{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 51.02040816326531, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.05102040816326531, "grad_norm": 10.402610778808594, "learning_rate": 1.8e-06, "loss": 0.8659, "step": 10 }, { "epoch": 0.10204081632653061, "grad_norm": 3.46895694732666, "learning_rate": 3.8e-06, "loss": 0.6358, "step": 20 }, { "epoch": 0.15306122448979592, "grad_norm": 2.355771780014038, "learning_rate": 5.8e-06, "loss": 0.3704, "step": 30 }, { "epoch": 0.20408163265306123, "grad_norm": 1.6672266721725464, "learning_rate": 7.8e-06, "loss": 0.2329, "step": 40 }, { "epoch": 0.25510204081632654, "grad_norm": 1.2819832563400269, "learning_rate": 9.800000000000001e-06, "loss": 0.1786, "step": 50 }, { "epoch": 0.30612244897959184, "grad_norm": 0.8300465941429138, "learning_rate": 1.18e-05, "loss": 0.1522, "step": 60 }, { "epoch": 0.35714285714285715, "grad_norm": 0.7117576003074646, "learning_rate": 1.3800000000000002e-05, "loss": 0.1388, "step": 70 }, { "epoch": 0.40816326530612246, "grad_norm": 0.755882978439331, "learning_rate": 1.58e-05, "loss": 0.115, "step": 80 }, { "epoch": 0.45918367346938777, "grad_norm": 0.599662721157074, "learning_rate": 1.78e-05, "loss": 0.1018, "step": 90 }, { "epoch": 0.5102040816326531, "grad_norm": 0.5327855348587036, "learning_rate": 1.9800000000000004e-05, "loss": 0.0865, "step": 100 }, { "epoch": 0.5612244897959183, "grad_norm": 0.5049394369125366, "learning_rate": 2.18e-05, "loss": 0.0862, "step": 110 }, { "epoch": 0.6122448979591837, "grad_norm": 0.6216588616371155, "learning_rate": 2.38e-05, "loss": 0.0742, "step": 120 }, { "epoch": 0.6632653061224489, "grad_norm": 0.5371184349060059, "learning_rate": 2.58e-05, "loss": 0.0627, "step": 130 }, { "epoch": 0.7142857142857143, "grad_norm": 0.6983064413070679, "learning_rate": 2.7800000000000005e-05, "loss": 0.0625, "step": 140 }, { "epoch": 0.7653061224489796, "grad_norm": 0.8011510372161865, "learning_rate": 2.98e-05, "loss": 0.0559, "step": 150 }, { "epoch": 0.8163265306122449, "grad_norm": 0.40350407361984253, "learning_rate": 3.18e-05, "loss": 0.0513, "step": 160 }, { "epoch": 0.8673469387755102, "grad_norm": 0.45743849873542786, "learning_rate": 3.38e-05, "loss": 0.0458, "step": 170 }, { "epoch": 0.9183673469387755, "grad_norm": 0.47269245982170105, "learning_rate": 3.58e-05, "loss": 0.0412, "step": 180 }, { "epoch": 0.9693877551020408, "grad_norm": 0.5295642018318176, "learning_rate": 3.7800000000000004e-05, "loss": 0.041, "step": 190 }, { "epoch": 1.0204081632653061, "grad_norm": 0.7207706570625305, "learning_rate": 3.9800000000000005e-05, "loss": 0.0387, "step": 200 }, { "epoch": 1.0714285714285714, "grad_norm": 0.7374840378761292, "learning_rate": 4.18e-05, "loss": 0.0391, "step": 210 }, { "epoch": 1.1224489795918366, "grad_norm": 0.5862524509429932, "learning_rate": 4.38e-05, "loss": 0.0384, "step": 220 }, { "epoch": 1.1734693877551021, "grad_norm": 0.4840109646320343, "learning_rate": 4.58e-05, "loss": 0.0388, "step": 230 }, { "epoch": 1.2244897959183674, "grad_norm": 0.6424559354782104, "learning_rate": 4.78e-05, "loss": 0.0399, "step": 240 }, { "epoch": 1.2755102040816326, "grad_norm": 0.2996210753917694, "learning_rate": 4.9800000000000004e-05, "loss": 0.0332, "step": 250 }, { "epoch": 1.3265306122448979, "grad_norm": 0.4844838082790375, "learning_rate": 5.1800000000000005e-05, "loss": 0.035, "step": 260 }, { "epoch": 1.3775510204081631, "grad_norm": 0.47933506965637207, "learning_rate": 5.380000000000001e-05, "loss": 0.0339, "step": 270 }, { "epoch": 1.4285714285714286, "grad_norm": 0.47784510254859924, "learning_rate": 5.580000000000001e-05, "loss": 0.0331, "step": 280 }, { "epoch": 1.4795918367346939, "grad_norm": 0.5544247031211853, "learning_rate": 5.7799999999999995e-05, "loss": 0.0318, "step": 290 }, { "epoch": 1.5306122448979593, "grad_norm": 0.7399377822875977, "learning_rate": 5.9800000000000003e-05, "loss": 0.0362, "step": 300 }, { "epoch": 1.5816326530612246, "grad_norm": 0.7247652411460876, "learning_rate": 6.18e-05, "loss": 0.0367, "step": 310 }, { "epoch": 1.6326530612244898, "grad_norm": 0.7861680388450623, "learning_rate": 6.38e-05, "loss": 0.0307, "step": 320 }, { "epoch": 1.683673469387755, "grad_norm": 0.4805033504962921, "learning_rate": 6.58e-05, "loss": 0.0329, "step": 330 }, { "epoch": 1.7346938775510203, "grad_norm": 0.4455874562263489, "learning_rate": 6.780000000000001e-05, "loss": 0.0352, "step": 340 }, { "epoch": 1.7857142857142856, "grad_norm": 0.43569931387901306, "learning_rate": 6.98e-05, "loss": 0.0295, "step": 350 }, { "epoch": 1.836734693877551, "grad_norm": 0.4896446466445923, "learning_rate": 7.18e-05, "loss": 0.0325, "step": 360 }, { "epoch": 1.8877551020408163, "grad_norm": 0.5620689392089844, "learning_rate": 7.38e-05, "loss": 0.0298, "step": 370 }, { "epoch": 1.9387755102040818, "grad_norm": 0.42494043707847595, "learning_rate": 7.58e-05, "loss": 0.03, "step": 380 }, { "epoch": 1.989795918367347, "grad_norm": 0.7252960801124573, "learning_rate": 7.780000000000001e-05, "loss": 0.0306, "step": 390 }, { "epoch": 2.0408163265306123, "grad_norm": 0.31677159667015076, "learning_rate": 7.98e-05, "loss": 0.0294, "step": 400 }, { "epoch": 2.0918367346938775, "grad_norm": 0.63107830286026, "learning_rate": 8.18e-05, "loss": 0.0332, "step": 410 }, { "epoch": 2.142857142857143, "grad_norm": 0.6301782727241516, "learning_rate": 8.38e-05, "loss": 0.0292, "step": 420 }, { "epoch": 2.193877551020408, "grad_norm": 0.7156728506088257, "learning_rate": 8.58e-05, "loss": 0.0317, "step": 430 }, { "epoch": 2.2448979591836733, "grad_norm": 0.5250526070594788, "learning_rate": 8.78e-05, "loss": 0.0282, "step": 440 }, { "epoch": 2.295918367346939, "grad_norm": 0.5226017832756042, "learning_rate": 8.98e-05, "loss": 0.0312, "step": 450 }, { "epoch": 2.3469387755102042, "grad_norm": 0.3892342150211334, "learning_rate": 9.180000000000001e-05, "loss": 0.0285, "step": 460 }, { "epoch": 2.3979591836734695, "grad_norm": 0.34976691007614136, "learning_rate": 9.38e-05, "loss": 0.0263, "step": 470 }, { "epoch": 2.4489795918367347, "grad_norm": 0.5273671746253967, "learning_rate": 9.58e-05, "loss": 0.0272, "step": 480 }, { "epoch": 2.5, "grad_norm": 0.7589205503463745, "learning_rate": 9.78e-05, "loss": 0.0289, "step": 490 }, { "epoch": 2.5510204081632653, "grad_norm": 0.4627077877521515, "learning_rate": 9.98e-05, "loss": 0.0305, "step": 500 }, { "epoch": 2.6020408163265305, "grad_norm": 0.4716263711452484, "learning_rate": 9.9999778549206e-05, "loss": 0.0282, "step": 510 }, { "epoch": 2.6530612244897958, "grad_norm": 0.4921065866947174, "learning_rate": 9.999901304280685e-05, "loss": 0.0284, "step": 520 }, { "epoch": 2.704081632653061, "grad_norm": 0.6303737163543701, "learning_rate": 9.999770075521164e-05, "loss": 0.0289, "step": 530 }, { "epoch": 2.7551020408163263, "grad_norm": 0.6243062019348145, "learning_rate": 9.99958417007713e-05, "loss": 0.0296, "step": 540 }, { "epoch": 2.806122448979592, "grad_norm": 0.39823251962661743, "learning_rate": 9.999343589981615e-05, "loss": 0.026, "step": 550 }, { "epoch": 2.857142857142857, "grad_norm": 0.6608560681343079, "learning_rate": 9.999048337865568e-05, "loss": 0.0302, "step": 560 }, { "epoch": 2.9081632653061225, "grad_norm": 0.7151702046394348, "learning_rate": 9.998698416957815e-05, "loss": 0.0328, "step": 570 }, { "epoch": 2.9591836734693877, "grad_norm": 0.4519721567630768, "learning_rate": 9.998293831085037e-05, "loss": 0.0313, "step": 580 }, { "epoch": 3.010204081632653, "grad_norm": 0.31802868843078613, "learning_rate": 9.997834584671719e-05, "loss": 0.0264, "step": 590 }, { "epoch": 3.061224489795918, "grad_norm": 0.3028657138347626, "learning_rate": 9.997320682740107e-05, "loss": 0.027, "step": 600 }, { "epoch": 3.1122448979591835, "grad_norm": 0.4726928472518921, "learning_rate": 9.996752130910149e-05, "loss": 0.0258, "step": 610 }, { "epoch": 3.163265306122449, "grad_norm": 0.20876920223236084, "learning_rate": 9.99612893539944e-05, "loss": 0.0247, "step": 620 }, { "epoch": 3.2142857142857144, "grad_norm": 0.4355166554450989, "learning_rate": 9.995451103023144e-05, "loss": 0.0264, "step": 630 }, { "epoch": 3.2653061224489797, "grad_norm": 0.3747992217540741, "learning_rate": 9.994718641193928e-05, "loss": 0.0269, "step": 640 }, { "epoch": 3.316326530612245, "grad_norm": 0.2605808675289154, "learning_rate": 9.993931557921874e-05, "loss": 0.0273, "step": 650 }, { "epoch": 3.36734693877551, "grad_norm": 0.31417006254196167, "learning_rate": 9.993089861814402e-05, "loss": 0.0257, "step": 660 }, { "epoch": 3.4183673469387754, "grad_norm": 0.4274827837944031, "learning_rate": 9.992193562076166e-05, "loss": 0.0249, "step": 670 }, { "epoch": 3.4693877551020407, "grad_norm": 0.3169364929199219, "learning_rate": 9.991242668508954e-05, "loss": 0.023, "step": 680 }, { "epoch": 3.520408163265306, "grad_norm": 0.37575116753578186, "learning_rate": 9.990237191511587e-05, "loss": 0.0215, "step": 690 }, { "epoch": 3.571428571428571, "grad_norm": 0.27361607551574707, "learning_rate": 9.989177142079802e-05, "loss": 0.0238, "step": 700 }, { "epoch": 3.622448979591837, "grad_norm": 0.4054713547229767, "learning_rate": 9.988062531806126e-05, "loss": 0.0237, "step": 710 }, { "epoch": 3.673469387755102, "grad_norm": 0.3417719900608063, "learning_rate": 9.986893372879762e-05, "loss": 0.0229, "step": 720 }, { "epoch": 3.7244897959183674, "grad_norm": 0.37635910511016846, "learning_rate": 9.985669678086443e-05, "loss": 0.0243, "step": 730 }, { "epoch": 3.7755102040816326, "grad_norm": 0.3380339741706848, "learning_rate": 9.984391460808298e-05, "loss": 0.0258, "step": 740 }, { "epoch": 3.826530612244898, "grad_norm": 0.38736411929130554, "learning_rate": 9.983058735023709e-05, "loss": 0.0251, "step": 750 }, { "epoch": 3.877551020408163, "grad_norm": 0.30285143852233887, "learning_rate": 9.98167151530715e-05, "loss": 0.0251, "step": 760 }, { "epoch": 3.928571428571429, "grad_norm": 0.36467981338500977, "learning_rate": 9.980229816829034e-05, "loss": 0.0255, "step": 770 }, { "epoch": 3.979591836734694, "grad_norm": 0.2766786515712738, "learning_rate": 9.978733655355544e-05, "loss": 0.0235, "step": 780 }, { "epoch": 4.030612244897959, "grad_norm": 0.30421629548072815, "learning_rate": 9.977183047248464e-05, "loss": 0.0219, "step": 790 }, { "epoch": 4.081632653061225, "grad_norm": 0.2804276645183563, "learning_rate": 9.975578009464992e-05, "loss": 0.0241, "step": 800 }, { "epoch": 4.13265306122449, "grad_norm": 0.36253225803375244, "learning_rate": 9.97391855955757e-05, "loss": 0.0205, "step": 810 }, { "epoch": 4.183673469387755, "grad_norm": 0.27331048250198364, "learning_rate": 9.972204715673669e-05, "loss": 0.0217, "step": 820 }, { "epoch": 4.23469387755102, "grad_norm": 0.19847875833511353, "learning_rate": 9.970436496555617e-05, "loss": 0.0219, "step": 830 }, { "epoch": 4.285714285714286, "grad_norm": 0.4231332540512085, "learning_rate": 9.968613921540373e-05, "loss": 0.0222, "step": 840 }, { "epoch": 4.336734693877551, "grad_norm": 0.26923805475234985, "learning_rate": 9.966737010559326e-05, "loss": 0.0231, "step": 850 }, { "epoch": 4.387755102040816, "grad_norm": 0.3702804744243622, "learning_rate": 9.964805784138072e-05, "loss": 0.0226, "step": 860 }, { "epoch": 4.438775510204081, "grad_norm": 0.3984103798866272, "learning_rate": 9.962820263396195e-05, "loss": 0.0226, "step": 870 }, { "epoch": 4.489795918367347, "grad_norm": 0.3307846784591675, "learning_rate": 9.960780470047033e-05, "loss": 0.0262, "step": 880 }, { "epoch": 4.540816326530612, "grad_norm": 0.31499937176704407, "learning_rate": 9.958686426397437e-05, "loss": 0.0234, "step": 890 }, { "epoch": 4.591836734693878, "grad_norm": 0.35475045442581177, "learning_rate": 9.956538155347534e-05, "loss": 0.0256, "step": 900 }, { "epoch": 4.642857142857143, "grad_norm": 0.38957640528678894, "learning_rate": 9.95433568039047e-05, "loss": 0.0223, "step": 910 }, { "epoch": 4.6938775510204085, "grad_norm": 0.2586624026298523, "learning_rate": 9.952079025612162e-05, "loss": 0.0221, "step": 920 }, { "epoch": 4.744897959183674, "grad_norm": 0.3217252492904663, "learning_rate": 9.949768215691022e-05, "loss": 0.0225, "step": 930 }, { "epoch": 4.795918367346939, "grad_norm": 0.37759119272232056, "learning_rate": 9.9474032758977e-05, "loss": 0.021, "step": 940 }, { "epoch": 4.846938775510204, "grad_norm": 0.3058706820011139, "learning_rate": 9.944984232094794e-05, "loss": 0.022, "step": 950 }, { "epoch": 4.8979591836734695, "grad_norm": 0.2515983283519745, "learning_rate": 9.942511110736584e-05, "loss": 0.0222, "step": 960 }, { "epoch": 4.948979591836735, "grad_norm": 0.3808935284614563, "learning_rate": 9.939983938868726e-05, "loss": 0.0231, "step": 970 }, { "epoch": 5.0, "grad_norm": 0.3011344373226166, "learning_rate": 9.93740274412797e-05, "loss": 0.0204, "step": 980 }, { "epoch": 5.051020408163265, "grad_norm": 0.396373987197876, "learning_rate": 9.934767554741846e-05, "loss": 0.0241, "step": 990 }, { "epoch": 5.1020408163265305, "grad_norm": 0.2780967354774475, "learning_rate": 9.932078399528361e-05, "loss": 0.0199, "step": 1000 }, { "epoch": 5.153061224489796, "grad_norm": 0.4155687689781189, "learning_rate": 9.929335307895689e-05, "loss": 0.0206, "step": 1010 }, { "epoch": 5.204081632653061, "grad_norm": 0.3944474756717682, "learning_rate": 9.926538309841839e-05, "loss": 0.0246, "step": 1020 }, { "epoch": 5.255102040816326, "grad_norm": 0.24255912005901337, "learning_rate": 9.923687435954334e-05, "loss": 0.0206, "step": 1030 }, { "epoch": 5.3061224489795915, "grad_norm": 0.26634329557418823, "learning_rate": 9.920782717409873e-05, "loss": 0.02, "step": 1040 }, { "epoch": 5.357142857142857, "grad_norm": 0.21066752076148987, "learning_rate": 9.917824185973994e-05, "loss": 0.0193, "step": 1050 }, { "epoch": 5.408163265306122, "grad_norm": 0.24421487748622894, "learning_rate": 9.914811874000723e-05, "loss": 0.0184, "step": 1060 }, { "epoch": 5.459183673469388, "grad_norm": 0.25356921553611755, "learning_rate": 9.911745814432218e-05, "loss": 0.0198, "step": 1070 }, { "epoch": 5.510204081632653, "grad_norm": 0.5104711651802063, "learning_rate": 9.90862604079842e-05, "loss": 0.0215, "step": 1080 }, { "epoch": 5.561224489795919, "grad_norm": 0.3930110037326813, "learning_rate": 9.90545258721667e-05, "loss": 0.0198, "step": 1090 }, { "epoch": 5.612244897959184, "grad_norm": 0.2771541178226471, "learning_rate": 9.90222548839135e-05, "loss": 0.0193, "step": 1100 }, { "epoch": 5.663265306122449, "grad_norm": 0.40832555294036865, "learning_rate": 9.898944779613495e-05, "loss": 0.0182, "step": 1110 }, { "epoch": 5.714285714285714, "grad_norm": 0.16354826092720032, "learning_rate": 9.89561049676041e-05, "loss": 0.0185, "step": 1120 }, { "epoch": 5.76530612244898, "grad_norm": 0.44468289613723755, "learning_rate": 9.89222267629528e-05, "loss": 0.0208, "step": 1130 }, { "epoch": 5.816326530612245, "grad_norm": 0.3538622260093689, "learning_rate": 9.888781355266763e-05, "loss": 0.0198, "step": 1140 }, { "epoch": 5.86734693877551, "grad_norm": 0.37022602558135986, "learning_rate": 9.885286571308598e-05, "loss": 0.0185, "step": 1150 }, { "epoch": 5.918367346938775, "grad_norm": 0.33385682106018066, "learning_rate": 9.881738362639182e-05, "loss": 0.0203, "step": 1160 }, { "epoch": 5.969387755102041, "grad_norm": 0.3826387822628021, "learning_rate": 9.878136768061154e-05, "loss": 0.0226, "step": 1170 }, { "epoch": 6.020408163265306, "grad_norm": 0.3458908498287201, "learning_rate": 9.874481826960979e-05, "loss": 0.0202, "step": 1180 }, { "epoch": 6.071428571428571, "grad_norm": 0.31151530146598816, "learning_rate": 9.870773579308503e-05, "loss": 0.0182, "step": 1190 }, { "epoch": 6.122448979591836, "grad_norm": 0.2888461947441101, "learning_rate": 9.867012065656533e-05, "loss": 0.0202, "step": 1200 }, { "epoch": 6.173469387755102, "grad_norm": 0.24357391893863678, "learning_rate": 9.863197327140376e-05, "loss": 0.0175, "step": 1210 }, { "epoch": 6.224489795918367, "grad_norm": 0.3024928569793701, "learning_rate": 9.859329405477403e-05, "loss": 0.0176, "step": 1220 }, { "epoch": 6.275510204081632, "grad_norm": 0.3485017716884613, "learning_rate": 9.855408342966585e-05, "loss": 0.0193, "step": 1230 }, { "epoch": 6.326530612244898, "grad_norm": 0.2527174949645996, "learning_rate": 9.851434182488033e-05, "loss": 0.0195, "step": 1240 }, { "epoch": 6.377551020408164, "grad_norm": 0.2952190339565277, "learning_rate": 9.84740696750253e-05, "loss": 0.0184, "step": 1250 }, { "epoch": 6.428571428571429, "grad_norm": 0.331484854221344, "learning_rate": 9.843326742051055e-05, "loss": 0.0186, "step": 1260 }, { "epoch": 6.479591836734694, "grad_norm": 0.35059869289398193, "learning_rate": 9.839193550754297e-05, "loss": 0.0184, "step": 1270 }, { "epoch": 6.530612244897959, "grad_norm": 0.2054317146539688, "learning_rate": 9.835007438812177e-05, "loss": 0.0201, "step": 1280 }, { "epoch": 6.581632653061225, "grad_norm": 0.3477015495300293, "learning_rate": 9.830768452003341e-05, "loss": 0.0193, "step": 1290 }, { "epoch": 6.63265306122449, "grad_norm": 0.2933940291404724, "learning_rate": 9.826476636684671e-05, "loss": 0.0187, "step": 1300 }, { "epoch": 6.683673469387755, "grad_norm": 0.3926159739494324, "learning_rate": 9.822132039790773e-05, "loss": 0.0174, "step": 1310 }, { "epoch": 6.73469387755102, "grad_norm": 0.3543316721916199, "learning_rate": 9.817734708833461e-05, "loss": 0.0163, "step": 1320 }, { "epoch": 6.785714285714286, "grad_norm": 0.35534197092056274, "learning_rate": 9.813284691901243e-05, "loss": 0.0211, "step": 1330 }, { "epoch": 6.836734693877551, "grad_norm": 0.31214722990989685, "learning_rate": 9.808782037658792e-05, "loss": 0.02, "step": 1340 }, { "epoch": 6.887755102040816, "grad_norm": 0.3505015969276428, "learning_rate": 9.804226795346411e-05, "loss": 0.0195, "step": 1350 }, { "epoch": 6.938775510204081, "grad_norm": 0.2676263451576233, "learning_rate": 9.799619014779503e-05, "loss": 0.0186, "step": 1360 }, { "epoch": 6.989795918367347, "grad_norm": 0.31860265135765076, "learning_rate": 9.794958746348013e-05, "loss": 0.0182, "step": 1370 }, { "epoch": 7.040816326530612, "grad_norm": 0.28607189655303955, "learning_rate": 9.790246041015896e-05, "loss": 0.0167, "step": 1380 }, { "epoch": 7.091836734693878, "grad_norm": 0.17099547386169434, "learning_rate": 9.785480950320538e-05, "loss": 0.0173, "step": 1390 }, { "epoch": 7.142857142857143, "grad_norm": 0.1480858027935028, "learning_rate": 9.78066352637221e-05, "loss": 0.017, "step": 1400 }, { "epoch": 7.1938775510204085, "grad_norm": 0.18142010271549225, "learning_rate": 9.775793821853488e-05, "loss": 0.0161, "step": 1410 }, { "epoch": 7.244897959183674, "grad_norm": 0.3372935354709625, "learning_rate": 9.77087189001868e-05, "loss": 0.0166, "step": 1420 }, { "epoch": 7.295918367346939, "grad_norm": 0.2566656768321991, "learning_rate": 9.765897784693243e-05, "loss": 0.0161, "step": 1430 }, { "epoch": 7.346938775510204, "grad_norm": 0.20847970247268677, "learning_rate": 9.760871560273197e-05, "loss": 0.0161, "step": 1440 }, { "epoch": 7.3979591836734695, "grad_norm": 0.21302850544452667, "learning_rate": 9.755793271724526e-05, "loss": 0.0166, "step": 1450 }, { "epoch": 7.448979591836735, "grad_norm": 0.28545570373535156, "learning_rate": 9.750662974582584e-05, "loss": 0.0194, "step": 1460 }, { "epoch": 7.5, "grad_norm": 0.3417540490627289, "learning_rate": 9.745480724951473e-05, "loss": 0.0162, "step": 1470 }, { "epoch": 7.551020408163265, "grad_norm": 0.27870655059814453, "learning_rate": 9.740246579503447e-05, "loss": 0.0182, "step": 1480 }, { "epoch": 7.6020408163265305, "grad_norm": 0.34729447960853577, "learning_rate": 9.734960595478284e-05, "loss": 0.0182, "step": 1490 }, { "epoch": 7.653061224489796, "grad_norm": 0.3608493208885193, "learning_rate": 9.729622830682657e-05, "loss": 0.0182, "step": 1500 }, { "epoch": 7.704081632653061, "grad_norm": 0.20140162110328674, "learning_rate": 9.724233343489504e-05, "loss": 0.0199, "step": 1510 }, { "epoch": 7.755102040816326, "grad_norm": 0.3673495054244995, "learning_rate": 9.718792192837396e-05, "loss": 0.0149, "step": 1520 }, { "epoch": 7.8061224489795915, "grad_norm": 0.40889599919319153, "learning_rate": 9.713299438229886e-05, "loss": 0.0179, "step": 1530 }, { "epoch": 7.857142857142857, "grad_norm": 0.29143819212913513, "learning_rate": 9.707755139734855e-05, "loss": 0.0183, "step": 1540 }, { "epoch": 7.908163265306122, "grad_norm": 0.3438901901245117, "learning_rate": 9.702159357983866e-05, "loss": 0.0179, "step": 1550 }, { "epoch": 7.959183673469388, "grad_norm": 0.3607460558414459, "learning_rate": 9.696512154171492e-05, "loss": 0.0178, "step": 1560 }, { "epoch": 8.010204081632653, "grad_norm": 0.29925620555877686, "learning_rate": 9.690813590054645e-05, "loss": 0.0187, "step": 1570 }, { "epoch": 8.061224489795919, "grad_norm": 0.5310401916503906, "learning_rate": 9.685063727951914e-05, "loss": 0.0176, "step": 1580 }, { "epoch": 8.112244897959183, "grad_norm": 0.35968977212905884, "learning_rate": 9.679262630742865e-05, "loss": 0.018, "step": 1590 }, { "epoch": 8.16326530612245, "grad_norm": 0.3266374468803406, "learning_rate": 9.673410361867373e-05, "loss": 0.0189, "step": 1600 }, { "epoch": 8.214285714285714, "grad_norm": 0.1696818619966507, "learning_rate": 9.667506985324909e-05, "loss": 0.0164, "step": 1610 }, { "epoch": 8.26530612244898, "grad_norm": 0.31890949606895447, "learning_rate": 9.661552565673855e-05, "loss": 0.0157, "step": 1620 }, { "epoch": 8.316326530612244, "grad_norm": 0.2842443287372589, "learning_rate": 9.655547168030789e-05, "loss": 0.0166, "step": 1630 }, { "epoch": 8.36734693877551, "grad_norm": 0.24909497797489166, "learning_rate": 9.649490858069777e-05, "loss": 0.015, "step": 1640 }, { "epoch": 8.418367346938776, "grad_norm": 0.2549191117286682, "learning_rate": 9.643383702021658e-05, "loss": 0.0158, "step": 1650 }, { "epoch": 8.46938775510204, "grad_norm": 0.4351326525211334, "learning_rate": 9.637225766673307e-05, "loss": 0.0149, "step": 1660 }, { "epoch": 8.520408163265307, "grad_norm": 0.546602189540863, "learning_rate": 9.631017119366922e-05, "loss": 0.0166, "step": 1670 }, { "epoch": 8.571428571428571, "grad_norm": 0.31468871235847473, "learning_rate": 9.624757827999273e-05, "loss": 0.0174, "step": 1680 }, { "epoch": 8.622448979591837, "grad_norm": 0.25562915205955505, "learning_rate": 9.618447961020971e-05, "loss": 0.0197, "step": 1690 }, { "epoch": 8.673469387755102, "grad_norm": 0.4273330569267273, "learning_rate": 9.612087587435707e-05, "loss": 0.018, "step": 1700 }, { "epoch": 8.724489795918368, "grad_norm": 0.30122387409210205, "learning_rate": 9.605676776799508e-05, "loss": 0.0176, "step": 1710 }, { "epoch": 8.775510204081632, "grad_norm": 0.3553696572780609, "learning_rate": 9.599215599219973e-05, "loss": 0.0167, "step": 1720 }, { "epoch": 8.826530612244898, "grad_norm": 0.34977346658706665, "learning_rate": 9.592704125355505e-05, "loss": 0.0169, "step": 1730 }, { "epoch": 8.877551020408163, "grad_norm": 0.24394738674163818, "learning_rate": 9.586142426414538e-05, "loss": 0.015, "step": 1740 }, { "epoch": 8.928571428571429, "grad_norm": 0.16736595332622528, "learning_rate": 9.57953057415476e-05, "loss": 0.0171, "step": 1750 }, { "epoch": 8.979591836734693, "grad_norm": 0.22622261941432953, "learning_rate": 9.572868640882328e-05, "loss": 0.0146, "step": 1760 }, { "epoch": 9.03061224489796, "grad_norm": 0.1895446628332138, "learning_rate": 9.56615669945108e-05, "loss": 0.0162, "step": 1770 }, { "epoch": 9.081632653061224, "grad_norm": 0.2732166349887848, "learning_rate": 9.55939482326173e-05, "loss": 0.017, "step": 1780 }, { "epoch": 9.13265306122449, "grad_norm": 0.21778897941112518, "learning_rate": 9.552583086261069e-05, "loss": 0.0176, "step": 1790 }, { "epoch": 9.183673469387756, "grad_norm": 0.26416125893592834, "learning_rate": 9.545721562941168e-05, "loss": 0.0161, "step": 1800 }, { "epoch": 9.23469387755102, "grad_norm": 0.45024582743644714, "learning_rate": 9.538810328338543e-05, "loss": 0.0146, "step": 1810 }, { "epoch": 9.285714285714286, "grad_norm": 0.378096342086792, "learning_rate": 9.531849458033349e-05, "loss": 0.0165, "step": 1820 }, { "epoch": 9.33673469387755, "grad_norm": 0.31736621260643005, "learning_rate": 9.524839028148547e-05, "loss": 0.0161, "step": 1830 }, { "epoch": 9.387755102040817, "grad_norm": 0.4405612647533417, "learning_rate": 9.517779115349077e-05, "loss": 0.0175, "step": 1840 }, { "epoch": 9.438775510204081, "grad_norm": 0.29838114976882935, "learning_rate": 9.510669796841014e-05, "loss": 0.0167, "step": 1850 }, { "epoch": 9.489795918367347, "grad_norm": 0.2295493185520172, "learning_rate": 9.503511150370727e-05, "loss": 0.0189, "step": 1860 }, { "epoch": 9.540816326530612, "grad_norm": 0.2769347131252289, "learning_rate": 9.496303254224024e-05, "loss": 0.0167, "step": 1870 }, { "epoch": 9.591836734693878, "grad_norm": 0.2015368789434433, "learning_rate": 9.489046187225306e-05, "loss": 0.0155, "step": 1880 }, { "epoch": 9.642857142857142, "grad_norm": 0.2741110920906067, "learning_rate": 9.481740028736692e-05, "loss": 0.0176, "step": 1890 }, { "epoch": 9.693877551020408, "grad_norm": 0.36915838718414307, "learning_rate": 9.474384858657164e-05, "loss": 0.0169, "step": 1900 }, { "epoch": 9.744897959183673, "grad_norm": 0.23773711919784546, "learning_rate": 9.466980757421679e-05, "loss": 0.0163, "step": 1910 }, { "epoch": 9.795918367346939, "grad_norm": 0.5074164867401123, "learning_rate": 9.459527806000305e-05, "loss": 0.0173, "step": 1920 }, { "epoch": 9.846938775510203, "grad_norm": 0.4252951741218567, "learning_rate": 9.452026085897325e-05, "loss": 0.017, "step": 1930 }, { "epoch": 9.89795918367347, "grad_norm": 0.35428088903427124, "learning_rate": 9.444475679150348e-05, "loss": 0.0167, "step": 1940 }, { "epoch": 9.948979591836736, "grad_norm": 0.26375457644462585, "learning_rate": 9.436876668329411e-05, "loss": 0.0163, "step": 1950 }, { "epoch": 10.0, "grad_norm": 0.18078748881816864, "learning_rate": 9.429229136536079e-05, "loss": 0.0167, "step": 1960 }, { "epoch": 10.051020408163266, "grad_norm": 0.2727520763874054, "learning_rate": 9.421533167402534e-05, "loss": 0.0169, "step": 1970 }, { "epoch": 10.10204081632653, "grad_norm": 0.16076399385929108, "learning_rate": 9.413788845090666e-05, "loss": 0.0161, "step": 1980 }, { "epoch": 10.153061224489797, "grad_norm": 0.2337694764137268, "learning_rate": 9.405996254291136e-05, "loss": 0.0153, "step": 1990 }, { "epoch": 10.204081632653061, "grad_norm": 0.3368735909461975, "learning_rate": 9.398155480222474e-05, "loss": 0.0164, "step": 2000 }, { "epoch": 10.255102040816327, "grad_norm": 0.2857428789138794, "learning_rate": 9.390266608630128e-05, "loss": 0.0152, "step": 2010 }, { "epoch": 10.306122448979592, "grad_norm": 0.2701791822910309, "learning_rate": 9.38232972578553e-05, "loss": 0.0171, "step": 2020 }, { "epoch": 10.357142857142858, "grad_norm": 0.2637172341346741, "learning_rate": 9.374344918485164e-05, "loss": 0.0165, "step": 2030 }, { "epoch": 10.408163265306122, "grad_norm": 0.19126808643341064, "learning_rate": 9.366312274049602e-05, "loss": 0.0148, "step": 2040 }, { "epoch": 10.459183673469388, "grad_norm": 0.23476313054561615, "learning_rate": 9.358231880322554e-05, "loss": 0.0155, "step": 2050 }, { "epoch": 10.510204081632653, "grad_norm": 0.17433427274227142, "learning_rate": 9.350103825669916e-05, "loss": 0.0151, "step": 2060 }, { "epoch": 10.561224489795919, "grad_norm": 0.3013644814491272, "learning_rate": 9.341928198978787e-05, "loss": 0.0155, "step": 2070 }, { "epoch": 10.612244897959183, "grad_norm": 0.21649682521820068, "learning_rate": 9.333705089656512e-05, "loss": 0.0135, "step": 2080 }, { "epoch": 10.66326530612245, "grad_norm": 0.3561263084411621, "learning_rate": 9.325434587629698e-05, "loss": 0.0177, "step": 2090 }, { "epoch": 10.714285714285714, "grad_norm": 0.20543578267097473, "learning_rate": 9.31711678334323e-05, "loss": 0.0161, "step": 2100 }, { "epoch": 10.76530612244898, "grad_norm": 0.18536753952503204, "learning_rate": 9.308751767759282e-05, "loss": 0.0132, "step": 2110 }, { "epoch": 10.816326530612244, "grad_norm": 0.26259610056877136, "learning_rate": 9.300339632356325e-05, "loss": 0.0143, "step": 2120 }, { "epoch": 10.86734693877551, "grad_norm": 0.2785841226577759, "learning_rate": 9.291880469128124e-05, "loss": 0.0166, "step": 2130 }, { "epoch": 10.918367346938776, "grad_norm": 0.14623740315437317, "learning_rate": 9.283374370582732e-05, "loss": 0.0142, "step": 2140 }, { "epoch": 10.96938775510204, "grad_norm": 0.1685599535703659, "learning_rate": 9.274821429741482e-05, "loss": 0.0144, "step": 2150 }, { "epoch": 11.020408163265307, "grad_norm": 0.2024644911289215, "learning_rate": 9.266221740137961e-05, "loss": 0.0141, "step": 2160 }, { "epoch": 11.071428571428571, "grad_norm": 0.38779518008232117, "learning_rate": 9.257575395817001e-05, "loss": 0.0144, "step": 2170 }, { "epoch": 11.122448979591837, "grad_norm": 0.28561678528785706, "learning_rate": 9.248882491333637e-05, "loss": 0.0147, "step": 2180 }, { "epoch": 11.173469387755102, "grad_norm": 0.2720696032047272, "learning_rate": 9.240143121752076e-05, "loss": 0.0139, "step": 2190 }, { "epoch": 11.224489795918368, "grad_norm": 0.3915072977542877, "learning_rate": 9.23135738264467e-05, "loss": 0.0183, "step": 2200 }, { "epoch": 11.275510204081632, "grad_norm": 0.19509173929691315, "learning_rate": 9.222525370090849e-05, "loss": 0.0158, "step": 2210 }, { "epoch": 11.326530612244898, "grad_norm": 0.24125432968139648, "learning_rate": 9.213647180676088e-05, "loss": 0.0179, "step": 2220 }, { "epoch": 11.377551020408163, "grad_norm": 0.3016866445541382, "learning_rate": 9.204722911490846e-05, "loss": 0.0153, "step": 2230 }, { "epoch": 11.428571428571429, "grad_norm": 0.21861532330513, "learning_rate": 9.1957526601295e-05, "loss": 0.0147, "step": 2240 }, { "epoch": 11.479591836734693, "grad_norm": 0.20517617464065552, "learning_rate": 9.186736524689281e-05, "loss": 0.0157, "step": 2250 }, { "epoch": 11.53061224489796, "grad_norm": 0.20268233120441437, "learning_rate": 9.177674603769204e-05, "loss": 0.0152, "step": 2260 }, { "epoch": 11.581632653061224, "grad_norm": 0.19584843516349792, "learning_rate": 9.168566996468983e-05, "loss": 0.015, "step": 2270 }, { "epoch": 11.63265306122449, "grad_norm": 0.3113105297088623, "learning_rate": 9.159413802387951e-05, "loss": 0.013, "step": 2280 }, { "epoch": 11.683673469387756, "grad_norm": 0.19416527450084686, "learning_rate": 9.150215121623974e-05, "loss": 0.0151, "step": 2290 }, { "epoch": 11.73469387755102, "grad_norm": 0.1969633251428604, "learning_rate": 9.140971054772349e-05, "loss": 0.014, "step": 2300 }, { "epoch": 11.785714285714286, "grad_norm": 0.32190269231796265, "learning_rate": 9.131681702924713e-05, "loss": 0.015, "step": 2310 }, { "epoch": 11.83673469387755, "grad_norm": 0.1873030662536621, "learning_rate": 9.122347167667926e-05, "loss": 0.0145, "step": 2320 }, { "epoch": 11.887755102040817, "grad_norm": 0.3558906316757202, "learning_rate": 9.112967551082973e-05, "loss": 0.0175, "step": 2330 }, { "epoch": 11.938775510204081, "grad_norm": 0.2542678117752075, "learning_rate": 9.103542955743835e-05, "loss": 0.0132, "step": 2340 }, { "epoch": 11.989795918367347, "grad_norm": 0.2868235409259796, "learning_rate": 9.094073484716381e-05, "loss": 0.017, "step": 2350 }, { "epoch": 12.040816326530612, "grad_norm": 0.3411066234111786, "learning_rate": 9.084559241557226e-05, "loss": 0.0144, "step": 2360 }, { "epoch": 12.091836734693878, "grad_norm": 0.2043617218732834, "learning_rate": 9.075000330312608e-05, "loss": 0.0143, "step": 2370 }, { "epoch": 12.142857142857142, "grad_norm": 0.32890254259109497, "learning_rate": 9.065396855517253e-05, "loss": 0.0151, "step": 2380 }, { "epoch": 12.193877551020408, "grad_norm": 0.1941353976726532, "learning_rate": 9.055748922193219e-05, "loss": 0.0142, "step": 2390 }, { "epoch": 12.244897959183673, "grad_norm": 0.28506866097450256, "learning_rate": 9.046056635848761e-05, "loss": 0.0142, "step": 2400 }, { "epoch": 12.295918367346939, "grad_norm": 0.32878702878952026, "learning_rate": 9.036320102477169e-05, "loss": 0.0121, "step": 2410 }, { "epoch": 12.346938775510203, "grad_norm": 0.2457260936498642, "learning_rate": 9.02653942855561e-05, "loss": 0.0188, "step": 2420 }, { "epoch": 12.39795918367347, "grad_norm": 0.21025685966014862, "learning_rate": 9.016714721043971e-05, "loss": 0.0155, "step": 2430 }, { "epoch": 12.448979591836734, "grad_norm": 0.17360110580921173, "learning_rate": 9.006846087383675e-05, "loss": 0.0151, "step": 2440 }, { "epoch": 12.5, "grad_norm": 0.21136237680912018, "learning_rate": 8.996933635496523e-05, "loss": 0.0146, "step": 2450 }, { "epoch": 12.551020408163264, "grad_norm": 0.13748657703399658, "learning_rate": 8.986977473783498e-05, "loss": 0.0133, "step": 2460 }, { "epoch": 12.60204081632653, "grad_norm": 0.22515879571437836, "learning_rate": 8.97697771112359e-05, "loss": 0.0137, "step": 2470 }, { "epoch": 12.653061224489797, "grad_norm": 0.21475346386432648, "learning_rate": 8.966934456872602e-05, "loss": 0.0132, "step": 2480 }, { "epoch": 12.704081632653061, "grad_norm": 0.17272284626960754, "learning_rate": 8.95684782086195e-05, "loss": 0.017, "step": 2490 }, { "epoch": 12.755102040816327, "grad_norm": 0.19078773260116577, "learning_rate": 8.946717913397476e-05, "loss": 0.0133, "step": 2500 }, { "epoch": 12.806122448979592, "grad_norm": 0.25016656517982483, "learning_rate": 8.93654484525822e-05, "loss": 0.015, "step": 2510 }, { "epoch": 12.857142857142858, "grad_norm": 0.39352351427078247, "learning_rate": 8.926328727695226e-05, "loss": 0.0141, "step": 2520 }, { "epoch": 12.908163265306122, "grad_norm": 0.22416472434997559, "learning_rate": 8.916069672430319e-05, "loss": 0.0144, "step": 2530 }, { "epoch": 12.959183673469388, "grad_norm": 0.2705744802951813, "learning_rate": 8.905767791654884e-05, "loss": 0.0133, "step": 2540 }, { "epoch": 13.010204081632653, "grad_norm": 0.3071611523628235, "learning_rate": 8.895423198028638e-05, "loss": 0.0164, "step": 2550 }, { "epoch": 13.061224489795919, "grad_norm": 0.2733868956565857, "learning_rate": 8.885036004678402e-05, "loss": 0.0161, "step": 2560 }, { "epoch": 13.112244897959183, "grad_norm": 0.23020657896995544, "learning_rate": 8.874606325196857e-05, "loss": 0.0141, "step": 2570 }, { "epoch": 13.16326530612245, "grad_norm": 0.28492066264152527, "learning_rate": 8.864134273641304e-05, "loss": 0.0142, "step": 2580 }, { "epoch": 13.214285714285714, "grad_norm": 0.29224517941474915, "learning_rate": 8.853619964532427e-05, "loss": 0.0128, "step": 2590 }, { "epoch": 13.26530612244898, "grad_norm": 0.2016928493976593, "learning_rate": 8.843063512853019e-05, "loss": 0.0166, "step": 2600 }, { "epoch": 13.316326530612244, "grad_norm": 0.2512950897216797, "learning_rate": 8.832465034046749e-05, "loss": 0.0151, "step": 2610 }, { "epoch": 13.36734693877551, "grad_norm": 0.2311781644821167, "learning_rate": 8.821824644016882e-05, "loss": 0.0138, "step": 2620 }, { "epoch": 13.418367346938776, "grad_norm": 0.1771167367696762, "learning_rate": 8.811142459125019e-05, "loss": 0.0135, "step": 2630 }, { "epoch": 13.46938775510204, "grad_norm": 0.255623459815979, "learning_rate": 8.800418596189822e-05, "loss": 0.0142, "step": 2640 }, { "epoch": 13.520408163265307, "grad_norm": 0.3858594596385956, "learning_rate": 8.789653172485737e-05, "loss": 0.014, "step": 2650 }, { "epoch": 13.571428571428571, "grad_norm": 0.31865981221199036, "learning_rate": 8.778846305741715e-05, "loss": 0.0138, "step": 2660 }, { "epoch": 13.622448979591837, "grad_norm": 0.21795199811458588, "learning_rate": 8.767998114139918e-05, "loss": 0.014, "step": 2670 }, { "epoch": 13.673469387755102, "grad_norm": 0.3378768861293793, "learning_rate": 8.757108716314429e-05, "loss": 0.0146, "step": 2680 }, { "epoch": 13.724489795918368, "grad_norm": 0.308027446269989, "learning_rate": 8.746178231349962e-05, "loss": 0.0155, "step": 2690 }, { "epoch": 13.775510204081632, "grad_norm": 0.256556898355484, "learning_rate": 8.735206778780549e-05, "loss": 0.0135, "step": 2700 }, { "epoch": 13.826530612244898, "grad_norm": 0.33553218841552734, "learning_rate": 8.724194478588234e-05, "loss": 0.0135, "step": 2710 }, { "epoch": 13.877551020408163, "grad_norm": 0.2881321609020233, "learning_rate": 8.713141451201772e-05, "loss": 0.0141, "step": 2720 }, { "epoch": 13.928571428571429, "grad_norm": 0.2434074729681015, "learning_rate": 8.702047817495295e-05, "loss": 0.0142, "step": 2730 }, { "epoch": 13.979591836734693, "grad_norm": 0.4652666747570038, "learning_rate": 8.69091369878701e-05, "loss": 0.013, "step": 2740 }, { "epoch": 14.03061224489796, "grad_norm": 0.2455073893070221, "learning_rate": 8.679739216837849e-05, "loss": 0.012, "step": 2750 }, { "epoch": 14.081632653061224, "grad_norm": 0.36831730604171753, "learning_rate": 8.66852449385016e-05, "loss": 0.013, "step": 2760 }, { "epoch": 14.13265306122449, "grad_norm": 0.14272406697273254, "learning_rate": 8.657269652466356e-05, "loss": 0.0152, "step": 2770 }, { "epoch": 14.183673469387756, "grad_norm": 0.38238051533699036, "learning_rate": 8.645974815767577e-05, "loss": 0.0124, "step": 2780 }, { "epoch": 14.23469387755102, "grad_norm": 0.2490607649087906, "learning_rate": 8.634640107272351e-05, "loss": 0.0124, "step": 2790 }, { "epoch": 14.285714285714286, "grad_norm": 0.2800692319869995, "learning_rate": 8.623265650935234e-05, "loss": 0.0131, "step": 2800 }, { "epoch": 14.33673469387755, "grad_norm": 0.2061230093240738, "learning_rate": 8.611851571145456e-05, "loss": 0.0147, "step": 2810 }, { "epoch": 14.387755102040817, "grad_norm": 0.17345425486564636, "learning_rate": 8.600397992725566e-05, "loss": 0.0145, "step": 2820 }, { "epoch": 14.438775510204081, "grad_norm": 0.3071790039539337, "learning_rate": 8.588905040930061e-05, "loss": 0.0116, "step": 2830 }, { "epoch": 14.489795918367347, "grad_norm": 0.27997928857803345, "learning_rate": 8.577372841444022e-05, "loss": 0.012, "step": 2840 }, { "epoch": 14.540816326530612, "grad_norm": 0.18748684227466583, "learning_rate": 8.565801520381736e-05, "loss": 0.0132, "step": 2850 }, { "epoch": 14.591836734693878, "grad_norm": 0.3445316255092621, "learning_rate": 8.554191204285313e-05, "loss": 0.0167, "step": 2860 }, { "epoch": 14.642857142857142, "grad_norm": 0.38402268290519714, "learning_rate": 8.542542020123315e-05, "loss": 0.0156, "step": 2870 }, { "epoch": 14.693877551020408, "grad_norm": 0.23096321523189545, "learning_rate": 8.530854095289347e-05, "loss": 0.0133, "step": 2880 }, { "epoch": 14.744897959183673, "grad_norm": 0.3270000219345093, "learning_rate": 8.519127557600688e-05, "loss": 0.0143, "step": 2890 }, { "epoch": 14.795918367346939, "grad_norm": 0.32618221640586853, "learning_rate": 8.507362535296871e-05, "loss": 0.0131, "step": 2900 }, { "epoch": 14.846938775510203, "grad_norm": 0.26457804441452026, "learning_rate": 8.495559157038299e-05, "loss": 0.0129, "step": 2910 }, { "epoch": 14.89795918367347, "grad_norm": 0.2867559492588043, "learning_rate": 8.483717551904823e-05, "loss": 0.0152, "step": 2920 }, { "epoch": 14.948979591836736, "grad_norm": 0.3372673988342285, "learning_rate": 8.47183784939434e-05, "loss": 0.016, "step": 2930 }, { "epoch": 15.0, "grad_norm": 0.2412089854478836, "learning_rate": 8.459920179421374e-05, "loss": 0.0153, "step": 2940 }, { "epoch": 15.051020408163266, "grad_norm": 0.20294897258281708, "learning_rate": 8.447964672315656e-05, "loss": 0.0128, "step": 2950 }, { "epoch": 15.10204081632653, "grad_norm": 0.31983137130737305, "learning_rate": 8.435971458820692e-05, "loss": 0.0145, "step": 2960 }, { "epoch": 15.153061224489797, "grad_norm": 0.2947094440460205, "learning_rate": 8.423940670092345e-05, "loss": 0.0129, "step": 2970 }, { "epoch": 15.204081632653061, "grad_norm": 0.24315378069877625, "learning_rate": 8.411872437697394e-05, "loss": 0.0128, "step": 2980 }, { "epoch": 15.255102040816327, "grad_norm": 0.2993694245815277, "learning_rate": 8.399766893612096e-05, "loss": 0.0127, "step": 2990 }, { "epoch": 15.306122448979592, "grad_norm": 0.27624186873435974, "learning_rate": 8.38762417022074e-05, "loss": 0.0123, "step": 3000 }, { "epoch": 15.357142857142858, "grad_norm": 0.2961912453174591, "learning_rate": 8.375444400314204e-05, "loss": 0.0121, "step": 3010 }, { "epoch": 15.408163265306122, "grad_norm": 0.14205600321292877, "learning_rate": 8.3632277170885e-05, "loss": 0.0126, "step": 3020 }, { "epoch": 15.459183673469388, "grad_norm": 0.15576381981372833, "learning_rate": 8.350974254143318e-05, "loss": 0.0106, "step": 3030 }, { "epoch": 15.510204081632653, "grad_norm": 0.1879970133304596, "learning_rate": 8.338684145480566e-05, "loss": 0.0128, "step": 3040 }, { "epoch": 15.561224489795919, "grad_norm": 0.18633995950222015, "learning_rate": 8.326357525502904e-05, "loss": 0.0118, "step": 3050 }, { "epoch": 15.612244897959183, "grad_norm": 0.4358859360218048, "learning_rate": 8.313994529012273e-05, "loss": 0.0121, "step": 3060 }, { "epoch": 15.66326530612245, "grad_norm": 0.2667674720287323, "learning_rate": 8.301595291208422e-05, "loss": 0.0122, "step": 3070 }, { "epoch": 15.714285714285714, "grad_norm": 0.19431917369365692, "learning_rate": 8.289159947687427e-05, "loss": 0.0129, "step": 3080 }, { "epoch": 15.76530612244898, "grad_norm": 0.184364452958107, "learning_rate": 8.276688634440216e-05, "loss": 0.0135, "step": 3090 }, { "epoch": 15.816326530612244, "grad_norm": 0.16020889580249786, "learning_rate": 8.26418148785107e-05, "loss": 0.0133, "step": 3100 }, { "epoch": 15.86734693877551, "grad_norm": 0.25338858366012573, "learning_rate": 8.251638644696141e-05, "loss": 0.0122, "step": 3110 }, { "epoch": 15.918367346938776, "grad_norm": 0.1335408091545105, "learning_rate": 8.23906024214195e-05, "loss": 0.0114, "step": 3120 }, { "epoch": 15.96938775510204, "grad_norm": 0.33032405376434326, "learning_rate": 8.226446417743897e-05, "loss": 0.0145, "step": 3130 }, { "epoch": 16.020408163265305, "grad_norm": 0.20884694159030914, "learning_rate": 8.213797309444742e-05, "loss": 0.0123, "step": 3140 }, { "epoch": 16.071428571428573, "grad_norm": 0.13955514132976532, "learning_rate": 8.201113055573105e-05, "loss": 0.0125, "step": 3150 }, { "epoch": 16.122448979591837, "grad_norm": 0.15403415262699127, "learning_rate": 8.188393794841958e-05, "loss": 0.0117, "step": 3160 }, { "epoch": 16.1734693877551, "grad_norm": 0.2488146722316742, "learning_rate": 8.175639666347094e-05, "loss": 0.0112, "step": 3170 }, { "epoch": 16.224489795918366, "grad_norm": 0.15036450326442719, "learning_rate": 8.162850809565623e-05, "loss": 0.0122, "step": 3180 }, { "epoch": 16.275510204081634, "grad_norm": 0.23091945052146912, "learning_rate": 8.150027364354431e-05, "loss": 0.0126, "step": 3190 }, { "epoch": 16.3265306122449, "grad_norm": 0.2231752574443817, "learning_rate": 8.137169470948662e-05, "loss": 0.0131, "step": 3200 }, { "epoch": 16.377551020408163, "grad_norm": 0.26042595505714417, "learning_rate": 8.124277269960179e-05, "loss": 0.0126, "step": 3210 }, { "epoch": 16.428571428571427, "grad_norm": 0.24597467482089996, "learning_rate": 8.111350902376023e-05, "loss": 0.0109, "step": 3220 }, { "epoch": 16.479591836734695, "grad_norm": 0.24575425684452057, "learning_rate": 8.098390509556883e-05, "loss": 0.0103, "step": 3230 }, { "epoch": 16.53061224489796, "grad_norm": 0.17975269258022308, "learning_rate": 8.085396233235536e-05, "loss": 0.0124, "step": 3240 }, { "epoch": 16.581632653061224, "grad_norm": 0.20594057440757751, "learning_rate": 8.072368215515306e-05, "loss": 0.0113, "step": 3250 }, { "epoch": 16.632653061224488, "grad_norm": 0.2970283329486847, "learning_rate": 8.059306598868506e-05, "loss": 0.014, "step": 3260 }, { "epoch": 16.683673469387756, "grad_norm": 0.14972592890262604, "learning_rate": 8.046211526134888e-05, "loss": 0.0121, "step": 3270 }, { "epoch": 16.73469387755102, "grad_norm": 0.12837564945220947, "learning_rate": 8.033083140520065e-05, "loss": 0.0125, "step": 3280 }, { "epoch": 16.785714285714285, "grad_norm": 0.21782033145427704, "learning_rate": 8.019921585593962e-05, "loss": 0.016, "step": 3290 }, { "epoch": 16.836734693877553, "grad_norm": 0.2740643620491028, "learning_rate": 8.006727005289232e-05, "loss": 0.0114, "step": 3300 }, { "epoch": 16.887755102040817, "grad_norm": 0.34749147295951843, "learning_rate": 7.993499543899692e-05, "loss": 0.0131, "step": 3310 }, { "epoch": 16.93877551020408, "grad_norm": 0.2707062363624573, "learning_rate": 7.980239346078742e-05, "loss": 0.0138, "step": 3320 }, { "epoch": 16.989795918367346, "grad_norm": 0.15632031857967377, "learning_rate": 7.966946556837778e-05, "loss": 0.011, "step": 3330 }, { "epoch": 17.040816326530614, "grad_norm": 0.2848604917526245, "learning_rate": 7.953621321544616e-05, "loss": 0.0122, "step": 3340 }, { "epoch": 17.091836734693878, "grad_norm": 0.25311529636383057, "learning_rate": 7.940263785921896e-05, "loss": 0.0129, "step": 3350 }, { "epoch": 17.142857142857142, "grad_norm": 0.22705887258052826, "learning_rate": 7.926874096045482e-05, "loss": 0.0123, "step": 3360 }, { "epoch": 17.193877551020407, "grad_norm": 0.1871379315853119, "learning_rate": 7.913452398342881e-05, "loss": 0.0132, "step": 3370 }, { "epoch": 17.244897959183675, "grad_norm": 0.26447540521621704, "learning_rate": 7.89999883959163e-05, "loss": 0.0114, "step": 3380 }, { "epoch": 17.29591836734694, "grad_norm": 0.20584167540073395, "learning_rate": 7.886513566917687e-05, "loss": 0.0105, "step": 3390 }, { "epoch": 17.346938775510203, "grad_norm": 0.160394549369812, "learning_rate": 7.872996727793838e-05, "loss": 0.0137, "step": 3400 }, { "epoch": 17.397959183673468, "grad_norm": 0.21800066530704498, "learning_rate": 7.859448470038069e-05, "loss": 0.0103, "step": 3410 }, { "epoch": 17.448979591836736, "grad_norm": 0.1987452656030655, "learning_rate": 7.845868941811956e-05, "loss": 0.0106, "step": 3420 }, { "epoch": 17.5, "grad_norm": 0.35021811723709106, "learning_rate": 7.832258291619043e-05, "loss": 0.0122, "step": 3430 }, { "epoch": 17.551020408163264, "grad_norm": 0.23994189500808716, "learning_rate": 7.81861666830322e-05, "loss": 0.0136, "step": 3440 }, { "epoch": 17.602040816326532, "grad_norm": 0.4185681641101837, "learning_rate": 7.804944221047097e-05, "loss": 0.0126, "step": 3450 }, { "epoch": 17.653061224489797, "grad_norm": 0.2237282395362854, "learning_rate": 7.791241099370364e-05, "loss": 0.0158, "step": 3460 }, { "epoch": 17.70408163265306, "grad_norm": 0.2274068146944046, "learning_rate": 7.777507453128163e-05, "loss": 0.0105, "step": 3470 }, { "epoch": 17.755102040816325, "grad_norm": 0.2978285849094391, "learning_rate": 7.763743432509451e-05, "loss": 0.0131, "step": 3480 }, { "epoch": 17.806122448979593, "grad_norm": 0.18854284286499023, "learning_rate": 7.749949188035353e-05, "loss": 0.0128, "step": 3490 }, { "epoch": 17.857142857142858, "grad_norm": 0.2800551950931549, "learning_rate": 7.736124870557516e-05, "loss": 0.012, "step": 3500 }, { "epoch": 17.908163265306122, "grad_norm": 0.15255339443683624, "learning_rate": 7.722270631256459e-05, "loss": 0.0101, "step": 3510 }, { "epoch": 17.959183673469386, "grad_norm": 0.32308852672576904, "learning_rate": 7.708386621639925e-05, "loss": 0.013, "step": 3520 }, { "epoch": 18.010204081632654, "grad_norm": 0.19822238385677338, "learning_rate": 7.694472993541219e-05, "loss": 0.0143, "step": 3530 }, { "epoch": 18.06122448979592, "grad_norm": 0.13634613156318665, "learning_rate": 7.680529899117547e-05, "loss": 0.012, "step": 3540 }, { "epoch": 18.112244897959183, "grad_norm": 0.2609889805316925, "learning_rate": 7.666557490848358e-05, "loss": 0.012, "step": 3550 }, { "epoch": 18.163265306122447, "grad_norm": 0.18496136367321014, "learning_rate": 7.65255592153367e-05, "loss": 0.0118, "step": 3560 }, { "epoch": 18.214285714285715, "grad_norm": 0.18766474723815918, "learning_rate": 7.638525344292402e-05, "loss": 0.012, "step": 3570 }, { "epoch": 18.26530612244898, "grad_norm": 0.21336659789085388, "learning_rate": 7.624465912560697e-05, "loss": 0.0126, "step": 3580 }, { "epoch": 18.316326530612244, "grad_norm": 0.2043648660182953, "learning_rate": 7.610377780090249e-05, "loss": 0.0113, "step": 3590 }, { "epoch": 18.367346938775512, "grad_norm": 0.20681460201740265, "learning_rate": 7.596261100946618e-05, "loss": 0.0122, "step": 3600 }, { "epoch": 18.418367346938776, "grad_norm": 0.19744859635829926, "learning_rate": 7.582116029507542e-05, "loss": 0.0109, "step": 3610 }, { "epoch": 18.46938775510204, "grad_norm": 0.16987404227256775, "learning_rate": 7.56794272046126e-05, "loss": 0.0102, "step": 3620 }, { "epoch": 18.520408163265305, "grad_norm": 0.14916828274726868, "learning_rate": 7.55374132880481e-05, "loss": 0.0109, "step": 3630 }, { "epoch": 18.571428571428573, "grad_norm": 0.20487569272518158, "learning_rate": 7.539512009842333e-05, "loss": 0.0115, "step": 3640 }, { "epoch": 18.622448979591837, "grad_norm": 0.24674925208091736, "learning_rate": 7.525254919183382e-05, "loss": 0.0119, "step": 3650 }, { "epoch": 18.6734693877551, "grad_norm": 0.1887817084789276, "learning_rate": 7.510970212741215e-05, "loss": 0.0131, "step": 3660 }, { "epoch": 18.724489795918366, "grad_norm": 0.18235325813293457, "learning_rate": 7.496658046731096e-05, "loss": 0.0113, "step": 3670 }, { "epoch": 18.775510204081634, "grad_norm": 0.12009734660387039, "learning_rate": 7.482318577668578e-05, "loss": 0.0107, "step": 3680 }, { "epoch": 18.8265306122449, "grad_norm": 0.1425887495279312, "learning_rate": 7.467951962367796e-05, "loss": 0.0112, "step": 3690 }, { "epoch": 18.877551020408163, "grad_norm": 0.14049017429351807, "learning_rate": 7.453558357939755e-05, "loss": 0.0126, "step": 3700 }, { "epoch": 18.928571428571427, "grad_norm": 0.33091381192207336, "learning_rate": 7.439137921790606e-05, "loss": 0.0111, "step": 3710 }, { "epoch": 18.979591836734695, "grad_norm": 0.19092345237731934, "learning_rate": 7.42469081161993e-05, "loss": 0.0128, "step": 3720 }, { "epoch": 19.03061224489796, "grad_norm": 0.20333991944789886, "learning_rate": 7.410217185419006e-05, "loss": 0.0119, "step": 3730 }, { "epoch": 19.081632653061224, "grad_norm": 0.1206803172826767, "learning_rate": 7.395717201469095e-05, "loss": 0.0118, "step": 3740 }, { "epoch": 19.132653061224488, "grad_norm": 0.1619206964969635, "learning_rate": 7.381191018339696e-05, "loss": 0.0112, "step": 3750 }, { "epoch": 19.183673469387756, "grad_norm": 0.24156610667705536, "learning_rate": 7.36663879488682e-05, "loss": 0.0114, "step": 3760 }, { "epoch": 19.23469387755102, "grad_norm": 0.23019012808799744, "learning_rate": 7.352060690251254e-05, "loss": 0.0112, "step": 3770 }, { "epoch": 19.285714285714285, "grad_norm": 0.22867348790168762, "learning_rate": 7.337456863856811e-05, "loss": 0.0103, "step": 3780 }, { "epoch": 19.336734693877553, "grad_norm": 0.22941316664218903, "learning_rate": 7.3228274754086e-05, "loss": 0.0124, "step": 3790 }, { "epoch": 19.387755102040817, "grad_norm": 0.22303640842437744, "learning_rate": 7.308172684891267e-05, "loss": 0.012, "step": 3800 }, { "epoch": 19.43877551020408, "grad_norm": 0.20708808302879333, "learning_rate": 7.293492652567255e-05, "loss": 0.01, "step": 3810 }, { "epoch": 19.489795918367346, "grad_norm": 0.2207169532775879, "learning_rate": 7.278787538975043e-05, "loss": 0.011, "step": 3820 }, { "epoch": 19.540816326530614, "grad_norm": 0.15825074911117554, "learning_rate": 7.2640575049274e-05, "loss": 0.0114, "step": 3830 }, { "epoch": 19.591836734693878, "grad_norm": 0.12295360118150711, "learning_rate": 7.249302711509616e-05, "loss": 0.0105, "step": 3840 }, { "epoch": 19.642857142857142, "grad_norm": 0.18383798003196716, "learning_rate": 7.23452332007775e-05, "loss": 0.0104, "step": 3850 }, { "epoch": 19.693877551020407, "grad_norm": 0.14068524539470673, "learning_rate": 7.219719492256858e-05, "loss": 0.0121, "step": 3860 }, { "epoch": 19.744897959183675, "grad_norm": 0.22027108073234558, "learning_rate": 7.20489138993923e-05, "loss": 0.0116, "step": 3870 }, { "epoch": 19.79591836734694, "grad_norm": 0.23714610934257507, "learning_rate": 7.190039175282614e-05, "loss": 0.0103, "step": 3880 }, { "epoch": 19.846938775510203, "grad_norm": 0.27144551277160645, "learning_rate": 7.175163010708455e-05, "loss": 0.0093, "step": 3890 }, { "epoch": 19.897959183673468, "grad_norm": 0.11050327867269516, "learning_rate": 7.1602630589001e-05, "loss": 0.0108, "step": 3900 }, { "epoch": 19.948979591836736, "grad_norm": 0.2859252095222473, "learning_rate": 7.14533948280104e-05, "loss": 0.0127, "step": 3910 }, { "epoch": 20.0, "grad_norm": 0.21518123149871826, "learning_rate": 7.130392445613109e-05, "loss": 0.0139, "step": 3920 }, { "epoch": 20.051020408163264, "grad_norm": 0.16452106833457947, "learning_rate": 7.115422110794711e-05, "loss": 0.0097, "step": 3930 }, { "epoch": 20.102040816326532, "grad_norm": 0.2540481388568878, "learning_rate": 7.100428642059033e-05, "loss": 0.0113, "step": 3940 }, { "epoch": 20.153061224489797, "grad_norm": 0.16517497599124908, "learning_rate": 7.08541220337224e-05, "loss": 0.01, "step": 3950 }, { "epoch": 20.20408163265306, "grad_norm": 0.17894187569618225, "learning_rate": 7.070372958951706e-05, "loss": 0.0124, "step": 3960 }, { "epoch": 20.255102040816325, "grad_norm": 0.16708102822303772, "learning_rate": 7.055311073264194e-05, "loss": 0.0115, "step": 3970 }, { "epoch": 20.306122448979593, "grad_norm": 0.25460880994796753, "learning_rate": 7.040226711024077e-05, "loss": 0.0163, "step": 3980 }, { "epoch": 20.357142857142858, "grad_norm": 0.1931772083044052, "learning_rate": 7.02512003719152e-05, "loss": 0.0151, "step": 3990 }, { "epoch": 20.408163265306122, "grad_norm": 0.21869954466819763, "learning_rate": 7.00999121697069e-05, "loss": 0.0115, "step": 4000 }, { "epoch": 20.459183673469386, "grad_norm": 0.15397238731384277, "learning_rate": 6.99484041580794e-05, "loss": 0.012, "step": 4010 }, { "epoch": 20.510204081632654, "grad_norm": 0.2852042317390442, "learning_rate": 6.979667799390004e-05, "loss": 0.0105, "step": 4020 }, { "epoch": 20.56122448979592, "grad_norm": 0.1933477222919464, "learning_rate": 6.964473533642185e-05, "loss": 0.0102, "step": 4030 }, { "epoch": 20.612244897959183, "grad_norm": 0.16715887188911438, "learning_rate": 6.949257784726539e-05, "loss": 0.0104, "step": 4040 }, { "epoch": 20.663265306122447, "grad_norm": 0.2258576601743698, "learning_rate": 6.934020719040056e-05, "loss": 0.0093, "step": 4050 }, { "epoch": 20.714285714285715, "grad_norm": 0.27509212493896484, "learning_rate": 6.918762503212848e-05, "loss": 0.011, "step": 4060 }, { "epoch": 20.76530612244898, "grad_norm": 0.2225647121667862, "learning_rate": 6.903483304106319e-05, "loss": 0.0103, "step": 4070 }, { "epoch": 20.816326530612244, "grad_norm": 0.26670223474502563, "learning_rate": 6.888183288811341e-05, "loss": 0.0112, "step": 4080 }, { "epoch": 20.867346938775512, "grad_norm": 0.2505559027194977, "learning_rate": 6.87286262464643e-05, "loss": 0.0119, "step": 4090 }, { "epoch": 20.918367346938776, "grad_norm": 0.2762985825538635, "learning_rate": 6.857521479155915e-05, "loss": 0.0109, "step": 4100 }, { "epoch": 20.96938775510204, "grad_norm": 0.24262341856956482, "learning_rate": 6.842160020108104e-05, "loss": 0.0116, "step": 4110 }, { "epoch": 21.020408163265305, "grad_norm": 0.2080129235982895, "learning_rate": 6.826778415493455e-05, "loss": 0.0101, "step": 4120 }, { "epoch": 21.071428571428573, "grad_norm": 0.17384271323680878, "learning_rate": 6.811376833522729e-05, "loss": 0.0093, "step": 4130 }, { "epoch": 21.122448979591837, "grad_norm": 0.23997440934181213, "learning_rate": 6.795955442625159e-05, "loss": 0.0127, "step": 4140 }, { "epoch": 21.1734693877551, "grad_norm": 0.23796118795871735, "learning_rate": 6.780514411446608e-05, "loss": 0.0104, "step": 4150 }, { "epoch": 21.224489795918366, "grad_norm": 0.17038597166538239, "learning_rate": 6.765053908847716e-05, "loss": 0.0123, "step": 4160 }, { "epoch": 21.275510204081634, "grad_norm": 0.143656387925148, "learning_rate": 6.749574103902064e-05, "loss": 0.0099, "step": 4170 }, { "epoch": 21.3265306122449, "grad_norm": 0.12225384265184402, "learning_rate": 6.734075165894317e-05, "loss": 0.0105, "step": 4180 }, { "epoch": 21.377551020408163, "grad_norm": 0.18611663579940796, "learning_rate": 6.71855726431838e-05, "loss": 0.0094, "step": 4190 }, { "epoch": 21.428571428571427, "grad_norm": 0.2547371983528137, "learning_rate": 6.703020568875538e-05, "loss": 0.0095, "step": 4200 }, { "epoch": 21.479591836734695, "grad_norm": 0.1599801778793335, "learning_rate": 6.687465249472603e-05, "loss": 0.0097, "step": 4210 }, { "epoch": 21.53061224489796, "grad_norm": 0.1612466424703598, "learning_rate": 6.671891476220055e-05, "loss": 0.0095, "step": 4220 }, { "epoch": 21.581632653061224, "grad_norm": 0.21456393599510193, "learning_rate": 6.656299419430183e-05, "loss": 0.0094, "step": 4230 }, { "epoch": 21.632653061224488, "grad_norm": 0.15586300194263458, "learning_rate": 6.640689249615223e-05, "loss": 0.0122, "step": 4240 }, { "epoch": 21.683673469387756, "grad_norm": 0.18530231714248657, "learning_rate": 6.625061137485491e-05, "loss": 0.0098, "step": 4250 }, { "epoch": 21.73469387755102, "grad_norm": 0.23475265502929688, "learning_rate": 6.609415253947517e-05, "loss": 0.0105, "step": 4260 }, { "epoch": 21.785714285714285, "grad_norm": 0.266766756772995, "learning_rate": 6.593751770102178e-05, "loss": 0.0094, "step": 4270 }, { "epoch": 21.836734693877553, "grad_norm": 0.3379665017127991, "learning_rate": 6.578070857242823e-05, "loss": 0.0102, "step": 4280 }, { "epoch": 21.887755102040817, "grad_norm": 0.1976608783006668, "learning_rate": 6.562372686853402e-05, "loss": 0.0107, "step": 4290 }, { "epoch": 21.93877551020408, "grad_norm": 0.2913855016231537, "learning_rate": 6.546657430606593e-05, "loss": 0.0103, "step": 4300 }, { "epoch": 21.989795918367346, "grad_norm": 0.2175789773464203, "learning_rate": 6.530925260361918e-05, "loss": 0.0119, "step": 4310 }, { "epoch": 22.040816326530614, "grad_norm": 0.24629560112953186, "learning_rate": 6.515176348163871e-05, "loss": 0.0097, "step": 4320 }, { "epoch": 22.091836734693878, "grad_norm": 0.17452183365821838, "learning_rate": 6.499410866240032e-05, "loss": 0.0115, "step": 4330 }, { "epoch": 22.142857142857142, "grad_norm": 0.2677079439163208, "learning_rate": 6.48362898699919e-05, "loss": 0.0118, "step": 4340 }, { "epoch": 22.193877551020407, "grad_norm": 0.20298388600349426, "learning_rate": 6.467830883029443e-05, "loss": 0.0097, "step": 4350 }, { "epoch": 22.244897959183675, "grad_norm": 0.18491944670677185, "learning_rate": 6.452016727096326e-05, "loss": 0.0094, "step": 4360 }, { "epoch": 22.29591836734694, "grad_norm": 0.29317691922187805, "learning_rate": 6.436186692140916e-05, "loss": 0.0104, "step": 4370 }, { "epoch": 22.346938775510203, "grad_norm": 0.20735745131969452, "learning_rate": 6.420340951277938e-05, "loss": 0.0112, "step": 4380 }, { "epoch": 22.397959183673468, "grad_norm": 0.22862356901168823, "learning_rate": 6.404479677793874e-05, "loss": 0.0128, "step": 4390 }, { "epoch": 22.448979591836736, "grad_norm": 0.26520586013793945, "learning_rate": 6.388603045145075e-05, "loss": 0.0127, "step": 4400 }, { "epoch": 22.5, "grad_norm": 0.21048423647880554, "learning_rate": 6.372711226955843e-05, "loss": 0.0129, "step": 4410 }, { "epoch": 22.551020408163264, "grad_norm": 0.22128242254257202, "learning_rate": 6.356804397016564e-05, "loss": 0.0098, "step": 4420 }, { "epoch": 22.602040816326532, "grad_norm": 0.21060210466384888, "learning_rate": 6.340882729281779e-05, "loss": 0.0112, "step": 4430 }, { "epoch": 22.653061224489797, "grad_norm": 0.388913631439209, "learning_rate": 6.324946397868294e-05, "loss": 0.0104, "step": 4440 }, { "epoch": 22.70408163265306, "grad_norm": 0.21065767109394073, "learning_rate": 6.308995577053276e-05, "loss": 0.0112, "step": 4450 }, { "epoch": 22.755102040816325, "grad_norm": 0.1968468427658081, "learning_rate": 6.293030441272347e-05, "loss": 0.0093, "step": 4460 }, { "epoch": 22.806122448979593, "grad_norm": 0.13169094920158386, "learning_rate": 6.277051165117677e-05, "loss": 0.0086, "step": 4470 }, { "epoch": 22.857142857142858, "grad_norm": 0.16332277655601501, "learning_rate": 6.261057923336064e-05, "loss": 0.0107, "step": 4480 }, { "epoch": 22.908163265306122, "grad_norm": 0.14051508903503418, "learning_rate": 6.245050890827042e-05, "loss": 0.0098, "step": 4490 }, { "epoch": 22.959183673469386, "grad_norm": 0.1796175092458725, "learning_rate": 6.229030242640952e-05, "loss": 0.0103, "step": 4500 }, { "epoch": 23.010204081632654, "grad_norm": 0.197859987616539, "learning_rate": 6.212996153977037e-05, "loss": 0.0088, "step": 4510 }, { "epoch": 23.06122448979592, "grad_norm": 0.21856550872325897, "learning_rate": 6.196948800181523e-05, "loss": 0.0092, "step": 4520 }, { "epoch": 23.112244897959183, "grad_norm": 0.18091683089733124, "learning_rate": 6.180888356745695e-05, "loss": 0.0105, "step": 4530 }, { "epoch": 23.163265306122447, "grad_norm": 0.21952562034130096, "learning_rate": 6.164814999303995e-05, "loss": 0.0103, "step": 4540 }, { "epoch": 23.214285714285715, "grad_norm": 0.31855741143226624, "learning_rate": 6.148728903632081e-05, "loss": 0.0096, "step": 4550 }, { "epoch": 23.26530612244898, "grad_norm": 0.1732335090637207, "learning_rate": 6.132630245644921e-05, "loss": 0.0103, "step": 4560 }, { "epoch": 23.316326530612244, "grad_norm": 0.3058975636959076, "learning_rate": 6.116519201394857e-05, "loss": 0.0089, "step": 4570 }, { "epoch": 23.367346938775512, "grad_norm": 0.17514942586421967, "learning_rate": 6.10039594706969e-05, "loss": 0.0107, "step": 4580 }, { "epoch": 23.418367346938776, "grad_norm": 0.16879461705684662, "learning_rate": 6.084260658990744e-05, "loss": 0.0115, "step": 4590 }, { "epoch": 23.46938775510204, "grad_norm": 0.1473907083272934, "learning_rate": 6.068113513610943e-05, "loss": 0.0097, "step": 4600 }, { "epoch": 23.520408163265305, "grad_norm": 0.16791348159313202, "learning_rate": 6.0519546875128876e-05, "loss": 0.0114, "step": 4610 }, { "epoch": 23.571428571428573, "grad_norm": 0.18700872361660004, "learning_rate": 6.035784357406906e-05, "loss": 0.0094, "step": 4620 }, { "epoch": 23.622448979591837, "grad_norm": 0.1533333659172058, "learning_rate": 6.01960270012914e-05, "loss": 0.011, "step": 4630 }, { "epoch": 23.6734693877551, "grad_norm": 0.1658274084329605, "learning_rate": 6.003409892639599e-05, "loss": 0.0096, "step": 4640 }, { "epoch": 23.724489795918366, "grad_norm": 0.24512982368469238, "learning_rate": 5.9872061120202336e-05, "loss": 0.0105, "step": 4650 }, { "epoch": 23.775510204081634, "grad_norm": 0.17851103842258453, "learning_rate": 5.9709915354729914e-05, "loss": 0.0101, "step": 4660 }, { "epoch": 23.8265306122449, "grad_norm": 0.19424092769622803, "learning_rate": 5.9547663403178824e-05, "loss": 0.0107, "step": 4670 }, { "epoch": 23.877551020408163, "grad_norm": 0.12847241759300232, "learning_rate": 5.9385307039910445e-05, "loss": 0.0104, "step": 4680 }, { "epoch": 23.928571428571427, "grad_norm": 0.20569926500320435, "learning_rate": 5.922284804042792e-05, "loss": 0.0123, "step": 4690 }, { "epoch": 23.979591836734695, "grad_norm": 0.18088386952877045, "learning_rate": 5.906028818135687e-05, "loss": 0.0095, "step": 4700 }, { "epoch": 24.03061224489796, "grad_norm": 0.13830965757369995, "learning_rate": 5.889762924042585e-05, "loss": 0.0095, "step": 4710 }, { "epoch": 24.081632653061224, "grad_norm": 0.13896656036376953, "learning_rate": 5.873487299644699e-05, "loss": 0.0114, "step": 4720 }, { "epoch": 24.132653061224488, "grad_norm": 0.21596750617027283, "learning_rate": 5.857202122929649e-05, "loss": 0.01, "step": 4730 }, { "epoch": 24.183673469387756, "grad_norm": 0.11036713421344757, "learning_rate": 5.840907571989518e-05, "loss": 0.009, "step": 4740 }, { "epoch": 24.23469387755102, "grad_norm": 0.18683020770549774, "learning_rate": 5.824603825018904e-05, "loss": 0.0112, "step": 4750 }, { "epoch": 24.285714285714285, "grad_norm": 0.2586684823036194, "learning_rate": 5.808291060312975e-05, "loss": 0.0109, "step": 4760 }, { "epoch": 24.336734693877553, "grad_norm": 0.25213420391082764, "learning_rate": 5.7919694562655083e-05, "loss": 0.0116, "step": 4770 }, { "epoch": 24.387755102040817, "grad_norm": 0.23867247998714447, "learning_rate": 5.775639191366954e-05, "loss": 0.011, "step": 4780 }, { "epoch": 24.43877551020408, "grad_norm": 0.24193839728832245, "learning_rate": 5.75930044420247e-05, "loss": 0.0113, "step": 4790 }, { "epoch": 24.489795918367346, "grad_norm": 0.20850181579589844, "learning_rate": 5.74295339344998e-05, "loss": 0.0097, "step": 4800 }, { "epoch": 24.540816326530614, "grad_norm": 0.22214262187480927, "learning_rate": 5.726598217878211e-05, "loss": 0.0116, "step": 4810 }, { "epoch": 24.591836734693878, "grad_norm": 0.3797580897808075, "learning_rate": 5.71023509634474e-05, "loss": 0.0103, "step": 4820 }, { "epoch": 24.642857142857142, "grad_norm": 0.1686123162508011, "learning_rate": 5.693864207794049e-05, "loss": 0.011, "step": 4830 }, { "epoch": 24.693877551020407, "grad_norm": 0.26540806889533997, "learning_rate": 5.677485731255545e-05, "loss": 0.0108, "step": 4840 }, { "epoch": 24.744897959183675, "grad_norm": 0.18972492218017578, "learning_rate": 5.6610998458416296e-05, "loss": 0.0091, "step": 4850 }, { "epoch": 24.79591836734694, "grad_norm": 0.13661795854568481, "learning_rate": 5.644706730745716e-05, "loss": 0.0104, "step": 4860 }, { "epoch": 24.846938775510203, "grad_norm": 0.17988818883895874, "learning_rate": 5.628306565240287e-05, "loss": 0.0103, "step": 4870 }, { "epoch": 24.897959183673468, "grad_norm": 0.24251392483711243, "learning_rate": 5.611899528674923e-05, "loss": 0.0096, "step": 4880 }, { "epoch": 24.948979591836736, "grad_norm": 0.32364219427108765, "learning_rate": 5.595485800474349e-05, "loss": 0.0094, "step": 4890 }, { "epoch": 25.0, "grad_norm": 0.29088330268859863, "learning_rate": 5.579065560136467e-05, "loss": 0.0103, "step": 4900 }, { "epoch": 25.051020408163264, "grad_norm": 0.28665196895599365, "learning_rate": 5.562638987230392e-05, "loss": 0.0095, "step": 4910 }, { "epoch": 25.102040816326532, "grad_norm": 0.14775583148002625, "learning_rate": 5.546206261394498e-05, "loss": 0.0127, "step": 4920 }, { "epoch": 25.153061224489797, "grad_norm": 0.22124029695987701, "learning_rate": 5.529767562334437e-05, "loss": 0.012, "step": 4930 }, { "epoch": 25.20408163265306, "grad_norm": 0.22517479956150055, "learning_rate": 5.5133230698211926e-05, "loss": 0.0097, "step": 4940 }, { "epoch": 25.255102040816325, "grad_norm": 0.24514521658420563, "learning_rate": 5.496872963689096e-05, "loss": 0.0113, "step": 4950 }, { "epoch": 25.306122448979593, "grad_norm": 0.27261433005332947, "learning_rate": 5.4804174238338756e-05, "loss": 0.0113, "step": 4960 }, { "epoch": 25.357142857142858, "grad_norm": 0.161941796541214, "learning_rate": 5.463956630210678e-05, "loss": 0.0094, "step": 4970 }, { "epoch": 25.408163265306122, "grad_norm": 0.1453172117471695, "learning_rate": 5.4474907628321046e-05, "loss": 0.01, "step": 4980 }, { "epoch": 25.459183673469386, "grad_norm": 0.15813565254211426, "learning_rate": 5.431020001766244e-05, "loss": 0.0092, "step": 4990 }, { "epoch": 25.510204081632654, "grad_norm": 0.1172117218375206, "learning_rate": 5.4145445271346986e-05, "loss": 0.0106, "step": 5000 }, { "epoch": 25.56122448979592, "grad_norm": 0.16811995208263397, "learning_rate": 5.398064519110622e-05, "loss": 0.0103, "step": 5010 }, { "epoch": 25.612244897959183, "grad_norm": 0.23049351572990417, "learning_rate": 5.3815801579167394e-05, "loss": 0.0104, "step": 5020 }, { "epoch": 25.663265306122447, "grad_norm": 0.21891969442367554, "learning_rate": 5.365091623823382e-05, "loss": 0.0093, "step": 5030 }, { "epoch": 25.714285714285715, "grad_norm": 0.2916574478149414, "learning_rate": 5.348599097146521e-05, "loss": 0.0099, "step": 5040 }, { "epoch": 25.76530612244898, "grad_norm": 0.14204153418540955, "learning_rate": 5.3321027582457836e-05, "loss": 0.0092, "step": 5050 }, { "epoch": 25.816326530612244, "grad_norm": 0.2285534143447876, "learning_rate": 5.315602787522491e-05, "loss": 0.0097, "step": 5060 }, { "epoch": 25.867346938775512, "grad_norm": 0.2207268923521042, "learning_rate": 5.299099365417678e-05, "loss": 0.0092, "step": 5070 }, { "epoch": 25.918367346938776, "grad_norm": 0.20298534631729126, "learning_rate": 5.2825926724101236e-05, "loss": 0.0082, "step": 5080 }, { "epoch": 25.96938775510204, "grad_norm": 0.20029844343662262, "learning_rate": 5.26608288901438e-05, "loss": 0.0095, "step": 5090 }, { "epoch": 26.020408163265305, "grad_norm": 0.15149113535881042, "learning_rate": 5.24957019577879e-05, "loss": 0.0087, "step": 5100 }, { "epoch": 26.071428571428573, "grad_norm": 0.1924610137939453, "learning_rate": 5.2330547732835266e-05, "loss": 0.009, "step": 5110 }, { "epoch": 26.122448979591837, "grad_norm": 0.2385406196117401, "learning_rate": 5.2165368021385996e-05, "loss": 0.0092, "step": 5120 }, { "epoch": 26.1734693877551, "grad_norm": 0.2380906194448471, "learning_rate": 5.200016462981897e-05, "loss": 0.0094, "step": 5130 }, { "epoch": 26.224489795918366, "grad_norm": 0.1740346997976303, "learning_rate": 5.1834939364772015e-05, "loss": 0.0106, "step": 5140 }, { "epoch": 26.275510204081634, "grad_norm": 0.1322096586227417, "learning_rate": 5.166969403312214e-05, "loss": 0.0094, "step": 5150 }, { "epoch": 26.3265306122449, "grad_norm": 0.15692533552646637, "learning_rate": 5.1504430441965844e-05, "loss": 0.0104, "step": 5160 }, { "epoch": 26.377551020408163, "grad_norm": 0.2436470240354538, "learning_rate": 5.133915039859923e-05, "loss": 0.0078, "step": 5170 }, { "epoch": 26.428571428571427, "grad_norm": 0.24672247469425201, "learning_rate": 5.1173855710498444e-05, "loss": 0.0076, "step": 5180 }, { "epoch": 26.479591836734695, "grad_norm": 0.2999960780143738, "learning_rate": 5.100854818529967e-05, "loss": 0.0097, "step": 5190 }, { "epoch": 26.53061224489796, "grad_norm": 0.2051677405834198, "learning_rate": 5.084322963077951e-05, "loss": 0.0094, "step": 5200 }, { "epoch": 26.581632653061224, "grad_norm": 0.31467360258102417, "learning_rate": 5.067790185483522e-05, "loss": 0.0089, "step": 5210 }, { "epoch": 26.632653061224488, "grad_norm": 0.2187862992286682, "learning_rate": 5.0512566665464844e-05, "loss": 0.0095, "step": 5220 }, { "epoch": 26.683673469387756, "grad_norm": 0.12864990532398224, "learning_rate": 5.034722587074755e-05, "loss": 0.0083, "step": 5230 }, { "epoch": 26.73469387755102, "grad_norm": 0.1356782764196396, "learning_rate": 5.018188127882375e-05, "loss": 0.0094, "step": 5240 }, { "epoch": 26.785714285714285, "grad_norm": 0.1458166539669037, "learning_rate": 5.0016534697875417e-05, "loss": 0.0086, "step": 5250 }, { "epoch": 26.836734693877553, "grad_norm": 0.14307551085948944, "learning_rate": 4.9851187936106294e-05, "loss": 0.0108, "step": 5260 }, { "epoch": 26.887755102040817, "grad_norm": 0.1394525170326233, "learning_rate": 4.968584280172206e-05, "loss": 0.0092, "step": 5270 }, { "epoch": 26.93877551020408, "grad_norm": 0.26238420605659485, "learning_rate": 4.95205011029106e-05, "loss": 0.0084, "step": 5280 }, { "epoch": 26.989795918367346, "grad_norm": 0.17156554758548737, "learning_rate": 4.935516464782227e-05, "loss": 0.0083, "step": 5290 }, { "epoch": 27.040816326530614, "grad_norm": 0.18224193155765533, "learning_rate": 4.918983524455003e-05, "loss": 0.0087, "step": 5300 }, { "epoch": 27.091836734693878, "grad_norm": 0.1758424937725067, "learning_rate": 4.9024514701109766e-05, "loss": 0.008, "step": 5310 }, { "epoch": 27.142857142857142, "grad_norm": 0.2172459214925766, "learning_rate": 4.885920482542043e-05, "loss": 0.0085, "step": 5320 }, { "epoch": 27.193877551020407, "grad_norm": 0.16209247708320618, "learning_rate": 4.869390742528438e-05, "loss": 0.0085, "step": 5330 }, { "epoch": 27.244897959183675, "grad_norm": 0.10191047191619873, "learning_rate": 4.852862430836744e-05, "loss": 0.0097, "step": 5340 }, { "epoch": 27.29591836734694, "grad_norm": 0.26966533064842224, "learning_rate": 4.836335728217933e-05, "loss": 0.0088, "step": 5350 }, { "epoch": 27.346938775510203, "grad_norm": 0.21062733232975006, "learning_rate": 4.819810815405379e-05, "loss": 0.0091, "step": 5360 }, { "epoch": 27.397959183673468, "grad_norm": 0.24097348749637604, "learning_rate": 4.803287873112877e-05, "loss": 0.0084, "step": 5370 }, { "epoch": 27.448979591836736, "grad_norm": 0.29177507758140564, "learning_rate": 4.786767082032681e-05, "loss": 0.0081, "step": 5380 }, { "epoch": 27.5, "grad_norm": 0.21912606060504913, "learning_rate": 4.77024862283351e-05, "loss": 0.0086, "step": 5390 }, { "epoch": 27.551020408163264, "grad_norm": 0.24101978540420532, "learning_rate": 4.753732676158593e-05, "loss": 0.0104, "step": 5400 }, { "epoch": 27.602040816326532, "grad_norm": 0.19426803290843964, "learning_rate": 4.737219422623672e-05, "loss": 0.0091, "step": 5410 }, { "epoch": 27.653061224489797, "grad_norm": 0.17346209287643433, "learning_rate": 4.720709042815044e-05, "loss": 0.0091, "step": 5420 }, { "epoch": 27.70408163265306, "grad_norm": 0.15763318538665771, "learning_rate": 4.704201717287578e-05, "loss": 0.0082, "step": 5430 }, { "epoch": 27.755102040816325, "grad_norm": 0.21148861944675446, "learning_rate": 4.6876976265627404e-05, "loss": 0.0075, "step": 5440 }, { "epoch": 27.806122448979593, "grad_norm": 0.24889998137950897, "learning_rate": 4.671196951126626e-05, "loss": 0.0076, "step": 5450 }, { "epoch": 27.857142857142858, "grad_norm": 0.230227991938591, "learning_rate": 4.654699871427971e-05, "loss": 0.0077, "step": 5460 }, { "epoch": 27.908163265306122, "grad_norm": 0.10753627121448517, "learning_rate": 4.6382065678762034e-05, "loss": 0.0087, "step": 5470 }, { "epoch": 27.959183673469386, "grad_norm": 0.16144831478595734, "learning_rate": 4.6217172208394424e-05, "loss": 0.0082, "step": 5480 }, { "epoch": 28.010204081632654, "grad_norm": 0.20871073007583618, "learning_rate": 4.605232010642549e-05, "loss": 0.0084, "step": 5490 }, { "epoch": 28.06122448979592, "grad_norm": 0.12219980359077454, "learning_rate": 4.588751117565142e-05, "loss": 0.0099, "step": 5500 }, { "epoch": 28.112244897959183, "grad_norm": 0.2575138807296753, "learning_rate": 4.5722747218396214e-05, "loss": 0.0082, "step": 5510 }, { "epoch": 28.163265306122447, "grad_norm": 0.14473095536231995, "learning_rate": 4.5558030036492194e-05, "loss": 0.0084, "step": 5520 }, { "epoch": 28.214285714285715, "grad_norm": 0.18454857170581818, "learning_rate": 4.539336143125999e-05, "loss": 0.0079, "step": 5530 }, { "epoch": 28.26530612244898, "grad_norm": 0.14465652406215668, "learning_rate": 4.522874320348916e-05, "loss": 0.0091, "step": 5540 }, { "epoch": 28.316326530612244, "grad_norm": 0.11447791755199432, "learning_rate": 4.506417715341821e-05, "loss": 0.0085, "step": 5550 }, { "epoch": 28.367346938775512, "grad_norm": 0.10874499380588531, "learning_rate": 4.489966508071511e-05, "loss": 0.0098, "step": 5560 }, { "epoch": 28.418367346938776, "grad_norm": 0.1765771210193634, "learning_rate": 4.4735208784457575e-05, "loss": 0.009, "step": 5570 }, { "epoch": 28.46938775510204, "grad_norm": 0.1845337599515915, "learning_rate": 4.457081006311325e-05, "loss": 0.0118, "step": 5580 }, { "epoch": 28.520408163265305, "grad_norm": 0.10803712159395218, "learning_rate": 4.440647071452027e-05, "loss": 0.0089, "step": 5590 }, { "epoch": 28.571428571428573, "grad_norm": 0.12594307959079742, "learning_rate": 4.424219253586737e-05, "loss": 0.0108, "step": 5600 }, { "epoch": 28.622448979591837, "grad_norm": 0.1718534380197525, "learning_rate": 4.407797732367443e-05, "loss": 0.0098, "step": 5610 }, { "epoch": 28.6734693877551, "grad_norm": 0.12491802126169205, "learning_rate": 4.391382687377268e-05, "loss": 0.0088, "step": 5620 }, { "epoch": 28.724489795918366, "grad_norm": 0.20640331506729126, "learning_rate": 4.374974298128512e-05, "loss": 0.0094, "step": 5630 }, { "epoch": 28.775510204081634, "grad_norm": 0.21418480575084686, "learning_rate": 4.358572744060699e-05, "loss": 0.0077, "step": 5640 }, { "epoch": 28.8265306122449, "grad_norm": 0.18439410626888275, "learning_rate": 4.342178204538588e-05, "loss": 0.0086, "step": 5650 }, { "epoch": 28.877551020408163, "grad_norm": 0.14028343558311462, "learning_rate": 4.325790858850241e-05, "loss": 0.0082, "step": 5660 }, { "epoch": 28.928571428571427, "grad_norm": 0.18500345945358276, "learning_rate": 4.309410886205043e-05, "loss": 0.0109, "step": 5670 }, { "epoch": 28.979591836734695, "grad_norm": 0.13544045388698578, "learning_rate": 4.293038465731752e-05, "loss": 0.0082, "step": 5680 }, { "epoch": 29.03061224489796, "grad_norm": 0.160054013133049, "learning_rate": 4.276673776476533e-05, "loss": 0.0089, "step": 5690 }, { "epoch": 29.081632653061224, "grad_norm": 0.14999930560588837, "learning_rate": 4.260316997401007e-05, "loss": 0.0077, "step": 5700 }, { "epoch": 29.132653061224488, "grad_norm": 0.1551632434129715, "learning_rate": 4.243968307380293e-05, "loss": 0.0093, "step": 5710 }, { "epoch": 29.183673469387756, "grad_norm": 0.1415146440267563, "learning_rate": 4.22762788520104e-05, "loss": 0.0082, "step": 5720 }, { "epoch": 29.23469387755102, "grad_norm": 0.23250897228717804, "learning_rate": 4.211295909559491e-05, "loss": 0.011, "step": 5730 }, { "epoch": 29.285714285714285, "grad_norm": 0.12581869959831238, "learning_rate": 4.194972559059511e-05, "loss": 0.0087, "step": 5740 }, { "epoch": 29.336734693877553, "grad_norm": 0.15572062134742737, "learning_rate": 4.178658012210651e-05, "loss": 0.0083, "step": 5750 }, { "epoch": 29.387755102040817, "grad_norm": 0.14380468428134918, "learning_rate": 4.162352447426177e-05, "loss": 0.0089, "step": 5760 }, { "epoch": 29.43877551020408, "grad_norm": 0.1303432732820511, "learning_rate": 4.146056043021135e-05, "loss": 0.0078, "step": 5770 }, { "epoch": 29.489795918367346, "grad_norm": 0.14290691912174225, "learning_rate": 4.1297689772103944e-05, "loss": 0.0083, "step": 5780 }, { "epoch": 29.540816326530614, "grad_norm": 0.210905060172081, "learning_rate": 4.113491428106694e-05, "loss": 0.0073, "step": 5790 }, { "epoch": 29.591836734693878, "grad_norm": 0.14722353219985962, "learning_rate": 4.0972235737187055e-05, "loss": 0.0083, "step": 5800 }, { "epoch": 29.642857142857142, "grad_norm": 0.2440355122089386, "learning_rate": 4.080965591949076e-05, "loss": 0.0091, "step": 5810 }, { "epoch": 29.693877551020407, "grad_norm": 0.17500843107700348, "learning_rate": 4.0647176605924924e-05, "loss": 0.0089, "step": 5820 }, { "epoch": 29.744897959183675, "grad_norm": 0.16860683262348175, "learning_rate": 4.0484799573337255e-05, "loss": 0.0072, "step": 5830 }, { "epoch": 29.79591836734694, "grad_norm": 0.18064527213573456, "learning_rate": 4.032252659745699e-05, "loss": 0.0077, "step": 5840 }, { "epoch": 29.846938775510203, "grad_norm": 0.16424335539340973, "learning_rate": 4.016035945287539e-05, "loss": 0.0081, "step": 5850 }, { "epoch": 29.897959183673468, "grad_norm": 0.13454100489616394, "learning_rate": 3.999829991302635e-05, "loss": 0.0083, "step": 5860 }, { "epoch": 29.948979591836736, "grad_norm": 0.16323933005332947, "learning_rate": 3.983634975016707e-05, "loss": 0.0073, "step": 5870 }, { "epoch": 30.0, "grad_norm": 0.08898316323757172, "learning_rate": 3.967451073535854e-05, "loss": 0.0074, "step": 5880 }, { "epoch": 30.051020408163264, "grad_norm": 0.13088096678256989, "learning_rate": 3.951278463844633e-05, "loss": 0.0058, "step": 5890 }, { "epoch": 30.102040816326532, "grad_norm": 0.18143142759799957, "learning_rate": 3.935117322804111e-05, "loss": 0.0087, "step": 5900 }, { "epoch": 30.153061224489797, "grad_norm": 0.18187865614891052, "learning_rate": 3.918967827149938e-05, "loss": 0.0087, "step": 5910 }, { "epoch": 30.20408163265306, "grad_norm": 0.24840015172958374, "learning_rate": 3.9028301534904094e-05, "loss": 0.0078, "step": 5920 }, { "epoch": 30.255102040816325, "grad_norm": 0.18876773118972778, "learning_rate": 3.88670447830454e-05, "loss": 0.01, "step": 5930 }, { "epoch": 30.306122448979593, "grad_norm": 0.31260260939598083, "learning_rate": 3.870590977940132e-05, "loss": 0.0098, "step": 5940 }, { "epoch": 30.357142857142858, "grad_norm": 0.19560395181179047, "learning_rate": 3.8544898286118404e-05, "loss": 0.0103, "step": 5950 }, { "epoch": 30.408163265306122, "grad_norm": 0.12452078610658646, "learning_rate": 3.838401206399257e-05, "loss": 0.0078, "step": 5960 }, { "epoch": 30.459183673469386, "grad_norm": 0.20444771647453308, "learning_rate": 3.822325287244975e-05, "loss": 0.0076, "step": 5970 }, { "epoch": 30.510204081632654, "grad_norm": 0.19818340241909027, "learning_rate": 3.8062622469526725e-05, "loss": 0.0092, "step": 5980 }, { "epoch": 30.56122448979592, "grad_norm": 0.2142488956451416, "learning_rate": 3.790212261185183e-05, "loss": 0.0085, "step": 5990 }, { "epoch": 30.612244897959183, "grad_norm": 0.14496754109859467, "learning_rate": 3.7741755054625794e-05, "loss": 0.007, "step": 6000 }, { "epoch": 30.663265306122447, "grad_norm": 0.1661590039730072, "learning_rate": 3.758152155160255e-05, "loss": 0.0097, "step": 6010 }, { "epoch": 30.714285714285715, "grad_norm": 0.40694093704223633, "learning_rate": 3.742142385506999e-05, "loss": 0.007, "step": 6020 }, { "epoch": 30.76530612244898, "grad_norm": 0.39330917596817017, "learning_rate": 3.72614637158309e-05, "loss": 0.009, "step": 6030 }, { "epoch": 30.816326530612244, "grad_norm": 0.1730642318725586, "learning_rate": 3.710164288318371e-05, "loss": 0.0089, "step": 6040 }, { "epoch": 30.867346938775512, "grad_norm": 0.1413699835538864, "learning_rate": 3.694196310490345e-05, "loss": 0.0085, "step": 6050 }, { "epoch": 30.918367346938776, "grad_norm": 0.1101151704788208, "learning_rate": 3.678242612722259e-05, "loss": 0.0075, "step": 6060 }, { "epoch": 30.96938775510204, "grad_norm": 0.1267882138490677, "learning_rate": 3.6623033694811953e-05, "loss": 0.0077, "step": 6070 }, { "epoch": 31.020408163265305, "grad_norm": 0.16635707020759583, "learning_rate": 3.6463787550761665e-05, "loss": 0.0108, "step": 6080 }, { "epoch": 31.071428571428573, "grad_norm": 0.25873029232025146, "learning_rate": 3.630468943656202e-05, "loss": 0.0082, "step": 6090 }, { "epoch": 31.122448979591837, "grad_norm": 0.225686714053154, "learning_rate": 3.6145741092084523e-05, "loss": 0.0081, "step": 6100 }, { "epoch": 31.1734693877551, "grad_norm": 0.18951626121997833, "learning_rate": 3.598694425556278e-05, "loss": 0.0087, "step": 6110 }, { "epoch": 31.224489795918366, "grad_norm": 0.2085665613412857, "learning_rate": 3.58283006635736e-05, "loss": 0.0092, "step": 6120 }, { "epoch": 31.275510204081634, "grad_norm": 0.15748299658298492, "learning_rate": 3.566981205101781e-05, "loss": 0.0084, "step": 6130 }, { "epoch": 31.3265306122449, "grad_norm": 0.7750831842422485, "learning_rate": 3.5511480151101556e-05, "loss": 0.0086, "step": 6140 }, { "epoch": 31.377551020408163, "grad_norm": 0.1870189607143402, "learning_rate": 3.5353306695317104e-05, "loss": 0.0071, "step": 6150 }, { "epoch": 31.428571428571427, "grad_norm": 0.21275849640369415, "learning_rate": 3.519529341342402e-05, "loss": 0.0073, "step": 6160 }, { "epoch": 31.479591836734695, "grad_norm": 0.12956885993480682, "learning_rate": 3.503744203343026e-05, "loss": 0.0066, "step": 6170 }, { "epoch": 31.53061224489796, "grad_norm": 0.16028520464897156, "learning_rate": 3.487975428157318e-05, "loss": 0.0068, "step": 6180 }, { "epoch": 31.581632653061224, "grad_norm": 0.2254897952079773, "learning_rate": 3.472223188230083e-05, "loss": 0.0077, "step": 6190 }, { "epoch": 31.632653061224488, "grad_norm": 0.2015315741300583, "learning_rate": 3.4564876558252866e-05, "loss": 0.0085, "step": 6200 }, { "epoch": 31.683673469387756, "grad_norm": 0.21496239304542542, "learning_rate": 3.440769003024195e-05, "loss": 0.0081, "step": 6210 }, { "epoch": 31.73469387755102, "grad_norm": 0.17642275989055634, "learning_rate": 3.425067401723477e-05, "loss": 0.0074, "step": 6220 }, { "epoch": 31.785714285714285, "grad_norm": 0.30243194103240967, "learning_rate": 3.409383023633325e-05, "loss": 0.0068, "step": 6230 }, { "epoch": 31.836734693877553, "grad_norm": 0.13930855691432953, "learning_rate": 3.3937160402755894e-05, "loss": 0.0078, "step": 6240 }, { "epoch": 31.887755102040817, "grad_norm": 0.19032539427280426, "learning_rate": 3.378066622981885e-05, "loss": 0.0074, "step": 6250 }, { "epoch": 31.93877551020408, "grad_norm": 0.19643543660640717, "learning_rate": 3.362434942891738e-05, "loss": 0.0068, "step": 6260 }, { "epoch": 31.989795918367346, "grad_norm": 0.17527306079864502, "learning_rate": 3.346821170950693e-05, "loss": 0.0068, "step": 6270 }, { "epoch": 32.04081632653061, "grad_norm": 0.1680750548839569, "learning_rate": 3.3312254779084585e-05, "loss": 0.0074, "step": 6280 }, { "epoch": 32.09183673469388, "grad_norm": 0.17037495970726013, "learning_rate": 3.315648034317039e-05, "loss": 0.0084, "step": 6290 }, { "epoch": 32.142857142857146, "grad_norm": 0.19460611045360565, "learning_rate": 3.3000890105288564e-05, "loss": 0.0073, "step": 6300 }, { "epoch": 32.19387755102041, "grad_norm": 0.14815102517604828, "learning_rate": 3.284548576694908e-05, "loss": 0.0066, "step": 6310 }, { "epoch": 32.244897959183675, "grad_norm": 0.1936333328485489, "learning_rate": 3.2690269027628815e-05, "loss": 0.0071, "step": 6320 }, { "epoch": 32.295918367346935, "grad_norm": 0.14736583828926086, "learning_rate": 3.253524158475324e-05, "loss": 0.0079, "step": 6330 }, { "epoch": 32.3469387755102, "grad_norm": 0.4856330454349518, "learning_rate": 3.238040513367757e-05, "loss": 0.0073, "step": 6340 }, { "epoch": 32.39795918367347, "grad_norm": 0.1489243358373642, "learning_rate": 3.222576136766843e-05, "loss": 0.0096, "step": 6350 }, { "epoch": 32.44897959183673, "grad_norm": 0.253857284784317, "learning_rate": 3.2071311977885324e-05, "loss": 0.0071, "step": 6360 }, { "epoch": 32.5, "grad_norm": 0.15831169486045837, "learning_rate": 3.191705865336197e-05, "loss": 0.0076, "step": 6370 }, { "epoch": 32.55102040816327, "grad_norm": 0.18786761164665222, "learning_rate": 3.1763003080988075e-05, "loss": 0.0074, "step": 6380 }, { "epoch": 32.60204081632653, "grad_norm": 0.10303844511508942, "learning_rate": 3.160914694549063e-05, "loss": 0.0083, "step": 6390 }, { "epoch": 32.6530612244898, "grad_norm": 0.16114811599254608, "learning_rate": 3.145549192941573e-05, "loss": 0.0087, "step": 6400 }, { "epoch": 32.704081632653065, "grad_norm": 0.18253128230571747, "learning_rate": 3.130203971310999e-05, "loss": 0.009, "step": 6410 }, { "epoch": 32.755102040816325, "grad_norm": 0.36176836490631104, "learning_rate": 3.114879197470225e-05, "loss": 0.0078, "step": 6420 }, { "epoch": 32.80612244897959, "grad_norm": 0.11460644751787186, "learning_rate": 3.0995750390085285e-05, "loss": 0.0063, "step": 6430 }, { "epoch": 32.857142857142854, "grad_norm": 0.15334132313728333, "learning_rate": 3.084291663289728e-05, "loss": 0.0081, "step": 6440 }, { "epoch": 32.90816326530612, "grad_norm": 0.12485190480947495, "learning_rate": 3.069029237450375e-05, "loss": 0.0066, "step": 6450 }, { "epoch": 32.95918367346939, "grad_norm": 0.17049705982208252, "learning_rate": 3.053787928397911e-05, "loss": 0.0072, "step": 6460 }, { "epoch": 33.01020408163265, "grad_norm": 0.13319264352321625, "learning_rate": 3.0385679028088526e-05, "loss": 0.0059, "step": 6470 }, { "epoch": 33.06122448979592, "grad_norm": 0.1867329329252243, "learning_rate": 3.023369327126959e-05, "loss": 0.007, "step": 6480 }, { "epoch": 33.11224489795919, "grad_norm": 0.44456103444099426, "learning_rate": 3.0081923675614198e-05, "loss": 0.0087, "step": 6490 }, { "epoch": 33.16326530612245, "grad_norm": 0.15839847922325134, "learning_rate": 2.993037190085034e-05, "loss": 0.0073, "step": 6500 }, { "epoch": 33.214285714285715, "grad_norm": 0.1503809690475464, "learning_rate": 2.977903960432392e-05, "loss": 0.0082, "step": 6510 }, { "epoch": 33.265306122448976, "grad_norm": 0.0889095589518547, "learning_rate": 2.9627928440980722e-05, "loss": 0.007, "step": 6520 }, { "epoch": 33.316326530612244, "grad_norm": 0.13975508511066437, "learning_rate": 2.9477040063348183e-05, "loss": 0.0076, "step": 6530 }, { "epoch": 33.36734693877551, "grad_norm": 0.1615682691335678, "learning_rate": 2.9326376121517456e-05, "loss": 0.0057, "step": 6540 }, { "epoch": 33.41836734693877, "grad_norm": 0.14844313263893127, "learning_rate": 2.9175938263125236e-05, "loss": 0.0076, "step": 6550 }, { "epoch": 33.46938775510204, "grad_norm": 0.11719641834497452, "learning_rate": 2.9025728133335873e-05, "loss": 0.0077, "step": 6560 }, { "epoch": 33.52040816326531, "grad_norm": 0.17526297271251678, "learning_rate": 2.8875747374823288e-05, "loss": 0.0081, "step": 6570 }, { "epoch": 33.57142857142857, "grad_norm": 0.17895475029945374, "learning_rate": 2.872599762775298e-05, "loss": 0.0062, "step": 6580 }, { "epoch": 33.62244897959184, "grad_norm": 0.200625479221344, "learning_rate": 2.857648052976425e-05, "loss": 0.0071, "step": 6590 }, { "epoch": 33.673469387755105, "grad_norm": 0.18703550100326538, "learning_rate": 2.8427197715952047e-05, "loss": 0.0083, "step": 6600 }, { "epoch": 33.724489795918366, "grad_norm": 0.16133534908294678, "learning_rate": 2.8278150818849393e-05, "loss": 0.0073, "step": 6610 }, { "epoch": 33.775510204081634, "grad_norm": 0.07263559848070145, "learning_rate": 2.812934146840922e-05, "loss": 0.0084, "step": 6620 }, { "epoch": 33.826530612244895, "grad_norm": 0.19839783012866974, "learning_rate": 2.7980771291986764e-05, "loss": 0.0088, "step": 6630 }, { "epoch": 33.87755102040816, "grad_norm": 0.14743749797344208, "learning_rate": 2.783244191432167e-05, "loss": 0.0083, "step": 6640 }, { "epoch": 33.92857142857143, "grad_norm": 0.17601127922534943, "learning_rate": 2.768435495752022e-05, "loss": 0.0086, "step": 6650 }, { "epoch": 33.97959183673469, "grad_norm": 0.14149773120880127, "learning_rate": 2.753651204103771e-05, "loss": 0.0066, "step": 6660 }, { "epoch": 34.03061224489796, "grad_norm": 0.12192489951848984, "learning_rate": 2.7388914781660523e-05, "loss": 0.0073, "step": 6670 }, { "epoch": 34.08163265306123, "grad_norm": 0.15470747649669647, "learning_rate": 2.7241564793488693e-05, "loss": 0.0072, "step": 6680 }, { "epoch": 34.13265306122449, "grad_norm": 0.1043485477566719, "learning_rate": 2.7094463687918037e-05, "loss": 0.0068, "step": 6690 }, { "epoch": 34.183673469387756, "grad_norm": 0.15143121778964996, "learning_rate": 2.694761307362268e-05, "loss": 0.0068, "step": 6700 }, { "epoch": 34.234693877551024, "grad_norm": 0.10546765476465225, "learning_rate": 2.6801014556537467e-05, "loss": 0.0074, "step": 6710 }, { "epoch": 34.285714285714285, "grad_norm": 0.17052751779556274, "learning_rate": 2.6654669739840243e-05, "loss": 0.0086, "step": 6720 }, { "epoch": 34.33673469387755, "grad_norm": 0.14397816359996796, "learning_rate": 2.650858022393451e-05, "loss": 0.0077, "step": 6730 }, { "epoch": 34.38775510204081, "grad_norm": 0.1201322004199028, "learning_rate": 2.6362747606431747e-05, "loss": 0.007, "step": 6740 }, { "epoch": 34.43877551020408, "grad_norm": 0.1957298368215561, "learning_rate": 2.6217173482134172e-05, "loss": 0.0095, "step": 6750 }, { "epoch": 34.48979591836735, "grad_norm": 0.12146192044019699, "learning_rate": 2.6071859443017044e-05, "loss": 0.0075, "step": 6760 }, { "epoch": 34.54081632653061, "grad_norm": 0.14981606602668762, "learning_rate": 2.5926807078211414e-05, "loss": 0.0084, "step": 6770 }, { "epoch": 34.59183673469388, "grad_norm": 0.1322251260280609, "learning_rate": 2.5782017973986728e-05, "loss": 0.0079, "step": 6780 }, { "epoch": 34.642857142857146, "grad_norm": 0.13431791961193085, "learning_rate": 2.5637493713733374e-05, "loss": 0.0076, "step": 6790 }, { "epoch": 34.69387755102041, "grad_norm": 0.0918695256114006, "learning_rate": 2.549323587794559e-05, "loss": 0.0073, "step": 6800 }, { "epoch": 34.744897959183675, "grad_norm": 0.13444942235946655, "learning_rate": 2.5349246044203895e-05, "loss": 0.0068, "step": 6810 }, { "epoch": 34.795918367346935, "grad_norm": 0.11196573078632355, "learning_rate": 2.520552578715808e-05, "loss": 0.0077, "step": 6820 }, { "epoch": 34.8469387755102, "grad_norm": 0.09248281270265579, "learning_rate": 2.506207667850981e-05, "loss": 0.0071, "step": 6830 }, { "epoch": 34.89795918367347, "grad_norm": 0.16308583319187164, "learning_rate": 2.4918900286995555e-05, "loss": 0.0078, "step": 6840 }, { "epoch": 34.94897959183673, "grad_norm": 0.10463730990886688, "learning_rate": 2.4775998178369458e-05, "loss": 0.0062, "step": 6850 }, { "epoch": 35.0, "grad_norm": 0.2387719452381134, "learning_rate": 2.4633371915386017e-05, "loss": 0.0071, "step": 6860 }, { "epoch": 35.05102040816327, "grad_norm": 0.2909601628780365, "learning_rate": 2.4491023057783235e-05, "loss": 0.0071, "step": 6870 }, { "epoch": 35.10204081632653, "grad_norm": 0.1645418107509613, "learning_rate": 2.4348953162265375e-05, "loss": 0.0086, "step": 6880 }, { "epoch": 35.1530612244898, "grad_norm": 0.1207936480641365, "learning_rate": 2.420716378248607e-05, "loss": 0.0066, "step": 6890 }, { "epoch": 35.204081632653065, "grad_norm": 0.11405135691165924, "learning_rate": 2.4065656469031266e-05, "loss": 0.0074, "step": 6900 }, { "epoch": 35.255102040816325, "grad_norm": 0.19770999252796173, "learning_rate": 2.3924432769402268e-05, "loss": 0.0085, "step": 6910 }, { "epoch": 35.30612244897959, "grad_norm": 0.12510912120342255, "learning_rate": 2.3783494227998844e-05, "loss": 0.0071, "step": 6920 }, { "epoch": 35.357142857142854, "grad_norm": 0.19053128361701965, "learning_rate": 2.3642842386102264e-05, "loss": 0.0082, "step": 6930 }, { "epoch": 35.40816326530612, "grad_norm": 0.1479298174381256, "learning_rate": 2.3502478781858567e-05, "loss": 0.0057, "step": 6940 }, { "epoch": 35.45918367346939, "grad_norm": 0.19798047840595245, "learning_rate": 2.3362404950261628e-05, "loss": 0.0071, "step": 6950 }, { "epoch": 35.51020408163265, "grad_norm": 0.12378936260938644, "learning_rate": 2.3222622423136458e-05, "loss": 0.0068, "step": 6960 }, { "epoch": 35.56122448979592, "grad_norm": 0.12424926459789276, "learning_rate": 2.3083132729122332e-05, "loss": 0.0071, "step": 6970 }, { "epoch": 35.61224489795919, "grad_norm": 0.08270122855901718, "learning_rate": 2.294393739365621e-05, "loss": 0.0072, "step": 6980 }, { "epoch": 35.66326530612245, "grad_norm": 0.1557644158601761, "learning_rate": 2.2805037938956e-05, "loss": 0.0061, "step": 6990 }, { "epoch": 35.714285714285715, "grad_norm": 0.18610058724880219, "learning_rate": 2.266643588400386e-05, "loss": 0.0065, "step": 7000 }, { "epoch": 35.765306122448976, "grad_norm": 0.12360028177499771, "learning_rate": 2.252813274452969e-05, "loss": 0.0073, "step": 7010 }, { "epoch": 35.816326530612244, "grad_norm": 0.1347714364528656, "learning_rate": 2.2390130032994427e-05, "loss": 0.0064, "step": 7020 }, { "epoch": 35.86734693877551, "grad_norm": 0.1429823338985443, "learning_rate": 2.2252429258573633e-05, "loss": 0.0063, "step": 7030 }, { "epoch": 35.91836734693877, "grad_norm": 0.10499073565006256, "learning_rate": 2.2115031927140904e-05, "loss": 0.0064, "step": 7040 }, { "epoch": 35.96938775510204, "grad_norm": 0.14435063302516937, "learning_rate": 2.1977939541251463e-05, "loss": 0.0065, "step": 7050 }, { "epoch": 36.02040816326531, "grad_norm": 0.11588199436664581, "learning_rate": 2.1841153600125684e-05, "loss": 0.0056, "step": 7060 }, { "epoch": 36.07142857142857, "grad_norm": 0.17952819168567657, "learning_rate": 2.170467559963267e-05, "loss": 0.0084, "step": 7070 }, { "epoch": 36.12244897959184, "grad_norm": 0.14735175669193268, "learning_rate": 2.1568507032273982e-05, "loss": 0.0066, "step": 7080 }, { "epoch": 36.173469387755105, "grad_norm": 0.19875267148017883, "learning_rate": 2.1432649387167264e-05, "loss": 0.0068, "step": 7090 }, { "epoch": 36.224489795918366, "grad_norm": 0.3810866177082062, "learning_rate": 2.1297104150029973e-05, "loss": 0.0059, "step": 7100 }, { "epoch": 36.275510204081634, "grad_norm": 0.15593960881233215, "learning_rate": 2.116187280316307e-05, "loss": 0.0069, "step": 7110 }, { "epoch": 36.326530612244895, "grad_norm": 0.14989875257015228, "learning_rate": 2.1026956825434908e-05, "loss": 0.0066, "step": 7120 }, { "epoch": 36.37755102040816, "grad_norm": 0.13820670545101166, "learning_rate": 2.0892357692265017e-05, "loss": 0.0071, "step": 7130 }, { "epoch": 36.42857142857143, "grad_norm": 0.14701518416404724, "learning_rate": 2.0758076875607947e-05, "loss": 0.0057, "step": 7140 }, { "epoch": 36.47959183673469, "grad_norm": 0.15851764380931854, "learning_rate": 2.0624115843937207e-05, "loss": 0.0052, "step": 7150 }, { "epoch": 36.53061224489796, "grad_norm": 0.16059838235378265, "learning_rate": 2.0490476062229157e-05, "loss": 0.007, "step": 7160 }, { "epoch": 36.58163265306123, "grad_norm": 0.12356032431125641, "learning_rate": 2.035715899194704e-05, "loss": 0.0072, "step": 7170 }, { "epoch": 36.63265306122449, "grad_norm": 0.15733937919139862, "learning_rate": 2.022416609102499e-05, "loss": 0.007, "step": 7180 }, { "epoch": 36.683673469387756, "grad_norm": 0.16273163259029388, "learning_rate": 2.009149881385205e-05, "loss": 0.0074, "step": 7190 }, { "epoch": 36.734693877551024, "grad_norm": 0.10188206285238266, "learning_rate": 1.995915861125634e-05, "loss": 0.0065, "step": 7200 }, { "epoch": 36.785714285714285, "grad_norm": 0.11456573754549026, "learning_rate": 1.9827146930489065e-05, "loss": 0.0068, "step": 7210 }, { "epoch": 36.83673469387755, "grad_norm": 0.08121509104967117, "learning_rate": 1.9695465215208848e-05, "loss": 0.0054, "step": 7220 }, { "epoch": 36.88775510204081, "grad_norm": 0.11486631631851196, "learning_rate": 1.9564114905465813e-05, "loss": 0.0079, "step": 7230 }, { "epoch": 36.93877551020408, "grad_norm": 0.1250721961259842, "learning_rate": 1.9433097437685936e-05, "loss": 0.0086, "step": 7240 }, { "epoch": 36.98979591836735, "grad_norm": 0.09356781840324402, "learning_rate": 1.930241424465521e-05, "loss": 0.0056, "step": 7250 }, { "epoch": 37.04081632653061, "grad_norm": 0.12055828422307968, "learning_rate": 1.9172066755504115e-05, "loss": 0.0058, "step": 7260 }, { "epoch": 37.09183673469388, "grad_norm": 0.1686311513185501, "learning_rate": 1.9042056395691914e-05, "loss": 0.0066, "step": 7270 }, { "epoch": 37.142857142857146, "grad_norm": 0.2329787164926529, "learning_rate": 1.8912384586991066e-05, "loss": 0.0068, "step": 7280 }, { "epoch": 37.19387755102041, "grad_norm": 0.13944311439990997, "learning_rate": 1.8783052747471717e-05, "loss": 0.0066, "step": 7290 }, { "epoch": 37.244897959183675, "grad_norm": 0.14261546730995178, "learning_rate": 1.865406229148611e-05, "loss": 0.0055, "step": 7300 }, { "epoch": 37.295918367346935, "grad_norm": 0.22354821860790253, "learning_rate": 1.8525414629653233e-05, "loss": 0.0065, "step": 7310 }, { "epoch": 37.3469387755102, "grad_norm": 0.12241596728563309, "learning_rate": 1.8397111168843255e-05, "loss": 0.0075, "step": 7320 }, { "epoch": 37.39795918367347, "grad_norm": 0.10310354828834534, "learning_rate": 1.8269153312162323e-05, "loss": 0.0054, "step": 7330 }, { "epoch": 37.44897959183673, "grad_norm": 0.08391445875167847, "learning_rate": 1.8141542458937054e-05, "loss": 0.0055, "step": 7340 }, { "epoch": 37.5, "grad_norm": 0.08221261948347092, "learning_rate": 1.8014280004699268e-05, "loss": 0.0058, "step": 7350 }, { "epoch": 37.55102040816327, "grad_norm": 0.11536182463169098, "learning_rate": 1.788736734117078e-05, "loss": 0.0101, "step": 7360 }, { "epoch": 37.60204081632653, "grad_norm": 0.1698896735906601, "learning_rate": 1.7760805856248152e-05, "loss": 0.0085, "step": 7370 }, { "epoch": 37.6530612244898, "grad_norm": 0.14559519290924072, "learning_rate": 1.7634596933987518e-05, "loss": 0.007, "step": 7380 }, { "epoch": 37.704081632653065, "grad_norm": 0.1524273008108139, "learning_rate": 1.7508741954589404e-05, "loss": 0.0069, "step": 7390 }, { "epoch": 37.755102040816325, "grad_norm": 0.12319854646921158, "learning_rate": 1.7383242294383717e-05, "loss": 0.006, "step": 7400 }, { "epoch": 37.80612244897959, "grad_norm": 0.15481755137443542, "learning_rate": 1.7258099325814632e-05, "loss": 0.0053, "step": 7410 }, { "epoch": 37.857142857142854, "grad_norm": 0.0687246322631836, "learning_rate": 1.7133314417425594e-05, "loss": 0.0056, "step": 7420 }, { "epoch": 37.90816326530612, "grad_norm": 0.09326207637786865, "learning_rate": 1.7008888933844408e-05, "loss": 0.0059, "step": 7430 }, { "epoch": 37.95918367346939, "grad_norm": 0.1630684733390808, "learning_rate": 1.6884824235768172e-05, "loss": 0.0056, "step": 7440 }, { "epoch": 38.01020408163265, "grad_norm": 0.10772176086902618, "learning_rate": 1.6761121679948592e-05, "loss": 0.0058, "step": 7450 }, { "epoch": 38.06122448979592, "grad_norm": 0.15577349066734314, "learning_rate": 1.663778261917695e-05, "loss": 0.0075, "step": 7460 }, { "epoch": 38.11224489795919, "grad_norm": 0.22791704535484314, "learning_rate": 1.651480840226952e-05, "loss": 0.006, "step": 7470 }, { "epoch": 38.16326530612245, "grad_norm": 0.11536995321512222, "learning_rate": 1.639220037405258e-05, "loss": 0.005, "step": 7480 }, { "epoch": 38.214285714285715, "grad_norm": 0.10460111498832703, "learning_rate": 1.6269959875347906e-05, "loss": 0.0051, "step": 7490 }, { "epoch": 38.265306122448976, "grad_norm": 0.21754498779773712, "learning_rate": 1.614808824295802e-05, "loss": 0.0065, "step": 7500 }, { "epoch": 38.316326530612244, "grad_norm": 0.35055291652679443, "learning_rate": 1.602658680965152e-05, "loss": 0.0054, "step": 7510 }, { "epoch": 38.36734693877551, "grad_norm": 0.09078759700059891, "learning_rate": 1.5905456904148686e-05, "loss": 0.0072, "step": 7520 }, { "epoch": 38.41836734693877, "grad_norm": 0.11175307631492615, "learning_rate": 1.57846998511067e-05, "loss": 0.0055, "step": 7530 }, { "epoch": 38.46938775510204, "grad_norm": 0.1477552056312561, "learning_rate": 1.566431697110538e-05, "loss": 0.0058, "step": 7540 }, { "epoch": 38.52040816326531, "grad_norm": 0.1321031004190445, "learning_rate": 1.554430958063259e-05, "loss": 0.0056, "step": 7550 }, { "epoch": 38.57142857142857, "grad_norm": 0.12381129711866379, "learning_rate": 1.5424678992069912e-05, "loss": 0.0077, "step": 7560 }, { "epoch": 38.62244897959184, "grad_norm": 0.25669065117836, "learning_rate": 1.5305426513678362e-05, "loss": 0.0053, "step": 7570 }, { "epoch": 38.673469387755105, "grad_norm": 0.1067737340927124, "learning_rate": 1.518655344958388e-05, "loss": 0.0057, "step": 7580 }, { "epoch": 38.724489795918366, "grad_norm": 0.15630072355270386, "learning_rate": 1.5068061099763275e-05, "loss": 0.0053, "step": 7590 }, { "epoch": 38.775510204081634, "grad_norm": 0.12403763830661774, "learning_rate": 1.494995076002988e-05, "loss": 0.0065, "step": 7600 }, { "epoch": 38.826530612244895, "grad_norm": 0.18981009721755981, "learning_rate": 1.4832223722019456e-05, "loss": 0.005, "step": 7610 }, { "epoch": 38.87755102040816, "grad_norm": 0.13081875443458557, "learning_rate": 1.4714881273176035e-05, "loss": 0.0062, "step": 7620 }, { "epoch": 38.92857142857143, "grad_norm": 0.23744253814220428, "learning_rate": 1.4597924696737835e-05, "loss": 0.0063, "step": 7630 }, { "epoch": 38.97959183673469, "grad_norm": 0.15743772685527802, "learning_rate": 1.4481355271723252e-05, "loss": 0.006, "step": 7640 }, { "epoch": 39.03061224489796, "grad_norm": 0.10736050456762314, "learning_rate": 1.4365174272916809e-05, "loss": 0.0065, "step": 7650 }, { "epoch": 39.08163265306123, "grad_norm": 0.1354866474866867, "learning_rate": 1.4249382970855319e-05, "loss": 0.0074, "step": 7660 }, { "epoch": 39.13265306122449, "grad_norm": 0.16383299231529236, "learning_rate": 1.4133982631813903e-05, "loss": 0.0063, "step": 7670 }, { "epoch": 39.183673469387756, "grad_norm": 0.16017812490463257, "learning_rate": 1.4018974517792194e-05, "loss": 0.0054, "step": 7680 }, { "epoch": 39.234693877551024, "grad_norm": 0.18350949883460999, "learning_rate": 1.390435988650048e-05, "loss": 0.0055, "step": 7690 }, { "epoch": 39.285714285714285, "grad_norm": 0.16759999096393585, "learning_rate": 1.3790139991346006e-05, "loss": 0.008, "step": 7700 }, { "epoch": 39.33673469387755, "grad_norm": 0.11148501187562943, "learning_rate": 1.367631608141926e-05, "loss": 0.007, "step": 7710 }, { "epoch": 39.38775510204081, "grad_norm": 0.1648229956626892, "learning_rate": 1.3562889401480278e-05, "loss": 0.0052, "step": 7720 }, { "epoch": 39.43877551020408, "grad_norm": 0.15598870813846588, "learning_rate": 1.3449861191945074e-05, "loss": 0.0059, "step": 7730 }, { "epoch": 39.48979591836735, "grad_norm": 0.13632188737392426, "learning_rate": 1.3337232688872009e-05, "loss": 0.0056, "step": 7740 }, { "epoch": 39.54081632653061, "grad_norm": 0.20490474998950958, "learning_rate": 1.3225005123948364e-05, "loss": 0.006, "step": 7750 }, { "epoch": 39.59183673469388, "grad_norm": 0.10257858037948608, "learning_rate": 1.311317972447681e-05, "loss": 0.0054, "step": 7760 }, { "epoch": 39.642857142857146, "grad_norm": 0.09683653712272644, "learning_rate": 1.3001757713361996e-05, "loss": 0.0089, "step": 7770 }, { "epoch": 39.69387755102041, "grad_norm": 0.07005012780427933, "learning_rate": 1.2890740309097204e-05, "loss": 0.0063, "step": 7780 }, { "epoch": 39.744897959183675, "grad_norm": 0.10676898807287216, "learning_rate": 1.2780128725750944e-05, "loss": 0.0047, "step": 7790 }, { "epoch": 39.795918367346935, "grad_norm": 0.05751161649823189, "learning_rate": 1.266992417295379e-05, "loss": 0.0061, "step": 7800 }, { "epoch": 39.8469387755102, "grad_norm": 0.09668602794408798, "learning_rate": 1.2560127855885073e-05, "loss": 0.005, "step": 7810 }, { "epoch": 39.89795918367347, "grad_norm": 0.16969068348407745, "learning_rate": 1.2450740975259745e-05, "loss": 0.005, "step": 7820 }, { "epoch": 39.94897959183673, "grad_norm": 0.10911020636558533, "learning_rate": 1.234176472731517e-05, "loss": 0.0074, "step": 7830 }, { "epoch": 40.0, "grad_norm": 0.11451397091150284, "learning_rate": 1.2233200303798158e-05, "loss": 0.0049, "step": 7840 }, { "epoch": 40.05102040816327, "grad_norm": 0.12042118608951569, "learning_rate": 1.2125048891951846e-05, "loss": 0.0054, "step": 7850 }, { "epoch": 40.10204081632653, "grad_norm": 0.0865764319896698, "learning_rate": 1.2017311674502745e-05, "loss": 0.0067, "step": 7860 }, { "epoch": 40.1530612244898, "grad_norm": 0.17693985998630524, "learning_rate": 1.1909989829647822e-05, "loss": 0.0056, "step": 7870 }, { "epoch": 40.204081632653065, "grad_norm": 0.2039027363061905, "learning_rate": 1.1803084531041553e-05, "loss": 0.0051, "step": 7880 }, { "epoch": 40.255102040816325, "grad_norm": 0.35931265354156494, "learning_rate": 1.1696596947783162e-05, "loss": 0.0063, "step": 7890 }, { "epoch": 40.30612244897959, "grad_norm": 0.26195549964904785, "learning_rate": 1.1590528244403803e-05, "loss": 0.006, "step": 7900 }, { "epoch": 40.357142857142854, "grad_norm": 0.18848632276058197, "learning_rate": 1.148487958085382e-05, "loss": 0.0064, "step": 7910 }, { "epoch": 40.40816326530612, "grad_norm": 0.15940047800540924, "learning_rate": 1.1379652112490086e-05, "loss": 0.0056, "step": 7920 }, { "epoch": 40.45918367346939, "grad_norm": 0.1535605490207672, "learning_rate": 1.1274846990063315e-05, "loss": 0.0059, "step": 7930 }, { "epoch": 40.51020408163265, "grad_norm": 0.10936044901609421, "learning_rate": 1.117046535970554e-05, "loss": 0.0063, "step": 7940 }, { "epoch": 40.56122448979592, "grad_norm": 0.12344779074192047, "learning_rate": 1.106650836291755e-05, "loss": 0.0061, "step": 7950 }, { "epoch": 40.61224489795919, "grad_norm": 0.1099054291844368, "learning_rate": 1.0962977136556418e-05, "loss": 0.0068, "step": 7960 }, { "epoch": 40.66326530612245, "grad_norm": 0.1716424524784088, "learning_rate": 1.0859872812823024e-05, "loss": 0.0076, "step": 7970 }, { "epoch": 40.714285714285715, "grad_norm": 0.13540196418762207, "learning_rate": 1.0757196519249747e-05, "loss": 0.0069, "step": 7980 }, { "epoch": 40.765306122448976, "grad_norm": 0.10966414213180542, "learning_rate": 1.0654949378688077e-05, "loss": 0.0051, "step": 7990 }, { "epoch": 40.816326530612244, "grad_norm": 0.06738685816526413, "learning_rate": 1.0553132509296376e-05, "loss": 0.0058, "step": 8000 }, { "epoch": 40.86734693877551, "grad_norm": 0.09318239986896515, "learning_rate": 1.0451747024527613e-05, "loss": 0.0046, "step": 8010 }, { "epoch": 40.91836734693877, "grad_norm": 0.10648567229509354, "learning_rate": 1.0350794033117189e-05, "loss": 0.0052, "step": 8020 }, { "epoch": 40.96938775510204, "grad_norm": 0.10357114672660828, "learning_rate": 1.0250274639070856e-05, "loss": 0.0056, "step": 8030 }, { "epoch": 41.02040816326531, "grad_norm": 0.07822016626596451, "learning_rate": 1.0150189941652599e-05, "loss": 0.0059, "step": 8040 }, { "epoch": 41.07142857142857, "grad_norm": 0.10909505188465118, "learning_rate": 1.0050541035372635e-05, "loss": 0.0054, "step": 8050 }, { "epoch": 41.12244897959184, "grad_norm": 0.16779902577400208, "learning_rate": 9.951329009975458e-06, "loss": 0.0058, "step": 8060 }, { "epoch": 41.173469387755105, "grad_norm": 0.12928882241249084, "learning_rate": 9.852554950427845e-06, "loss": 0.0056, "step": 8070 }, { "epoch": 41.224489795918366, "grad_norm": 0.08801032602787018, "learning_rate": 9.754219936907105e-06, "loss": 0.0066, "step": 8080 }, { "epoch": 41.275510204081634, "grad_norm": 0.1047731414437294, "learning_rate": 9.656325044789194e-06, "loss": 0.0056, "step": 8090 }, { "epoch": 41.326530612244895, "grad_norm": 0.11543439328670502, "learning_rate": 9.55887134463697e-06, "loss": 0.0064, "step": 8100 }, { "epoch": 41.37755102040816, "grad_norm": 0.12027120590209961, "learning_rate": 9.461859902188475e-06, "loss": 0.0051, "step": 8110 }, { "epoch": 41.42857142857143, "grad_norm": 0.1366007924079895, "learning_rate": 9.365291778345303e-06, "loss": 0.0056, "step": 8120 }, { "epoch": 41.47959183673469, "grad_norm": 0.13182061910629272, "learning_rate": 9.269168029160991e-06, "loss": 0.006, "step": 8130 }, { "epoch": 41.53061224489796, "grad_norm": 0.09126241505146027, "learning_rate": 9.173489705829447e-06, "loss": 0.005, "step": 8140 }, { "epoch": 41.58163265306123, "grad_norm": 0.09983145445585251, "learning_rate": 9.078257854673516e-06, "loss": 0.0058, "step": 8150 }, { "epoch": 41.63265306122449, "grad_norm": 0.09555571526288986, "learning_rate": 8.983473517133429e-06, "loss": 0.0057, "step": 8160 }, { "epoch": 41.683673469387756, "grad_norm": 0.10230182111263275, "learning_rate": 8.889137729755537e-06, "loss": 0.0052, "step": 8170 }, { "epoch": 41.734693877551024, "grad_norm": 0.22840645909309387, "learning_rate": 8.79525152418087e-06, "loss": 0.0058, "step": 8180 }, { "epoch": 41.785714285714285, "grad_norm": 0.1212320327758789, "learning_rate": 8.701815927133961e-06, "loss": 0.0059, "step": 8190 }, { "epoch": 41.83673469387755, "grad_norm": 0.10512705147266388, "learning_rate": 8.608831960411534e-06, "loss": 0.0053, "step": 8200 }, { "epoch": 41.88775510204081, "grad_norm": 0.14518438279628754, "learning_rate": 8.516300640871321e-06, "loss": 0.0053, "step": 8210 }, { "epoch": 41.93877551020408, "grad_norm": 0.1743626445531845, "learning_rate": 8.424222980421038e-06, "loss": 0.0046, "step": 8220 }, { "epoch": 41.98979591836735, "grad_norm": 0.12512050569057465, "learning_rate": 8.332599986007184e-06, "loss": 0.0049, "step": 8230 }, { "epoch": 42.04081632653061, "grad_norm": 0.31048712134361267, "learning_rate": 8.241432659604203e-06, "loss": 0.006, "step": 8240 }, { "epoch": 42.09183673469388, "grad_norm": 0.09587971121072769, "learning_rate": 8.150721998203331e-06, "loss": 0.0062, "step": 8250 }, { "epoch": 42.142857142857146, "grad_norm": 0.12172222137451172, "learning_rate": 8.06046899380184e-06, "loss": 0.0045, "step": 8260 }, { "epoch": 42.19387755102041, "grad_norm": 0.11157670617103577, "learning_rate": 7.970674633392133e-06, "loss": 0.0049, "step": 8270 }, { "epoch": 42.244897959183675, "grad_norm": 0.44152021408081055, "learning_rate": 7.881339898950924e-06, "loss": 0.0051, "step": 8280 }, { "epoch": 42.295918367346935, "grad_norm": 0.12573958933353424, "learning_rate": 7.792465767428597e-06, "loss": 0.0057, "step": 8290 }, { "epoch": 42.3469387755102, "grad_norm": 0.21653640270233154, "learning_rate": 7.704053210738376e-06, "loss": 0.005, "step": 8300 }, { "epoch": 42.39795918367347, "grad_norm": 0.13924400508403778, "learning_rate": 7.6161031957458494e-06, "loss": 0.0062, "step": 8310 }, { "epoch": 42.44897959183673, "grad_norm": 0.13284781575202942, "learning_rate": 7.5286166842582605e-06, "loss": 0.0045, "step": 8320 }, { "epoch": 42.5, "grad_norm": 0.12139739096164703, "learning_rate": 7.4415946330140814e-06, "loss": 0.0053, "step": 8330 }, { "epoch": 42.55102040816327, "grad_norm": 0.2873706519603729, "learning_rate": 7.3550379936725644e-06, "loss": 0.0052, "step": 8340 }, { "epoch": 42.60204081632653, "grad_norm": 0.21512092649936676, "learning_rate": 7.2689477128032035e-06, "loss": 0.0064, "step": 8350 }, { "epoch": 42.6530612244898, "grad_norm": 1.0677272081375122, "learning_rate": 7.183324731875551e-06, "loss": 0.0057, "step": 8360 }, { "epoch": 42.704081632653065, "grad_norm": 0.11112631857395172, "learning_rate": 7.098169987248782e-06, "loss": 0.0043, "step": 8370 }, { "epoch": 42.755102040816325, "grad_norm": 0.09391681104898453, "learning_rate": 7.013484410161553e-06, "loss": 0.0049, "step": 8380 }, { "epoch": 42.80612244897959, "grad_norm": 0.12046920508146286, "learning_rate": 6.92926892672176e-06, "loss": 0.0053, "step": 8390 }, { "epoch": 42.857142857142854, "grad_norm": 0.05072518065571785, "learning_rate": 6.845524457896446e-06, "loss": 0.0052, "step": 8400 }, { "epoch": 42.90816326530612, "grad_norm": 0.1401190310716629, "learning_rate": 6.7622519195017165e-06, "loss": 0.0065, "step": 8410 }, { "epoch": 42.95918367346939, "grad_norm": 0.16936041414737701, "learning_rate": 6.679452222192684e-06, "loss": 0.0048, "step": 8420 }, { "epoch": 43.01020408163265, "grad_norm": 0.17046530544757843, "learning_rate": 6.597126271453579e-06, "loss": 0.0047, "step": 8430 }, { "epoch": 43.06122448979592, "grad_norm": 0.14290176331996918, "learning_rate": 6.51527496758782e-06, "loss": 0.0053, "step": 8440 }, { "epoch": 43.11224489795919, "grad_norm": 0.16380253434181213, "learning_rate": 6.433899205708155e-06, "loss": 0.0075, "step": 8450 }, { "epoch": 43.16326530612245, "grad_norm": 0.12436424940824509, "learning_rate": 6.352999875726856e-06, "loss": 0.0047, "step": 8460 }, { "epoch": 43.214285714285715, "grad_norm": 0.2028365135192871, "learning_rate": 6.272577862346052e-06, "loss": 0.0051, "step": 8470 }, { "epoch": 43.265306122448976, "grad_norm": 0.18097545206546783, "learning_rate": 6.192634045047996e-06, "loss": 0.005, "step": 8480 }, { "epoch": 43.316326530612244, "grad_norm": 0.08532506227493286, "learning_rate": 6.113169298085458e-06, "loss": 0.0058, "step": 8490 }, { "epoch": 43.36734693877551, "grad_norm": 0.1312270611524582, "learning_rate": 6.034184490472195e-06, "loss": 0.0069, "step": 8500 }, { "epoch": 43.41836734693877, "grad_norm": 0.0766311064362526, "learning_rate": 5.955680485973386e-06, "loss": 0.0048, "step": 8510 }, { "epoch": 43.46938775510204, "grad_norm": 0.10970660299062729, "learning_rate": 5.877658143096265e-06, "loss": 0.0046, "step": 8520 }, { "epoch": 43.52040816326531, "grad_norm": 0.11848579347133636, "learning_rate": 5.800118315080661e-06, "loss": 0.0046, "step": 8530 }, { "epoch": 43.57142857142857, "grad_norm": 0.07426905632019043, "learning_rate": 5.723061849889716e-06, "loss": 0.0063, "step": 8540 }, { "epoch": 43.62244897959184, "grad_norm": 0.11027754098176956, "learning_rate": 5.646489590200604e-06, "loss": 0.0048, "step": 8550 }, { "epoch": 43.673469387755105, "grad_norm": 0.13325151801109314, "learning_rate": 5.570402373395256e-06, "loss": 0.0057, "step": 8560 }, { "epoch": 43.724489795918366, "grad_norm": 0.25764501094818115, "learning_rate": 5.494801031551305e-06, "loss": 0.0079, "step": 8570 }, { "epoch": 43.775510204081634, "grad_norm": 0.13249604403972626, "learning_rate": 5.41968639143291e-06, "loss": 0.0051, "step": 8580 }, { "epoch": 43.826530612244895, "grad_norm": 0.07403765618801117, "learning_rate": 5.345059274481751e-06, "loss": 0.005, "step": 8590 }, { "epoch": 43.87755102040816, "grad_norm": 0.08898995071649551, "learning_rate": 5.270920496808002e-06, "loss": 0.0049, "step": 8600 }, { "epoch": 43.92857142857143, "grad_norm": 0.12128536403179169, "learning_rate": 5.1972708691814695e-06, "loss": 0.0065, "step": 8610 }, { "epoch": 43.97959183673469, "grad_norm": 0.1151263564825058, "learning_rate": 5.124111197022674e-06, "loss": 0.0048, "step": 8620 }, { "epoch": 44.03061224489796, "grad_norm": 0.11438912898302078, "learning_rate": 5.051442280394081e-06, "loss": 0.006, "step": 8630 }, { "epoch": 44.08163265306123, "grad_norm": 0.14923195540905, "learning_rate": 4.979264913991322e-06, "loss": 0.0055, "step": 8640 }, { "epoch": 44.13265306122449, "grad_norm": 0.08234194666147232, "learning_rate": 4.907579887134489e-06, "loss": 0.0055, "step": 8650 }, { "epoch": 44.183673469387756, "grad_norm": 0.10806439071893692, "learning_rate": 4.836387983759572e-06, "loss": 0.0051, "step": 8660 }, { "epoch": 44.234693877551024, "grad_norm": 0.08062957227230072, "learning_rate": 4.765689982409816e-06, "loss": 0.0056, "step": 8670 }, { "epoch": 44.285714285714285, "grad_norm": 0.09889788180589676, "learning_rate": 4.695486656227233e-06, "loss": 0.0056, "step": 8680 }, { "epoch": 44.33673469387755, "grad_norm": 0.15418440103530884, "learning_rate": 4.625778772944156e-06, "loss": 0.0045, "step": 8690 }, { "epoch": 44.38775510204081, "grad_norm": 0.10153764486312866, "learning_rate": 4.556567094874825e-06, "loss": 0.0039, "step": 8700 }, { "epoch": 44.43877551020408, "grad_norm": 0.11475788056850433, "learning_rate": 4.487852378907059e-06, "loss": 0.0056, "step": 8710 }, { "epoch": 44.48979591836735, "grad_norm": 0.15989309549331665, "learning_rate": 4.419635376493986e-06, "loss": 0.0058, "step": 8720 }, { "epoch": 44.54081632653061, "grad_norm": 0.17842482030391693, "learning_rate": 4.351916833645825e-06, "loss": 0.0046, "step": 8730 }, { "epoch": 44.59183673469388, "grad_norm": 0.10781311988830566, "learning_rate": 4.284697490921691e-06, "loss": 0.005, "step": 8740 }, { "epoch": 44.642857142857146, "grad_norm": 0.08490155637264252, "learning_rate": 4.2179780834215585e-06, "loss": 0.0039, "step": 8750 }, { "epoch": 44.69387755102041, "grad_norm": 0.09688279777765274, "learning_rate": 4.151759340778178e-06, "loss": 0.0047, "step": 8760 }, { "epoch": 44.744897959183675, "grad_norm": 0.08814825862646103, "learning_rate": 4.086041987149109e-06, "loss": 0.0058, "step": 8770 }, { "epoch": 44.795918367346935, "grad_norm": 0.11854441463947296, "learning_rate": 4.020826741208811e-06, "loss": 0.0057, "step": 8780 }, { "epoch": 44.8469387755102, "grad_norm": 0.11184152215719223, "learning_rate": 3.956114316140746e-06, "loss": 0.0053, "step": 8790 }, { "epoch": 44.89795918367347, "grad_norm": 0.08430422097444534, "learning_rate": 3.891905419629643e-06, "loss": 0.0039, "step": 8800 }, { "epoch": 44.94897959183673, "grad_norm": 0.40296468138694763, "learning_rate": 3.8282007538536946e-06, "loss": 0.0052, "step": 8810 }, { "epoch": 45.0, "grad_norm": 0.09073133021593094, "learning_rate": 3.7650010154769265e-06, "loss": 0.005, "step": 8820 }, { "epoch": 45.05102040816327, "grad_norm": 0.08713637292385101, "learning_rate": 3.7023068956415608e-06, "loss": 0.0047, "step": 8830 }, { "epoch": 45.10204081632653, "grad_norm": 0.1528698205947876, "learning_rate": 3.6401190799604303e-06, "loss": 0.0061, "step": 8840 }, { "epoch": 45.1530612244898, "grad_norm": 0.18639759719371796, "learning_rate": 3.578438248509536e-06, "loss": 0.0057, "step": 8850 }, { "epoch": 45.204081632653065, "grad_norm": 0.17868518829345703, "learning_rate": 3.5172650758205583e-06, "loss": 0.005, "step": 8860 }, { "epoch": 45.255102040816325, "grad_norm": 0.07096349447965622, "learning_rate": 3.45660023087353e-06, "loss": 0.0054, "step": 8870 }, { "epoch": 45.30612244897959, "grad_norm": 0.2474617063999176, "learning_rate": 3.3964443770894528e-06, "loss": 0.0049, "step": 8880 }, { "epoch": 45.357142857142854, "grad_norm": 0.19123110175132751, "learning_rate": 3.3367981723231245e-06, "loss": 0.0062, "step": 8890 }, { "epoch": 45.40816326530612, "grad_norm": 0.14141562581062317, "learning_rate": 3.2776622688558746e-06, "loss": 0.0053, "step": 8900 }, { "epoch": 45.45918367346939, "grad_norm": 0.1695157289505005, "learning_rate": 3.2190373133884677e-06, "loss": 0.0044, "step": 8910 }, { "epoch": 45.51020408163265, "grad_norm": 0.16435100138187408, "learning_rate": 3.1609239470340446e-06, "loss": 0.0055, "step": 8920 }, { "epoch": 45.56122448979592, "grad_norm": 0.10433336347341537, "learning_rate": 3.1033228053110373e-06, "loss": 0.0068, "step": 8930 }, { "epoch": 45.61224489795919, "grad_norm": 1.2468801736831665, "learning_rate": 3.0462345181363314e-06, "loss": 0.0052, "step": 8940 }, { "epoch": 45.66326530612245, "grad_norm": 0.09109508246183395, "learning_rate": 2.9896597098182654e-06, "loss": 0.0051, "step": 8950 }, { "epoch": 45.714285714285715, "grad_norm": 0.09234023094177246, "learning_rate": 2.933598999049891e-06, "loss": 0.006, "step": 8960 }, { "epoch": 45.765306122448976, "grad_norm": 0.16320347785949707, "learning_rate": 2.8780529989021697e-06, "loss": 0.0054, "step": 8970 }, { "epoch": 45.816326530612244, "grad_norm": 0.14051592350006104, "learning_rate": 2.823022316817242e-06, "loss": 0.0047, "step": 8980 }, { "epoch": 45.86734693877551, "grad_norm": 0.059627026319503784, "learning_rate": 2.7685075546018456e-06, "loss": 0.0059, "step": 8990 }, { "epoch": 45.91836734693877, "grad_norm": 0.10166656970977783, "learning_rate": 2.7145093084206598e-06, "loss": 0.0061, "step": 9000 }, { "epoch": 45.96938775510204, "grad_norm": 0.1093280166387558, "learning_rate": 2.661028168789892e-06, "loss": 0.0059, "step": 9010 }, { "epoch": 46.02040816326531, "grad_norm": 0.08354687690734863, "learning_rate": 2.6080647205706855e-06, "loss": 0.0065, "step": 9020 }, { "epoch": 46.07142857142857, "grad_norm": 0.09814265370368958, "learning_rate": 2.555619542962834e-06, "loss": 0.0067, "step": 9030 }, { "epoch": 46.12244897959184, "grad_norm": 0.11660932749509811, "learning_rate": 2.503693209498409e-06, "loss": 0.0048, "step": 9040 }, { "epoch": 46.173469387755105, "grad_norm": 0.12881217896938324, "learning_rate": 2.452286288035449e-06, "loss": 0.0043, "step": 9050 }, { "epoch": 46.224489795918366, "grad_norm": 0.11460472643375397, "learning_rate": 2.4013993407518363e-06, "loss": 0.0049, "step": 9060 }, { "epoch": 46.275510204081634, "grad_norm": 0.14519673585891724, "learning_rate": 2.351032924139063e-06, "loss": 0.0049, "step": 9070 }, { "epoch": 46.326530612244895, "grad_norm": 0.23098841309547424, "learning_rate": 2.30118758899619e-06, "loss": 0.0048, "step": 9080 }, { "epoch": 46.37755102040816, "grad_norm": 0.17174358665943146, "learning_rate": 2.2518638804238157e-06, "loss": 0.0055, "step": 9090 }, { "epoch": 46.42857142857143, "grad_norm": 0.214328333735466, "learning_rate": 2.203062337818118e-06, "loss": 0.0047, "step": 9100 }, { "epoch": 46.47959183673469, "grad_norm": 0.2349383682012558, "learning_rate": 2.1547834948649483e-06, "loss": 0.0041, "step": 9110 }, { "epoch": 46.53061224489796, "grad_norm": 0.16913291811943054, "learning_rate": 2.1070278795340017e-06, "loss": 0.0057, "step": 9120 }, { "epoch": 46.58163265306123, "grad_norm": 0.3802333176136017, "learning_rate": 2.059796014073029e-06, "loss": 0.0053, "step": 9130 }, { "epoch": 46.63265306122449, "grad_norm": 0.11178506910800934, "learning_rate": 2.01308841500214e-06, "loss": 0.0044, "step": 9140 }, { "epoch": 46.683673469387756, "grad_norm": 0.13485543429851532, "learning_rate": 1.9669055931081704e-06, "loss": 0.0056, "step": 9150 }, { "epoch": 46.734693877551024, "grad_norm": 0.10500293970108032, "learning_rate": 1.9212480534390507e-06, "loss": 0.0039, "step": 9160 }, { "epoch": 46.785714285714285, "grad_norm": 0.14204591512680054, "learning_rate": 1.8761162952983246e-06, "loss": 0.0049, "step": 9170 }, { "epoch": 46.83673469387755, "grad_norm": 0.16562332212924957, "learning_rate": 1.8315108122396618e-06, "loss": 0.0047, "step": 9180 }, { "epoch": 46.88775510204081, "grad_norm": 0.1518019437789917, "learning_rate": 1.787432092061475e-06, "loss": 0.0057, "step": 9190 }, { "epoch": 46.93877551020408, "grad_norm": 0.18393941223621368, "learning_rate": 1.743880616801602e-06, "loss": 0.005, "step": 9200 }, { "epoch": 46.98979591836735, "grad_norm": 0.07821141183376312, "learning_rate": 1.7008568627319865e-06, "loss": 0.0063, "step": 9210 }, { "epoch": 47.04081632653061, "grad_norm": 0.10665757209062576, "learning_rate": 1.6583613003535226e-06, "loss": 0.0046, "step": 9220 }, { "epoch": 47.09183673469388, "grad_norm": 0.11643277108669281, "learning_rate": 1.6163943943908522e-06, "loss": 0.0052, "step": 9230 }, { "epoch": 47.142857142857146, "grad_norm": 0.1593342125415802, "learning_rate": 1.5749566037873476e-06, "loss": 0.0039, "step": 9240 }, { "epoch": 47.19387755102041, "grad_norm": 0.08654476702213287, "learning_rate": 1.5340483817000428e-06, "loss": 0.0052, "step": 9250 }, { "epoch": 47.244897959183675, "grad_norm": 0.13065624237060547, "learning_rate": 1.4936701754947101e-06, "loss": 0.0064, "step": 9260 }, { "epoch": 47.295918367346935, "grad_norm": 0.13620144128799438, "learning_rate": 1.4538224267409361e-06, "loss": 0.0068, "step": 9270 }, { "epoch": 47.3469387755102, "grad_norm": 0.10501694679260254, "learning_rate": 1.414505571207314e-06, "loss": 0.0057, "step": 9280 }, { "epoch": 47.39795918367347, "grad_norm": 0.11845658719539642, "learning_rate": 1.3757200388566816e-06, "loss": 0.0047, "step": 9290 }, { "epoch": 47.44897959183673, "grad_norm": 0.2210085093975067, "learning_rate": 1.3374662538414074e-06, "loss": 0.0054, "step": 9300 }, { "epoch": 47.5, "grad_norm": 0.08250414580106735, "learning_rate": 1.2997446344987617e-06, "loss": 0.0044, "step": 9310 }, { "epoch": 47.55102040816327, "grad_norm": 0.3145185708999634, "learning_rate": 1.262555593346315e-06, "loss": 0.0041, "step": 9320 }, { "epoch": 47.60204081632653, "grad_norm": 0.1367759108543396, "learning_rate": 1.2258995370774685e-06, "loss": 0.0042, "step": 9330 }, { "epoch": 47.6530612244898, "grad_norm": 0.15927579998970032, "learning_rate": 1.1897768665569798e-06, "loss": 0.0081, "step": 9340 }, { "epoch": 47.704081632653065, "grad_norm": 0.11456169188022614, "learning_rate": 1.1541879768165954e-06, "loss": 0.0056, "step": 9350 }, { "epoch": 47.755102040816325, "grad_norm": 0.08145991712808609, "learning_rate": 1.1191332570507085e-06, "loss": 0.0047, "step": 9360 }, { "epoch": 47.80612244897959, "grad_norm": 0.1047838032245636, "learning_rate": 1.0846130906121132e-06, "loss": 0.0047, "step": 9370 }, { "epoch": 47.857142857142854, "grad_norm": 0.08225499093532562, "learning_rate": 1.0506278550078131e-06, "loss": 0.0041, "step": 9380 }, { "epoch": 47.90816326530612, "grad_norm": 0.12090161442756653, "learning_rate": 1.0171779218949185e-06, "loss": 0.0053, "step": 9390 }, { "epoch": 47.95918367346939, "grad_norm": 0.1431620717048645, "learning_rate": 9.842636570765174e-07, "loss": 0.0049, "step": 9400 }, { "epoch": 48.01020408163265, "grad_norm": 0.20520652830600739, "learning_rate": 9.518854204977612e-07, "loss": 0.0046, "step": 9410 }, { "epoch": 48.06122448979592, "grad_norm": 0.42215585708618164, "learning_rate": 9.200435662418349e-07, "loss": 0.0039, "step": 9420 }, { "epoch": 48.11224489795919, "grad_norm": 0.07625243812799454, "learning_rate": 8.887384425261658e-07, "loss": 0.0041, "step": 9430 }, { "epoch": 48.16326530612245, "grad_norm": 0.19119127094745636, "learning_rate": 8.579703916985648e-07, "loss": 0.0043, "step": 9440 }, { "epoch": 48.214285714285715, "grad_norm": 0.141096830368042, "learning_rate": 8.277397502335194e-07, "loss": 0.0049, "step": 9450 }, { "epoch": 48.265306122448976, "grad_norm": 0.25900956988334656, "learning_rate": 7.980468487284675e-07, "loss": 0.0049, "step": 9460 }, { "epoch": 48.316326530612244, "grad_norm": 0.14170551300048828, "learning_rate": 7.688920119002297e-07, "loss": 0.0042, "step": 9470 }, { "epoch": 48.36734693877551, "grad_norm": 0.10030486434698105, "learning_rate": 7.402755585814269e-07, "loss": 0.0042, "step": 9480 }, { "epoch": 48.41836734693877, "grad_norm": 0.061988670378923416, "learning_rate": 7.121978017170073e-07, "loss": 0.004, "step": 9490 }, { "epoch": 48.46938775510204, "grad_norm": 0.21009737253189087, "learning_rate": 6.846590483608306e-07, "loss": 0.0035, "step": 9500 }, { "epoch": 48.52040816326531, "grad_norm": 0.21758700907230377, "learning_rate": 6.576595996722834e-07, "loss": 0.0047, "step": 9510 }, { "epoch": 48.57142857142857, "grad_norm": 0.0872144103050232, "learning_rate": 6.311997509130141e-07, "loss": 0.005, "step": 9520 }, { "epoch": 48.62244897959184, "grad_norm": 0.17584608495235443, "learning_rate": 6.052797914436803e-07, "loss": 0.004, "step": 9530 }, { "epoch": 48.673469387755105, "grad_norm": 0.11404914408922195, "learning_rate": 5.799000047208181e-07, "loss": 0.0046, "step": 9540 }, { "epoch": 48.724489795918366, "grad_norm": 0.23156529664993286, "learning_rate": 5.550606682937054e-07, "loss": 0.004, "step": 9550 }, { "epoch": 48.775510204081634, "grad_norm": 0.06839218735694885, "learning_rate": 5.307620538013481e-07, "loss": 0.0038, "step": 9560 }, { "epoch": 48.826530612244895, "grad_norm": 0.06368494033813477, "learning_rate": 5.070044269694874e-07, "loss": 0.0053, "step": 9570 }, { "epoch": 48.87755102040816, "grad_norm": 0.1766105443239212, "learning_rate": 4.837880476077417e-07, "loss": 0.0049, "step": 9580 }, { "epoch": 48.92857142857143, "grad_norm": 0.0877106636762619, "learning_rate": 4.6111316960670835e-07, "loss": 0.0051, "step": 9590 }, { "epoch": 48.97959183673469, "grad_norm": 0.06350372731685638, "learning_rate": 4.389800409352218e-07, "loss": 0.0048, "step": 9600 }, { "epoch": 49.03061224489796, "grad_norm": 0.06393253058195114, "learning_rate": 4.173889036376277e-07, "loss": 0.0039, "step": 9610 }, { "epoch": 49.08163265306123, "grad_norm": 0.23827730119228363, "learning_rate": 3.963399938311463e-07, "loss": 0.0052, "step": 9620 }, { "epoch": 49.13265306122449, "grad_norm": 0.08909648656845093, "learning_rate": 3.7583354170328545e-07, "loss": 0.0044, "step": 9630 }, { "epoch": 49.183673469387756, "grad_norm": 0.09459241479635239, "learning_rate": 3.558697715093207e-07, "loss": 0.0044, "step": 9640 }, { "epoch": 49.234693877551024, "grad_norm": 0.08130288124084473, "learning_rate": 3.3644890156983576e-07, "loss": 0.0044, "step": 9650 }, { "epoch": 49.285714285714285, "grad_norm": 0.10264245420694351, "learning_rate": 3.175711442683638e-07, "loss": 0.0041, "step": 9660 }, { "epoch": 49.33673469387755, "grad_norm": 0.11238948255777359, "learning_rate": 2.9923670604902197e-07, "loss": 0.0051, "step": 9670 }, { "epoch": 49.38775510204081, "grad_norm": 0.11828438937664032, "learning_rate": 2.814457874143028e-07, "loss": 0.0048, "step": 9680 }, { "epoch": 49.43877551020408, "grad_norm": 0.09683988988399506, "learning_rate": 2.641985829228366e-07, "loss": 0.0041, "step": 9690 }, { "epoch": 49.48979591836735, "grad_norm": 0.2072741538286209, "learning_rate": 2.474952811872877e-07, "loss": 0.0047, "step": 9700 }, { "epoch": 49.54081632653061, "grad_norm": 0.08927561342716217, "learning_rate": 2.3133606487228397e-07, "loss": 0.005, "step": 9710 }, { "epoch": 49.59183673469388, "grad_norm": 0.09228548407554626, "learning_rate": 2.157211106924295e-07, "loss": 0.0041, "step": 9720 }, { "epoch": 49.642857142857146, "grad_norm": 0.07332868129014969, "learning_rate": 2.006505894103672e-07, "loss": 0.004, "step": 9730 }, { "epoch": 49.69387755102041, "grad_norm": 0.1933111995458603, "learning_rate": 1.8612466583489696e-07, "loss": 0.0056, "step": 9740 }, { "epoch": 49.744897959183675, "grad_norm": 0.1359729766845703, "learning_rate": 1.7214349881918834e-07, "loss": 0.0055, "step": 9750 }, { "epoch": 49.795918367346935, "grad_norm": 0.05880662426352501, "learning_rate": 1.5870724125904845e-07, "loss": 0.0048, "step": 9760 }, { "epoch": 49.8469387755102, "grad_norm": 0.10643438249826431, "learning_rate": 1.4581604009124006e-07, "loss": 0.0043, "step": 9770 }, { "epoch": 49.89795918367347, "grad_norm": 0.09929107129573822, "learning_rate": 1.334700362918717e-07, "loss": 0.0067, "step": 9780 }, { "epoch": 49.94897959183673, "grad_norm": 0.09181753545999527, "learning_rate": 1.2166936487486015e-07, "loss": 0.0053, "step": 9790 }, { "epoch": 50.0, "grad_norm": 0.5141326785087585, "learning_rate": 1.1041415489045914e-07, "loss": 0.0042, "step": 9800 }, { "epoch": 50.05102040816327, "grad_norm": 0.09939754009246826, "learning_rate": 9.970452942384412e-08, "loss": 0.0036, "step": 9810 }, { "epoch": 50.10204081632653, "grad_norm": 0.07293524593114853, "learning_rate": 8.954060559375754e-08, "loss": 0.0039, "step": 9820 }, { "epoch": 50.1530612244898, "grad_norm": 0.0818442851305008, "learning_rate": 7.99224945512489e-08, "loss": 0.0041, "step": 9830 }, { "epoch": 50.204081632653065, "grad_norm": 0.2200051248073578, "learning_rate": 7.085030147843675e-08, "loss": 0.0045, "step": 9840 }, { "epoch": 50.255102040816325, "grad_norm": 0.11378786712884903, "learning_rate": 6.232412558736523e-08, "loss": 0.007, "step": 9850 }, { "epoch": 50.30612244897959, "grad_norm": 0.0833803191781044, "learning_rate": 5.434406011893822e-08, "loss": 0.0043, "step": 9860 }, { "epoch": 50.357142857142854, "grad_norm": 0.17778050899505615, "learning_rate": 4.6910192341864664e-08, "loss": 0.0058, "step": 9870 }, { "epoch": 50.40816326530612, "grad_norm": 0.13979117572307587, "learning_rate": 4.0022603551737035e-08, "loss": 0.0048, "step": 9880 }, { "epoch": 50.45918367346939, "grad_norm": 0.1155349388718605, "learning_rate": 3.3681369070120985e-08, "loss": 0.0053, "step": 9890 }, { "epoch": 50.51020408163265, "grad_norm": 0.1303984820842743, "learning_rate": 2.7886558243744866e-08, "loss": 0.0056, "step": 9900 }, { "epoch": 50.56122448979592, "grad_norm": 0.1775132715702057, "learning_rate": 2.2638234443722596e-08, "loss": 0.0043, "step": 9910 }, { "epoch": 50.61224489795919, "grad_norm": 0.11080679297447205, "learning_rate": 1.7936455064887504e-08, "loss": 0.0062, "step": 9920 }, { "epoch": 50.66326530612245, "grad_norm": 0.09285643696784973, "learning_rate": 1.378127152514841e-08, "loss": 0.006, "step": 9930 }, { "epoch": 50.714285714285715, "grad_norm": 0.04935538023710251, "learning_rate": 1.0172729264917857e-08, "loss": 0.0047, "step": 9940 }, { "epoch": 50.765306122448976, "grad_norm": 0.16407077014446259, "learning_rate": 7.1108677466458215e-09, "loss": 0.0049, "step": 9950 }, { "epoch": 50.816326530612244, "grad_norm": 0.1501726508140564, "learning_rate": 4.595720454353414e-09, "loss": 0.0044, "step": 9960 }, { "epoch": 50.86734693877551, "grad_norm": 0.10269220918416977, "learning_rate": 2.627314893294264e-09, "loss": 0.0039, "step": 9970 }, { "epoch": 50.91836734693877, "grad_norm": 0.1023826152086258, "learning_rate": 1.2056725896270048e-09, "loss": 0.0048, "step": 9980 }, { "epoch": 50.96938775510204, "grad_norm": 0.0942123755812645, "learning_rate": 3.308090902098826e-10, "loss": 0.0043, "step": 9990 }, { "epoch": 51.02040816326531, "grad_norm": 0.06581837683916092, "learning_rate": 2.7339624120159555e-12, "loss": 0.0047, "step": 10000 }, { "epoch": 51.02040816326531, "step": 10000, "total_flos": 0.0, "train_loss": 0.014424933608621359, "train_runtime": 26667.941, "train_samples_per_second": 95.995, "train_steps_per_second": 0.375 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 52, "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": 64, "trial_name": null, "trial_params": null }