{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 3.0322396755218506, "learning_rate": 1.8e-06, "loss": 1.1669, "step": 10 }, { "grad_norm": 2.2011330127716064, "learning_rate": 3.8e-06, "loss": 1.1512, "step": 20 }, { "grad_norm": 0.9810276627540588, "learning_rate": 5.8e-06, "loss": 1.0873, "step": 30 }, { "grad_norm": 0.7856635451316833, "learning_rate": 7.8e-06, "loss": 1.0517, "step": 40 }, { "grad_norm": 1.1428941488265991, "learning_rate": 9.800000000000001e-06, "loss": 1.0432, "step": 50 }, { "grad_norm": 0.6241004467010498, "learning_rate": 1.18e-05, "loss": 1.0278, "step": 60 }, { "grad_norm": 0.6585308313369751, "learning_rate": 1.3800000000000002e-05, "loss": 1.0476, "step": 70 }, { "grad_norm": 0.6832582354545593, "learning_rate": 1.58e-05, "loss": 1.0192, "step": 80 }, { "grad_norm": 0.5609439015388489, "learning_rate": 1.78e-05, "loss": 1.0228, "step": 90 }, { "grad_norm": 0.5982586145401001, "learning_rate": 1.9800000000000004e-05, "loss": 1.0239, "step": 100 }, { "grad_norm": 0.5948628187179565, "learning_rate": 2.18e-05, "loss": 1.0278, "step": 110 }, { "grad_norm": 1.6620466709136963, "learning_rate": 2.38e-05, "loss": 1.0299, "step": 120 }, { "grad_norm": 2.551886558532715, "learning_rate": 2.58e-05, "loss": 0.9877, "step": 130 }, { "grad_norm": 1.5303394794464111, "learning_rate": 2.7800000000000005e-05, "loss": 0.9712, "step": 140 }, { "grad_norm": 1.7442747354507446, "learning_rate": 2.98e-05, "loss": 0.933, "step": 150 }, { "grad_norm": 1.0566720962524414, "learning_rate": 3.18e-05, "loss": 0.9119, "step": 160 }, { "grad_norm": 0.9604274034500122, "learning_rate": 3.38e-05, "loss": 0.9157, "step": 170 }, { "grad_norm": 1.1089171171188354, "learning_rate": 3.58e-05, "loss": 0.894, "step": 180 }, { "grad_norm": 0.7986904978752136, "learning_rate": 3.7800000000000004e-05, "loss": 0.8734, "step": 190 }, { "grad_norm": 1.449533224105835, "learning_rate": 3.9800000000000005e-05, "loss": 0.8631, "step": 200 }, { "grad_norm": 2.0386946201324463, "learning_rate": 4.18e-05, "loss": 0.8436, "step": 210 }, { "grad_norm": 2.5494658946990967, "learning_rate": 4.38e-05, "loss": 0.7814, "step": 220 }, { "grad_norm": 1.8697882890701294, "learning_rate": 4.58e-05, "loss": 0.7094, "step": 230 }, { "grad_norm": 2.805925130844116, "learning_rate": 4.78e-05, "loss": 0.613, "step": 240 }, { "grad_norm": 1.5028965473175049, "learning_rate": 4.9800000000000004e-05, "loss": 0.5028, "step": 250 }, { "grad_norm": 2.5280911922454834, "learning_rate": 5.1800000000000005e-05, "loss": 0.3741, "step": 260 }, { "grad_norm": 1.6380298137664795, "learning_rate": 5.380000000000001e-05, "loss": 0.2821, "step": 270 }, { "grad_norm": 2.2940752506256104, "learning_rate": 5.580000000000001e-05, "loss": 0.2392, "step": 280 }, { "grad_norm": 2.5730438232421875, "learning_rate": 5.7799999999999995e-05, "loss": 0.1933, "step": 290 }, { "grad_norm": 1.6670491695404053, "learning_rate": 5.9800000000000003e-05, "loss": 0.1773, "step": 300 }, { "grad_norm": 2.1963818073272705, "learning_rate": 6.18e-05, "loss": 0.1546, "step": 310 }, { "grad_norm": 1.7032819986343384, "learning_rate": 6.38e-05, "loss": 0.1487, "step": 320 }, { "grad_norm": 1.1375164985656738, "learning_rate": 6.58e-05, "loss": 0.1362, "step": 330 }, { "grad_norm": 1.1723397970199585, "learning_rate": 6.780000000000001e-05, "loss": 0.1257, "step": 340 }, { "grad_norm": 1.0788356065750122, "learning_rate": 6.98e-05, "loss": 0.1297, "step": 350 }, { "grad_norm": 1.0764281749725342, "learning_rate": 7.18e-05, "loss": 0.1123, "step": 360 }, { "grad_norm": 1.1485093832015991, "learning_rate": 7.38e-05, "loss": 0.1257, "step": 370 }, { "grad_norm": 1.501458764076233, "learning_rate": 7.58e-05, "loss": 0.1019, "step": 380 }, { "grad_norm": 1.155227541923523, "learning_rate": 7.780000000000001e-05, "loss": 0.1145, "step": 390 }, { "grad_norm": 1.2736395597457886, "learning_rate": 7.98e-05, "loss": 0.1081, "step": 400 }, { "grad_norm": 1.0844621658325195, "learning_rate": 8.18e-05, "loss": 0.1018, "step": 410 }, { "grad_norm": 1.2185215950012207, "learning_rate": 8.38e-05, "loss": 0.0963, "step": 420 }, { "grad_norm": 6.255167007446289, "learning_rate": 8.58e-05, "loss": 0.1077, "step": 430 }, { "grad_norm": 1.9335203170776367, "learning_rate": 8.78e-05, "loss": 0.1231, "step": 440 }, { "grad_norm": 0.8321042656898499, "learning_rate": 8.98e-05, "loss": 0.1138, "step": 450 }, { "grad_norm": 0.8995205163955688, "learning_rate": 9.180000000000001e-05, "loss": 0.0854, "step": 460 }, { "grad_norm": 0.8628952503204346, "learning_rate": 9.38e-05, "loss": 0.0984, "step": 470 }, { "grad_norm": 1.0661661624908447, "learning_rate": 9.58e-05, "loss": 0.1078, "step": 480 }, { "grad_norm": 0.9602690935134888, "learning_rate": 9.78e-05, "loss": 0.0943, "step": 490 }, { "grad_norm": 1.296250581741333, "learning_rate": 9.98e-05, "loss": 0.0958, "step": 500 }, { "grad_norm": 1.1472097635269165, "learning_rate": 9.9999778549206e-05, "loss": 0.1081, "step": 510 }, { "grad_norm": 1.076284408569336, "learning_rate": 9.999901304280685e-05, "loss": 0.1067, "step": 520 }, { "grad_norm": 0.6838881969451904, "learning_rate": 9.999770075521164e-05, "loss": 0.0945, "step": 530 }, { "grad_norm": 1.5203118324279785, "learning_rate": 9.99958417007713e-05, "loss": 0.0975, "step": 540 }, { "grad_norm": 1.0812902450561523, "learning_rate": 9.999343589981615e-05, "loss": 0.1121, "step": 550 }, { "grad_norm": 0.7028939127922058, "learning_rate": 9.999048337865568e-05, "loss": 0.0995, "step": 560 }, { "grad_norm": 0.6051813960075378, "learning_rate": 9.998698416957815e-05, "loss": 0.0952, "step": 570 }, { "grad_norm": 0.8162695169448853, "learning_rate": 9.998293831085037e-05, "loss": 0.0803, "step": 580 }, { "grad_norm": 0.5644664764404297, "learning_rate": 9.997834584671719e-05, "loss": 0.0843, "step": 590 }, { "grad_norm": 0.6265770792961121, "learning_rate": 9.997320682740107e-05, "loss": 0.0866, "step": 600 }, { "grad_norm": 0.6384676694869995, "learning_rate": 9.996752130910149e-05, "loss": 0.0894, "step": 610 }, { "grad_norm": 0.8044371604919434, "learning_rate": 9.99612893539944e-05, "loss": 0.088, "step": 620 }, { "grad_norm": 0.7883325219154358, "learning_rate": 9.995451103023144e-05, "loss": 0.0807, "step": 630 }, { "grad_norm": 0.8214014172554016, "learning_rate": 9.994718641193928e-05, "loss": 0.0794, "step": 640 }, { "grad_norm": 0.7955110669136047, "learning_rate": 9.993931557921874e-05, "loss": 0.0653, "step": 650 }, { "grad_norm": 1.0301965475082397, "learning_rate": 9.993089861814402e-05, "loss": 0.0872, "step": 660 }, { "grad_norm": 0.6119454503059387, "learning_rate": 9.992193562076166e-05, "loss": 0.0814, "step": 670 }, { "grad_norm": 0.7859804034233093, "learning_rate": 9.991242668508954e-05, "loss": 0.0773, "step": 680 }, { "grad_norm": 0.9492737650871277, "learning_rate": 9.990237191511587e-05, "loss": 0.0761, "step": 690 }, { "grad_norm": 0.7229829430580139, "learning_rate": 9.989177142079802e-05, "loss": 0.0771, "step": 700 }, { "grad_norm": 0.6185887455940247, "learning_rate": 9.988062531806126e-05, "loss": 0.0696, "step": 710 }, { "grad_norm": 0.9210008382797241, "learning_rate": 9.986893372879762e-05, "loss": 0.086, "step": 720 }, { "grad_norm": 0.7393677234649658, "learning_rate": 9.985669678086443e-05, "loss": 0.0736, "step": 730 }, { "grad_norm": 1.0680969953536987, "learning_rate": 9.984391460808298e-05, "loss": 0.0768, "step": 740 }, { "grad_norm": 0.7507827281951904, "learning_rate": 9.983058735023709e-05, "loss": 0.0813, "step": 750 }, { "grad_norm": 1.2693743705749512, "learning_rate": 9.98167151530715e-05, "loss": 0.0719, "step": 760 }, { "grad_norm": 1.532736897468567, "learning_rate": 9.980229816829034e-05, "loss": 0.0668, "step": 770 }, { "grad_norm": 0.7922891974449158, "learning_rate": 9.978733655355544e-05, "loss": 0.0833, "step": 780 }, { "grad_norm": 0.5808783769607544, "learning_rate": 9.977183047248464e-05, "loss": 0.0695, "step": 790 }, { "grad_norm": 0.5857202410697937, "learning_rate": 9.975578009464992e-05, "loss": 0.0583, "step": 800 }, { "grad_norm": 0.7801036834716797, "learning_rate": 9.97391855955757e-05, "loss": 0.0633, "step": 810 }, { "grad_norm": 0.5595137476921082, "learning_rate": 9.972204715673669e-05, "loss": 0.0771, "step": 820 }, { "grad_norm": 0.6050302982330322, "learning_rate": 9.970436496555617e-05, "loss": 0.0776, "step": 830 }, { "grad_norm": 0.6795267462730408, "learning_rate": 9.968613921540373e-05, "loss": 0.075, "step": 840 }, { "grad_norm": 0.9821692705154419, "learning_rate": 9.966737010559326e-05, "loss": 0.0558, "step": 850 }, { "grad_norm": 0.6710021495819092, "learning_rate": 9.964805784138072e-05, "loss": 0.0722, "step": 860 }, { "grad_norm": 0.5773111581802368, "learning_rate": 9.962820263396195e-05, "loss": 0.066, "step": 870 }, { "grad_norm": 0.4262138903141022, "learning_rate": 9.960780470047033e-05, "loss": 0.0516, "step": 880 }, { "grad_norm": 0.4067871570587158, "learning_rate": 9.958686426397437e-05, "loss": 0.0652, "step": 890 }, { "grad_norm": 0.7921264171600342, "learning_rate": 9.956538155347534e-05, "loss": 0.0794, "step": 900 }, { "grad_norm": 0.6045096516609192, "learning_rate": 9.95433568039047e-05, "loss": 0.0706, "step": 910 }, { "grad_norm": 0.6204010844230652, "learning_rate": 9.952079025612162e-05, "loss": 0.0565, "step": 920 }, { "grad_norm": 46.013484954833984, "learning_rate": 9.949768215691022e-05, "loss": 0.0719, "step": 930 }, { "grad_norm": 0.8080171346664429, "learning_rate": 9.9474032758977e-05, "loss": 0.0724, "step": 940 }, { "grad_norm": 0.6104950308799744, "learning_rate": 9.944984232094794e-05, "loss": 0.0618, "step": 950 }, { "grad_norm": 0.6578022241592407, "learning_rate": 9.942511110736584e-05, "loss": 0.0695, "step": 960 }, { "grad_norm": 0.5351039171218872, "learning_rate": 9.939983938868726e-05, "loss": 0.0625, "step": 970 }, { "grad_norm": 0.5968323349952698, "learning_rate": 9.93740274412797e-05, "loss": 0.069, "step": 980 }, { "grad_norm": 0.4959365427494049, "learning_rate": 9.934767554741846e-05, "loss": 0.0597, "step": 990 }, { "grad_norm": 0.4687926769256592, "learning_rate": 9.932078399528361e-05, "loss": 0.0574, "step": 1000 }, { "grad_norm": 0.7520908713340759, "learning_rate": 9.929335307895689e-05, "loss": 0.0577, "step": 1010 }, { "grad_norm": 0.4894666373729706, "learning_rate": 9.926538309841839e-05, "loss": 0.0588, "step": 1020 }, { "grad_norm": 0.5658991932868958, "learning_rate": 9.923687435954334e-05, "loss": 0.0524, "step": 1030 }, { "grad_norm": 0.547452986240387, "learning_rate": 9.920782717409873e-05, "loss": 0.0612, "step": 1040 }, { "grad_norm": 1.6426231861114502, "learning_rate": 9.917824185973994e-05, "loss": 0.0612, "step": 1050 }, { "grad_norm": 0.6883339881896973, "learning_rate": 9.914811874000723e-05, "loss": 0.0642, "step": 1060 }, { "grad_norm": 0.5833764672279358, "learning_rate": 9.911745814432218e-05, "loss": 0.0639, "step": 1070 }, { "grad_norm": 0.8547801971435547, "learning_rate": 9.90862604079842e-05, "loss": 0.0616, "step": 1080 }, { "grad_norm": 0.5691959261894226, "learning_rate": 9.90545258721667e-05, "loss": 0.0656, "step": 1090 }, { "grad_norm": 0.5417507290840149, "learning_rate": 9.90222548839135e-05, "loss": 0.0511, "step": 1100 }, { "grad_norm": 0.4010971784591675, "learning_rate": 9.898944779613495e-05, "loss": 0.0598, "step": 1110 }, { "grad_norm": 0.5401006937026978, "learning_rate": 9.89561049676041e-05, "loss": 0.0656, "step": 1120 }, { "grad_norm": 0.33658120036125183, "learning_rate": 9.89222267629528e-05, "loss": 0.0584, "step": 1130 }, { "grad_norm": 0.6328154802322388, "learning_rate": 9.888781355266763e-05, "loss": 0.0542, "step": 1140 }, { "grad_norm": 0.6131583452224731, "learning_rate": 9.885286571308598e-05, "loss": 0.0596, "step": 1150 }, { "grad_norm": 0.8274643421173096, "learning_rate": 9.881738362639182e-05, "loss": 0.0502, "step": 1160 }, { "grad_norm": 0.6295856833457947, "learning_rate": 9.878136768061154e-05, "loss": 0.0619, "step": 1170 }, { "grad_norm": 0.47114869952201843, "learning_rate": 9.874481826960979e-05, "loss": 0.0562, "step": 1180 }, { "grad_norm": 0.8196890354156494, "learning_rate": 9.870773579308503e-05, "loss": 0.0624, "step": 1190 }, { "grad_norm": 0.4601079821586609, "learning_rate": 9.867012065656533e-05, "loss": 0.0554, "step": 1200 }, { "grad_norm": 0.4027715027332306, "learning_rate": 9.863197327140376e-05, "loss": 0.052, "step": 1210 }, { "grad_norm": 0.38178929686546326, "learning_rate": 9.859329405477403e-05, "loss": 0.0503, "step": 1220 }, { "grad_norm": 0.6494172811508179, "learning_rate": 9.855408342966585e-05, "loss": 0.05, "step": 1230 }, { "grad_norm": 0.5392621755599976, "learning_rate": 9.851434182488033e-05, "loss": 0.0634, "step": 1240 }, { "grad_norm": 0.5887909531593323, "learning_rate": 9.84740696750253e-05, "loss": 0.0582, "step": 1250 }, { "grad_norm": 0.6925702691078186, "learning_rate": 9.843326742051055e-05, "loss": 0.0675, "step": 1260 }, { "grad_norm": 0.6437531113624573, "learning_rate": 9.839193550754297e-05, "loss": 0.0651, "step": 1270 }, { "grad_norm": 0.3562054932117462, "learning_rate": 9.835007438812177e-05, "loss": 0.0456, "step": 1280 }, { "grad_norm": 0.6808228492736816, "learning_rate": 9.830768452003341e-05, "loss": 0.0664, "step": 1290 }, { "grad_norm": 0.666236400604248, "learning_rate": 9.826476636684671e-05, "loss": 0.0532, "step": 1300 }, { "grad_norm": 0.4259832203388214, "learning_rate": 9.822132039790773e-05, "loss": 0.0594, "step": 1310 }, { "grad_norm": 0.6460294127464294, "learning_rate": 9.817734708833461e-05, "loss": 0.0564, "step": 1320 }, { "grad_norm": 0.6385717988014221, "learning_rate": 9.813284691901243e-05, "loss": 0.0538, "step": 1330 }, { "grad_norm": 0.4378400146961212, "learning_rate": 9.808782037658792e-05, "loss": 0.0514, "step": 1340 }, { "grad_norm": 0.5851609706878662, "learning_rate": 9.804226795346411e-05, "loss": 0.0664, "step": 1350 }, { "grad_norm": 0.4828128218650818, "learning_rate": 9.799619014779503e-05, "loss": 0.0461, "step": 1360 }, { "grad_norm": 0.6621993184089661, "learning_rate": 9.794958746348013e-05, "loss": 0.0657, "step": 1370 }, { "grad_norm": 1.0745426416397095, "learning_rate": 9.790246041015896e-05, "loss": 0.0551, "step": 1380 }, { "grad_norm": 0.46200865507125854, "learning_rate": 9.785480950320538e-05, "loss": 0.056, "step": 1390 }, { "grad_norm": 0.5264564156532288, "learning_rate": 9.78066352637221e-05, "loss": 0.0478, "step": 1400 }, { "grad_norm": 0.7886044979095459, "learning_rate": 9.775793821853488e-05, "loss": 0.0569, "step": 1410 }, { "grad_norm": 0.5127342343330383, "learning_rate": 9.77087189001868e-05, "loss": 0.0517, "step": 1420 }, { "grad_norm": 0.545864999294281, "learning_rate": 9.765897784693243e-05, "loss": 0.061, "step": 1430 }, { "grad_norm": 0.35494253039360046, "learning_rate": 9.760871560273197e-05, "loss": 0.0478, "step": 1440 }, { "grad_norm": 0.523431122303009, "learning_rate": 9.755793271724526e-05, "loss": 0.0562, "step": 1450 }, { "grad_norm": 0.49055546522140503, "learning_rate": 9.750662974582584e-05, "loss": 0.05, "step": 1460 }, { "grad_norm": 0.49891480803489685, "learning_rate": 9.745480724951473e-05, "loss": 0.0522, "step": 1470 }, { "grad_norm": 0.48990434408187866, "learning_rate": 9.740246579503447e-05, "loss": 0.0529, "step": 1480 }, { "grad_norm": 0.46830251812934875, "learning_rate": 9.734960595478284e-05, "loss": 0.0433, "step": 1490 }, { "grad_norm": 1.3716031312942505, "learning_rate": 9.729622830682657e-05, "loss": 0.0442, "step": 1500 }, { "grad_norm": 0.39105504751205444, "learning_rate": 9.724233343489504e-05, "loss": 0.0587, "step": 1510 }, { "grad_norm": 2.06432843208313, "learning_rate": 9.718792192837396e-05, "loss": 0.0659, "step": 1520 }, { "grad_norm": 0.6027862429618835, "learning_rate": 9.713299438229886e-05, "loss": 0.0736, "step": 1530 }, { "grad_norm": 1.9449260234832764, "learning_rate": 9.707755139734855e-05, "loss": 0.0602, "step": 1540 }, { "grad_norm": 0.41933107376098633, "learning_rate": 9.702159357983866e-05, "loss": 0.0671, "step": 1550 }, { "grad_norm": 0.54278564453125, "learning_rate": 9.696512154171492e-05, "loss": 0.0563, "step": 1560 }, { "grad_norm": 0.4629799425601959, "learning_rate": 9.690813590054645e-05, "loss": 0.048, "step": 1570 }, { "grad_norm": 0.7710604667663574, "learning_rate": 9.685063727951914e-05, "loss": 0.0576, "step": 1580 }, { "grad_norm": 0.3388233780860901, "learning_rate": 9.679262630742865e-05, "loss": 0.0444, "step": 1590 }, { "grad_norm": 0.43013179302215576, "learning_rate": 9.673410361867373e-05, "loss": 0.0574, "step": 1600 }, { "grad_norm": 0.4889190196990967, "learning_rate": 9.667506985324909e-05, "loss": 0.0481, "step": 1610 }, { "grad_norm": 0.5401610136032104, "learning_rate": 9.661552565673855e-05, "loss": 0.0515, "step": 1620 }, { "grad_norm": 1.047282338142395, "learning_rate": 9.655547168030789e-05, "loss": 0.0484, "step": 1630 }, { "grad_norm": 0.6536716222763062, "learning_rate": 9.649490858069777e-05, "loss": 0.0467, "step": 1640 }, { "grad_norm": 0.6523647308349609, "learning_rate": 9.643383702021658e-05, "loss": 0.0481, "step": 1650 }, { "grad_norm": 0.6695061326026917, "learning_rate": 9.637225766673307e-05, "loss": 0.0517, "step": 1660 }, { "grad_norm": 0.45191293954849243, "learning_rate": 9.631017119366922e-05, "loss": 0.0485, "step": 1670 }, { "grad_norm": 0.9498205780982971, "learning_rate": 9.624757827999273e-05, "loss": 0.0559, "step": 1680 }, { "grad_norm": 0.917064905166626, "learning_rate": 9.618447961020971e-05, "loss": 0.0544, "step": 1690 }, { "grad_norm": 0.7450239658355713, "learning_rate": 9.612087587435707e-05, "loss": 0.0683, "step": 1700 }, { "grad_norm": 1.322640299797058, "learning_rate": 9.605676776799508e-05, "loss": 0.0455, "step": 1710 }, { "grad_norm": 0.6333397030830383, "learning_rate": 9.599215599219973e-05, "loss": 0.0566, "step": 1720 }, { "grad_norm": 0.4215209186077118, "learning_rate": 9.592704125355505e-05, "loss": 0.0483, "step": 1730 }, { "grad_norm": 0.4425746500492096, "learning_rate": 9.586142426414538e-05, "loss": 0.0402, "step": 1740 }, { "grad_norm": 0.6782955527305603, "learning_rate": 9.57953057415476e-05, "loss": 0.0467, "step": 1750 }, { "grad_norm": 0.5909518003463745, "learning_rate": 9.572868640882328e-05, "loss": 0.0523, "step": 1760 }, { "grad_norm": 0.4209151566028595, "learning_rate": 9.56615669945108e-05, "loss": 0.0559, "step": 1770 }, { "grad_norm": 0.3657643496990204, "learning_rate": 9.55939482326173e-05, "loss": 0.0454, "step": 1780 }, { "grad_norm": 0.3392142057418823, "learning_rate": 9.552583086261069e-05, "loss": 0.0418, "step": 1790 }, { "grad_norm": 0.44981077313423157, "learning_rate": 9.545721562941168e-05, "loss": 0.0465, "step": 1800 }, { "grad_norm": 0.33901557326316833, "learning_rate": 9.538810328338543e-05, "loss": 0.0578, "step": 1810 }, { "grad_norm": 0.441211074590683, "learning_rate": 9.531849458033349e-05, "loss": 0.0489, "step": 1820 }, { "grad_norm": 0.5280428528785706, "learning_rate": 9.524839028148547e-05, "loss": 0.0448, "step": 1830 }, { "grad_norm": 0.3850110173225403, "learning_rate": 9.517779115349077e-05, "loss": 0.0473, "step": 1840 }, { "grad_norm": 0.4975970983505249, "learning_rate": 9.510669796841014e-05, "loss": 0.0378, "step": 1850 }, { "grad_norm": 0.3681924343109131, "learning_rate": 9.503511150370727e-05, "loss": 0.0457, "step": 1860 }, { "grad_norm": 0.46341457962989807, "learning_rate": 9.496303254224024e-05, "loss": 0.0519, "step": 1870 }, { "grad_norm": 0.44314858317375183, "learning_rate": 9.489046187225306e-05, "loss": 0.0513, "step": 1880 }, { "grad_norm": 0.369530588388443, "learning_rate": 9.481740028736692e-05, "loss": 0.0452, "step": 1890 }, { "grad_norm": 0.4352402091026306, "learning_rate": 9.474384858657164e-05, "loss": 0.0479, "step": 1900 }, { "grad_norm": 0.7376493215560913, "learning_rate": 9.466980757421679e-05, "loss": 0.051, "step": 1910 }, { "grad_norm": 0.4161694347858429, "learning_rate": 9.459527806000305e-05, "loss": 0.0508, "step": 1920 }, { "grad_norm": 0.6254855990409851, "learning_rate": 9.452026085897325e-05, "loss": 0.0439, "step": 1930 }, { "grad_norm": 0.5774593353271484, "learning_rate": 9.444475679150348e-05, "loss": 0.0477, "step": 1940 }, { "grad_norm": 0.6692832112312317, "learning_rate": 9.436876668329411e-05, "loss": 0.0564, "step": 1950 }, { "grad_norm": 1.5172719955444336, "learning_rate": 9.429229136536079e-05, "loss": 0.0507, "step": 1960 }, { "grad_norm": 0.6628224849700928, "learning_rate": 9.421533167402534e-05, "loss": 0.0539, "step": 1970 }, { "grad_norm": 0.7655863165855408, "learning_rate": 9.413788845090666e-05, "loss": 0.0551, "step": 1980 }, { "grad_norm": 0.6128696799278259, "learning_rate": 9.405996254291136e-05, "loss": 0.0593, "step": 1990 }, { "grad_norm": 0.5116813778877258, "learning_rate": 9.398155480222474e-05, "loss": 0.05, "step": 2000 }, { "grad_norm": 0.3826529383659363, "learning_rate": 9.390266608630128e-05, "loss": 0.0503, "step": 2010 }, { "grad_norm": 0.937923014163971, "learning_rate": 9.38232972578553e-05, "loss": 0.0539, "step": 2020 }, { "grad_norm": 0.3956811726093292, "learning_rate": 9.374344918485164e-05, "loss": 0.0502, "step": 2030 }, { "grad_norm": 0.5579599142074585, "learning_rate": 9.366312274049602e-05, "loss": 0.052, "step": 2040 }, { "grad_norm": 0.5786269307136536, "learning_rate": 9.358231880322554e-05, "loss": 0.0459, "step": 2050 }, { "grad_norm": 0.48672574758529663, "learning_rate": 9.350103825669916e-05, "loss": 0.0543, "step": 2060 }, { "grad_norm": 0.42453330755233765, "learning_rate": 9.341928198978787e-05, "loss": 0.0459, "step": 2070 }, { "grad_norm": 0.5340445637702942, "learning_rate": 9.333705089656512e-05, "loss": 0.0502, "step": 2080 }, { "grad_norm": 0.4847047030925751, "learning_rate": 9.325434587629698e-05, "loss": 0.0445, "step": 2090 }, { "grad_norm": 0.3306330740451813, "learning_rate": 9.31711678334323e-05, "loss": 0.0523, "step": 2100 }, { "grad_norm": 0.6150120496749878, "learning_rate": 9.308751767759282e-05, "loss": 0.0428, "step": 2110 }, { "grad_norm": 0.4795934855937958, "learning_rate": 9.300339632356325e-05, "loss": 0.0509, "step": 2120 }, { "grad_norm": 0.3605153262615204, "learning_rate": 9.291880469128124e-05, "loss": 0.0427, "step": 2130 }, { "grad_norm": 0.41115236282348633, "learning_rate": 9.283374370582732e-05, "loss": 0.0454, "step": 2140 }, { "grad_norm": 0.5758653283119202, "learning_rate": 9.274821429741482e-05, "loss": 0.0477, "step": 2150 }, { "grad_norm": 0.3515821099281311, "learning_rate": 9.266221740137961e-05, "loss": 0.048, "step": 2160 }, { "grad_norm": 0.3458983600139618, "learning_rate": 9.257575395817001e-05, "loss": 0.0423, "step": 2170 }, { "grad_norm": 0.702502429485321, "learning_rate": 9.248882491333637e-05, "loss": 0.0475, "step": 2180 }, { "grad_norm": 0.6328100562095642, "learning_rate": 9.240143121752076e-05, "loss": 0.0508, "step": 2190 }, { "grad_norm": 0.3577858507633209, "learning_rate": 9.23135738264467e-05, "loss": 0.0472, "step": 2200 }, { "grad_norm": 0.8081894516944885, "learning_rate": 9.222525370090849e-05, "loss": 0.0464, "step": 2210 }, { "grad_norm": 0.27200502157211304, "learning_rate": 9.213647180676088e-05, "loss": 0.0424, "step": 2220 }, { "grad_norm": 0.6085584759712219, "learning_rate": 9.204722911490846e-05, "loss": 0.0455, "step": 2230 }, { "grad_norm": 0.4615256190299988, "learning_rate": 9.1957526601295e-05, "loss": 0.0501, "step": 2240 }, { "grad_norm": 0.573671281337738, "learning_rate": 9.186736524689281e-05, "loss": 0.0441, "step": 2250 }, { "grad_norm": 0.45691442489624023, "learning_rate": 9.177674603769204e-05, "loss": 0.0447, "step": 2260 }, { "grad_norm": 0.47951534390449524, "learning_rate": 9.168566996468983e-05, "loss": 0.0464, "step": 2270 }, { "grad_norm": 0.44436559081077576, "learning_rate": 9.159413802387951e-05, "loss": 0.0387, "step": 2280 }, { "grad_norm": 0.7563621401786804, "learning_rate": 9.150215121623974e-05, "loss": 0.0379, "step": 2290 }, { "grad_norm": 0.43905118107795715, "learning_rate": 9.140971054772349e-05, "loss": 0.0422, "step": 2300 }, { "grad_norm": 0.49906694889068604, "learning_rate": 9.131681702924713e-05, "loss": 0.0433, "step": 2310 }, { "grad_norm": 0.5140312910079956, "learning_rate": 9.122347167667926e-05, "loss": 0.0418, "step": 2320 }, { "grad_norm": 0.4769602417945862, "learning_rate": 9.112967551082973e-05, "loss": 0.0455, "step": 2330 }, { "grad_norm": 0.645749032497406, "learning_rate": 9.103542955743835e-05, "loss": 0.0442, "step": 2340 }, { "grad_norm": 0.41193607449531555, "learning_rate": 9.094073484716381e-05, "loss": 0.0494, "step": 2350 }, { "grad_norm": 0.5554569959640503, "learning_rate": 9.084559241557226e-05, "loss": 0.0503, "step": 2360 }, { "grad_norm": 0.46796873211860657, "learning_rate": 9.075000330312608e-05, "loss": 0.0407, "step": 2370 }, { "grad_norm": 0.44065243005752563, "learning_rate": 9.065396855517253e-05, "loss": 0.0497, "step": 2380 }, { "grad_norm": 0.6189984679222107, "learning_rate": 9.055748922193219e-05, "loss": 0.0478, "step": 2390 }, { "grad_norm": 0.28249165415763855, "learning_rate": 9.046056635848761e-05, "loss": 0.0542, "step": 2400 }, { "grad_norm": 0.4959670901298523, "learning_rate": 9.036320102477169e-05, "loss": 0.0415, "step": 2410 }, { "grad_norm": 0.42937374114990234, "learning_rate": 9.02653942855561e-05, "loss": 0.0471, "step": 2420 }, { "grad_norm": 0.4387403428554535, "learning_rate": 9.016714721043971e-05, "loss": 0.0507, "step": 2430 }, { "grad_norm": 0.4716196060180664, "learning_rate": 9.006846087383675e-05, "loss": 0.0518, "step": 2440 }, { "grad_norm": 0.4211229383945465, "learning_rate": 8.996933635496523e-05, "loss": 0.05, "step": 2450 }, { "grad_norm": 0.5069652199745178, "learning_rate": 8.986977473783498e-05, "loss": 0.0455, "step": 2460 }, { "grad_norm": 0.732395350933075, "learning_rate": 8.97697771112359e-05, "loss": 0.0517, "step": 2470 }, { "grad_norm": 0.5265001654624939, "learning_rate": 8.966934456872602e-05, "loss": 0.0411, "step": 2480 }, { "grad_norm": 0.3218359351158142, "learning_rate": 8.95684782086195e-05, "loss": 0.0412, "step": 2490 }, { "grad_norm": 0.41282373666763306, "learning_rate": 8.946717913397476e-05, "loss": 0.0617, "step": 2500 }, { "grad_norm": 0.33135777711868286, "learning_rate": 8.93654484525822e-05, "loss": 0.0513, "step": 2510 }, { "grad_norm": 0.5424937605857849, "learning_rate": 8.926328727695226e-05, "loss": 0.0397, "step": 2520 }, { "grad_norm": 0.6015431880950928, "learning_rate": 8.916069672430319e-05, "loss": 0.0406, "step": 2530 }, { "grad_norm": 0.4396994709968567, "learning_rate": 8.905767791654884e-05, "loss": 0.0371, "step": 2540 }, { "grad_norm": 0.4122588038444519, "learning_rate": 8.895423198028638e-05, "loss": 0.0434, "step": 2550 }, { "grad_norm": 0.363315224647522, "learning_rate": 8.885036004678402e-05, "loss": 0.0386, "step": 2560 }, { "grad_norm": 0.5324111580848694, "learning_rate": 8.874606325196857e-05, "loss": 0.0458, "step": 2570 }, { "grad_norm": 0.43670472502708435, "learning_rate": 8.864134273641304e-05, "loss": 0.0579, "step": 2580 }, { "grad_norm": 0.30471140146255493, "learning_rate": 8.853619964532427e-05, "loss": 0.0417, "step": 2590 }, { "grad_norm": 0.6511371731758118, "learning_rate": 8.843063512853019e-05, "loss": 0.0451, "step": 2600 }, { "grad_norm": 0.454571008682251, "learning_rate": 8.832465034046749e-05, "loss": 0.0572, "step": 2610 }, { "grad_norm": 0.38319551944732666, "learning_rate": 8.821824644016882e-05, "loss": 0.0509, "step": 2620 }, { "grad_norm": 1.85652494430542, "learning_rate": 8.811142459125019e-05, "loss": 0.0462, "step": 2630 }, { "grad_norm": 0.3159908056259155, "learning_rate": 8.800418596189822e-05, "loss": 0.0493, "step": 2640 }, { "grad_norm": 0.5272867679595947, "learning_rate": 8.789653172485737e-05, "loss": 0.0401, "step": 2650 }, { "grad_norm": 0.6878018379211426, "learning_rate": 8.778846305741715e-05, "loss": 0.0409, "step": 2660 }, { "grad_norm": 0.4739729166030884, "learning_rate": 8.767998114139918e-05, "loss": 0.0493, "step": 2670 }, { "grad_norm": 0.45318254828453064, "learning_rate": 8.757108716314429e-05, "loss": 0.0416, "step": 2680 }, { "grad_norm": 0.4417954087257385, "learning_rate": 8.746178231349962e-05, "loss": 0.0411, "step": 2690 }, { "grad_norm": 0.3224528729915619, "learning_rate": 8.735206778780549e-05, "loss": 0.0476, "step": 2700 }, { "grad_norm": 1.2496780157089233, "learning_rate": 8.724194478588234e-05, "loss": 0.0489, "step": 2710 }, { "grad_norm": 0.4489833116531372, "learning_rate": 8.713141451201772e-05, "loss": 0.0433, "step": 2720 }, { "grad_norm": 0.5782540440559387, "learning_rate": 8.702047817495295e-05, "loss": 0.0409, "step": 2730 }, { "grad_norm": 0.33869582414627075, "learning_rate": 8.69091369878701e-05, "loss": 0.0455, "step": 2740 }, { "grad_norm": 0.6005355715751648, "learning_rate": 8.679739216837849e-05, "loss": 0.0475, "step": 2750 }, { "grad_norm": 0.5582241415977478, "learning_rate": 8.66852449385016e-05, "loss": 0.0431, "step": 2760 }, { "grad_norm": 0.5648136138916016, "learning_rate": 8.657269652466356e-05, "loss": 0.0477, "step": 2770 }, { "grad_norm": 0.44452986121177673, "learning_rate": 8.645974815767577e-05, "loss": 0.0468, "step": 2780 }, { "grad_norm": 0.33636653423309326, "learning_rate": 8.634640107272351e-05, "loss": 0.0373, "step": 2790 }, { "grad_norm": 0.5762991905212402, "learning_rate": 8.623265650935234e-05, "loss": 0.0402, "step": 2800 }, { "grad_norm": 0.5141375064849854, "learning_rate": 8.611851571145456e-05, "loss": 0.0378, "step": 2810 }, { "grad_norm": 0.4655013084411621, "learning_rate": 8.600397992725566e-05, "loss": 0.0363, "step": 2820 }, { "grad_norm": 0.7830609083175659, "learning_rate": 8.588905040930061e-05, "loss": 0.0382, "step": 2830 }, { "grad_norm": 0.4160034954547882, "learning_rate": 8.577372841444022e-05, "loss": 0.0423, "step": 2840 }, { "grad_norm": 0.3848729729652405, "learning_rate": 8.565801520381736e-05, "loss": 0.0472, "step": 2850 }, { "grad_norm": 0.4083252251148224, "learning_rate": 8.554191204285313e-05, "loss": 0.0556, "step": 2860 }, { "grad_norm": 0.41059887409210205, "learning_rate": 8.542542020123315e-05, "loss": 0.0382, "step": 2870 }, { "grad_norm": 0.42460960149765015, "learning_rate": 8.530854095289347e-05, "loss": 0.0429, "step": 2880 }, { "grad_norm": 0.29321345686912537, "learning_rate": 8.519127557600688e-05, "loss": 0.0403, "step": 2890 }, { "grad_norm": 0.4358566105365753, "learning_rate": 8.507362535296871e-05, "loss": 0.0358, "step": 2900 }, { "grad_norm": 0.3429378867149353, "learning_rate": 8.495559157038299e-05, "loss": 0.0503, "step": 2910 }, { "grad_norm": 0.3100273609161377, "learning_rate": 8.483717551904823e-05, "loss": 0.0402, "step": 2920 }, { "grad_norm": 0.3722005784511566, "learning_rate": 8.47183784939434e-05, "loss": 0.0393, "step": 2930 }, { "grad_norm": 0.48891112208366394, "learning_rate": 8.459920179421374e-05, "loss": 0.0462, "step": 2940 }, { "grad_norm": 0.37187764048576355, "learning_rate": 8.447964672315656e-05, "loss": 0.0355, "step": 2950 }, { "grad_norm": 0.3038395345211029, "learning_rate": 8.435971458820692e-05, "loss": 0.0334, "step": 2960 }, { "grad_norm": 0.5634755492210388, "learning_rate": 8.423940670092345e-05, "loss": 0.0482, "step": 2970 }, { "grad_norm": 0.5036453008651733, "learning_rate": 8.411872437697394e-05, "loss": 0.038, "step": 2980 }, { "grad_norm": 0.4866580069065094, "learning_rate": 8.399766893612096e-05, "loss": 0.0348, "step": 2990 }, { "grad_norm": 0.5499194860458374, "learning_rate": 8.38762417022074e-05, "loss": 0.0438, "step": 3000 }, { "grad_norm": 0.677514910697937, "learning_rate": 8.375444400314204e-05, "loss": 0.0377, "step": 3010 }, { "grad_norm": 0.32236215472221375, "learning_rate": 8.3632277170885e-05, "loss": 0.0392, "step": 3020 }, { "grad_norm": 0.37690380215644836, "learning_rate": 8.350974254143318e-05, "loss": 0.0384, "step": 3030 }, { "grad_norm": 0.7834134697914124, "learning_rate": 8.338684145480566e-05, "loss": 0.0338, "step": 3040 }, { "grad_norm": 0.5244308710098267, "learning_rate": 8.326357525502904e-05, "loss": 0.0367, "step": 3050 }, { "grad_norm": 0.3667405843734741, "learning_rate": 8.313994529012273e-05, "loss": 0.0298, "step": 3060 }, { "grad_norm": 0.3164689838886261, "learning_rate": 8.301595291208422e-05, "loss": 0.039, "step": 3070 }, { "grad_norm": 0.5000348091125488, "learning_rate": 8.289159947687427e-05, "loss": 0.0399, "step": 3080 }, { "grad_norm": 0.5161594748497009, "learning_rate": 8.276688634440216e-05, "loss": 0.0414, "step": 3090 }, { "grad_norm": 0.41306668519973755, "learning_rate": 8.26418148785107e-05, "loss": 0.0465, "step": 3100 }, { "grad_norm": 0.4972187280654907, "learning_rate": 8.251638644696141e-05, "loss": 0.0464, "step": 3110 }, { "grad_norm": 0.40276795625686646, "learning_rate": 8.23906024214195e-05, "loss": 0.0493, "step": 3120 }, { "grad_norm": 0.6121402978897095, "learning_rate": 8.226446417743897e-05, "loss": 0.0531, "step": 3130 }, { "grad_norm": 0.4457087814807892, "learning_rate": 8.213797309444742e-05, "loss": 0.0547, "step": 3140 }, { "grad_norm": 0.48042023181915283, "learning_rate": 8.201113055573105e-05, "loss": 0.044, "step": 3150 }, { "grad_norm": 0.4738384783267975, "learning_rate": 8.188393794841958e-05, "loss": 0.0447, "step": 3160 }, { "grad_norm": 0.5618190169334412, "learning_rate": 8.175639666347094e-05, "loss": 0.0412, "step": 3170 }, { "grad_norm": 0.5391665697097778, "learning_rate": 8.162850809565623e-05, "loss": 0.0406, "step": 3180 }, { "grad_norm": 0.45960521697998047, "learning_rate": 8.150027364354431e-05, "loss": 0.0395, "step": 3190 }, { "grad_norm": 1.1940240859985352, "learning_rate": 8.137169470948662e-05, "loss": 0.0412, "step": 3200 }, { "grad_norm": 0.8628629446029663, "learning_rate": 8.124277269960179e-05, "loss": 0.0511, "step": 3210 }, { "grad_norm": 0.4749574661254883, "learning_rate": 8.111350902376023e-05, "loss": 0.04, "step": 3220 }, { "grad_norm": 0.5520264506340027, "learning_rate": 8.098390509556883e-05, "loss": 0.0472, "step": 3230 }, { "grad_norm": 0.7796582579612732, "learning_rate": 8.085396233235536e-05, "loss": 0.046, "step": 3240 }, { "grad_norm": 0.37452542781829834, "learning_rate": 8.072368215515306e-05, "loss": 0.0443, "step": 3250 }, { "grad_norm": 0.6499863266944885, "learning_rate": 8.059306598868506e-05, "loss": 0.036, "step": 3260 }, { "grad_norm": 0.47941604256629944, "learning_rate": 8.046211526134888e-05, "loss": 0.048, "step": 3270 }, { "grad_norm": 1.3007968664169312, "learning_rate": 8.033083140520065e-05, "loss": 0.0485, "step": 3280 }, { "grad_norm": 0.3602273762226105, "learning_rate": 8.019921585593962e-05, "loss": 0.044, "step": 3290 }, { "grad_norm": 0.48344120383262634, "learning_rate": 8.006727005289232e-05, "loss": 0.0413, "step": 3300 }, { "grad_norm": 0.35201558470726013, "learning_rate": 7.993499543899692e-05, "loss": 0.0498, "step": 3310 }, { "grad_norm": 0.3839559853076935, "learning_rate": 7.980239346078742e-05, "loss": 0.0388, "step": 3320 }, { "grad_norm": 0.4483330547809601, "learning_rate": 7.966946556837778e-05, "loss": 0.0387, "step": 3330 }, { "grad_norm": 0.539341151714325, "learning_rate": 7.953621321544616e-05, "loss": 0.0383, "step": 3340 }, { "grad_norm": 0.6192002892494202, "learning_rate": 7.940263785921896e-05, "loss": 0.0366, "step": 3350 }, { "grad_norm": 0.5550284385681152, "learning_rate": 7.926874096045482e-05, "loss": 0.0416, "step": 3360 }, { "grad_norm": 0.33168432116508484, "learning_rate": 7.913452398342881e-05, "loss": 0.0433, "step": 3370 }, { "grad_norm": 0.3350684940814972, "learning_rate": 7.89999883959163e-05, "loss": 0.0521, "step": 3380 }, { "grad_norm": 0.5977014303207397, "learning_rate": 7.886513566917687e-05, "loss": 0.0367, "step": 3390 }, { "grad_norm": 0.32529667019844055, "learning_rate": 7.872996727793838e-05, "loss": 0.0322, "step": 3400 }, { "grad_norm": 0.45305153727531433, "learning_rate": 7.859448470038069e-05, "loss": 0.0384, "step": 3410 }, { "grad_norm": 0.8965548872947693, "learning_rate": 7.845868941811956e-05, "loss": 0.0493, "step": 3420 }, { "grad_norm": 0.44502463936805725, "learning_rate": 7.832258291619043e-05, "loss": 0.0375, "step": 3430 }, { "grad_norm": 0.5077293515205383, "learning_rate": 7.81861666830322e-05, "loss": 0.0334, "step": 3440 }, { "grad_norm": 0.6367596387863159, "learning_rate": 7.804944221047097e-05, "loss": 0.0334, "step": 3450 }, { "grad_norm": 0.6158141493797302, "learning_rate": 7.791241099370364e-05, "loss": 0.0332, "step": 3460 }, { "grad_norm": 0.4262571632862091, "learning_rate": 7.777507453128163e-05, "loss": 0.0365, "step": 3470 }, { "grad_norm": 0.40228477120399475, "learning_rate": 7.763743432509451e-05, "loss": 0.0319, "step": 3480 }, { "grad_norm": 0.28088390827178955, "learning_rate": 7.749949188035353e-05, "loss": 0.0456, "step": 3490 }, { "grad_norm": 0.5139207243919373, "learning_rate": 7.736124870557516e-05, "loss": 0.0388, "step": 3500 }, { "grad_norm": 0.6150637269020081, "learning_rate": 7.722270631256459e-05, "loss": 0.0338, "step": 3510 }, { "grad_norm": 0.45171260833740234, "learning_rate": 7.708386621639925e-05, "loss": 0.0357, "step": 3520 }, { "grad_norm": 0.46733397245407104, "learning_rate": 7.694472993541219e-05, "loss": 0.0302, "step": 3530 }, { "grad_norm": 0.37751492857933044, "learning_rate": 7.680529899117547e-05, "loss": 0.0375, "step": 3540 }, { "grad_norm": 0.5415570735931396, "learning_rate": 7.666557490848358e-05, "loss": 0.0454, "step": 3550 }, { "grad_norm": 0.25973278284072876, "learning_rate": 7.65255592153367e-05, "loss": 0.0362, "step": 3560 }, { "grad_norm": 0.42604586482048035, "learning_rate": 7.638525344292402e-05, "loss": 0.0391, "step": 3570 }, { "grad_norm": 0.32205280661582947, "learning_rate": 7.624465912560697e-05, "loss": 0.0439, "step": 3580 }, { "grad_norm": 0.4151095747947693, "learning_rate": 7.610377780090249e-05, "loss": 0.0429, "step": 3590 }, { "grad_norm": 0.4224369525909424, "learning_rate": 7.596261100946618e-05, "loss": 0.038, "step": 3600 }, { "grad_norm": 0.41639986634254456, "learning_rate": 7.582116029507542e-05, "loss": 0.0411, "step": 3610 }, { "grad_norm": 0.3002557158470154, "learning_rate": 7.56794272046126e-05, "loss": 0.0426, "step": 3620 }, { "grad_norm": 0.5755050778388977, "learning_rate": 7.55374132880481e-05, "loss": 0.0462, "step": 3630 }, { "grad_norm": 0.43357717990875244, "learning_rate": 7.539512009842333e-05, "loss": 0.0379, "step": 3640 }, { "grad_norm": 0.6342546939849854, "learning_rate": 7.525254919183382e-05, "loss": 0.0402, "step": 3650 }, { "grad_norm": 0.3681163191795349, "learning_rate": 7.510970212741215e-05, "loss": 0.0334, "step": 3660 }, { "grad_norm": 0.31042253971099854, "learning_rate": 7.496658046731096e-05, "loss": 0.0434, "step": 3670 }, { "grad_norm": 0.3157895803451538, "learning_rate": 7.482318577668578e-05, "loss": 0.035, "step": 3680 }, { "grad_norm": 0.6451574563980103, "learning_rate": 7.467951962367796e-05, "loss": 0.0406, "step": 3690 }, { "grad_norm": 0.45665308833122253, "learning_rate": 7.453558357939755e-05, "loss": 0.0428, "step": 3700 }, { "grad_norm": 0.8346014022827148, "learning_rate": 7.439137921790606e-05, "loss": 0.0427, "step": 3710 }, { "grad_norm": 0.7972629070281982, "learning_rate": 7.42469081161993e-05, "loss": 0.0485, "step": 3720 }, { "grad_norm": 0.4633433520793915, "learning_rate": 7.410217185419006e-05, "loss": 0.033, "step": 3730 }, { "grad_norm": 0.45173871517181396, "learning_rate": 7.395717201469095e-05, "loss": 0.0431, "step": 3740 }, { "grad_norm": 0.4394107758998871, "learning_rate": 7.381191018339696e-05, "loss": 0.0359, "step": 3750 }, { "grad_norm": 0.5095446705818176, "learning_rate": 7.36663879488682e-05, "loss": 0.0364, "step": 3760 }, { "grad_norm": 0.619640052318573, "learning_rate": 7.352060690251254e-05, "loss": 0.0395, "step": 3770 }, { "grad_norm": 0.399773508310318, "learning_rate": 7.337456863856811e-05, "loss": 0.0342, "step": 3780 }, { "grad_norm": 0.8159610033035278, "learning_rate": 7.3228274754086e-05, "loss": 0.043, "step": 3790 }, { "grad_norm": 0.6848739981651306, "learning_rate": 7.308172684891267e-05, "loss": 0.04, "step": 3800 }, { "grad_norm": 0.6661529541015625, "learning_rate": 7.293492652567255e-05, "loss": 0.032, "step": 3810 }, { "grad_norm": 0.8737601041793823, "learning_rate": 7.278787538975043e-05, "loss": 0.041, "step": 3820 }, { "grad_norm": 0.3100826144218445, "learning_rate": 7.2640575049274e-05, "loss": 0.0366, "step": 3830 }, { "grad_norm": 0.4944784939289093, "learning_rate": 7.249302711509616e-05, "loss": 0.0357, "step": 3840 }, { "grad_norm": 0.3880489468574524, "learning_rate": 7.23452332007775e-05, "loss": 0.0395, "step": 3850 }, { "grad_norm": 0.44571420550346375, "learning_rate": 7.219719492256858e-05, "loss": 0.0372, "step": 3860 }, { "grad_norm": 0.511232316493988, "learning_rate": 7.20489138993923e-05, "loss": 0.0415, "step": 3870 }, { "grad_norm": 0.3063661456108093, "learning_rate": 7.190039175282614e-05, "loss": 0.0349, "step": 3880 }, { "grad_norm": 0.4156712293624878, "learning_rate": 7.175163010708455e-05, "loss": 0.0492, "step": 3890 }, { "grad_norm": 0.7749978303909302, "learning_rate": 7.1602630589001e-05, "loss": 0.0433, "step": 3900 }, { "grad_norm": 0.35543760657310486, "learning_rate": 7.14533948280104e-05, "loss": 0.0361, "step": 3910 }, { "grad_norm": 0.49793151021003723, "learning_rate": 7.130392445613109e-05, "loss": 0.0406, "step": 3920 }, { "grad_norm": 0.5292438864707947, "learning_rate": 7.115422110794711e-05, "loss": 0.0416, "step": 3930 }, { "grad_norm": 0.3980823755264282, "learning_rate": 7.100428642059033e-05, "loss": 0.0423, "step": 3940 }, { "grad_norm": 0.24081678688526154, "learning_rate": 7.08541220337224e-05, "loss": 0.0381, "step": 3950 }, { "grad_norm": 0.4224522113800049, "learning_rate": 7.070372958951706e-05, "loss": 0.0363, "step": 3960 }, { "grad_norm": 1.3326479196548462, "learning_rate": 7.055311073264194e-05, "loss": 0.0416, "step": 3970 }, { "grad_norm": 0.37363454699516296, "learning_rate": 7.040226711024077e-05, "loss": 0.0437, "step": 3980 }, { "grad_norm": 0.6010105609893799, "learning_rate": 7.02512003719152e-05, "loss": 0.0429, "step": 3990 }, { "grad_norm": 0.5243247151374817, "learning_rate": 7.00999121697069e-05, "loss": 0.0412, "step": 4000 }, { "grad_norm": 0.6635685563087463, "learning_rate": 6.99484041580794e-05, "loss": 0.0372, "step": 4010 }, { "grad_norm": 0.5024315118789673, "learning_rate": 6.979667799390004e-05, "loss": 0.0384, "step": 4020 }, { "grad_norm": 0.48179805278778076, "learning_rate": 6.964473533642185e-05, "loss": 0.042, "step": 4030 }, { "grad_norm": 0.3786553144454956, "learning_rate": 6.949257784726539e-05, "loss": 0.0411, "step": 4040 }, { "grad_norm": 0.504101574420929, "learning_rate": 6.934020719040056e-05, "loss": 0.0389, "step": 4050 }, { "grad_norm": 0.49691036343574524, "learning_rate": 6.918762503212848e-05, "loss": 0.0362, "step": 4060 }, { "grad_norm": 0.8493368029594421, "learning_rate": 6.903483304106319e-05, "loss": 0.0351, "step": 4070 }, { "grad_norm": 0.3916228413581848, "learning_rate": 6.888183288811341e-05, "loss": 0.0435, "step": 4080 }, { "grad_norm": 0.3568408191204071, "learning_rate": 6.87286262464643e-05, "loss": 0.0397, "step": 4090 }, { "grad_norm": 0.4437256157398224, "learning_rate": 6.857521479155915e-05, "loss": 0.0319, "step": 4100 }, { "grad_norm": 0.7760893702507019, "learning_rate": 6.842160020108104e-05, "loss": 0.0349, "step": 4110 }, { "grad_norm": 0.2659861147403717, "learning_rate": 6.826778415493455e-05, "loss": 0.0422, "step": 4120 }, { "grad_norm": 0.3792249262332916, "learning_rate": 6.811376833522729e-05, "loss": 0.0287, "step": 4130 }, { "grad_norm": 0.5829780101776123, "learning_rate": 6.795955442625159e-05, "loss": 0.0407, "step": 4140 }, { "grad_norm": 0.448262095451355, "learning_rate": 6.780514411446608e-05, "loss": 0.041, "step": 4150 }, { "grad_norm": 0.5682547688484192, "learning_rate": 6.765053908847716e-05, "loss": 0.0313, "step": 4160 }, { "grad_norm": 0.3665095567703247, "learning_rate": 6.749574103902064e-05, "loss": 0.0376, "step": 4170 }, { "grad_norm": 0.4075768291950226, "learning_rate": 6.734075165894317e-05, "loss": 0.0375, "step": 4180 }, { "grad_norm": 0.4387996792793274, "learning_rate": 6.71855726431838e-05, "loss": 0.0355, "step": 4190 }, { "grad_norm": 0.44973334670066833, "learning_rate": 6.703020568875538e-05, "loss": 0.0428, "step": 4200 }, { "grad_norm": 0.687142014503479, "learning_rate": 6.687465249472603e-05, "loss": 0.0423, "step": 4210 }, { "grad_norm": 0.6020686030387878, "learning_rate": 6.671891476220055e-05, "loss": 0.0286, "step": 4220 }, { "grad_norm": 0.310160368680954, "learning_rate": 6.656299419430183e-05, "loss": 0.0328, "step": 4230 }, { "grad_norm": 0.30104905366897583, "learning_rate": 6.640689249615223e-05, "loss": 0.0313, "step": 4240 }, { "grad_norm": 0.38839849829673767, "learning_rate": 6.625061137485491e-05, "loss": 0.0488, "step": 4250 }, { "grad_norm": 0.4284921884536743, "learning_rate": 6.609415253947517e-05, "loss": 0.0316, "step": 4260 }, { "grad_norm": 0.5058846473693848, "learning_rate": 6.593751770102178e-05, "loss": 0.028, "step": 4270 }, { "grad_norm": 0.3010232150554657, "learning_rate": 6.578070857242823e-05, "loss": 0.0332, "step": 4280 }, { "grad_norm": 0.4770226776599884, "learning_rate": 6.562372686853402e-05, "loss": 0.032, "step": 4290 }, { "grad_norm": 0.3739854395389557, "learning_rate": 6.546657430606593e-05, "loss": 0.0314, "step": 4300 }, { "grad_norm": 0.29421791434288025, "learning_rate": 6.530925260361918e-05, "loss": 0.0336, "step": 4310 }, { "grad_norm": 0.41391998529434204, "learning_rate": 6.515176348163871e-05, "loss": 0.0349, "step": 4320 }, { "grad_norm": 0.6603525280952454, "learning_rate": 6.499410866240032e-05, "loss": 0.0322, "step": 4330 }, { "grad_norm": 0.3882249593734741, "learning_rate": 6.48362898699919e-05, "loss": 0.0338, "step": 4340 }, { "grad_norm": 0.5888624787330627, "learning_rate": 6.467830883029443e-05, "loss": 0.045, "step": 4350 }, { "grad_norm": 1.6214359998703003, "learning_rate": 6.452016727096326e-05, "loss": 0.0476, "step": 4360 }, { "grad_norm": 0.30280470848083496, "learning_rate": 6.436186692140916e-05, "loss": 0.0379, "step": 4370 }, { "grad_norm": 0.3055734932422638, "learning_rate": 6.420340951277938e-05, "loss": 0.0428, "step": 4380 }, { "grad_norm": 0.47408896684646606, "learning_rate": 6.404479677793874e-05, "loss": 0.0494, "step": 4390 }, { "grad_norm": 0.3791012167930603, "learning_rate": 6.388603045145075e-05, "loss": 0.0434, "step": 4400 }, { "grad_norm": 0.3020074665546417, "learning_rate": 6.372711226955843e-05, "loss": 0.0351, "step": 4410 }, { "grad_norm": 0.6280782222747803, "learning_rate": 6.356804397016564e-05, "loss": 0.0343, "step": 4420 }, { "grad_norm": 0.3259531855583191, "learning_rate": 6.340882729281779e-05, "loss": 0.0362, "step": 4430 }, { "grad_norm": 0.4053093194961548, "learning_rate": 6.324946397868294e-05, "loss": 0.0344, "step": 4440 }, { "grad_norm": 0.8536799550056458, "learning_rate": 6.308995577053276e-05, "loss": 0.0498, "step": 4450 }, { "grad_norm": 0.5600643157958984, "learning_rate": 6.293030441272347e-05, "loss": 0.0361, "step": 4460 }, { "grad_norm": 1.0406831502914429, "learning_rate": 6.277051165117677e-05, "loss": 0.0418, "step": 4470 }, { "grad_norm": 0.2640027403831482, "learning_rate": 6.261057923336064e-05, "loss": 0.0288, "step": 4480 }, { "grad_norm": 0.413507878780365, "learning_rate": 6.245050890827042e-05, "loss": 0.0306, "step": 4490 }, { "grad_norm": 0.4595297873020172, "learning_rate": 6.229030242640952e-05, "loss": 0.034, "step": 4500 }, { "grad_norm": 0.4358069598674774, "learning_rate": 6.212996153977037e-05, "loss": 0.0369, "step": 4510 }, { "grad_norm": 0.24214838445186615, "learning_rate": 6.196948800181523e-05, "loss": 0.0324, "step": 4520 }, { "grad_norm": 0.6787503957748413, "learning_rate": 6.180888356745695e-05, "loss": 0.0312, "step": 4530 }, { "grad_norm": 0.6001297235488892, "learning_rate": 6.164814999303995e-05, "loss": 0.0388, "step": 4540 }, { "grad_norm": 0.3514004945755005, "learning_rate": 6.148728903632081e-05, "loss": 0.039, "step": 4550 }, { "grad_norm": 0.37345126271247864, "learning_rate": 6.132630245644921e-05, "loss": 0.0464, "step": 4560 }, { "grad_norm": 0.5382639765739441, "learning_rate": 6.116519201394857e-05, "loss": 0.0477, "step": 4570 }, { "grad_norm": 0.39688408374786377, "learning_rate": 6.10039594706969e-05, "loss": 0.0344, "step": 4580 }, { "grad_norm": 0.42196446657180786, "learning_rate": 6.084260658990744e-05, "loss": 0.0337, "step": 4590 }, { "grad_norm": 0.9615933299064636, "learning_rate": 6.068113513610943e-05, "loss": 0.0361, "step": 4600 }, { "grad_norm": 0.4376397728919983, "learning_rate": 6.0519546875128876e-05, "loss": 0.0421, "step": 4610 }, { "grad_norm": 0.4625268280506134, "learning_rate": 6.035784357406906e-05, "loss": 0.0427, "step": 4620 }, { "grad_norm": 0.5378472208976746, "learning_rate": 6.01960270012914e-05, "loss": 0.0397, "step": 4630 }, { "grad_norm": 0.39504826068878174, "learning_rate": 6.003409892639599e-05, "loss": 0.0407, "step": 4640 }, { "grad_norm": 0.4013054370880127, "learning_rate": 5.9872061120202336e-05, "loss": 0.0356, "step": 4650 }, { "grad_norm": 0.39824575185775757, "learning_rate": 5.9709915354729914e-05, "loss": 0.0348, "step": 4660 }, { "grad_norm": 0.31824424862861633, "learning_rate": 5.9547663403178824e-05, "loss": 0.0373, "step": 4670 }, { "grad_norm": 0.646044909954071, "learning_rate": 5.9385307039910445e-05, "loss": 0.0346, "step": 4680 }, { "grad_norm": 0.4102570414543152, "learning_rate": 5.922284804042792e-05, "loss": 0.0293, "step": 4690 }, { "grad_norm": 0.3224484622478485, "learning_rate": 5.906028818135687e-05, "loss": 0.0322, "step": 4700 }, { "grad_norm": 0.4025365114212036, "learning_rate": 5.889762924042585e-05, "loss": 0.0333, "step": 4710 }, { "grad_norm": 0.19355787336826324, "learning_rate": 5.873487299644699e-05, "loss": 0.0273, "step": 4720 }, { "grad_norm": 0.8183116912841797, "learning_rate": 5.857202122929649e-05, "loss": 0.0351, "step": 4730 }, { "grad_norm": 0.38802966475486755, "learning_rate": 5.840907571989518e-05, "loss": 0.0401, "step": 4740 }, { "grad_norm": 0.47111502289772034, "learning_rate": 5.824603825018904e-05, "loss": 0.0303, "step": 4750 }, { "grad_norm": 0.5380892753601074, "learning_rate": 5.808291060312975e-05, "loss": 0.037, "step": 4760 }, { "grad_norm": 0.3510511517524719, "learning_rate": 5.7919694562655083e-05, "loss": 0.0306, "step": 4770 }, { "grad_norm": 1.1914355754852295, "learning_rate": 5.775639191366954e-05, "loss": 0.0285, "step": 4780 }, { "grad_norm": 0.4094200134277344, "learning_rate": 5.75930044420247e-05, "loss": 0.0279, "step": 4790 }, { "grad_norm": 0.20832709968090057, "learning_rate": 5.74295339344998e-05, "loss": 0.0258, "step": 4800 }, { "grad_norm": 0.3135875463485718, "learning_rate": 5.726598217878211e-05, "loss": 0.0292, "step": 4810 }, { "grad_norm": 0.47905635833740234, "learning_rate": 5.71023509634474e-05, "loss": 0.0304, "step": 4820 }, { "grad_norm": 0.4722461402416229, "learning_rate": 5.693864207794049e-05, "loss": 0.0414, "step": 4830 }, { "grad_norm": 0.5215014815330505, "learning_rate": 5.677485731255545e-05, "loss": 0.0356, "step": 4840 }, { "grad_norm": 0.42558372020721436, "learning_rate": 5.6610998458416296e-05, "loss": 0.0341, "step": 4850 }, { "grad_norm": 0.33624106645584106, "learning_rate": 5.644706730745716e-05, "loss": 0.0351, "step": 4860 }, { "grad_norm": 1.5069810152053833, "learning_rate": 5.628306565240287e-05, "loss": 0.04, "step": 4870 }, { "grad_norm": 0.5245074033737183, "learning_rate": 5.611899528674923e-05, "loss": 0.0419, "step": 4880 }, { "grad_norm": 0.36282992362976074, "learning_rate": 5.595485800474349e-05, "loss": 0.0405, "step": 4890 }, { "grad_norm": 0.3294852674007416, "learning_rate": 5.579065560136467e-05, "loss": 0.0295, "step": 4900 }, { "grad_norm": 0.4048248529434204, "learning_rate": 5.562638987230392e-05, "loss": 0.0274, "step": 4910 }, { "grad_norm": 0.6282249093055725, "learning_rate": 5.546206261394498e-05, "loss": 0.0347, "step": 4920 }, { "grad_norm": 0.4325050711631775, "learning_rate": 5.529767562334437e-05, "loss": 0.0328, "step": 4930 }, { "grad_norm": 0.4243454039096832, "learning_rate": 5.5133230698211926e-05, "loss": 0.0445, "step": 4940 }, { "grad_norm": 0.30980050563812256, "learning_rate": 5.496872963689096e-05, "loss": 0.0449, "step": 4950 }, { "grad_norm": 0.4702783226966858, "learning_rate": 5.4804174238338756e-05, "loss": 0.0394, "step": 4960 }, { "grad_norm": 0.35800647735595703, "learning_rate": 5.463956630210678e-05, "loss": 0.0303, "step": 4970 }, { "grad_norm": 0.38920822739601135, "learning_rate": 5.4474907628321046e-05, "loss": 0.0381, "step": 4980 }, { "grad_norm": 0.25693026185035706, "learning_rate": 5.431020001766244e-05, "loss": 0.0324, "step": 4990 }, { "grad_norm": 0.29267191886901855, "learning_rate": 5.4145445271346986e-05, "loss": 0.0334, "step": 5000 }, { "grad_norm": 0.3412073850631714, "learning_rate": 5.398064519110622e-05, "loss": 0.0385, "step": 5010 }, { "grad_norm": 0.42189228534698486, "learning_rate": 5.3815801579167394e-05, "loss": 0.0351, "step": 5020 }, { "grad_norm": 0.3074387311935425, "learning_rate": 5.365091623823382e-05, "loss": 0.0367, "step": 5030 }, { "grad_norm": 0.291652649641037, "learning_rate": 5.348599097146521e-05, "loss": 0.0306, "step": 5040 }, { "grad_norm": 0.2838083505630493, "learning_rate": 5.3321027582457836e-05, "loss": 0.0329, "step": 5050 }, { "grad_norm": 0.3992389142513275, "learning_rate": 5.315602787522491e-05, "loss": 0.0361, "step": 5060 }, { "grad_norm": 0.7992575168609619, "learning_rate": 5.299099365417678e-05, "loss": 0.0303, "step": 5070 }, { "grad_norm": 0.5181387662887573, "learning_rate": 5.2825926724101236e-05, "loss": 0.0307, "step": 5080 }, { "grad_norm": 0.7917940616607666, "learning_rate": 5.26608288901438e-05, "loss": 0.0332, "step": 5090 }, { "grad_norm": 0.45531561970710754, "learning_rate": 5.24957019577879e-05, "loss": 0.0324, "step": 5100 }, { "grad_norm": 0.4324059784412384, "learning_rate": 5.2330547732835266e-05, "loss": 0.0383, "step": 5110 }, { "grad_norm": 0.4848370850086212, "learning_rate": 5.2165368021385996e-05, "loss": 0.0366, "step": 5120 }, { "grad_norm": 0.39073967933654785, "learning_rate": 5.200016462981897e-05, "loss": 0.0261, "step": 5130 }, { "grad_norm": 1.231094241142273, "learning_rate": 5.1834939364772015e-05, "loss": 0.0352, "step": 5140 }, { "grad_norm": 0.396804541349411, "learning_rate": 5.166969403312214e-05, "loss": 0.0366, "step": 5150 }, { "grad_norm": 0.27492445707321167, "learning_rate": 5.1504430441965844e-05, "loss": 0.0254, "step": 5160 }, { "grad_norm": 1.2446248531341553, "learning_rate": 5.133915039859923e-05, "loss": 0.0381, "step": 5170 }, { "grad_norm": 0.29994118213653564, "learning_rate": 5.1173855710498444e-05, "loss": 0.0291, "step": 5180 }, { "grad_norm": 0.297990620136261, "learning_rate": 5.100854818529967e-05, "loss": 0.0314, "step": 5190 }, { "grad_norm": 0.27462977170944214, "learning_rate": 5.084322963077951e-05, "loss": 0.0257, "step": 5200 }, { "grad_norm": 0.3500578701496124, "learning_rate": 5.067790185483522e-05, "loss": 0.0291, "step": 5210 }, { "grad_norm": 0.44426748156547546, "learning_rate": 5.0512566665464844e-05, "loss": 0.0273, "step": 5220 }, { "grad_norm": 0.6191573143005371, "learning_rate": 5.034722587074755e-05, "loss": 0.0352, "step": 5230 }, { "grad_norm": 0.3504461348056793, "learning_rate": 5.018188127882375e-05, "loss": 0.028, "step": 5240 }, { "grad_norm": 0.5457392930984497, "learning_rate": 5.0016534697875417e-05, "loss": 0.027, "step": 5250 }, { "grad_norm": 0.33397963643074036, "learning_rate": 4.9851187936106294e-05, "loss": 0.0341, "step": 5260 }, { "grad_norm": 0.5520656704902649, "learning_rate": 4.968584280172206e-05, "loss": 0.0311, "step": 5270 }, { "grad_norm": 0.5733316540718079, "learning_rate": 4.95205011029106e-05, "loss": 0.0303, "step": 5280 }, { "grad_norm": 0.5954337120056152, "learning_rate": 4.935516464782227e-05, "loss": 0.03, "step": 5290 }, { "grad_norm": 0.4418424069881439, "learning_rate": 4.918983524455003e-05, "loss": 0.0341, "step": 5300 }, { "grad_norm": 0.3413905203342438, "learning_rate": 4.9024514701109766e-05, "loss": 0.0313, "step": 5310 }, { "grad_norm": 0.37635546922683716, "learning_rate": 4.885920482542043e-05, "loss": 0.034, "step": 5320 }, { "grad_norm": 0.6968426704406738, "learning_rate": 4.869390742528438e-05, "loss": 0.0265, "step": 5330 }, { "grad_norm": 0.5819005370140076, "learning_rate": 4.852862430836744e-05, "loss": 0.0343, "step": 5340 }, { "grad_norm": 0.7570703029632568, "learning_rate": 4.836335728217933e-05, "loss": 0.0382, "step": 5350 }, { "grad_norm": 0.5112927556037903, "learning_rate": 4.819810815405379e-05, "loss": 0.0365, "step": 5360 }, { "grad_norm": 0.475203275680542, "learning_rate": 4.803287873112877e-05, "loss": 0.0278, "step": 5370 }, { "grad_norm": 0.3377304673194885, "learning_rate": 4.786767082032681e-05, "loss": 0.0293, "step": 5380 }, { "grad_norm": 0.4347391426563263, "learning_rate": 4.77024862283351e-05, "loss": 0.0287, "step": 5390 }, { "grad_norm": 0.2404053956270218, "learning_rate": 4.753732676158593e-05, "loss": 0.0324, "step": 5400 }, { "grad_norm": 0.29286518692970276, "learning_rate": 4.737219422623672e-05, "loss": 0.0311, "step": 5410 }, { "grad_norm": 1.0225869417190552, "learning_rate": 4.720709042815044e-05, "loss": 0.0328, "step": 5420 }, { "grad_norm": 0.3591194152832031, "learning_rate": 4.704201717287578e-05, "loss": 0.0319, "step": 5430 }, { "grad_norm": 0.5703968405723572, "learning_rate": 4.6876976265627404e-05, "loss": 0.0292, "step": 5440 }, { "grad_norm": 0.3349083364009857, "learning_rate": 4.671196951126626e-05, "loss": 0.0264, "step": 5450 }, { "grad_norm": 0.32466208934783936, "learning_rate": 4.654699871427971e-05, "loss": 0.0298, "step": 5460 }, { "grad_norm": 0.5054230093955994, "learning_rate": 4.6382065678762034e-05, "loss": 0.0286, "step": 5470 }, { "grad_norm": 2.2849557399749756, "learning_rate": 4.6217172208394424e-05, "loss": 0.0305, "step": 5480 }, { "grad_norm": 0.3523491322994232, "learning_rate": 4.605232010642549e-05, "loss": 0.0297, "step": 5490 }, { "grad_norm": 0.30912867188453674, "learning_rate": 4.588751117565142e-05, "loss": 0.0283, "step": 5500 }, { "grad_norm": 0.7438023090362549, "learning_rate": 4.5722747218396214e-05, "loss": 0.031, "step": 5510 }, { "grad_norm": 0.43762657046318054, "learning_rate": 4.5558030036492194e-05, "loss": 0.0341, "step": 5520 }, { "grad_norm": 0.687526524066925, "learning_rate": 4.539336143125999e-05, "loss": 0.0373, "step": 5530 }, { "grad_norm": 0.2577524483203888, "learning_rate": 4.522874320348916e-05, "loss": 0.0359, "step": 5540 }, { "grad_norm": 0.6617255210876465, "learning_rate": 4.506417715341821e-05, "loss": 0.0296, "step": 5550 }, { "grad_norm": 0.33408111333847046, "learning_rate": 4.489966508071511e-05, "loss": 0.0348, "step": 5560 }, { "grad_norm": 0.4379657804965973, "learning_rate": 4.4735208784457575e-05, "loss": 0.0345, "step": 5570 }, { "grad_norm": 0.373502254486084, "learning_rate": 4.457081006311325e-05, "loss": 0.0312, "step": 5580 }, { "grad_norm": 0.2966044843196869, "learning_rate": 4.440647071452027e-05, "loss": 0.028, "step": 5590 }, { "grad_norm": 0.38492828607559204, "learning_rate": 4.424219253586737e-05, "loss": 0.0352, "step": 5600 }, { "grad_norm": 0.34989821910858154, "learning_rate": 4.407797732367443e-05, "loss": 0.0289, "step": 5610 }, { "grad_norm": 0.287028431892395, "learning_rate": 4.391382687377268e-05, "loss": 0.0278, "step": 5620 }, { "grad_norm": 0.3059958517551422, "learning_rate": 4.374974298128512e-05, "loss": 0.0344, "step": 5630 }, { "grad_norm": 0.25095704197883606, "learning_rate": 4.358572744060699e-05, "loss": 0.0299, "step": 5640 }, { "grad_norm": 0.38532835245132446, "learning_rate": 4.342178204538588e-05, "loss": 0.0245, "step": 5650 }, { "grad_norm": 0.38667604327201843, "learning_rate": 4.325790858850241e-05, "loss": 0.0313, "step": 5660 }, { "grad_norm": 0.24977676570415497, "learning_rate": 4.309410886205043e-05, "loss": 0.0271, "step": 5670 }, { "grad_norm": 0.2900142967700958, "learning_rate": 4.293038465731752e-05, "loss": 0.0344, "step": 5680 }, { "grad_norm": 0.3448508381843567, "learning_rate": 4.276673776476533e-05, "loss": 0.0312, "step": 5690 }, { "grad_norm": 0.8138441443443298, "learning_rate": 4.260316997401007e-05, "loss": 0.0288, "step": 5700 }, { "grad_norm": 0.25529804825782776, "learning_rate": 4.243968307380293e-05, "loss": 0.0281, "step": 5710 }, { "grad_norm": 0.38965433835983276, "learning_rate": 4.22762788520104e-05, "loss": 0.0305, "step": 5720 }, { "grad_norm": 0.7432264685630798, "learning_rate": 4.211295909559491e-05, "loss": 0.0293, "step": 5730 }, { "grad_norm": 0.26739630103111267, "learning_rate": 4.194972559059511e-05, "loss": 0.0324, "step": 5740 }, { "grad_norm": 0.5407295823097229, "learning_rate": 4.178658012210651e-05, "loss": 0.0377, "step": 5750 }, { "grad_norm": 1.8566784858703613, "learning_rate": 4.162352447426177e-05, "loss": 0.0319, "step": 5760 }, { "grad_norm": 0.39516475796699524, "learning_rate": 4.146056043021135e-05, "loss": 0.0298, "step": 5770 }, { "grad_norm": 0.25355854630470276, "learning_rate": 4.1297689772103944e-05, "loss": 0.026, "step": 5780 }, { "grad_norm": 0.33691245317459106, "learning_rate": 4.113491428106694e-05, "loss": 0.0262, "step": 5790 }, { "grad_norm": 0.3872177004814148, "learning_rate": 4.0972235737187055e-05, "loss": 0.0268, "step": 5800 }, { "grad_norm": 0.44735443592071533, "learning_rate": 4.080965591949076e-05, "loss": 0.026, "step": 5810 }, { "grad_norm": 0.3078666627407074, "learning_rate": 4.0647176605924924e-05, "loss": 0.0245, "step": 5820 }, { "grad_norm": 0.36718639731407166, "learning_rate": 4.0484799573337255e-05, "loss": 0.0305, "step": 5830 }, { "grad_norm": 0.3599289357662201, "learning_rate": 4.032252659745699e-05, "loss": 0.026, "step": 5840 }, { "grad_norm": 0.3766120672225952, "learning_rate": 4.016035945287539e-05, "loss": 0.0327, "step": 5850 }, { "grad_norm": 0.3527063727378845, "learning_rate": 3.999829991302635e-05, "loss": 0.0236, "step": 5860 }, { "grad_norm": 0.2938133776187897, "learning_rate": 3.983634975016707e-05, "loss": 0.0267, "step": 5870 }, { "grad_norm": 0.5224175453186035, "learning_rate": 3.967451073535854e-05, "loss": 0.0302, "step": 5880 }, { "grad_norm": 0.34892210364341736, "learning_rate": 3.951278463844633e-05, "loss": 0.0236, "step": 5890 }, { "grad_norm": 0.6942933201789856, "learning_rate": 3.935117322804111e-05, "loss": 0.0267, "step": 5900 }, { "grad_norm": 0.2897878587245941, "learning_rate": 3.918967827149938e-05, "loss": 0.0318, "step": 5910 }, { "grad_norm": 0.3707078993320465, "learning_rate": 3.9028301534904094e-05, "loss": 0.031, "step": 5920 }, { "grad_norm": 0.3392997682094574, "learning_rate": 3.88670447830454e-05, "loss": 0.0297, "step": 5930 }, { "grad_norm": 0.6112711429595947, "learning_rate": 3.870590977940132e-05, "loss": 0.0272, "step": 5940 }, { "grad_norm": 0.286777138710022, "learning_rate": 3.8544898286118404e-05, "loss": 0.0259, "step": 5950 }, { "grad_norm": 0.31893420219421387, "learning_rate": 3.838401206399257e-05, "loss": 0.0333, "step": 5960 }, { "grad_norm": 0.5282262563705444, "learning_rate": 3.822325287244975e-05, "loss": 0.0349, "step": 5970 }, { "grad_norm": 0.4044519066810608, "learning_rate": 3.8062622469526725e-05, "loss": 0.0364, "step": 5980 }, { "grad_norm": 0.42226627469062805, "learning_rate": 3.790212261185183e-05, "loss": 0.034, "step": 5990 }, { "grad_norm": 1.5693583488464355, "learning_rate": 3.7741755054625794e-05, "loss": 0.0294, "step": 6000 }, { "grad_norm": 0.48320475220680237, "learning_rate": 3.758152155160255e-05, "loss": 0.028, "step": 6010 }, { "grad_norm": 0.44498103857040405, "learning_rate": 3.742142385506999e-05, "loss": 0.0333, "step": 6020 }, { "grad_norm": 0.637031078338623, "learning_rate": 3.72614637158309e-05, "loss": 0.0273, "step": 6030 }, { "grad_norm": 0.2873016893863678, "learning_rate": 3.710164288318371e-05, "loss": 0.0273, "step": 6040 }, { "grad_norm": 0.8845828175544739, "learning_rate": 3.694196310490345e-05, "loss": 0.0348, "step": 6050 }, { "grad_norm": 0.6289351582527161, "learning_rate": 3.678242612722259e-05, "loss": 0.0242, "step": 6060 }, { "grad_norm": 0.3801291584968567, "learning_rate": 3.6623033694811953e-05, "loss": 0.031, "step": 6070 }, { "grad_norm": 0.3094252645969391, "learning_rate": 3.6463787550761665e-05, "loss": 0.024, "step": 6080 }, { "grad_norm": 0.35827335715293884, "learning_rate": 3.630468943656202e-05, "loss": 0.0241, "step": 6090 }, { "grad_norm": 0.652944803237915, "learning_rate": 3.6145741092084523e-05, "loss": 0.0272, "step": 6100 }, { "grad_norm": 0.35197532176971436, "learning_rate": 3.598694425556278e-05, "loss": 0.0323, "step": 6110 }, { "grad_norm": 0.8375176787376404, "learning_rate": 3.58283006635736e-05, "loss": 0.0326, "step": 6120 }, { "grad_norm": 0.6728918552398682, "learning_rate": 3.566981205101781e-05, "loss": 0.038, "step": 6130 }, { "grad_norm": 0.2188553512096405, "learning_rate": 3.5511480151101556e-05, "loss": 0.027, "step": 6140 }, { "grad_norm": 0.350041002035141, "learning_rate": 3.5353306695317104e-05, "loss": 0.029, "step": 6150 }, { "grad_norm": 0.5239511728286743, "learning_rate": 3.519529341342402e-05, "loss": 0.0274, "step": 6160 }, { "grad_norm": 0.7617340087890625, "learning_rate": 3.503744203343026e-05, "loss": 0.0287, "step": 6170 }, { "grad_norm": 0.566163182258606, "learning_rate": 3.487975428157318e-05, "loss": 0.0303, "step": 6180 }, { "grad_norm": 0.5162640810012817, "learning_rate": 3.472223188230083e-05, "loss": 0.0238, "step": 6190 }, { "grad_norm": 0.26109224557876587, "learning_rate": 3.4564876558252866e-05, "loss": 0.0285, "step": 6200 }, { "grad_norm": 0.4273248314857483, "learning_rate": 3.440769003024195e-05, "loss": 0.0342, "step": 6210 }, { "grad_norm": 0.3802716135978699, "learning_rate": 3.425067401723477e-05, "loss": 0.0303, "step": 6220 }, { "grad_norm": 0.3939044177532196, "learning_rate": 3.409383023633325e-05, "loss": 0.0362, "step": 6230 }, { "grad_norm": 0.41104981303215027, "learning_rate": 3.3937160402755894e-05, "loss": 0.0324, "step": 6240 }, { "grad_norm": 0.6202765107154846, "learning_rate": 3.378066622981885e-05, "loss": 0.0355, "step": 6250 }, { "grad_norm": 0.4759131073951721, "learning_rate": 3.362434942891738e-05, "loss": 0.0344, "step": 6260 }, { "grad_norm": 0.5869651436805725, "learning_rate": 3.346821170950693e-05, "loss": 0.0321, "step": 6270 }, { "grad_norm": 0.4614925980567932, "learning_rate": 3.3312254779084585e-05, "loss": 0.0286, "step": 6280 }, { "grad_norm": 0.44577866792678833, "learning_rate": 3.315648034317039e-05, "loss": 0.0315, "step": 6290 }, { "grad_norm": 0.23973332345485687, "learning_rate": 3.3000890105288564e-05, "loss": 0.0277, "step": 6300 }, { "grad_norm": 0.5095074772834778, "learning_rate": 3.284548576694908e-05, "loss": 0.0309, "step": 6310 }, { "grad_norm": 0.48157602548599243, "learning_rate": 3.2690269027628815e-05, "loss": 0.0261, "step": 6320 }, { "grad_norm": 0.41792169213294983, "learning_rate": 3.253524158475324e-05, "loss": 0.0422, "step": 6330 }, { "grad_norm": 0.38813546299934387, "learning_rate": 3.238040513367757e-05, "loss": 0.0341, "step": 6340 }, { "grad_norm": 0.6386147737503052, "learning_rate": 3.222576136766843e-05, "loss": 0.0331, "step": 6350 }, { "grad_norm": 0.31393110752105713, "learning_rate": 3.2071311977885324e-05, "loss": 0.0227, "step": 6360 }, { "grad_norm": 0.3807489573955536, "learning_rate": 3.191705865336197e-05, "loss": 0.0276, "step": 6370 }, { "grad_norm": 0.40429434180259705, "learning_rate": 3.1763003080988075e-05, "loss": 0.0268, "step": 6380 }, { "grad_norm": 0.2683643400669098, "learning_rate": 3.160914694549063e-05, "loss": 0.024, "step": 6390 }, { "grad_norm": 0.4885174632072449, "learning_rate": 3.145549192941573e-05, "loss": 0.0297, "step": 6400 }, { "grad_norm": 0.5081372261047363, "learning_rate": 3.130203971310999e-05, "loss": 0.0318, "step": 6410 }, { "grad_norm": 0.39594805240631104, "learning_rate": 3.114879197470225e-05, "loss": 0.026, "step": 6420 }, { "grad_norm": 0.36854487657546997, "learning_rate": 3.0995750390085285e-05, "loss": 0.0279, "step": 6430 }, { "grad_norm": 0.20717976987361908, "learning_rate": 3.084291663289728e-05, "loss": 0.0318, "step": 6440 }, { "grad_norm": 0.229153111577034, "learning_rate": 3.069029237450375e-05, "loss": 0.0264, "step": 6450 }, { "grad_norm": 0.42472395300865173, "learning_rate": 3.053787928397911e-05, "loss": 0.0316, "step": 6460 }, { "grad_norm": 0.26687610149383545, "learning_rate": 3.0385679028088526e-05, "loss": 0.0326, "step": 6470 }, { "grad_norm": 0.2529737651348114, "learning_rate": 3.023369327126959e-05, "loss": 0.026, "step": 6480 }, { "grad_norm": 0.30746614933013916, "learning_rate": 3.0081923675614198e-05, "loss": 0.0235, "step": 6490 }, { "grad_norm": 0.2654724717140198, "learning_rate": 2.993037190085034e-05, "loss": 0.0267, "step": 6500 }, { "grad_norm": 0.31287702918052673, "learning_rate": 2.977903960432392e-05, "loss": 0.0268, "step": 6510 }, { "grad_norm": 0.3534926474094391, "learning_rate": 2.9627928440980722e-05, "loss": 0.0262, "step": 6520 }, { "grad_norm": 0.33883389830589294, "learning_rate": 2.9477040063348183e-05, "loss": 0.0242, "step": 6530 }, { "grad_norm": 0.2922869324684143, "learning_rate": 2.9326376121517456e-05, "loss": 0.0279, "step": 6540 }, { "grad_norm": 0.35121944546699524, "learning_rate": 2.9175938263125236e-05, "loss": 0.0205, "step": 6550 }, { "grad_norm": 0.4071419835090637, "learning_rate": 2.9025728133335873e-05, "loss": 0.0219, "step": 6560 }, { "grad_norm": 0.18294961750507355, "learning_rate": 2.8875747374823288e-05, "loss": 0.023, "step": 6570 }, { "grad_norm": 0.6878488659858704, "learning_rate": 2.872599762775298e-05, "loss": 0.0257, "step": 6580 }, { "grad_norm": 0.2973591983318329, "learning_rate": 2.857648052976425e-05, "loss": 0.0217, "step": 6590 }, { "grad_norm": 0.31356048583984375, "learning_rate": 2.8427197715952047e-05, "loss": 0.0197, "step": 6600 }, { "grad_norm": 0.303670734167099, "learning_rate": 2.8278150818849393e-05, "loss": 0.0276, "step": 6610 }, { "grad_norm": 0.46224260330200195, "learning_rate": 2.812934146840922e-05, "loss": 0.0252, "step": 6620 }, { "grad_norm": 0.39281874895095825, "learning_rate": 2.7980771291986764e-05, "loss": 0.0267, "step": 6630 }, { "grad_norm": 0.5297064781188965, "learning_rate": 2.783244191432167e-05, "loss": 0.0259, "step": 6640 }, { "grad_norm": 0.5946605801582336, "learning_rate": 2.768435495752022e-05, "loss": 0.0279, "step": 6650 }, { "grad_norm": 0.3251868188381195, "learning_rate": 2.753651204103771e-05, "loss": 0.0337, "step": 6660 }, { "grad_norm": 0.5797733664512634, "learning_rate": 2.7388914781660523e-05, "loss": 0.0273, "step": 6670 }, { "grad_norm": 0.548480749130249, "learning_rate": 2.7241564793488693e-05, "loss": 0.0262, "step": 6680 }, { "grad_norm": 0.41308483481407166, "learning_rate": 2.7094463687918037e-05, "loss": 0.0253, "step": 6690 }, { "grad_norm": 0.3282982110977173, "learning_rate": 2.694761307362268e-05, "loss": 0.0254, "step": 6700 }, { "grad_norm": 0.43327638506889343, "learning_rate": 2.6801014556537467e-05, "loss": 0.032, "step": 6710 }, { "grad_norm": 0.5334053039550781, "learning_rate": 2.6654669739840243e-05, "loss": 0.028, "step": 6720 }, { "grad_norm": 0.6647353172302246, "learning_rate": 2.650858022393451e-05, "loss": 0.026, "step": 6730 }, { "grad_norm": 0.4013376832008362, "learning_rate": 2.6362747606431747e-05, "loss": 0.0246, "step": 6740 }, { "grad_norm": 0.3450195789337158, "learning_rate": 2.6217173482134172e-05, "loss": 0.0255, "step": 6750 }, { "grad_norm": 0.4422897696495056, "learning_rate": 2.6071859443017044e-05, "loss": 0.027, "step": 6760 }, { "grad_norm": 0.4555124044418335, "learning_rate": 2.5926807078211414e-05, "loss": 0.0302, "step": 6770 }, { "grad_norm": 0.7389348149299622, "learning_rate": 2.5782017973986728e-05, "loss": 0.0328, "step": 6780 }, { "grad_norm": 0.3251645863056183, "learning_rate": 2.5637493713733374e-05, "loss": 0.0262, "step": 6790 }, { "grad_norm": 0.5132068991661072, "learning_rate": 2.549323587794559e-05, "loss": 0.0314, "step": 6800 }, { "grad_norm": 0.22303754091262817, "learning_rate": 2.5349246044203895e-05, "loss": 0.0325, "step": 6810 }, { "grad_norm": 0.4640345573425293, "learning_rate": 2.520552578715808e-05, "loss": 0.0242, "step": 6820 }, { "grad_norm": 0.2909422516822815, "learning_rate": 2.506207667850981e-05, "loss": 0.0254, "step": 6830 }, { "grad_norm": 1.5867938995361328, "learning_rate": 2.4918900286995555e-05, "loss": 0.0299, "step": 6840 }, { "grad_norm": 1.7177354097366333, "learning_rate": 2.4775998178369458e-05, "loss": 0.028, "step": 6850 }, { "grad_norm": 0.26830941438674927, "learning_rate": 2.4633371915386017e-05, "loss": 0.0303, "step": 6860 }, { "grad_norm": 0.8601913452148438, "learning_rate": 2.4491023057783235e-05, "loss": 0.0321, "step": 6870 }, { "grad_norm": 0.44798222184181213, "learning_rate": 2.4348953162265375e-05, "loss": 0.0258, "step": 6880 }, { "grad_norm": 0.5396066308021545, "learning_rate": 2.420716378248607e-05, "loss": 0.0235, "step": 6890 }, { "grad_norm": 0.3758165240287781, "learning_rate": 2.4065656469031266e-05, "loss": 0.025, "step": 6900 }, { "grad_norm": 0.46708133816719055, "learning_rate": 2.3924432769402268e-05, "loss": 0.0365, "step": 6910 }, { "grad_norm": 0.4811571538448334, "learning_rate": 2.3783494227998844e-05, "loss": 0.027, "step": 6920 }, { "grad_norm": 0.5201545357704163, "learning_rate": 2.3642842386102264e-05, "loss": 0.0293, "step": 6930 }, { "grad_norm": 0.35595959424972534, "learning_rate": 2.3502478781858567e-05, "loss": 0.0281, "step": 6940 }, { "grad_norm": 0.9493057131767273, "learning_rate": 2.3362404950261628e-05, "loss": 0.0318, "step": 6950 }, { "grad_norm": 0.42703843116760254, "learning_rate": 2.3222622423136458e-05, "loss": 0.0233, "step": 6960 }, { "grad_norm": 0.3913417160511017, "learning_rate": 2.3083132729122332e-05, "loss": 0.0296, "step": 6970 }, { "grad_norm": 0.5639697909355164, "learning_rate": 2.294393739365621e-05, "loss": 0.0237, "step": 6980 }, { "grad_norm": 0.40071967244148254, "learning_rate": 2.2805037938956e-05, "loss": 0.0245, "step": 6990 }, { "grad_norm": 0.31974515318870544, "learning_rate": 2.266643588400386e-05, "loss": 0.0322, "step": 7000 }, { "grad_norm": 0.30382561683654785, "learning_rate": 2.252813274452969e-05, "loss": 0.0213, "step": 7010 }, { "grad_norm": 0.38035860657691956, "learning_rate": 2.2390130032994427e-05, "loss": 0.0375, "step": 7020 }, { "grad_norm": 0.4806108772754669, "learning_rate": 2.2252429258573633e-05, "loss": 0.0238, "step": 7030 }, { "grad_norm": 0.43503281474113464, "learning_rate": 2.2115031927140904e-05, "loss": 0.0262, "step": 7040 }, { "grad_norm": 0.29157617688179016, "learning_rate": 2.1977939541251463e-05, "loss": 0.0215, "step": 7050 }, { "grad_norm": 0.3817867934703827, "learning_rate": 2.1841153600125684e-05, "loss": 0.031, "step": 7060 }, { "grad_norm": 0.5023079514503479, "learning_rate": 2.170467559963267e-05, "loss": 0.0317, "step": 7070 }, { "grad_norm": 0.7835108041763306, "learning_rate": 2.1568507032273982e-05, "loss": 0.0303, "step": 7080 }, { "grad_norm": 0.4332734942436218, "learning_rate": 2.1432649387167264e-05, "loss": 0.027, "step": 7090 }, { "grad_norm": 0.5897580981254578, "learning_rate": 2.1297104150029973e-05, "loss": 0.0255, "step": 7100 }, { "grad_norm": 0.27596065402030945, "learning_rate": 2.116187280316307e-05, "loss": 0.0289, "step": 7110 }, { "grad_norm": 0.3004196882247925, "learning_rate": 2.1026956825434908e-05, "loss": 0.0244, "step": 7120 }, { "grad_norm": 0.41873565316200256, "learning_rate": 2.0892357692265017e-05, "loss": 0.0332, "step": 7130 }, { "grad_norm": 0.5549617409706116, "learning_rate": 2.0758076875607947e-05, "loss": 0.0264, "step": 7140 }, { "grad_norm": 0.47440993785858154, "learning_rate": 2.0624115843937207e-05, "loss": 0.0303, "step": 7150 }, { "grad_norm": 0.24260501563549042, "learning_rate": 2.0490476062229157e-05, "loss": 0.0212, "step": 7160 }, { "grad_norm": 0.2648763954639435, "learning_rate": 2.035715899194704e-05, "loss": 0.0295, "step": 7170 }, { "grad_norm": 0.21949419379234314, "learning_rate": 2.022416609102499e-05, "loss": 0.0229, "step": 7180 }, { "grad_norm": 0.3179526925086975, "learning_rate": 2.009149881385205e-05, "loss": 0.0264, "step": 7190 }, { "grad_norm": 0.26673057675361633, "learning_rate": 1.995915861125634e-05, "loss": 0.0301, "step": 7200 }, { "grad_norm": 0.37167781591415405, "learning_rate": 1.9827146930489065e-05, "loss": 0.0233, "step": 7210 }, { "grad_norm": 0.2561936676502228, "learning_rate": 1.9695465215208848e-05, "loss": 0.0227, "step": 7220 }, { "grad_norm": 0.3161408007144928, "learning_rate": 1.9564114905465813e-05, "loss": 0.0232, "step": 7230 }, { "grad_norm": 0.19553589820861816, "learning_rate": 1.9433097437685936e-05, "loss": 0.0303, "step": 7240 }, { "grad_norm": 0.3304789066314697, "learning_rate": 1.930241424465521e-05, "loss": 0.0236, "step": 7250 }, { "grad_norm": 0.41526198387145996, "learning_rate": 1.9172066755504115e-05, "loss": 0.0239, "step": 7260 }, { "grad_norm": 0.2929757535457611, "learning_rate": 1.9042056395691914e-05, "loss": 0.0322, "step": 7270 }, { "grad_norm": 0.43654054403305054, "learning_rate": 1.8912384586991066e-05, "loss": 0.0314, "step": 7280 }, { "grad_norm": 0.44934678077697754, "learning_rate": 1.8783052747471717e-05, "loss": 0.0312, "step": 7290 }, { "grad_norm": 0.3265818655490875, "learning_rate": 1.865406229148611e-05, "loss": 0.0242, "step": 7300 }, { "grad_norm": 0.33041590452194214, "learning_rate": 1.8525414629653233e-05, "loss": 0.03, "step": 7310 }, { "grad_norm": 0.3144224286079407, "learning_rate": 1.8397111168843255e-05, "loss": 0.0252, "step": 7320 }, { "grad_norm": 0.28946441411972046, "learning_rate": 1.8269153312162323e-05, "loss": 0.0192, "step": 7330 }, { "grad_norm": 0.4116132855415344, "learning_rate": 1.8141542458937054e-05, "loss": 0.0234, "step": 7340 }, { "grad_norm": 0.4250207543373108, "learning_rate": 1.8014280004699268e-05, "loss": 0.0233, "step": 7350 }, { "grad_norm": 0.6355520486831665, "learning_rate": 1.788736734117078e-05, "loss": 0.0248, "step": 7360 }, { "grad_norm": 0.3089202642440796, "learning_rate": 1.7760805856248152e-05, "loss": 0.0219, "step": 7370 }, { "grad_norm": 0.38890331983566284, "learning_rate": 1.7634596933987518e-05, "loss": 0.0234, "step": 7380 }, { "grad_norm": 0.6232661604881287, "learning_rate": 1.7508741954589404e-05, "loss": 0.0224, "step": 7390 }, { "grad_norm": 0.8616970777511597, "learning_rate": 1.7383242294383717e-05, "loss": 0.0243, "step": 7400 }, { "grad_norm": 0.43286487460136414, "learning_rate": 1.7258099325814632e-05, "loss": 0.0265, "step": 7410 }, { "grad_norm": 0.8192644119262695, "learning_rate": 1.7133314417425594e-05, "loss": 0.0308, "step": 7420 }, { "grad_norm": 0.23756588995456696, "learning_rate": 1.7008888933844408e-05, "loss": 0.0276, "step": 7430 }, { "grad_norm": 0.6211453676223755, "learning_rate": 1.6884824235768172e-05, "loss": 0.0229, "step": 7440 }, { "grad_norm": 0.34216466546058655, "learning_rate": 1.6761121679948592e-05, "loss": 0.0309, "step": 7450 }, { "grad_norm": 0.3565342128276825, "learning_rate": 1.663778261917695e-05, "loss": 0.0243, "step": 7460 }, { "grad_norm": 0.2935832142829895, "learning_rate": 1.651480840226952e-05, "loss": 0.0363, "step": 7470 }, { "grad_norm": 0.21819259226322174, "learning_rate": 1.639220037405258e-05, "loss": 0.0283, "step": 7480 }, { "grad_norm": 0.3889438807964325, "learning_rate": 1.6269959875347906e-05, "loss": 0.0258, "step": 7490 }, { "grad_norm": 0.6047902703285217, "learning_rate": 1.614808824295802e-05, "loss": 0.0304, "step": 7500 }, { "grad_norm": 0.4127240478992462, "learning_rate": 1.602658680965152e-05, "loss": 0.0255, "step": 7510 }, { "grad_norm": 0.3865024745464325, "learning_rate": 1.5905456904148686e-05, "loss": 0.0242, "step": 7520 }, { "grad_norm": 0.2503134310245514, "learning_rate": 1.57846998511067e-05, "loss": 0.0249, "step": 7530 }, { "grad_norm": 1.0493054389953613, "learning_rate": 1.566431697110538e-05, "loss": 0.0291, "step": 7540 }, { "grad_norm": 0.46878376603126526, "learning_rate": 1.554430958063259e-05, "loss": 0.0335, "step": 7550 }, { "grad_norm": 0.3137737214565277, "learning_rate": 1.5424678992069912e-05, "loss": 0.0259, "step": 7560 }, { "grad_norm": 0.25972068309783936, "learning_rate": 1.5305426513678362e-05, "loss": 0.0284, "step": 7570 }, { "grad_norm": 0.7701228260993958, "learning_rate": 1.518655344958388e-05, "loss": 0.0263, "step": 7580 }, { "grad_norm": 0.476356565952301, "learning_rate": 1.5068061099763275e-05, "loss": 0.0219, "step": 7590 }, { "grad_norm": 0.35269349813461304, "learning_rate": 1.494995076002988e-05, "loss": 0.0234, "step": 7600 }, { "grad_norm": 0.4179874062538147, "learning_rate": 1.4832223722019456e-05, "loss": 0.0209, "step": 7610 }, { "grad_norm": 0.39547452330589294, "learning_rate": 1.4714881273176035e-05, "loss": 0.0268, "step": 7620 }, { "grad_norm": 0.5685515403747559, "learning_rate": 1.4597924696737835e-05, "loss": 0.0293, "step": 7630 }, { "grad_norm": 0.5017313361167908, "learning_rate": 1.4481355271723252e-05, "loss": 0.0231, "step": 7640 }, { "grad_norm": 0.2977704703807831, "learning_rate": 1.4365174272916809e-05, "loss": 0.0225, "step": 7650 }, { "grad_norm": 0.22654534876346588, "learning_rate": 1.4249382970855319e-05, "loss": 0.0254, "step": 7660 }, { "grad_norm": 0.2606799900531769, "learning_rate": 1.4133982631813903e-05, "loss": 0.0219, "step": 7670 }, { "grad_norm": 0.30454108119010925, "learning_rate": 1.4018974517792194e-05, "loss": 0.0193, "step": 7680 }, { "grad_norm": 0.21303042769432068, "learning_rate": 1.390435988650048e-05, "loss": 0.0222, "step": 7690 }, { "grad_norm": 0.32101476192474365, "learning_rate": 1.3790139991346006e-05, "loss": 0.0222, "step": 7700 }, { "grad_norm": 0.2826872169971466, "learning_rate": 1.367631608141926e-05, "loss": 0.0246, "step": 7710 }, { "grad_norm": 0.99654620885849, "learning_rate": 1.3562889401480278e-05, "loss": 0.0229, "step": 7720 }, { "grad_norm": 0.2744136154651642, "learning_rate": 1.3449861191945074e-05, "loss": 0.0292, "step": 7730 }, { "grad_norm": 0.42959925532341003, "learning_rate": 1.3337232688872009e-05, "loss": 0.0245, "step": 7740 }, { "grad_norm": 0.23847398161888123, "learning_rate": 1.3225005123948364e-05, "loss": 0.0189, "step": 7750 }, { "grad_norm": 0.6085127592086792, "learning_rate": 1.311317972447681e-05, "loss": 0.0231, "step": 7760 }, { "grad_norm": 0.36748650670051575, "learning_rate": 1.3001757713361996e-05, "loss": 0.025, "step": 7770 }, { "grad_norm": 0.22233383357524872, "learning_rate": 1.2890740309097204e-05, "loss": 0.0213, "step": 7780 }, { "grad_norm": 0.5203900933265686, "learning_rate": 1.2780128725750944e-05, "loss": 0.0239, "step": 7790 }, { "grad_norm": 1.0815848112106323, "learning_rate": 1.266992417295379e-05, "loss": 0.0264, "step": 7800 }, { "grad_norm": 0.37504705786705017, "learning_rate": 1.2560127855885073e-05, "loss": 0.0226, "step": 7810 }, { "grad_norm": 0.28880059719085693, "learning_rate": 1.2450740975259745e-05, "loss": 0.0257, "step": 7820 }, { "grad_norm": 0.23325610160827637, "learning_rate": 1.234176472731517e-05, "loss": 0.0232, "step": 7830 }, { "grad_norm": 0.26797088980674744, "learning_rate": 1.2233200303798158e-05, "loss": 0.0182, "step": 7840 }, { "grad_norm": 0.5015556216239929, "learning_rate": 1.2125048891951846e-05, "loss": 0.0201, "step": 7850 }, { "grad_norm": 0.23454511165618896, "learning_rate": 1.2017311674502745e-05, "loss": 0.03, "step": 7860 }, { "grad_norm": 1.1543573141098022, "learning_rate": 1.1909989829647822e-05, "loss": 0.0262, "step": 7870 }, { "grad_norm": 0.27224382758140564, "learning_rate": 1.1803084531041553e-05, "loss": 0.0184, "step": 7880 }, { "grad_norm": 0.30252301692962646, "learning_rate": 1.1696596947783162e-05, "loss": 0.0238, "step": 7890 }, { "grad_norm": 0.5748562812805176, "learning_rate": 1.1590528244403803e-05, "loss": 0.02, "step": 7900 }, { "grad_norm": 0.30584320425987244, "learning_rate": 1.148487958085382e-05, "loss": 0.0224, "step": 7910 }, { "grad_norm": 0.6170525550842285, "learning_rate": 1.1379652112490086e-05, "loss": 0.0213, "step": 7920 }, { "grad_norm": 0.29798707365989685, "learning_rate": 1.1274846990063315e-05, "loss": 0.0255, "step": 7930 }, { "grad_norm": 0.30941957235336304, "learning_rate": 1.117046535970554e-05, "loss": 0.0264, "step": 7940 }, { "grad_norm": 0.41302111744880676, "learning_rate": 1.106650836291755e-05, "loss": 0.0232, "step": 7950 }, { "grad_norm": 0.25537002086639404, "learning_rate": 1.0962977136556418e-05, "loss": 0.037, "step": 7960 }, { "grad_norm": 0.22938591241836548, "learning_rate": 1.0859872812823024e-05, "loss": 0.0211, "step": 7970 }, { "grad_norm": 0.3744678795337677, "learning_rate": 1.0757196519249747e-05, "loss": 0.0227, "step": 7980 }, { "grad_norm": 0.30523186922073364, "learning_rate": 1.0654949378688077e-05, "loss": 0.0188, "step": 7990 }, { "grad_norm": 0.20380137860774994, "learning_rate": 1.0553132509296376e-05, "loss": 0.0216, "step": 8000 }, { "grad_norm": 0.15640057623386383, "learning_rate": 1.0451747024527613e-05, "loss": 0.0188, "step": 8010 }, { "grad_norm": 0.43110358715057373, "learning_rate": 1.0350794033117189e-05, "loss": 0.0242, "step": 8020 }, { "grad_norm": 1.2533138990402222, "learning_rate": 1.0250274639070856e-05, "loss": 0.0252, "step": 8030 }, { "grad_norm": 0.33290621638298035, "learning_rate": 1.0150189941652599e-05, "loss": 0.0236, "step": 8040 }, { "grad_norm": 0.545903742313385, "learning_rate": 1.0050541035372635e-05, "loss": 0.0281, "step": 8050 }, { "grad_norm": 0.3848140239715576, "learning_rate": 9.951329009975458e-06, "loss": 0.0213, "step": 8060 }, { "grad_norm": 0.34551799297332764, "learning_rate": 9.852554950427845e-06, "loss": 0.0267, "step": 8070 }, { "grad_norm": 1.2099828720092773, "learning_rate": 9.754219936907105e-06, "loss": 0.0182, "step": 8080 }, { "grad_norm": 0.3048720061779022, "learning_rate": 9.656325044789194e-06, "loss": 0.0175, "step": 8090 }, { "grad_norm": 0.20810218155384064, "learning_rate": 9.55887134463697e-06, "loss": 0.0219, "step": 8100 }, { "grad_norm": 0.4905526340007782, "learning_rate": 9.461859902188475e-06, "loss": 0.0284, "step": 8110 }, { "grad_norm": 0.48985356092453003, "learning_rate": 9.365291778345303e-06, "loss": 0.0189, "step": 8120 }, { "grad_norm": 0.2609074115753174, "learning_rate": 9.269168029160991e-06, "loss": 0.0217, "step": 8130 }, { "grad_norm": 0.23836171627044678, "learning_rate": 9.173489705829447e-06, "loss": 0.0238, "step": 8140 }, { "grad_norm": 0.2788395881652832, "learning_rate": 9.078257854673516e-06, "loss": 0.0214, "step": 8150 }, { "grad_norm": 0.36261677742004395, "learning_rate": 8.983473517133429e-06, "loss": 0.022, "step": 8160 }, { "grad_norm": 0.2682543098926544, "learning_rate": 8.889137729755537e-06, "loss": 0.0238, "step": 8170 }, { "grad_norm": 0.26575934886932373, "learning_rate": 8.79525152418087e-06, "loss": 0.0237, "step": 8180 }, { "grad_norm": 0.3463025987148285, "learning_rate": 8.701815927133961e-06, "loss": 0.025, "step": 8190 }, { "grad_norm": 0.21904657781124115, "learning_rate": 8.608831960411534e-06, "loss": 0.0185, "step": 8200 }, { "grad_norm": 0.44770997762680054, "learning_rate": 8.516300640871321e-06, "loss": 0.0276, "step": 8210 }, { "grad_norm": 0.17561765015125275, "learning_rate": 8.424222980421038e-06, "loss": 0.0263, "step": 8220 }, { "grad_norm": 0.2333260327577591, "learning_rate": 8.332599986007184e-06, "loss": 0.0238, "step": 8230 }, { "grad_norm": 0.3016291856765747, "learning_rate": 8.241432659604203e-06, "loss": 0.0242, "step": 8240 }, { "grad_norm": 0.3232913613319397, "learning_rate": 8.150721998203331e-06, "loss": 0.0293, "step": 8250 }, { "grad_norm": 0.1869889348745346, "learning_rate": 8.06046899380184e-06, "loss": 0.0235, "step": 8260 }, { "grad_norm": 0.23543646931648254, "learning_rate": 7.970674633392133e-06, "loss": 0.0234, "step": 8270 }, { "grad_norm": 0.3160131871700287, "learning_rate": 7.881339898950924e-06, "loss": 0.0253, "step": 8280 }, { "grad_norm": 0.8385632634162903, "learning_rate": 7.792465767428597e-06, "loss": 0.0207, "step": 8290 }, { "grad_norm": 0.18377253413200378, "learning_rate": 7.704053210738376e-06, "loss": 0.0185, "step": 8300 }, { "grad_norm": 0.21611380577087402, "learning_rate": 7.6161031957458494e-06, "loss": 0.0239, "step": 8310 }, { "grad_norm": 0.5535539388656616, "learning_rate": 7.5286166842582605e-06, "loss": 0.0218, "step": 8320 }, { "grad_norm": 0.30015742778778076, "learning_rate": 7.4415946330140814e-06, "loss": 0.0327, "step": 8330 }, { "grad_norm": 0.1893400400876999, "learning_rate": 7.3550379936725644e-06, "loss": 0.0218, "step": 8340 }, { "grad_norm": 0.38621601462364197, "learning_rate": 7.2689477128032035e-06, "loss": 0.0207, "step": 8350 }, { "grad_norm": 0.25465279817581177, "learning_rate": 7.183324731875551e-06, "loss": 0.0243, "step": 8360 }, { "grad_norm": 0.45982834696769714, "learning_rate": 7.098169987248782e-06, "loss": 0.0239, "step": 8370 }, { "grad_norm": 0.4492223560810089, "learning_rate": 7.013484410161553e-06, "loss": 0.0326, "step": 8380 }, { "grad_norm": 0.4831450283527374, "learning_rate": 6.92926892672176e-06, "loss": 0.0243, "step": 8390 }, { "grad_norm": 0.29467639327049255, "learning_rate": 6.845524457896446e-06, "loss": 0.0201, "step": 8400 }, { "grad_norm": 0.30270370841026306, "learning_rate": 6.7622519195017165e-06, "loss": 0.0256, "step": 8410 }, { "grad_norm": 0.33148401975631714, "learning_rate": 6.679452222192684e-06, "loss": 0.02, "step": 8420 }, { "grad_norm": 1.1087185144424438, "learning_rate": 6.597126271453579e-06, "loss": 0.0244, "step": 8430 }, { "grad_norm": 0.4420481324195862, "learning_rate": 6.51527496758782e-06, "loss": 0.0234, "step": 8440 }, { "grad_norm": 0.6271085739135742, "learning_rate": 6.433899205708155e-06, "loss": 0.026, "step": 8450 }, { "grad_norm": 0.3207818567752838, "learning_rate": 6.352999875726856e-06, "loss": 0.0227, "step": 8460 }, { "grad_norm": 0.25023460388183594, "learning_rate": 6.272577862346052e-06, "loss": 0.0218, "step": 8470 }, { "grad_norm": 1.022682785987854, "learning_rate": 6.192634045047996e-06, "loss": 0.0239, "step": 8480 }, { "grad_norm": 0.4599818289279938, "learning_rate": 6.113169298085458e-06, "loss": 0.0252, "step": 8490 }, { "grad_norm": 0.3383421301841736, "learning_rate": 6.034184490472195e-06, "loss": 0.0217, "step": 8500 }, { "grad_norm": 0.5999664068222046, "learning_rate": 5.955680485973386e-06, "loss": 0.0217, "step": 8510 }, { "grad_norm": 0.4964524507522583, "learning_rate": 5.877658143096265e-06, "loss": 0.0221, "step": 8520 }, { "grad_norm": 0.319999635219574, "learning_rate": 5.800118315080661e-06, "loss": 0.0236, "step": 8530 }, { "grad_norm": 0.3427635431289673, "learning_rate": 5.723061849889716e-06, "loss": 0.023, "step": 8540 }, { "grad_norm": 0.33376508951187134, "learning_rate": 5.646489590200604e-06, "loss": 0.0237, "step": 8550 }, { "grad_norm": 0.27630865573883057, "learning_rate": 5.570402373395256e-06, "loss": 0.0228, "step": 8560 }, { "grad_norm": 0.30539679527282715, "learning_rate": 5.494801031551305e-06, "loss": 0.0237, "step": 8570 }, { "grad_norm": 0.6860610842704773, "learning_rate": 5.41968639143291e-06, "loss": 0.031, "step": 8580 }, { "grad_norm": 0.48348191380500793, "learning_rate": 5.345059274481751e-06, "loss": 0.0202, "step": 8590 }, { "grad_norm": 0.30793142318725586, "learning_rate": 5.270920496808002e-06, "loss": 0.0255, "step": 8600 }, { "grad_norm": 0.3083607256412506, "learning_rate": 5.1972708691814695e-06, "loss": 0.0242, "step": 8610 }, { "grad_norm": 0.2135923057794571, "learning_rate": 5.124111197022674e-06, "loss": 0.0222, "step": 8620 }, { "grad_norm": 0.32382744550704956, "learning_rate": 5.051442280394081e-06, "loss": 0.0232, "step": 8630 }, { "grad_norm": 1.2330256700515747, "learning_rate": 4.979264913991322e-06, "loss": 0.0228, "step": 8640 }, { "grad_norm": 0.26222866773605347, "learning_rate": 4.907579887134489e-06, "loss": 0.023, "step": 8650 }, { "grad_norm": 0.2501695156097412, "learning_rate": 4.836387983759572e-06, "loss": 0.02, "step": 8660 }, { "grad_norm": 0.19905517995357513, "learning_rate": 4.765689982409816e-06, "loss": 0.0207, "step": 8670 }, { "grad_norm": 0.6085242629051208, "learning_rate": 4.695486656227233e-06, "loss": 0.0203, "step": 8680 }, { "grad_norm": 0.29833275079727173, "learning_rate": 4.625778772944156e-06, "loss": 0.0235, "step": 8690 }, { "grad_norm": 0.17381176352500916, "learning_rate": 4.556567094874825e-06, "loss": 0.0158, "step": 8700 }, { "grad_norm": 0.4643753170967102, "learning_rate": 4.487852378907059e-06, "loss": 0.0209, "step": 8710 }, { "grad_norm": 0.4169240891933441, "learning_rate": 4.419635376493986e-06, "loss": 0.0253, "step": 8720 }, { "grad_norm": 0.3325566053390503, "learning_rate": 4.351916833645825e-06, "loss": 0.0208, "step": 8730 }, { "grad_norm": 0.9973950982093811, "learning_rate": 4.284697490921691e-06, "loss": 0.0224, "step": 8740 }, { "grad_norm": 0.32427507638931274, "learning_rate": 4.2179780834215585e-06, "loss": 0.0139, "step": 8750 }, { "grad_norm": 0.2685390114784241, "learning_rate": 4.151759340778178e-06, "loss": 0.0264, "step": 8760 }, { "grad_norm": 0.3046257197856903, "learning_rate": 4.086041987149109e-06, "loss": 0.0188, "step": 8770 }, { "grad_norm": 0.8612890839576721, "learning_rate": 4.020826741208811e-06, "loss": 0.0221, "step": 8780 }, { "grad_norm": 0.15918666124343872, "learning_rate": 3.956114316140746e-06, "loss": 0.0184, "step": 8790 }, { "grad_norm": 0.23081634938716888, "learning_rate": 3.891905419629643e-06, "loss": 0.0183, "step": 8800 }, { "grad_norm": 1.370868444442749, "learning_rate": 3.8282007538536946e-06, "loss": 0.0209, "step": 8810 }, { "grad_norm": 0.6476850509643555, "learning_rate": 3.7650010154769265e-06, "loss": 0.0249, "step": 8820 }, { "grad_norm": 0.3161841332912445, "learning_rate": 3.7023068956415608e-06, "loss": 0.0216, "step": 8830 }, { "grad_norm": 0.40090954303741455, "learning_rate": 3.6401190799604303e-06, "loss": 0.0312, "step": 8840 }, { "grad_norm": 0.37668895721435547, "learning_rate": 3.578438248509536e-06, "loss": 0.0204, "step": 8850 }, { "grad_norm": 0.30760154128074646, "learning_rate": 3.5172650758205583e-06, "loss": 0.0189, "step": 8860 }, { "grad_norm": 0.5577331781387329, "learning_rate": 3.45660023087353e-06, "loss": 0.0242, "step": 8870 }, { "grad_norm": 1.0352036952972412, "learning_rate": 3.3964443770894528e-06, "loss": 0.0204, "step": 8880 }, { "grad_norm": 0.7902668118476868, "learning_rate": 3.3367981723231245e-06, "loss": 0.0186, "step": 8890 }, { "grad_norm": 0.3904038071632385, "learning_rate": 3.2776622688558746e-06, "loss": 0.0195, "step": 8900 }, { "grad_norm": 0.6345455050468445, "learning_rate": 3.2190373133884677e-06, "loss": 0.0232, "step": 8910 }, { "grad_norm": 0.2836427092552185, "learning_rate": 3.1609239470340446e-06, "loss": 0.0192, "step": 8920 }, { "grad_norm": 1.1556655168533325, "learning_rate": 3.1033228053110373e-06, "loss": 0.0301, "step": 8930 }, { "grad_norm": 0.19598965346813202, "learning_rate": 3.0462345181363314e-06, "loss": 0.0222, "step": 8940 }, { "grad_norm": 1.738374948501587, "learning_rate": 2.9896597098182654e-06, "loss": 0.025, "step": 8950 }, { "grad_norm": 0.5447664856910706, "learning_rate": 2.933598999049891e-06, "loss": 0.0253, "step": 8960 }, { "grad_norm": 1.4435174465179443, "learning_rate": 2.8780529989021697e-06, "loss": 0.0228, "step": 8970 }, { "grad_norm": 0.8878384232521057, "learning_rate": 2.823022316817242e-06, "loss": 0.0225, "step": 8980 }, { "grad_norm": 0.38206616044044495, "learning_rate": 2.7685075546018456e-06, "loss": 0.0232, "step": 8990 }, { "grad_norm": 1.064813256263733, "learning_rate": 2.7145093084206598e-06, "loss": 0.0231, "step": 9000 }, { "grad_norm": 0.3807591497898102, "learning_rate": 2.661028168789892e-06, "loss": 0.0295, "step": 9010 }, { "grad_norm": 0.5200569033622742, "learning_rate": 2.6080647205706855e-06, "loss": 0.0202, "step": 9020 }, { "grad_norm": 0.3985135853290558, "learning_rate": 2.555619542962834e-06, "loss": 0.0213, "step": 9030 }, { "grad_norm": 0.4842459261417389, "learning_rate": 2.503693209498409e-06, "loss": 0.0194, "step": 9040 }, { "grad_norm": 0.43309152126312256, "learning_rate": 2.452286288035449e-06, "loss": 0.0214, "step": 9050 }, { "grad_norm": 0.27750229835510254, "learning_rate": 2.4013993407518363e-06, "loss": 0.0204, "step": 9060 }, { "grad_norm": 0.42672085762023926, "learning_rate": 2.351032924139063e-06, "loss": 0.0269, "step": 9070 }, { "grad_norm": 0.2665941119194031, "learning_rate": 2.30118758899619e-06, "loss": 0.0282, "step": 9080 }, { "grad_norm": 2.5250210762023926, "learning_rate": 2.2518638804238157e-06, "loss": 0.0256, "step": 9090 }, { "grad_norm": 0.4180643856525421, "learning_rate": 2.203062337818118e-06, "loss": 0.0249, "step": 9100 }, { "grad_norm": 0.5180522799491882, "learning_rate": 2.1547834948649483e-06, "loss": 0.0264, "step": 9110 }, { "grad_norm": 0.9126828908920288, "learning_rate": 2.1070278795340017e-06, "loss": 0.0236, "step": 9120 }, { "grad_norm": 0.5253291726112366, "learning_rate": 2.059796014073029e-06, "loss": 0.0273, "step": 9130 }, { "grad_norm": 0.37783849239349365, "learning_rate": 2.01308841500214e-06, "loss": 0.0329, "step": 9140 }, { "grad_norm": 0.38792023062705994, "learning_rate": 1.9669055931081704e-06, "loss": 0.028, "step": 9150 }, { "grad_norm": 0.4780908226966858, "learning_rate": 1.9212480534390507e-06, "loss": 0.023, "step": 9160 }, { "grad_norm": 0.8774855136871338, "learning_rate": 1.8761162952983246e-06, "loss": 0.0299, "step": 9170 }, { "grad_norm": 0.5942392945289612, "learning_rate": 1.8315108122396618e-06, "loss": 0.0203, "step": 9180 }, { "grad_norm": 0.2625359892845154, "learning_rate": 1.787432092061475e-06, "loss": 0.0231, "step": 9190 }, { "grad_norm": 0.3467889428138733, "learning_rate": 1.743880616801602e-06, "loss": 0.0271, "step": 9200 }, { "grad_norm": 0.27616366744041443, "learning_rate": 1.7008568627319865e-06, "loss": 0.0238, "step": 9210 }, { "grad_norm": 0.3191075921058655, "learning_rate": 1.6583613003535226e-06, "loss": 0.022, "step": 9220 }, { "grad_norm": 0.3378906548023224, "learning_rate": 1.6163943943908522e-06, "loss": 0.0204, "step": 9230 }, { "grad_norm": 1.242722511291504, "learning_rate": 1.5749566037873476e-06, "loss": 0.0204, "step": 9240 }, { "grad_norm": 1.0863956212997437, "learning_rate": 1.5340483817000428e-06, "loss": 0.0235, "step": 9250 }, { "grad_norm": 0.19269999861717224, "learning_rate": 1.4936701754947101e-06, "loss": 0.0158, "step": 9260 }, { "grad_norm": 0.3165123760700226, "learning_rate": 1.4538224267409361e-06, "loss": 0.0214, "step": 9270 }, { "grad_norm": 0.28326520323753357, "learning_rate": 1.414505571207314e-06, "loss": 0.0193, "step": 9280 }, { "grad_norm": 0.24328787624835968, "learning_rate": 1.3757200388566816e-06, "loss": 0.0216, "step": 9290 }, { "grad_norm": 0.5796698927879333, "learning_rate": 1.3374662538414074e-06, "loss": 0.0218, "step": 9300 }, { "grad_norm": 0.3439558446407318, "learning_rate": 1.2997446344987617e-06, "loss": 0.0221, "step": 9310 }, { "grad_norm": 0.3295418322086334, "learning_rate": 1.262555593346315e-06, "loss": 0.0254, "step": 9320 }, { "grad_norm": 0.33723556995391846, "learning_rate": 1.2258995370774685e-06, "loss": 0.0207, "step": 9330 }, { "grad_norm": 0.395887166261673, "learning_rate": 1.1897768665569798e-06, "loss": 0.0214, "step": 9340 }, { "grad_norm": 0.27041298151016235, "learning_rate": 1.1541879768165954e-06, "loss": 0.0154, "step": 9350 }, { "grad_norm": 0.20727092027664185, "learning_rate": 1.1191332570507085e-06, "loss": 0.0234, "step": 9360 }, { "grad_norm": 0.5107284188270569, "learning_rate": 1.0846130906121132e-06, "loss": 0.0252, "step": 9370 }, { "grad_norm": 0.5912117958068848, "learning_rate": 1.0506278550078131e-06, "loss": 0.0227, "step": 9380 }, { "grad_norm": 0.6253002882003784, "learning_rate": 1.0171779218949185e-06, "loss": 0.0247, "step": 9390 }, { "grad_norm": 0.20585118234157562, "learning_rate": 9.842636570765174e-07, "loss": 0.0281, "step": 9400 }, { "grad_norm": 0.29934990406036377, "learning_rate": 9.518854204977612e-07, "loss": 0.0211, "step": 9410 }, { "grad_norm": 0.40449780225753784, "learning_rate": 9.200435662418349e-07, "loss": 0.0237, "step": 9420 }, { "grad_norm": 0.19402848184108734, "learning_rate": 8.887384425261658e-07, "loss": 0.0219, "step": 9430 }, { "grad_norm": 0.3355835974216461, "learning_rate": 8.579703916985648e-07, "loss": 0.0199, "step": 9440 }, { "grad_norm": 0.7711787819862366, "learning_rate": 8.277397502335194e-07, "loss": 0.0256, "step": 9450 }, { "grad_norm": 0.4123276472091675, "learning_rate": 7.980468487284675e-07, "loss": 0.022, "step": 9460 }, { "grad_norm": 0.19559307396411896, "learning_rate": 7.688920119002297e-07, "loss": 0.0234, "step": 9470 }, { "grad_norm": 0.18713486194610596, "learning_rate": 7.402755585814269e-07, "loss": 0.0198, "step": 9480 }, { "grad_norm": 0.29147613048553467, "learning_rate": 7.121978017170073e-07, "loss": 0.0234, "step": 9490 }, { "grad_norm": 0.5416914224624634, "learning_rate": 6.846590483608306e-07, "loss": 0.0256, "step": 9500 }, { "grad_norm": 0.6483994722366333, "learning_rate": 6.576595996722834e-07, "loss": 0.0201, "step": 9510 }, { "grad_norm": 0.4099307656288147, "learning_rate": 6.311997509130141e-07, "loss": 0.019, "step": 9520 }, { "grad_norm": 0.42260485887527466, "learning_rate": 6.052797914436803e-07, "loss": 0.0232, "step": 9530 }, { "grad_norm": 0.2795911133289337, "learning_rate": 5.799000047208181e-07, "loss": 0.0216, "step": 9540 }, { "grad_norm": 0.2981329560279846, "learning_rate": 5.550606682937054e-07, "loss": 0.0241, "step": 9550 }, { "grad_norm": 0.3402572274208069, "learning_rate": 5.307620538013481e-07, "loss": 0.0203, "step": 9560 }, { "grad_norm": 0.1646828055381775, "learning_rate": 5.070044269694874e-07, "loss": 0.0188, "step": 9570 }, { "grad_norm": 0.3287215828895569, "learning_rate": 4.837880476077417e-07, "loss": 0.0248, "step": 9580 }, { "grad_norm": 0.5527787208557129, "learning_rate": 4.6111316960670835e-07, "loss": 0.0227, "step": 9590 }, { "grad_norm": 0.26895472407341003, "learning_rate": 4.389800409352218e-07, "loss": 0.0199, "step": 9600 }, { "grad_norm": 0.8149322867393494, "learning_rate": 4.173889036376277e-07, "loss": 0.0265, "step": 9610 }, { "grad_norm": 0.28635114431381226, "learning_rate": 3.963399938311463e-07, "loss": 0.0222, "step": 9620 }, { "grad_norm": 0.3597462773323059, "learning_rate": 3.7583354170328545e-07, "loss": 0.0204, "step": 9630 }, { "grad_norm": 0.33199068903923035, "learning_rate": 3.558697715093207e-07, "loss": 0.0226, "step": 9640 }, { "grad_norm": 0.46178677678108215, "learning_rate": 3.3644890156983576e-07, "loss": 0.0223, "step": 9650 }, { "grad_norm": 0.485688716173172, "learning_rate": 3.175711442683638e-07, "loss": 0.0179, "step": 9660 }, { "grad_norm": 1.4376778602600098, "learning_rate": 2.9923670604902197e-07, "loss": 0.0177, "step": 9670 }, { "grad_norm": 0.5326407551765442, "learning_rate": 2.814457874143028e-07, "loss": 0.0231, "step": 9680 }, { "grad_norm": 0.39845171570777893, "learning_rate": 2.641985829228366e-07, "loss": 0.0239, "step": 9690 }, { "grad_norm": 0.3926568925380707, "learning_rate": 2.474952811872877e-07, "loss": 0.0214, "step": 9700 }, { "grad_norm": 0.4270414412021637, "learning_rate": 2.3133606487228397e-07, "loss": 0.0229, "step": 9710 }, { "grad_norm": 0.3791038990020752, "learning_rate": 2.157211106924295e-07, "loss": 0.0235, "step": 9720 }, { "grad_norm": 0.5995451211929321, "learning_rate": 2.006505894103672e-07, "loss": 0.0285, "step": 9730 }, { "grad_norm": 3.323338508605957, "learning_rate": 1.8612466583489696e-07, "loss": 0.0174, "step": 9740 }, { "grad_norm": 0.39331528544425964, "learning_rate": 1.7214349881918834e-07, "loss": 0.0246, "step": 9750 }, { "grad_norm": 0.3890715539455414, "learning_rate": 1.5870724125904845e-07, "loss": 0.0223, "step": 9760 }, { "grad_norm": 0.6677830815315247, "learning_rate": 1.4581604009124006e-07, "loss": 0.0217, "step": 9770 }, { "grad_norm": 0.5347046256065369, "learning_rate": 1.334700362918717e-07, "loss": 0.0248, "step": 9780 }, { "grad_norm": 0.28236719965934753, "learning_rate": 1.2166936487486015e-07, "loss": 0.0182, "step": 9790 }, { "grad_norm": 0.46528416872024536, "learning_rate": 1.1041415489045914e-07, "loss": 0.0335, "step": 9800 }, { "grad_norm": 0.3817199468612671, "learning_rate": 9.970452942384412e-08, "loss": 0.0224, "step": 9810 }, { "grad_norm": 0.924498438835144, "learning_rate": 8.954060559375754e-08, "loss": 0.0246, "step": 9820 }, { "grad_norm": 0.7273439168930054, "learning_rate": 7.99224945512489e-08, "loss": 0.0192, "step": 9830 }, { "grad_norm": 0.4562165439128876, "learning_rate": 7.085030147843675e-08, "loss": 0.0246, "step": 9840 }, { "grad_norm": 0.366067498922348, "learning_rate": 6.232412558736523e-08, "loss": 0.0192, "step": 9850 }, { "grad_norm": 0.3241680860519409, "learning_rate": 5.434406011893822e-08, "loss": 0.0179, "step": 9860 }, { "grad_norm": 0.451852411031723, "learning_rate": 4.6910192341864664e-08, "loss": 0.026, "step": 9870 }, { "grad_norm": 0.2274736613035202, "learning_rate": 4.0022603551737035e-08, "loss": 0.0205, "step": 9880 }, { "grad_norm": 0.4411798119544983, "learning_rate": 3.3681369070120985e-08, "loss": 0.0231, "step": 9890 }, { "grad_norm": 0.2563296854496002, "learning_rate": 2.7886558243744866e-08, "loss": 0.0211, "step": 9900 }, { "grad_norm": 0.26643574237823486, "learning_rate": 2.2638234443722596e-08, "loss": 0.018, "step": 9910 }, { "grad_norm": 0.6394485235214233, "learning_rate": 1.7936455064887504e-08, "loss": 0.0173, "step": 9920 }, { "grad_norm": 0.26663437485694885, "learning_rate": 1.378127152514841e-08, "loss": 0.0192, "step": 9930 }, { "grad_norm": 3.856313467025757, "learning_rate": 1.0172729264917857e-08, "loss": 0.0253, "step": 9940 }, { "grad_norm": 0.7363432049751282, "learning_rate": 7.1108677466458215e-09, "loss": 0.0217, "step": 9950 }, { "grad_norm": 0.45511987805366516, "learning_rate": 4.595720454353414e-09, "loss": 0.0193, "step": 9960 }, { "grad_norm": 0.3037374019622803, "learning_rate": 2.627314893294264e-09, "loss": 0.0245, "step": 9970 }, { "grad_norm": 0.3044417202472687, "learning_rate": 1.2056725896270048e-09, "loss": 0.0158, "step": 9980 }, { "grad_norm": 0.4187524616718292, "learning_rate": 3.308090902098826e-10, "loss": 0.0148, "step": 9990 }, { "grad_norm": 0.3531603515148163, "learning_rate": 2.7339624120159555e-12, "loss": 0.02, "step": 10000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 3000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }