09668052196502686, + "learning_rate": 2.3218931176738847e-06, + "loss": 0.0017, + "step": 54450 + }, + { + "grad_norm": 0.10153709352016449, + "learning_rate": 2.313600041486469e-06, + "loss": 0.0016, + "step": 54460 + }, + { + "grad_norm": 0.07403513789176941, + "learning_rate": 2.3053214511682743e-06, + "loss": 0.0015, + "step": 54470 + }, + { + "grad_norm": 0.07765621691942215, + "learning_rate": 2.2970573492341163e-06, + "loss": 0.0026, + "step": 54480 + }, + { + "grad_norm": 0.1168711856007576, + "learning_rate": 2.288807738194415e-06, + "loss": 0.0024, + "step": 54490 + }, + { + "grad_norm": 0.18862630426883698, + "learning_rate": 2.2805726205551768e-06, + "loss": 0.0033, + "step": 54500 + }, + { + "grad_norm": 0.06827455013990402, + "learning_rate": 2.272351998818023e-06, + "loss": 0.0025, + "step": 54510 + }, + { + "grad_norm": 0.05661782622337341, + "learning_rate": 2.2641458754801505e-06, + "loss": 0.0018, + "step": 54520 + }, + { + "grad_norm": 0.07161679863929749, + "learning_rate": 2.2559542530343756e-06, + "loss": 0.0018, + "step": 54530 + }, + { + "grad_norm": 0.04397375509142876, + "learning_rate": 2.247777133969087e-06, + "loss": 0.0017, + "step": 54540 + }, + { + "grad_norm": 0.058602336794137955, + "learning_rate": 2.2396145207682795e-06, + "loss": 0.0019, + "step": 54550 + }, + { + "grad_norm": 0.06978696584701538, + "learning_rate": 2.231466415911543e-06, + "loss": 0.0014, + "step": 54560 + }, + { + "grad_norm": 0.057848718017339706, + "learning_rate": 2.2233328218740524e-06, + "loss": 0.0016, + "step": 54570 + }, + { + "grad_norm": 0.05732574686408043, + "learning_rate": 2.215213741126576e-06, + "loss": 0.0022, + "step": 54580 + }, + { + "grad_norm": 0.09724109619855881, + "learning_rate": 2.2071091761354912e-06, + "loss": 0.0046, + "step": 54590 + }, + { + "grad_norm": 0.09100291877985, + "learning_rate": 2.1990191293627337e-06, + "loss": 0.0017, + "step": 54600 + }, + { + "grad_norm": 0.0882551372051239, + "learning_rate": 2.1909436032658548e-06, + "loss": 0.0018, + "step": 54610 + }, + { + "grad_norm": 0.1002589613199234, + "learning_rate": 2.1828826002979806e-06, + "loss": 0.003, + "step": 54620 + }, + { + "grad_norm": 0.08412455767393112, + "learning_rate": 2.1748361229078362e-06, + "loss": 0.0017, + "step": 54630 + }, + { + "grad_norm": 0.1040043756365776, + "learning_rate": 2.1668041735397327e-06, + "loss": 0.0022, + "step": 54640 + }, + { + "grad_norm": 0.07975007593631744, + "learning_rate": 2.1587867546335514e-06, + "loss": 0.0014, + "step": 54650 + }, + { + "grad_norm": 0.03711052983999252, + "learning_rate": 2.1507838686247894e-06, + "loss": 0.0029, + "step": 54660 + }, + { + "grad_norm": 0.06248718127608299, + "learning_rate": 2.1427955179444957e-06, + "loss": 0.0016, + "step": 54670 + }, + { + "grad_norm": 0.04533735662698746, + "learning_rate": 2.13482170501933e-06, + "loss": 0.002, + "step": 54680 + }, + { + "grad_norm": 0.05845937877893448, + "learning_rate": 2.1268624322715202e-06, + "loss": 0.0016, + "step": 54690 + }, + { + "grad_norm": 0.07439695298671722, + "learning_rate": 2.118917702118889e-06, + "loss": 0.0022, + "step": 54700 + }, + { + "grad_norm": 0.0428188256919384, + "learning_rate": 2.1109875169748327e-06, + "loss": 0.0029, + "step": 54710 + }, + { + "grad_norm": 0.05631381645798683, + "learning_rate": 2.103071879248336e-06, + "loss": 0.0027, + "step": 54720 + }, + { + "grad_norm": 0.07006761431694031, + "learning_rate": 2.0951707913439478e-06, + "loss": 0.0035, + "step": 54730 + }, + { + "grad_norm": 0.05126510560512543, + "learning_rate": 2.0872842556618255e-06, + "loss": 0.0018, + "step": 54740 + }, + { + "grad_norm": 0.04206765070557594, + "learning_rate": 2.079412274597686e-06, + "loss": 0.002, + "step": 54750 + }, + { + "grad_norm": 0.06521360576152802, + "learning_rate": 2.0715548505428284e-06, + "loss": 0.0018, + "step": 54760 + }, + { + "grad_norm": 0.09802603721618652, + "learning_rate": 2.0637119858841258e-06, + "loss": 0.0014, + "step": 54770 + }, + { + "grad_norm": 0.04573657363653183, + "learning_rate": 2.055883683004034e-06, + "loss": 0.0019, + "step": 54780 + }, + { + "grad_norm": 0.0576496422290802, + "learning_rate": 2.0480699442806006e-06, + "loss": 0.0025, + "step": 54790 + }, + { + "grad_norm": 0.0925966128706932, + "learning_rate": 2.0402707720874105e-06, + "loss": 0.0017, + "step": 54800 + }, + { + "grad_norm": 0.06119620054960251, + "learning_rate": 2.032486168793668e-06, + "loss": 0.0049, + "step": 54810 + }, + { + "grad_norm": 0.11019590497016907, + "learning_rate": 2.024716136764104e-06, + "loss": 0.0026, + "step": 54820 + }, + { + "grad_norm": 0.08656821399927139, + "learning_rate": 2.0169606783590757e-06, + "loss": 0.0023, + "step": 54830 + }, + { + "grad_norm": 0.10480524599552155, + "learning_rate": 2.0092197959344638e-06, + "loss": 0.0017, + "step": 54840 + }, + { + "grad_norm": 0.1236996054649353, + "learning_rate": 2.0014934918417606e-06, + "loss": 0.0026, + "step": 54850 + }, + { + "grad_norm": 0.08634085208177567, + "learning_rate": 1.99378176842801e-06, + "loss": 0.0018, + "step": 54860 + }, + { + "grad_norm": 0.08186342567205429, + "learning_rate": 1.9860846280358224e-06, + "loss": 0.0016, + "step": 54870 + }, + { + "grad_norm": 0.054384924471378326, + "learning_rate": 1.978402073003394e-06, + "loss": 0.0016, + "step": 54880 + }, + { + "grad_norm": 0.06738172471523285, + "learning_rate": 1.9707341056644736e-06, + "loss": 0.0019, + "step": 54890 + }, + { + "grad_norm": 0.17576219141483307, + "learning_rate": 1.963080728348399e-06, + "loss": 0.0016, + "step": 54900 + }, + { + "grad_norm": 0.10620465129613876, + "learning_rate": 1.955441943380054e-06, + "loss": 0.0043, + "step": 54910 + }, + { + "grad_norm": 0.05007859319448471, + "learning_rate": 1.947817753079906e-06, + "loss": 0.0028, + "step": 54920 + }, + { + "grad_norm": 0.04470022767782211, + "learning_rate": 1.9402081597639785e-06, + "loss": 0.0017, + "step": 54930 + }, + { + "grad_norm": 0.0445830412209034, + "learning_rate": 1.9326131657438683e-06, + "loss": 0.0016, + "step": 54940 + }, + { + "grad_norm": 0.0491841658949852, + "learning_rate": 1.925032773326724e-06, + "loss": 0.0016, + "step": 54950 + }, + { + "grad_norm": 0.05340069159865379, + "learning_rate": 1.9174669848152916e-06, + "loss": 0.0021, + "step": 54960 + }, + { + "grad_norm": 0.060233116149902344, + "learning_rate": 1.9099158025078336e-06, + "loss": 0.002, + "step": 54970 + }, + { + "grad_norm": 0.08575578778982162, + "learning_rate": 1.90237922869822e-06, + "loss": 0.0016, + "step": 54980 + }, + { + "grad_norm": 0.04427783936262131, + "learning_rate": 1.8948572656758367e-06, + "loss": 0.0016, + "step": 54990 + }, + { + "grad_norm": 0.06264875829219818, + "learning_rate": 1.8873499157256834e-06, + "loss": 0.0017, + "step": 55000 + }, + { + "grad_norm": 0.04102405160665512, + "learning_rate": 1.879857181128286e-06, + "loss": 0.0033, + "step": 55010 + }, + { + "grad_norm": 0.08111399412155151, + "learning_rate": 1.8723790641597349e-06, + "loss": 0.0028, + "step": 55020 + }, + { + "grad_norm": 0.06301755458116531, + "learning_rate": 1.8649155670916906e-06, + "loss": 0.0017, + "step": 55030 + }, + { + "grad_norm": 0.09790223091840744, + "learning_rate": 1.8574666921913565e-06, + "loss": 0.0021, + "step": 55040 + }, + { + "grad_norm": 0.16046346724033356, + "learning_rate": 1.8500324417215166e-06, + "loss": 0.0021, + "step": 55050 + }, + { + "grad_norm": 0.05608521029353142, + "learning_rate": 1.8426128179404977e-06, + "loss": 0.0022, + "step": 55060 + }, + { + "grad_norm": 0.10060373693704605, + "learning_rate": 1.8352078231021807e-06, + "loss": 0.0022, + "step": 55070 + }, + { + "grad_norm": 0.06295880675315857, + "learning_rate": 1.8278174594560049e-06, + "loss": 0.002, + "step": 55080 + }, + { + "grad_norm": 0.08510169386863708, + "learning_rate": 1.82044172924698e-06, + "loss": 0.0017, + "step": 55090 + }, + { + "grad_norm": 0.10186333954334259, + "learning_rate": 1.813080634715636e-06, + "loss": 0.0026, + "step": 55100 + }, + { + "grad_norm": 0.059168219566345215, + "learning_rate": 1.8057341780981119e-06, + "loss": 0.0027, + "step": 55110 + }, + { + "grad_norm": 0.11927201598882675, + "learning_rate": 1.7984023616260338e-06, + "loss": 0.0019, + "step": 55120 + }, + { + "grad_norm": 0.05390871688723564, + "learning_rate": 1.7910851875266421e-06, + "loss": 0.0033, + "step": 55130 + }, + { + "grad_norm": 0.06839112937450409, + "learning_rate": 1.7837826580226757e-06, + "loss": 0.0018, + "step": 55140 + }, + { + "grad_norm": 0.06399611383676529, + "learning_rate": 1.7764947753324656e-06, + "loss": 0.0018, + "step": 55150 + }, + { + "grad_norm": 0.06992144137620926, + "learning_rate": 1.7692215416698799e-06, + "loss": 0.0022, + "step": 55160 + }, + { + "grad_norm": 0.11019055545330048, + "learning_rate": 1.7619629592443233e-06, + "loss": 0.0018, + "step": 55170 + }, + { + "grad_norm": 0.09786662459373474, + "learning_rate": 1.7547190302607709e-06, + "loss": 0.003, + "step": 55180 + }, + { + "grad_norm": 0.06553267687559128, + "learning_rate": 1.747489756919729e-06, + "loss": 0.0024, + "step": 55190 + }, + { + "grad_norm": 0.06700772047042847, + "learning_rate": 1.7402751414172802e-06, + "loss": 0.003, + "step": 55200 + }, + { + "grad_norm": 0.30077818036079407, + "learning_rate": 1.7330751859450044e-06, + "loss": 0.0017, + "step": 55210 + }, + { + "grad_norm": 0.062278904020786285, + "learning_rate": 1.7258898926900801e-06, + "loss": 0.0017, + "step": 55220 + }, + { + "grad_norm": 0.15523922443389893, + "learning_rate": 1.7187192638352002e-06, + "loss": 0.0016, + "step": 55230 + }, + { + "grad_norm": 0.04551497474312782, + "learning_rate": 1.7115633015586163e-06, + "loss": 0.0017, + "step": 55240 + }, + { + "grad_norm": 0.05989941954612732, + "learning_rate": 1.7044220080341178e-06, + "loss": 0.0021, + "step": 55250 + }, + { + "grad_norm": 0.06417721509933472, + "learning_rate": 1.6972953854310414e-06, + "loss": 0.0021, + "step": 55260 + }, + { + "grad_norm": 0.038663625717163086, + "learning_rate": 1.690183435914261e-06, + "loss": 0.0013, + "step": 55270 + }, + { + "grad_norm": 0.05625640228390694, + "learning_rate": 1.6830861616442206e-06, + "loss": 0.0019, + "step": 55280 + }, + { + "grad_norm": 0.059589698910713196, + "learning_rate": 1.6760035647768568e-06, + "loss": 0.0027, + "step": 55290 + }, + { + "grad_norm": 0.05900309979915619, + "learning_rate": 1.6689356474636985e-06, + "loss": 0.002, + "step": 55300 + }, + { + "grad_norm": 0.10812920331954956, + "learning_rate": 1.6618824118517784e-06, + "loss": 0.0019, + "step": 55310 + }, + { + "grad_norm": 0.07384541630744934, + "learning_rate": 1.654843860083688e-06, + "loss": 0.0019, + "step": 55320 + }, + { + "grad_norm": 0.055850252509117126, + "learning_rate": 1.647819994297556e-06, + "loss": 0.0013, + "step": 55330 + }, + { + "grad_norm": 0.06195919215679169, + "learning_rate": 1.6408108166270363e-06, + "loss": 0.0016, + "step": 55340 + }, + { + "grad_norm": 0.05866905301809311, + "learning_rate": 1.6338163292013486e-06, + "loss": 0.0033, + "step": 55350 + }, + { + "grad_norm": 0.06020285189151764, + "learning_rate": 1.6268365341452208e-06, + "loss": 0.0013, + "step": 55360 + }, + { + "grad_norm": 0.10561487078666687, + "learning_rate": 1.6198714335789345e-06, + "loss": 0.0019, + "step": 55370 + }, + { + "grad_norm": 0.06198647990822792, + "learning_rate": 1.612921029618303e-06, + "loss": 0.0015, + "step": 55380 + }, + { + "grad_norm": 0.056997813284397125, + "learning_rate": 1.6059853243746815e-06, + "loss": 0.0018, + "step": 55390 + }, + { + "grad_norm": 0.04382305592298508, + "learning_rate": 1.5990643199549404e-06, + "loss": 0.0021, + "step": 55400 + }, + { + "grad_norm": 0.09489727765321732, + "learning_rate": 1.5921580184615147e-06, + "loss": 0.0023, + "step": 55410 + }, + { + "grad_norm": 0.03671587258577347, + "learning_rate": 1.5852664219923374e-06, + "loss": 0.0014, + "step": 55420 + }, + { + "grad_norm": 0.06806158274412155, + "learning_rate": 1.5783895326409172e-06, + "loss": 0.002, + "step": 55430 + }, + { + "grad_norm": 0.04714014753699303, + "learning_rate": 1.5715273524962438e-06, + "loss": 0.0019, + "step": 55440 + }, + { + "grad_norm": 0.10765212029218674, + "learning_rate": 1.564679883642889e-06, + "loss": 0.0017, + "step": 55450 + }, + { + "grad_norm": 0.054942239075899124, + "learning_rate": 1.5578471281609276e-06, + "loss": 0.0016, + "step": 55460 + }, + { + "grad_norm": 0.06616050750017166, + "learning_rate": 1.551029088125966e-06, + "loss": 0.0017, + "step": 55470 + }, + { + "grad_norm": 0.06391224265098572, + "learning_rate": 1.544225765609142e-06, + "loss": 0.0014, + "step": 55480 + }, + { + "grad_norm": 0.052012715488672256, + "learning_rate": 1.53743716267713e-06, + "loss": 0.0017, + "step": 55490 + }, + { + "grad_norm": 0.06327026337385178, + "learning_rate": 1.5306632813921361e-06, + "loss": 0.0032, + "step": 55500 + }, + { + "grad_norm": 0.05464669689536095, + "learning_rate": 1.52390412381187e-06, + "loss": 0.0015, + "step": 55510 + }, + { + "grad_norm": 0.1308482438325882, + "learning_rate": 1.5171596919895948e-06, + "loss": 0.0024, + "step": 55520 + }, + { + "grad_norm": 0.07215800136327744, + "learning_rate": 1.5104299879740936e-06, + "loss": 0.0012, + "step": 55530 + }, + { + "grad_norm": 0.07016236335039139, + "learning_rate": 1.5037150138096701e-06, + "loss": 0.0029, + "step": 55540 + }, + { + "grad_norm": 0.03346060588955879, + "learning_rate": 1.4970147715361538e-06, + "loss": 0.0014, + "step": 55550 + }, + { + "grad_norm": 0.07614639401435852, + "learning_rate": 1.4903292631889054e-06, + "loss": 0.0024, + "step": 55560 + }, + { + "grad_norm": 0.03969057649374008, + "learning_rate": 1.483658490798795e-06, + "loss": 0.0014, + "step": 55570 + }, + { + "grad_norm": 0.04972626641392708, + "learning_rate": 1.4770024563922457e-06, + "loss": 0.0018, + "step": 55580 + }, + { + "grad_norm": 0.09137051552534103, + "learning_rate": 1.4703611619911628e-06, + "loss": 0.002, + "step": 55590 + }, + { + "grad_norm": 0.13551737368106842, + "learning_rate": 1.4637346096130155e-06, + "loss": 0.0025, + "step": 55600 + }, + { + "grad_norm": 0.04698452353477478, + "learning_rate": 1.4571228012707662e-06, + "loss": 0.0017, + "step": 55610 + }, + { + "grad_norm": 0.05467325448989868, + "learning_rate": 1.4505257389729132e-06, + "loss": 0.0014, + "step": 55620 + }, + { + "grad_norm": 0.04777725785970688, + "learning_rate": 1.4439434247234596e-06, + "loss": 0.0017, + "step": 55630 + }, + { + "grad_norm": 0.06600505113601685, + "learning_rate": 1.4373758605219445e-06, + "loss": 0.0022, + "step": 55640 + }, + { + "grad_norm": 0.07806967198848724, + "learning_rate": 1.4308230483634333e-06, + "loss": 0.0023, + "step": 55650 + }, + { + "grad_norm": 0.06175815314054489, + "learning_rate": 1.4242849902384724e-06, + "loss": 0.0012, + "step": 55660 + }, + { + "grad_norm": 0.07971997559070587, + "learning_rate": 1.4177616881331734e-06, + "loss": 0.002, + "step": 55670 + }, + { + "grad_norm": 0.09943245351314545, + "learning_rate": 1.4112531440291233e-06, + "loss": 0.0023, + "step": 55680 + }, + { + "grad_norm": 0.05424543842673302, + "learning_rate": 1.4047593599034624e-06, + "loss": 0.0036, + "step": 55690 + }, + { + "grad_norm": 0.05612809211015701, + "learning_rate": 1.3982803377288246e-06, + "loss": 0.0024, + "step": 55700 + }, + { + "grad_norm": 0.037639278918504715, + "learning_rate": 1.3918160794733681e-06, + "loss": 0.0019, + "step": 55710 + }, + { + "grad_norm": 0.10856914520263672, + "learning_rate": 1.3853665871007615e-06, + "loss": 0.0019, + "step": 55720 + }, + { + "grad_norm": 0.053919292986392975, + "learning_rate": 1.378931862570193e-06, + "loss": 0.0014, + "step": 55730 + }, + { + "grad_norm": 0.04358123242855072, + "learning_rate": 1.372511907836349e-06, + "loss": 0.0017, + "step": 55740 + }, + { + "grad_norm": 0.03900609537959099, + "learning_rate": 1.3661067248494586e-06, + "loss": 0.0015, + "step": 55750 + }, + { + "grad_norm": 0.06084549054503441, + "learning_rate": 1.3597163155552429e-06, + "loss": 0.0018, + "step": 55760 + }, + { + "grad_norm": 0.06153925880789757, + "learning_rate": 1.3533406818949434e-06, + "loss": 0.0021, + "step": 55770 + }, + { + "grad_norm": 0.0626777708530426, + "learning_rate": 1.3469798258053002e-06, + "loss": 0.0035, + "step": 55780 + }, + { + "grad_norm": 0.0658852681517601, + "learning_rate": 1.3406337492185672e-06, + "loss": 0.0034, + "step": 55790 + }, + { + "grad_norm": 0.07038184255361557, + "learning_rate": 1.3343024540625414e-06, + "loss": 0.0021, + "step": 55800 + }, + { + "grad_norm": 0.06997904926538467, + "learning_rate": 1.3279859422604735e-06, + "loss": 0.0026, + "step": 55810 + }, + { + "grad_norm": 0.0552855022251606, + "learning_rate": 1.3216842157311781e-06, + "loss": 0.0013, + "step": 55820 + }, + { + "grad_norm": 0.12711909413337708, + "learning_rate": 1.3153972763889355e-06, + "loss": 0.0021, + "step": 55830 + }, + { + "grad_norm": 0.06278539448976517, + "learning_rate": 1.3091251261435566e-06, + "loss": 0.0022, + "step": 55840 + }, + { + "grad_norm": 0.06203445419669151, + "learning_rate": 1.3028677669003564e-06, + "loss": 0.0017, + "step": 55850 + }, + { + "grad_norm": 0.06531459838151932, + "learning_rate": 1.2966252005601587e-06, + "loss": 0.0019, + "step": 55860 + }, + { + "grad_norm": 0.06296014040708542, + "learning_rate": 1.2903974290192855e-06, + "loss": 0.0027, + "step": 55870 + }, + { + "grad_norm": 0.04302409291267395, + "learning_rate": 1.284184454169568e-06, + "loss": 0.0022, + "step": 55880 + }, + { + "grad_norm": 0.04586528614163399, + "learning_rate": 1.2779862778983464e-06, + "loss": 0.0015, + "step": 55890 + }, + { + "grad_norm": 0.07205494493246078, + "learning_rate": 1.2718029020884647e-06, + "loss": 0.002, + "step": 55900 + }, + { + "grad_norm": 0.06471510231494904, + "learning_rate": 1.2656343286182703e-06, + "loss": 0.0023, + "step": 55910 + }, + { + "grad_norm": 0.07436578720808029, + "learning_rate": 1.2594805593616088e-06, + "loss": 0.0016, + "step": 55920 + }, + { + "grad_norm": 0.1629922240972519, + "learning_rate": 1.2533415961878404e-06, + "loss": 0.0016, + "step": 55930 + }, + { + "grad_norm": 0.07494357973337173, + "learning_rate": 1.2472174409618009e-06, + "loss": 0.0021, + "step": 55940 + }, + { + "grad_norm": 0.0752205103635788, + "learning_rate": 1.2411080955438747e-06, + "loss": 0.002, + "step": 55950 + }, + { + "grad_norm": 0.049530670046806335, + "learning_rate": 1.235013561789894e-06, + "loss": 0.0037, + "step": 55960 + }, + { + "grad_norm": 0.10082685202360153, + "learning_rate": 1.2289338415512385e-06, + "loss": 0.0018, + "step": 55970 + }, + { + "grad_norm": 0.054863251745700836, + "learning_rate": 1.222868936674748e-06, + "loss": 0.0012, + "step": 55980 + }, + { + "grad_norm": 0.057741425931453705, + "learning_rate": 1.2168188490027876e-06, + "loss": 0.0019, + "step": 55990 + }, + { + "grad_norm": 0.030309228226542473, + "learning_rate": 1.2107835803732204e-06, + "loss": 0.0012, + "step": 56000 + }, + { + "grad_norm": 0.03874985873699188, + "learning_rate": 1.2047631326193964e-06, + "loss": 0.0027, + "step": 56010 + }, + { + "grad_norm": 0.08432970941066742, + "learning_rate": 1.1987575075701696e-06, + "loss": 0.002, + "step": 56020 + }, + { + "grad_norm": 0.041321296244859695, + "learning_rate": 1.1927667070498916e-06, + "loss": 0.0015, + "step": 56030 + }, + { + "grad_norm": 0.09103959053754807, + "learning_rate": 1.1867907328784067e-06, + "loss": 0.0015, + "step": 56040 + }, + { + "grad_norm": 0.12052654474973679, + "learning_rate": 1.1808295868710518e-06, + "loss": 0.0017, + "step": 56050 + }, + { + "grad_norm": 0.04596017673611641, + "learning_rate": 1.174883270838678e-06, + "loss": 0.0014, + "step": 56060 + }, + { + "grad_norm": 0.09621170908212662, + "learning_rate": 1.1689517865876187e-06, + "loss": 0.0035, + "step": 56070 + }, + { + "grad_norm": 0.051761675626039505, + "learning_rate": 1.1630351359196933e-06, + "loss": 0.0026, + "step": 56080 + }, + { + "grad_norm": 0.12357889115810394, + "learning_rate": 1.1571333206322255e-06, + "loss": 0.0044, + "step": 56090 + }, + { + "grad_norm": 0.07829266041517258, + "learning_rate": 1.1512463425180365e-06, + "loss": 0.0024, + "step": 56100 + }, + { + "grad_norm": 0.07623567432165146, + "learning_rate": 1.145374203365429e-06, + "loss": 0.0018, + "step": 56110 + }, + { + "grad_norm": 0.13896743953227997, + "learning_rate": 1.1395169049582155e-06, + "loss": 0.0018, + "step": 56120 + }, + { + "grad_norm": 0.0428655743598938, + "learning_rate": 1.1336744490756722e-06, + "loss": 0.0014, + "step": 56130 + }, + { + "grad_norm": 0.10808618366718292, + "learning_rate": 1.1278468374925967e-06, + "loss": 0.0017, + "step": 56140 + }, + { + "grad_norm": 0.11178108304738998, + "learning_rate": 1.12203407197925e-06, + "loss": 0.0021, + "step": 56150 + }, + { + "grad_norm": 0.06898672133684158, + "learning_rate": 1.116236154301409e-06, + "loss": 0.0016, + "step": 56160 + }, + { + "grad_norm": 0.036168038845062256, + "learning_rate": 1.11045308622032e-06, + "loss": 0.0014, + "step": 56170 + }, + { + "grad_norm": 0.03508417308330536, + "learning_rate": 1.1046848694927337e-06, + "loss": 0.0022, + "step": 56180 + }, + { + "grad_norm": 0.07750049978494644, + "learning_rate": 1.098931505870876e-06, + "loss": 0.002, + "step": 56190 + }, + { + "grad_norm": 0.04469189792871475, + "learning_rate": 1.0931929971024657e-06, + "loss": 0.0024, + "step": 56200 + }, + { + "grad_norm": 0.09542711824178696, + "learning_rate": 1.0874693449307193e-06, + "loss": 0.0021, + "step": 56210 + }, + { + "grad_norm": 0.04920828714966774, + "learning_rate": 1.081760551094324e-06, + "loss": 0.0019, + "step": 56220 + }, + { + "grad_norm": 0.08028838038444519, + "learning_rate": 1.0760666173274592e-06, + "loss": 0.0028, + "step": 56230 + }, + { + "grad_norm": 0.04915463551878929, + "learning_rate": 1.0703875453597967e-06, + "loss": 0.0014, + "step": 56240 + }, + { + "grad_norm": 0.04620865732431412, + "learning_rate": 1.0647233369164845e-06, + "loss": 0.002, + "step": 56250 + }, + { + "grad_norm": 0.06789971888065338, + "learning_rate": 1.0590739937181625e-06, + "loss": 0.0018, + "step": 56260 + }, + { + "grad_norm": 0.07244718819856644, + "learning_rate": 1.053439517480953e-06, + "loss": 0.0022, + "step": 56270 + }, + { + "grad_norm": 0.06584817916154861, + "learning_rate": 1.047819909916453e-06, + "loss": 0.0019, + "step": 56280 + }, + { + "grad_norm": 0.0596274696290493, + "learning_rate": 1.0422151727317697e-06, + "loss": 0.0022, + "step": 56290 + }, + { + "grad_norm": 0.06513825803995132, + "learning_rate": 1.0366253076294462e-06, + "loss": 0.0017, + "step": 56300 + }, + { + "grad_norm": 0.050875693559646606, + "learning_rate": 1.031050316307558e-06, + "loss": 0.0012, + "step": 56310 + }, + { + "grad_norm": 0.11654133349657059, + "learning_rate": 1.0254902004596333e-06, + "loss": 0.0019, + "step": 56320 + }, + { + "grad_norm": 0.09935937821865082, + "learning_rate": 1.0199449617746882e-06, + "loss": 0.0017, + "step": 56330 + }, + { + "grad_norm": 0.050849977880716324, + "learning_rate": 1.0144146019372247e-06, + "loss": 0.0022, + "step": 56340 + }, + { + "grad_norm": 0.06618990749120712, + "learning_rate": 1.0088991226272048e-06, + "loss": 0.0017, + "step": 56350 + }, + { + "grad_norm": 0.05514708161354065, + "learning_rate": 1.003398525520105e-06, + "loss": 0.0018, + "step": 56360 + }, + { + "grad_norm": 0.07978447526693344, + "learning_rate": 9.979128122868552e-07, + "loss": 0.0018, + "step": 56370 + }, + { + "grad_norm": 0.03807741031050682, + "learning_rate": 9.924419845938614e-07, + "loss": 0.0017, + "step": 56380 + }, + { + "grad_norm": 0.07537122070789337, + "learning_rate": 9.869860441030276e-07, + "loss": 0.0017, + "step": 56390 + }, + { + "grad_norm": 0.07776734977960587, + "learning_rate": 9.815449924717169e-07, + "loss": 0.0017, + "step": 56400 + }, + { + "grad_norm": 0.06905701756477356, + "learning_rate": 9.761188313527791e-07, + "loss": 0.002, + "step": 56410 + }, + { + "grad_norm": 0.08049152046442032, + "learning_rate": 9.70707562394546e-07, + "loss": 0.0025, + "step": 56420 + }, + { + "grad_norm": 0.11917521059513092, + "learning_rate": 9.653111872408027e-07, + "loss": 0.002, + "step": 56430 + }, + { + "grad_norm": 0.07971474528312683, + "learning_rate": 9.599297075308434e-07, + "loss": 0.0022, + "step": 56440 + }, + { + "grad_norm": 0.03614998236298561, + "learning_rate": 9.545631248994048e-07, + "loss": 0.0015, + "step": 56450 + }, + { + "grad_norm": 0.0247368011623621, + "learning_rate": 9.492114409767217e-07, + "loss": 0.0018, + "step": 56460 + }, + { + "grad_norm": 0.05091584846377373, + "learning_rate": 9.438746573884938e-07, + "loss": 0.002, + "step": 56470 + }, + { + "grad_norm": 0.05932494252920151, + "learning_rate": 9.385527757558909e-07, + "loss": 0.0016, + "step": 56480 + }, + { + "grad_norm": 0.07539341598749161, + "learning_rate": 9.332457976955644e-07, + "loss": 0.0039, + "step": 56490 + }, + { + "grad_norm": 0.17069612443447113, + "learning_rate": 9.279537248196247e-07, + "loss": 0.0032, + "step": 56500 + }, + { + "grad_norm": 0.04852021113038063, + "learning_rate": 9.2267655873568e-07, + "loss": 0.0016, + "step": 56510 + }, + { + "grad_norm": 0.05712112411856651, + "learning_rate": 9.174143010467762e-07, + "loss": 0.0018, + "step": 56520 + }, + { + "grad_norm": 0.06710833311080933, + "learning_rate": 9.12166953351462e-07, + "loss": 0.0013, + "step": 56530 + }, + { + "grad_norm": 0.058544598519802094, + "learning_rate": 9.069345172437404e-07, + "loss": 0.0025, + "step": 56540 + }, + { + "grad_norm": 0.05729273334145546, + "learning_rate": 9.017169943130843e-07, + "loss": 0.0019, + "step": 56550 + }, + { + "grad_norm": 0.05204881727695465, + "learning_rate": 8.965143861444425e-07, + "loss": 0.0019, + "step": 56560 + }, + { + "grad_norm": 0.05346278101205826, + "learning_rate": 8.91326694318223e-07, + "loss": 0.0042, + "step": 56570 + }, + { + "grad_norm": 0.04623052105307579, + "learning_rate": 8.861539204103098e-07, + "loss": 0.0025, + "step": 56580 + }, + { + "grad_norm": 0.046255018562078476, + "learning_rate": 8.809960659920735e-07, + "loss": 0.002, + "step": 56590 + }, + { + "grad_norm": 0.03455488383769989, + "learning_rate": 8.758531326303055e-07, + "loss": 0.0039, + "step": 56600 + }, + { + "grad_norm": 0.05625394359230995, + "learning_rate": 8.707251218873169e-07, + "loss": 0.0016, + "step": 56610 + }, + { + "grad_norm": 0.09812338650226593, + "learning_rate": 8.656120353208507e-07, + "loss": 0.002, + "step": 56620 + }, + { + "grad_norm": 0.08530284464359283, + "learning_rate": 8.605138744841312e-07, + "loss": 0.0021, + "step": 56630 + }, + { + "grad_norm": 0.044126808643341064, + "learning_rate": 8.554306409258417e-07, + "loss": 0.0021, + "step": 56640 + }, + { + "grad_norm": 0.04840657487511635, + "learning_rate": 8.503623361901358e-07, + "loss": 0.0018, + "step": 56650 + }, + { + "grad_norm": 0.06996975094079971, + "learning_rate": 8.453089618166377e-07, + "loss": 0.0027, + "step": 56660 + }, + { + "grad_norm": 0.0683833658695221, + "learning_rate": 8.402705193404137e-07, + "loss": 0.0017, + "step": 56670 + }, + { + "grad_norm": 0.06494048982858658, + "learning_rate": 8.352470102920174e-07, + "loss": 0.0015, + "step": 56680 + }, + { + "grad_norm": 0.05499904975295067, + "learning_rate": 8.302384361974669e-07, + "loss": 0.0031, + "step": 56690 + }, + { + "grad_norm": 0.285785436630249, + "learning_rate": 8.252447985782231e-07, + "loss": 0.0025, + "step": 56700 + }, + { + "grad_norm": 0.05633128434419632, + "learning_rate": 8.202660989512279e-07, + "loss": 0.0021, + "step": 56710 + }, + { + "grad_norm": 0.06436681747436523, + "learning_rate": 8.153023388288772e-07, + "loss": 0.0023, + "step": 56720 + }, + { + "grad_norm": 0.099615179002285, + "learning_rate": 8.103535197190204e-07, + "loss": 0.0026, + "step": 56730 + }, + { + "grad_norm": 0.06618428975343704, + "learning_rate": 8.054196431249993e-07, + "loss": 0.0019, + "step": 56740 + }, + { + "grad_norm": 0.12161441892385483, + "learning_rate": 8.005007105455709e-07, + "loss": 0.0019, + "step": 56750 + }, + { + "grad_norm": 0.06798754632472992, + "learning_rate": 7.95596723474995e-07, + "loss": 0.0014, + "step": 56760 + }, + { + "grad_norm": 0.035703256726264954, + "learning_rate": 7.907076834029692e-07, + "loss": 0.0013, + "step": 56770 + }, + { + "grad_norm": 0.05448156222701073, + "learning_rate": 7.858335918146498e-07, + "loss": 0.0022, + "step": 56780 + }, + { + "grad_norm": 0.04690316319465637, + "learning_rate": 7.809744501906635e-07, + "loss": 0.0021, + "step": 56790 + }, + { + "grad_norm": 0.046710677444934845, + "learning_rate": 7.761302600070797e-07, + "loss": 0.0019, + "step": 56800 + }, + { + "grad_norm": 0.05189380422234535, + "learning_rate": 7.713010227354545e-07, + "loss": 0.0018, + "step": 56810 + }, + { + "grad_norm": 0.04737582430243492, + "learning_rate": 7.664867398427589e-07, + "loss": 0.0014, + "step": 56820 + }, + { + "grad_norm": 0.09771882742643356, + "learning_rate": 7.616874127914619e-07, + "loss": 0.0018, + "step": 56830 + }, + { + "grad_norm": 0.09414016455411911, + "learning_rate": 7.569030430394641e-07, + "loss": 0.0028, + "step": 56840 + }, + { + "grad_norm": 0.048337697982788086, + "learning_rate": 7.521336320401306e-07, + "loss": 0.0019, + "step": 56850 + }, + { + "grad_norm": 0.0667913556098938, + "learning_rate": 7.473791812422915e-07, + "loss": 0.0018, + "step": 56860 + }, + { + "grad_norm": 0.04438870772719383, + "learning_rate": 7.426396920902134e-07, + "loss": 0.0019, + "step": 56870 + }, + { + "grad_norm": 0.06593619287014008, + "learning_rate": 7.379151660236283e-07, + "loss": 0.0039, + "step": 56880 + }, + { + "grad_norm": 0.0466214120388031, + "learning_rate": 7.332056044777324e-07, + "loss": 0.002, + "step": 56890 + }, + { + "grad_norm": 0.05017360299825668, + "learning_rate": 7.285110088831537e-07, + "loss": 0.0024, + "step": 56900 + }, + { + "grad_norm": 0.17466281354427338, + "learning_rate": 7.238313806659902e-07, + "loss": 0.0025, + "step": 56910 + }, + { + "grad_norm": 0.05216871201992035, + "learning_rate": 7.191667212477993e-07, + "loss": 0.0027, + "step": 56920 + }, + { + "grad_norm": 0.0463014654815197, + "learning_rate": 7.145170320455697e-07, + "loss": 0.0019, + "step": 56930 + }, + { + "grad_norm": 0.054463278502225876, + "learning_rate": 7.098823144717604e-07, + "loss": 0.0022, + "step": 56940 + }, + { + "grad_norm": 0.08143548667430878, + "learning_rate": 7.052625699342674e-07, + "loss": 0.0017, + "step": 56950 + }, + { + "grad_norm": 0.02746555395424366, + "learning_rate": 7.006577998364628e-07, + "loss": 0.0024, + "step": 56960 + }, + { + "grad_norm": 0.1048780083656311, + "learning_rate": 6.96068005577133e-07, + "loss": 0.0021, + "step": 56970 + }, + { + "grad_norm": 0.16140638291835785, + "learning_rate": 6.914931885505627e-07, + "loss": 0.0015, + "step": 56980 + }, + { + "grad_norm": 0.03862803801894188, + "learning_rate": 6.869333501464347e-07, + "loss": 0.0018, + "step": 56990 + }, + { + "grad_norm": 0.051892779767513275, + "learning_rate": 6.823884917499246e-07, + "loss": 0.0035, + "step": 57000 + }, + { + "grad_norm": 0.08538272976875305, + "learning_rate": 6.778586147416278e-07, + "loss": 0.0014, + "step": 57010 + }, + { + "grad_norm": 0.03845364972949028, + "learning_rate": 6.733437204976156e-07, + "loss": 0.001, + "step": 57020 + }, + { + "grad_norm": 0.10623897612094879, + "learning_rate": 6.688438103893857e-07, + "loss": 0.0042, + "step": 57030 + }, + { + "grad_norm": 0.09667661786079407, + "learning_rate": 6.64358885783889e-07, + "loss": 0.0026, + "step": 57040 + }, + { + "grad_norm": 0.051102884113788605, + "learning_rate": 6.598889480435299e-07, + "loss": 0.0019, + "step": 57050 + }, + { + "grad_norm": 0.06881003826856613, + "learning_rate": 6.554339985261615e-07, + "loss": 0.0016, + "step": 57060 + }, + { + "grad_norm": 0.061392948031425476, + "learning_rate": 6.509940385850733e-07, + "loss": 0.0016, + "step": 57070 + }, + { + "grad_norm": 0.10281597077846527, + "learning_rate": 6.465690695690141e-07, + "loss": 0.0018, + "step": 57080 + }, + { + "grad_norm": 0.042001206427812576, + "learning_rate": 6.421590928221699e-07, + "loss": 0.0014, + "step": 57090 + }, + { + "grad_norm": 0.05065804347395897, + "learning_rate": 6.377641096841691e-07, + "loss": 0.0014, + "step": 57100 + }, + { + "grad_norm": 0.030709296464920044, + "learning_rate": 6.333841214901048e-07, + "loss": 0.0017, + "step": 57110 + }, + { + "grad_norm": 0.04658455401659012, + "learning_rate": 6.290191295704906e-07, + "loss": 0.002, + "step": 57120 + }, + { + "grad_norm": 0.09820462018251419, + "learning_rate": 6.246691352513046e-07, + "loss": 0.0032, + "step": 57130 + }, + { + "grad_norm": 0.06811260432004929, + "learning_rate": 6.203341398539452e-07, + "loss": 0.0015, + "step": 57140 + }, + { + "grad_norm": 0.03459732607007027, + "learning_rate": 6.160141446952872e-07, + "loss": 0.0014, + "step": 57150 + }, + { + "grad_norm": 0.04866626858711243, + "learning_rate": 6.11709151087625e-07, + "loss": 0.0016, + "step": 57160 + }, + { + "grad_norm": 0.041527654975652695, + "learning_rate": 6.074191603386958e-07, + "loss": 0.0014, + "step": 57170 + }, + { + "grad_norm": 0.03288004919886589, + "learning_rate": 6.031441737516907e-07, + "loss": 0.0013, + "step": 57180 + }, + { + "grad_norm": 0.060558710247278214, + "learning_rate": 5.988841926252431e-07, + "loss": 0.0028, + "step": 57190 + }, + { + "grad_norm": 0.08243124186992645, + "learning_rate": 5.946392182534066e-07, + "loss": 0.0019, + "step": 57200 + }, + { + "grad_norm": 0.05785783380270004, + "learning_rate": 5.90409251925711e-07, + "loss": 0.0016, + "step": 57210 + }, + { + "grad_norm": 0.04605141282081604, + "learning_rate": 5.861942949270949e-07, + "loss": 0.005, + "step": 57220 + }, + { + "grad_norm": 0.031177379190921783, + "learning_rate": 5.819943485379564e-07, + "loss": 0.0016, + "step": 57230 + }, + { + "grad_norm": 0.042267605662345886, + "learning_rate": 5.778094140341306e-07, + "loss": 0.0018, + "step": 57240 + }, + { + "grad_norm": 0.05929802730679512, + "learning_rate": 5.736394926868893e-07, + "loss": 0.0015, + "step": 57250 + }, + { + "grad_norm": 0.04832405596971512, + "learning_rate": 5.694845857629416e-07, + "loss": 0.0018, + "step": 57260 + }, + { + "grad_norm": 0.072139210999012, + "learning_rate": 5.653446945244334e-07, + "loss": 0.0014, + "step": 57270 + }, + { + "grad_norm": 0.05935519561171532, + "learning_rate": 5.612198202289698e-07, + "loss": 0.0023, + "step": 57280 + }, + { + "grad_norm": 0.045992977917194366, + "learning_rate": 5.571099641295596e-07, + "loss": 0.0026, + "step": 57290 + }, + { + "grad_norm": 0.05129547789692879, + "learning_rate": 5.530151274746875e-07, + "loss": 0.0014, + "step": 57300 + }, + { + "grad_norm": 0.03503677621483803, + "learning_rate": 5.489353115082418e-07, + "loss": 0.0015, + "step": 57310 + }, + { + "grad_norm": 0.0757785364985466, + "learning_rate": 5.4487051746957e-07, + "loss": 0.0021, + "step": 57320 + }, + { + "grad_norm": 0.033482201397418976, + "learning_rate": 5.408207465934511e-07, + "loss": 0.0021, + "step": 57330 + }, + { + "grad_norm": 0.06210426613688469, + "learning_rate": 5.3678600011009e-07, + "loss": 0.0013, + "step": 57340 + }, + { + "grad_norm": 0.05117350071668625, + "learning_rate": 5.327662792451449e-07, + "loss": 0.0028, + "step": 57350 + }, + { + "grad_norm": 0.04117773473262787, + "learning_rate": 5.287615852196947e-07, + "loss": 0.0018, + "step": 57360 + }, + { + "grad_norm": 0.05955406650900841, + "learning_rate": 5.247719192502665e-07, + "loss": 0.0016, + "step": 57370 + }, + { + "grad_norm": 0.029013805091381073, + "learning_rate": 5.207972825488128e-07, + "loss": 0.0026, + "step": 57380 + }, + { + "grad_norm": 0.03855122998356819, + "learning_rate": 5.168376763227178e-07, + "loss": 0.002, + "step": 57390 + }, + { + "grad_norm": 0.03751670569181442, + "learning_rate": 5.12893101774814e-07, + "loss": 0.002, + "step": 57400 + }, + { + "grad_norm": 0.12298837304115295, + "learning_rate": 5.089635601033483e-07, + "loss": 0.0014, + "step": 57410 + }, + { + "grad_norm": 0.044152356684207916, + "learning_rate": 5.050490525020213e-07, + "loss": 0.002, + "step": 57420 + }, + { + "grad_norm": 0.03840535134077072, + "learning_rate": 5.011495801599541e-07, + "loss": 0.0021, + "step": 57430 + }, + { + "grad_norm": 0.04537270963191986, + "learning_rate": 4.972651442616994e-07, + "loss": 0.0019, + "step": 57440 + }, + { + "grad_norm": 0.036976296454668045, + "learning_rate": 4.933957459872574e-07, + "loss": 0.0034, + "step": 57450 + }, + { + "grad_norm": 0.06119784340262413, + "learning_rate": 4.895413865120324e-07, + "loss": 0.0016, + "step": 57460 + }, + { + "grad_norm": 0.06850314140319824, + "learning_rate": 4.857020670068935e-07, + "loss": 0.002, + "step": 57470 + }, + { + "grad_norm": 0.026897938922047615, + "learning_rate": 4.818777886381132e-07, + "loss": 0.0011, + "step": 57480 + }, + { + "grad_norm": 0.04337281733751297, + "learning_rate": 4.780685525674122e-07, + "loss": 0.0018, + "step": 57490 + }, + { + "grad_norm": 0.034391701221466064, + "learning_rate": 4.7427435995193724e-07, + "loss": 0.0017, + "step": 57500 + }, + { + "grad_norm": 0.03225114941596985, + "learning_rate": 4.7049521194425515e-07, + "loss": 0.0021, + "step": 57510 + }, + { + "grad_norm": 0.024797815829515457, + "learning_rate": 4.667311096923754e-07, + "loss": 0.0019, + "step": 57520 + }, + { + "grad_norm": 0.05999648571014404, + "learning_rate": 4.6298205433973895e-07, + "loss": 0.0018, + "step": 57530 + }, + { + "grad_norm": 0.06903325766324997, + "learning_rate": 4.5924804702520696e-07, + "loss": 0.0027, + "step": 57540 + }, + { + "grad_norm": 0.05884411931037903, + "learning_rate": 4.5552908888306655e-07, + "loss": 0.0026, + "step": 57550 + }, + { + "grad_norm": 0.029378321021795273, + "learning_rate": 4.5182518104304185e-07, + "loss": 0.0042, + "step": 57560 + }, + { + "grad_norm": 0.0833432748913765, + "learning_rate": 4.4813632463028277e-07, + "loss": 0.0015, + "step": 57570 + }, + { + "grad_norm": 0.03921504318714142, + "learning_rate": 4.444625207653763e-07, + "loss": 0.0017, + "step": 57580 + }, + { + "grad_norm": 0.04709751904010773, + "learning_rate": 4.4080377056430753e-07, + "loss": 0.0018, + "step": 57590 + }, + { + "grad_norm": 0.040002815425395966, + "learning_rate": 4.371600751385263e-07, + "loss": 0.0023, + "step": 57600 + }, + { + "grad_norm": 0.04730124771595001, + "learning_rate": 4.3353143559487495e-07, + "loss": 0.0012, + "step": 57610 + }, + { + "grad_norm": 0.07143151760101318, + "learning_rate": 4.2991785303565513e-07, + "loss": 0.0031, + "step": 57620 + }, + { + "grad_norm": 0.07616531848907471, + "learning_rate": 4.2631932855856647e-07, + "loss": 0.0018, + "step": 57630 + }, + { + "grad_norm": 0.03569434583187103, + "learning_rate": 4.2273586325674576e-07, + "loss": 0.0031, + "step": 57640 + }, + { + "grad_norm": 0.04422670602798462, + "learning_rate": 4.1916745821876103e-07, + "loss": 0.0035, + "step": 57650 + }, + { + "grad_norm": 0.03967597335577011, + "learning_rate": 4.156141145285897e-07, + "loss": 0.0015, + "step": 57660 + }, + { + "grad_norm": 0.05007532611489296, + "learning_rate": 4.1207583326566267e-07, + "loss": 0.0018, + "step": 57670 + }, + { + "grad_norm": 0.03677612170577049, + "learning_rate": 4.085526155047925e-07, + "loss": 0.0016, + "step": 57680 + }, + { + "grad_norm": 0.10075671970844269, + "learning_rate": 4.050444623162564e-07, + "loss": 0.003, + "step": 57690 + }, + { + "grad_norm": 0.030502984300255775, + "learning_rate": 4.015513747657351e-07, + "loss": 0.0022, + "step": 57700 + }, + { + "grad_norm": 0.06658079475164413, + "learning_rate": 3.9807335391433554e-07, + "loss": 0.0038, + "step": 57710 + }, + { + "grad_norm": 0.029139438644051552, + "learning_rate": 3.946104008185847e-07, + "loss": 0.0015, + "step": 57720 + }, + { + "grad_norm": 0.061930879950523376, + "learning_rate": 3.9116251653044113e-07, + "loss": 0.0018, + "step": 57730 + }, + { + "grad_norm": 0.05950324982404709, + "learning_rate": 3.877297020972781e-07, + "loss": 0.0022, + "step": 57740 + }, + { + "grad_norm": 0.05202266946434975, + "learning_rate": 3.8431195856190036e-07, + "loss": 0.0015, + "step": 57750 + }, + { + "grad_norm": 0.042618997395038605, + "learning_rate": 3.8090928696251635e-07, + "loss": 0.0024, + "step": 57760 + }, + { + "grad_norm": 0.0490972101688385, + "learning_rate": 3.77521688332777e-07, + "loss": 0.0019, + "step": 57770 + }, + { + "grad_norm": 0.0705968514084816, + "learning_rate": 3.741491637017425e-07, + "loss": 0.0015, + "step": 57780 + }, + { + "grad_norm": 0.04218076914548874, + "learning_rate": 3.707917140939043e-07, + "loss": 0.0019, + "step": 57790 + }, + { + "grad_norm": 0.027439618483185768, + "learning_rate": 3.6744934052915235e-07, + "loss": 0.0026, + "step": 57800 + }, + { + "grad_norm": 0.08499769866466522, + "learning_rate": 3.641220440228188e-07, + "loss": 0.0015, + "step": 57810 + }, + { + "grad_norm": 0.07947226613759995, + "learning_rate": 3.608098255856562e-07, + "loss": 0.0029, + "step": 57820 + }, + { + "grad_norm": 0.035306788980960846, + "learning_rate": 3.575126862238154e-07, + "loss": 0.0019, + "step": 57830 + }, + { + "grad_norm": 0.07461398839950562, + "learning_rate": 3.5423062693888955e-07, + "loss": 0.003, + "step": 57840 + }, + { + "grad_norm": 0.05418021231889725, + "learning_rate": 3.509636487278756e-07, + "loss": 0.0013, + "step": 57850 + }, + { + "grad_norm": 0.03223951905965805, + "learning_rate": 3.4771175258320186e-07, + "loss": 0.0015, + "step": 57860 + }, + { + "grad_norm": 0.10770311951637268, + "learning_rate": 3.4447493949270047e-07, + "loss": 0.0022, + "step": 57870 + }, + { + "grad_norm": 0.0410153791308403, + "learning_rate": 3.4125321043964045e-07, + "loss": 0.0014, + "step": 57880 + }, + { + "grad_norm": 0.036266446113586426, + "learning_rate": 3.380465664026833e-07, + "loss": 0.002, + "step": 57890 + }, + { + "grad_norm": 0.04671436920762062, + "learning_rate": 3.348550083559388e-07, + "loss": 0.0013, + "step": 57900 + }, + { + "grad_norm": 0.036069322377443314, + "learning_rate": 3.316785372689091e-07, + "loss": 0.0016, + "step": 57910 + }, + { + "grad_norm": 0.05417667701840401, + "learning_rate": 3.285171541065224e-07, + "loss": 0.0015, + "step": 57920 + }, + { + "grad_norm": 0.035045161843299866, + "learning_rate": 3.253708598291272e-07, + "loss": 0.0021, + "step": 57930 + }, + { + "grad_norm": 0.0699935331940651, + "learning_rate": 3.2223965539248116e-07, + "loss": 0.0014, + "step": 57940 + }, + { + "grad_norm": 0.03070792928338051, + "learning_rate": 3.1912354174776227e-07, + "loss": 0.0014, + "step": 57950 + }, + { + "grad_norm": 0.037877488881349564, + "learning_rate": 3.1602251984155786e-07, + "loss": 0.0014, + "step": 57960 + }, + { + "grad_norm": 0.11186408996582031, + "learning_rate": 3.1293659061589207e-07, + "loss": 0.0024, + "step": 57970 + }, + { + "grad_norm": 0.039434753358364105, + "learning_rate": 3.098657550081707e-07, + "loss": 0.0014, + "step": 57980 + }, + { + "grad_norm": 0.04078147932887077, + "learning_rate": 3.068100139512475e-07, + "loss": 0.0026, + "step": 57990 + }, + { + "grad_norm": 0.033300504088401794, + "learning_rate": 3.037693683733689e-07, + "loss": 0.002, + "step": 58000 + }, + { + "grad_norm": 0.028592251241207123, + "learning_rate": 3.007438191982015e-07, + "loss": 0.0023, + "step": 58010 + }, + { + "grad_norm": 0.14351193606853485, + "learning_rate": 2.9773336734482684e-07, + "loss": 0.0014, + "step": 58020 + }, + { + "grad_norm": 0.03563110530376434, + "learning_rate": 2.9473801372774667e-07, + "loss": 0.0015, + "step": 58030 + }, + { + "grad_norm": 0.04076527804136276, + "learning_rate": 2.91757759256861e-07, + "loss": 0.0018, + "step": 58040 + }, + { + "grad_norm": 0.04505648463964462, + "learning_rate": 2.887926048375067e-07, + "loss": 0.0018, + "step": 58050 + }, + { + "grad_norm": 0.08677639067173004, + "learning_rate": 2.858425513704022e-07, + "loss": 0.0023, + "step": 58060 + }, + { + "grad_norm": 0.024720491841435432, + "learning_rate": 2.8290759975170834e-07, + "loss": 0.0024, + "step": 58070 + }, + { + "grad_norm": 0.10004983097314835, + "learning_rate": 2.799877508729787e-07, + "loss": 0.0048, + "step": 58080 + }, + { + "grad_norm": 0.04419567808508873, + "learning_rate": 2.770830056211926e-07, + "loss": 0.0014, + "step": 58090 + }, + { + "grad_norm": 0.12463217228651047, + "learning_rate": 2.741933648787331e-07, + "loss": 0.0023, + "step": 58100 + }, + { + "grad_norm": 0.02443358488380909, + "learning_rate": 2.7131882952339263e-07, + "loss": 0.002, + "step": 58110 + }, + { + "grad_norm": 0.04266953840851784, + "learning_rate": 2.684594004283836e-07, + "loss": 0.0022, + "step": 58120 + }, + { + "grad_norm": 0.04970623552799225, + "learning_rate": 2.6561507846232234e-07, + "loss": 0.0037, + "step": 58130 + }, + { + "grad_norm": 0.05502042546868324, + "learning_rate": 2.6278586448924005e-07, + "loss": 0.0024, + "step": 58140 + }, + { + "grad_norm": 0.053655046969652176, + "learning_rate": 2.5997175936857685e-07, + "loss": 0.0016, + "step": 58150 + }, + { + "grad_norm": 0.057660698890686035, + "learning_rate": 2.57172763955188e-07, + "loss": 0.0022, + "step": 58160 + }, + { + "grad_norm": 0.06838681548833847, + "learning_rate": 2.543888790993265e-07, + "loss": 0.0019, + "step": 58170 + }, + { + "grad_norm": 0.060685619711875916, + "learning_rate": 2.5162010564666607e-07, + "loss": 0.0017, + "step": 58180 + }, + { + "grad_norm": 0.14120259881019592, + "learning_rate": 2.488664444382893e-07, + "loss": 0.0018, + "step": 58190 + }, + { + "grad_norm": 0.06408952176570892, + "learning_rate": 2.461278963106828e-07, + "loss": 0.0022, + "step": 58200 + }, + { + "grad_norm": 0.036832425743341446, + "learning_rate": 2.434044620957421e-07, + "loss": 0.0026, + "step": 58210 + }, + { + "grad_norm": 0.03708605840802193, + "learning_rate": 2.406961426207832e-07, + "loss": 0.0014, + "step": 58220 + }, + { + "grad_norm": 0.025073770433664322, + "learning_rate": 2.380029387085203e-07, + "loss": 0.0014, + "step": 58230 + }, + { + "grad_norm": 0.040609683841466904, + "learning_rate": 2.353248511770767e-07, + "loss": 0.0017, + "step": 58240 + }, + { + "grad_norm": 0.026789173483848572, + "learning_rate": 2.3266188083997942e-07, + "loss": 0.0017, + "step": 58250 + }, + { + "grad_norm": 0.05461307615041733, + "learning_rate": 2.3001402850617027e-07, + "loss": 0.0027, + "step": 58260 + }, + { + "grad_norm": 0.03069292940199375, + "learning_rate": 2.2738129498000581e-07, + "loss": 0.002, + "step": 58270 + }, + { + "grad_norm": 0.03969322144985199, + "learning_rate": 2.2476368106122414e-07, + "loss": 0.0018, + "step": 58280 + }, + { + "grad_norm": 0.10775385051965714, + "learning_rate": 2.2216118754500582e-07, + "loss": 0.0023, + "step": 58290 + }, + { + "grad_norm": 0.07354037463665009, + "learning_rate": 2.1957381522190735e-07, + "loss": 0.0016, + "step": 58300 + }, + { + "grad_norm": 0.039317790418863297, + "learning_rate": 2.1700156487790558e-07, + "loss": 0.0018, + "step": 58310 + }, + { + "grad_norm": 0.026262616738677025, + "learning_rate": 2.1444443729439213e-07, + "loss": 0.0012, + "step": 58320 + }, + { + "grad_norm": 0.06699593365192413, + "learning_rate": 2.1190243324814007e-07, + "loss": 0.0021, + "step": 58330 + }, + { + "grad_norm": 0.10579883307218552, + "learning_rate": 2.0937555351135395e-07, + "loss": 0.0022, + "step": 58340 + }, + { + "grad_norm": 0.08626481890678406, + "learning_rate": 2.0686379885162532e-07, + "loss": 0.0021, + "step": 58350 + }, + { + "grad_norm": 0.04022153094410896, + "learning_rate": 2.0436717003196604e-07, + "loss": 0.0015, + "step": 58360 + }, + { + "grad_norm": 0.0913291722536087, + "learning_rate": 2.0188566781078057e-07, + "loss": 0.0019, + "step": 58370 + }, + { + "grad_norm": 0.030319202691316605, + "learning_rate": 1.994192929418881e-07, + "loss": 0.0015, + "step": 58380 + }, + { + "grad_norm": 0.03326492756605148, + "learning_rate": 1.9696804617451158e-07, + "loss": 0.0018, + "step": 58390 + }, + { + "grad_norm": 0.039654720574617386, + "learning_rate": 1.9453192825326093e-07, + "loss": 0.0033, + "step": 58400 + }, + { + "grad_norm": 0.15418066084384918, + "learning_rate": 1.9211093991817753e-07, + "loss": 0.0024, + "step": 58410 + }, + { + "grad_norm": 0.04638614505529404, + "learning_rate": 1.8970508190468973e-07, + "loss": 0.0016, + "step": 58420 + }, + { + "grad_norm": 0.05325758457183838, + "learning_rate": 1.8731435494362958e-07, + "loss": 0.0012, + "step": 58430 + }, + { + "grad_norm": 0.08779995888471603, + "learning_rate": 1.849387597612495e-07, + "loss": 0.0016, + "step": 58440 + }, + { + "grad_norm": 0.05808218568563461, + "learning_rate": 1.8257829707917228e-07, + "loss": 0.0019, + "step": 58450 + }, + { + "grad_norm": 0.025890076532959938, + "learning_rate": 1.8023296761446317e-07, + "loss": 0.0029, + "step": 58460 + }, + { + "grad_norm": 0.0482134073972702, + "learning_rate": 1.7790277207956341e-07, + "loss": 0.0019, + "step": 58470 + }, + { + "grad_norm": 0.0647921934723854, + "learning_rate": 1.7558771118232343e-07, + "loss": 0.0016, + "step": 58480 + }, + { + "grad_norm": 0.048067670315504074, + "learning_rate": 1.7328778562599734e-07, + "loss": 0.0021, + "step": 58490 + }, + { + "grad_norm": 0.0886903703212738, + "learning_rate": 1.7100299610924298e-07, + "loss": 0.0022, + "step": 58500 + }, + { + "grad_norm": 0.04870017617940903, + "learning_rate": 1.6873334332612733e-07, + "loss": 0.0015, + "step": 58510 + }, + { + "grad_norm": 0.04579079523682594, + "learning_rate": 1.6647882796609894e-07, + "loss": 0.0029, + "step": 58520 + }, + { + "grad_norm": 0.07362257689237595, + "learning_rate": 1.6423945071402102e-07, + "loss": 0.0016, + "step": 58530 + }, + { + "grad_norm": 0.064451202750206, + "learning_rate": 1.6201521225016614e-07, + "loss": 0.0031, + "step": 58540 + }, + { + "grad_norm": 0.086424820125103, + "learning_rate": 1.598061132501938e-07, + "loss": 0.0022, + "step": 58550 + }, + { + "grad_norm": 0.032967064529657364, + "learning_rate": 1.576121543851672e-07, + "loss": 0.0018, + "step": 58560 + }, + { + "grad_norm": 0.06411159038543701, + "learning_rate": 1.5543333632155876e-07, + "loss": 0.0017, + "step": 58570 + }, + { + "grad_norm": 0.055354807525873184, + "learning_rate": 1.5326965972123352e-07, + "loss": 0.0026, + "step": 58580 + }, + { + "grad_norm": 0.028955847024917603, + "learning_rate": 1.5112112524146016e-07, + "loss": 0.0019, + "step": 58590 + }, + { + "grad_norm": 0.04863233119249344, + "learning_rate": 1.4898773353489992e-07, + "loss": 0.0015, + "step": 58600 + }, + { + "grad_norm": 0.03290300816297531, + "learning_rate": 1.4686948524962884e-07, + "loss": 0.002, + "step": 58610 + }, + { + "grad_norm": 0.06219300627708435, + "learning_rate": 1.4476638102911556e-07, + "loss": 0.0021, + "step": 58620 + }, + { + "grad_norm": 0.06562041491270065, + "learning_rate": 1.4267842151222123e-07, + "loss": 0.0021, + "step": 58630 + }, + { + "grad_norm": 0.036848705261945724, + "learning_rate": 1.4060560733321626e-07, + "loss": 0.0022, + "step": 58640 + }, + { + "grad_norm": 0.06273098289966583, + "learning_rate": 1.385479391217692e-07, + "loss": 0.0015, + "step": 58650 + }, + { + "grad_norm": 0.0530925877392292, + "learning_rate": 1.365054175029412e-07, + "loss": 0.0027, + "step": 58660 + }, + { + "grad_norm": 0.048895213752985, + "learning_rate": 1.3447804309719702e-07, + "loss": 0.0013, + "step": 58670 + }, + { + "grad_norm": 0.028824251145124435, + "learning_rate": 1.3246581652040512e-07, + "loss": 0.0013, + "step": 58680 + }, + { + "grad_norm": 0.028173640370368958, + "learning_rate": 1.3046873838381546e-07, + "loss": 0.0018, + "step": 58690 + }, + { + "grad_norm": 0.031453412026166916, + "learning_rate": 1.2848680929409828e-07, + "loss": 0.002, + "step": 58700 + }, + { + "grad_norm": 0.06685216724872589, + "learning_rate": 1.2652002985331091e-07, + "loss": 0.0014, + "step": 58710 + }, + { + "grad_norm": 0.05645514279603958, + "learning_rate": 1.2456840065889764e-07, + "loss": 0.0016, + "step": 58720 + }, + { + "grad_norm": 0.053065910935401917, + "learning_rate": 1.226319223037231e-07, + "loss": 0.0026, + "step": 58730 + }, + { + "grad_norm": 0.054537683725357056, + "learning_rate": 1.2071059537603902e-07, + "loss": 0.0018, + "step": 58740 + }, + { + "grad_norm": 0.05256020650267601, + "learning_rate": 1.1880442045948403e-07, + "loss": 0.0019, + "step": 58750 + }, + { + "grad_norm": 0.1144891232252121, + "learning_rate": 1.1691339813311164e-07, + "loss": 0.0018, + "step": 58760 + }, + { + "grad_norm": 0.03753571957349777, + "learning_rate": 1.1503752897136233e-07, + "loss": 0.0023, + "step": 58770 + }, + { + "grad_norm": 0.07236386090517044, + "learning_rate": 1.1317681354407472e-07, + "loss": 0.0027, + "step": 58780 + }, + { + "grad_norm": 0.04716004803776741, + "learning_rate": 1.1133125241649111e-07, + "loss": 0.0015, + "step": 58790 + }, + { + "grad_norm": 0.1047186553478241, + "learning_rate": 1.0950084614922973e-07, + "loss": 0.0031, + "step": 58800 + }, + { + "grad_norm": 0.10258243978023529, + "learning_rate": 1.0768559529834021e-07, + "loss": 0.0029, + "step": 58810 + }, + { + "grad_norm": 0.0595087893307209, + "learning_rate": 1.0588550041522594e-07, + "loss": 0.0013, + "step": 58820 + }, + { + "grad_norm": 0.08410868048667908, + "learning_rate": 1.0410056204672169e-07, + "loss": 0.0033, + "step": 58830 + }, + { + "grad_norm": 0.03289599344134331, + "learning_rate": 1.0233078073504376e-07, + "loss": 0.0017, + "step": 58840 + }, + { + "grad_norm": 0.07193457335233688, + "learning_rate": 1.0057615701780654e-07, + "loss": 0.0019, + "step": 58850 + }, + { + "grad_norm": 0.10032462328672409, + "learning_rate": 9.883669142801144e-08, + "loss": 0.0023, + "step": 58860 + }, + { + "grad_norm": 0.05668910965323448, + "learning_rate": 9.711238449406356e-08, + "loss": 0.0027, + "step": 58870 + }, + { + "grad_norm": 0.08455364406108856, + "learning_rate": 9.540323673976614e-08, + "loss": 0.0026, + "step": 58880 + }, + { + "grad_norm": 0.08835799247026443, + "learning_rate": 9.370924868430942e-08, + "loss": 0.0026, + "step": 58890 + }, + { + "grad_norm": 0.02371213398873806, + "learning_rate": 9.203042084228175e-08, + "loss": 0.0018, + "step": 58900 + }, + { + "grad_norm": 0.02446998469531536, + "learning_rate": 9.036675372366965e-08, + "loss": 0.0016, + "step": 58910 + }, + { + "grad_norm": 0.03439794480800629, + "learning_rate": 8.871824783385218e-08, + "loss": 0.0012, + "step": 58920 + }, + { + "grad_norm": 0.06286773830652237, + "learning_rate": 8.7084903673601e-08, + "loss": 0.0023, + "step": 58930 + }, + { + "grad_norm": 0.037395279854536057, + "learning_rate": 8.546672173908032e-08, + "loss": 0.0014, + "step": 58940 + }, + { + "grad_norm": 0.05481887608766556, + "learning_rate": 8.386370252185249e-08, + "loss": 0.0044, + "step": 58950 + }, + { + "grad_norm": 0.05990765243768692, + "learning_rate": 8.227584650887243e-08, + "loss": 0.0024, + "step": 58960 + }, + { + "grad_norm": 0.10528717935085297, + "learning_rate": 8.070315418249319e-08, + "loss": 0.0017, + "step": 58970 + }, + { + "grad_norm": 0.03271739184856415, + "learning_rate": 7.914562602044929e-08, + "loss": 0.0024, + "step": 58980 + }, + { + "grad_norm": 0.039924196898937225, + "learning_rate": 7.76032624958789e-08, + "loss": 0.0036, + "step": 58990 + }, + { + "grad_norm": 0.07227745652198792, + "learning_rate": 7.607606407731282e-08, + "loss": 0.0021, + "step": 59000 + }, + { + "grad_norm": 0.033133938908576965, + "learning_rate": 7.45640312286744e-08, + "loss": 0.0023, + "step": 59010 + }, + { + "grad_norm": 0.04029903933405876, + "learning_rate": 7.306716440927952e-08, + "loss": 0.0027, + "step": 59020 + }, + { + "grad_norm": 0.08826343715190887, + "learning_rate": 7.15854640738367e-08, + "loss": 0.0024, + "step": 59030 + }, + { + "grad_norm": 0.14936865866184235, + "learning_rate": 7.011893067244701e-08, + "loss": 0.0025, + "step": 59040 + }, + { + "grad_norm": 0.04518784210085869, + "learning_rate": 6.866756465060408e-08, + "loss": 0.0018, + "step": 59050 + }, + { + "grad_norm": 0.0326828695833683, + "learning_rate": 6.723136644918859e-08, + "loss": 0.0017, + "step": 59060 + }, + { + "grad_norm": 0.03347652778029442, + "learning_rate": 6.581033650449042e-08, + "loss": 0.0013, + "step": 59070 + }, + { + "grad_norm": 0.047391701489686966, + "learning_rate": 6.440447524817539e-08, + "loss": 0.0013, + "step": 59080 + }, + { + "grad_norm": 0.036413270980119705, + "learning_rate": 6.301378310730743e-08, + "loss": 0.0018, + "step": 59090 + }, + { + "grad_norm": 0.08045385777950287, + "learning_rate": 6.163826050434307e-08, + "loss": 0.0021, + "step": 59100 + }, + { + "grad_norm": 0.061411552131175995, + "learning_rate": 6.027790785713139e-08, + "loss": 0.0019, + "step": 59110 + }, + { + "grad_norm": 0.08067141473293304, + "learning_rate": 5.89327255789085e-08, + "loss": 0.0016, + "step": 59120 + }, + { + "grad_norm": 0.032342471182346344, + "learning_rate": 5.7602714078303085e-08, + "loss": 0.0015, + "step": 59130 + }, + { + "grad_norm": 0.07008244842290878, + "learning_rate": 5.628787375934197e-08, + "loss": 0.0025, + "step": 59140 + }, + { + "grad_norm": 0.03568591922521591, + "learning_rate": 5.498820502143898e-08, + "loss": 0.0019, + "step": 59150 + }, + { + "grad_norm": 0.02357313223183155, + "learning_rate": 5.370370825939497e-08, + "loss": 0.0014, + "step": 59160 + }, + { + "grad_norm": 0.0540211983025074, + "learning_rate": 5.243438386340893e-08, + "loss": 0.0017, + "step": 59170 + }, + { + "grad_norm": 0.03671962395310402, + "learning_rate": 5.118023221907242e-08, + "loss": 0.0018, + "step": 59180 + }, + { + "grad_norm": 0.05022633075714111, + "learning_rate": 4.994125370735292e-08, + "loss": 0.0018, + "step": 59190 + }, + { + "grad_norm": 0.02455817349255085, + "learning_rate": 4.871744870462713e-08, + "loss": 0.0018, + "step": 59200 + }, + { + "grad_norm": 0.07708315551280975, + "learning_rate": 4.7508817582658794e-08, + "loss": 0.0017, + "step": 59210 + }, + { + "grad_norm": 0.033112216740846634, + "learning_rate": 4.631536070858755e-08, + "loss": 0.0023, + "step": 59220 + }, + { + "grad_norm": 0.034977734088897705, + "learning_rate": 4.513707844495674e-08, + "loss": 0.0022, + "step": 59230 + }, + { + "grad_norm": 0.06507162749767303, + "learning_rate": 4.3973971149702255e-08, + "loss": 0.0016, + "step": 59240 + }, + { + "grad_norm": 0.054538268595933914, + "learning_rate": 4.2826039176147025e-08, + "loss": 0.0016, + "step": 59250 + }, + { + "grad_norm": 0.04904944822192192, + "learning_rate": 4.169328287299545e-08, + "loss": 0.002, + "step": 59260 + }, + { + "grad_norm": 0.07659133523702621, + "learning_rate": 4.057570258435006e-08, + "loss": 0.0025, + "step": 59270 + }, + { + "grad_norm": 0.02684958651661873, + "learning_rate": 3.947329864970595e-08, + "loss": 0.0016, + "step": 59280 + }, + { + "grad_norm": 0.06208379194140434, + "learning_rate": 3.8386071403939686e-08, + "loss": 0.0017, + "step": 59290 + }, + { + "grad_norm": 0.03156406059861183, + "learning_rate": 3.731402117733152e-08, + "loss": 0.0015, + "step": 59300 + }, + { + "grad_norm": 0.07186618447303772, + "learning_rate": 3.625714829552651e-08, + "loss": 0.0021, + "step": 59310 + }, + { + "grad_norm": 0.06333423405885696, + "learning_rate": 3.5215453079590065e-08, + "loss": 0.0014, + "step": 59320 + }, + { + "grad_norm": 0.08392772823572159, + "learning_rate": 3.4188935845952396e-08, + "loss": 0.002, + "step": 59330 + }, + { + "grad_norm": 0.03192640095949173, + "learning_rate": 3.3177596906447396e-08, + "loss": 0.0017, + "step": 59340 + }, + { + "grad_norm": 0.07749873399734497, + "learning_rate": 3.218143656829043e-08, + "loss": 0.002, + "step": 59350 + }, + { + "grad_norm": 0.026460831984877586, + "learning_rate": 3.120045513408387e-08, + "loss": 0.0011, + "step": 59360 + }, + { + "grad_norm": 0.032364800572395325, + "learning_rate": 3.02346529018338e-08, + "loss": 0.0037, + "step": 59370 + }, + { + "grad_norm": 0.04694696143269539, + "learning_rate": 2.9284030164922204e-08, + "loss": 0.0019, + "step": 59380 + }, + { + "grad_norm": 0.028310921043157578, + "learning_rate": 2.8348587212123634e-08, + "loss": 0.0028, + "step": 59390 + }, + { + "grad_norm": 0.12441930174827576, + "learning_rate": 2.7428324327594125e-08, + "loss": 0.0022, + "step": 59400 + }, + { + "grad_norm": 0.03990059718489647, + "learning_rate": 2.6523241790893383e-08, + "loss": 0.0021, + "step": 59410 + }, + { + "grad_norm": 0.042742300778627396, + "learning_rate": 2.563333987695704e-08, + "loss": 0.0021, + "step": 59420 + }, + { + "grad_norm": 0.09080519527196884, + "learning_rate": 2.4758618856118852e-08, + "loss": 0.0031, + "step": 59430 + }, + { + "grad_norm": 0.02846841886639595, + "learning_rate": 2.3899078994088497e-08, + "loss": 0.0025, + "step": 59440 + }, + { + "grad_norm": 0.048731427639722824, + "learning_rate": 2.3054720551973775e-08, + "loss": 0.0015, + "step": 59450 + }, + { + "grad_norm": 0.03363591060042381, + "learning_rate": 2.222554378627506e-08, + "loss": 0.0012, + "step": 59460 + }, + { + "grad_norm": 0.020043428987264633, + "learning_rate": 2.1411548948868653e-08, + "loss": 0.0012, + "step": 59470 + }, + { + "grad_norm": 0.21088097989559174, + "learning_rate": 2.0612736287023426e-08, + "loss": 0.0026, + "step": 59480 + }, + { + "grad_norm": 0.0361279733479023, + "learning_rate": 1.9829106043400826e-08, + "loss": 0.0017, + "step": 59490 + }, + { + "grad_norm": 0.02286909893155098, + "learning_rate": 1.9060658456043768e-08, + "loss": 0.0015, + "step": 59500 + }, + { + "grad_norm": 0.031134311109781265, + "learning_rate": 1.830739375838775e-08, + "loss": 0.0013, + "step": 59510 + }, + { + "grad_norm": 0.09717237204313278, + "learning_rate": 1.7569312179260832e-08, + "loss": 0.0013, + "step": 59520 + }, + { + "grad_norm": 0.08429041504859924, + "learning_rate": 1.684641394286146e-08, + "loss": 0.0022, + "step": 59530 + }, + { + "grad_norm": 0.045938536524772644, + "learning_rate": 1.6138699268797296e-08, + "loss": 0.0013, + "step": 59540 + }, + { + "grad_norm": 0.042852360755205154, + "learning_rate": 1.5446168372046376e-08, + "loss": 0.0019, + "step": 59550 + }, + { + "grad_norm": 0.07190746068954468, + "learning_rate": 1.4768821462984861e-08, + "loss": 0.0017, + "step": 59560 + }, + { + "grad_norm": 0.027321411296725273, + "learning_rate": 1.4106658747370383e-08, + "loss": 0.0043, + "step": 59570 + }, + { + "grad_norm": 0.05060608312487602, + "learning_rate": 1.3459680426353149e-08, + "loss": 0.0032, + "step": 59580 + }, + { + "grad_norm": 0.04011053591966629, + "learning_rate": 1.2827886696464841e-08, + "loss": 0.0016, + "step": 59590 + }, + { + "grad_norm": 0.06210273504257202, + "learning_rate": 1.2211277749635264e-08, + "loss": 0.0024, + "step": 59600 + }, + { + "grad_norm": 0.04004385694861412, + "learning_rate": 1.1609853773164592e-08, + "loss": 0.0016, + "step": 59610 + }, + { + "grad_norm": 0.0310466680675745, + "learning_rate": 1.1023614949751127e-08, + "loss": 0.0024, + "step": 59620 + }, + { + "grad_norm": 0.03447660803794861, + "learning_rate": 1.0452561457485744e-08, + "loss": 0.0034, + "step": 59630 + }, + { + "grad_norm": 0.06696496903896332, + "learning_rate": 9.896693469829688e-09, + "loss": 0.0014, + "step": 59640 + }, + { + "grad_norm": 0.028710516169667244, + "learning_rate": 9.35601115564788e-09, + "loss": 0.0014, + "step": 59650 + }, + { + "grad_norm": 0.05129091069102287, + "learning_rate": 8.830514679186719e-09, + "loss": 0.0017, + "step": 59660 + }, + { + "grad_norm": 0.10318212956190109, + "learning_rate": 8.320204200074066e-09, + "loss": 0.0019, + "step": 59670 + }, + { + "grad_norm": 0.061326656490564346, + "learning_rate": 7.825079873324814e-09, + "loss": 0.0011, + "step": 59680 + }, + { + "grad_norm": 0.05653013661503792, + "learning_rate": 7.345141849351977e-09, + "loss": 0.0015, + "step": 59690 + }, + { + "grad_norm": 0.05302873253822327, + "learning_rate": 6.880390273944493e-09, + "loss": 0.0014, + "step": 59700 + }, + { + "grad_norm": 0.0874003916978836, + "learning_rate": 6.4308252882838704e-09, + "loss": 0.0014, + "step": 59710 + }, + { + "grad_norm": 0.08187376707792282, + "learning_rate": 5.9964470289386455e-09, + "loss": 0.0013, + "step": 59720 + }, + { + "grad_norm": 0.07469252496957779, + "learning_rate": 5.577255627853273e-09, + "loss": 0.0016, + "step": 59730 + }, + { + "grad_norm": 0.10403530299663544, + "learning_rate": 5.173251212370334e-09, + "loss": 0.0024, + "step": 59740 + }, + { + "grad_norm": 0.026600105687975883, + "learning_rate": 4.784433905219432e-09, + "loss": 0.0025, + "step": 59750 + }, + { + "grad_norm": 0.06172705069184303, + "learning_rate": 4.4108038245060934e-09, + "loss": 0.002, + "step": 59760 + }, + { + "grad_norm": 0.02699715830385685, + "learning_rate": 4.05236108373952e-09, + "loss": 0.0019, + "step": 59770 + }, + { + "grad_norm": 0.051058750599622726, + "learning_rate": 3.7091057917937324e-09, + "loss": 0.0021, + "step": 59780 + }, + { + "grad_norm": 0.05513815954327583, + "learning_rate": 3.381038052946428e-09, + "loss": 0.0019, + "step": 59790 + }, + { + "grad_norm": 0.06892674416303635, + "learning_rate": 3.0681579668623283e-09, + "loss": 0.0018, + "step": 59800 + }, + { + "grad_norm": 0.11579745262861252, + "learning_rate": 2.7704656285709727e-09, + "loss": 0.0019, + "step": 59810 + }, + { + "grad_norm": 0.031310658901929855, + "learning_rate": 2.4879611285166803e-09, + "loss": 0.0013, + "step": 59820 + }, + { + "grad_norm": 0.0371757410466671, + "learning_rate": 2.2206445525085883e-09, + "loss": 0.0014, + "step": 59830 + }, + { + "grad_norm": 0.04418807476758957, + "learning_rate": 1.9685159817595112e-09, + "loss": 0.0017, + "step": 59840 + }, + { + "grad_norm": 0.05054246261715889, + "learning_rate": 1.7315754928470817e-09, + "loss": 0.0019, + "step": 59850 + }, + { + "grad_norm": 0.06414949148893356, + "learning_rate": 1.5098231577581611e-09, + "loss": 0.0023, + "step": 59860 + }, + { + "grad_norm": 0.023615892976522446, + "learning_rate": 1.3032590438499804e-09, + "loss": 0.0019, + "step": 59870 + }, + { + "grad_norm": 0.09313354641199112, + "learning_rate": 1.1118832138723444e-09, + "loss": 0.0029, + "step": 59880 + }, + { + "grad_norm": 0.08225814998149872, + "learning_rate": 9.356957259620825e-10, + "loss": 0.0032, + "step": 59890 + }, + { + "grad_norm": 0.03423801064491272, + "learning_rate": 7.74696633637495e-10, + "loss": 0.0023, + "step": 59900 + }, + { + "grad_norm": 0.025128740817308426, + "learning_rate": 6.288859858039064e-10, + "loss": 0.0017, + "step": 59910 + }, + { + "grad_norm": 0.03376555070281029, + "learning_rate": 4.982638267647665e-10, + "loss": 0.0026, + "step": 59920 + }, + { + "grad_norm": 0.040921635925769806, + "learning_rate": 3.8283019618834406e-10, + "loss": 0.0016, + "step": 59930 + }, + { + "grad_norm": 0.02604549750685692, + "learning_rate": 2.825851291410331e-10, + "loss": 0.0019, + "step": 59940 + }, + { + "grad_norm": 0.029058320447802544, + "learning_rate": 1.975286560873535e-10, + "loss": 0.0012, + "step": 59950 + }, + { + "grad_norm": 0.021468672901391983, + "learning_rate": 1.2766080285109284e-10, + "loss": 0.0015, + "step": 59960 + }, + { + "grad_norm": 0.061713118106126785, + "learning_rate": 7.298159065971533e-11, + "loss": 0.0015, + "step": 59970 + }, + { + "grad_norm": 0.057619862258434296, + "learning_rate": 3.349103612770854e-11, + "loss": 0.0036, + "step": 59980 + }, + { + "grad_norm": 0.0311071015894413, + "learning_rate": 9.189151245481142e-12, + "loss": 0.0017, + "step": 59990 + }, + { + "grad_norm": 0.06286779046058655, + "learning_rate": 7.594340156735769e-14, + "loss": 0.0028, + "step": 60000 } ], "logging_steps": 10, - "max_steps": 20000, + "max_steps": 60000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 1000,