{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 52.91005291005291, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.05291005291005291, "grad_norm": 9.126862525939941, "learning_rate": 1.8e-06, "loss": 1.2067, "step": 10 }, { "epoch": 0.10582010582010581, "grad_norm": 4.3329243659973145, "learning_rate": 3.8e-06, "loss": 0.9948, "step": 20 }, { "epoch": 0.15873015873015872, "grad_norm": 1.4057083129882812, "learning_rate": 5.8e-06, "loss": 0.4462, "step": 30 }, { "epoch": 0.21164021164021163, "grad_norm": 0.748518168926239, "learning_rate": 7.8e-06, "loss": 0.1997, "step": 40 }, { "epoch": 0.26455026455026454, "grad_norm": 0.580504298210144, "learning_rate": 9.800000000000001e-06, "loss": 0.1223, "step": 50 }, { "epoch": 0.31746031746031744, "grad_norm": 0.5996851921081543, "learning_rate": 1.18e-05, "loss": 0.095, "step": 60 }, { "epoch": 0.37037037037037035, "grad_norm": 0.5250024795532227, "learning_rate": 1.3800000000000002e-05, "loss": 0.0894, "step": 70 }, { "epoch": 0.42328042328042326, "grad_norm": 0.2574264705181122, "learning_rate": 1.58e-05, "loss": 0.0781, "step": 80 }, { "epoch": 0.47619047619047616, "grad_norm": 0.38775157928466797, "learning_rate": 1.78e-05, "loss": 0.0709, "step": 90 }, { "epoch": 0.5291005291005291, "grad_norm": 0.4114510118961334, "learning_rate": 1.9800000000000004e-05, "loss": 0.0643, "step": 100 }, { "epoch": 0.582010582010582, "grad_norm": 0.2912083864212036, "learning_rate": 2.18e-05, "loss": 0.0598, "step": 110 }, { "epoch": 0.6349206349206349, "grad_norm": 0.3235427439212799, "learning_rate": 2.38e-05, "loss": 0.0534, "step": 120 }, { "epoch": 0.6878306878306878, "grad_norm": 0.41158467531204224, "learning_rate": 2.58e-05, "loss": 0.0517, "step": 130 }, { "epoch": 0.7407407407407407, "grad_norm": 0.30140212178230286, "learning_rate": 2.7800000000000005e-05, "loss": 0.0521, "step": 140 }, { "epoch": 0.7936507936507936, "grad_norm": 0.30355292558670044, "learning_rate": 2.98e-05, "loss": 0.047, "step": 150 }, { "epoch": 0.8465608465608465, "grad_norm": 0.4104730486869812, "learning_rate": 3.18e-05, "loss": 0.0426, "step": 160 }, { "epoch": 0.8994708994708994, "grad_norm": 0.35002169013023376, "learning_rate": 3.38e-05, "loss": 0.0397, "step": 170 }, { "epoch": 0.9523809523809523, "grad_norm": 0.32858532667160034, "learning_rate": 3.58e-05, "loss": 0.04, "step": 180 }, { "epoch": 1.0052910052910053, "grad_norm": 0.33840125799179077, "learning_rate": 3.7800000000000004e-05, "loss": 0.0397, "step": 190 }, { "epoch": 1.0582010582010581, "grad_norm": 0.310943603515625, "learning_rate": 3.9800000000000005e-05, "loss": 0.0387, "step": 200 }, { "epoch": 1.1111111111111112, "grad_norm": 0.41826796531677246, "learning_rate": 4.18e-05, "loss": 0.0334, "step": 210 }, { "epoch": 1.164021164021164, "grad_norm": 0.34721270203590393, "learning_rate": 4.38e-05, "loss": 0.0316, "step": 220 }, { "epoch": 1.216931216931217, "grad_norm": 0.28007030487060547, "learning_rate": 4.58e-05, "loss": 0.0282, "step": 230 }, { "epoch": 1.2698412698412698, "grad_norm": 0.28840744495391846, "learning_rate": 4.78e-05, "loss": 0.0298, "step": 240 }, { "epoch": 1.3227513227513228, "grad_norm": 0.30962225794792175, "learning_rate": 4.9800000000000004e-05, "loss": 0.0267, "step": 250 }, { "epoch": 1.3756613756613756, "grad_norm": 0.23388192057609558, "learning_rate": 5.1800000000000005e-05, "loss": 0.0247, "step": 260 }, { "epoch": 1.4285714285714286, "grad_norm": 0.293412446975708, "learning_rate": 5.380000000000001e-05, "loss": 0.0278, "step": 270 }, { "epoch": 1.4814814814814814, "grad_norm": 0.2939009964466095, "learning_rate": 5.580000000000001e-05, "loss": 0.0252, "step": 280 }, { "epoch": 1.5343915343915344, "grad_norm": 0.2336420863866806, "learning_rate": 5.7799999999999995e-05, "loss": 0.0231, "step": 290 }, { "epoch": 1.5873015873015874, "grad_norm": 0.3097730278968811, "learning_rate": 5.9800000000000003e-05, "loss": 0.0277, "step": 300 }, { "epoch": 1.6402116402116402, "grad_norm": 0.3215116560459137, "learning_rate": 6.18e-05, "loss": 0.0224, "step": 310 }, { "epoch": 1.693121693121693, "grad_norm": 0.3378458619117737, "learning_rate": 6.38e-05, "loss": 0.0209, "step": 320 }, { "epoch": 1.746031746031746, "grad_norm": 0.3013512194156647, "learning_rate": 6.58e-05, "loss": 0.0223, "step": 330 }, { "epoch": 1.798941798941799, "grad_norm": 0.36562272906303406, "learning_rate": 6.780000000000001e-05, "loss": 0.0223, "step": 340 }, { "epoch": 1.8518518518518519, "grad_norm": 0.3270028531551361, "learning_rate": 6.98e-05, "loss": 0.02, "step": 350 }, { "epoch": 1.9047619047619047, "grad_norm": 0.2906351685523987, "learning_rate": 7.18e-05, "loss": 0.0186, "step": 360 }, { "epoch": 1.9576719576719577, "grad_norm": 0.24939194321632385, "learning_rate": 7.38e-05, "loss": 0.0183, "step": 370 }, { "epoch": 2.0105820105820107, "grad_norm": 0.25886496901512146, "learning_rate": 7.58e-05, "loss": 0.019, "step": 380 }, { "epoch": 2.0634920634920633, "grad_norm": 0.30217093229293823, "learning_rate": 7.780000000000001e-05, "loss": 0.0204, "step": 390 }, { "epoch": 2.1164021164021163, "grad_norm": 0.1779695302248001, "learning_rate": 7.98e-05, "loss": 0.0168, "step": 400 }, { "epoch": 2.1693121693121693, "grad_norm": 0.22402387857437134, "learning_rate": 8.18e-05, "loss": 0.0154, "step": 410 }, { "epoch": 2.2222222222222223, "grad_norm": 0.2873242497444153, "learning_rate": 8.38e-05, "loss": 0.0157, "step": 420 }, { "epoch": 2.2751322751322753, "grad_norm": 0.3185768723487854, "learning_rate": 8.58e-05, "loss": 0.0205, "step": 430 }, { "epoch": 2.328042328042328, "grad_norm": 0.2523813247680664, "learning_rate": 8.78e-05, "loss": 0.0181, "step": 440 }, { "epoch": 2.380952380952381, "grad_norm": 0.21842168271541595, "learning_rate": 8.98e-05, "loss": 0.0192, "step": 450 }, { "epoch": 2.433862433862434, "grad_norm": 0.32591164112091064, "learning_rate": 9.180000000000001e-05, "loss": 0.0182, "step": 460 }, { "epoch": 2.4867724867724865, "grad_norm": 0.22708426415920258, "learning_rate": 9.38e-05, "loss": 0.0166, "step": 470 }, { "epoch": 2.5396825396825395, "grad_norm": 0.3629976511001587, "learning_rate": 9.58e-05, "loss": 0.016, "step": 480 }, { "epoch": 2.5925925925925926, "grad_norm": 0.25041821599006653, "learning_rate": 9.78e-05, "loss": 0.0145, "step": 490 }, { "epoch": 2.6455026455026456, "grad_norm": 0.36431604623794556, "learning_rate": 9.98e-05, "loss": 0.0141, "step": 500 }, { "epoch": 2.6984126984126986, "grad_norm": 0.29010534286499023, "learning_rate": 9.9999778549206e-05, "loss": 0.0164, "step": 510 }, { "epoch": 2.751322751322751, "grad_norm": 0.2596469223499298, "learning_rate": 9.999901304280685e-05, "loss": 0.0182, "step": 520 }, { "epoch": 2.804232804232804, "grad_norm": 0.3801230788230896, "learning_rate": 9.999770075521164e-05, "loss": 0.0151, "step": 530 }, { "epoch": 2.857142857142857, "grad_norm": 0.359813928604126, "learning_rate": 9.99958417007713e-05, "loss": 0.0167, "step": 540 }, { "epoch": 2.91005291005291, "grad_norm": 0.322299063205719, "learning_rate": 9.999343589981615e-05, "loss": 0.0176, "step": 550 }, { "epoch": 2.962962962962963, "grad_norm": 0.2990533411502838, "learning_rate": 9.999048337865568e-05, "loss": 0.0155, "step": 560 }, { "epoch": 3.015873015873016, "grad_norm": 0.35644400119781494, "learning_rate": 9.998698416957815e-05, "loss": 0.0147, "step": 570 }, { "epoch": 3.068783068783069, "grad_norm": 0.2201194167137146, "learning_rate": 9.998293831085037e-05, "loss": 0.016, "step": 580 }, { "epoch": 3.121693121693122, "grad_norm": 0.22524341940879822, "learning_rate": 9.997834584671719e-05, "loss": 0.0134, "step": 590 }, { "epoch": 3.1746031746031744, "grad_norm": 0.24721431732177734, "learning_rate": 9.997320682740107e-05, "loss": 0.0133, "step": 600 }, { "epoch": 3.2275132275132274, "grad_norm": 0.22271420061588287, "learning_rate": 9.996752130910149e-05, "loss": 0.015, "step": 610 }, { "epoch": 3.2804232804232805, "grad_norm": 0.2773207426071167, "learning_rate": 9.99612893539944e-05, "loss": 0.0121, "step": 620 }, { "epoch": 3.3333333333333335, "grad_norm": 0.22422675788402557, "learning_rate": 9.995451103023144e-05, "loss": 0.0137, "step": 630 }, { "epoch": 3.386243386243386, "grad_norm": 0.2582796812057495, "learning_rate": 9.994718641193928e-05, "loss": 0.0144, "step": 640 }, { "epoch": 3.439153439153439, "grad_norm": 0.22184346616268158, "learning_rate": 9.993931557921874e-05, "loss": 0.0121, "step": 650 }, { "epoch": 3.492063492063492, "grad_norm": 0.28815826773643494, "learning_rate": 9.993089861814402e-05, "loss": 0.013, "step": 660 }, { "epoch": 3.544973544973545, "grad_norm": 0.29933735728263855, "learning_rate": 9.992193562076166e-05, "loss": 0.0124, "step": 670 }, { "epoch": 3.597883597883598, "grad_norm": 0.35792794823646545, "learning_rate": 9.991242668508954e-05, "loss": 0.0154, "step": 680 }, { "epoch": 3.6507936507936507, "grad_norm": 0.2978787124156952, "learning_rate": 9.990237191511587e-05, "loss": 0.0129, "step": 690 }, { "epoch": 3.7037037037037037, "grad_norm": 0.29716867208480835, "learning_rate": 9.989177142079802e-05, "loss": 0.0107, "step": 700 }, { "epoch": 3.7566137566137567, "grad_norm": 0.3191848397254944, "learning_rate": 9.988062531806126e-05, "loss": 0.0133, "step": 710 }, { "epoch": 3.8095238095238093, "grad_norm": 0.21713797748088837, "learning_rate": 9.986893372879762e-05, "loss": 0.0125, "step": 720 }, { "epoch": 3.8624338624338623, "grad_norm": 0.32239532470703125, "learning_rate": 9.985669678086443e-05, "loss": 0.0153, "step": 730 }, { "epoch": 3.9153439153439153, "grad_norm": 0.36931267380714417, "learning_rate": 9.984391460808298e-05, "loss": 0.0117, "step": 740 }, { "epoch": 3.9682539682539684, "grad_norm": 0.23314177989959717, "learning_rate": 9.983058735023709e-05, "loss": 0.0126, "step": 750 }, { "epoch": 4.021164021164021, "grad_norm": 0.1878504902124405, "learning_rate": 9.98167151530715e-05, "loss": 0.012, "step": 760 }, { "epoch": 4.074074074074074, "grad_norm": 0.26781243085861206, "learning_rate": 9.980229816829034e-05, "loss": 0.013, "step": 770 }, { "epoch": 4.1269841269841265, "grad_norm": 0.24909581243991852, "learning_rate": 9.978733655355544e-05, "loss": 0.0118, "step": 780 }, { "epoch": 4.1798941798941796, "grad_norm": 0.25832822918891907, "learning_rate": 9.977183047248464e-05, "loss": 0.013, "step": 790 }, { "epoch": 4.232804232804233, "grad_norm": 0.25693967938423157, "learning_rate": 9.975578009464992e-05, "loss": 0.0144, "step": 800 }, { "epoch": 4.285714285714286, "grad_norm": 0.3252553343772888, "learning_rate": 9.97391855955757e-05, "loss": 0.0116, "step": 810 }, { "epoch": 4.338624338624339, "grad_norm": 0.20883044600486755, "learning_rate": 9.972204715673669e-05, "loss": 0.0133, "step": 820 }, { "epoch": 4.391534391534392, "grad_norm": 0.25589728355407715, "learning_rate": 9.970436496555617e-05, "loss": 0.0129, "step": 830 }, { "epoch": 4.444444444444445, "grad_norm": 0.20333099365234375, "learning_rate": 9.968613921540373e-05, "loss": 0.0115, "step": 840 }, { "epoch": 4.497354497354498, "grad_norm": 0.2376997172832489, "learning_rate": 9.966737010559326e-05, "loss": 0.0116, "step": 850 }, { "epoch": 4.550264550264551, "grad_norm": 0.25048720836639404, "learning_rate": 9.964805784138072e-05, "loss": 0.0113, "step": 860 }, { "epoch": 4.603174603174603, "grad_norm": 0.37046122550964355, "learning_rate": 9.962820263396195e-05, "loss": 0.011, "step": 870 }, { "epoch": 4.656084656084656, "grad_norm": 0.20802900195121765, "learning_rate": 9.960780470047033e-05, "loss": 0.0094, "step": 880 }, { "epoch": 4.708994708994709, "grad_norm": 0.20334285497665405, "learning_rate": 9.958686426397437e-05, "loss": 0.0099, "step": 890 }, { "epoch": 4.761904761904762, "grad_norm": 0.2225441336631775, "learning_rate": 9.956538155347534e-05, "loss": 0.0107, "step": 900 }, { "epoch": 4.814814814814815, "grad_norm": 0.27975761890411377, "learning_rate": 9.95433568039047e-05, "loss": 0.012, "step": 910 }, { "epoch": 4.867724867724868, "grad_norm": 0.2021826058626175, "learning_rate": 9.952079025612162e-05, "loss": 0.0111, "step": 920 }, { "epoch": 4.920634920634921, "grad_norm": 0.1899338811635971, "learning_rate": 9.949768215691022e-05, "loss": 0.0113, "step": 930 }, { "epoch": 4.973544973544973, "grad_norm": 0.23938655853271484, "learning_rate": 9.9474032758977e-05, "loss": 0.0106, "step": 940 }, { "epoch": 5.026455026455026, "grad_norm": 0.20454533398151398, "learning_rate": 9.944984232094794e-05, "loss": 0.0127, "step": 950 }, { "epoch": 5.079365079365079, "grad_norm": 0.24551042914390564, "learning_rate": 9.942511110736584e-05, "loss": 0.0133, "step": 960 }, { "epoch": 5.132275132275132, "grad_norm": 0.21976925432682037, "learning_rate": 9.939983938868726e-05, "loss": 0.0108, "step": 970 }, { "epoch": 5.185185185185185, "grad_norm": 0.31280621886253357, "learning_rate": 9.93740274412797e-05, "loss": 0.0111, "step": 980 }, { "epoch": 5.238095238095238, "grad_norm": 0.30914512276649475, "learning_rate": 9.934767554741846e-05, "loss": 0.0137, "step": 990 }, { "epoch": 5.291005291005291, "grad_norm": 0.18038618564605713, "learning_rate": 9.932078399528361e-05, "loss": 0.0109, "step": 1000 }, { "epoch": 5.343915343915344, "grad_norm": 0.2399945855140686, "learning_rate": 9.929335307895689e-05, "loss": 0.0103, "step": 1010 }, { "epoch": 5.396825396825397, "grad_norm": 0.2483704537153244, "learning_rate": 9.926538309841839e-05, "loss": 0.0106, "step": 1020 }, { "epoch": 5.449735449735449, "grad_norm": 0.23170380294322968, "learning_rate": 9.923687435954334e-05, "loss": 0.0102, "step": 1030 }, { "epoch": 5.502645502645502, "grad_norm": 0.22171145677566528, "learning_rate": 9.920782717409873e-05, "loss": 0.0086, "step": 1040 }, { "epoch": 5.555555555555555, "grad_norm": 0.16594019532203674, "learning_rate": 9.917824185973994e-05, "loss": 0.0096, "step": 1050 }, { "epoch": 5.608465608465608, "grad_norm": 0.23909465968608856, "learning_rate": 9.914811874000723e-05, "loss": 0.0112, "step": 1060 }, { "epoch": 5.661375661375661, "grad_norm": 0.17024555802345276, "learning_rate": 9.911745814432218e-05, "loss": 0.0098, "step": 1070 }, { "epoch": 5.714285714285714, "grad_norm": 0.23646000027656555, "learning_rate": 9.90862604079842e-05, "loss": 0.0117, "step": 1080 }, { "epoch": 5.767195767195767, "grad_norm": 0.20663554966449738, "learning_rate": 9.90545258721667e-05, "loss": 0.0095, "step": 1090 }, { "epoch": 5.8201058201058204, "grad_norm": 0.3044997453689575, "learning_rate": 9.90222548839135e-05, "loss": 0.0119, "step": 1100 }, { "epoch": 5.8730158730158735, "grad_norm": 0.2594415545463562, "learning_rate": 9.898944779613495e-05, "loss": 0.0118, "step": 1110 }, { "epoch": 5.925925925925926, "grad_norm": 0.229117751121521, "learning_rate": 9.89561049676041e-05, "loss": 0.0112, "step": 1120 }, { "epoch": 5.978835978835979, "grad_norm": 0.2768831253051758, "learning_rate": 9.89222267629528e-05, "loss": 0.011, "step": 1130 }, { "epoch": 6.031746031746032, "grad_norm": 0.19585560262203217, "learning_rate": 9.888781355266763e-05, "loss": 0.011, "step": 1140 }, { "epoch": 6.084656084656085, "grad_norm": 0.2453078180551529, "learning_rate": 9.885286571308598e-05, "loss": 0.0136, "step": 1150 }, { "epoch": 6.137566137566138, "grad_norm": 0.225248321890831, "learning_rate": 9.881738362639182e-05, "loss": 0.0099, "step": 1160 }, { "epoch": 6.190476190476191, "grad_norm": 0.1603901982307434, "learning_rate": 9.878136768061154e-05, "loss": 0.0089, "step": 1170 }, { "epoch": 6.243386243386244, "grad_norm": 0.17369356751441956, "learning_rate": 9.874481826960979e-05, "loss": 0.0086, "step": 1180 }, { "epoch": 6.296296296296296, "grad_norm": 0.26182791590690613, "learning_rate": 9.870773579308503e-05, "loss": 0.0104, "step": 1190 }, { "epoch": 6.349206349206349, "grad_norm": 0.22114048898220062, "learning_rate": 9.867012065656533e-05, "loss": 0.0102, "step": 1200 }, { "epoch": 6.402116402116402, "grad_norm": 0.16951113939285278, "learning_rate": 9.863197327140376e-05, "loss": 0.008, "step": 1210 }, { "epoch": 6.455026455026455, "grad_norm": 0.24596001207828522, "learning_rate": 9.859329405477403e-05, "loss": 0.0116, "step": 1220 }, { "epoch": 6.507936507936508, "grad_norm": 0.19753050804138184, "learning_rate": 9.855408342966585e-05, "loss": 0.009, "step": 1230 }, { "epoch": 6.560846560846561, "grad_norm": 0.26562634110450745, "learning_rate": 9.851434182488033e-05, "loss": 0.01, "step": 1240 }, { "epoch": 6.613756613756614, "grad_norm": 0.16184137761592865, "learning_rate": 9.84740696750253e-05, "loss": 0.0094, "step": 1250 }, { "epoch": 6.666666666666667, "grad_norm": 0.18835291266441345, "learning_rate": 9.843326742051055e-05, "loss": 0.0091, "step": 1260 }, { "epoch": 6.71957671957672, "grad_norm": 0.22983013093471527, "learning_rate": 9.839193550754297e-05, "loss": 0.0113, "step": 1270 }, { "epoch": 6.772486772486772, "grad_norm": 0.21129533648490906, "learning_rate": 9.835007438812177e-05, "loss": 0.0086, "step": 1280 }, { "epoch": 6.825396825396825, "grad_norm": 0.20113864541053772, "learning_rate": 9.830768452003341e-05, "loss": 0.0078, "step": 1290 }, { "epoch": 6.878306878306878, "grad_norm": 0.2174144685268402, "learning_rate": 9.826476636684671e-05, "loss": 0.0114, "step": 1300 }, { "epoch": 6.931216931216931, "grad_norm": 0.18579621613025665, "learning_rate": 9.822132039790773e-05, "loss": 0.0111, "step": 1310 }, { "epoch": 6.984126984126984, "grad_norm": 0.22104878723621368, "learning_rate": 9.817734708833461e-05, "loss": 0.0089, "step": 1320 }, { "epoch": 7.037037037037037, "grad_norm": 0.17224571108818054, "learning_rate": 9.813284691901243e-05, "loss": 0.0087, "step": 1330 }, { "epoch": 7.08994708994709, "grad_norm": 0.20241349935531616, "learning_rate": 9.808782037658792e-05, "loss": 0.0073, "step": 1340 }, { "epoch": 7.142857142857143, "grad_norm": 0.14239327609539032, "learning_rate": 9.804226795346411e-05, "loss": 0.0072, "step": 1350 }, { "epoch": 7.195767195767195, "grad_norm": 0.2131844460964203, "learning_rate": 9.799619014779503e-05, "loss": 0.0078, "step": 1360 }, { "epoch": 7.248677248677248, "grad_norm": 0.1776800900697708, "learning_rate": 9.794958746348013e-05, "loss": 0.0091, "step": 1370 }, { "epoch": 7.301587301587301, "grad_norm": 0.20831404626369476, "learning_rate": 9.790246041015896e-05, "loss": 0.0083, "step": 1380 }, { "epoch": 7.354497354497354, "grad_norm": 0.13879866898059845, "learning_rate": 9.785480950320538e-05, "loss": 0.0077, "step": 1390 }, { "epoch": 7.407407407407407, "grad_norm": 0.18246503174304962, "learning_rate": 9.78066352637221e-05, "loss": 0.0088, "step": 1400 }, { "epoch": 7.4603174603174605, "grad_norm": 0.22751545906066895, "learning_rate": 9.775793821853488e-05, "loss": 0.0085, "step": 1410 }, { "epoch": 7.5132275132275135, "grad_norm": 0.15783411264419556, "learning_rate": 9.77087189001868e-05, "loss": 0.0075, "step": 1420 }, { "epoch": 7.5661375661375665, "grad_norm": 0.26348280906677246, "learning_rate": 9.765897784693243e-05, "loss": 0.0085, "step": 1430 }, { "epoch": 7.619047619047619, "grad_norm": 0.17351244390010834, "learning_rate": 9.760871560273197e-05, "loss": 0.009, "step": 1440 }, { "epoch": 7.671957671957672, "grad_norm": 0.21902164816856384, "learning_rate": 9.755793271724526e-05, "loss": 0.0093, "step": 1450 }, { "epoch": 7.724867724867725, "grad_norm": 0.22974957525730133, "learning_rate": 9.750662974582584e-05, "loss": 0.01, "step": 1460 }, { "epoch": 7.777777777777778, "grad_norm": 0.17598550021648407, "learning_rate": 9.745480724951473e-05, "loss": 0.007, "step": 1470 }, { "epoch": 7.830687830687831, "grad_norm": 0.16767877340316772, "learning_rate": 9.740246579503447e-05, "loss": 0.0103, "step": 1480 }, { "epoch": 7.883597883597884, "grad_norm": 0.18469084799289703, "learning_rate": 9.734960595478284e-05, "loss": 0.0097, "step": 1490 }, { "epoch": 7.936507936507937, "grad_norm": 0.19406293332576752, "learning_rate": 9.729622830682657e-05, "loss": 0.0108, "step": 1500 }, { "epoch": 7.98941798941799, "grad_norm": 0.1934950202703476, "learning_rate": 9.724233343489504e-05, "loss": 0.0086, "step": 1510 }, { "epoch": 8.042328042328043, "grad_norm": 0.16742374002933502, "learning_rate": 9.718792192837396e-05, "loss": 0.008, "step": 1520 }, { "epoch": 8.095238095238095, "grad_norm": 0.2543432116508484, "learning_rate": 9.713299438229886e-05, "loss": 0.0083, "step": 1530 }, { "epoch": 8.148148148148149, "grad_norm": 0.16946610808372498, "learning_rate": 9.707755139734855e-05, "loss": 0.0088, "step": 1540 }, { "epoch": 8.201058201058201, "grad_norm": 0.19916781783103943, "learning_rate": 9.702159357983866e-05, "loss": 0.0074, "step": 1550 }, { "epoch": 8.253968253968253, "grad_norm": 0.14717981219291687, "learning_rate": 9.696512154171492e-05, "loss": 0.0077, "step": 1560 }, { "epoch": 8.306878306878307, "grad_norm": 0.23303870856761932, "learning_rate": 9.690813590054645e-05, "loss": 0.0082, "step": 1570 }, { "epoch": 8.359788359788359, "grad_norm": 0.181081160902977, "learning_rate": 9.685063727951914e-05, "loss": 0.0078, "step": 1580 }, { "epoch": 8.412698412698413, "grad_norm": 0.2447252869606018, "learning_rate": 9.679262630742865e-05, "loss": 0.0078, "step": 1590 }, { "epoch": 8.465608465608465, "grad_norm": 0.14809352159500122, "learning_rate": 9.673410361867373e-05, "loss": 0.0077, "step": 1600 }, { "epoch": 8.518518518518519, "grad_norm": 0.2121380716562271, "learning_rate": 9.667506985324909e-05, "loss": 0.0086, "step": 1610 }, { "epoch": 8.571428571428571, "grad_norm": 0.12314649671316147, "learning_rate": 9.661552565673855e-05, "loss": 0.0069, "step": 1620 }, { "epoch": 8.624338624338625, "grad_norm": 0.16369786858558655, "learning_rate": 9.655547168030789e-05, "loss": 0.01, "step": 1630 }, { "epoch": 8.677248677248677, "grad_norm": 0.20343919098377228, "learning_rate": 9.649490858069777e-05, "loss": 0.0089, "step": 1640 }, { "epoch": 8.73015873015873, "grad_norm": 0.15146712958812714, "learning_rate": 9.643383702021658e-05, "loss": 0.0084, "step": 1650 }, { "epoch": 8.783068783068783, "grad_norm": 0.1656627506017685, "learning_rate": 9.637225766673307e-05, "loss": 0.008, "step": 1660 }, { "epoch": 8.835978835978835, "grad_norm": 0.1543031483888626, "learning_rate": 9.631017119366922e-05, "loss": 0.0073, "step": 1670 }, { "epoch": 8.88888888888889, "grad_norm": 0.17806491255760193, "learning_rate": 9.624757827999273e-05, "loss": 0.0071, "step": 1680 }, { "epoch": 8.941798941798941, "grad_norm": 0.16285857558250427, "learning_rate": 9.618447961020971e-05, "loss": 0.0067, "step": 1690 }, { "epoch": 8.994708994708995, "grad_norm": 0.1509881317615509, "learning_rate": 9.612087587435707e-05, "loss": 0.0104, "step": 1700 }, { "epoch": 9.047619047619047, "grad_norm": 0.17392238974571228, "learning_rate": 9.605676776799508e-05, "loss": 0.0079, "step": 1710 }, { "epoch": 9.100529100529101, "grad_norm": 0.17451897263526917, "learning_rate": 9.599215599219973e-05, "loss": 0.0087, "step": 1720 }, { "epoch": 9.153439153439153, "grad_norm": 0.2740662395954132, "learning_rate": 9.592704125355505e-05, "loss": 0.009, "step": 1730 }, { "epoch": 9.206349206349206, "grad_norm": 0.2265627086162567, "learning_rate": 9.586142426414538e-05, "loss": 0.0085, "step": 1740 }, { "epoch": 9.25925925925926, "grad_norm": 0.26739341020584106, "learning_rate": 9.57953057415476e-05, "loss": 0.0087, "step": 1750 }, { "epoch": 9.312169312169312, "grad_norm": 0.2655901610851288, "learning_rate": 9.572868640882328e-05, "loss": 0.0074, "step": 1760 }, { "epoch": 9.365079365079366, "grad_norm": 0.1802184134721756, "learning_rate": 9.56615669945108e-05, "loss": 0.0076, "step": 1770 }, { "epoch": 9.417989417989418, "grad_norm": 0.19627468287944794, "learning_rate": 9.55939482326173e-05, "loss": 0.0078, "step": 1780 }, { "epoch": 9.470899470899472, "grad_norm": 0.2125983089208603, "learning_rate": 9.552583086261069e-05, "loss": 0.0065, "step": 1790 }, { "epoch": 9.523809523809524, "grad_norm": 0.1243014857172966, "learning_rate": 9.545721562941168e-05, "loss": 0.0075, "step": 1800 }, { "epoch": 9.576719576719576, "grad_norm": 0.18863843381404877, "learning_rate": 9.538810328338543e-05, "loss": 0.007, "step": 1810 }, { "epoch": 9.62962962962963, "grad_norm": 0.22842098772525787, "learning_rate": 9.531849458033349e-05, "loss": 0.0071, "step": 1820 }, { "epoch": 9.682539682539682, "grad_norm": 0.18659493327140808, "learning_rate": 9.524839028148547e-05, "loss": 0.0077, "step": 1830 }, { "epoch": 9.735449735449736, "grad_norm": 0.2362309992313385, "learning_rate": 9.517779115349077e-05, "loss": 0.0076, "step": 1840 }, { "epoch": 9.788359788359788, "grad_norm": 0.24288448691368103, "learning_rate": 9.510669796841014e-05, "loss": 0.007, "step": 1850 }, { "epoch": 9.841269841269842, "grad_norm": 0.28122323751449585, "learning_rate": 9.503511150370727e-05, "loss": 0.01, "step": 1860 }, { "epoch": 9.894179894179894, "grad_norm": 0.144936203956604, "learning_rate": 9.496303254224024e-05, "loss": 0.008, "step": 1870 }, { "epoch": 9.947089947089948, "grad_norm": 0.19548813998699188, "learning_rate": 9.489046187225306e-05, "loss": 0.0078, "step": 1880 }, { "epoch": 10.0, "grad_norm": 0.11918188631534576, "learning_rate": 9.481740028736692e-05, "loss": 0.0075, "step": 1890 }, { "epoch": 10.052910052910052, "grad_norm": 0.2927192747592926, "learning_rate": 9.474384858657164e-05, "loss": 0.0088, "step": 1900 }, { "epoch": 10.105820105820106, "grad_norm": 0.2122175246477127, "learning_rate": 9.466980757421679e-05, "loss": 0.0069, "step": 1910 }, { "epoch": 10.158730158730158, "grad_norm": 0.20151248574256897, "learning_rate": 9.459527806000305e-05, "loss": 0.0079, "step": 1920 }, { "epoch": 10.211640211640212, "grad_norm": 0.1929294615983963, "learning_rate": 9.452026085897325e-05, "loss": 0.008, "step": 1930 }, { "epoch": 10.264550264550264, "grad_norm": 0.18397334218025208, "learning_rate": 9.444475679150348e-05, "loss": 0.0085, "step": 1940 }, { "epoch": 10.317460317460318, "grad_norm": 0.18000218272209167, "learning_rate": 9.436876668329411e-05, "loss": 0.0073, "step": 1950 }, { "epoch": 10.37037037037037, "grad_norm": 0.2253434956073761, "learning_rate": 9.429229136536079e-05, "loss": 0.0098, "step": 1960 }, { "epoch": 10.423280423280424, "grad_norm": 0.17696668207645416, "learning_rate": 9.421533167402534e-05, "loss": 0.008, "step": 1970 }, { "epoch": 10.476190476190476, "grad_norm": 0.16151362657546997, "learning_rate": 9.413788845090666e-05, "loss": 0.0075, "step": 1980 }, { "epoch": 10.529100529100528, "grad_norm": 0.1504540592432022, "learning_rate": 9.405996254291136e-05, "loss": 0.0094, "step": 1990 }, { "epoch": 10.582010582010582, "grad_norm": 0.14837105572223663, "learning_rate": 9.398155480222474e-05, "loss": 0.0069, "step": 2000 }, { "epoch": 10.634920634920634, "grad_norm": 0.24493512511253357, "learning_rate": 9.390266608630128e-05, "loss": 0.0084, "step": 2010 }, { "epoch": 10.687830687830688, "grad_norm": 0.11716434359550476, "learning_rate": 9.38232972578553e-05, "loss": 0.0062, "step": 2020 }, { "epoch": 10.74074074074074, "grad_norm": 0.19078141450881958, "learning_rate": 9.374344918485164e-05, "loss": 0.0081, "step": 2030 }, { "epoch": 10.793650793650794, "grad_norm": 0.11472897231578827, "learning_rate": 9.366312274049602e-05, "loss": 0.0059, "step": 2040 }, { "epoch": 10.846560846560847, "grad_norm": 0.16060738265514374, "learning_rate": 9.358231880322554e-05, "loss": 0.0086, "step": 2050 }, { "epoch": 10.899470899470899, "grad_norm": 0.211285799741745, "learning_rate": 9.350103825669916e-05, "loss": 0.0071, "step": 2060 }, { "epoch": 10.952380952380953, "grad_norm": 0.16725730895996094, "learning_rate": 9.341928198978787e-05, "loss": 0.0072, "step": 2070 }, { "epoch": 11.005291005291005, "grad_norm": 0.15011338889598846, "learning_rate": 9.333705089656512e-05, "loss": 0.0057, "step": 2080 }, { "epoch": 11.058201058201059, "grad_norm": 0.16925115883350372, "learning_rate": 9.325434587629698e-05, "loss": 0.0089, "step": 2090 }, { "epoch": 11.11111111111111, "grad_norm": 0.144941508769989, "learning_rate": 9.31711678334323e-05, "loss": 0.0074, "step": 2100 }, { "epoch": 11.164021164021165, "grad_norm": 0.14591579139232635, "learning_rate": 9.308751767759282e-05, "loss": 0.0061, "step": 2110 }, { "epoch": 11.216931216931217, "grad_norm": 0.14999443292617798, "learning_rate": 9.300339632356325e-05, "loss": 0.0065, "step": 2120 }, { "epoch": 11.26984126984127, "grad_norm": 0.15272104740142822, "learning_rate": 9.291880469128124e-05, "loss": 0.0062, "step": 2130 }, { "epoch": 11.322751322751323, "grad_norm": 0.164956733584404, "learning_rate": 9.283374370582732e-05, "loss": 0.0069, "step": 2140 }, { "epoch": 11.375661375661375, "grad_norm": 0.24659952521324158, "learning_rate": 9.274821429741482e-05, "loss": 0.0078, "step": 2150 }, { "epoch": 11.428571428571429, "grad_norm": 0.1309472769498825, "learning_rate": 9.266221740137961e-05, "loss": 0.0094, "step": 2160 }, { "epoch": 11.481481481481481, "grad_norm": 0.2406492531299591, "learning_rate": 9.257575395817001e-05, "loss": 0.0088, "step": 2170 }, { "epoch": 11.534391534391535, "grad_norm": 0.18980859220027924, "learning_rate": 9.248882491333637e-05, "loss": 0.0087, "step": 2180 }, { "epoch": 11.587301587301587, "grad_norm": 0.16797028481960297, "learning_rate": 9.240143121752076e-05, "loss": 0.0069, "step": 2190 }, { "epoch": 11.640211640211641, "grad_norm": 0.17111560702323914, "learning_rate": 9.23135738264467e-05, "loss": 0.0061, "step": 2200 }, { "epoch": 11.693121693121693, "grad_norm": 0.17521721124649048, "learning_rate": 9.222525370090849e-05, "loss": 0.0085, "step": 2210 }, { "epoch": 11.746031746031747, "grad_norm": 0.19105973839759827, "learning_rate": 9.213647180676088e-05, "loss": 0.0075, "step": 2220 }, { "epoch": 11.798941798941799, "grad_norm": 0.17100267112255096, "learning_rate": 9.204722911490846e-05, "loss": 0.0069, "step": 2230 }, { "epoch": 11.851851851851851, "grad_norm": 0.17743617296218872, "learning_rate": 9.1957526601295e-05, "loss": 0.0093, "step": 2240 }, { "epoch": 11.904761904761905, "grad_norm": 0.18579012155532837, "learning_rate": 9.186736524689281e-05, "loss": 0.0073, "step": 2250 }, { "epoch": 11.957671957671957, "grad_norm": 0.1838773638010025, "learning_rate": 9.177674603769204e-05, "loss": 0.0076, "step": 2260 }, { "epoch": 12.010582010582011, "grad_norm": 0.11034994572401047, "learning_rate": 9.168566996468983e-05, "loss": 0.0085, "step": 2270 }, { "epoch": 12.063492063492063, "grad_norm": 0.1641353815793991, "learning_rate": 9.159413802387951e-05, "loss": 0.0077, "step": 2280 }, { "epoch": 12.116402116402117, "grad_norm": 0.2112085074186325, "learning_rate": 9.150215121623974e-05, "loss": 0.007, "step": 2290 }, { "epoch": 12.16931216931217, "grad_norm": 0.20550212264060974, "learning_rate": 9.140971054772349e-05, "loss": 0.0085, "step": 2300 }, { "epoch": 12.222222222222221, "grad_norm": 0.183491513133049, "learning_rate": 9.131681702924713e-05, "loss": 0.0072, "step": 2310 }, { "epoch": 12.275132275132275, "grad_norm": 0.1910405308008194, "learning_rate": 9.122347167667926e-05, "loss": 0.007, "step": 2320 }, { "epoch": 12.328042328042327, "grad_norm": 0.16887152194976807, "learning_rate": 9.112967551082973e-05, "loss": 0.0074, "step": 2330 }, { "epoch": 12.380952380952381, "grad_norm": 0.17609399557113647, "learning_rate": 9.103542955743835e-05, "loss": 0.0072, "step": 2340 }, { "epoch": 12.433862433862434, "grad_norm": 0.19146792590618134, "learning_rate": 9.094073484716381e-05, "loss": 0.0068, "step": 2350 }, { "epoch": 12.486772486772487, "grad_norm": 0.15603068470954895, "learning_rate": 9.084559241557226e-05, "loss": 0.0085, "step": 2360 }, { "epoch": 12.53968253968254, "grad_norm": 0.17193616926670074, "learning_rate": 9.075000330312608e-05, "loss": 0.0065, "step": 2370 }, { "epoch": 12.592592592592592, "grad_norm": 0.14418815076351166, "learning_rate": 9.065396855517253e-05, "loss": 0.0098, "step": 2380 }, { "epoch": 12.645502645502646, "grad_norm": 0.1780521124601364, "learning_rate": 9.055748922193219e-05, "loss": 0.007, "step": 2390 }, { "epoch": 12.698412698412698, "grad_norm": 0.21523050963878632, "learning_rate": 9.046056635848761e-05, "loss": 0.0095, "step": 2400 }, { "epoch": 12.751322751322752, "grad_norm": 0.1339305192232132, "learning_rate": 9.036320102477169e-05, "loss": 0.0089, "step": 2410 }, { "epoch": 12.804232804232804, "grad_norm": 0.1643170416355133, "learning_rate": 9.02653942855561e-05, "loss": 0.0071, "step": 2420 }, { "epoch": 12.857142857142858, "grad_norm": 0.15023145079612732, "learning_rate": 9.016714721043971e-05, "loss": 0.007, "step": 2430 }, { "epoch": 12.91005291005291, "grad_norm": 0.1806618571281433, "learning_rate": 9.006846087383675e-05, "loss": 0.0089, "step": 2440 }, { "epoch": 12.962962962962964, "grad_norm": 0.20346465706825256, "learning_rate": 8.996933635496523e-05, "loss": 0.0077, "step": 2450 }, { "epoch": 13.015873015873016, "grad_norm": 0.18316155672073364, "learning_rate": 8.986977473783498e-05, "loss": 0.0066, "step": 2460 }, { "epoch": 13.068783068783068, "grad_norm": 0.16884809732437134, "learning_rate": 8.97697771112359e-05, "loss": 0.0074, "step": 2470 }, { "epoch": 13.121693121693122, "grad_norm": 0.201015904545784, "learning_rate": 8.966934456872602e-05, "loss": 0.0075, "step": 2480 }, { "epoch": 13.174603174603174, "grad_norm": 0.2114284187555313, "learning_rate": 8.95684782086195e-05, "loss": 0.0064, "step": 2490 }, { "epoch": 13.227513227513228, "grad_norm": 0.1519557535648346, "learning_rate": 8.946717913397476e-05, "loss": 0.0068, "step": 2500 }, { "epoch": 13.28042328042328, "grad_norm": 0.16507606208324432, "learning_rate": 8.93654484525822e-05, "loss": 0.0064, "step": 2510 }, { "epoch": 13.333333333333334, "grad_norm": 0.16865520179271698, "learning_rate": 8.926328727695226e-05, "loss": 0.0071, "step": 2520 }, { "epoch": 13.386243386243386, "grad_norm": 0.17065347731113434, "learning_rate": 8.916069672430319e-05, "loss": 0.0076, "step": 2530 }, { "epoch": 13.43915343915344, "grad_norm": 0.15293556451797485, "learning_rate": 8.905767791654884e-05, "loss": 0.0071, "step": 2540 }, { "epoch": 13.492063492063492, "grad_norm": 0.19155821204185486, "learning_rate": 8.895423198028638e-05, "loss": 0.0064, "step": 2550 }, { "epoch": 13.544973544973544, "grad_norm": 0.16175182163715363, "learning_rate": 8.885036004678402e-05, "loss": 0.0064, "step": 2560 }, { "epoch": 13.597883597883598, "grad_norm": 0.15450112521648407, "learning_rate": 8.874606325196857e-05, "loss": 0.0064, "step": 2570 }, { "epoch": 13.65079365079365, "grad_norm": 0.14070011675357819, "learning_rate": 8.864134273641304e-05, "loss": 0.007, "step": 2580 }, { "epoch": 13.703703703703704, "grad_norm": 0.14170874655246735, "learning_rate": 8.853619964532427e-05, "loss": 0.0064, "step": 2590 }, { "epoch": 13.756613756613756, "grad_norm": 0.20005124807357788, "learning_rate": 8.843063512853019e-05, "loss": 0.0078, "step": 2600 }, { "epoch": 13.80952380952381, "grad_norm": 0.16936667263507843, "learning_rate": 8.832465034046749e-05, "loss": 0.0067, "step": 2610 }, { "epoch": 13.862433862433862, "grad_norm": 0.20712535083293915, "learning_rate": 8.821824644016882e-05, "loss": 0.0072, "step": 2620 }, { "epoch": 13.915343915343914, "grad_norm": 0.18267923593521118, "learning_rate": 8.811142459125019e-05, "loss": 0.0072, "step": 2630 }, { "epoch": 13.968253968253968, "grad_norm": 0.16764400899410248, "learning_rate": 8.800418596189822e-05, "loss": 0.0071, "step": 2640 }, { "epoch": 14.02116402116402, "grad_norm": 0.24446216225624084, "learning_rate": 8.789653172485737e-05, "loss": 0.0076, "step": 2650 }, { "epoch": 14.074074074074074, "grad_norm": 0.17392943799495697, "learning_rate": 8.778846305741715e-05, "loss": 0.0055, "step": 2660 }, { "epoch": 14.126984126984127, "grad_norm": 0.22672197222709656, "learning_rate": 8.767998114139918e-05, "loss": 0.0086, "step": 2670 }, { "epoch": 14.17989417989418, "grad_norm": 0.17057965695858002, "learning_rate": 8.757108716314429e-05, "loss": 0.0088, "step": 2680 }, { "epoch": 14.232804232804233, "grad_norm": 0.1652648150920868, "learning_rate": 8.746178231349962e-05, "loss": 0.0055, "step": 2690 }, { "epoch": 14.285714285714286, "grad_norm": 0.14276280999183655, "learning_rate": 8.735206778780549e-05, "loss": 0.0056, "step": 2700 }, { "epoch": 14.338624338624339, "grad_norm": 0.17826801538467407, "learning_rate": 8.724194478588234e-05, "loss": 0.0059, "step": 2710 }, { "epoch": 14.39153439153439, "grad_norm": 0.1443377435207367, "learning_rate": 8.713141451201772e-05, "loss": 0.0078, "step": 2720 }, { "epoch": 14.444444444444445, "grad_norm": 0.15001343190670013, "learning_rate": 8.702047817495295e-05, "loss": 0.0083, "step": 2730 }, { "epoch": 14.497354497354497, "grad_norm": 0.18860018253326416, "learning_rate": 8.69091369878701e-05, "loss": 0.0055, "step": 2740 }, { "epoch": 14.55026455026455, "grad_norm": 0.1230684444308281, "learning_rate": 8.679739216837849e-05, "loss": 0.0059, "step": 2750 }, { "epoch": 14.603174603174603, "grad_norm": 0.13554492592811584, "learning_rate": 8.66852449385016e-05, "loss": 0.0069, "step": 2760 }, { "epoch": 14.656084656084657, "grad_norm": 0.1218063160777092, "learning_rate": 8.657269652466356e-05, "loss": 0.0051, "step": 2770 }, { "epoch": 14.708994708994709, "grad_norm": 0.12417104095220566, "learning_rate": 8.645974815767577e-05, "loss": 0.0057, "step": 2780 }, { "epoch": 14.761904761904763, "grad_norm": 0.17231541872024536, "learning_rate": 8.634640107272351e-05, "loss": 0.0054, "step": 2790 }, { "epoch": 14.814814814814815, "grad_norm": 0.16462373733520508, "learning_rate": 8.623265650935234e-05, "loss": 0.0051, "step": 2800 }, { "epoch": 14.867724867724867, "grad_norm": 0.24479028582572937, "learning_rate": 8.611851571145456e-05, "loss": 0.0076, "step": 2810 }, { "epoch": 14.920634920634921, "grad_norm": 0.18936169147491455, "learning_rate": 8.600397992725566e-05, "loss": 0.0082, "step": 2820 }, { "epoch": 14.973544973544973, "grad_norm": 0.20124199986457825, "learning_rate": 8.588905040930061e-05, "loss": 0.0076, "step": 2830 }, { "epoch": 15.026455026455027, "grad_norm": 0.19626690447330475, "learning_rate": 8.577372841444022e-05, "loss": 0.0091, "step": 2840 }, { "epoch": 15.079365079365079, "grad_norm": 0.1948651373386383, "learning_rate": 8.565801520381736e-05, "loss": 0.0068, "step": 2850 }, { "epoch": 15.132275132275133, "grad_norm": 0.18579399585723877, "learning_rate": 8.554191204285313e-05, "loss": 0.0062, "step": 2860 }, { "epoch": 15.185185185185185, "grad_norm": 0.1961190104484558, "learning_rate": 8.542542020123315e-05, "loss": 0.0064, "step": 2870 }, { "epoch": 15.238095238095237, "grad_norm": 0.17042799293994904, "learning_rate": 8.530854095289347e-05, "loss": 0.0074, "step": 2880 }, { "epoch": 15.291005291005291, "grad_norm": 0.1990518718957901, "learning_rate": 8.519127557600688e-05, "loss": 0.0079, "step": 2890 }, { "epoch": 15.343915343915343, "grad_norm": 0.23110555112361908, "learning_rate": 8.507362535296871e-05, "loss": 0.0064, "step": 2900 }, { "epoch": 15.396825396825397, "grad_norm": 0.29159578680992126, "learning_rate": 8.495559157038299e-05, "loss": 0.0065, "step": 2910 }, { "epoch": 15.44973544973545, "grad_norm": 0.21433016657829285, "learning_rate": 8.483717551904823e-05, "loss": 0.0059, "step": 2920 }, { "epoch": 15.502645502645503, "grad_norm": 0.1564585566520691, "learning_rate": 8.47183784939434e-05, "loss": 0.0061, "step": 2930 }, { "epoch": 15.555555555555555, "grad_norm": 0.19648239016532898, "learning_rate": 8.459920179421374e-05, "loss": 0.0059, "step": 2940 }, { "epoch": 15.60846560846561, "grad_norm": 0.1594851016998291, "learning_rate": 8.447964672315656e-05, "loss": 0.0061, "step": 2950 }, { "epoch": 15.661375661375661, "grad_norm": 0.11531036347150803, "learning_rate": 8.435971458820692e-05, "loss": 0.007, "step": 2960 }, { "epoch": 15.714285714285714, "grad_norm": 0.12807506322860718, "learning_rate": 8.423940670092345e-05, "loss": 0.0062, "step": 2970 }, { "epoch": 15.767195767195767, "grad_norm": 0.10728122293949127, "learning_rate": 8.411872437697394e-05, "loss": 0.0087, "step": 2980 }, { "epoch": 15.82010582010582, "grad_norm": 0.13496579229831696, "learning_rate": 8.399766893612096e-05, "loss": 0.0069, "step": 2990 }, { "epoch": 15.873015873015873, "grad_norm": 0.1787007600069046, "learning_rate": 8.38762417022074e-05, "loss": 0.0063, "step": 3000 }, { "epoch": 15.925925925925926, "grad_norm": 0.12794695794582367, "learning_rate": 8.375444400314204e-05, "loss": 0.0081, "step": 3010 }, { "epoch": 15.97883597883598, "grad_norm": 0.15977424383163452, "learning_rate": 8.3632277170885e-05, "loss": 0.0064, "step": 3020 }, { "epoch": 16.03174603174603, "grad_norm": 0.12448996305465698, "learning_rate": 8.350974254143318e-05, "loss": 0.0066, "step": 3030 }, { "epoch": 16.084656084656086, "grad_norm": 0.17275062203407288, "learning_rate": 8.338684145480566e-05, "loss": 0.0059, "step": 3040 }, { "epoch": 16.137566137566136, "grad_norm": 0.11005474627017975, "learning_rate": 8.326357525502904e-05, "loss": 0.0062, "step": 3050 }, { "epoch": 16.19047619047619, "grad_norm": 0.1260443478822708, "learning_rate": 8.313994529012273e-05, "loss": 0.0064, "step": 3060 }, { "epoch": 16.243386243386244, "grad_norm": 0.15497419238090515, "learning_rate": 8.301595291208422e-05, "loss": 0.007, "step": 3070 }, { "epoch": 16.296296296296298, "grad_norm": 0.15739837288856506, "learning_rate": 8.289159947687427e-05, "loss": 0.0051, "step": 3080 }, { "epoch": 16.349206349206348, "grad_norm": 0.13926488161087036, "learning_rate": 8.276688634440216e-05, "loss": 0.0054, "step": 3090 }, { "epoch": 16.402116402116402, "grad_norm": 0.19562622904777527, "learning_rate": 8.26418148785107e-05, "loss": 0.0107, "step": 3100 }, { "epoch": 16.455026455026456, "grad_norm": 0.22759263217449188, "learning_rate": 8.251638644696141e-05, "loss": 0.0087, "step": 3110 }, { "epoch": 16.507936507936506, "grad_norm": 0.19835254549980164, "learning_rate": 8.23906024214195e-05, "loss": 0.0074, "step": 3120 }, { "epoch": 16.56084656084656, "grad_norm": 0.1921726018190384, "learning_rate": 8.226446417743897e-05, "loss": 0.007, "step": 3130 }, { "epoch": 16.613756613756614, "grad_norm": 0.24273419380187988, "learning_rate": 8.213797309444742e-05, "loss": 0.007, "step": 3140 }, { "epoch": 16.666666666666668, "grad_norm": 0.2878647744655609, "learning_rate": 8.201113055573105e-05, "loss": 0.0059, "step": 3150 }, { "epoch": 16.719576719576718, "grad_norm": 0.15777450799942017, "learning_rate": 8.188393794841958e-05, "loss": 0.0058, "step": 3160 }, { "epoch": 16.772486772486772, "grad_norm": 0.15833404660224915, "learning_rate": 8.175639666347094e-05, "loss": 0.007, "step": 3170 }, { "epoch": 16.825396825396826, "grad_norm": 0.11987781524658203, "learning_rate": 8.162850809565623e-05, "loss": 0.0074, "step": 3180 }, { "epoch": 16.87830687830688, "grad_norm": 0.1330547034740448, "learning_rate": 8.150027364354431e-05, "loss": 0.0075, "step": 3190 }, { "epoch": 16.93121693121693, "grad_norm": 0.25050437450408936, "learning_rate": 8.137169470948662e-05, "loss": 0.0066, "step": 3200 }, { "epoch": 16.984126984126984, "grad_norm": 0.17232587933540344, "learning_rate": 8.124277269960179e-05, "loss": 0.0075, "step": 3210 }, { "epoch": 17.037037037037038, "grad_norm": 0.2330465465784073, "learning_rate": 8.111350902376023e-05, "loss": 0.0069, "step": 3220 }, { "epoch": 17.08994708994709, "grad_norm": 0.18809424340724945, "learning_rate": 8.098390509556883e-05, "loss": 0.0067, "step": 3230 }, { "epoch": 17.142857142857142, "grad_norm": 0.15625116229057312, "learning_rate": 8.085396233235536e-05, "loss": 0.0059, "step": 3240 }, { "epoch": 17.195767195767196, "grad_norm": 0.1564536839723587, "learning_rate": 8.072368215515306e-05, "loss": 0.0059, "step": 3250 }, { "epoch": 17.24867724867725, "grad_norm": 0.13552048802375793, "learning_rate": 8.059306598868506e-05, "loss": 0.0066, "step": 3260 }, { "epoch": 17.3015873015873, "grad_norm": 0.26676738262176514, "learning_rate": 8.046211526134888e-05, "loss": 0.0074, "step": 3270 }, { "epoch": 17.354497354497354, "grad_norm": 0.13536971807479858, "learning_rate": 8.033083140520065e-05, "loss": 0.008, "step": 3280 }, { "epoch": 17.40740740740741, "grad_norm": 0.1637164056301117, "learning_rate": 8.019921585593962e-05, "loss": 0.0057, "step": 3290 }, { "epoch": 17.46031746031746, "grad_norm": 0.15398825705051422, "learning_rate": 8.006727005289232e-05, "loss": 0.006, "step": 3300 }, { "epoch": 17.513227513227513, "grad_norm": 0.20822389423847198, "learning_rate": 7.993499543899692e-05, "loss": 0.0047, "step": 3310 }, { "epoch": 17.566137566137566, "grad_norm": 0.1953025460243225, "learning_rate": 7.980239346078742e-05, "loss": 0.0084, "step": 3320 }, { "epoch": 17.61904761904762, "grad_norm": 0.12149490416049957, "learning_rate": 7.966946556837778e-05, "loss": 0.0073, "step": 3330 }, { "epoch": 17.67195767195767, "grad_norm": 0.1526612490415573, "learning_rate": 7.953621321544616e-05, "loss": 0.0067, "step": 3340 }, { "epoch": 17.724867724867725, "grad_norm": 0.19760744273662567, "learning_rate": 7.940263785921896e-05, "loss": 0.006, "step": 3350 }, { "epoch": 17.77777777777778, "grad_norm": 0.16924995183944702, "learning_rate": 7.926874096045482e-05, "loss": 0.0055, "step": 3360 }, { "epoch": 17.83068783068783, "grad_norm": 0.1346261352300644, "learning_rate": 7.913452398342881e-05, "loss": 0.007, "step": 3370 }, { "epoch": 17.883597883597883, "grad_norm": 0.16452911496162415, "learning_rate": 7.89999883959163e-05, "loss": 0.0059, "step": 3380 }, { "epoch": 17.936507936507937, "grad_norm": 0.1587088704109192, "learning_rate": 7.886513566917687e-05, "loss": 0.0059, "step": 3390 }, { "epoch": 17.98941798941799, "grad_norm": 0.15406270325183868, "learning_rate": 7.872996727793838e-05, "loss": 0.0054, "step": 3400 }, { "epoch": 18.04232804232804, "grad_norm": 0.1580214500427246, "learning_rate": 7.859448470038069e-05, "loss": 0.0057, "step": 3410 }, { "epoch": 18.095238095238095, "grad_norm": 0.11246869713068008, "learning_rate": 7.845868941811956e-05, "loss": 0.0059, "step": 3420 }, { "epoch": 18.14814814814815, "grad_norm": 0.12457894533872604, "learning_rate": 7.832258291619043e-05, "loss": 0.0061, "step": 3430 }, { "epoch": 18.201058201058203, "grad_norm": 0.15718570351600647, "learning_rate": 7.81861666830322e-05, "loss": 0.0062, "step": 3440 }, { "epoch": 18.253968253968253, "grad_norm": 0.12985922396183014, "learning_rate": 7.804944221047097e-05, "loss": 0.0053, "step": 3450 }, { "epoch": 18.306878306878307, "grad_norm": 0.1498313546180725, "learning_rate": 7.791241099370364e-05, "loss": 0.0063, "step": 3460 }, { "epoch": 18.35978835978836, "grad_norm": 0.1876976490020752, "learning_rate": 7.777507453128163e-05, "loss": 0.0048, "step": 3470 }, { "epoch": 18.41269841269841, "grad_norm": 0.21113644540309906, "learning_rate": 7.763743432509451e-05, "loss": 0.0056, "step": 3480 }, { "epoch": 18.465608465608465, "grad_norm": 0.2325858622789383, "learning_rate": 7.749949188035353e-05, "loss": 0.0079, "step": 3490 }, { "epoch": 18.51851851851852, "grad_norm": 0.212503582239151, "learning_rate": 7.736124870557516e-05, "loss": 0.0063, "step": 3500 }, { "epoch": 18.571428571428573, "grad_norm": 0.17650046944618225, "learning_rate": 7.722270631256459e-05, "loss": 0.0053, "step": 3510 }, { "epoch": 18.624338624338623, "grad_norm": 0.17883922159671783, "learning_rate": 7.708386621639925e-05, "loss": 0.005, "step": 3520 }, { "epoch": 18.677248677248677, "grad_norm": 0.17638792097568512, "learning_rate": 7.694472993541219e-05, "loss": 0.0055, "step": 3530 }, { "epoch": 18.73015873015873, "grad_norm": 0.12859579920768738, "learning_rate": 7.680529899117547e-05, "loss": 0.0085, "step": 3540 }, { "epoch": 18.78306878306878, "grad_norm": 0.15070609748363495, "learning_rate": 7.666557490848358e-05, "loss": 0.0044, "step": 3550 }, { "epoch": 18.835978835978835, "grad_norm": 0.09971506148576736, "learning_rate": 7.65255592153367e-05, "loss": 0.0048, "step": 3560 }, { "epoch": 18.88888888888889, "grad_norm": 0.12707024812698364, "learning_rate": 7.638525344292402e-05, "loss": 0.005, "step": 3570 }, { "epoch": 18.941798941798943, "grad_norm": 0.12521536648273468, "learning_rate": 7.624465912560697e-05, "loss": 0.0068, "step": 3580 }, { "epoch": 18.994708994708994, "grad_norm": 0.1582782119512558, "learning_rate": 7.610377780090249e-05, "loss": 0.0063, "step": 3590 }, { "epoch": 19.047619047619047, "grad_norm": 0.14891067147254944, "learning_rate": 7.596261100946618e-05, "loss": 0.0055, "step": 3600 }, { "epoch": 19.1005291005291, "grad_norm": 0.20990151166915894, "learning_rate": 7.582116029507542e-05, "loss": 0.0075, "step": 3610 }, { "epoch": 19.15343915343915, "grad_norm": 0.14947167038917542, "learning_rate": 7.56794272046126e-05, "loss": 0.0042, "step": 3620 }, { "epoch": 19.206349206349206, "grad_norm": 0.1930917501449585, "learning_rate": 7.55374132880481e-05, "loss": 0.0061, "step": 3630 }, { "epoch": 19.25925925925926, "grad_norm": 0.2090824395418167, "learning_rate": 7.539512009842333e-05, "loss": 0.0063, "step": 3640 }, { "epoch": 19.312169312169313, "grad_norm": 0.20530176162719727, "learning_rate": 7.525254919183382e-05, "loss": 0.0055, "step": 3650 }, { "epoch": 19.365079365079364, "grad_norm": 0.1702355295419693, "learning_rate": 7.510970212741215e-05, "loss": 0.0079, "step": 3660 }, { "epoch": 19.417989417989418, "grad_norm": 0.13657717406749725, "learning_rate": 7.496658046731096e-05, "loss": 0.0069, "step": 3670 }, { "epoch": 19.47089947089947, "grad_norm": 0.1847354769706726, "learning_rate": 7.482318577668578e-05, "loss": 0.006, "step": 3680 }, { "epoch": 19.523809523809526, "grad_norm": 0.12479028850793839, "learning_rate": 7.467951962367796e-05, "loss": 0.0059, "step": 3690 }, { "epoch": 19.576719576719576, "grad_norm": 0.2056093066930771, "learning_rate": 7.453558357939755e-05, "loss": 0.0062, "step": 3700 }, { "epoch": 19.62962962962963, "grad_norm": 0.1579756736755371, "learning_rate": 7.439137921790606e-05, "loss": 0.0067, "step": 3710 }, { "epoch": 19.682539682539684, "grad_norm": 0.11352311074733734, "learning_rate": 7.42469081161993e-05, "loss": 0.0051, "step": 3720 }, { "epoch": 19.735449735449734, "grad_norm": 0.15483702719211578, "learning_rate": 7.410217185419006e-05, "loss": 0.0066, "step": 3730 }, { "epoch": 19.788359788359788, "grad_norm": 0.13532264530658722, "learning_rate": 7.395717201469095e-05, "loss": 0.0059, "step": 3740 }, { "epoch": 19.841269841269842, "grad_norm": 0.17482222616672516, "learning_rate": 7.381191018339696e-05, "loss": 0.008, "step": 3750 }, { "epoch": 19.894179894179896, "grad_norm": 0.1695830523967743, "learning_rate": 7.36663879488682e-05, "loss": 0.0067, "step": 3760 }, { "epoch": 19.947089947089946, "grad_norm": 0.21444471180438995, "learning_rate": 7.352060690251254e-05, "loss": 0.0071, "step": 3770 }, { "epoch": 20.0, "grad_norm": 0.14765210449695587, "learning_rate": 7.337456863856811e-05, "loss": 0.0065, "step": 3780 }, { "epoch": 20.052910052910054, "grad_norm": 0.1627817302942276, "learning_rate": 7.3228274754086e-05, "loss": 0.006, "step": 3790 }, { "epoch": 20.105820105820104, "grad_norm": 0.14007431268692017, "learning_rate": 7.308172684891267e-05, "loss": 0.005, "step": 3800 }, { "epoch": 20.158730158730158, "grad_norm": 0.1487901657819748, "learning_rate": 7.293492652567255e-05, "loss": 0.0051, "step": 3810 }, { "epoch": 20.211640211640212, "grad_norm": 0.14542068541049957, "learning_rate": 7.278787538975043e-05, "loss": 0.0044, "step": 3820 }, { "epoch": 20.264550264550266, "grad_norm": 0.13495437800884247, "learning_rate": 7.2640575049274e-05, "loss": 0.0049, "step": 3830 }, { "epoch": 20.317460317460316, "grad_norm": 0.17136886715888977, "learning_rate": 7.249302711509616e-05, "loss": 0.0059, "step": 3840 }, { "epoch": 20.37037037037037, "grad_norm": 0.16152210533618927, "learning_rate": 7.23452332007775e-05, "loss": 0.0049, "step": 3850 }, { "epoch": 20.423280423280424, "grad_norm": 0.09862919896841049, "learning_rate": 7.219719492256858e-05, "loss": 0.005, "step": 3860 }, { "epoch": 20.476190476190474, "grad_norm": 0.11863570660352707, "learning_rate": 7.20489138993923e-05, "loss": 0.0051, "step": 3870 }, { "epoch": 20.52910052910053, "grad_norm": 0.17006359994411469, "learning_rate": 7.190039175282614e-05, "loss": 0.0051, "step": 3880 }, { "epoch": 20.582010582010582, "grad_norm": 0.13077835738658905, "learning_rate": 7.175163010708455e-05, "loss": 0.0066, "step": 3890 }, { "epoch": 20.634920634920636, "grad_norm": 0.11507479846477509, "learning_rate": 7.1602630589001e-05, "loss": 0.0048, "step": 3900 }, { "epoch": 20.687830687830687, "grad_norm": 0.1295529454946518, "learning_rate": 7.14533948280104e-05, "loss": 0.0043, "step": 3910 }, { "epoch": 20.74074074074074, "grad_norm": 0.15012744069099426, "learning_rate": 7.130392445613109e-05, "loss": 0.006, "step": 3920 }, { "epoch": 20.793650793650794, "grad_norm": 0.12511366605758667, "learning_rate": 7.115422110794711e-05, "loss": 0.005, "step": 3930 }, { "epoch": 20.84656084656085, "grad_norm": 0.10185888409614563, "learning_rate": 7.100428642059033e-05, "loss": 0.0054, "step": 3940 }, { "epoch": 20.8994708994709, "grad_norm": 0.14203502237796783, "learning_rate": 7.08541220337224e-05, "loss": 0.0057, "step": 3950 }, { "epoch": 20.952380952380953, "grad_norm": 0.18597184121608734, "learning_rate": 7.070372958951706e-05, "loss": 0.0061, "step": 3960 }, { "epoch": 21.005291005291006, "grad_norm": 0.1796630620956421, "learning_rate": 7.055311073264194e-05, "loss": 0.0058, "step": 3970 }, { "epoch": 21.058201058201057, "grad_norm": 0.20042525231838226, "learning_rate": 7.040226711024077e-05, "loss": 0.0057, "step": 3980 }, { "epoch": 21.11111111111111, "grad_norm": 0.15560920536518097, "learning_rate": 7.02512003719152e-05, "loss": 0.0045, "step": 3990 }, { "epoch": 21.164021164021165, "grad_norm": 0.11000833660364151, "learning_rate": 7.00999121697069e-05, "loss": 0.007, "step": 4000 }, { "epoch": 21.21693121693122, "grad_norm": 0.1113070547580719, "learning_rate": 6.99484041580794e-05, "loss": 0.0068, "step": 4010 }, { "epoch": 21.26984126984127, "grad_norm": 0.14219734072685242, "learning_rate": 6.979667799390004e-05, "loss": 0.0074, "step": 4020 }, { "epoch": 21.322751322751323, "grad_norm": 0.15107855200767517, "learning_rate": 6.964473533642185e-05, "loss": 0.0054, "step": 4030 }, { "epoch": 21.375661375661377, "grad_norm": 0.1830112338066101, "learning_rate": 6.949257784726539e-05, "loss": 0.005, "step": 4040 }, { "epoch": 21.428571428571427, "grad_norm": 0.25000301003456116, "learning_rate": 6.934020719040056e-05, "loss": 0.0052, "step": 4050 }, { "epoch": 21.48148148148148, "grad_norm": 0.1867487132549286, "learning_rate": 6.918762503212848e-05, "loss": 0.0055, "step": 4060 }, { "epoch": 21.534391534391535, "grad_norm": 0.1459341198205948, "learning_rate": 6.903483304106319e-05, "loss": 0.0051, "step": 4070 }, { "epoch": 21.58730158730159, "grad_norm": 0.22725768387317657, "learning_rate": 6.888183288811341e-05, "loss": 0.0061, "step": 4080 }, { "epoch": 21.64021164021164, "grad_norm": 0.22197817265987396, "learning_rate": 6.87286262464643e-05, "loss": 0.0057, "step": 4090 }, { "epoch": 21.693121693121693, "grad_norm": 0.14944735169410706, "learning_rate": 6.857521479155915e-05, "loss": 0.0052, "step": 4100 }, { "epoch": 21.746031746031747, "grad_norm": 0.166793555021286, "learning_rate": 6.842160020108104e-05, "loss": 0.0051, "step": 4110 }, { "epoch": 21.798941798941797, "grad_norm": 0.1651468276977539, "learning_rate": 6.826778415493455e-05, "loss": 0.0048, "step": 4120 }, { "epoch": 21.85185185185185, "grad_norm": 0.12476926296949387, "learning_rate": 6.811376833522729e-05, "loss": 0.0041, "step": 4130 }, { "epoch": 21.904761904761905, "grad_norm": 0.1487903743982315, "learning_rate": 6.795955442625159e-05, "loss": 0.0058, "step": 4140 }, { "epoch": 21.95767195767196, "grad_norm": 0.09915949404239655, "learning_rate": 6.780514411446608e-05, "loss": 0.0045, "step": 4150 }, { "epoch": 22.01058201058201, "grad_norm": 0.13605885207653046, "learning_rate": 6.765053908847716e-05, "loss": 0.0047, "step": 4160 }, { "epoch": 22.063492063492063, "grad_norm": 0.182969868183136, "learning_rate": 6.749574103902064e-05, "loss": 0.0055, "step": 4170 }, { "epoch": 22.116402116402117, "grad_norm": 0.1381661742925644, "learning_rate": 6.734075165894317e-05, "loss": 0.0068, "step": 4180 }, { "epoch": 22.16931216931217, "grad_norm": 0.12793253362178802, "learning_rate": 6.71855726431838e-05, "loss": 0.005, "step": 4190 }, { "epoch": 22.22222222222222, "grad_norm": 0.12822803854942322, "learning_rate": 6.703020568875538e-05, "loss": 0.0054, "step": 4200 }, { "epoch": 22.275132275132275, "grad_norm": 0.261322945356369, "learning_rate": 6.687465249472603e-05, "loss": 0.009, "step": 4210 }, { "epoch": 22.32804232804233, "grad_norm": 0.13759928941726685, "learning_rate": 6.671891476220055e-05, "loss": 0.0053, "step": 4220 }, { "epoch": 22.38095238095238, "grad_norm": 0.1728186309337616, "learning_rate": 6.656299419430183e-05, "loss": 0.0056, "step": 4230 }, { "epoch": 22.433862433862434, "grad_norm": 0.1503288298845291, "learning_rate": 6.640689249615223e-05, "loss": 0.0056, "step": 4240 }, { "epoch": 22.486772486772487, "grad_norm": 0.08736085146665573, "learning_rate": 6.625061137485491e-05, "loss": 0.0058, "step": 4250 }, { "epoch": 22.53968253968254, "grad_norm": 0.10541359335184097, "learning_rate": 6.609415253947517e-05, "loss": 0.0049, "step": 4260 }, { "epoch": 22.59259259259259, "grad_norm": 0.1013571172952652, "learning_rate": 6.593751770102178e-05, "loss": 0.0042, "step": 4270 }, { "epoch": 22.645502645502646, "grad_norm": 0.1516013741493225, "learning_rate": 6.578070857242823e-05, "loss": 0.0057, "step": 4280 }, { "epoch": 22.6984126984127, "grad_norm": 0.1406678706407547, "learning_rate": 6.562372686853402e-05, "loss": 0.0047, "step": 4290 }, { "epoch": 22.75132275132275, "grad_norm": 0.12260832637548447, "learning_rate": 6.546657430606593e-05, "loss": 0.0048, "step": 4300 }, { "epoch": 22.804232804232804, "grad_norm": 0.23470881581306458, "learning_rate": 6.530925260361918e-05, "loss": 0.0062, "step": 4310 }, { "epoch": 22.857142857142858, "grad_norm": 0.19619087874889374, "learning_rate": 6.515176348163871e-05, "loss": 0.0057, "step": 4320 }, { "epoch": 22.91005291005291, "grad_norm": 0.1563940942287445, "learning_rate": 6.499410866240032e-05, "loss": 0.0046, "step": 4330 }, { "epoch": 22.962962962962962, "grad_norm": 0.1268693059682846, "learning_rate": 6.48362898699919e-05, "loss": 0.0043, "step": 4340 }, { "epoch": 23.015873015873016, "grad_norm": 0.11784721910953522, "learning_rate": 6.467830883029443e-05, "loss": 0.0056, "step": 4350 }, { "epoch": 23.06878306878307, "grad_norm": 0.11858135461807251, "learning_rate": 6.452016727096326e-05, "loss": 0.0054, "step": 4360 }, { "epoch": 23.12169312169312, "grad_norm": 0.1471864879131317, "learning_rate": 6.436186692140916e-05, "loss": 0.0037, "step": 4370 }, { "epoch": 23.174603174603174, "grad_norm": 0.11541644483804703, "learning_rate": 6.420340951277938e-05, "loss": 0.0043, "step": 4380 }, { "epoch": 23.227513227513228, "grad_norm": 0.13914278149604797, "learning_rate": 6.404479677793874e-05, "loss": 0.0047, "step": 4390 }, { "epoch": 23.280423280423282, "grad_norm": 0.14987073838710785, "learning_rate": 6.388603045145075e-05, "loss": 0.0064, "step": 4400 }, { "epoch": 23.333333333333332, "grad_norm": 0.16585387289524078, "learning_rate": 6.372711226955843e-05, "loss": 0.006, "step": 4410 }, { "epoch": 23.386243386243386, "grad_norm": 0.14283493161201477, "learning_rate": 6.356804397016564e-05, "loss": 0.0054, "step": 4420 }, { "epoch": 23.43915343915344, "grad_norm": 0.15132249891757965, "learning_rate": 6.340882729281779e-05, "loss": 0.0069, "step": 4430 }, { "epoch": 23.49206349206349, "grad_norm": 0.1349671483039856, "learning_rate": 6.324946397868294e-05, "loss": 0.0039, "step": 4440 }, { "epoch": 23.544973544973544, "grad_norm": 0.21060624718666077, "learning_rate": 6.308995577053276e-05, "loss": 0.0055, "step": 4450 }, { "epoch": 23.597883597883598, "grad_norm": 0.1185762956738472, "learning_rate": 6.293030441272347e-05, "loss": 0.0042, "step": 4460 }, { "epoch": 23.650793650793652, "grad_norm": 0.16911469399929047, "learning_rate": 6.277051165117677e-05, "loss": 0.0066, "step": 4470 }, { "epoch": 23.703703703703702, "grad_norm": 0.15484260022640228, "learning_rate": 6.261057923336064e-05, "loss": 0.0056, "step": 4480 }, { "epoch": 23.756613756613756, "grad_norm": 0.19421520829200745, "learning_rate": 6.245050890827042e-05, "loss": 0.0046, "step": 4490 }, { "epoch": 23.80952380952381, "grad_norm": 0.1930636316537857, "learning_rate": 6.229030242640952e-05, "loss": 0.0047, "step": 4500 }, { "epoch": 23.862433862433864, "grad_norm": 0.1315583437681198, "learning_rate": 6.212996153977037e-05, "loss": 0.0075, "step": 4510 }, { "epoch": 23.915343915343914, "grad_norm": 0.19039970636367798, "learning_rate": 6.196948800181523e-05, "loss": 0.0051, "step": 4520 }, { "epoch": 23.96825396825397, "grad_norm": 0.16167831420898438, "learning_rate": 6.180888356745695e-05, "loss": 0.0045, "step": 4530 }, { "epoch": 24.021164021164022, "grad_norm": 0.09198671579360962, "learning_rate": 6.164814999303995e-05, "loss": 0.0043, "step": 4540 }, { "epoch": 24.074074074074073, "grad_norm": 0.11823570728302002, "learning_rate": 6.148728903632081e-05, "loss": 0.0049, "step": 4550 }, { "epoch": 24.126984126984127, "grad_norm": 0.15359170734882355, "learning_rate": 6.132630245644921e-05, "loss": 0.0045, "step": 4560 }, { "epoch": 24.17989417989418, "grad_norm": 0.1803666353225708, "learning_rate": 6.116519201394857e-05, "loss": 0.0056, "step": 4570 }, { "epoch": 24.232804232804234, "grad_norm": 0.20210173726081848, "learning_rate": 6.10039594706969e-05, "loss": 0.0071, "step": 4580 }, { "epoch": 24.285714285714285, "grad_norm": 0.2114834189414978, "learning_rate": 6.084260658990744e-05, "loss": 0.0048, "step": 4590 }, { "epoch": 24.33862433862434, "grad_norm": 0.13181151449680328, "learning_rate": 6.068113513610943e-05, "loss": 0.006, "step": 4600 }, { "epoch": 24.391534391534393, "grad_norm": 0.13241025805473328, "learning_rate": 6.0519546875128876e-05, "loss": 0.0049, "step": 4610 }, { "epoch": 24.444444444444443, "grad_norm": 0.09654827415943146, "learning_rate": 6.035784357406906e-05, "loss": 0.0053, "step": 4620 }, { "epoch": 24.497354497354497, "grad_norm": 0.12895146012306213, "learning_rate": 6.01960270012914e-05, "loss": 0.0039, "step": 4630 }, { "epoch": 24.55026455026455, "grad_norm": 0.08805560320615768, "learning_rate": 6.003409892639599e-05, "loss": 0.0044, "step": 4640 }, { "epoch": 24.603174603174605, "grad_norm": 0.13976435363292694, "learning_rate": 5.9872061120202336e-05, "loss": 0.005, "step": 4650 }, { "epoch": 24.656084656084655, "grad_norm": 0.13314101099967957, "learning_rate": 5.9709915354729914e-05, "loss": 0.0042, "step": 4660 }, { "epoch": 24.70899470899471, "grad_norm": 0.1388780176639557, "learning_rate": 5.9547663403178824e-05, "loss": 0.0056, "step": 4670 }, { "epoch": 24.761904761904763, "grad_norm": 0.10985314846038818, "learning_rate": 5.9385307039910445e-05, "loss": 0.0062, "step": 4680 }, { "epoch": 24.814814814814813, "grad_norm": 0.1164306178689003, "learning_rate": 5.922284804042792e-05, "loss": 0.005, "step": 4690 }, { "epoch": 24.867724867724867, "grad_norm": 0.10957598686218262, "learning_rate": 5.906028818135687e-05, "loss": 0.0044, "step": 4700 }, { "epoch": 24.92063492063492, "grad_norm": 0.09089328348636627, "learning_rate": 5.889762924042585e-05, "loss": 0.0039, "step": 4710 }, { "epoch": 24.973544973544975, "grad_norm": 0.11004319787025452, "learning_rate": 5.873487299644699e-05, "loss": 0.0045, "step": 4720 }, { "epoch": 25.026455026455025, "grad_norm": 0.07459110021591187, "learning_rate": 5.857202122929649e-05, "loss": 0.0055, "step": 4730 }, { "epoch": 25.07936507936508, "grad_norm": 0.12694111466407776, "learning_rate": 5.840907571989518e-05, "loss": 0.0046, "step": 4740 }, { "epoch": 25.132275132275133, "grad_norm": 0.16302569210529327, "learning_rate": 5.824603825018904e-05, "loss": 0.0049, "step": 4750 }, { "epoch": 25.185185185185187, "grad_norm": 0.15223564207553864, "learning_rate": 5.808291060312975e-05, "loss": 0.0056, "step": 4760 }, { "epoch": 25.238095238095237, "grad_norm": 0.12582972645759583, "learning_rate": 5.7919694562655083e-05, "loss": 0.0037, "step": 4770 }, { "epoch": 25.29100529100529, "grad_norm": 0.17052477598190308, "learning_rate": 5.775639191366954e-05, "loss": 0.0061, "step": 4780 }, { "epoch": 25.343915343915345, "grad_norm": 0.13960470259189606, "learning_rate": 5.75930044420247e-05, "loss": 0.0065, "step": 4790 }, { "epoch": 25.396825396825395, "grad_norm": 0.15274673700332642, "learning_rate": 5.74295339344998e-05, "loss": 0.0045, "step": 4800 }, { "epoch": 25.44973544973545, "grad_norm": 0.12514567375183105, "learning_rate": 5.726598217878211e-05, "loss": 0.0051, "step": 4810 }, { "epoch": 25.502645502645503, "grad_norm": 0.1788453757762909, "learning_rate": 5.71023509634474e-05, "loss": 0.0054, "step": 4820 }, { "epoch": 25.555555555555557, "grad_norm": 0.12110009789466858, "learning_rate": 5.693864207794049e-05, "loss": 0.004, "step": 4830 }, { "epoch": 25.608465608465607, "grad_norm": 0.15937726199626923, "learning_rate": 5.677485731255545e-05, "loss": 0.0041, "step": 4840 }, { "epoch": 25.66137566137566, "grad_norm": 0.16388045251369476, "learning_rate": 5.6610998458416296e-05, "loss": 0.0062, "step": 4850 }, { "epoch": 25.714285714285715, "grad_norm": 0.17359450459480286, "learning_rate": 5.644706730745716e-05, "loss": 0.0057, "step": 4860 }, { "epoch": 25.767195767195766, "grad_norm": 0.14966924488544464, "learning_rate": 5.628306565240287e-05, "loss": 0.0048, "step": 4870 }, { "epoch": 25.82010582010582, "grad_norm": 0.1611105352640152, "learning_rate": 5.611899528674923e-05, "loss": 0.0049, "step": 4880 }, { "epoch": 25.873015873015873, "grad_norm": 0.1644187569618225, "learning_rate": 5.595485800474349e-05, "loss": 0.0052, "step": 4890 }, { "epoch": 25.925925925925927, "grad_norm": 0.19155435264110565, "learning_rate": 5.579065560136467e-05, "loss": 0.0049, "step": 4900 }, { "epoch": 25.978835978835978, "grad_norm": 0.16636909544467926, "learning_rate": 5.562638987230392e-05, "loss": 0.0045, "step": 4910 }, { "epoch": 26.03174603174603, "grad_norm": 0.12824350595474243, "learning_rate": 5.546206261394498e-05, "loss": 0.0043, "step": 4920 }, { "epoch": 26.084656084656086, "grad_norm": 0.10657738149166107, "learning_rate": 5.529767562334437e-05, "loss": 0.004, "step": 4930 }, { "epoch": 26.137566137566136, "grad_norm": 0.09692290425300598, "learning_rate": 5.5133230698211926e-05, "loss": 0.0044, "step": 4940 }, { "epoch": 26.19047619047619, "grad_norm": 0.11259174346923828, "learning_rate": 5.496872963689096e-05, "loss": 0.0047, "step": 4950 }, { "epoch": 26.243386243386244, "grad_norm": 0.0912521556019783, "learning_rate": 5.4804174238338756e-05, "loss": 0.0045, "step": 4960 }, { "epoch": 26.296296296296298, "grad_norm": 0.12382809817790985, "learning_rate": 5.463956630210678e-05, "loss": 0.0038, "step": 4970 }, { "epoch": 26.349206349206348, "grad_norm": 0.10942231863737106, "learning_rate": 5.4474907628321046e-05, "loss": 0.0042, "step": 4980 }, { "epoch": 26.402116402116402, "grad_norm": 0.11941510438919067, "learning_rate": 5.431020001766244e-05, "loss": 0.0045, "step": 4990 }, { "epoch": 26.455026455026456, "grad_norm": 0.11378979682922363, "learning_rate": 5.4145445271346986e-05, "loss": 0.0043, "step": 5000 }, { "epoch": 26.507936507936506, "grad_norm": 0.1344377100467682, "learning_rate": 5.398064519110622e-05, "loss": 0.0073, "step": 5010 }, { "epoch": 26.56084656084656, "grad_norm": 0.1562003791332245, "learning_rate": 5.3815801579167394e-05, "loss": 0.0047, "step": 5020 }, { "epoch": 26.613756613756614, "grad_norm": 0.10757927596569061, "learning_rate": 5.365091623823382e-05, "loss": 0.0036, "step": 5030 }, { "epoch": 26.666666666666668, "grad_norm": 0.11341147124767303, "learning_rate": 5.348599097146521e-05, "loss": 0.0044, "step": 5040 }, { "epoch": 26.719576719576718, "grad_norm": 0.13083405792713165, "learning_rate": 5.3321027582457836e-05, "loss": 0.004, "step": 5050 }, { "epoch": 26.772486772486772, "grad_norm": 0.1499222069978714, "learning_rate": 5.315602787522491e-05, "loss": 0.0047, "step": 5060 }, { "epoch": 26.825396825396826, "grad_norm": 0.17251653969287872, "learning_rate": 5.299099365417678e-05, "loss": 0.0045, "step": 5070 }, { "epoch": 26.87830687830688, "grad_norm": 0.12605343759059906, "learning_rate": 5.2825926724101236e-05, "loss": 0.0048, "step": 5080 }, { "epoch": 26.93121693121693, "grad_norm": 0.14382816851139069, "learning_rate": 5.26608288901438e-05, "loss": 0.0091, "step": 5090 }, { "epoch": 26.984126984126984, "grad_norm": 0.1619909703731537, "learning_rate": 5.24957019577879e-05, "loss": 0.0057, "step": 5100 }, { "epoch": 27.037037037037038, "grad_norm": 0.1273384988307953, "learning_rate": 5.2330547732835266e-05, "loss": 0.0078, "step": 5110 }, { "epoch": 27.08994708994709, "grad_norm": 0.12588146328926086, "learning_rate": 5.2165368021385996e-05, "loss": 0.0052, "step": 5120 }, { "epoch": 27.142857142857142, "grad_norm": 0.19856271147727966, "learning_rate": 5.200016462981897e-05, "loss": 0.0054, "step": 5130 }, { "epoch": 27.195767195767196, "grad_norm": 0.1304505616426468, "learning_rate": 5.1834939364772015e-05, "loss": 0.0046, "step": 5140 }, { "epoch": 27.24867724867725, "grad_norm": 0.13570591807365417, "learning_rate": 5.166969403312214e-05, "loss": 0.0037, "step": 5150 }, { "epoch": 27.3015873015873, "grad_norm": 0.1991366147994995, "learning_rate": 5.1504430441965844e-05, "loss": 0.005, "step": 5160 }, { "epoch": 27.354497354497354, "grad_norm": 0.19475103914737701, "learning_rate": 5.133915039859923e-05, "loss": 0.0039, "step": 5170 }, { "epoch": 27.40740740740741, "grad_norm": 0.13262182474136353, "learning_rate": 5.1173855710498444e-05, "loss": 0.0035, "step": 5180 }, { "epoch": 27.46031746031746, "grad_norm": 0.1417936533689499, "learning_rate": 5.100854818529967e-05, "loss": 0.0042, "step": 5190 }, { "epoch": 27.513227513227513, "grad_norm": 0.08601425588130951, "learning_rate": 5.084322963077951e-05, "loss": 0.0042, "step": 5200 }, { "epoch": 27.566137566137566, "grad_norm": 0.19597119092941284, "learning_rate": 5.067790185483522e-05, "loss": 0.0046, "step": 5210 }, { "epoch": 27.61904761904762, "grad_norm": 0.14126229286193848, "learning_rate": 5.0512566665464844e-05, "loss": 0.0044, "step": 5220 }, { "epoch": 27.67195767195767, "grad_norm": 0.11521231383085251, "learning_rate": 5.034722587074755e-05, "loss": 0.0039, "step": 5230 }, { "epoch": 27.724867724867725, "grad_norm": 0.13098078966140747, "learning_rate": 5.018188127882375e-05, "loss": 0.005, "step": 5240 }, { "epoch": 27.77777777777778, "grad_norm": 0.08162485808134079, "learning_rate": 5.0016534697875417e-05, "loss": 0.004, "step": 5250 }, { "epoch": 27.83068783068783, "grad_norm": 0.10995332151651382, "learning_rate": 4.9851187936106294e-05, "loss": 0.0052, "step": 5260 }, { "epoch": 27.883597883597883, "grad_norm": 0.13082946836948395, "learning_rate": 4.968584280172206e-05, "loss": 0.0044, "step": 5270 }, { "epoch": 27.936507936507937, "grad_norm": 0.12464485317468643, "learning_rate": 4.95205011029106e-05, "loss": 0.0035, "step": 5280 }, { "epoch": 27.98941798941799, "grad_norm": 0.1584559828042984, "learning_rate": 4.935516464782227e-05, "loss": 0.0048, "step": 5290 }, { "epoch": 28.04232804232804, "grad_norm": 0.1352042853832245, "learning_rate": 4.918983524455003e-05, "loss": 0.0057, "step": 5300 }, { "epoch": 28.095238095238095, "grad_norm": 0.1795787215232849, "learning_rate": 4.9024514701109766e-05, "loss": 0.0041, "step": 5310 }, { "epoch": 28.14814814814815, "grad_norm": 0.17419162392616272, "learning_rate": 4.885920482542043e-05, "loss": 0.0057, "step": 5320 }, { "epoch": 28.201058201058203, "grad_norm": 0.08271908760070801, "learning_rate": 4.869390742528438e-05, "loss": 0.0042, "step": 5330 }, { "epoch": 28.253968253968253, "grad_norm": 0.08289580792188644, "learning_rate": 4.852862430836744e-05, "loss": 0.0043, "step": 5340 }, { "epoch": 28.306878306878307, "grad_norm": 0.13012798130512238, "learning_rate": 4.836335728217933e-05, "loss": 0.0038, "step": 5350 }, { "epoch": 28.35978835978836, "grad_norm": 0.11444774270057678, "learning_rate": 4.819810815405379e-05, "loss": 0.0032, "step": 5360 }, { "epoch": 28.41269841269841, "grad_norm": 0.10862157493829727, "learning_rate": 4.803287873112877e-05, "loss": 0.0045, "step": 5370 }, { "epoch": 28.465608465608465, "grad_norm": 0.16293944418430328, "learning_rate": 4.786767082032681e-05, "loss": 0.0035, "step": 5380 }, { "epoch": 28.51851851851852, "grad_norm": 0.10642692446708679, "learning_rate": 4.77024862283351e-05, "loss": 0.0035, "step": 5390 }, { "epoch": 28.571428571428573, "grad_norm": 0.11997364461421967, "learning_rate": 4.753732676158593e-05, "loss": 0.0035, "step": 5400 }, { "epoch": 28.624338624338623, "grad_norm": 0.10173264890909195, "learning_rate": 4.737219422623672e-05, "loss": 0.0037, "step": 5410 }, { "epoch": 28.677248677248677, "grad_norm": 0.12275197356939316, "learning_rate": 4.720709042815044e-05, "loss": 0.0034, "step": 5420 }, { "epoch": 28.73015873015873, "grad_norm": 0.09017963707447052, "learning_rate": 4.704201717287578e-05, "loss": 0.0059, "step": 5430 }, { "epoch": 28.78306878306878, "grad_norm": 0.14553631842136383, "learning_rate": 4.6876976265627404e-05, "loss": 0.0048, "step": 5440 }, { "epoch": 28.835978835978835, "grad_norm": 0.13454687595367432, "learning_rate": 4.671196951126626e-05, "loss": 0.004, "step": 5450 }, { "epoch": 28.88888888888889, "grad_norm": 0.13795746862888336, "learning_rate": 4.654699871427971e-05, "loss": 0.0033, "step": 5460 }, { "epoch": 28.941798941798943, "grad_norm": 0.11427441984415054, "learning_rate": 4.6382065678762034e-05, "loss": 0.0037, "step": 5470 }, { "epoch": 28.994708994708994, "grad_norm": 0.12775473296642303, "learning_rate": 4.6217172208394424e-05, "loss": 0.0037, "step": 5480 }, { "epoch": 29.047619047619047, "grad_norm": 0.12013862282037735, "learning_rate": 4.605232010642549e-05, "loss": 0.0042, "step": 5490 }, { "epoch": 29.1005291005291, "grad_norm": 0.08739227801561356, "learning_rate": 4.588751117565142e-05, "loss": 0.0035, "step": 5500 }, { "epoch": 29.15343915343915, "grad_norm": 0.12889230251312256, "learning_rate": 4.5722747218396214e-05, "loss": 0.0047, "step": 5510 }, { "epoch": 29.206349206349206, "grad_norm": 0.0809677243232727, "learning_rate": 4.5558030036492194e-05, "loss": 0.0043, "step": 5520 }, { "epoch": 29.25925925925926, "grad_norm": 0.12085974216461182, "learning_rate": 4.539336143125999e-05, "loss": 0.0035, "step": 5530 }, { "epoch": 29.312169312169313, "grad_norm": 0.1498689353466034, "learning_rate": 4.522874320348916e-05, "loss": 0.0048, "step": 5540 }, { "epoch": 29.365079365079364, "grad_norm": 0.12123317271471024, "learning_rate": 4.506417715341821e-05, "loss": 0.0041, "step": 5550 }, { "epoch": 29.417989417989418, "grad_norm": 0.09922231733798981, "learning_rate": 4.489966508071511e-05, "loss": 0.0049, "step": 5560 }, { "epoch": 29.47089947089947, "grad_norm": 0.16680902242660522, "learning_rate": 4.4735208784457575e-05, "loss": 0.0044, "step": 5570 }, { "epoch": 29.523809523809526, "grad_norm": 0.10641931742429733, "learning_rate": 4.457081006311325e-05, "loss": 0.0035, "step": 5580 }, { "epoch": 29.576719576719576, "grad_norm": 0.10134295374155045, "learning_rate": 4.440647071452027e-05, "loss": 0.0037, "step": 5590 }, { "epoch": 29.62962962962963, "grad_norm": 0.09935961663722992, "learning_rate": 4.424219253586737e-05, "loss": 0.0045, "step": 5600 }, { "epoch": 29.682539682539684, "grad_norm": 0.15346187353134155, "learning_rate": 4.407797732367443e-05, "loss": 0.0041, "step": 5610 }, { "epoch": 29.735449735449734, "grad_norm": 0.11558949202299118, "learning_rate": 4.391382687377268e-05, "loss": 0.0046, "step": 5620 }, { "epoch": 29.788359788359788, "grad_norm": 0.13318535685539246, "learning_rate": 4.374974298128512e-05, "loss": 0.0045, "step": 5630 }, { "epoch": 29.841269841269842, "grad_norm": 0.11273392289876938, "learning_rate": 4.358572744060699e-05, "loss": 0.0047, "step": 5640 }, { "epoch": 29.894179894179896, "grad_norm": 0.12689395248889923, "learning_rate": 4.342178204538588e-05, "loss": 0.0041, "step": 5650 }, { "epoch": 29.947089947089946, "grad_norm": 0.11726480722427368, "learning_rate": 4.325790858850241e-05, "loss": 0.003, "step": 5660 }, { "epoch": 30.0, "grad_norm": 0.148945152759552, "learning_rate": 4.309410886205043e-05, "loss": 0.0049, "step": 5670 }, { "epoch": 30.052910052910054, "grad_norm": 0.1381072849035263, "learning_rate": 4.293038465731752e-05, "loss": 0.0033, "step": 5680 }, { "epoch": 30.105820105820104, "grad_norm": 0.12408214807510376, "learning_rate": 4.276673776476533e-05, "loss": 0.004, "step": 5690 }, { "epoch": 30.158730158730158, "grad_norm": 0.09706781804561615, "learning_rate": 4.260316997401007e-05, "loss": 0.0045, "step": 5700 }, { "epoch": 30.211640211640212, "grad_norm": 0.11160382628440857, "learning_rate": 4.243968307380293e-05, "loss": 0.0034, "step": 5710 }, { "epoch": 30.264550264550266, "grad_norm": 0.08478042483329773, "learning_rate": 4.22762788520104e-05, "loss": 0.0033, "step": 5720 }, { "epoch": 30.317460317460316, "grad_norm": 0.09233059734106064, "learning_rate": 4.211295909559491e-05, "loss": 0.0045, "step": 5730 }, { "epoch": 30.37037037037037, "grad_norm": 0.12063518911600113, "learning_rate": 4.194972559059511e-05, "loss": 0.0034, "step": 5740 }, { "epoch": 30.423280423280424, "grad_norm": 0.17353114485740662, "learning_rate": 4.178658012210651e-05, "loss": 0.0058, "step": 5750 }, { "epoch": 30.476190476190474, "grad_norm": 0.15166908502578735, "learning_rate": 4.162352447426177e-05, "loss": 0.0048, "step": 5760 }, { "epoch": 30.52910052910053, "grad_norm": 0.08267221599817276, "learning_rate": 4.146056043021135e-05, "loss": 0.0033, "step": 5770 }, { "epoch": 30.582010582010582, "grad_norm": 0.09959477186203003, "learning_rate": 4.1297689772103944e-05, "loss": 0.0034, "step": 5780 }, { "epoch": 30.634920634920636, "grad_norm": 0.11318983137607574, "learning_rate": 4.113491428106694e-05, "loss": 0.0035, "step": 5790 }, { "epoch": 30.687830687830687, "grad_norm": 0.10690370947122574, "learning_rate": 4.0972235737187055e-05, "loss": 0.0036, "step": 5800 }, { "epoch": 30.74074074074074, "grad_norm": 0.12049101293087006, "learning_rate": 4.080965591949076e-05, "loss": 0.0048, "step": 5810 }, { "epoch": 30.793650793650794, "grad_norm": 0.14609123766422272, "learning_rate": 4.0647176605924924e-05, "loss": 0.0033, "step": 5820 }, { "epoch": 30.84656084656085, "grad_norm": 0.14186571538448334, "learning_rate": 4.0484799573337255e-05, "loss": 0.0045, "step": 5830 }, { "epoch": 30.8994708994709, "grad_norm": 0.1085847020149231, "learning_rate": 4.032252659745699e-05, "loss": 0.0032, "step": 5840 }, { "epoch": 30.952380952380953, "grad_norm": 0.10414460301399231, "learning_rate": 4.016035945287539e-05, "loss": 0.0034, "step": 5850 }, { "epoch": 31.005291005291006, "grad_norm": 0.11237084120512009, "learning_rate": 3.999829991302635e-05, "loss": 0.0044, "step": 5860 }, { "epoch": 31.058201058201057, "grad_norm": 0.1936102658510208, "learning_rate": 3.983634975016707e-05, "loss": 0.0049, "step": 5870 }, { "epoch": 31.11111111111111, "grad_norm": 0.13109681010246277, "learning_rate": 3.967451073535854e-05, "loss": 0.0036, "step": 5880 }, { "epoch": 31.164021164021165, "grad_norm": 0.15107570588588715, "learning_rate": 3.951278463844633e-05, "loss": 0.004, "step": 5890 }, { "epoch": 31.21693121693122, "grad_norm": 0.11617960035800934, "learning_rate": 3.935117322804111e-05, "loss": 0.0041, "step": 5900 }, { "epoch": 31.26984126984127, "grad_norm": 0.11706047505140305, "learning_rate": 3.918967827149938e-05, "loss": 0.0047, "step": 5910 }, { "epoch": 31.322751322751323, "grad_norm": 0.23876994848251343, "learning_rate": 3.9028301534904094e-05, "loss": 0.0046, "step": 5920 }, { "epoch": 31.375661375661377, "grad_norm": 0.11635509878396988, "learning_rate": 3.88670447830454e-05, "loss": 0.0039, "step": 5930 }, { "epoch": 31.428571428571427, "grad_norm": 0.10454533249139786, "learning_rate": 3.870590977940132e-05, "loss": 0.0034, "step": 5940 }, { "epoch": 31.48148148148148, "grad_norm": 0.12024228274822235, "learning_rate": 3.8544898286118404e-05, "loss": 0.0051, "step": 5950 }, { "epoch": 31.534391534391535, "grad_norm": 0.1343003511428833, "learning_rate": 3.838401206399257e-05, "loss": 0.0037, "step": 5960 }, { "epoch": 31.58730158730159, "grad_norm": 0.15652774274349213, "learning_rate": 3.822325287244975e-05, "loss": 0.004, "step": 5970 }, { "epoch": 31.64021164021164, "grad_norm": 0.11422162503004074, "learning_rate": 3.8062622469526725e-05, "loss": 0.0038, "step": 5980 }, { "epoch": 31.693121693121693, "grad_norm": 0.12600815296173096, "learning_rate": 3.790212261185183e-05, "loss": 0.0042, "step": 5990 }, { "epoch": 31.746031746031747, "grad_norm": 0.08925296366214752, "learning_rate": 3.7741755054625794e-05, "loss": 0.0052, "step": 6000 }, { "epoch": 31.798941798941797, "grad_norm": 0.12995077669620514, "learning_rate": 3.758152155160255e-05, "loss": 0.0045, "step": 6010 }, { "epoch": 31.85185185185185, "grad_norm": 0.09381058067083359, "learning_rate": 3.742142385506999e-05, "loss": 0.0031, "step": 6020 }, { "epoch": 31.904761904761905, "grad_norm": 0.14041894674301147, "learning_rate": 3.72614637158309e-05, "loss": 0.0039, "step": 6030 }, { "epoch": 31.95767195767196, "grad_norm": 0.10743750631809235, "learning_rate": 3.710164288318371e-05, "loss": 0.0035, "step": 6040 }, { "epoch": 32.01058201058201, "grad_norm": 0.18516381084918976, "learning_rate": 3.694196310490345e-05, "loss": 0.0044, "step": 6050 }, { "epoch": 32.06349206349206, "grad_norm": 0.13609465956687927, "learning_rate": 3.678242612722259e-05, "loss": 0.0041, "step": 6060 }, { "epoch": 32.116402116402114, "grad_norm": 0.13547858595848083, "learning_rate": 3.6623033694811953e-05, "loss": 0.0034, "step": 6070 }, { "epoch": 32.16931216931217, "grad_norm": 0.18592754006385803, "learning_rate": 3.6463787550761665e-05, "loss": 0.0056, "step": 6080 }, { "epoch": 32.22222222222222, "grad_norm": 0.08751697093248367, "learning_rate": 3.630468943656202e-05, "loss": 0.0044, "step": 6090 }, { "epoch": 32.27513227513227, "grad_norm": 0.09516023844480515, "learning_rate": 3.6145741092084523e-05, "loss": 0.0035, "step": 6100 }, { "epoch": 32.32804232804233, "grad_norm": 0.12655679881572723, "learning_rate": 3.598694425556278e-05, "loss": 0.0053, "step": 6110 }, { "epoch": 32.38095238095238, "grad_norm": 0.08323931694030762, "learning_rate": 3.58283006635736e-05, "loss": 0.0039, "step": 6120 }, { "epoch": 32.43386243386244, "grad_norm": 0.09887446463108063, "learning_rate": 3.566981205101781e-05, "loss": 0.0043, "step": 6130 }, { "epoch": 32.48677248677249, "grad_norm": 0.17184188961982727, "learning_rate": 3.5511480151101556e-05, "loss": 0.0037, "step": 6140 }, { "epoch": 32.53968253968254, "grad_norm": 0.15427406132221222, "learning_rate": 3.5353306695317104e-05, "loss": 0.0042, "step": 6150 }, { "epoch": 32.592592592592595, "grad_norm": 0.19764907658100128, "learning_rate": 3.519529341342402e-05, "loss": 0.0055, "step": 6160 }, { "epoch": 32.645502645502646, "grad_norm": 0.11758694052696228, "learning_rate": 3.503744203343026e-05, "loss": 0.0035, "step": 6170 }, { "epoch": 32.698412698412696, "grad_norm": 0.10999855399131775, "learning_rate": 3.487975428157318e-05, "loss": 0.0044, "step": 6180 }, { "epoch": 32.75132275132275, "grad_norm": 0.10582797974348068, "learning_rate": 3.472223188230083e-05, "loss": 0.0034, "step": 6190 }, { "epoch": 32.804232804232804, "grad_norm": 0.11475761234760284, "learning_rate": 3.4564876558252866e-05, "loss": 0.0036, "step": 6200 }, { "epoch": 32.857142857142854, "grad_norm": 0.08803584426641464, "learning_rate": 3.440769003024195e-05, "loss": 0.0033, "step": 6210 }, { "epoch": 32.91005291005291, "grad_norm": 0.15942472219467163, "learning_rate": 3.425067401723477e-05, "loss": 0.0044, "step": 6220 }, { "epoch": 32.96296296296296, "grad_norm": 0.15383343398571014, "learning_rate": 3.409383023633325e-05, "loss": 0.0043, "step": 6230 }, { "epoch": 33.01587301587302, "grad_norm": 0.11310908198356628, "learning_rate": 3.3937160402755894e-05, "loss": 0.0032, "step": 6240 }, { "epoch": 33.06878306878307, "grad_norm": 0.0895925760269165, "learning_rate": 3.378066622981885e-05, "loss": 0.0033, "step": 6250 }, { "epoch": 33.12169312169312, "grad_norm": 0.12127673625946045, "learning_rate": 3.362434942891738e-05, "loss": 0.003, "step": 6260 }, { "epoch": 33.17460317460318, "grad_norm": 0.08445215970277786, "learning_rate": 3.346821170950693e-05, "loss": 0.0037, "step": 6270 }, { "epoch": 33.22751322751323, "grad_norm": 0.07515180855989456, "learning_rate": 3.3312254779084585e-05, "loss": 0.0034, "step": 6280 }, { "epoch": 33.28042328042328, "grad_norm": 0.10749741643667221, "learning_rate": 3.315648034317039e-05, "loss": 0.0046, "step": 6290 }, { "epoch": 33.333333333333336, "grad_norm": 0.12727117538452148, "learning_rate": 3.3000890105288564e-05, "loss": 0.0044, "step": 6300 }, { "epoch": 33.386243386243386, "grad_norm": 0.1169172152876854, "learning_rate": 3.284548576694908e-05, "loss": 0.0056, "step": 6310 }, { "epoch": 33.439153439153436, "grad_norm": 0.14694686233997345, "learning_rate": 3.2690269027628815e-05, "loss": 0.0042, "step": 6320 }, { "epoch": 33.492063492063494, "grad_norm": 0.14343777298927307, "learning_rate": 3.253524158475324e-05, "loss": 0.0045, "step": 6330 }, { "epoch": 33.544973544973544, "grad_norm": 0.12558597326278687, "learning_rate": 3.238040513367757e-05, "loss": 0.0041, "step": 6340 }, { "epoch": 33.597883597883595, "grad_norm": 0.12548616528511047, "learning_rate": 3.222576136766843e-05, "loss": 0.0048, "step": 6350 }, { "epoch": 33.65079365079365, "grad_norm": 0.1444886028766632, "learning_rate": 3.2071311977885324e-05, "loss": 0.0036, "step": 6360 }, { "epoch": 33.7037037037037, "grad_norm": 0.11942130327224731, "learning_rate": 3.191705865336197e-05, "loss": 0.0042, "step": 6370 }, { "epoch": 33.75661375661376, "grad_norm": 0.16597044467926025, "learning_rate": 3.1763003080988075e-05, "loss": 0.0035, "step": 6380 }, { "epoch": 33.80952380952381, "grad_norm": 0.18611551821231842, "learning_rate": 3.160914694549063e-05, "loss": 0.0044, "step": 6390 }, { "epoch": 33.86243386243386, "grad_norm": 0.11668522655963898, "learning_rate": 3.145549192941573e-05, "loss": 0.0044, "step": 6400 }, { "epoch": 33.91534391534392, "grad_norm": 0.17806777358055115, "learning_rate": 3.130203971310999e-05, "loss": 0.0041, "step": 6410 }, { "epoch": 33.96825396825397, "grad_norm": 0.0968608483672142, "learning_rate": 3.114879197470225e-05, "loss": 0.0039, "step": 6420 }, { "epoch": 34.02116402116402, "grad_norm": 0.12741924822330475, "learning_rate": 3.0995750390085285e-05, "loss": 0.0047, "step": 6430 }, { "epoch": 34.074074074074076, "grad_norm": 0.07951290905475616, "learning_rate": 3.084291663289728e-05, "loss": 0.0035, "step": 6440 }, { "epoch": 34.12698412698413, "grad_norm": 0.11628319323062897, "learning_rate": 3.069029237450375e-05, "loss": 0.0044, "step": 6450 }, { "epoch": 34.17989417989418, "grad_norm": 0.1083105131983757, "learning_rate": 3.053787928397911e-05, "loss": 0.0029, "step": 6460 }, { "epoch": 34.232804232804234, "grad_norm": 0.0716942548751831, "learning_rate": 3.0385679028088526e-05, "loss": 0.0036, "step": 6470 }, { "epoch": 34.285714285714285, "grad_norm": 0.08762670308351517, "learning_rate": 3.023369327126959e-05, "loss": 0.0028, "step": 6480 }, { "epoch": 34.33862433862434, "grad_norm": 0.10415586084127426, "learning_rate": 3.0081923675614198e-05, "loss": 0.0029, "step": 6490 }, { "epoch": 34.39153439153439, "grad_norm": 0.058047302067279816, "learning_rate": 2.993037190085034e-05, "loss": 0.0028, "step": 6500 }, { "epoch": 34.44444444444444, "grad_norm": 0.10042303055524826, "learning_rate": 2.977903960432392e-05, "loss": 0.0035, "step": 6510 }, { "epoch": 34.4973544973545, "grad_norm": 0.09083425998687744, "learning_rate": 2.9627928440980722e-05, "loss": 0.003, "step": 6520 }, { "epoch": 34.55026455026455, "grad_norm": 0.07457281649112701, "learning_rate": 2.9477040063348183e-05, "loss": 0.0038, "step": 6530 }, { "epoch": 34.6031746031746, "grad_norm": 0.10889340937137604, "learning_rate": 2.9326376121517456e-05, "loss": 0.0031, "step": 6540 }, { "epoch": 34.65608465608466, "grad_norm": 0.13216157257556915, "learning_rate": 2.9175938263125236e-05, "loss": 0.0038, "step": 6550 }, { "epoch": 34.70899470899471, "grad_norm": 0.2216378003358841, "learning_rate": 2.9025728133335873e-05, "loss": 0.0033, "step": 6560 }, { "epoch": 34.76190476190476, "grad_norm": 0.16318024694919586, "learning_rate": 2.8875747374823288e-05, "loss": 0.0037, "step": 6570 }, { "epoch": 34.81481481481482, "grad_norm": 0.1352868676185608, "learning_rate": 2.872599762775298e-05, "loss": 0.0046, "step": 6580 }, { "epoch": 34.86772486772487, "grad_norm": 0.13937005400657654, "learning_rate": 2.857648052976425e-05, "loss": 0.0039, "step": 6590 }, { "epoch": 34.92063492063492, "grad_norm": 0.13573339581489563, "learning_rate": 2.8427197715952047e-05, "loss": 0.0034, "step": 6600 }, { "epoch": 34.973544973544975, "grad_norm": 0.1036810576915741, "learning_rate": 2.8278150818849393e-05, "loss": 0.0044, "step": 6610 }, { "epoch": 35.026455026455025, "grad_norm": 0.11498697102069855, "learning_rate": 2.812934146840922e-05, "loss": 0.0029, "step": 6620 }, { "epoch": 35.07936507936508, "grad_norm": 0.055324409157037735, "learning_rate": 2.7980771291986764e-05, "loss": 0.0021, "step": 6630 }, { "epoch": 35.13227513227513, "grad_norm": 0.07279566675424576, "learning_rate": 2.783244191432167e-05, "loss": 0.0034, "step": 6640 }, { "epoch": 35.18518518518518, "grad_norm": 0.08215716481208801, "learning_rate": 2.768435495752022e-05, "loss": 0.0028, "step": 6650 }, { "epoch": 35.23809523809524, "grad_norm": 0.09474048018455505, "learning_rate": 2.753651204103771e-05, "loss": 0.0046, "step": 6660 }, { "epoch": 35.29100529100529, "grad_norm": 0.09787704795598984, "learning_rate": 2.7388914781660523e-05, "loss": 0.003, "step": 6670 }, { "epoch": 35.34391534391534, "grad_norm": 0.12245072424411774, "learning_rate": 2.7241564793488693e-05, "loss": 0.0033, "step": 6680 }, { "epoch": 35.3968253968254, "grad_norm": 0.07514521479606628, "learning_rate": 2.7094463687918037e-05, "loss": 0.0029, "step": 6690 }, { "epoch": 35.44973544973545, "grad_norm": 0.0708879753947258, "learning_rate": 2.694761307362268e-05, "loss": 0.0038, "step": 6700 }, { "epoch": 35.5026455026455, "grad_norm": 0.07997647672891617, "learning_rate": 2.6801014556537467e-05, "loss": 0.0029, "step": 6710 }, { "epoch": 35.55555555555556, "grad_norm": 0.0956379771232605, "learning_rate": 2.6654669739840243e-05, "loss": 0.0033, "step": 6720 }, { "epoch": 35.60846560846561, "grad_norm": 0.09623092412948608, "learning_rate": 2.650858022393451e-05, "loss": 0.0049, "step": 6730 }, { "epoch": 35.66137566137566, "grad_norm": 0.10295127332210541, "learning_rate": 2.6362747606431747e-05, "loss": 0.0038, "step": 6740 }, { "epoch": 35.714285714285715, "grad_norm": 0.1539899706840515, "learning_rate": 2.6217173482134172e-05, "loss": 0.0029, "step": 6750 }, { "epoch": 35.767195767195766, "grad_norm": 0.07674749940633774, "learning_rate": 2.6071859443017044e-05, "loss": 0.003, "step": 6760 }, { "epoch": 35.82010582010582, "grad_norm": 0.0886496901512146, "learning_rate": 2.5926807078211414e-05, "loss": 0.0029, "step": 6770 }, { "epoch": 35.87301587301587, "grad_norm": 0.07741867005825043, "learning_rate": 2.5782017973986728e-05, "loss": 0.0033, "step": 6780 }, { "epoch": 35.925925925925924, "grad_norm": 0.14688631892204285, "learning_rate": 2.5637493713733374e-05, "loss": 0.003, "step": 6790 }, { "epoch": 35.97883597883598, "grad_norm": 0.09790076315402985, "learning_rate": 2.549323587794559e-05, "loss": 0.0028, "step": 6800 }, { "epoch": 36.03174603174603, "grad_norm": 0.06711734086275101, "learning_rate": 2.5349246044203895e-05, "loss": 0.0025, "step": 6810 }, { "epoch": 36.08465608465608, "grad_norm": 0.06354604661464691, "learning_rate": 2.520552578715808e-05, "loss": 0.0031, "step": 6820 }, { "epoch": 36.13756613756614, "grad_norm": 0.11155734211206436, "learning_rate": 2.506207667850981e-05, "loss": 0.0031, "step": 6830 }, { "epoch": 36.19047619047619, "grad_norm": 0.07272658497095108, "learning_rate": 2.4918900286995555e-05, "loss": 0.0034, "step": 6840 }, { "epoch": 36.24338624338624, "grad_norm": 0.11255387961864471, "learning_rate": 2.4775998178369458e-05, "loss": 0.0025, "step": 6850 }, { "epoch": 36.2962962962963, "grad_norm": 0.08785462379455566, "learning_rate": 2.4633371915386017e-05, "loss": 0.0036, "step": 6860 }, { "epoch": 36.34920634920635, "grad_norm": 0.09526315331459045, "learning_rate": 2.4491023057783235e-05, "loss": 0.0029, "step": 6870 }, { "epoch": 36.402116402116405, "grad_norm": 0.10234445333480835, "learning_rate": 2.4348953162265375e-05, "loss": 0.0028, "step": 6880 }, { "epoch": 36.455026455026456, "grad_norm": 0.06426072120666504, "learning_rate": 2.420716378248607e-05, "loss": 0.0031, "step": 6890 }, { "epoch": 36.507936507936506, "grad_norm": 0.0744519904255867, "learning_rate": 2.4065656469031266e-05, "loss": 0.0029, "step": 6900 }, { "epoch": 36.560846560846564, "grad_norm": 0.09663981944322586, "learning_rate": 2.3924432769402268e-05, "loss": 0.0037, "step": 6910 }, { "epoch": 36.613756613756614, "grad_norm": 0.13265518844127655, "learning_rate": 2.3783494227998844e-05, "loss": 0.0029, "step": 6920 }, { "epoch": 36.666666666666664, "grad_norm": 0.1177295446395874, "learning_rate": 2.3642842386102264e-05, "loss": 0.0033, "step": 6930 }, { "epoch": 36.71957671957672, "grad_norm": 0.11080149561166763, "learning_rate": 2.3502478781858567e-05, "loss": 0.0033, "step": 6940 }, { "epoch": 36.77248677248677, "grad_norm": 0.11145573109388351, "learning_rate": 2.3362404950261628e-05, "loss": 0.0029, "step": 6950 }, { "epoch": 36.82539682539682, "grad_norm": 0.11042343080043793, "learning_rate": 2.3222622423136458e-05, "loss": 0.0026, "step": 6960 }, { "epoch": 36.87830687830688, "grad_norm": 0.1116756871342659, "learning_rate": 2.3083132729122332e-05, "loss": 0.0028, "step": 6970 }, { "epoch": 36.93121693121693, "grad_norm": 0.12313203513622284, "learning_rate": 2.294393739365621e-05, "loss": 0.003, "step": 6980 }, { "epoch": 36.98412698412698, "grad_norm": 0.09837013483047485, "learning_rate": 2.2805037938956e-05, "loss": 0.0026, "step": 6990 }, { "epoch": 37.03703703703704, "grad_norm": 0.10853338986635208, "learning_rate": 2.266643588400386e-05, "loss": 0.0026, "step": 7000 }, { "epoch": 37.08994708994709, "grad_norm": 0.07267850637435913, "learning_rate": 2.252813274452969e-05, "loss": 0.0025, "step": 7010 }, { "epoch": 37.142857142857146, "grad_norm": 0.08111244440078735, "learning_rate": 2.2390130032994427e-05, "loss": 0.0024, "step": 7020 }, { "epoch": 37.195767195767196, "grad_norm": 0.10990393906831741, "learning_rate": 2.2252429258573633e-05, "loss": 0.0029, "step": 7030 }, { "epoch": 37.24867724867725, "grad_norm": 0.10033424198627472, "learning_rate": 2.2115031927140904e-05, "loss": 0.0043, "step": 7040 }, { "epoch": 37.301587301587304, "grad_norm": 0.08440393954515457, "learning_rate": 2.1977939541251463e-05, "loss": 0.0024, "step": 7050 }, { "epoch": 37.354497354497354, "grad_norm": 0.14266985654830933, "learning_rate": 2.1841153600125684e-05, "loss": 0.0033, "step": 7060 }, { "epoch": 37.407407407407405, "grad_norm": 0.11893881112337112, "learning_rate": 2.170467559963267e-05, "loss": 0.0029, "step": 7070 }, { "epoch": 37.46031746031746, "grad_norm": 0.09068603813648224, "learning_rate": 2.1568507032273982e-05, "loss": 0.0025, "step": 7080 }, { "epoch": 37.51322751322751, "grad_norm": 0.08722154796123505, "learning_rate": 2.1432649387167264e-05, "loss": 0.0033, "step": 7090 }, { "epoch": 37.56613756613756, "grad_norm": 0.11743749678134918, "learning_rate": 2.1297104150029973e-05, "loss": 0.0028, "step": 7100 }, { "epoch": 37.61904761904762, "grad_norm": 0.06725412607192993, "learning_rate": 2.116187280316307e-05, "loss": 0.0038, "step": 7110 }, { "epoch": 37.67195767195767, "grad_norm": 0.0862404853105545, "learning_rate": 2.1026956825434908e-05, "loss": 0.0026, "step": 7120 }, { "epoch": 37.72486772486773, "grad_norm": 0.07607479393482208, "learning_rate": 2.0892357692265017e-05, "loss": 0.0044, "step": 7130 }, { "epoch": 37.77777777777778, "grad_norm": 0.12202905863523483, "learning_rate": 2.0758076875607947e-05, "loss": 0.0049, "step": 7140 }, { "epoch": 37.83068783068783, "grad_norm": 0.09695997089147568, "learning_rate": 2.0624115843937207e-05, "loss": 0.0036, "step": 7150 }, { "epoch": 37.883597883597886, "grad_norm": 0.1274048089981079, "learning_rate": 2.0490476062229157e-05, "loss": 0.004, "step": 7160 }, { "epoch": 37.93650793650794, "grad_norm": 0.1332247108221054, "learning_rate": 2.035715899194704e-05, "loss": 0.0043, "step": 7170 }, { "epoch": 37.98941798941799, "grad_norm": 0.10100788623094559, "learning_rate": 2.022416609102499e-05, "loss": 0.004, "step": 7180 }, { "epoch": 38.042328042328045, "grad_norm": 0.13027310371398926, "learning_rate": 2.009149881385205e-05, "loss": 0.0032, "step": 7190 }, { "epoch": 38.095238095238095, "grad_norm": 0.10778859257698059, "learning_rate": 1.995915861125634e-05, "loss": 0.0036, "step": 7200 }, { "epoch": 38.148148148148145, "grad_norm": 0.07033434510231018, "learning_rate": 1.9827146930489065e-05, "loss": 0.0024, "step": 7210 }, { "epoch": 38.2010582010582, "grad_norm": 0.05678543075919151, "learning_rate": 1.9695465215208848e-05, "loss": 0.0023, "step": 7220 }, { "epoch": 38.25396825396825, "grad_norm": 0.10585637390613556, "learning_rate": 1.9564114905465813e-05, "loss": 0.0033, "step": 7230 }, { "epoch": 38.3068783068783, "grad_norm": 0.10460075736045837, "learning_rate": 1.9433097437685936e-05, "loss": 0.0028, "step": 7240 }, { "epoch": 38.35978835978836, "grad_norm": 0.12966494262218475, "learning_rate": 1.930241424465521e-05, "loss": 0.0035, "step": 7250 }, { "epoch": 38.41269841269841, "grad_norm": 0.10625789314508438, "learning_rate": 1.9172066755504115e-05, "loss": 0.0032, "step": 7260 }, { "epoch": 38.46560846560847, "grad_norm": 0.13831879198551178, "learning_rate": 1.9042056395691914e-05, "loss": 0.0029, "step": 7270 }, { "epoch": 38.51851851851852, "grad_norm": 0.12166795879602432, "learning_rate": 1.8912384586991066e-05, "loss": 0.0037, "step": 7280 }, { "epoch": 38.57142857142857, "grad_norm": 0.07098385691642761, "learning_rate": 1.8783052747471717e-05, "loss": 0.0023, "step": 7290 }, { "epoch": 38.62433862433863, "grad_norm": 0.0631115511059761, "learning_rate": 1.865406229148611e-05, "loss": 0.0025, "step": 7300 }, { "epoch": 38.67724867724868, "grad_norm": 0.09456410259008408, "learning_rate": 1.8525414629653233e-05, "loss": 0.0029, "step": 7310 }, { "epoch": 38.73015873015873, "grad_norm": 0.09200344979763031, "learning_rate": 1.8397111168843255e-05, "loss": 0.0027, "step": 7320 }, { "epoch": 38.783068783068785, "grad_norm": 0.09153919667005539, "learning_rate": 1.8269153312162323e-05, "loss": 0.0032, "step": 7330 }, { "epoch": 38.835978835978835, "grad_norm": 0.12154002487659454, "learning_rate": 1.8141542458937054e-05, "loss": 0.0044, "step": 7340 }, { "epoch": 38.888888888888886, "grad_norm": 0.07170013338327408, "learning_rate": 1.8014280004699268e-05, "loss": 0.0028, "step": 7350 }, { "epoch": 38.94179894179894, "grad_norm": 0.09592019766569138, "learning_rate": 1.788736734117078e-05, "loss": 0.0033, "step": 7360 }, { "epoch": 38.99470899470899, "grad_norm": 0.10066182166337967, "learning_rate": 1.7760805856248152e-05, "loss": 0.0025, "step": 7370 }, { "epoch": 39.04761904761905, "grad_norm": 0.0638168528676033, "learning_rate": 1.7634596933987518e-05, "loss": 0.0043, "step": 7380 }, { "epoch": 39.1005291005291, "grad_norm": 0.08771741390228271, "learning_rate": 1.7508741954589404e-05, "loss": 0.0027, "step": 7390 }, { "epoch": 39.15343915343915, "grad_norm": 0.08649582415819168, "learning_rate": 1.7383242294383717e-05, "loss": 0.0029, "step": 7400 }, { "epoch": 39.20634920634921, "grad_norm": 0.076836958527565, "learning_rate": 1.7258099325814632e-05, "loss": 0.0025, "step": 7410 }, { "epoch": 39.25925925925926, "grad_norm": 0.05890624597668648, "learning_rate": 1.7133314417425594e-05, "loss": 0.0029, "step": 7420 }, { "epoch": 39.31216931216931, "grad_norm": 0.06317563354969025, "learning_rate": 1.7008888933844408e-05, "loss": 0.0027, "step": 7430 }, { "epoch": 39.36507936507937, "grad_norm": 0.0679774358868599, "learning_rate": 1.6884824235768172e-05, "loss": 0.0027, "step": 7440 }, { "epoch": 39.41798941798942, "grad_norm": 0.05621592700481415, "learning_rate": 1.6761121679948592e-05, "loss": 0.0033, "step": 7450 }, { "epoch": 39.47089947089947, "grad_norm": 0.0619661882519722, "learning_rate": 1.663778261917695e-05, "loss": 0.0023, "step": 7460 }, { "epoch": 39.523809523809526, "grad_norm": 0.06733821332454681, "learning_rate": 1.651480840226952e-05, "loss": 0.0028, "step": 7470 }, { "epoch": 39.576719576719576, "grad_norm": 0.06399443000555038, "learning_rate": 1.639220037405258e-05, "loss": 0.0022, "step": 7480 }, { "epoch": 39.629629629629626, "grad_norm": 0.129105806350708, "learning_rate": 1.6269959875347906e-05, "loss": 0.0046, "step": 7490 }, { "epoch": 39.682539682539684, "grad_norm": 0.08309172093868256, "learning_rate": 1.614808824295802e-05, "loss": 0.0023, "step": 7500 }, { "epoch": 39.735449735449734, "grad_norm": 0.13913407921791077, "learning_rate": 1.602658680965152e-05, "loss": 0.0028, "step": 7510 }, { "epoch": 39.78835978835979, "grad_norm": 0.12876394391059875, "learning_rate": 1.5905456904148686e-05, "loss": 0.0024, "step": 7520 }, { "epoch": 39.84126984126984, "grad_norm": 0.09393489360809326, "learning_rate": 1.57846998511067e-05, "loss": 0.0027, "step": 7530 }, { "epoch": 39.89417989417989, "grad_norm": 0.07630272209644318, "learning_rate": 1.566431697110538e-05, "loss": 0.0019, "step": 7540 }, { "epoch": 39.94708994708995, "grad_norm": 0.055215246975421906, "learning_rate": 1.554430958063259e-05, "loss": 0.003, "step": 7550 }, { "epoch": 40.0, "grad_norm": 0.09851783514022827, "learning_rate": 1.5424678992069912e-05, "loss": 0.0024, "step": 7560 }, { "epoch": 40.05291005291005, "grad_norm": 0.0872630849480629, "learning_rate": 1.5305426513678362e-05, "loss": 0.0031, "step": 7570 }, { "epoch": 40.10582010582011, "grad_norm": 0.08486311137676239, "learning_rate": 1.518655344958388e-05, "loss": 0.0033, "step": 7580 }, { "epoch": 40.15873015873016, "grad_norm": 0.10787244141101837, "learning_rate": 1.5068061099763275e-05, "loss": 0.0031, "step": 7590 }, { "epoch": 40.21164021164021, "grad_norm": 0.11103319376707077, "learning_rate": 1.494995076002988e-05, "loss": 0.0036, "step": 7600 }, { "epoch": 40.264550264550266, "grad_norm": 0.1186375766992569, "learning_rate": 1.4832223722019456e-05, "loss": 0.0028, "step": 7610 }, { "epoch": 40.317460317460316, "grad_norm": 0.12691158056259155, "learning_rate": 1.4714881273176035e-05, "loss": 0.0047, "step": 7620 }, { "epoch": 40.370370370370374, "grad_norm": 0.07471556216478348, "learning_rate": 1.4597924696737835e-05, "loss": 0.0022, "step": 7630 }, { "epoch": 40.423280423280424, "grad_norm": 0.07035799324512482, "learning_rate": 1.4481355271723252e-05, "loss": 0.0024, "step": 7640 }, { "epoch": 40.476190476190474, "grad_norm": 0.04878492280840874, "learning_rate": 1.4365174272916809e-05, "loss": 0.0022, "step": 7650 }, { "epoch": 40.52910052910053, "grad_norm": 0.09836369007825851, "learning_rate": 1.4249382970855319e-05, "loss": 0.004, "step": 7660 }, { "epoch": 40.58201058201058, "grad_norm": 0.05819159001111984, "learning_rate": 1.4133982631813903e-05, "loss": 0.0022, "step": 7670 }, { "epoch": 40.63492063492063, "grad_norm": 0.04416554048657417, "learning_rate": 1.4018974517792194e-05, "loss": 0.0023, "step": 7680 }, { "epoch": 40.68783068783069, "grad_norm": 0.09750942885875702, "learning_rate": 1.390435988650048e-05, "loss": 0.003, "step": 7690 }, { "epoch": 40.74074074074074, "grad_norm": 0.062239550054073334, "learning_rate": 1.3790139991346006e-05, "loss": 0.0029, "step": 7700 }, { "epoch": 40.79365079365079, "grad_norm": 0.0827951431274414, "learning_rate": 1.367631608141926e-05, "loss": 0.0023, "step": 7710 }, { "epoch": 40.84656084656085, "grad_norm": 0.09299186617136002, "learning_rate": 1.3562889401480278e-05, "loss": 0.0034, "step": 7720 }, { "epoch": 40.8994708994709, "grad_norm": 0.08352316915988922, "learning_rate": 1.3449861191945074e-05, "loss": 0.0021, "step": 7730 }, { "epoch": 40.95238095238095, "grad_norm": 0.07465235888957977, "learning_rate": 1.3337232688872009e-05, "loss": 0.0023, "step": 7740 }, { "epoch": 41.00529100529101, "grad_norm": 0.05501871183514595, "learning_rate": 1.3225005123948364e-05, "loss": 0.0026, "step": 7750 }, { "epoch": 41.05820105820106, "grad_norm": 0.06581036001443863, "learning_rate": 1.311317972447681e-05, "loss": 0.0024, "step": 7760 }, { "epoch": 41.111111111111114, "grad_norm": 0.05128316953778267, "learning_rate": 1.3001757713361996e-05, "loss": 0.0024, "step": 7770 }, { "epoch": 41.164021164021165, "grad_norm": 0.077063649892807, "learning_rate": 1.2890740309097204e-05, "loss": 0.0055, "step": 7780 }, { "epoch": 41.216931216931215, "grad_norm": 0.1039019227027893, "learning_rate": 1.2780128725750944e-05, "loss": 0.0023, "step": 7790 }, { "epoch": 41.26984126984127, "grad_norm": 0.05210014432668686, "learning_rate": 1.266992417295379e-05, "loss": 0.0032, "step": 7800 }, { "epoch": 41.32275132275132, "grad_norm": 0.08479829132556915, "learning_rate": 1.2560127855885073e-05, "loss": 0.0022, "step": 7810 }, { "epoch": 41.37566137566137, "grad_norm": 0.14084255695343018, "learning_rate": 1.2450740975259745e-05, "loss": 0.0031, "step": 7820 }, { "epoch": 41.42857142857143, "grad_norm": 0.07982408255338669, "learning_rate": 1.234176472731517e-05, "loss": 0.0033, "step": 7830 }, { "epoch": 41.48148148148148, "grad_norm": 0.10496889054775238, "learning_rate": 1.2233200303798158e-05, "loss": 0.0028, "step": 7840 }, { "epoch": 41.53439153439153, "grad_norm": 0.07431522011756897, "learning_rate": 1.2125048891951846e-05, "loss": 0.0027, "step": 7850 }, { "epoch": 41.58730158730159, "grad_norm": 0.14783434569835663, "learning_rate": 1.2017311674502745e-05, "loss": 0.0031, "step": 7860 }, { "epoch": 41.64021164021164, "grad_norm": 0.07344204932451248, "learning_rate": 1.1909989829647822e-05, "loss": 0.0021, "step": 7870 }, { "epoch": 41.6931216931217, "grad_norm": 0.05640701577067375, "learning_rate": 1.1803084531041553e-05, "loss": 0.0022, "step": 7880 }, { "epoch": 41.74603174603175, "grad_norm": 0.08225259184837341, "learning_rate": 1.1696596947783162e-05, "loss": 0.0018, "step": 7890 }, { "epoch": 41.7989417989418, "grad_norm": 0.07301223278045654, "learning_rate": 1.1590528244403803e-05, "loss": 0.0021, "step": 7900 }, { "epoch": 41.851851851851855, "grad_norm": 0.0678814947605133, "learning_rate": 1.148487958085382e-05, "loss": 0.0022, "step": 7910 }, { "epoch": 41.904761904761905, "grad_norm": 0.12124235928058624, "learning_rate": 1.1379652112490086e-05, "loss": 0.0023, "step": 7920 }, { "epoch": 41.957671957671955, "grad_norm": 0.04354817420244217, "learning_rate": 1.1274846990063315e-05, "loss": 0.0025, "step": 7930 }, { "epoch": 42.01058201058201, "grad_norm": 0.06643577665090561, "learning_rate": 1.117046535970554e-05, "loss": 0.0028, "step": 7940 }, { "epoch": 42.06349206349206, "grad_norm": 0.06463734060525894, "learning_rate": 1.106650836291755e-05, "loss": 0.0025, "step": 7950 }, { "epoch": 42.116402116402114, "grad_norm": 0.06024490296840668, "learning_rate": 1.0962977136556418e-05, "loss": 0.0025, "step": 7960 }, { "epoch": 42.16931216931217, "grad_norm": 0.044284887611866, "learning_rate": 1.0859872812823024e-05, "loss": 0.0018, "step": 7970 }, { "epoch": 42.22222222222222, "grad_norm": 0.053850866854190826, "learning_rate": 1.0757196519249747e-05, "loss": 0.0025, "step": 7980 }, { "epoch": 42.27513227513227, "grad_norm": 0.06391821801662445, "learning_rate": 1.0654949378688077e-05, "loss": 0.0028, "step": 7990 }, { "epoch": 42.32804232804233, "grad_norm": 0.061592113226652145, "learning_rate": 1.0553132509296376e-05, "loss": 0.0026, "step": 8000 }, { "epoch": 42.38095238095238, "grad_norm": 0.039215922355651855, "learning_rate": 1.0451747024527613e-05, "loss": 0.0019, "step": 8010 }, { "epoch": 42.43386243386244, "grad_norm": 0.07505115866661072, "learning_rate": 1.0350794033117189e-05, "loss": 0.002, "step": 8020 }, { "epoch": 42.48677248677249, "grad_norm": 0.05983835458755493, "learning_rate": 1.0250274639070856e-05, "loss": 0.0019, "step": 8030 }, { "epoch": 42.53968253968254, "grad_norm": 0.04498854652047157, "learning_rate": 1.0150189941652599e-05, "loss": 0.0024, "step": 8040 }, { "epoch": 42.592592592592595, "grad_norm": 0.06539206206798553, "learning_rate": 1.0050541035372635e-05, "loss": 0.0028, "step": 8050 }, { "epoch": 42.645502645502646, "grad_norm": 0.06036480888724327, "learning_rate": 9.951329009975458e-06, "loss": 0.0019, "step": 8060 }, { "epoch": 42.698412698412696, "grad_norm": 0.07967852056026459, "learning_rate": 9.852554950427845e-06, "loss": 0.0024, "step": 8070 }, { "epoch": 42.75132275132275, "grad_norm": 0.06078115478157997, "learning_rate": 9.754219936907105e-06, "loss": 0.0024, "step": 8080 }, { "epoch": 42.804232804232804, "grad_norm": 0.045221682637929916, "learning_rate": 9.656325044789194e-06, "loss": 0.0021, "step": 8090 }, { "epoch": 42.857142857142854, "grad_norm": 0.04402315244078636, "learning_rate": 9.55887134463697e-06, "loss": 0.002, "step": 8100 }, { "epoch": 42.91005291005291, "grad_norm": 0.051226384937763214, "learning_rate": 9.461859902188475e-06, "loss": 0.0035, "step": 8110 }, { "epoch": 42.96296296296296, "grad_norm": 0.08825559169054031, "learning_rate": 9.365291778345303e-06, "loss": 0.0025, "step": 8120 }, { "epoch": 43.01587301587302, "grad_norm": 0.06666417419910431, "learning_rate": 9.269168029160991e-06, "loss": 0.0032, "step": 8130 }, { "epoch": 43.06878306878307, "grad_norm": 0.11924991011619568, "learning_rate": 9.173489705829447e-06, "loss": 0.0022, "step": 8140 }, { "epoch": 43.12169312169312, "grad_norm": 0.05916042998433113, "learning_rate": 9.078257854673516e-06, "loss": 0.0021, "step": 8150 }, { "epoch": 43.17460317460318, "grad_norm": 0.07745632529258728, "learning_rate": 8.983473517133429e-06, "loss": 0.0026, "step": 8160 }, { "epoch": 43.22751322751323, "grad_norm": 0.036119505763053894, "learning_rate": 8.889137729755537e-06, "loss": 0.0026, "step": 8170 }, { "epoch": 43.28042328042328, "grad_norm": 0.07312419265508652, "learning_rate": 8.79525152418087e-06, "loss": 0.0033, "step": 8180 }, { "epoch": 43.333333333333336, "grad_norm": 0.06665989011526108, "learning_rate": 8.701815927133961e-06, "loss": 0.0036, "step": 8190 }, { "epoch": 43.386243386243386, "grad_norm": 0.08487293124198914, "learning_rate": 8.608831960411534e-06, "loss": 0.0034, "step": 8200 }, { "epoch": 43.439153439153436, "grad_norm": 0.05715225264430046, "learning_rate": 8.516300640871321e-06, "loss": 0.0021, "step": 8210 }, { "epoch": 43.492063492063494, "grad_norm": 0.039052270352840424, "learning_rate": 8.424222980421038e-06, "loss": 0.0025, "step": 8220 }, { "epoch": 43.544973544973544, "grad_norm": 0.12846902012825012, "learning_rate": 8.332599986007184e-06, "loss": 0.002, "step": 8230 }, { "epoch": 43.597883597883595, "grad_norm": 0.06003214418888092, "learning_rate": 8.241432659604203e-06, "loss": 0.0023, "step": 8240 }, { "epoch": 43.65079365079365, "grad_norm": 0.06788913905620575, "learning_rate": 8.150721998203331e-06, "loss": 0.0023, "step": 8250 }, { "epoch": 43.7037037037037, "grad_norm": 0.04894135892391205, "learning_rate": 8.06046899380184e-06, "loss": 0.0019, "step": 8260 }, { "epoch": 43.75661375661376, "grad_norm": 0.05002419650554657, "learning_rate": 7.970674633392133e-06, "loss": 0.0024, "step": 8270 }, { "epoch": 43.80952380952381, "grad_norm": 0.04428821802139282, "learning_rate": 7.881339898950924e-06, "loss": 0.0016, "step": 8280 }, { "epoch": 43.86243386243386, "grad_norm": 0.07982371747493744, "learning_rate": 7.792465767428597e-06, "loss": 0.0021, "step": 8290 }, { "epoch": 43.91534391534392, "grad_norm": 0.10252171009778976, "learning_rate": 7.704053210738376e-06, "loss": 0.0027, "step": 8300 }, { "epoch": 43.96825396825397, "grad_norm": 0.046775735914707184, "learning_rate": 7.6161031957458494e-06, "loss": 0.0028, "step": 8310 }, { "epoch": 44.02116402116402, "grad_norm": 0.0593828409910202, "learning_rate": 7.5286166842582605e-06, "loss": 0.003, "step": 8320 }, { "epoch": 44.074074074074076, "grad_norm": 0.04723280668258667, "learning_rate": 7.4415946330140814e-06, "loss": 0.0031, "step": 8330 }, { "epoch": 44.12698412698413, "grad_norm": 0.03772735595703125, "learning_rate": 7.3550379936725644e-06, "loss": 0.0021, "step": 8340 }, { "epoch": 44.17989417989418, "grad_norm": 0.061364173889160156, "learning_rate": 7.2689477128032035e-06, "loss": 0.0021, "step": 8350 }, { "epoch": 44.232804232804234, "grad_norm": 0.03763009235262871, "learning_rate": 7.183324731875551e-06, "loss": 0.0024, "step": 8360 }, { "epoch": 44.285714285714285, "grad_norm": 0.04696877300739288, "learning_rate": 7.098169987248782e-06, "loss": 0.0015, "step": 8370 }, { "epoch": 44.33862433862434, "grad_norm": 0.0779181718826294, "learning_rate": 7.013484410161553e-06, "loss": 0.0026, "step": 8380 }, { "epoch": 44.39153439153439, "grad_norm": 0.059823594987392426, "learning_rate": 6.92926892672176e-06, "loss": 0.0022, "step": 8390 }, { "epoch": 44.44444444444444, "grad_norm": 0.07204800844192505, "learning_rate": 6.845524457896446e-06, "loss": 0.0047, "step": 8400 }, { "epoch": 44.4973544973545, "grad_norm": 0.07600213587284088, "learning_rate": 6.7622519195017165e-06, "loss": 0.0019, "step": 8410 }, { "epoch": 44.55026455026455, "grad_norm": 0.04590420797467232, "learning_rate": 6.679452222192684e-06, "loss": 0.0018, "step": 8420 }, { "epoch": 44.6031746031746, "grad_norm": 0.05781931057572365, "learning_rate": 6.597126271453579e-06, "loss": 0.0022, "step": 8430 }, { "epoch": 44.65608465608466, "grad_norm": 0.05953796207904816, "learning_rate": 6.51527496758782e-06, "loss": 0.0019, "step": 8440 }, { "epoch": 44.70899470899471, "grad_norm": 0.059320174157619476, "learning_rate": 6.433899205708155e-06, "loss": 0.0033, "step": 8450 }, { "epoch": 44.76190476190476, "grad_norm": 0.06743332743644714, "learning_rate": 6.352999875726856e-06, "loss": 0.0022, "step": 8460 }, { "epoch": 44.81481481481482, "grad_norm": 0.09310760349035263, "learning_rate": 6.272577862346052e-06, "loss": 0.0028, "step": 8470 }, { "epoch": 44.86772486772487, "grad_norm": 0.04635382443666458, "learning_rate": 6.192634045047996e-06, "loss": 0.0032, "step": 8480 }, { "epoch": 44.92063492063492, "grad_norm": 0.08243785798549652, "learning_rate": 6.113169298085458e-06, "loss": 0.003, "step": 8490 }, { "epoch": 44.973544973544975, "grad_norm": 0.043838564306497574, "learning_rate": 6.034184490472195e-06, "loss": 0.0026, "step": 8500 }, { "epoch": 45.026455026455025, "grad_norm": 0.07907702028751373, "learning_rate": 5.955680485973386e-06, "loss": 0.0039, "step": 8510 }, { "epoch": 45.07936507936508, "grad_norm": 0.07482165098190308, "learning_rate": 5.877658143096265e-06, "loss": 0.0019, "step": 8520 }, { "epoch": 45.13227513227513, "grad_norm": 0.11568737775087357, "learning_rate": 5.800118315080661e-06, "loss": 0.0028, "step": 8530 }, { "epoch": 45.18518518518518, "grad_norm": 0.07257694005966187, "learning_rate": 5.723061849889716e-06, "loss": 0.0018, "step": 8540 }, { "epoch": 45.23809523809524, "grad_norm": 0.05349784344434738, "learning_rate": 5.646489590200604e-06, "loss": 0.002, "step": 8550 }, { "epoch": 45.29100529100529, "grad_norm": 0.03407958149909973, "learning_rate": 5.570402373395256e-06, "loss": 0.0023, "step": 8560 }, { "epoch": 45.34391534391534, "grad_norm": 0.03151976689696312, "learning_rate": 5.494801031551305e-06, "loss": 0.0028, "step": 8570 }, { "epoch": 45.3968253968254, "grad_norm": 0.06410853564739227, "learning_rate": 5.41968639143291e-06, "loss": 0.0018, "step": 8580 }, { "epoch": 45.44973544973545, "grad_norm": 0.1213647648692131, "learning_rate": 5.345059274481751e-06, "loss": 0.0024, "step": 8590 }, { "epoch": 45.5026455026455, "grad_norm": 0.10514358431100845, "learning_rate": 5.270920496808002e-06, "loss": 0.0024, "step": 8600 }, { "epoch": 45.55555555555556, "grad_norm": 0.07078886777162552, "learning_rate": 5.1972708691814695e-06, "loss": 0.003, "step": 8610 }, { "epoch": 45.60846560846561, "grad_norm": 0.04438650608062744, "learning_rate": 5.124111197022674e-06, "loss": 0.003, "step": 8620 }, { "epoch": 45.66137566137566, "grad_norm": 0.0978962630033493, "learning_rate": 5.051442280394081e-06, "loss": 0.0027, "step": 8630 }, { "epoch": 45.714285714285715, "grad_norm": 0.0430341437458992, "learning_rate": 4.979264913991322e-06, "loss": 0.0039, "step": 8640 }, { "epoch": 45.767195767195766, "grad_norm": 0.08823809772729874, "learning_rate": 4.907579887134489e-06, "loss": 0.003, "step": 8650 }, { "epoch": 45.82010582010582, "grad_norm": 0.11226730048656464, "learning_rate": 4.836387983759572e-06, "loss": 0.0021, "step": 8660 }, { "epoch": 45.87301587301587, "grad_norm": 0.08951160311698914, "learning_rate": 4.765689982409816e-06, "loss": 0.0029, "step": 8670 }, { "epoch": 45.925925925925924, "grad_norm": 0.060601674020290375, "learning_rate": 4.695486656227233e-06, "loss": 0.0029, "step": 8680 }, { "epoch": 45.97883597883598, "grad_norm": 0.07224203646183014, "learning_rate": 4.625778772944156e-06, "loss": 0.0033, "step": 8690 }, { "epoch": 46.03174603174603, "grad_norm": 0.08406931161880493, "learning_rate": 4.556567094874825e-06, "loss": 0.0027, "step": 8700 }, { "epoch": 46.08465608465608, "grad_norm": 0.03635375201702118, "learning_rate": 4.487852378907059e-06, "loss": 0.0026, "step": 8710 }, { "epoch": 46.13756613756614, "grad_norm": 0.08861076831817627, "learning_rate": 4.419635376493986e-06, "loss": 0.0024, "step": 8720 }, { "epoch": 46.19047619047619, "grad_norm": 0.03910231590270996, "learning_rate": 4.351916833645825e-06, "loss": 0.0028, "step": 8730 }, { "epoch": 46.24338624338624, "grad_norm": 0.07249995321035385, "learning_rate": 4.284697490921691e-06, "loss": 0.0023, "step": 8740 }, { "epoch": 46.2962962962963, "grad_norm": 0.11680825054645538, "learning_rate": 4.2179780834215585e-06, "loss": 0.0019, "step": 8750 }, { "epoch": 46.34920634920635, "grad_norm": 0.044864822179079056, "learning_rate": 4.151759340778178e-06, "loss": 0.0024, "step": 8760 }, { "epoch": 46.402116402116405, "grad_norm": 0.03768221661448479, "learning_rate": 4.086041987149109e-06, "loss": 0.0019, "step": 8770 }, { "epoch": 46.455026455026456, "grad_norm": 0.061667632311582565, "learning_rate": 4.020826741208811e-06, "loss": 0.002, "step": 8780 }, { "epoch": 46.507936507936506, "grad_norm": 0.03395885229110718, "learning_rate": 3.956114316140746e-06, "loss": 0.0019, "step": 8790 }, { "epoch": 46.560846560846564, "grad_norm": 0.09251205623149872, "learning_rate": 3.891905419629643e-06, "loss": 0.0033, "step": 8800 }, { "epoch": 46.613756613756614, "grad_norm": 0.08876454085111618, "learning_rate": 3.8282007538536946e-06, "loss": 0.0019, "step": 8810 }, { "epoch": 46.666666666666664, "grad_norm": 0.06236439570784569, "learning_rate": 3.7650010154769265e-06, "loss": 0.0019, "step": 8820 }, { "epoch": 46.71957671957672, "grad_norm": 0.06314297765493393, "learning_rate": 3.7023068956415608e-06, "loss": 0.0022, "step": 8830 }, { "epoch": 46.77248677248677, "grad_norm": 0.0358610637485981, "learning_rate": 3.6401190799604303e-06, "loss": 0.0035, "step": 8840 }, { "epoch": 46.82539682539682, "grad_norm": 0.0616828016936779, "learning_rate": 3.578438248509536e-06, "loss": 0.0025, "step": 8850 }, { "epoch": 46.87830687830688, "grad_norm": 0.08215471357107162, "learning_rate": 3.5172650758205583e-06, "loss": 0.0021, "step": 8860 }, { "epoch": 46.93121693121693, "grad_norm": 0.09664836525917053, "learning_rate": 3.45660023087353e-06, "loss": 0.0028, "step": 8870 }, { "epoch": 46.98412698412698, "grad_norm": 0.08146372437477112, "learning_rate": 3.3964443770894528e-06, "loss": 0.0021, "step": 8880 }, { "epoch": 47.03703703703704, "grad_norm": 0.041112929582595825, "learning_rate": 3.3367981723231245e-06, "loss": 0.0018, "step": 8890 }, { "epoch": 47.08994708994709, "grad_norm": 0.10833887755870819, "learning_rate": 3.2776622688558746e-06, "loss": 0.0024, "step": 8900 }, { "epoch": 47.142857142857146, "grad_norm": 0.07685001194477081, "learning_rate": 3.2190373133884677e-06, "loss": 0.0023, "step": 8910 }, { "epoch": 47.195767195767196, "grad_norm": 0.08453325182199478, "learning_rate": 3.1609239470340446e-06, "loss": 0.0023, "step": 8920 }, { "epoch": 47.24867724867725, "grad_norm": 0.052381958812475204, "learning_rate": 3.1033228053110373e-06, "loss": 0.0022, "step": 8930 }, { "epoch": 47.301587301587304, "grad_norm": 0.041455794125795364, "learning_rate": 3.0462345181363314e-06, "loss": 0.0016, "step": 8940 }, { "epoch": 47.354497354497354, "grad_norm": 0.08272495865821838, "learning_rate": 2.9896597098182654e-06, "loss": 0.002, "step": 8950 }, { "epoch": 47.407407407407405, "grad_norm": 0.026947449892759323, "learning_rate": 2.933598999049891e-06, "loss": 0.0017, "step": 8960 }, { "epoch": 47.46031746031746, "grad_norm": 0.038717083632946014, "learning_rate": 2.8780529989021697e-06, "loss": 0.002, "step": 8970 }, { "epoch": 47.51322751322751, "grad_norm": 0.03269005939364433, "learning_rate": 2.823022316817242e-06, "loss": 0.0021, "step": 8980 }, { "epoch": 47.56613756613756, "grad_norm": 0.042242106050252914, "learning_rate": 2.7685075546018456e-06, "loss": 0.0016, "step": 8990 }, { "epoch": 47.61904761904762, "grad_norm": 0.03139335662126541, "learning_rate": 2.7145093084206598e-06, "loss": 0.0018, "step": 9000 }, { "epoch": 47.67195767195767, "grad_norm": 0.02996080182492733, "learning_rate": 2.661028168789892e-06, "loss": 0.002, "step": 9010 }, { "epoch": 47.72486772486773, "grad_norm": 0.09505295753479004, "learning_rate": 2.6080647205706855e-06, "loss": 0.0017, "step": 9020 }, { "epoch": 47.77777777777778, "grad_norm": 0.05809766799211502, "learning_rate": 2.555619542962834e-06, "loss": 0.0019, "step": 9030 }, { "epoch": 47.83068783068783, "grad_norm": 0.09852910786867142, "learning_rate": 2.503693209498409e-06, "loss": 0.002, "step": 9040 }, { "epoch": 47.883597883597886, "grad_norm": 0.031182926148176193, "learning_rate": 2.452286288035449e-06, "loss": 0.0027, "step": 9050 }, { "epoch": 47.93650793650794, "grad_norm": 0.028979627415537834, "learning_rate": 2.4013993407518363e-06, "loss": 0.0024, "step": 9060 }, { "epoch": 47.98941798941799, "grad_norm": 0.04749450087547302, "learning_rate": 2.351032924139063e-06, "loss": 0.0024, "step": 9070 }, { "epoch": 48.042328042328045, "grad_norm": 0.07246076315641403, "learning_rate": 2.30118758899619e-06, "loss": 0.002, "step": 9080 }, { "epoch": 48.095238095238095, "grad_norm": 0.049553729593753815, "learning_rate": 2.2518638804238157e-06, "loss": 0.002, "step": 9090 }, { "epoch": 48.148148148148145, "grad_norm": 0.022088665515184402, "learning_rate": 2.203062337818118e-06, "loss": 0.0025, "step": 9100 }, { "epoch": 48.2010582010582, "grad_norm": 0.03272722288966179, "learning_rate": 2.1547834948649483e-06, "loss": 0.002, "step": 9110 }, { "epoch": 48.25396825396825, "grad_norm": 0.033946167677640915, "learning_rate": 2.1070278795340017e-06, "loss": 0.0019, "step": 9120 }, { "epoch": 48.3068783068783, "grad_norm": 0.02590041048824787, "learning_rate": 2.059796014073029e-06, "loss": 0.0024, "step": 9130 }, { "epoch": 48.35978835978836, "grad_norm": 0.020455658435821533, "learning_rate": 2.01308841500214e-06, "loss": 0.0015, "step": 9140 }, { "epoch": 48.41269841269841, "grad_norm": 0.025815559551119804, "learning_rate": 1.9669055931081704e-06, "loss": 0.0016, "step": 9150 }, { "epoch": 48.46560846560847, "grad_norm": 0.03265366330742836, "learning_rate": 1.9212480534390507e-06, "loss": 0.0019, "step": 9160 }, { "epoch": 48.51851851851852, "grad_norm": 0.078748419880867, "learning_rate": 1.8761162952983246e-06, "loss": 0.0028, "step": 9170 }, { "epoch": 48.57142857142857, "grad_norm": 0.0266819279640913, "learning_rate": 1.8315108122396618e-06, "loss": 0.0023, "step": 9180 }, { "epoch": 48.62433862433863, "grad_norm": 0.06296074390411377, "learning_rate": 1.787432092061475e-06, "loss": 0.0028, "step": 9190 }, { "epoch": 48.67724867724868, "grad_norm": 0.05147553235292435, "learning_rate": 1.743880616801602e-06, "loss": 0.0025, "step": 9200 }, { "epoch": 48.73015873015873, "grad_norm": 0.04126429557800293, "learning_rate": 1.7008568627319865e-06, "loss": 0.0025, "step": 9210 }, { "epoch": 48.783068783068785, "grad_norm": 0.13614654541015625, "learning_rate": 1.6583613003535226e-06, "loss": 0.0023, "step": 9220 }, { "epoch": 48.835978835978835, "grad_norm": 0.04965079203248024, "learning_rate": 1.6163943943908522e-06, "loss": 0.0019, "step": 9230 }, { "epoch": 48.888888888888886, "grad_norm": 0.030159763991832733, "learning_rate": 1.5749566037873476e-06, "loss": 0.0015, "step": 9240 }, { "epoch": 48.94179894179894, "grad_norm": 0.05381500720977783, "learning_rate": 1.5340483817000428e-06, "loss": 0.0032, "step": 9250 }, { "epoch": 48.99470899470899, "grad_norm": 0.05476432293653488, "learning_rate": 1.4936701754947101e-06, "loss": 0.0022, "step": 9260 }, { "epoch": 49.04761904761905, "grad_norm": 0.041640669107437134, "learning_rate": 1.4538224267409361e-06, "loss": 0.0042, "step": 9270 }, { "epoch": 49.1005291005291, "grad_norm": 0.033190030604600906, "learning_rate": 1.414505571207314e-06, "loss": 0.0015, "step": 9280 }, { "epoch": 49.15343915343915, "grad_norm": 0.041642751544713974, "learning_rate": 1.3757200388566816e-06, "loss": 0.0017, "step": 9290 }, { "epoch": 49.20634920634921, "grad_norm": 0.08033628016710281, "learning_rate": 1.3374662538414074e-06, "loss": 0.0023, "step": 9300 }, { "epoch": 49.25925925925926, "grad_norm": 0.029303984716534615, "learning_rate": 1.2997446344987617e-06, "loss": 0.0018, "step": 9310 }, { "epoch": 49.31216931216931, "grad_norm": 0.03676794469356537, "learning_rate": 1.262555593346315e-06, "loss": 0.0029, "step": 9320 }, { "epoch": 49.36507936507937, "grad_norm": 0.061546072363853455, "learning_rate": 1.2258995370774685e-06, "loss": 0.0024, "step": 9330 }, { "epoch": 49.41798941798942, "grad_norm": 0.0675833523273468, "learning_rate": 1.1897768665569798e-06, "loss": 0.0018, "step": 9340 }, { "epoch": 49.47089947089947, "grad_norm": 0.04244307428598404, "learning_rate": 1.1541879768165954e-06, "loss": 0.002, "step": 9350 }, { "epoch": 49.523809523809526, "grad_norm": 0.06433667987585068, "learning_rate": 1.1191332570507085e-06, "loss": 0.0026, "step": 9360 }, { "epoch": 49.576719576719576, "grad_norm": 0.04198251664638519, "learning_rate": 1.0846130906121132e-06, "loss": 0.002, "step": 9370 }, { "epoch": 49.629629629629626, "grad_norm": 0.05469237267971039, "learning_rate": 1.0506278550078131e-06, "loss": 0.0018, "step": 9380 }, { "epoch": 49.682539682539684, "grad_norm": 0.04273342341184616, "learning_rate": 1.0171779218949185e-06, "loss": 0.0014, "step": 9390 }, { "epoch": 49.735449735449734, "grad_norm": 0.049993354827165604, "learning_rate": 9.842636570765174e-07, "loss": 0.0019, "step": 9400 }, { "epoch": 49.78835978835979, "grad_norm": 0.09222929924726486, "learning_rate": 9.518854204977612e-07, "loss": 0.0041, "step": 9410 }, { "epoch": 49.84126984126984, "grad_norm": 0.052687615156173706, "learning_rate": 9.200435662418349e-07, "loss": 0.003, "step": 9420 }, { "epoch": 49.89417989417989, "grad_norm": 0.06168532371520996, "learning_rate": 8.887384425261658e-07, "loss": 0.0026, "step": 9430 }, { "epoch": 49.94708994708995, "grad_norm": 0.06320830434560776, "learning_rate": 8.579703916985648e-07, "loss": 0.0027, "step": 9440 }, { "epoch": 50.0, "grad_norm": 0.06489460915327072, "learning_rate": 8.277397502335194e-07, "loss": 0.0026, "step": 9450 }, { "epoch": 50.05291005291005, "grad_norm": 0.096370629966259, "learning_rate": 7.980468487284675e-07, "loss": 0.0022, "step": 9460 }, { "epoch": 50.10582010582011, "grad_norm": 0.028497254475951195, "learning_rate": 7.688920119002297e-07, "loss": 0.002, "step": 9470 }, { "epoch": 50.15873015873016, "grad_norm": 0.02044384554028511, "learning_rate": 7.402755585814269e-07, "loss": 0.0016, "step": 9480 }, { "epoch": 50.21164021164021, "grad_norm": 0.06846970319747925, "learning_rate": 7.121978017170073e-07, "loss": 0.0018, "step": 9490 }, { "epoch": 50.264550264550266, "grad_norm": 0.03613802045583725, "learning_rate": 6.846590483608306e-07, "loss": 0.0017, "step": 9500 }, { "epoch": 50.317460317460316, "grad_norm": 0.03778558969497681, "learning_rate": 6.576595996722834e-07, "loss": 0.0018, "step": 9510 }, { "epoch": 50.370370370370374, "grad_norm": 0.056209102272987366, "learning_rate": 6.311997509130141e-07, "loss": 0.0019, "step": 9520 }, { "epoch": 50.423280423280424, "grad_norm": 0.022677993401885033, "learning_rate": 6.052797914436803e-07, "loss": 0.0019, "step": 9530 }, { "epoch": 50.476190476190474, "grad_norm": 0.023270627483725548, "learning_rate": 5.799000047208181e-07, "loss": 0.0017, "step": 9540 }, { "epoch": 50.52910052910053, "grad_norm": 0.031920816749334335, "learning_rate": 5.550606682937054e-07, "loss": 0.0027, "step": 9550 }, { "epoch": 50.58201058201058, "grad_norm": 0.02740374766290188, "learning_rate": 5.307620538013481e-07, "loss": 0.0019, "step": 9560 }, { "epoch": 50.63492063492063, "grad_norm": 0.03158656507730484, "learning_rate": 5.070044269694874e-07, "loss": 0.0019, "step": 9570 }, { "epoch": 50.68783068783069, "grad_norm": 0.03670842945575714, "learning_rate": 4.837880476077417e-07, "loss": 0.0014, "step": 9580 }, { "epoch": 50.74074074074074, "grad_norm": 0.021545149385929108, "learning_rate": 4.6111316960670835e-07, "loss": 0.002, "step": 9590 }, { "epoch": 50.79365079365079, "grad_norm": 0.033108413219451904, "learning_rate": 4.389800409352218e-07, "loss": 0.002, "step": 9600 }, { "epoch": 50.84656084656085, "grad_norm": 0.049538418650627136, "learning_rate": 4.173889036376277e-07, "loss": 0.0016, "step": 9610 }, { "epoch": 50.8994708994709, "grad_norm": 0.03392045199871063, "learning_rate": 3.963399938311463e-07, "loss": 0.0023, "step": 9620 }, { "epoch": 50.95238095238095, "grad_norm": 0.044099390506744385, "learning_rate": 3.7583354170328545e-07, "loss": 0.0028, "step": 9630 }, { "epoch": 51.00529100529101, "grad_norm": 0.03674205392599106, "learning_rate": 3.558697715093207e-07, "loss": 0.0016, "step": 9640 }, { "epoch": 51.05820105820106, "grad_norm": 0.039752259850502014, "learning_rate": 3.3644890156983576e-07, "loss": 0.0018, "step": 9650 }, { "epoch": 51.111111111111114, "grad_norm": 0.05322498828172684, "learning_rate": 3.175711442683638e-07, "loss": 0.0019, "step": 9660 }, { "epoch": 51.164021164021165, "grad_norm": 0.062151405960321426, "learning_rate": 2.9923670604902197e-07, "loss": 0.0022, "step": 9670 }, { "epoch": 51.216931216931215, "grad_norm": 0.023568203672766685, "learning_rate": 2.814457874143028e-07, "loss": 0.003, "step": 9680 }, { "epoch": 51.26984126984127, "grad_norm": 0.021968849003314972, "learning_rate": 2.641985829228366e-07, "loss": 0.0014, "step": 9690 }, { "epoch": 51.32275132275132, "grad_norm": 0.024687567725777626, "learning_rate": 2.474952811872877e-07, "loss": 0.0017, "step": 9700 }, { "epoch": 51.37566137566137, "grad_norm": 0.022805940359830856, "learning_rate": 2.3133606487228397e-07, "loss": 0.0017, "step": 9710 }, { "epoch": 51.42857142857143, "grad_norm": 0.06309483200311661, "learning_rate": 2.157211106924295e-07, "loss": 0.0021, "step": 9720 }, { "epoch": 51.48148148148148, "grad_norm": 0.03799191489815712, "learning_rate": 2.006505894103672e-07, "loss": 0.0023, "step": 9730 }, { "epoch": 51.53439153439153, "grad_norm": 0.024295153096318245, "learning_rate": 1.8612466583489696e-07, "loss": 0.0015, "step": 9740 }, { "epoch": 51.58730158730159, "grad_norm": 0.030950749292969704, "learning_rate": 1.7214349881918834e-07, "loss": 0.0022, "step": 9750 }, { "epoch": 51.64021164021164, "grad_norm": 0.06668223440647125, "learning_rate": 1.5870724125904845e-07, "loss": 0.0025, "step": 9760 }, { "epoch": 51.6931216931217, "grad_norm": 0.08100949972867966, "learning_rate": 1.4581604009124006e-07, "loss": 0.0022, "step": 9770 }, { "epoch": 51.74603174603175, "grad_norm": 0.04216459020972252, "learning_rate": 1.334700362918717e-07, "loss": 0.0016, "step": 9780 }, { "epoch": 51.7989417989418, "grad_norm": 0.03631274774670601, "learning_rate": 1.2166936487486015e-07, "loss": 0.0026, "step": 9790 }, { "epoch": 51.851851851851855, "grad_norm": 0.024479085579514503, "learning_rate": 1.1041415489045914e-07, "loss": 0.0022, "step": 9800 }, { "epoch": 51.904761904761905, "grad_norm": 0.034816134721040726, "learning_rate": 9.970452942384412e-08, "loss": 0.0028, "step": 9810 }, { "epoch": 51.957671957671955, "grad_norm": 0.10182525962591171, "learning_rate": 8.954060559375754e-08, "loss": 0.0018, "step": 9820 }, { "epoch": 52.01058201058201, "grad_norm": 0.07205993682146072, "learning_rate": 7.99224945512489e-08, "loss": 0.0018, "step": 9830 }, { "epoch": 52.06349206349206, "grad_norm": 0.06608884781599045, "learning_rate": 7.085030147843675e-08, "loss": 0.0014, "step": 9840 }, { "epoch": 52.116402116402114, "grad_norm": 0.053027331829071045, "learning_rate": 6.232412558736523e-08, "loss": 0.003, "step": 9850 }, { "epoch": 52.16931216931217, "grad_norm": 0.02169005200266838, "learning_rate": 5.434406011893822e-08, "loss": 0.0023, "step": 9860 }, { "epoch": 52.22222222222222, "grad_norm": 0.01737486943602562, "learning_rate": 4.6910192341864664e-08, "loss": 0.0025, "step": 9870 }, { "epoch": 52.27513227513227, "grad_norm": 0.03450096771121025, "learning_rate": 4.0022603551737035e-08, "loss": 0.0015, "step": 9880 }, { "epoch": 52.32804232804233, "grad_norm": 0.07834013551473618, "learning_rate": 3.3681369070120985e-08, "loss": 0.0024, "step": 9890 }, { "epoch": 52.38095238095238, "grad_norm": 0.01643853634595871, "learning_rate": 2.7886558243744866e-08, "loss": 0.0017, "step": 9900 }, { "epoch": 52.43386243386244, "grad_norm": 0.040502045303583145, "learning_rate": 2.2638234443722596e-08, "loss": 0.0015, "step": 9910 }, { "epoch": 52.48677248677249, "grad_norm": 0.03804892301559448, "learning_rate": 1.7936455064887504e-08, "loss": 0.0022, "step": 9920 }, { "epoch": 52.53968253968254, "grad_norm": 0.0233614481985569, "learning_rate": 1.378127152514841e-08, "loss": 0.0016, "step": 9930 }, { "epoch": 52.592592592592595, "grad_norm": 0.05919909104704857, "learning_rate": 1.0172729264917857e-08, "loss": 0.0044, "step": 9940 }, { "epoch": 52.645502645502646, "grad_norm": 0.02121146209537983, "learning_rate": 7.1108677466458215e-09, "loss": 0.0023, "step": 9950 }, { "epoch": 52.698412698412696, "grad_norm": 0.029287297278642654, "learning_rate": 4.595720454353414e-09, "loss": 0.0021, "step": 9960 }, { "epoch": 52.75132275132275, "grad_norm": 0.024320444092154503, "learning_rate": 2.627314893294264e-09, "loss": 0.0016, "step": 9970 }, { "epoch": 52.804232804232804, "grad_norm": 0.015103484503924847, "learning_rate": 1.2056725896270048e-09, "loss": 0.0019, "step": 9980 }, { "epoch": 52.857142857142854, "grad_norm": 0.07831604033708572, "learning_rate": 3.308090902098826e-10, "loss": 0.0019, "step": 9990 }, { "epoch": 52.91005291005291, "grad_norm": 0.028462175279855728, "learning_rate": 2.7339624120159555e-12, "loss": 0.0015, "step": 10000 }, { "epoch": 52.91005291005291, "step": 10000, "total_flos": 0.0, "train_loss": 0.009403320060111582, "train_runtime": 20710.9751, "train_samples_per_second": 193.134, "train_steps_per_second": 0.483 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 53, "save_steps": 1000, "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": 50, "trial_name": null, "trial_params": null }