{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 9.191176470588236, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.009191176470588236, "grad_norm": 0.9019544720649719, "learning_rate": 1.8e-06, "loss": 1.3307, "step": 10 }, { "epoch": 0.01838235294117647, "grad_norm": 0.7220153212547302, "learning_rate": 3.8e-06, "loss": 1.2437, "step": 20 }, { "epoch": 0.027573529411764705, "grad_norm": 0.6808254718780518, "learning_rate": 5.8e-06, "loss": 1.2337, "step": 30 }, { "epoch": 0.03676470588235294, "grad_norm": 0.5643627047538757, "learning_rate": 7.8e-06, "loss": 1.2192, "step": 40 }, { "epoch": 0.04595588235294118, "grad_norm": 0.6695625185966492, "learning_rate": 9.800000000000001e-06, "loss": 1.2177, "step": 50 }, { "epoch": 0.05514705882352941, "grad_norm": 0.9542908072471619, "learning_rate": 1.18e-05, "loss": 1.2157, "step": 60 }, { "epoch": 0.06433823529411764, "grad_norm": 1.3575010299682617, "learning_rate": 1.3800000000000002e-05, "loss": 1.1989, "step": 70 }, { "epoch": 0.07352941176470588, "grad_norm": 1.4558545351028442, "learning_rate": 1.58e-05, "loss": 1.1998, "step": 80 }, { "epoch": 0.08272058823529412, "grad_norm": 0.5149608850479126, "learning_rate": 1.78e-05, "loss": 1.1992, "step": 90 }, { "epoch": 0.09191176470588236, "grad_norm": 0.5245566368103027, "learning_rate": 1.9800000000000004e-05, "loss": 1.1909, "step": 100 }, { "epoch": 0.10110294117647059, "grad_norm": 0.5645520091056824, "learning_rate": 2.18e-05, "loss": 1.1816, "step": 110 }, { "epoch": 0.11029411764705882, "grad_norm": 0.7525063753128052, "learning_rate": 2.38e-05, "loss": 1.0925, "step": 120 }, { "epoch": 0.11948529411764706, "grad_norm": 2.5766584873199463, "learning_rate": 2.58e-05, "loss": 0.9617, "step": 130 }, { "epoch": 0.12867647058823528, "grad_norm": 2.395406484603882, "learning_rate": 2.7800000000000005e-05, "loss": 0.8142, "step": 140 }, { "epoch": 0.13786764705882354, "grad_norm": 3.3057961463928223, "learning_rate": 2.98e-05, "loss": 0.6805, "step": 150 }, { "epoch": 0.14705882352941177, "grad_norm": 1.9755018949508667, "learning_rate": 3.18e-05, "loss": 0.5546, "step": 160 }, { "epoch": 0.15625, "grad_norm": 2.6880404949188232, "learning_rate": 3.38e-05, "loss": 0.4259, "step": 170 }, { "epoch": 0.16544117647058823, "grad_norm": 1.927709937095642, "learning_rate": 3.58e-05, "loss": 0.3554, "step": 180 }, { "epoch": 0.17463235294117646, "grad_norm": 1.318535327911377, "learning_rate": 3.7800000000000004e-05, "loss": 0.2731, "step": 190 }, { "epoch": 0.18382352941176472, "grad_norm": 4.075316429138184, "learning_rate": 3.9800000000000005e-05, "loss": 0.2609, "step": 200 }, { "epoch": 0.19301470588235295, "grad_norm": 1.0927462577819824, "learning_rate": 4.18e-05, "loss": 0.2227, "step": 210 }, { "epoch": 0.20220588235294118, "grad_norm": 1.0717206001281738, "learning_rate": 4.38e-05, "loss": 0.1973, "step": 220 }, { "epoch": 0.2113970588235294, "grad_norm": 1.5133775472640991, "learning_rate": 4.58e-05, "loss": 0.1788, "step": 230 }, { "epoch": 0.22058823529411764, "grad_norm": 1.4684120416641235, "learning_rate": 4.78e-05, "loss": 0.1639, "step": 240 }, { "epoch": 0.22977941176470587, "grad_norm": 1.1788971424102783, "learning_rate": 4.9800000000000004e-05, "loss": 0.1512, "step": 250 }, { "epoch": 0.23897058823529413, "grad_norm": 1.1925201416015625, "learning_rate": 5.1800000000000005e-05, "loss": 0.1383, "step": 260 }, { "epoch": 0.24816176470588236, "grad_norm": 1.1428803205490112, "learning_rate": 5.380000000000001e-05, "loss": 0.1555, "step": 270 }, { "epoch": 0.25735294117647056, "grad_norm": 1.7445504665374756, "learning_rate": 5.580000000000001e-05, "loss": 0.1332, "step": 280 }, { "epoch": 0.2665441176470588, "grad_norm": 1.5160346031188965, "learning_rate": 5.7799999999999995e-05, "loss": 0.1545, "step": 290 }, { "epoch": 0.2757352941176471, "grad_norm": 1.615963339805603, "learning_rate": 5.9800000000000003e-05, "loss": 0.1299, "step": 300 }, { "epoch": 0.2849264705882353, "grad_norm": 1.1847264766693115, "learning_rate": 6.18e-05, "loss": 0.1403, "step": 310 }, { "epoch": 0.29411764705882354, "grad_norm": 1.2159889936447144, "learning_rate": 6.38e-05, "loss": 0.1197, "step": 320 }, { "epoch": 0.30330882352941174, "grad_norm": 1.4204158782958984, "learning_rate": 6.58e-05, "loss": 0.1188, "step": 330 }, { "epoch": 0.3125, "grad_norm": 1.3493999242782593, "learning_rate": 6.780000000000001e-05, "loss": 0.1111, "step": 340 }, { "epoch": 0.32169117647058826, "grad_norm": 1.2892401218414307, "learning_rate": 6.98e-05, "loss": 0.1024, "step": 350 }, { "epoch": 0.33088235294117646, "grad_norm": 1.270029902458191, "learning_rate": 7.18e-05, "loss": 0.1061, "step": 360 }, { "epoch": 0.3400735294117647, "grad_norm": 1.1763330698013306, "learning_rate": 7.38e-05, "loss": 0.1127, "step": 370 }, { "epoch": 0.3492647058823529, "grad_norm": 1.3401509523391724, "learning_rate": 7.58e-05, "loss": 0.1017, "step": 380 }, { "epoch": 0.3584558823529412, "grad_norm": 1.2815041542053223, "learning_rate": 7.780000000000001e-05, "loss": 0.1061, "step": 390 }, { "epoch": 0.36764705882352944, "grad_norm": 1.5589075088500977, "learning_rate": 7.98e-05, "loss": 0.0993, "step": 400 }, { "epoch": 0.37683823529411764, "grad_norm": 1.6591590642929077, "learning_rate": 8.18e-05, "loss": 0.1061, "step": 410 }, { "epoch": 0.3860294117647059, "grad_norm": 1.2315337657928467, "learning_rate": 8.38e-05, "loss": 0.1001, "step": 420 }, { "epoch": 0.3952205882352941, "grad_norm": 1.168882131576538, "learning_rate": 8.58e-05, "loss": 0.1016, "step": 430 }, { "epoch": 0.40441176470588236, "grad_norm": 1.2543108463287354, "learning_rate": 8.78e-05, "loss": 0.0968, "step": 440 }, { "epoch": 0.41360294117647056, "grad_norm": 1.2993816137313843, "learning_rate": 8.98e-05, "loss": 0.0986, "step": 450 }, { "epoch": 0.4227941176470588, "grad_norm": 1.1530307531356812, "learning_rate": 9.180000000000001e-05, "loss": 0.0975, "step": 460 }, { "epoch": 0.4319852941176471, "grad_norm": 1.2659258842468262, "learning_rate": 9.38e-05, "loss": 0.1016, "step": 470 }, { "epoch": 0.4411764705882353, "grad_norm": 1.0709675550460815, "learning_rate": 9.58e-05, "loss": 0.0969, "step": 480 }, { "epoch": 0.45036764705882354, "grad_norm": 0.8813609480857849, "learning_rate": 9.78e-05, "loss": 0.0911, "step": 490 }, { "epoch": 0.45955882352941174, "grad_norm": 0.8756126761436462, "learning_rate": 9.98e-05, "loss": 0.0882, "step": 500 }, { "epoch": 0.46875, "grad_norm": 1.101743221282959, "learning_rate": 9.9999778549206e-05, "loss": 0.0933, "step": 510 }, { "epoch": 0.47794117647058826, "grad_norm": 1.1696802377700806, "learning_rate": 9.999901304280685e-05, "loss": 0.0842, "step": 520 }, { "epoch": 0.48713235294117646, "grad_norm": 1.209069848060608, "learning_rate": 9.999770075521164e-05, "loss": 0.0889, "step": 530 }, { "epoch": 0.4963235294117647, "grad_norm": 1.0431294441223145, "learning_rate": 9.99958417007713e-05, "loss": 0.0834, "step": 540 }, { "epoch": 0.5055147058823529, "grad_norm": 0.9448124170303345, "learning_rate": 9.999343589981615e-05, "loss": 0.0825, "step": 550 }, { "epoch": 0.5147058823529411, "grad_norm": 1.0365020036697388, "learning_rate": 9.999048337865568e-05, "loss": 0.0766, "step": 560 }, { "epoch": 0.5238970588235294, "grad_norm": 1.275138020515442, "learning_rate": 9.998698416957815e-05, "loss": 0.0756, "step": 570 }, { "epoch": 0.5330882352941176, "grad_norm": 0.8591288924217224, "learning_rate": 9.998293831085037e-05, "loss": 0.0774, "step": 580 }, { "epoch": 0.5422794117647058, "grad_norm": 0.9954205751419067, "learning_rate": 9.997834584671719e-05, "loss": 0.0833, "step": 590 }, { "epoch": 0.5514705882352942, "grad_norm": 1.1502774953842163, "learning_rate": 9.997320682740107e-05, "loss": 0.0833, "step": 600 }, { "epoch": 0.5606617647058824, "grad_norm": 1.0054662227630615, "learning_rate": 9.996752130910149e-05, "loss": 0.089, "step": 610 }, { "epoch": 0.5698529411764706, "grad_norm": 0.8506768345832825, "learning_rate": 9.99612893539944e-05, "loss": 0.0733, "step": 620 }, { "epoch": 0.5790441176470589, "grad_norm": 1.1216528415679932, "learning_rate": 9.995451103023144e-05, "loss": 0.0649, "step": 630 }, { "epoch": 0.5882352941176471, "grad_norm": 0.9082152247428894, "learning_rate": 9.994718641193928e-05, "loss": 0.0686, "step": 640 }, { "epoch": 0.5974264705882353, "grad_norm": 1.013120174407959, "learning_rate": 9.993931557921874e-05, "loss": 0.0766, "step": 650 }, { "epoch": 0.6066176470588235, "grad_norm": 1.0689529180526733, "learning_rate": 9.993089861814402e-05, "loss": 0.074, "step": 660 }, { "epoch": 0.6158088235294118, "grad_norm": 0.8483502864837646, "learning_rate": 9.992193562076166e-05, "loss": 0.0857, "step": 670 }, { "epoch": 0.625, "grad_norm": 0.747599720954895, "learning_rate": 9.991242668508954e-05, "loss": 0.0745, "step": 680 }, { "epoch": 0.6341911764705882, "grad_norm": 0.5499546527862549, "learning_rate": 9.990237191511587e-05, "loss": 0.0673, "step": 690 }, { "epoch": 0.6433823529411765, "grad_norm": 0.6998416185379028, "learning_rate": 9.989177142079802e-05, "loss": 0.0698, "step": 700 }, { "epoch": 0.6525735294117647, "grad_norm": 1.0077720880508423, "learning_rate": 9.988062531806126e-05, "loss": 0.0793, "step": 710 }, { "epoch": 0.6617647058823529, "grad_norm": 0.8169732689857483, "learning_rate": 9.986893372879762e-05, "loss": 0.067, "step": 720 }, { "epoch": 0.6709558823529411, "grad_norm": 0.7328667044639587, "learning_rate": 9.985669678086443e-05, "loss": 0.0679, "step": 730 }, { "epoch": 0.6801470588235294, "grad_norm": 0.8013283014297485, "learning_rate": 9.984391460808298e-05, "loss": 0.0586, "step": 740 }, { "epoch": 0.6893382352941176, "grad_norm": 0.8508779406547546, "learning_rate": 9.983058735023709e-05, "loss": 0.0656, "step": 750 }, { "epoch": 0.6985294117647058, "grad_norm": 0.8518447875976562, "learning_rate": 9.98167151530715e-05, "loss": 0.0573, "step": 760 }, { "epoch": 0.7077205882352942, "grad_norm": 0.6712861657142639, "learning_rate": 9.980229816829034e-05, "loss": 0.0616, "step": 770 }, { "epoch": 0.7169117647058824, "grad_norm": 0.8240413665771484, "learning_rate": 9.978733655355544e-05, "loss": 0.0609, "step": 780 }, { "epoch": 0.7261029411764706, "grad_norm": 0.8067315220832825, "learning_rate": 9.977183047248464e-05, "loss": 0.0609, "step": 790 }, { "epoch": 0.7352941176470589, "grad_norm": 0.7908912301063538, "learning_rate": 9.975578009464992e-05, "loss": 0.0647, "step": 800 }, { "epoch": 0.7444852941176471, "grad_norm": 0.7603898644447327, "learning_rate": 9.97391855955757e-05, "loss": 0.0673, "step": 810 }, { "epoch": 0.7536764705882353, "grad_norm": 0.735567033290863, "learning_rate": 9.972204715673669e-05, "loss": 0.0624, "step": 820 }, { "epoch": 0.7628676470588235, "grad_norm": 0.5817774534225464, "learning_rate": 9.970436496555617e-05, "loss": 0.0616, "step": 830 }, { "epoch": 0.7720588235294118, "grad_norm": 0.7370566725730896, "learning_rate": 9.968613921540373e-05, "loss": 0.0575, "step": 840 }, { "epoch": 0.78125, "grad_norm": 0.7867374420166016, "learning_rate": 9.966737010559326e-05, "loss": 0.0698, "step": 850 }, { "epoch": 0.7904411764705882, "grad_norm": 0.8004860877990723, "learning_rate": 9.964805784138072e-05, "loss": 0.061, "step": 860 }, { "epoch": 0.7996323529411765, "grad_norm": 0.9462243914604187, "learning_rate": 9.962820263396195e-05, "loss": 0.06, "step": 870 }, { "epoch": 0.8088235294117647, "grad_norm": 0.7862832546234131, "learning_rate": 9.960780470047033e-05, "loss": 0.0551, "step": 880 }, { "epoch": 0.8180147058823529, "grad_norm": 0.6072902679443359, "learning_rate": 9.958686426397437e-05, "loss": 0.0527, "step": 890 }, { "epoch": 0.8272058823529411, "grad_norm": 0.7157135009765625, "learning_rate": 9.956538155347534e-05, "loss": 0.0636, "step": 900 }, { "epoch": 0.8363970588235294, "grad_norm": 0.7437492609024048, "learning_rate": 9.95433568039047e-05, "loss": 0.0514, "step": 910 }, { "epoch": 0.8455882352941176, "grad_norm": 0.8708986043930054, "learning_rate": 9.952079025612162e-05, "loss": 0.0589, "step": 920 }, { "epoch": 0.8547794117647058, "grad_norm": 0.7011709213256836, "learning_rate": 9.949768215691022e-05, "loss": 0.0545, "step": 930 }, { "epoch": 0.8639705882352942, "grad_norm": 0.7934194207191467, "learning_rate": 9.9474032758977e-05, "loss": 0.0518, "step": 940 }, { "epoch": 0.8731617647058824, "grad_norm": 0.7175838947296143, "learning_rate": 9.944984232094794e-05, "loss": 0.0628, "step": 950 }, { "epoch": 0.8823529411764706, "grad_norm": 0.6982050538063049, "learning_rate": 9.942511110736584e-05, "loss": 0.0495, "step": 960 }, { "epoch": 0.8915441176470589, "grad_norm": 0.5045246481895447, "learning_rate": 9.939983938868726e-05, "loss": 0.0524, "step": 970 }, { "epoch": 0.9007352941176471, "grad_norm": 0.8361206650733948, "learning_rate": 9.93740274412797e-05, "loss": 0.0587, "step": 980 }, { "epoch": 0.9099264705882353, "grad_norm": 0.7167977690696716, "learning_rate": 9.934767554741846e-05, "loss": 0.0572, "step": 990 }, { "epoch": 0.9191176470588235, "grad_norm": 0.9074218273162842, "learning_rate": 9.932078399528361e-05, "loss": 0.0564, "step": 1000 }, { "epoch": 0.9283088235294118, "grad_norm": 0.6889597177505493, "learning_rate": 9.929335307895689e-05, "loss": 0.0508, "step": 1010 }, { "epoch": 0.9375, "grad_norm": 0.6695316433906555, "learning_rate": 9.926538309841839e-05, "loss": 0.0532, "step": 1020 }, { "epoch": 0.9466911764705882, "grad_norm": 0.6929487586021423, "learning_rate": 9.923687435954334e-05, "loss": 0.0529, "step": 1030 }, { "epoch": 0.9558823529411765, "grad_norm": 0.7124028205871582, "learning_rate": 9.920782717409873e-05, "loss": 0.0562, "step": 1040 }, { "epoch": 0.9650735294117647, "grad_norm": 0.6877460479736328, "learning_rate": 9.917824185973994e-05, "loss": 0.0493, "step": 1050 }, { "epoch": 0.9742647058823529, "grad_norm": 0.5176821947097778, "learning_rate": 9.914811874000723e-05, "loss": 0.0597, "step": 1060 }, { "epoch": 0.9834558823529411, "grad_norm": 0.7132368683815002, "learning_rate": 9.911745814432218e-05, "loss": 0.0542, "step": 1070 }, { "epoch": 0.9926470588235294, "grad_norm": 0.8198514580726624, "learning_rate": 9.90862604079842e-05, "loss": 0.0544, "step": 1080 }, { "epoch": 1.0018382352941178, "grad_norm": 0.75013267993927, "learning_rate": 9.90545258721667e-05, "loss": 0.0435, "step": 1090 }, { "epoch": 1.0110294117647058, "grad_norm": 0.6192613840103149, "learning_rate": 9.90222548839135e-05, "loss": 0.0488, "step": 1100 }, { "epoch": 1.0202205882352942, "grad_norm": 0.6941893696784973, "learning_rate": 9.898944779613495e-05, "loss": 0.0518, "step": 1110 }, { "epoch": 1.0294117647058822, "grad_norm": 0.4875585734844208, "learning_rate": 9.89561049676041e-05, "loss": 0.0505, "step": 1120 }, { "epoch": 1.0386029411764706, "grad_norm": 0.4896945655345917, "learning_rate": 9.89222267629528e-05, "loss": 0.0545, "step": 1130 }, { "epoch": 1.0477941176470589, "grad_norm": 0.5555590987205505, "learning_rate": 9.888781355266763e-05, "loss": 0.0526, "step": 1140 }, { "epoch": 1.056985294117647, "grad_norm": 0.6745964884757996, "learning_rate": 9.885286571308598e-05, "loss": 0.0498, "step": 1150 }, { "epoch": 1.0661764705882353, "grad_norm": 0.6715954542160034, "learning_rate": 9.881738362639182e-05, "loss": 0.047, "step": 1160 }, { "epoch": 1.0753676470588236, "grad_norm": 0.72353196144104, "learning_rate": 9.878136768061154e-05, "loss": 0.0464, "step": 1170 }, { "epoch": 1.0845588235294117, "grad_norm": 0.7114365696907043, "learning_rate": 9.874481826960979e-05, "loss": 0.0462, "step": 1180 }, { "epoch": 1.09375, "grad_norm": 0.9372706413269043, "learning_rate": 9.870773579308503e-05, "loss": 0.0493, "step": 1190 }, { "epoch": 1.1029411764705883, "grad_norm": 0.668539822101593, "learning_rate": 9.867012065656533e-05, "loss": 0.0495, "step": 1200 }, { "epoch": 1.1121323529411764, "grad_norm": 0.5126788020133972, "learning_rate": 9.863197327140376e-05, "loss": 0.0465, "step": 1210 }, { "epoch": 1.1213235294117647, "grad_norm": 0.7585374116897583, "learning_rate": 9.859329405477403e-05, "loss": 0.0414, "step": 1220 }, { "epoch": 1.130514705882353, "grad_norm": 0.676013708114624, "learning_rate": 9.855408342966585e-05, "loss": 0.0459, "step": 1230 }, { "epoch": 1.1397058823529411, "grad_norm": 0.6031401753425598, "learning_rate": 9.851434182488033e-05, "loss": 0.0399, "step": 1240 }, { "epoch": 1.1488970588235294, "grad_norm": 0.7307896614074707, "learning_rate": 9.84740696750253e-05, "loss": 0.0462, "step": 1250 }, { "epoch": 1.1580882352941178, "grad_norm": 0.5204982757568359, "learning_rate": 9.843326742051055e-05, "loss": 0.0395, "step": 1260 }, { "epoch": 1.1672794117647058, "grad_norm": 0.4351061284542084, "learning_rate": 9.839193550754297e-05, "loss": 0.0425, "step": 1270 }, { "epoch": 1.1764705882352942, "grad_norm": 0.6187815070152283, "learning_rate": 9.835007438812177e-05, "loss": 0.0415, "step": 1280 }, { "epoch": 1.1856617647058822, "grad_norm": 0.6585433483123779, "learning_rate": 9.830768452003341e-05, "loss": 0.0431, "step": 1290 }, { "epoch": 1.1948529411764706, "grad_norm": 0.440751850605011, "learning_rate": 9.826476636684671e-05, "loss": 0.0405, "step": 1300 }, { "epoch": 1.2040441176470589, "grad_norm": 0.3778493106365204, "learning_rate": 9.822132039790773e-05, "loss": 0.0381, "step": 1310 }, { "epoch": 1.213235294117647, "grad_norm": 0.6102190613746643, "learning_rate": 9.817734708833461e-05, "loss": 0.041, "step": 1320 }, { "epoch": 1.2224264705882353, "grad_norm": 0.6824349761009216, "learning_rate": 9.813284691901243e-05, "loss": 0.0366, "step": 1330 }, { "epoch": 1.2316176470588236, "grad_norm": 0.5893781185150146, "learning_rate": 9.808782037658792e-05, "loss": 0.0399, "step": 1340 }, { "epoch": 1.2408088235294117, "grad_norm": 0.6804303526878357, "learning_rate": 9.804226795346411e-05, "loss": 0.0342, "step": 1350 }, { "epoch": 1.25, "grad_norm": 0.6493552327156067, "learning_rate": 9.799619014779503e-05, "loss": 0.0367, "step": 1360 }, { "epoch": 1.2591911764705883, "grad_norm": 0.6825845241546631, "learning_rate": 9.794958746348013e-05, "loss": 0.0308, "step": 1370 }, { "epoch": 1.2683823529411764, "grad_norm": 0.6365014910697937, "learning_rate": 9.790246041015896e-05, "loss": 0.0366, "step": 1380 }, { "epoch": 1.2775735294117647, "grad_norm": 0.507903516292572, "learning_rate": 9.785480950320538e-05, "loss": 0.0329, "step": 1390 }, { "epoch": 1.2867647058823528, "grad_norm": 0.5516229271888733, "learning_rate": 9.78066352637221e-05, "loss": 0.0383, "step": 1400 }, { "epoch": 1.2959558823529411, "grad_norm": 0.6448059678077698, "learning_rate": 9.775793821853488e-05, "loss": 0.0305, "step": 1410 }, { "epoch": 1.3051470588235294, "grad_norm": 0.4839947819709778, "learning_rate": 9.77087189001868e-05, "loss": 0.0371, "step": 1420 }, { "epoch": 1.3143382352941178, "grad_norm": 0.42556673288345337, "learning_rate": 9.765897784693243e-05, "loss": 0.0332, "step": 1430 }, { "epoch": 1.3235294117647058, "grad_norm": 0.530942440032959, "learning_rate": 9.760871560273197e-05, "loss": 0.0323, "step": 1440 }, { "epoch": 1.3327205882352942, "grad_norm": 0.536531388759613, "learning_rate": 9.755793271724526e-05, "loss": 0.0349, "step": 1450 }, { "epoch": 1.3419117647058822, "grad_norm": 0.6450103521347046, "learning_rate": 9.750662974582584e-05, "loss": 0.0335, "step": 1460 }, { "epoch": 1.3511029411764706, "grad_norm": 0.7727478742599487, "learning_rate": 9.745480724951473e-05, "loss": 0.0348, "step": 1470 }, { "epoch": 1.3602941176470589, "grad_norm": 0.6717798709869385, "learning_rate": 9.740246579503447e-05, "loss": 0.0392, "step": 1480 }, { "epoch": 1.3694852941176472, "grad_norm": 0.6734787821769714, "learning_rate": 9.734960595478284e-05, "loss": 0.0313, "step": 1490 }, { "epoch": 1.3786764705882353, "grad_norm": 0.6354987621307373, "learning_rate": 9.729622830682657e-05, "loss": 0.0417, "step": 1500 }, { "epoch": 1.3878676470588236, "grad_norm": 0.6215519309043884, "learning_rate": 9.724233343489504e-05, "loss": 0.0341, "step": 1510 }, { "epoch": 1.3970588235294117, "grad_norm": 0.5931591987609863, "learning_rate": 9.718792192837396e-05, "loss": 0.0339, "step": 1520 }, { "epoch": 1.40625, "grad_norm": 0.586018443107605, "learning_rate": 9.713299438229886e-05, "loss": 0.0337, "step": 1530 }, { "epoch": 1.4154411764705883, "grad_norm": 0.7602446675300598, "learning_rate": 9.707755139734855e-05, "loss": 0.0378, "step": 1540 }, { "epoch": 1.4246323529411764, "grad_norm": 0.45805832743644714, "learning_rate": 9.702159357983866e-05, "loss": 0.0361, "step": 1550 }, { "epoch": 1.4338235294117647, "grad_norm": 0.6129457354545593, "learning_rate": 9.696512154171492e-05, "loss": 0.0342, "step": 1560 }, { "epoch": 1.4430147058823528, "grad_norm": 0.635168731212616, "learning_rate": 9.690813590054645e-05, "loss": 0.0323, "step": 1570 }, { "epoch": 1.4522058823529411, "grad_norm": 0.5320306420326233, "learning_rate": 9.685063727951914e-05, "loss": 0.0318, "step": 1580 }, { "epoch": 1.4613970588235294, "grad_norm": 0.4128386974334717, "learning_rate": 9.679262630742865e-05, "loss": 0.0335, "step": 1590 }, { "epoch": 1.4705882352941178, "grad_norm": 0.6510125994682312, "learning_rate": 9.673410361867373e-05, "loss": 0.0317, "step": 1600 }, { "epoch": 1.4797794117647058, "grad_norm": 0.6175063252449036, "learning_rate": 9.667506985324909e-05, "loss": 0.0281, "step": 1610 }, { "epoch": 1.4889705882352942, "grad_norm": 0.5576162338256836, "learning_rate": 9.661552565673855e-05, "loss": 0.0276, "step": 1620 }, { "epoch": 1.4981617647058822, "grad_norm": 0.5612936019897461, "learning_rate": 9.655547168030789e-05, "loss": 0.0277, "step": 1630 }, { "epoch": 1.5073529411764706, "grad_norm": 0.7814056873321533, "learning_rate": 9.649490858069777e-05, "loss": 0.0313, "step": 1640 }, { "epoch": 1.5165441176470589, "grad_norm": 0.6130262613296509, "learning_rate": 9.643383702021658e-05, "loss": 0.0344, "step": 1650 }, { "epoch": 1.5257352941176472, "grad_norm": 0.6517163515090942, "learning_rate": 9.637225766673307e-05, "loss": 0.0247, "step": 1660 }, { "epoch": 1.5349264705882353, "grad_norm": 0.5730800628662109, "learning_rate": 9.631017119366922e-05, "loss": 0.0299, "step": 1670 }, { "epoch": 1.5441176470588234, "grad_norm": 0.6556923389434814, "learning_rate": 9.624757827999273e-05, "loss": 0.0322, "step": 1680 }, { "epoch": 1.5533088235294117, "grad_norm": 0.56232750415802, "learning_rate": 9.618447961020971e-05, "loss": 0.0332, "step": 1690 }, { "epoch": 1.5625, "grad_norm": 0.6679795980453491, "learning_rate": 9.612087587435707e-05, "loss": 0.0295, "step": 1700 }, { "epoch": 1.5716911764705883, "grad_norm": 0.6611357927322388, "learning_rate": 9.605676776799508e-05, "loss": 0.0304, "step": 1710 }, { "epoch": 1.5808823529411766, "grad_norm": 0.610569417476654, "learning_rate": 9.599215599219973e-05, "loss": 0.0331, "step": 1720 }, { "epoch": 1.5900735294117647, "grad_norm": 0.5791309475898743, "learning_rate": 9.592704125355505e-05, "loss": 0.0323, "step": 1730 }, { "epoch": 1.5992647058823528, "grad_norm": 0.5949656367301941, "learning_rate": 9.586142426414538e-05, "loss": 0.0273, "step": 1740 }, { "epoch": 1.6084558823529411, "grad_norm": 0.5446546077728271, "learning_rate": 9.57953057415476e-05, "loss": 0.0312, "step": 1750 }, { "epoch": 1.6176470588235294, "grad_norm": 0.6060889959335327, "learning_rate": 9.572868640882328e-05, "loss": 0.0338, "step": 1760 }, { "epoch": 1.6268382352941178, "grad_norm": 0.49856460094451904, "learning_rate": 9.56615669945108e-05, "loss": 0.0302, "step": 1770 }, { "epoch": 1.6360294117647058, "grad_norm": 0.4326664209365845, "learning_rate": 9.55939482326173e-05, "loss": 0.0274, "step": 1780 }, { "epoch": 1.6452205882352942, "grad_norm": 0.556348979473114, "learning_rate": 9.552583086261069e-05, "loss": 0.0343, "step": 1790 }, { "epoch": 1.6544117647058822, "grad_norm": 0.4957214295864105, "learning_rate": 9.545721562941168e-05, "loss": 0.0293, "step": 1800 }, { "epoch": 1.6636029411764706, "grad_norm": 0.5637266039848328, "learning_rate": 9.538810328338543e-05, "loss": 0.0295, "step": 1810 }, { "epoch": 1.6727941176470589, "grad_norm": 0.4914757013320923, "learning_rate": 9.531849458033349e-05, "loss": 0.0252, "step": 1820 }, { "epoch": 1.6819852941176472, "grad_norm": 0.5111973285675049, "learning_rate": 9.524839028148547e-05, "loss": 0.0263, "step": 1830 }, { "epoch": 1.6911764705882353, "grad_norm": 0.7454209327697754, "learning_rate": 9.517779115349077e-05, "loss": 0.027, "step": 1840 }, { "epoch": 1.7003676470588234, "grad_norm": 0.7003049254417419, "learning_rate": 9.510669796841014e-05, "loss": 0.0276, "step": 1850 }, { "epoch": 1.7095588235294117, "grad_norm": 0.5316200256347656, "learning_rate": 9.503511150370727e-05, "loss": 0.032, "step": 1860 }, { "epoch": 1.71875, "grad_norm": 0.4544483423233032, "learning_rate": 9.496303254224024e-05, "loss": 0.0321, "step": 1870 }, { "epoch": 1.7279411764705883, "grad_norm": 0.5831987857818604, "learning_rate": 9.489046187225306e-05, "loss": 0.0263, "step": 1880 }, { "epoch": 1.7371323529411766, "grad_norm": 0.45511719584465027, "learning_rate": 9.481740028736692e-05, "loss": 0.0269, "step": 1890 }, { "epoch": 1.7463235294117647, "grad_norm": 0.6097242832183838, "learning_rate": 9.474384858657164e-05, "loss": 0.0345, "step": 1900 }, { "epoch": 1.7555147058823528, "grad_norm": 0.5893768668174744, "learning_rate": 9.466980757421679e-05, "loss": 0.0312, "step": 1910 }, { "epoch": 1.7647058823529411, "grad_norm": 0.7778833508491516, "learning_rate": 9.459527806000305e-05, "loss": 0.0277, "step": 1920 }, { "epoch": 1.7738970588235294, "grad_norm": 0.3030601739883423, "learning_rate": 9.452026085897325e-05, "loss": 0.0287, "step": 1930 }, { "epoch": 1.7830882352941178, "grad_norm": 0.540002167224884, "learning_rate": 9.444475679150348e-05, "loss": 0.0271, "step": 1940 }, { "epoch": 1.7922794117647058, "grad_norm": 0.4380544424057007, "learning_rate": 9.436876668329411e-05, "loss": 0.0277, "step": 1950 }, { "epoch": 1.8014705882352942, "grad_norm": 0.5024845600128174, "learning_rate": 9.429229136536079e-05, "loss": 0.0224, "step": 1960 }, { "epoch": 1.8106617647058822, "grad_norm": 0.4664299786090851, "learning_rate": 9.421533167402534e-05, "loss": 0.0286, "step": 1970 }, { "epoch": 1.8198529411764706, "grad_norm": 0.5278037190437317, "learning_rate": 9.413788845090666e-05, "loss": 0.0345, "step": 1980 }, { "epoch": 1.8290441176470589, "grad_norm": 0.49221664667129517, "learning_rate": 9.405996254291136e-05, "loss": 0.0254, "step": 1990 }, { "epoch": 1.8382352941176472, "grad_norm": 0.39088916778564453, "learning_rate": 9.398155480222474e-05, "loss": 0.0229, "step": 2000 }, { "epoch": 1.8474264705882353, "grad_norm": 0.37824904918670654, "learning_rate": 9.390266608630128e-05, "loss": 0.0226, "step": 2010 }, { "epoch": 1.8566176470588234, "grad_norm": 0.5605826377868652, "learning_rate": 9.38232972578553e-05, "loss": 0.0271, "step": 2020 }, { "epoch": 1.8658088235294117, "grad_norm": 0.588459312915802, "learning_rate": 9.374344918485164e-05, "loss": 0.0281, "step": 2030 }, { "epoch": 1.875, "grad_norm": 0.493164598941803, "learning_rate": 9.366312274049602e-05, "loss": 0.0298, "step": 2040 }, { "epoch": 1.8841911764705883, "grad_norm": 0.6307514905929565, "learning_rate": 9.358231880322554e-05, "loss": 0.0255, "step": 2050 }, { "epoch": 1.8933823529411766, "grad_norm": 0.577612042427063, "learning_rate": 9.350103825669916e-05, "loss": 0.0226, "step": 2060 }, { "epoch": 1.9025735294117647, "grad_norm": 0.38431042432785034, "learning_rate": 9.341928198978787e-05, "loss": 0.0236, "step": 2070 }, { "epoch": 1.9117647058823528, "grad_norm": 0.5789857506752014, "learning_rate": 9.333705089656512e-05, "loss": 0.0226, "step": 2080 }, { "epoch": 1.9209558823529411, "grad_norm": 0.46185600757598877, "learning_rate": 9.325434587629698e-05, "loss": 0.0243, "step": 2090 }, { "epoch": 1.9301470588235294, "grad_norm": 0.6314039826393127, "learning_rate": 9.31711678334323e-05, "loss": 0.0234, "step": 2100 }, { "epoch": 1.9393382352941178, "grad_norm": 0.6181092858314514, "learning_rate": 9.308751767759282e-05, "loss": 0.0233, "step": 2110 }, { "epoch": 1.9485294117647058, "grad_norm": 0.5274204611778259, "learning_rate": 9.300339632356325e-05, "loss": 0.0215, "step": 2120 }, { "epoch": 1.9577205882352942, "grad_norm": 0.6873094439506531, "learning_rate": 9.291880469128124e-05, "loss": 0.0239, "step": 2130 }, { "epoch": 1.9669117647058822, "grad_norm": 0.5622841715812683, "learning_rate": 9.283374370582732e-05, "loss": 0.0243, "step": 2140 }, { "epoch": 1.9761029411764706, "grad_norm": 0.45631369948387146, "learning_rate": 9.274821429741482e-05, "loss": 0.0288, "step": 2150 }, { "epoch": 1.9852941176470589, "grad_norm": 0.4745766222476959, "learning_rate": 9.266221740137961e-05, "loss": 0.024, "step": 2160 }, { "epoch": 1.9944852941176472, "grad_norm": 0.5218350887298584, "learning_rate": 9.257575395817001e-05, "loss": 0.0226, "step": 2170 }, { "epoch": 2.0036764705882355, "grad_norm": 0.8137276768684387, "learning_rate": 9.248882491333637e-05, "loss": 0.0273, "step": 2180 }, { "epoch": 2.0128676470588234, "grad_norm": 0.4195311963558197, "learning_rate": 9.240143121752076e-05, "loss": 0.0229, "step": 2190 }, { "epoch": 2.0220588235294117, "grad_norm": 0.33945387601852417, "learning_rate": 9.23135738264467e-05, "loss": 0.0237, "step": 2200 }, { "epoch": 2.03125, "grad_norm": 0.5028133392333984, "learning_rate": 9.222525370090849e-05, "loss": 0.0216, "step": 2210 }, { "epoch": 2.0404411764705883, "grad_norm": 0.5451264381408691, "learning_rate": 9.213647180676088e-05, "loss": 0.0214, "step": 2220 }, { "epoch": 2.0496323529411766, "grad_norm": 0.4600425362586975, "learning_rate": 9.204722911490846e-05, "loss": 0.0244, "step": 2230 }, { "epoch": 2.0588235294117645, "grad_norm": 0.4658612608909607, "learning_rate": 9.1957526601295e-05, "loss": 0.0208, "step": 2240 }, { "epoch": 2.068014705882353, "grad_norm": 0.4521946310997009, "learning_rate": 9.186736524689281e-05, "loss": 0.0252, "step": 2250 }, { "epoch": 2.077205882352941, "grad_norm": 0.3975313901901245, "learning_rate": 9.177674603769204e-05, "loss": 0.0234, "step": 2260 }, { "epoch": 2.0863970588235294, "grad_norm": 0.4542566239833832, "learning_rate": 9.168566996468983e-05, "loss": 0.0208, "step": 2270 }, { "epoch": 2.0955882352941178, "grad_norm": 0.4895295798778534, "learning_rate": 9.159413802387951e-05, "loss": 0.0254, "step": 2280 }, { "epoch": 2.104779411764706, "grad_norm": 0.5705728530883789, "learning_rate": 9.150215121623974e-05, "loss": 0.0228, "step": 2290 }, { "epoch": 2.113970588235294, "grad_norm": 0.48782315850257874, "learning_rate": 9.140971054772349e-05, "loss": 0.02, "step": 2300 }, { "epoch": 2.1231617647058822, "grad_norm": 0.4150049388408661, "learning_rate": 9.131681702924713e-05, "loss": 0.0283, "step": 2310 }, { "epoch": 2.1323529411764706, "grad_norm": 0.4983762204647064, "learning_rate": 9.122347167667926e-05, "loss": 0.0202, "step": 2320 }, { "epoch": 2.141544117647059, "grad_norm": 0.48840656876564026, "learning_rate": 9.112967551082973e-05, "loss": 0.0298, "step": 2330 }, { "epoch": 2.150735294117647, "grad_norm": 0.764106810092926, "learning_rate": 9.103542955743835e-05, "loss": 0.0218, "step": 2340 }, { "epoch": 2.1599264705882355, "grad_norm": 0.5620117783546448, "learning_rate": 9.094073484716381e-05, "loss": 0.0273, "step": 2350 }, { "epoch": 2.1691176470588234, "grad_norm": 0.5343868136405945, "learning_rate": 9.084559241557226e-05, "loss": 0.0222, "step": 2360 }, { "epoch": 2.1783088235294117, "grad_norm": 0.503672182559967, "learning_rate": 9.075000330312608e-05, "loss": 0.0233, "step": 2370 }, { "epoch": 2.1875, "grad_norm": 0.487435907125473, "learning_rate": 9.065396855517253e-05, "loss": 0.0211, "step": 2380 }, { "epoch": 2.1966911764705883, "grad_norm": 0.28125518560409546, "learning_rate": 9.055748922193219e-05, "loss": 0.0251, "step": 2390 }, { "epoch": 2.2058823529411766, "grad_norm": 0.6199022531509399, "learning_rate": 9.046056635848761e-05, "loss": 0.0251, "step": 2400 }, { "epoch": 2.2150735294117645, "grad_norm": 0.5764445662498474, "learning_rate": 9.036320102477169e-05, "loss": 0.022, "step": 2410 }, { "epoch": 2.224264705882353, "grad_norm": 0.5314090847969055, "learning_rate": 9.02653942855561e-05, "loss": 0.023, "step": 2420 }, { "epoch": 2.233455882352941, "grad_norm": 0.6271440982818604, "learning_rate": 9.016714721043971e-05, "loss": 0.0209, "step": 2430 }, { "epoch": 2.2426470588235294, "grad_norm": 0.5442946553230286, "learning_rate": 9.006846087383675e-05, "loss": 0.024, "step": 2440 }, { "epoch": 2.2518382352941178, "grad_norm": 0.48201119899749756, "learning_rate": 8.996933635496523e-05, "loss": 0.0215, "step": 2450 }, { "epoch": 2.261029411764706, "grad_norm": 0.3904911279678345, "learning_rate": 8.986977473783498e-05, "loss": 0.0242, "step": 2460 }, { "epoch": 2.270220588235294, "grad_norm": 0.4735877811908722, "learning_rate": 8.97697771112359e-05, "loss": 0.0215, "step": 2470 }, { "epoch": 2.2794117647058822, "grad_norm": 0.5196784734725952, "learning_rate": 8.966934456872602e-05, "loss": 0.022, "step": 2480 }, { "epoch": 2.2886029411764706, "grad_norm": 0.4864409565925598, "learning_rate": 8.95684782086195e-05, "loss": 0.0185, "step": 2490 }, { "epoch": 2.297794117647059, "grad_norm": 0.4121498763561249, "learning_rate": 8.946717913397476e-05, "loss": 0.025, "step": 2500 }, { "epoch": 2.306985294117647, "grad_norm": 0.43099141120910645, "learning_rate": 8.93654484525822e-05, "loss": 0.0222, "step": 2510 }, { "epoch": 2.3161764705882355, "grad_norm": 0.46026450395584106, "learning_rate": 8.926328727695226e-05, "loss": 0.0209, "step": 2520 }, { "epoch": 2.3253676470588234, "grad_norm": 0.5679813027381897, "learning_rate": 8.916069672430319e-05, "loss": 0.0191, "step": 2530 }, { "epoch": 2.3345588235294117, "grad_norm": 0.5221530199050903, "learning_rate": 8.905767791654884e-05, "loss": 0.0228, "step": 2540 }, { "epoch": 2.34375, "grad_norm": 0.5912122130393982, "learning_rate": 8.895423198028638e-05, "loss": 0.0274, "step": 2550 }, { "epoch": 2.3529411764705883, "grad_norm": 0.5581372976303101, "learning_rate": 8.885036004678402e-05, "loss": 0.0235, "step": 2560 }, { "epoch": 2.3621323529411766, "grad_norm": 0.4881454408168793, "learning_rate": 8.874606325196857e-05, "loss": 0.0216, "step": 2570 }, { "epoch": 2.3713235294117645, "grad_norm": 0.571555495262146, "learning_rate": 8.864134273641304e-05, "loss": 0.0235, "step": 2580 }, { "epoch": 2.380514705882353, "grad_norm": 0.4265335202217102, "learning_rate": 8.853619964532427e-05, "loss": 0.0221, "step": 2590 }, { "epoch": 2.389705882352941, "grad_norm": 0.5070019364356995, "learning_rate": 8.843063512853019e-05, "loss": 0.0205, "step": 2600 }, { "epoch": 2.3988970588235294, "grad_norm": 0.39745596051216125, "learning_rate": 8.832465034046749e-05, "loss": 0.0199, "step": 2610 }, { "epoch": 2.4080882352941178, "grad_norm": 0.39655235409736633, "learning_rate": 8.821824644016882e-05, "loss": 0.0211, "step": 2620 }, { "epoch": 2.417279411764706, "grad_norm": 0.581464409828186, "learning_rate": 8.811142459125019e-05, "loss": 0.0187, "step": 2630 }, { "epoch": 2.426470588235294, "grad_norm": 0.528378963470459, "learning_rate": 8.800418596189822e-05, "loss": 0.0178, "step": 2640 }, { "epoch": 2.4356617647058822, "grad_norm": 0.4996073544025421, "learning_rate": 8.789653172485737e-05, "loss": 0.0224, "step": 2650 }, { "epoch": 2.4448529411764706, "grad_norm": 0.4523492455482483, "learning_rate": 8.778846305741715e-05, "loss": 0.0276, "step": 2660 }, { "epoch": 2.454044117647059, "grad_norm": 0.3444526493549347, "learning_rate": 8.767998114139918e-05, "loss": 0.0225, "step": 2670 }, { "epoch": 2.463235294117647, "grad_norm": 0.42765453457832336, "learning_rate": 8.757108716314429e-05, "loss": 0.0251, "step": 2680 }, { "epoch": 2.4724264705882355, "grad_norm": 0.5003244876861572, "learning_rate": 8.746178231349962e-05, "loss": 0.0196, "step": 2690 }, { "epoch": 2.4816176470588234, "grad_norm": 0.349907249212265, "learning_rate": 8.735206778780549e-05, "loss": 0.0239, "step": 2700 }, { "epoch": 2.4908088235294117, "grad_norm": 0.42867523431777954, "learning_rate": 8.724194478588234e-05, "loss": 0.0219, "step": 2710 }, { "epoch": 2.5, "grad_norm": 0.4790487289428711, "learning_rate": 8.713141451201772e-05, "loss": 0.0207, "step": 2720 }, { "epoch": 2.5091911764705883, "grad_norm": 0.5373687148094177, "learning_rate": 8.702047817495295e-05, "loss": 0.0204, "step": 2730 }, { "epoch": 2.5183823529411766, "grad_norm": 0.5957093238830566, "learning_rate": 8.69091369878701e-05, "loss": 0.0213, "step": 2740 }, { "epoch": 2.5275735294117645, "grad_norm": 0.4312717020511627, "learning_rate": 8.679739216837849e-05, "loss": 0.0177, "step": 2750 }, { "epoch": 2.536764705882353, "grad_norm": 0.476746141910553, "learning_rate": 8.66852449385016e-05, "loss": 0.0195, "step": 2760 }, { "epoch": 2.545955882352941, "grad_norm": 0.37591588497161865, "learning_rate": 8.657269652466356e-05, "loss": 0.0197, "step": 2770 }, { "epoch": 2.5551470588235294, "grad_norm": 0.38740065693855286, "learning_rate": 8.645974815767577e-05, "loss": 0.0206, "step": 2780 }, { "epoch": 2.5643382352941178, "grad_norm": 0.4049248993396759, "learning_rate": 8.634640107272351e-05, "loss": 0.0187, "step": 2790 }, { "epoch": 2.5735294117647056, "grad_norm": 0.4691837430000305, "learning_rate": 8.623265650935234e-05, "loss": 0.0202, "step": 2800 }, { "epoch": 2.5827205882352944, "grad_norm": 0.4181932210922241, "learning_rate": 8.611851571145456e-05, "loss": 0.0214, "step": 2810 }, { "epoch": 2.5919117647058822, "grad_norm": 0.5595068335533142, "learning_rate": 8.600397992725566e-05, "loss": 0.0192, "step": 2820 }, { "epoch": 2.6011029411764706, "grad_norm": 0.494282990694046, "learning_rate": 8.588905040930061e-05, "loss": 0.0193, "step": 2830 }, { "epoch": 2.610294117647059, "grad_norm": 0.45018383860588074, "learning_rate": 8.577372841444022e-05, "loss": 0.0208, "step": 2840 }, { "epoch": 2.619485294117647, "grad_norm": 0.5019036531448364, "learning_rate": 8.565801520381736e-05, "loss": 0.0197, "step": 2850 }, { "epoch": 2.6286764705882355, "grad_norm": 0.4007212817668915, "learning_rate": 8.554191204285313e-05, "loss": 0.0196, "step": 2860 }, { "epoch": 2.6378676470588234, "grad_norm": 0.471796452999115, "learning_rate": 8.542542020123315e-05, "loss": 0.0191, "step": 2870 }, { "epoch": 2.6470588235294117, "grad_norm": 0.4328274130821228, "learning_rate": 8.530854095289347e-05, "loss": 0.0192, "step": 2880 }, { "epoch": 2.65625, "grad_norm": 0.4316680431365967, "learning_rate": 8.519127557600688e-05, "loss": 0.0191, "step": 2890 }, { "epoch": 2.6654411764705883, "grad_norm": 0.43830040097236633, "learning_rate": 8.507362535296871e-05, "loss": 0.0195, "step": 2900 }, { "epoch": 2.6746323529411766, "grad_norm": 0.5127837657928467, "learning_rate": 8.495559157038299e-05, "loss": 0.0276, "step": 2910 }, { "epoch": 2.6838235294117645, "grad_norm": 0.49351948499679565, "learning_rate": 8.483717551904823e-05, "loss": 0.0196, "step": 2920 }, { "epoch": 2.693014705882353, "grad_norm": 0.4956419765949249, "learning_rate": 8.47183784939434e-05, "loss": 0.019, "step": 2930 }, { "epoch": 2.702205882352941, "grad_norm": 0.46160534024238586, "learning_rate": 8.459920179421374e-05, "loss": 0.0205, "step": 2940 }, { "epoch": 2.7113970588235294, "grad_norm": 0.38040125370025635, "learning_rate": 8.447964672315656e-05, "loss": 0.0188, "step": 2950 }, { "epoch": 2.7205882352941178, "grad_norm": 0.4619915187358856, "learning_rate": 8.435971458820692e-05, "loss": 0.0228, "step": 2960 }, { "epoch": 2.7297794117647056, "grad_norm": 0.42629915475845337, "learning_rate": 8.423940670092345e-05, "loss": 0.022, "step": 2970 }, { "epoch": 2.7389705882352944, "grad_norm": 0.5419662594795227, "learning_rate": 8.411872437697394e-05, "loss": 0.0187, "step": 2980 }, { "epoch": 2.7481617647058822, "grad_norm": 0.3723689019680023, "learning_rate": 8.399766893612096e-05, "loss": 0.0168, "step": 2990 }, { "epoch": 2.7573529411764706, "grad_norm": 0.4072543978691101, "learning_rate": 8.38762417022074e-05, "loss": 0.0211, "step": 3000 }, { "epoch": 2.766544117647059, "grad_norm": 0.4469057023525238, "learning_rate": 8.375444400314204e-05, "loss": 0.0183, "step": 3010 }, { "epoch": 2.775735294117647, "grad_norm": 0.38040122389793396, "learning_rate": 8.3632277170885e-05, "loss": 0.0227, "step": 3020 }, { "epoch": 2.7849264705882355, "grad_norm": 0.41659340262413025, "learning_rate": 8.350974254143318e-05, "loss": 0.0178, "step": 3030 }, { "epoch": 2.7941176470588234, "grad_norm": 0.3125268518924713, "learning_rate": 8.338684145480566e-05, "loss": 0.0164, "step": 3040 }, { "epoch": 2.8033088235294117, "grad_norm": 0.3497525453567505, "learning_rate": 8.326357525502904e-05, "loss": 0.0204, "step": 3050 }, { "epoch": 2.8125, "grad_norm": 0.3184756338596344, "learning_rate": 8.313994529012273e-05, "loss": 0.02, "step": 3060 }, { "epoch": 2.8216911764705883, "grad_norm": 0.42762669920921326, "learning_rate": 8.301595291208422e-05, "loss": 0.0215, "step": 3070 }, { "epoch": 2.8308823529411766, "grad_norm": 0.3894162178039551, "learning_rate": 8.289159947687427e-05, "loss": 0.0189, "step": 3080 }, { "epoch": 2.8400735294117645, "grad_norm": 0.33726659417152405, "learning_rate": 8.276688634440216e-05, "loss": 0.0189, "step": 3090 }, { "epoch": 2.849264705882353, "grad_norm": 0.3626478612422943, "learning_rate": 8.26418148785107e-05, "loss": 0.019, "step": 3100 }, { "epoch": 2.858455882352941, "grad_norm": 0.47513875365257263, "learning_rate": 8.251638644696141e-05, "loss": 0.0221, "step": 3110 }, { "epoch": 2.8676470588235294, "grad_norm": 0.4034767746925354, "learning_rate": 8.23906024214195e-05, "loss": 0.0181, "step": 3120 }, { "epoch": 2.8768382352941178, "grad_norm": 0.5271226763725281, "learning_rate": 8.226446417743897e-05, "loss": 0.0214, "step": 3130 }, { "epoch": 2.8860294117647056, "grad_norm": 0.5263476967811584, "learning_rate": 8.213797309444742e-05, "loss": 0.0195, "step": 3140 }, { "epoch": 2.8952205882352944, "grad_norm": 0.5458633899688721, "learning_rate": 8.201113055573105e-05, "loss": 0.0203, "step": 3150 }, { "epoch": 2.9044117647058822, "grad_norm": 0.375986248254776, "learning_rate": 8.188393794841958e-05, "loss": 0.0182, "step": 3160 }, { "epoch": 2.9136029411764706, "grad_norm": 0.4179709851741791, "learning_rate": 8.175639666347094e-05, "loss": 0.0204, "step": 3170 }, { "epoch": 2.922794117647059, "grad_norm": 0.4425836205482483, "learning_rate": 8.162850809565623e-05, "loss": 0.0183, "step": 3180 }, { "epoch": 2.931985294117647, "grad_norm": 0.45844489336013794, "learning_rate": 8.150027364354431e-05, "loss": 0.0193, "step": 3190 }, { "epoch": 2.9411764705882355, "grad_norm": 0.5459668040275574, "learning_rate": 8.137169470948662e-05, "loss": 0.0218, "step": 3200 }, { "epoch": 2.9503676470588234, "grad_norm": 0.5563957691192627, "learning_rate": 8.124277269960179e-05, "loss": 0.0198, "step": 3210 }, { "epoch": 2.9595588235294117, "grad_norm": 0.5459716320037842, "learning_rate": 8.111350902376023e-05, "loss": 0.0177, "step": 3220 }, { "epoch": 2.96875, "grad_norm": 0.41941261291503906, "learning_rate": 8.098390509556883e-05, "loss": 0.0181, "step": 3230 }, { "epoch": 2.9779411764705883, "grad_norm": 0.4461565911769867, "learning_rate": 8.085396233235536e-05, "loss": 0.0158, "step": 3240 }, { "epoch": 2.9871323529411766, "grad_norm": 0.3183092176914215, "learning_rate": 8.072368215515306e-05, "loss": 0.0153, "step": 3250 }, { "epoch": 2.9963235294117645, "grad_norm": 0.3780999481678009, "learning_rate": 8.059306598868506e-05, "loss": 0.0237, "step": 3260 }, { "epoch": 3.005514705882353, "grad_norm": 0.3313882648944855, "learning_rate": 8.046211526134888e-05, "loss": 0.0183, "step": 3270 }, { "epoch": 3.014705882352941, "grad_norm": 0.3809677064418793, "learning_rate": 8.033083140520065e-05, "loss": 0.0163, "step": 3280 }, { "epoch": 3.0238970588235294, "grad_norm": 0.45595690608024597, "learning_rate": 8.019921585593962e-05, "loss": 0.0182, "step": 3290 }, { "epoch": 3.0330882352941178, "grad_norm": 0.43894991278648376, "learning_rate": 8.006727005289232e-05, "loss": 0.0175, "step": 3300 }, { "epoch": 3.042279411764706, "grad_norm": 0.5120572447776794, "learning_rate": 7.993499543899692e-05, "loss": 0.0181, "step": 3310 }, { "epoch": 3.051470588235294, "grad_norm": 0.38996589183807373, "learning_rate": 7.980239346078742e-05, "loss": 0.0163, "step": 3320 }, { "epoch": 3.0606617647058822, "grad_norm": 0.3946544826030731, "learning_rate": 7.966946556837778e-05, "loss": 0.0179, "step": 3330 }, { "epoch": 3.0698529411764706, "grad_norm": 0.40009406208992004, "learning_rate": 7.953621321544616e-05, "loss": 0.0162, "step": 3340 }, { "epoch": 3.079044117647059, "grad_norm": 0.4754481911659241, "learning_rate": 7.940263785921896e-05, "loss": 0.0215, "step": 3350 }, { "epoch": 3.088235294117647, "grad_norm": 0.4598515033721924, "learning_rate": 7.926874096045482e-05, "loss": 0.0159, "step": 3360 }, { "epoch": 3.0974264705882355, "grad_norm": 0.3667127192020416, "learning_rate": 7.913452398342881e-05, "loss": 0.0242, "step": 3370 }, { "epoch": 3.1066176470588234, "grad_norm": 0.3552553057670593, "learning_rate": 7.89999883959163e-05, "loss": 0.0193, "step": 3380 }, { "epoch": 3.1158088235294117, "grad_norm": 0.3895982503890991, "learning_rate": 7.886513566917687e-05, "loss": 0.0204, "step": 3390 }, { "epoch": 3.125, "grad_norm": 0.39475199580192566, "learning_rate": 7.872996727793838e-05, "loss": 0.0223, "step": 3400 }, { "epoch": 3.1341911764705883, "grad_norm": 0.3694431781768799, "learning_rate": 7.859448470038069e-05, "loss": 0.0187, "step": 3410 }, { "epoch": 3.1433823529411766, "grad_norm": 0.36823663115501404, "learning_rate": 7.845868941811956e-05, "loss": 0.0193, "step": 3420 }, { "epoch": 3.1525735294117645, "grad_norm": 0.4742307960987091, "learning_rate": 7.832258291619043e-05, "loss": 0.0172, "step": 3430 }, { "epoch": 3.161764705882353, "grad_norm": 0.3917524516582489, "learning_rate": 7.81861666830322e-05, "loss": 0.0198, "step": 3440 }, { "epoch": 3.170955882352941, "grad_norm": 0.4982855021953583, "learning_rate": 7.804944221047097e-05, "loss": 0.0164, "step": 3450 }, { "epoch": 3.1801470588235294, "grad_norm": 0.4138641059398651, "learning_rate": 7.791241099370364e-05, "loss": 0.0219, "step": 3460 }, { "epoch": 3.1893382352941178, "grad_norm": 0.27899065613746643, "learning_rate": 7.777507453128163e-05, "loss": 0.0164, "step": 3470 }, { "epoch": 3.198529411764706, "grad_norm": 0.3251478970050812, "learning_rate": 7.763743432509451e-05, "loss": 0.0186, "step": 3480 }, { "epoch": 3.207720588235294, "grad_norm": 0.4538785219192505, "learning_rate": 7.749949188035353e-05, "loss": 0.0177, "step": 3490 }, { "epoch": 3.2169117647058822, "grad_norm": 0.3044322729110718, "learning_rate": 7.736124870557516e-05, "loss": 0.024, "step": 3500 }, { "epoch": 3.2261029411764706, "grad_norm": 0.3238984942436218, "learning_rate": 7.722270631256459e-05, "loss": 0.0184, "step": 3510 }, { "epoch": 3.235294117647059, "grad_norm": 0.35984793305397034, "learning_rate": 7.708386621639925e-05, "loss": 0.0166, "step": 3520 }, { "epoch": 3.244485294117647, "grad_norm": 0.38111382722854614, "learning_rate": 7.694472993541219e-05, "loss": 0.0167, "step": 3530 }, { "epoch": 3.2536764705882355, "grad_norm": 0.4613358974456787, "learning_rate": 7.680529899117547e-05, "loss": 0.021, "step": 3540 }, { "epoch": 3.2628676470588234, "grad_norm": 0.42021623253822327, "learning_rate": 7.666557490848358e-05, "loss": 0.018, "step": 3550 }, { "epoch": 3.2720588235294117, "grad_norm": 0.519267737865448, "learning_rate": 7.65255592153367e-05, "loss": 0.0192, "step": 3560 }, { "epoch": 3.28125, "grad_norm": 0.5877336859703064, "learning_rate": 7.638525344292402e-05, "loss": 0.0207, "step": 3570 }, { "epoch": 3.2904411764705883, "grad_norm": 0.5215348601341248, "learning_rate": 7.624465912560697e-05, "loss": 0.021, "step": 3580 }, { "epoch": 3.2996323529411766, "grad_norm": 0.3409864008426666, "learning_rate": 7.610377780090249e-05, "loss": 0.0173, "step": 3590 }, { "epoch": 3.3088235294117645, "grad_norm": 0.49105897545814514, "learning_rate": 7.596261100946618e-05, "loss": 0.0196, "step": 3600 }, { "epoch": 3.318014705882353, "grad_norm": 0.510987401008606, "learning_rate": 7.582116029507542e-05, "loss": 0.0158, "step": 3610 }, { "epoch": 3.327205882352941, "grad_norm": 0.5016774535179138, "learning_rate": 7.56794272046126e-05, "loss": 0.018, "step": 3620 }, { "epoch": 3.3363970588235294, "grad_norm": 0.3931902348995209, "learning_rate": 7.55374132880481e-05, "loss": 0.0185, "step": 3630 }, { "epoch": 3.3455882352941178, "grad_norm": 0.31327125430107117, "learning_rate": 7.539512009842333e-05, "loss": 0.0196, "step": 3640 }, { "epoch": 3.354779411764706, "grad_norm": 0.4895755648612976, "learning_rate": 7.525254919183382e-05, "loss": 0.0202, "step": 3650 }, { "epoch": 3.363970588235294, "grad_norm": 0.5705951452255249, "learning_rate": 7.510970212741215e-05, "loss": 0.0148, "step": 3660 }, { "epoch": 3.3731617647058822, "grad_norm": 0.49122869968414307, "learning_rate": 7.496658046731096e-05, "loss": 0.0177, "step": 3670 }, { "epoch": 3.3823529411764706, "grad_norm": 0.30971559882164, "learning_rate": 7.482318577668578e-05, "loss": 0.0191, "step": 3680 }, { "epoch": 3.391544117647059, "grad_norm": 0.4256956875324249, "learning_rate": 7.467951962367796e-05, "loss": 0.019, "step": 3690 }, { "epoch": 3.400735294117647, "grad_norm": 0.3174140155315399, "learning_rate": 7.453558357939755e-05, "loss": 0.0163, "step": 3700 }, { "epoch": 3.4099264705882355, "grad_norm": 0.47344061732292175, "learning_rate": 7.439137921790606e-05, "loss": 0.0213, "step": 3710 }, { "epoch": 3.4191176470588234, "grad_norm": 0.506260335445404, "learning_rate": 7.42469081161993e-05, "loss": 0.0194, "step": 3720 }, { "epoch": 3.4283088235294117, "grad_norm": 0.4764038622379303, "learning_rate": 7.410217185419006e-05, "loss": 0.0155, "step": 3730 }, { "epoch": 3.4375, "grad_norm": 0.5739851593971252, "learning_rate": 7.395717201469095e-05, "loss": 0.0224, "step": 3740 }, { "epoch": 3.4466911764705883, "grad_norm": 0.4754575788974762, "learning_rate": 7.381191018339696e-05, "loss": 0.0184, "step": 3750 }, { "epoch": 3.4558823529411766, "grad_norm": 0.42088451981544495, "learning_rate": 7.36663879488682e-05, "loss": 0.0184, "step": 3760 }, { "epoch": 3.4650735294117645, "grad_norm": 0.33826860785484314, "learning_rate": 7.352060690251254e-05, "loss": 0.0215, "step": 3770 }, { "epoch": 3.474264705882353, "grad_norm": 0.6039459109306335, "learning_rate": 7.337456863856811e-05, "loss": 0.0167, "step": 3780 }, { "epoch": 3.483455882352941, "grad_norm": 0.3867538869380951, "learning_rate": 7.3228274754086e-05, "loss": 0.0154, "step": 3790 }, { "epoch": 3.4926470588235294, "grad_norm": 0.3827327489852905, "learning_rate": 7.308172684891267e-05, "loss": 0.0189, "step": 3800 }, { "epoch": 3.5018382352941178, "grad_norm": 0.4909115135669708, "learning_rate": 7.293492652567255e-05, "loss": 0.015, "step": 3810 }, { "epoch": 3.5110294117647056, "grad_norm": 0.37434259057044983, "learning_rate": 7.278787538975043e-05, "loss": 0.0199, "step": 3820 }, { "epoch": 3.5202205882352944, "grad_norm": 0.3311546742916107, "learning_rate": 7.2640575049274e-05, "loss": 0.0193, "step": 3830 }, { "epoch": 3.5294117647058822, "grad_norm": 0.3393455743789673, "learning_rate": 7.249302711509616e-05, "loss": 0.0165, "step": 3840 }, { "epoch": 3.5386029411764706, "grad_norm": 0.4709666669368744, "learning_rate": 7.23452332007775e-05, "loss": 0.0164, "step": 3850 }, { "epoch": 3.547794117647059, "grad_norm": 0.47920647263526917, "learning_rate": 7.219719492256858e-05, "loss": 0.0194, "step": 3860 }, { "epoch": 3.556985294117647, "grad_norm": 0.5136281847953796, "learning_rate": 7.20489138993923e-05, "loss": 0.0176, "step": 3870 }, { "epoch": 3.5661764705882355, "grad_norm": 0.32562685012817383, "learning_rate": 7.190039175282614e-05, "loss": 0.0167, "step": 3880 }, { "epoch": 3.5753676470588234, "grad_norm": 0.4345424771308899, "learning_rate": 7.175163010708455e-05, "loss": 0.0158, "step": 3890 }, { "epoch": 3.5845588235294117, "grad_norm": 0.4432017505168915, "learning_rate": 7.1602630589001e-05, "loss": 0.017, "step": 3900 }, { "epoch": 3.59375, "grad_norm": 0.3689446449279785, "learning_rate": 7.14533948280104e-05, "loss": 0.016, "step": 3910 }, { "epoch": 3.6029411764705883, "grad_norm": 0.4982896149158478, "learning_rate": 7.130392445613109e-05, "loss": 0.015, "step": 3920 }, { "epoch": 3.6121323529411766, "grad_norm": 0.4646291136741638, "learning_rate": 7.115422110794711e-05, "loss": 0.0164, "step": 3930 }, { "epoch": 3.6213235294117645, "grad_norm": 0.6014279127120972, "learning_rate": 7.100428642059033e-05, "loss": 0.0182, "step": 3940 }, { "epoch": 3.630514705882353, "grad_norm": 0.6391748189926147, "learning_rate": 7.08541220337224e-05, "loss": 0.0193, "step": 3950 }, { "epoch": 3.639705882352941, "grad_norm": 0.5698084235191345, "learning_rate": 7.070372958951706e-05, "loss": 0.0164, "step": 3960 }, { "epoch": 3.6488970588235294, "grad_norm": 0.4046298861503601, "learning_rate": 7.055311073264194e-05, "loss": 0.019, "step": 3970 }, { "epoch": 3.6580882352941178, "grad_norm": 0.5220407247543335, "learning_rate": 7.040226711024077e-05, "loss": 0.016, "step": 3980 }, { "epoch": 3.6672794117647056, "grad_norm": 0.3931867778301239, "learning_rate": 7.02512003719152e-05, "loss": 0.0163, "step": 3990 }, { "epoch": 3.6764705882352944, "grad_norm": 0.5839005708694458, "learning_rate": 7.00999121697069e-05, "loss": 0.0174, "step": 4000 }, { "epoch": 3.6856617647058822, "grad_norm": 0.3060004711151123, "learning_rate": 6.99484041580794e-05, "loss": 0.016, "step": 4010 }, { "epoch": 3.6948529411764706, "grad_norm": 0.47329533100128174, "learning_rate": 6.979667799390004e-05, "loss": 0.0152, "step": 4020 }, { "epoch": 3.704044117647059, "grad_norm": 0.4491497874259949, "learning_rate": 6.964473533642185e-05, "loss": 0.0171, "step": 4030 }, { "epoch": 3.713235294117647, "grad_norm": 0.236502543091774, "learning_rate": 6.949257784726539e-05, "loss": 0.0169, "step": 4040 }, { "epoch": 3.7224264705882355, "grad_norm": 0.31270214915275574, "learning_rate": 6.934020719040056e-05, "loss": 0.0144, "step": 4050 }, { "epoch": 3.7316176470588234, "grad_norm": 0.302190899848938, "learning_rate": 6.918762503212848e-05, "loss": 0.019, "step": 4060 }, { "epoch": 3.7408088235294117, "grad_norm": 0.34432461857795715, "learning_rate": 6.903483304106319e-05, "loss": 0.0169, "step": 4070 }, { "epoch": 3.75, "grad_norm": 0.4090389609336853, "learning_rate": 6.888183288811341e-05, "loss": 0.0151, "step": 4080 }, { "epoch": 3.7591911764705883, "grad_norm": 0.4263298511505127, "learning_rate": 6.87286262464643e-05, "loss": 0.0152, "step": 4090 }, { "epoch": 3.7683823529411766, "grad_norm": 0.3984905481338501, "learning_rate": 6.857521479155915e-05, "loss": 0.0174, "step": 4100 }, { "epoch": 3.7775735294117645, "grad_norm": 0.48731496930122375, "learning_rate": 6.842160020108104e-05, "loss": 0.0158, "step": 4110 }, { "epoch": 3.786764705882353, "grad_norm": 0.31711211800575256, "learning_rate": 6.826778415493455e-05, "loss": 0.0162, "step": 4120 }, { "epoch": 3.795955882352941, "grad_norm": 0.41115257143974304, "learning_rate": 6.811376833522729e-05, "loss": 0.0163, "step": 4130 }, { "epoch": 3.8051470588235294, "grad_norm": 0.3736651837825775, "learning_rate": 6.795955442625159e-05, "loss": 0.0171, "step": 4140 }, { "epoch": 3.8143382352941178, "grad_norm": 0.40125811100006104, "learning_rate": 6.780514411446608e-05, "loss": 0.0145, "step": 4150 }, { "epoch": 3.8235294117647056, "grad_norm": 0.5432369709014893, "learning_rate": 6.765053908847716e-05, "loss": 0.0142, "step": 4160 }, { "epoch": 3.8327205882352944, "grad_norm": 0.6269651055335999, "learning_rate": 6.749574103902064e-05, "loss": 0.0195, "step": 4170 }, { "epoch": 3.8419117647058822, "grad_norm": 0.41522759199142456, "learning_rate": 6.734075165894317e-05, "loss": 0.0184, "step": 4180 }, { "epoch": 3.8511029411764706, "grad_norm": 0.456086665391922, "learning_rate": 6.71855726431838e-05, "loss": 0.018, "step": 4190 }, { "epoch": 3.860294117647059, "grad_norm": 0.41022390127182007, "learning_rate": 6.703020568875538e-05, "loss": 0.0144, "step": 4200 }, { "epoch": 3.869485294117647, "grad_norm": 0.3123449385166168, "learning_rate": 6.687465249472603e-05, "loss": 0.0134, "step": 4210 }, { "epoch": 3.8786764705882355, "grad_norm": 0.4034487009048462, "learning_rate": 6.671891476220055e-05, "loss": 0.015, "step": 4220 }, { "epoch": 3.8878676470588234, "grad_norm": 0.36322689056396484, "learning_rate": 6.656299419430183e-05, "loss": 0.0138, "step": 4230 }, { "epoch": 3.8970588235294117, "grad_norm": 0.47680172324180603, "learning_rate": 6.640689249615223e-05, "loss": 0.0168, "step": 4240 }, { "epoch": 3.90625, "grad_norm": 0.36782121658325195, "learning_rate": 6.625061137485491e-05, "loss": 0.0167, "step": 4250 }, { "epoch": 3.9154411764705883, "grad_norm": 0.33727946877479553, "learning_rate": 6.609415253947517e-05, "loss": 0.0179, "step": 4260 }, { "epoch": 3.9246323529411766, "grad_norm": 0.37037405371665955, "learning_rate": 6.593751770102178e-05, "loss": 0.0164, "step": 4270 }, { "epoch": 3.9338235294117645, "grad_norm": 0.4587080478668213, "learning_rate": 6.578070857242823e-05, "loss": 0.0139, "step": 4280 }, { "epoch": 3.943014705882353, "grad_norm": 0.37842676043510437, "learning_rate": 6.562372686853402e-05, "loss": 0.0138, "step": 4290 }, { "epoch": 3.952205882352941, "grad_norm": 0.43958625197410583, "learning_rate": 6.546657430606593e-05, "loss": 0.0161, "step": 4300 }, { "epoch": 3.9613970588235294, "grad_norm": 0.5308313965797424, "learning_rate": 6.530925260361918e-05, "loss": 0.0136, "step": 4310 }, { "epoch": 3.9705882352941178, "grad_norm": 0.35479310154914856, "learning_rate": 6.515176348163871e-05, "loss": 0.016, "step": 4320 }, { "epoch": 3.9797794117647056, "grad_norm": 0.45303773880004883, "learning_rate": 6.499410866240032e-05, "loss": 0.0133, "step": 4330 }, { "epoch": 3.9889705882352944, "grad_norm": 0.33201804757118225, "learning_rate": 6.48362898699919e-05, "loss": 0.0146, "step": 4340 }, { "epoch": 3.9981617647058822, "grad_norm": 0.43322840332984924, "learning_rate": 6.467830883029443e-05, "loss": 0.015, "step": 4350 }, { "epoch": 4.007352941176471, "grad_norm": 0.3870246410369873, "learning_rate": 6.452016727096326e-05, "loss": 0.0137, "step": 4360 }, { "epoch": 4.016544117647059, "grad_norm": 0.29066815972328186, "learning_rate": 6.436186692140916e-05, "loss": 0.014, "step": 4370 }, { "epoch": 4.025735294117647, "grad_norm": 0.36254173517227173, "learning_rate": 6.420340951277938e-05, "loss": 0.0135, "step": 4380 }, { "epoch": 4.0349264705882355, "grad_norm": 0.4960012137889862, "learning_rate": 6.404479677793874e-05, "loss": 0.0183, "step": 4390 }, { "epoch": 4.044117647058823, "grad_norm": 0.4960069954395294, "learning_rate": 6.388603045145075e-05, "loss": 0.0188, "step": 4400 }, { "epoch": 4.053308823529412, "grad_norm": 0.3833214342594147, "learning_rate": 6.372711226955843e-05, "loss": 0.0153, "step": 4410 }, { "epoch": 4.0625, "grad_norm": 0.43608716130256653, "learning_rate": 6.356804397016564e-05, "loss": 0.0182, "step": 4420 }, { "epoch": 4.071691176470588, "grad_norm": 0.3791906237602234, "learning_rate": 6.340882729281779e-05, "loss": 0.0161, "step": 4430 }, { "epoch": 4.080882352941177, "grad_norm": 0.4668147563934326, "learning_rate": 6.324946397868294e-05, "loss": 0.0207, "step": 4440 }, { "epoch": 4.0900735294117645, "grad_norm": 0.34103018045425415, "learning_rate": 6.308995577053276e-05, "loss": 0.0143, "step": 4450 }, { "epoch": 4.099264705882353, "grad_norm": 0.4538170397281647, "learning_rate": 6.293030441272347e-05, "loss": 0.0155, "step": 4460 }, { "epoch": 4.108455882352941, "grad_norm": 0.5139859318733215, "learning_rate": 6.277051165117677e-05, "loss": 0.0142, "step": 4470 }, { "epoch": 4.117647058823529, "grad_norm": 0.3032166659832001, "learning_rate": 6.261057923336064e-05, "loss": 0.015, "step": 4480 }, { "epoch": 4.126838235294118, "grad_norm": 0.4086776673793793, "learning_rate": 6.245050890827042e-05, "loss": 0.0179, "step": 4490 }, { "epoch": 4.136029411764706, "grad_norm": 0.29072219133377075, "learning_rate": 6.229030242640952e-05, "loss": 0.0158, "step": 4500 }, { "epoch": 4.145220588235294, "grad_norm": 0.41120445728302, "learning_rate": 6.212996153977037e-05, "loss": 0.0183, "step": 4510 }, { "epoch": 4.154411764705882, "grad_norm": 0.36678290367126465, "learning_rate": 6.196948800181523e-05, "loss": 0.0148, "step": 4520 }, { "epoch": 4.163602941176471, "grad_norm": 0.5274208784103394, "learning_rate": 6.180888356745695e-05, "loss": 0.015, "step": 4530 }, { "epoch": 4.172794117647059, "grad_norm": 0.36312589049339294, "learning_rate": 6.164814999303995e-05, "loss": 0.0145, "step": 4540 }, { "epoch": 4.181985294117647, "grad_norm": 0.5565803647041321, "learning_rate": 6.148728903632081e-05, "loss": 0.0136, "step": 4550 }, { "epoch": 4.1911764705882355, "grad_norm": 0.2668694257736206, "learning_rate": 6.132630245644921e-05, "loss": 0.0123, "step": 4560 }, { "epoch": 4.200367647058823, "grad_norm": 0.4848521649837494, "learning_rate": 6.116519201394857e-05, "loss": 0.015, "step": 4570 }, { "epoch": 4.209558823529412, "grad_norm": 0.32921454310417175, "learning_rate": 6.10039594706969e-05, "loss": 0.0142, "step": 4580 }, { "epoch": 4.21875, "grad_norm": 0.25363287329673767, "learning_rate": 6.084260658990744e-05, "loss": 0.0135, "step": 4590 }, { "epoch": 4.227941176470588, "grad_norm": 0.2757788300514221, "learning_rate": 6.068113513610943e-05, "loss": 0.0142, "step": 4600 }, { "epoch": 4.237132352941177, "grad_norm": 0.3910484313964844, "learning_rate": 6.0519546875128876e-05, "loss": 0.0141, "step": 4610 }, { "epoch": 4.2463235294117645, "grad_norm": 0.34375420212745667, "learning_rate": 6.035784357406906e-05, "loss": 0.017, "step": 4620 }, { "epoch": 4.255514705882353, "grad_norm": 0.40266719460487366, "learning_rate": 6.01960270012914e-05, "loss": 0.0147, "step": 4630 }, { "epoch": 4.264705882352941, "grad_norm": 0.42728865146636963, "learning_rate": 6.003409892639599e-05, "loss": 0.0164, "step": 4640 }, { "epoch": 4.273897058823529, "grad_norm": 0.4151376187801361, "learning_rate": 5.9872061120202336e-05, "loss": 0.016, "step": 4650 }, { "epoch": 4.283088235294118, "grad_norm": 0.5265789031982422, "learning_rate": 5.9709915354729914e-05, "loss": 0.0219, "step": 4660 }, { "epoch": 4.292279411764706, "grad_norm": 0.2692059278488159, "learning_rate": 5.9547663403178824e-05, "loss": 0.0142, "step": 4670 }, { "epoch": 4.301470588235294, "grad_norm": 0.3371903598308563, "learning_rate": 5.9385307039910445e-05, "loss": 0.0116, "step": 4680 }, { "epoch": 4.310661764705882, "grad_norm": 0.373264878988266, "learning_rate": 5.922284804042792e-05, "loss": 0.0169, "step": 4690 }, { "epoch": 4.319852941176471, "grad_norm": 0.527279257774353, "learning_rate": 5.906028818135687e-05, "loss": 0.0199, "step": 4700 }, { "epoch": 4.329044117647059, "grad_norm": 0.37420180439949036, "learning_rate": 5.889762924042585e-05, "loss": 0.0142, "step": 4710 }, { "epoch": 4.338235294117647, "grad_norm": 0.4379793405532837, "learning_rate": 5.873487299644699e-05, "loss": 0.0155, "step": 4720 }, { "epoch": 4.3474264705882355, "grad_norm": 0.4763709008693695, "learning_rate": 5.857202122929649e-05, "loss": 0.0155, "step": 4730 }, { "epoch": 4.356617647058823, "grad_norm": 0.3225118815898895, "learning_rate": 5.840907571989518e-05, "loss": 0.0139, "step": 4740 }, { "epoch": 4.365808823529412, "grad_norm": 0.3212355971336365, "learning_rate": 5.824603825018904e-05, "loss": 0.0141, "step": 4750 }, { "epoch": 4.375, "grad_norm": 0.34345975518226624, "learning_rate": 5.808291060312975e-05, "loss": 0.0144, "step": 4760 }, { "epoch": 4.384191176470588, "grad_norm": 0.333039790391922, "learning_rate": 5.7919694562655083e-05, "loss": 0.015, "step": 4770 }, { "epoch": 4.393382352941177, "grad_norm": 0.3856314420700073, "learning_rate": 5.775639191366954e-05, "loss": 0.0159, "step": 4780 }, { "epoch": 4.4025735294117645, "grad_norm": 0.3945397436618805, "learning_rate": 5.75930044420247e-05, "loss": 0.014, "step": 4790 }, { "epoch": 4.411764705882353, "grad_norm": 0.4219256043434143, "learning_rate": 5.74295339344998e-05, "loss": 0.0165, "step": 4800 }, { "epoch": 4.420955882352941, "grad_norm": 0.38971999287605286, "learning_rate": 5.726598217878211e-05, "loss": 0.0134, "step": 4810 }, { "epoch": 4.430147058823529, "grad_norm": 0.2993128001689911, "learning_rate": 5.71023509634474e-05, "loss": 0.0121, "step": 4820 }, { "epoch": 4.439338235294118, "grad_norm": 0.31463080644607544, "learning_rate": 5.693864207794049e-05, "loss": 0.0168, "step": 4830 }, { "epoch": 4.448529411764706, "grad_norm": 0.2755267918109894, "learning_rate": 5.677485731255545e-05, "loss": 0.0216, "step": 4840 }, { "epoch": 4.457720588235294, "grad_norm": 0.3567773103713989, "learning_rate": 5.6610998458416296e-05, "loss": 0.0154, "step": 4850 }, { "epoch": 4.466911764705882, "grad_norm": 0.43718230724334717, "learning_rate": 5.644706730745716e-05, "loss": 0.0152, "step": 4860 }, { "epoch": 4.476102941176471, "grad_norm": 0.29651567339897156, "learning_rate": 5.628306565240287e-05, "loss": 0.013, "step": 4870 }, { "epoch": 4.485294117647059, "grad_norm": 0.3140612542629242, "learning_rate": 5.611899528674923e-05, "loss": 0.0168, "step": 4880 }, { "epoch": 4.494485294117647, "grad_norm": 0.36880287528038025, "learning_rate": 5.595485800474349e-05, "loss": 0.0144, "step": 4890 }, { "epoch": 4.5036764705882355, "grad_norm": 0.3191617429256439, "learning_rate": 5.579065560136467e-05, "loss": 0.014, "step": 4900 }, { "epoch": 4.512867647058823, "grad_norm": 0.43635985255241394, "learning_rate": 5.562638987230392e-05, "loss": 0.0129, "step": 4910 }, { "epoch": 4.522058823529412, "grad_norm": 0.33788228034973145, "learning_rate": 5.546206261394498e-05, "loss": 0.012, "step": 4920 }, { "epoch": 4.53125, "grad_norm": 0.38356906175613403, "learning_rate": 5.529767562334437e-05, "loss": 0.0149, "step": 4930 }, { "epoch": 4.540441176470588, "grad_norm": 0.36063456535339355, "learning_rate": 5.5133230698211926e-05, "loss": 0.0134, "step": 4940 }, { "epoch": 4.549632352941177, "grad_norm": 0.42338675260543823, "learning_rate": 5.496872963689096e-05, "loss": 0.0147, "step": 4950 }, { "epoch": 4.5588235294117645, "grad_norm": 0.3563850522041321, "learning_rate": 5.4804174238338756e-05, "loss": 0.0118, "step": 4960 }, { "epoch": 4.568014705882353, "grad_norm": 0.3230605721473694, "learning_rate": 5.463956630210678e-05, "loss": 0.0188, "step": 4970 }, { "epoch": 4.577205882352941, "grad_norm": 0.3216652274131775, "learning_rate": 5.4474907628321046e-05, "loss": 0.0169, "step": 4980 }, { "epoch": 4.586397058823529, "grad_norm": 0.4549153447151184, "learning_rate": 5.431020001766244e-05, "loss": 0.0136, "step": 4990 }, { "epoch": 4.595588235294118, "grad_norm": 0.39283961057662964, "learning_rate": 5.4145445271346986e-05, "loss": 0.0141, "step": 5000 }, { "epoch": 4.604779411764706, "grad_norm": 0.30578985810279846, "learning_rate": 5.398064519110622e-05, "loss": 0.0156, "step": 5010 }, { "epoch": 4.613970588235294, "grad_norm": 0.3042331039905548, "learning_rate": 5.3815801579167394e-05, "loss": 0.0138, "step": 5020 }, { "epoch": 4.623161764705882, "grad_norm": 0.2997395694255829, "learning_rate": 5.365091623823382e-05, "loss": 0.0137, "step": 5030 }, { "epoch": 4.632352941176471, "grad_norm": 0.3186718821525574, "learning_rate": 5.348599097146521e-05, "loss": 0.0127, "step": 5040 }, { "epoch": 4.641544117647059, "grad_norm": 0.36802321672439575, "learning_rate": 5.3321027582457836e-05, "loss": 0.0145, "step": 5050 }, { "epoch": 4.650735294117647, "grad_norm": 0.5434386134147644, "learning_rate": 5.315602787522491e-05, "loss": 0.0136, "step": 5060 }, { "epoch": 4.6599264705882355, "grad_norm": 0.35909056663513184, "learning_rate": 5.299099365417678e-05, "loss": 0.011, "step": 5070 }, { "epoch": 4.669117647058823, "grad_norm": 0.504350483417511, "learning_rate": 5.2825926724101236e-05, "loss": 0.0139, "step": 5080 }, { "epoch": 4.678308823529412, "grad_norm": 0.371775358915329, "learning_rate": 5.26608288901438e-05, "loss": 0.0141, "step": 5090 }, { "epoch": 4.6875, "grad_norm": 0.4863434433937073, "learning_rate": 5.24957019577879e-05, "loss": 0.0162, "step": 5100 }, { "epoch": 4.696691176470588, "grad_norm": 0.4539608359336853, "learning_rate": 5.2330547732835266e-05, "loss": 0.0126, "step": 5110 }, { "epoch": 4.705882352941177, "grad_norm": 0.48757702112197876, "learning_rate": 5.2165368021385996e-05, "loss": 0.0189, "step": 5120 }, { "epoch": 4.7150735294117645, "grad_norm": 0.33881187438964844, "learning_rate": 5.200016462981897e-05, "loss": 0.0116, "step": 5130 }, { "epoch": 4.724264705882353, "grad_norm": 0.37330809235572815, "learning_rate": 5.1834939364772015e-05, "loss": 0.013, "step": 5140 }, { "epoch": 4.733455882352941, "grad_norm": 0.37893691658973694, "learning_rate": 5.166969403312214e-05, "loss": 0.0124, "step": 5150 }, { "epoch": 4.742647058823529, "grad_norm": 0.3603902757167816, "learning_rate": 5.1504430441965844e-05, "loss": 0.0143, "step": 5160 }, { "epoch": 4.751838235294118, "grad_norm": 0.3397335410118103, "learning_rate": 5.133915039859923e-05, "loss": 0.0108, "step": 5170 }, { "epoch": 4.761029411764706, "grad_norm": 0.40848270058631897, "learning_rate": 5.1173855710498444e-05, "loss": 0.0126, "step": 5180 }, { "epoch": 4.770220588235294, "grad_norm": 0.37968817353248596, "learning_rate": 5.100854818529967e-05, "loss": 0.0154, "step": 5190 }, { "epoch": 4.779411764705882, "grad_norm": 0.38904112577438354, "learning_rate": 5.084322963077951e-05, "loss": 0.0162, "step": 5200 }, { "epoch": 4.788602941176471, "grad_norm": 0.43914979696273804, "learning_rate": 5.067790185483522e-05, "loss": 0.0131, "step": 5210 }, { "epoch": 4.797794117647059, "grad_norm": 0.44638633728027344, "learning_rate": 5.0512566665464844e-05, "loss": 0.0145, "step": 5220 }, { "epoch": 4.806985294117647, "grad_norm": 0.30461907386779785, "learning_rate": 5.034722587074755e-05, "loss": 0.0135, "step": 5230 }, { "epoch": 4.8161764705882355, "grad_norm": 0.3084958791732788, "learning_rate": 5.018188127882375e-05, "loss": 0.0124, "step": 5240 }, { "epoch": 4.825367647058823, "grad_norm": 0.4256832003593445, "learning_rate": 5.0016534697875417e-05, "loss": 0.0115, "step": 5250 }, { "epoch": 4.834558823529412, "grad_norm": 0.28911206126213074, "learning_rate": 4.9851187936106294e-05, "loss": 0.0143, "step": 5260 }, { "epoch": 4.84375, "grad_norm": 0.31694406270980835, "learning_rate": 4.968584280172206e-05, "loss": 0.0138, "step": 5270 }, { "epoch": 4.852941176470588, "grad_norm": 0.33373701572418213, "learning_rate": 4.95205011029106e-05, "loss": 0.0137, "step": 5280 }, { "epoch": 4.862132352941177, "grad_norm": 0.4766901731491089, "learning_rate": 4.935516464782227e-05, "loss": 0.0143, "step": 5290 }, { "epoch": 4.8713235294117645, "grad_norm": 0.33348438143730164, "learning_rate": 4.918983524455003e-05, "loss": 0.0128, "step": 5300 }, { "epoch": 4.880514705882353, "grad_norm": 0.428833931684494, "learning_rate": 4.9024514701109766e-05, "loss": 0.0125, "step": 5310 }, { "epoch": 4.889705882352941, "grad_norm": 0.3463518023490906, "learning_rate": 4.885920482542043e-05, "loss": 0.014, "step": 5320 }, { "epoch": 4.898897058823529, "grad_norm": 0.29322266578674316, "learning_rate": 4.869390742528438e-05, "loss": 0.0114, "step": 5330 }, { "epoch": 4.908088235294118, "grad_norm": 0.3245830535888672, "learning_rate": 4.852862430836744e-05, "loss": 0.0139, "step": 5340 }, { "epoch": 4.917279411764706, "grad_norm": 0.41826868057250977, "learning_rate": 4.836335728217933e-05, "loss": 0.0126, "step": 5350 }, { "epoch": 4.926470588235294, "grad_norm": 0.2758456766605377, "learning_rate": 4.819810815405379e-05, "loss": 0.0133, "step": 5360 }, { "epoch": 4.935661764705882, "grad_norm": 0.48234865069389343, "learning_rate": 4.803287873112877e-05, "loss": 0.0156, "step": 5370 }, { "epoch": 4.944852941176471, "grad_norm": 0.34228360652923584, "learning_rate": 4.786767082032681e-05, "loss": 0.0136, "step": 5380 }, { "epoch": 4.954044117647059, "grad_norm": 0.37777504324913025, "learning_rate": 4.77024862283351e-05, "loss": 0.0118, "step": 5390 }, { "epoch": 4.963235294117647, "grad_norm": 0.29665762186050415, "learning_rate": 4.753732676158593e-05, "loss": 0.0121, "step": 5400 }, { "epoch": 4.9724264705882355, "grad_norm": 0.2718345522880554, "learning_rate": 4.737219422623672e-05, "loss": 0.0131, "step": 5410 }, { "epoch": 4.981617647058823, "grad_norm": 0.4217487871646881, "learning_rate": 4.720709042815044e-05, "loss": 0.0109, "step": 5420 }, { "epoch": 4.990808823529412, "grad_norm": 0.43348756432533264, "learning_rate": 4.704201717287578e-05, "loss": 0.0128, "step": 5430 }, { "epoch": 5.0, "grad_norm": 0.2921617329120636, "learning_rate": 4.6876976265627404e-05, "loss": 0.0129, "step": 5440 }, { "epoch": 5.009191176470588, "grad_norm": 0.393252432346344, "learning_rate": 4.671196951126626e-05, "loss": 0.0161, "step": 5450 }, { "epoch": 5.018382352941177, "grad_norm": 0.4955695867538452, "learning_rate": 4.654699871427971e-05, "loss": 0.0109, "step": 5460 }, { "epoch": 5.0275735294117645, "grad_norm": 0.277001291513443, "learning_rate": 4.6382065678762034e-05, "loss": 0.0121, "step": 5470 }, { "epoch": 5.036764705882353, "grad_norm": 0.33374783396720886, "learning_rate": 4.6217172208394424e-05, "loss": 0.0116, "step": 5480 }, { "epoch": 5.045955882352941, "grad_norm": 0.2532215714454651, "learning_rate": 4.605232010642549e-05, "loss": 0.0101, "step": 5490 }, { "epoch": 5.055147058823529, "grad_norm": 0.3549438714981079, "learning_rate": 4.588751117565142e-05, "loss": 0.0113, "step": 5500 }, { "epoch": 5.064338235294118, "grad_norm": 0.37103039026260376, "learning_rate": 4.5722747218396214e-05, "loss": 0.0122, "step": 5510 }, { "epoch": 5.073529411764706, "grad_norm": 0.3276868462562561, "learning_rate": 4.5558030036492194e-05, "loss": 0.0097, "step": 5520 }, { "epoch": 5.082720588235294, "grad_norm": 0.2742269039154053, "learning_rate": 4.539336143125999e-05, "loss": 0.0168, "step": 5530 }, { "epoch": 5.091911764705882, "grad_norm": 0.3826009929180145, "learning_rate": 4.522874320348916e-05, "loss": 0.0112, "step": 5540 }, { "epoch": 5.101102941176471, "grad_norm": 0.27396705746650696, "learning_rate": 4.506417715341821e-05, "loss": 0.0096, "step": 5550 }, { "epoch": 5.110294117647059, "grad_norm": 0.33643218874931335, "learning_rate": 4.489966508071511e-05, "loss": 0.0114, "step": 5560 }, { "epoch": 5.119485294117647, "grad_norm": 0.3408394753932953, "learning_rate": 4.4735208784457575e-05, "loss": 0.0105, "step": 5570 }, { "epoch": 5.1286764705882355, "grad_norm": 0.39617693424224854, "learning_rate": 4.457081006311325e-05, "loss": 0.0125, "step": 5580 }, { "epoch": 5.137867647058823, "grad_norm": 0.34442245960235596, "learning_rate": 4.440647071452027e-05, "loss": 0.013, "step": 5590 }, { "epoch": 5.147058823529412, "grad_norm": 0.49555668234825134, "learning_rate": 4.424219253586737e-05, "loss": 0.0143, "step": 5600 }, { "epoch": 5.15625, "grad_norm": 0.3962758779525757, "learning_rate": 4.407797732367443e-05, "loss": 0.0144, "step": 5610 }, { "epoch": 5.165441176470588, "grad_norm": 0.5033507943153381, "learning_rate": 4.391382687377268e-05, "loss": 0.0139, "step": 5620 }, { "epoch": 5.174632352941177, "grad_norm": 0.3246171176433563, "learning_rate": 4.374974298128512e-05, "loss": 0.0126, "step": 5630 }, { "epoch": 5.1838235294117645, "grad_norm": 0.3496526777744293, "learning_rate": 4.358572744060699e-05, "loss": 0.0172, "step": 5640 }, { "epoch": 5.193014705882353, "grad_norm": 0.20570752024650574, "learning_rate": 4.342178204538588e-05, "loss": 0.0102, "step": 5650 }, { "epoch": 5.202205882352941, "grad_norm": 0.33615973591804504, "learning_rate": 4.325790858850241e-05, "loss": 0.0127, "step": 5660 }, { "epoch": 5.211397058823529, "grad_norm": 0.3672693371772766, "learning_rate": 4.309410886205043e-05, "loss": 0.0119, "step": 5670 }, { "epoch": 5.220588235294118, "grad_norm": 0.29862841963768005, "learning_rate": 4.293038465731752e-05, "loss": 0.0127, "step": 5680 }, { "epoch": 5.229779411764706, "grad_norm": 0.5310109853744507, "learning_rate": 4.276673776476533e-05, "loss": 0.0145, "step": 5690 }, { "epoch": 5.238970588235294, "grad_norm": 0.34983697533607483, "learning_rate": 4.260316997401007e-05, "loss": 0.0097, "step": 5700 }, { "epoch": 5.248161764705882, "grad_norm": 0.4464215040206909, "learning_rate": 4.243968307380293e-05, "loss": 0.0137, "step": 5710 }, { "epoch": 5.257352941176471, "grad_norm": 0.37648046016693115, "learning_rate": 4.22762788520104e-05, "loss": 0.0157, "step": 5720 }, { "epoch": 5.266544117647059, "grad_norm": 0.39762797951698303, "learning_rate": 4.211295909559491e-05, "loss": 0.0154, "step": 5730 }, { "epoch": 5.275735294117647, "grad_norm": 0.44506195187568665, "learning_rate": 4.194972559059511e-05, "loss": 0.0125, "step": 5740 }, { "epoch": 5.2849264705882355, "grad_norm": 0.4453440308570862, "learning_rate": 4.178658012210651e-05, "loss": 0.0113, "step": 5750 }, { "epoch": 5.294117647058823, "grad_norm": 0.3255781829357147, "learning_rate": 4.162352447426177e-05, "loss": 0.0111, "step": 5760 }, { "epoch": 5.303308823529412, "grad_norm": 0.5013770461082458, "learning_rate": 4.146056043021135e-05, "loss": 0.0122, "step": 5770 }, { "epoch": 5.3125, "grad_norm": 0.3653419017791748, "learning_rate": 4.1297689772103944e-05, "loss": 0.0139, "step": 5780 }, { "epoch": 5.321691176470588, "grad_norm": 0.38849711418151855, "learning_rate": 4.113491428106694e-05, "loss": 0.0133, "step": 5790 }, { "epoch": 5.330882352941177, "grad_norm": 0.3178305923938751, "learning_rate": 4.0972235737187055e-05, "loss": 0.0113, "step": 5800 }, { "epoch": 5.3400735294117645, "grad_norm": 0.28631338477134705, "learning_rate": 4.080965591949076e-05, "loss": 0.011, "step": 5810 }, { "epoch": 5.349264705882353, "grad_norm": 0.4599073529243469, "learning_rate": 4.0647176605924924e-05, "loss": 0.0113, "step": 5820 }, { "epoch": 5.358455882352941, "grad_norm": 0.4044896960258484, "learning_rate": 4.0484799573337255e-05, "loss": 0.0142, "step": 5830 }, { "epoch": 5.367647058823529, "grad_norm": 0.4467540979385376, "learning_rate": 4.032252659745699e-05, "loss": 0.0119, "step": 5840 }, { "epoch": 5.376838235294118, "grad_norm": 0.38581493496894836, "learning_rate": 4.016035945287539e-05, "loss": 0.0172, "step": 5850 }, { "epoch": 5.386029411764706, "grad_norm": 0.3222886025905609, "learning_rate": 3.999829991302635e-05, "loss": 0.0119, "step": 5860 }, { "epoch": 5.395220588235294, "grad_norm": 0.2662450969219208, "learning_rate": 3.983634975016707e-05, "loss": 0.0113, "step": 5870 }, { "epoch": 5.404411764705882, "grad_norm": 0.3706592917442322, "learning_rate": 3.967451073535854e-05, "loss": 0.0177, "step": 5880 }, { "epoch": 5.413602941176471, "grad_norm": 0.4469154179096222, "learning_rate": 3.951278463844633e-05, "loss": 0.0108, "step": 5890 }, { "epoch": 5.422794117647059, "grad_norm": 0.39254462718963623, "learning_rate": 3.935117322804111e-05, "loss": 0.0135, "step": 5900 }, { "epoch": 5.431985294117647, "grad_norm": 0.33926770091056824, "learning_rate": 3.918967827149938e-05, "loss": 0.0125, "step": 5910 }, { "epoch": 5.4411764705882355, "grad_norm": 0.43745139241218567, "learning_rate": 3.9028301534904094e-05, "loss": 0.0128, "step": 5920 }, { "epoch": 5.450367647058823, "grad_norm": 0.29478392004966736, "learning_rate": 3.88670447830454e-05, "loss": 0.0105, "step": 5930 }, { "epoch": 5.459558823529412, "grad_norm": 0.3247849643230438, "learning_rate": 3.870590977940132e-05, "loss": 0.0112, "step": 5940 }, { "epoch": 5.46875, "grad_norm": 0.33039912581443787, "learning_rate": 3.8544898286118404e-05, "loss": 0.0103, "step": 5950 }, { "epoch": 5.477941176470588, "grad_norm": 0.2854500710964203, "learning_rate": 3.838401206399257e-05, "loss": 0.01, "step": 5960 }, { "epoch": 5.487132352941177, "grad_norm": 0.30698370933532715, "learning_rate": 3.822325287244975e-05, "loss": 0.0097, "step": 5970 }, { "epoch": 5.4963235294117645, "grad_norm": 0.3080834448337555, "learning_rate": 3.8062622469526725e-05, "loss": 0.0114, "step": 5980 }, { "epoch": 5.505514705882353, "grad_norm": 0.428431898355484, "learning_rate": 3.790212261185183e-05, "loss": 0.0117, "step": 5990 }, { "epoch": 5.514705882352941, "grad_norm": 0.487800657749176, "learning_rate": 3.7741755054625794e-05, "loss": 0.0107, "step": 6000 }, { "epoch": 5.523897058823529, "grad_norm": 0.5118886828422546, "learning_rate": 3.758152155160255e-05, "loss": 0.0096, "step": 6010 }, { "epoch": 5.533088235294118, "grad_norm": 0.265516996383667, "learning_rate": 3.742142385506999e-05, "loss": 0.0117, "step": 6020 }, { "epoch": 5.542279411764706, "grad_norm": 0.4596225619316101, "learning_rate": 3.72614637158309e-05, "loss": 0.01, "step": 6030 }, { "epoch": 5.551470588235294, "grad_norm": 0.3256239891052246, "learning_rate": 3.710164288318371e-05, "loss": 0.0133, "step": 6040 }, { "epoch": 5.560661764705882, "grad_norm": 0.3839151859283447, "learning_rate": 3.694196310490345e-05, "loss": 0.0091, "step": 6050 }, { "epoch": 5.569852941176471, "grad_norm": 0.3249642550945282, "learning_rate": 3.678242612722259e-05, "loss": 0.0113, "step": 6060 }, { "epoch": 5.579044117647059, "grad_norm": 0.2995487451553345, "learning_rate": 3.6623033694811953e-05, "loss": 0.0102, "step": 6070 }, { "epoch": 5.588235294117647, "grad_norm": 0.2826014459133148, "learning_rate": 3.6463787550761665e-05, "loss": 0.0128, "step": 6080 }, { "epoch": 5.5974264705882355, "grad_norm": 0.2017374485731125, "learning_rate": 3.630468943656202e-05, "loss": 0.0105, "step": 6090 }, { "epoch": 5.606617647058823, "grad_norm": 0.33068224787712097, "learning_rate": 3.6145741092084523e-05, "loss": 0.0088, "step": 6100 }, { "epoch": 5.615808823529412, "grad_norm": 0.3728877902030945, "learning_rate": 3.598694425556278e-05, "loss": 0.009, "step": 6110 }, { "epoch": 5.625, "grad_norm": 0.3442837595939636, "learning_rate": 3.58283006635736e-05, "loss": 0.0119, "step": 6120 }, { "epoch": 5.634191176470588, "grad_norm": 0.3973132371902466, "learning_rate": 3.566981205101781e-05, "loss": 0.0111, "step": 6130 }, { "epoch": 5.643382352941177, "grad_norm": 0.35627076029777527, "learning_rate": 3.5511480151101556e-05, "loss": 0.0091, "step": 6140 }, { "epoch": 5.6525735294117645, "grad_norm": 0.2650625705718994, "learning_rate": 3.5353306695317104e-05, "loss": 0.009, "step": 6150 }, { "epoch": 5.661764705882353, "grad_norm": 0.3626503646373749, "learning_rate": 3.519529341342402e-05, "loss": 0.0125, "step": 6160 }, { "epoch": 5.670955882352941, "grad_norm": 0.38252466917037964, "learning_rate": 3.503744203343026e-05, "loss": 0.012, "step": 6170 }, { "epoch": 5.680147058823529, "grad_norm": 0.37305331230163574, "learning_rate": 3.487975428157318e-05, "loss": 0.0121, "step": 6180 }, { "epoch": 5.689338235294118, "grad_norm": 0.31039735674858093, "learning_rate": 3.472223188230083e-05, "loss": 0.0122, "step": 6190 }, { "epoch": 5.698529411764706, "grad_norm": 0.4086906909942627, "learning_rate": 3.4564876558252866e-05, "loss": 0.0106, "step": 6200 }, { "epoch": 5.707720588235294, "grad_norm": 0.3948252201080322, "learning_rate": 3.440769003024195e-05, "loss": 0.011, "step": 6210 }, { "epoch": 5.716911764705882, "grad_norm": 0.34615403413772583, "learning_rate": 3.425067401723477e-05, "loss": 0.0111, "step": 6220 }, { "epoch": 5.726102941176471, "grad_norm": 0.37720340490341187, "learning_rate": 3.409383023633325e-05, "loss": 0.0103, "step": 6230 }, { "epoch": 5.735294117647059, "grad_norm": 0.29859015345573425, "learning_rate": 3.3937160402755894e-05, "loss": 0.0108, "step": 6240 }, { "epoch": 5.744485294117647, "grad_norm": 0.3456728756427765, "learning_rate": 3.378066622981885e-05, "loss": 0.0129, "step": 6250 }, { "epoch": 5.7536764705882355, "grad_norm": 0.3022277355194092, "learning_rate": 3.362434942891738e-05, "loss": 0.0131, "step": 6260 }, { "epoch": 5.762867647058823, "grad_norm": 0.2379634976387024, "learning_rate": 3.346821170950693e-05, "loss": 0.0101, "step": 6270 }, { "epoch": 5.772058823529412, "grad_norm": 0.35081371665000916, "learning_rate": 3.3312254779084585e-05, "loss": 0.0112, "step": 6280 }, { "epoch": 5.78125, "grad_norm": 0.29423052072525024, "learning_rate": 3.315648034317039e-05, "loss": 0.0134, "step": 6290 }, { "epoch": 5.790441176470588, "grad_norm": 0.26410460472106934, "learning_rate": 3.3000890105288564e-05, "loss": 0.0099, "step": 6300 }, { "epoch": 5.799632352941177, "grad_norm": 0.3515719473361969, "learning_rate": 3.284548576694908e-05, "loss": 0.0108, "step": 6310 }, { "epoch": 5.8088235294117645, "grad_norm": 0.36386480927467346, "learning_rate": 3.2690269027628815e-05, "loss": 0.0092, "step": 6320 }, { "epoch": 5.818014705882353, "grad_norm": 0.4917903542518616, "learning_rate": 3.253524158475324e-05, "loss": 0.0103, "step": 6330 }, { "epoch": 5.827205882352941, "grad_norm": 0.3447128236293793, "learning_rate": 3.238040513367757e-05, "loss": 0.0112, "step": 6340 }, { "epoch": 5.836397058823529, "grad_norm": 0.38725847005844116, "learning_rate": 3.222576136766843e-05, "loss": 0.0093, "step": 6350 }, { "epoch": 5.845588235294118, "grad_norm": 0.38221919536590576, "learning_rate": 3.2071311977885324e-05, "loss": 0.0107, "step": 6360 }, { "epoch": 5.854779411764706, "grad_norm": 0.3518013060092926, "learning_rate": 3.191705865336197e-05, "loss": 0.0125, "step": 6370 }, { "epoch": 5.863970588235294, "grad_norm": 0.30311593413352966, "learning_rate": 3.1763003080988075e-05, "loss": 0.0097, "step": 6380 }, { "epoch": 5.873161764705882, "grad_norm": 0.2982126474380493, "learning_rate": 3.160914694549063e-05, "loss": 0.0109, "step": 6390 }, { "epoch": 5.882352941176471, "grad_norm": 0.26803097128868103, "learning_rate": 3.145549192941573e-05, "loss": 0.01, "step": 6400 }, { "epoch": 5.891544117647059, "grad_norm": 0.2977125346660614, "learning_rate": 3.130203971310999e-05, "loss": 0.0108, "step": 6410 }, { "epoch": 5.900735294117647, "grad_norm": 0.3927449584007263, "learning_rate": 3.114879197470225e-05, "loss": 0.0109, "step": 6420 }, { "epoch": 5.9099264705882355, "grad_norm": 0.3034214675426483, "learning_rate": 3.0995750390085285e-05, "loss": 0.0087, "step": 6430 }, { "epoch": 5.919117647058823, "grad_norm": 0.5464165210723877, "learning_rate": 3.084291663289728e-05, "loss": 0.0098, "step": 6440 }, { "epoch": 5.928308823529412, "grad_norm": 0.537919819355011, "learning_rate": 3.069029237450375e-05, "loss": 0.0083, "step": 6450 }, { "epoch": 5.9375, "grad_norm": 0.3033539652824402, "learning_rate": 3.053787928397911e-05, "loss": 0.0092, "step": 6460 }, { "epoch": 5.946691176470588, "grad_norm": 0.3915115296840668, "learning_rate": 3.0385679028088526e-05, "loss": 0.011, "step": 6470 }, { "epoch": 5.955882352941177, "grad_norm": 0.4144446551799774, "learning_rate": 3.023369327126959e-05, "loss": 0.0101, "step": 6480 }, { "epoch": 5.9650735294117645, "grad_norm": 0.2815667688846588, "learning_rate": 3.0081923675614198e-05, "loss": 0.0093, "step": 6490 }, { "epoch": 5.974264705882353, "grad_norm": 0.341545432806015, "learning_rate": 2.993037190085034e-05, "loss": 0.009, "step": 6500 }, { "epoch": 5.983455882352941, "grad_norm": 0.3925587236881256, "learning_rate": 2.977903960432392e-05, "loss": 0.0104, "step": 6510 }, { "epoch": 5.992647058823529, "grad_norm": 0.4770376980304718, "learning_rate": 2.9627928440980722e-05, "loss": 0.0131, "step": 6520 }, { "epoch": 6.001838235294118, "grad_norm": 0.33325669169425964, "learning_rate": 2.9477040063348183e-05, "loss": 0.0107, "step": 6530 }, { "epoch": 6.011029411764706, "grad_norm": 0.2520659565925598, "learning_rate": 2.9326376121517456e-05, "loss": 0.0117, "step": 6540 }, { "epoch": 6.020220588235294, "grad_norm": 0.2785213887691498, "learning_rate": 2.9175938263125236e-05, "loss": 0.0106, "step": 6550 }, { "epoch": 6.029411764705882, "grad_norm": 0.40615612268447876, "learning_rate": 2.9025728133335873e-05, "loss": 0.0107, "step": 6560 }, { "epoch": 6.038602941176471, "grad_norm": 0.35487574338912964, "learning_rate": 2.8875747374823288e-05, "loss": 0.0144, "step": 6570 }, { "epoch": 6.047794117647059, "grad_norm": 0.40217965841293335, "learning_rate": 2.872599762775298e-05, "loss": 0.0148, "step": 6580 }, { "epoch": 6.056985294117647, "grad_norm": 0.4580880403518677, "learning_rate": 2.857648052976425e-05, "loss": 0.0104, "step": 6590 }, { "epoch": 6.0661764705882355, "grad_norm": 0.36139973998069763, "learning_rate": 2.8427197715952047e-05, "loss": 0.0097, "step": 6600 }, { "epoch": 6.075367647058823, "grad_norm": 0.36743980646133423, "learning_rate": 2.8278150818849393e-05, "loss": 0.01, "step": 6610 }, { "epoch": 6.084558823529412, "grad_norm": 0.3188106417655945, "learning_rate": 2.812934146840922e-05, "loss": 0.0101, "step": 6620 }, { "epoch": 6.09375, "grad_norm": 0.3576509952545166, "learning_rate": 2.7980771291986764e-05, "loss": 0.0124, "step": 6630 }, { "epoch": 6.102941176470588, "grad_norm": 0.26821884512901306, "learning_rate": 2.783244191432167e-05, "loss": 0.0101, "step": 6640 }, { "epoch": 6.112132352941177, "grad_norm": 0.23525021970272064, "learning_rate": 2.768435495752022e-05, "loss": 0.009, "step": 6650 }, { "epoch": 6.1213235294117645, "grad_norm": 0.4137275218963623, "learning_rate": 2.753651204103771e-05, "loss": 0.0099, "step": 6660 }, { "epoch": 6.130514705882353, "grad_norm": 0.38588467240333557, "learning_rate": 2.7388914781660523e-05, "loss": 0.0098, "step": 6670 }, { "epoch": 6.139705882352941, "grad_norm": 0.3559839427471161, "learning_rate": 2.7241564793488693e-05, "loss": 0.0111, "step": 6680 }, { "epoch": 6.148897058823529, "grad_norm": 0.3329114019870758, "learning_rate": 2.7094463687918037e-05, "loss": 0.0098, "step": 6690 }, { "epoch": 6.158088235294118, "grad_norm": 0.27006974816322327, "learning_rate": 2.694761307362268e-05, "loss": 0.0093, "step": 6700 }, { "epoch": 6.167279411764706, "grad_norm": 0.3701556622982025, "learning_rate": 2.6801014556537467e-05, "loss": 0.0094, "step": 6710 }, { "epoch": 6.176470588235294, "grad_norm": 0.38370680809020996, "learning_rate": 2.6654669739840243e-05, "loss": 0.0095, "step": 6720 }, { "epoch": 6.185661764705882, "grad_norm": 0.2997688353061676, "learning_rate": 2.650858022393451e-05, "loss": 0.0105, "step": 6730 }, { "epoch": 6.194852941176471, "grad_norm": 0.3103492558002472, "learning_rate": 2.6362747606431747e-05, "loss": 0.012, "step": 6740 }, { "epoch": 6.204044117647059, "grad_norm": 0.2602502107620239, "learning_rate": 2.6217173482134172e-05, "loss": 0.0091, "step": 6750 }, { "epoch": 6.213235294117647, "grad_norm": 0.2754170596599579, "learning_rate": 2.6071859443017044e-05, "loss": 0.0091, "step": 6760 }, { "epoch": 6.2224264705882355, "grad_norm": 0.32339054346084595, "learning_rate": 2.5926807078211414e-05, "loss": 0.0082, "step": 6770 }, { "epoch": 6.231617647058823, "grad_norm": 0.3423077464103699, "learning_rate": 2.5782017973986728e-05, "loss": 0.009, "step": 6780 }, { "epoch": 6.240808823529412, "grad_norm": 0.47386062145233154, "learning_rate": 2.5637493713733374e-05, "loss": 0.0094, "step": 6790 }, { "epoch": 6.25, "grad_norm": 0.3446521461009979, "learning_rate": 2.549323587794559e-05, "loss": 0.0127, "step": 6800 }, { "epoch": 6.259191176470588, "grad_norm": 0.3277866244316101, "learning_rate": 2.5349246044203895e-05, "loss": 0.0078, "step": 6810 }, { "epoch": 6.268382352941177, "grad_norm": 0.2986924350261688, "learning_rate": 2.520552578715808e-05, "loss": 0.0093, "step": 6820 }, { "epoch": 6.2775735294117645, "grad_norm": 0.30999377369880676, "learning_rate": 2.506207667850981e-05, "loss": 0.0081, "step": 6830 }, { "epoch": 6.286764705882353, "grad_norm": 0.34954744577407837, "learning_rate": 2.4918900286995555e-05, "loss": 0.0096, "step": 6840 }, { "epoch": 6.295955882352941, "grad_norm": 0.38755038380622864, "learning_rate": 2.4775998178369458e-05, "loss": 0.0078, "step": 6850 }, { "epoch": 6.305147058823529, "grad_norm": 0.31793591380119324, "learning_rate": 2.4633371915386017e-05, "loss": 0.0109, "step": 6860 }, { "epoch": 6.314338235294118, "grad_norm": 0.3524736166000366, "learning_rate": 2.4491023057783235e-05, "loss": 0.0112, "step": 6870 }, { "epoch": 6.323529411764706, "grad_norm": 0.3981841802597046, "learning_rate": 2.4348953162265375e-05, "loss": 0.0133, "step": 6880 }, { "epoch": 6.332720588235294, "grad_norm": 0.2546362280845642, "learning_rate": 2.420716378248607e-05, "loss": 0.01, "step": 6890 }, { "epoch": 6.341911764705882, "grad_norm": 0.3159024119377136, "learning_rate": 2.4065656469031266e-05, "loss": 0.0092, "step": 6900 }, { "epoch": 6.351102941176471, "grad_norm": 0.37902170419692993, "learning_rate": 2.3924432769402268e-05, "loss": 0.0098, "step": 6910 }, { "epoch": 6.360294117647059, "grad_norm": 0.2722319960594177, "learning_rate": 2.3783494227998844e-05, "loss": 0.0142, "step": 6920 }, { "epoch": 6.369485294117647, "grad_norm": 0.2230018824338913, "learning_rate": 2.3642842386102264e-05, "loss": 0.0075, "step": 6930 }, { "epoch": 6.3786764705882355, "grad_norm": 0.3412310481071472, "learning_rate": 2.3502478781858567e-05, "loss": 0.0076, "step": 6940 }, { "epoch": 6.387867647058823, "grad_norm": 0.44032323360443115, "learning_rate": 2.3362404950261628e-05, "loss": 0.0112, "step": 6950 }, { "epoch": 6.397058823529412, "grad_norm": 0.45974090695381165, "learning_rate": 2.3222622423136458e-05, "loss": 0.0094, "step": 6960 }, { "epoch": 6.40625, "grad_norm": 0.26098760962486267, "learning_rate": 2.3083132729122332e-05, "loss": 0.0111, "step": 6970 }, { "epoch": 6.415441176470588, "grad_norm": 0.4639066755771637, "learning_rate": 2.294393739365621e-05, "loss": 0.0089, "step": 6980 }, { "epoch": 6.424632352941177, "grad_norm": 0.29276183247566223, "learning_rate": 2.2805037938956e-05, "loss": 0.0111, "step": 6990 }, { "epoch": 6.4338235294117645, "grad_norm": 0.30953171849250793, "learning_rate": 2.266643588400386e-05, "loss": 0.0083, "step": 7000 }, { "epoch": 6.443014705882353, "grad_norm": 0.20741379261016846, "learning_rate": 2.252813274452969e-05, "loss": 0.0092, "step": 7010 }, { "epoch": 6.452205882352941, "grad_norm": 0.24684885144233704, "learning_rate": 2.2390130032994427e-05, "loss": 0.0073, "step": 7020 }, { "epoch": 6.461397058823529, "grad_norm": 0.291708767414093, "learning_rate": 2.2252429258573633e-05, "loss": 0.0089, "step": 7030 }, { "epoch": 6.470588235294118, "grad_norm": 0.23906491696834564, "learning_rate": 2.2115031927140904e-05, "loss": 0.01, "step": 7040 }, { "epoch": 6.479779411764706, "grad_norm": 0.33487898111343384, "learning_rate": 2.1977939541251463e-05, "loss": 0.0139, "step": 7050 }, { "epoch": 6.488970588235294, "grad_norm": 0.26053792238235474, "learning_rate": 2.1841153600125684e-05, "loss": 0.0085, "step": 7060 }, { "epoch": 6.498161764705882, "grad_norm": 0.28365546464920044, "learning_rate": 2.170467559963267e-05, "loss": 0.0107, "step": 7070 }, { "epoch": 6.507352941176471, "grad_norm": 0.32671281695365906, "learning_rate": 2.1568507032273982e-05, "loss": 0.0074, "step": 7080 }, { "epoch": 6.516544117647059, "grad_norm": 0.2789180874824524, "learning_rate": 2.1432649387167264e-05, "loss": 0.0081, "step": 7090 }, { "epoch": 6.525735294117647, "grad_norm": 0.25455355644226074, "learning_rate": 2.1297104150029973e-05, "loss": 0.0086, "step": 7100 }, { "epoch": 6.5349264705882355, "grad_norm": 0.3030979335308075, "learning_rate": 2.116187280316307e-05, "loss": 0.0093, "step": 7110 }, { "epoch": 6.544117647058823, "grad_norm": 0.26105305552482605, "learning_rate": 2.1026956825434908e-05, "loss": 0.0092, "step": 7120 }, { "epoch": 6.553308823529412, "grad_norm": 0.2904008626937866, "learning_rate": 2.0892357692265017e-05, "loss": 0.0086, "step": 7130 }, { "epoch": 6.5625, "grad_norm": 0.29898276925086975, "learning_rate": 2.0758076875607947e-05, "loss": 0.0097, "step": 7140 }, { "epoch": 6.571691176470588, "grad_norm": 0.25379547476768494, "learning_rate": 2.0624115843937207e-05, "loss": 0.0109, "step": 7150 }, { "epoch": 6.580882352941177, "grad_norm": 0.26838940382003784, "learning_rate": 2.0490476062229157e-05, "loss": 0.0106, "step": 7160 }, { "epoch": 6.5900735294117645, "grad_norm": 0.24734647572040558, "learning_rate": 2.035715899194704e-05, "loss": 0.0084, "step": 7170 }, { "epoch": 6.599264705882353, "grad_norm": 0.3488079309463501, "learning_rate": 2.022416609102499e-05, "loss": 0.0091, "step": 7180 }, { "epoch": 6.608455882352941, "grad_norm": 0.3336898982524872, "learning_rate": 2.009149881385205e-05, "loss": 0.0127, "step": 7190 }, { "epoch": 6.617647058823529, "grad_norm": 0.25673535466194153, "learning_rate": 1.995915861125634e-05, "loss": 0.0086, "step": 7200 }, { "epoch": 6.626838235294118, "grad_norm": 0.2218119502067566, "learning_rate": 1.9827146930489065e-05, "loss": 0.0069, "step": 7210 }, { "epoch": 6.636029411764706, "grad_norm": 0.38927024602890015, "learning_rate": 1.9695465215208848e-05, "loss": 0.0106, "step": 7220 }, { "epoch": 6.645220588235294, "grad_norm": 0.3131321668624878, "learning_rate": 1.9564114905465813e-05, "loss": 0.0078, "step": 7230 }, { "epoch": 6.654411764705882, "grad_norm": 0.22971443831920624, "learning_rate": 1.9433097437685936e-05, "loss": 0.0084, "step": 7240 }, { "epoch": 6.663602941176471, "grad_norm": 0.31655681133270264, "learning_rate": 1.930241424465521e-05, "loss": 0.0073, "step": 7250 }, { "epoch": 6.672794117647059, "grad_norm": 0.3079403340816498, "learning_rate": 1.9172066755504115e-05, "loss": 0.0085, "step": 7260 }, { "epoch": 6.681985294117647, "grad_norm": 0.27573585510253906, "learning_rate": 1.9042056395691914e-05, "loss": 0.0083, "step": 7270 }, { "epoch": 6.6911764705882355, "grad_norm": 0.18852144479751587, "learning_rate": 1.8912384586991066e-05, "loss": 0.0076, "step": 7280 }, { "epoch": 6.700367647058823, "grad_norm": 0.2407255321741104, "learning_rate": 1.8783052747471717e-05, "loss": 0.0086, "step": 7290 }, { "epoch": 6.709558823529412, "grad_norm": 0.23558451235294342, "learning_rate": 1.865406229148611e-05, "loss": 0.0097, "step": 7300 }, { "epoch": 6.71875, "grad_norm": 0.2761400043964386, "learning_rate": 1.8525414629653233e-05, "loss": 0.0119, "step": 7310 }, { "epoch": 6.727941176470588, "grad_norm": 0.36148083209991455, "learning_rate": 1.8397111168843255e-05, "loss": 0.0077, "step": 7320 }, { "epoch": 6.737132352941177, "grad_norm": 0.2554199993610382, "learning_rate": 1.8269153312162323e-05, "loss": 0.0099, "step": 7330 }, { "epoch": 6.7463235294117645, "grad_norm": 0.27136483788490295, "learning_rate": 1.8141542458937054e-05, "loss": 0.0087, "step": 7340 }, { "epoch": 6.755514705882353, "grad_norm": 0.23154033720493317, "learning_rate": 1.8014280004699268e-05, "loss": 0.0077, "step": 7350 }, { "epoch": 6.764705882352941, "grad_norm": 0.307264506816864, "learning_rate": 1.788736734117078e-05, "loss": 0.0079, "step": 7360 }, { "epoch": 6.773897058823529, "grad_norm": 0.2880013585090637, "learning_rate": 1.7760805856248152e-05, "loss": 0.0085, "step": 7370 }, { "epoch": 6.783088235294118, "grad_norm": 0.28776559233665466, "learning_rate": 1.7634596933987518e-05, "loss": 0.0086, "step": 7380 }, { "epoch": 6.792279411764706, "grad_norm": 0.298885703086853, "learning_rate": 1.7508741954589404e-05, "loss": 0.011, "step": 7390 }, { "epoch": 6.801470588235294, "grad_norm": 0.2450408786535263, "learning_rate": 1.7383242294383717e-05, "loss": 0.0081, "step": 7400 }, { "epoch": 6.810661764705882, "grad_norm": 0.24446910619735718, "learning_rate": 1.7258099325814632e-05, "loss": 0.0072, "step": 7410 }, { "epoch": 6.819852941176471, "grad_norm": 0.23296265304088593, "learning_rate": 1.7133314417425594e-05, "loss": 0.0098, "step": 7420 }, { "epoch": 6.829044117647059, "grad_norm": 0.30054357647895813, "learning_rate": 1.7008888933844408e-05, "loss": 0.0084, "step": 7430 }, { "epoch": 6.838235294117647, "grad_norm": 0.2629185616970062, "learning_rate": 1.6884824235768172e-05, "loss": 0.0104, "step": 7440 }, { "epoch": 6.8474264705882355, "grad_norm": 0.2844895124435425, "learning_rate": 1.6761121679948592e-05, "loss": 0.0072, "step": 7450 }, { "epoch": 6.856617647058823, "grad_norm": 0.33059990406036377, "learning_rate": 1.663778261917695e-05, "loss": 0.0101, "step": 7460 }, { "epoch": 6.865808823529412, "grad_norm": 0.2688612937927246, "learning_rate": 1.651480840226952e-05, "loss": 0.0081, "step": 7470 }, { "epoch": 6.875, "grad_norm": 0.2447384148836136, "learning_rate": 1.639220037405258e-05, "loss": 0.0098, "step": 7480 }, { "epoch": 6.884191176470588, "grad_norm": 0.19501954317092896, "learning_rate": 1.6269959875347906e-05, "loss": 0.0095, "step": 7490 }, { "epoch": 6.893382352941177, "grad_norm": 0.2788120210170746, "learning_rate": 1.614808824295802e-05, "loss": 0.0078, "step": 7500 }, { "epoch": 6.9025735294117645, "grad_norm": 0.3465050756931305, "learning_rate": 1.602658680965152e-05, "loss": 0.0071, "step": 7510 }, { "epoch": 6.911764705882353, "grad_norm": 0.2406514286994934, "learning_rate": 1.5905456904148686e-05, "loss": 0.01, "step": 7520 }, { "epoch": 6.920955882352941, "grad_norm": 0.2917358875274658, "learning_rate": 1.57846998511067e-05, "loss": 0.0092, "step": 7530 }, { "epoch": 6.930147058823529, "grad_norm": 0.3689867854118347, "learning_rate": 1.566431697110538e-05, "loss": 0.0075, "step": 7540 }, { "epoch": 6.939338235294118, "grad_norm": 0.29593798518180847, "learning_rate": 1.554430958063259e-05, "loss": 0.0072, "step": 7550 }, { "epoch": 6.948529411764706, "grad_norm": 0.27510392665863037, "learning_rate": 1.5424678992069912e-05, "loss": 0.0071, "step": 7560 }, { "epoch": 6.957720588235294, "grad_norm": 0.2821337878704071, "learning_rate": 1.5305426513678362e-05, "loss": 0.0094, "step": 7570 }, { "epoch": 6.966911764705882, "grad_norm": 0.3060770630836487, "learning_rate": 1.518655344958388e-05, "loss": 0.0096, "step": 7580 }, { "epoch": 6.976102941176471, "grad_norm": 0.4188516438007355, "learning_rate": 1.5068061099763275e-05, "loss": 0.0088, "step": 7590 }, { "epoch": 6.985294117647059, "grad_norm": 0.2774643003940582, "learning_rate": 1.494995076002988e-05, "loss": 0.0089, "step": 7600 }, { "epoch": 6.994485294117647, "grad_norm": 0.3032972812652588, "learning_rate": 1.4832223722019456e-05, "loss": 0.007, "step": 7610 }, { "epoch": 7.0036764705882355, "grad_norm": 0.21218299865722656, "learning_rate": 1.4714881273176035e-05, "loss": 0.0072, "step": 7620 }, { "epoch": 7.012867647058823, "grad_norm": 0.31260043382644653, "learning_rate": 1.4597924696737835e-05, "loss": 0.0077, "step": 7630 }, { "epoch": 7.022058823529412, "grad_norm": 0.26134800910949707, "learning_rate": 1.4481355271723252e-05, "loss": 0.0077, "step": 7640 }, { "epoch": 7.03125, "grad_norm": 0.19740882515907288, "learning_rate": 1.4365174272916809e-05, "loss": 0.0093, "step": 7650 }, { "epoch": 7.040441176470588, "grad_norm": 0.3442723751068115, "learning_rate": 1.4249382970855319e-05, "loss": 0.0078, "step": 7660 }, { "epoch": 7.049632352941177, "grad_norm": 0.2376275360584259, "learning_rate": 1.4133982631813903e-05, "loss": 0.0083, "step": 7670 }, { "epoch": 7.0588235294117645, "grad_norm": 0.1998441219329834, "learning_rate": 1.4018974517792194e-05, "loss": 0.0069, "step": 7680 }, { "epoch": 7.068014705882353, "grad_norm": 0.33761832118034363, "learning_rate": 1.390435988650048e-05, "loss": 0.0074, "step": 7690 }, { "epoch": 7.077205882352941, "grad_norm": 0.2662760615348816, "learning_rate": 1.3790139991346006e-05, "loss": 0.0074, "step": 7700 }, { "epoch": 7.086397058823529, "grad_norm": 0.43220290541648865, "learning_rate": 1.367631608141926e-05, "loss": 0.0078, "step": 7710 }, { "epoch": 7.095588235294118, "grad_norm": 0.257103830575943, "learning_rate": 1.3562889401480278e-05, "loss": 0.0072, "step": 7720 }, { "epoch": 7.104779411764706, "grad_norm": 0.2577866017818451, "learning_rate": 1.3449861191945074e-05, "loss": 0.0092, "step": 7730 }, { "epoch": 7.113970588235294, "grad_norm": 0.25710979104042053, "learning_rate": 1.3337232688872009e-05, "loss": 0.0076, "step": 7740 }, { "epoch": 7.123161764705882, "grad_norm": 0.35710468888282776, "learning_rate": 1.3225005123948364e-05, "loss": 0.0079, "step": 7750 }, { "epoch": 7.132352941176471, "grad_norm": 0.2810736298561096, "learning_rate": 1.311317972447681e-05, "loss": 0.0074, "step": 7760 }, { "epoch": 7.141544117647059, "grad_norm": 0.26390713453292847, "learning_rate": 1.3001757713361996e-05, "loss": 0.0058, "step": 7770 }, { "epoch": 7.150735294117647, "grad_norm": 0.36639297008514404, "learning_rate": 1.2890740309097204e-05, "loss": 0.0084, "step": 7780 }, { "epoch": 7.1599264705882355, "grad_norm": 0.38648366928100586, "learning_rate": 1.2780128725750944e-05, "loss": 0.0071, "step": 7790 }, { "epoch": 7.169117647058823, "grad_norm": 0.28881534934043884, "learning_rate": 1.266992417295379e-05, "loss": 0.0078, "step": 7800 }, { "epoch": 7.178308823529412, "grad_norm": 0.24816535413265228, "learning_rate": 1.2560127855885073e-05, "loss": 0.0084, "step": 7810 }, { "epoch": 7.1875, "grad_norm": 0.3702505826950073, "learning_rate": 1.2450740975259745e-05, "loss": 0.011, "step": 7820 }, { "epoch": 7.196691176470588, "grad_norm": 0.334357887506485, "learning_rate": 1.234176472731517e-05, "loss": 0.0117, "step": 7830 }, { "epoch": 7.205882352941177, "grad_norm": 0.35254141688346863, "learning_rate": 1.2233200303798158e-05, "loss": 0.0083, "step": 7840 }, { "epoch": 7.2150735294117645, "grad_norm": 0.3218490183353424, "learning_rate": 1.2125048891951846e-05, "loss": 0.0069, "step": 7850 }, { "epoch": 7.224264705882353, "grad_norm": 0.32505226135253906, "learning_rate": 1.2017311674502745e-05, "loss": 0.0065, "step": 7860 }, { "epoch": 7.233455882352941, "grad_norm": 0.18104083836078644, "learning_rate": 1.1909989829647822e-05, "loss": 0.0078, "step": 7870 }, { "epoch": 7.242647058823529, "grad_norm": 0.49740689992904663, "learning_rate": 1.1803084531041553e-05, "loss": 0.0081, "step": 7880 }, { "epoch": 7.251838235294118, "grad_norm": 0.2313297986984253, "learning_rate": 1.1696596947783162e-05, "loss": 0.0074, "step": 7890 }, { "epoch": 7.261029411764706, "grad_norm": 0.2114240825176239, "learning_rate": 1.1590528244403803e-05, "loss": 0.007, "step": 7900 }, { "epoch": 7.270220588235294, "grad_norm": 0.32324767112731934, "learning_rate": 1.148487958085382e-05, "loss": 0.0094, "step": 7910 }, { "epoch": 7.279411764705882, "grad_norm": 0.20598939061164856, "learning_rate": 1.1379652112490086e-05, "loss": 0.0075, "step": 7920 }, { "epoch": 7.288602941176471, "grad_norm": 0.31917813420295715, "learning_rate": 1.1274846990063315e-05, "loss": 0.0085, "step": 7930 }, { "epoch": 7.297794117647059, "grad_norm": 0.2507917881011963, "learning_rate": 1.117046535970554e-05, "loss": 0.0075, "step": 7940 }, { "epoch": 7.306985294117647, "grad_norm": 0.24764913320541382, "learning_rate": 1.106650836291755e-05, "loss": 0.012, "step": 7950 }, { "epoch": 7.3161764705882355, "grad_norm": 0.28983357548713684, "learning_rate": 1.0962977136556418e-05, "loss": 0.0119, "step": 7960 }, { "epoch": 7.325367647058823, "grad_norm": 0.4287269711494446, "learning_rate": 1.0859872812823024e-05, "loss": 0.0081, "step": 7970 }, { "epoch": 7.334558823529412, "grad_norm": 0.3013823628425598, "learning_rate": 1.0757196519249747e-05, "loss": 0.014, "step": 7980 }, { "epoch": 7.34375, "grad_norm": 0.23420350253582, "learning_rate": 1.0654949378688077e-05, "loss": 0.0082, "step": 7990 }, { "epoch": 7.352941176470588, "grad_norm": 0.2517244517803192, "learning_rate": 1.0553132509296376e-05, "loss": 0.0075, "step": 8000 }, { "epoch": 7.362132352941177, "grad_norm": 0.23548376560211182, "learning_rate": 1.0451747024527613e-05, "loss": 0.0086, "step": 8010 }, { "epoch": 7.3713235294117645, "grad_norm": 0.2219359278678894, "learning_rate": 1.0350794033117189e-05, "loss": 0.0071, "step": 8020 }, { "epoch": 7.380514705882353, "grad_norm": 0.19729235768318176, "learning_rate": 1.0250274639070856e-05, "loss": 0.0082, "step": 8030 }, { "epoch": 7.389705882352941, "grad_norm": 0.21196654438972473, "learning_rate": 1.0150189941652599e-05, "loss": 0.0066, "step": 8040 }, { "epoch": 7.398897058823529, "grad_norm": 0.242101788520813, "learning_rate": 1.0050541035372635e-05, "loss": 0.0061, "step": 8050 }, { "epoch": 7.408088235294118, "grad_norm": 0.2300061136484146, "learning_rate": 9.951329009975458e-06, "loss": 0.0078, "step": 8060 }, { "epoch": 7.417279411764706, "grad_norm": 0.20957432687282562, "learning_rate": 9.852554950427845e-06, "loss": 0.0064, "step": 8070 }, { "epoch": 7.426470588235294, "grad_norm": 0.24981945753097534, "learning_rate": 9.754219936907105e-06, "loss": 0.0069, "step": 8080 }, { "epoch": 7.435661764705882, "grad_norm": 0.2942031919956207, "learning_rate": 9.656325044789194e-06, "loss": 0.0064, "step": 8090 }, { "epoch": 7.444852941176471, "grad_norm": 0.2001383751630783, "learning_rate": 9.55887134463697e-06, "loss": 0.0064, "step": 8100 }, { "epoch": 7.454044117647059, "grad_norm": 0.20538246631622314, "learning_rate": 9.461859902188475e-06, "loss": 0.0077, "step": 8110 }, { "epoch": 7.463235294117647, "grad_norm": 0.15733985602855682, "learning_rate": 9.365291778345303e-06, "loss": 0.0071, "step": 8120 }, { "epoch": 7.4724264705882355, "grad_norm": 0.16448861360549927, "learning_rate": 9.269168029160991e-06, "loss": 0.0058, "step": 8130 }, { "epoch": 7.481617647058823, "grad_norm": 0.39105620980262756, "learning_rate": 9.173489705829447e-06, "loss": 0.0083, "step": 8140 }, { "epoch": 7.490808823529412, "grad_norm": 0.2678716480731964, "learning_rate": 9.078257854673516e-06, "loss": 0.007, "step": 8150 }, { "epoch": 7.5, "grad_norm": 0.27058714628219604, "learning_rate": 8.983473517133429e-06, "loss": 0.007, "step": 8160 }, { "epoch": 7.509191176470588, "grad_norm": 0.25341692566871643, "learning_rate": 8.889137729755537e-06, "loss": 0.0067, "step": 8170 }, { "epoch": 7.518382352941177, "grad_norm": 0.17718660831451416, "learning_rate": 8.79525152418087e-06, "loss": 0.0082, "step": 8180 }, { "epoch": 7.5275735294117645, "grad_norm": 0.2768092155456543, "learning_rate": 8.701815927133961e-06, "loss": 0.0072, "step": 8190 }, { "epoch": 7.536764705882353, "grad_norm": 0.2564014196395874, "learning_rate": 8.608831960411534e-06, "loss": 0.0074, "step": 8200 }, { "epoch": 7.545955882352941, "grad_norm": 0.217429056763649, "learning_rate": 8.516300640871321e-06, "loss": 0.0085, "step": 8210 }, { "epoch": 7.555147058823529, "grad_norm": 0.1757505238056183, "learning_rate": 8.424222980421038e-06, "loss": 0.0081, "step": 8220 }, { "epoch": 7.564338235294118, "grad_norm": 0.32166725397109985, "learning_rate": 8.332599986007184e-06, "loss": 0.008, "step": 8230 }, { "epoch": 7.573529411764706, "grad_norm": 0.2559928297996521, "learning_rate": 8.241432659604203e-06, "loss": 0.0063, "step": 8240 }, { "epoch": 7.582720588235294, "grad_norm": 0.21301086246967316, "learning_rate": 8.150721998203331e-06, "loss": 0.0063, "step": 8250 }, { "epoch": 7.591911764705882, "grad_norm": 0.3428524136543274, "learning_rate": 8.06046899380184e-06, "loss": 0.0066, "step": 8260 }, { "epoch": 7.601102941176471, "grad_norm": 0.24014942348003387, "learning_rate": 7.970674633392133e-06, "loss": 0.0113, "step": 8270 }, { "epoch": 7.610294117647059, "grad_norm": 0.2688608169555664, "learning_rate": 7.881339898950924e-06, "loss": 0.0072, "step": 8280 }, { "epoch": 7.619485294117647, "grad_norm": 0.25965574383735657, "learning_rate": 7.792465767428597e-06, "loss": 0.0074, "step": 8290 }, { "epoch": 7.6286764705882355, "grad_norm": 0.2161405235528946, "learning_rate": 7.704053210738376e-06, "loss": 0.0058, "step": 8300 }, { "epoch": 7.637867647058823, "grad_norm": 0.39968928694725037, "learning_rate": 7.6161031957458494e-06, "loss": 0.01, "step": 8310 }, { "epoch": 7.647058823529412, "grad_norm": 0.20113994181156158, "learning_rate": 7.5286166842582605e-06, "loss": 0.0073, "step": 8320 }, { "epoch": 7.65625, "grad_norm": 0.31141898036003113, "learning_rate": 7.4415946330140814e-06, "loss": 0.0072, "step": 8330 }, { "epoch": 7.665441176470588, "grad_norm": 0.25661611557006836, "learning_rate": 7.3550379936725644e-06, "loss": 0.0065, "step": 8340 }, { "epoch": 7.674632352941177, "grad_norm": 0.2811231315135956, "learning_rate": 7.2689477128032035e-06, "loss": 0.0073, "step": 8350 }, { "epoch": 7.6838235294117645, "grad_norm": 0.19742384552955627, "learning_rate": 7.183324731875551e-06, "loss": 0.0063, "step": 8360 }, { "epoch": 7.693014705882353, "grad_norm": 0.18516254425048828, "learning_rate": 7.098169987248782e-06, "loss": 0.005, "step": 8370 }, { "epoch": 7.702205882352941, "grad_norm": 0.19681328535079956, "learning_rate": 7.013484410161553e-06, "loss": 0.0078, "step": 8380 }, { "epoch": 7.711397058823529, "grad_norm": 0.1999165415763855, "learning_rate": 6.92926892672176e-06, "loss": 0.0076, "step": 8390 }, { "epoch": 7.720588235294118, "grad_norm": 0.17462976276874542, "learning_rate": 6.845524457896446e-06, "loss": 0.0058, "step": 8400 }, { "epoch": 7.729779411764706, "grad_norm": 0.31693708896636963, "learning_rate": 6.7622519195017165e-06, "loss": 0.0068, "step": 8410 }, { "epoch": 7.738970588235294, "grad_norm": 0.1995631456375122, "learning_rate": 6.679452222192684e-06, "loss": 0.0065, "step": 8420 }, { "epoch": 7.748161764705882, "grad_norm": 0.25386086106300354, "learning_rate": 6.597126271453579e-06, "loss": 0.0064, "step": 8430 }, { "epoch": 7.757352941176471, "grad_norm": 0.21546174585819244, "learning_rate": 6.51527496758782e-06, "loss": 0.0062, "step": 8440 }, { "epoch": 7.766544117647059, "grad_norm": 0.21080993115901947, "learning_rate": 6.433899205708155e-06, "loss": 0.0061, "step": 8450 }, { "epoch": 7.775735294117647, "grad_norm": 0.19203387200832367, "learning_rate": 6.352999875726856e-06, "loss": 0.0057, "step": 8460 }, { "epoch": 7.7849264705882355, "grad_norm": 0.19412247836589813, "learning_rate": 6.272577862346052e-06, "loss": 0.0062, "step": 8470 }, { "epoch": 7.794117647058823, "grad_norm": 0.28905171155929565, "learning_rate": 6.192634045047996e-06, "loss": 0.0108, "step": 8480 }, { "epoch": 7.803308823529412, "grad_norm": 0.2105361670255661, "learning_rate": 6.113169298085458e-06, "loss": 0.0083, "step": 8490 }, { "epoch": 7.8125, "grad_norm": 0.2784077823162079, "learning_rate": 6.034184490472195e-06, "loss": 0.0057, "step": 8500 }, { "epoch": 7.821691176470588, "grad_norm": 0.34908047318458557, "learning_rate": 5.955680485973386e-06, "loss": 0.007, "step": 8510 }, { "epoch": 7.830882352941177, "grad_norm": 0.2582102417945862, "learning_rate": 5.877658143096265e-06, "loss": 0.0063, "step": 8520 }, { "epoch": 7.8400735294117645, "grad_norm": 0.1995401680469513, "learning_rate": 5.800118315080661e-06, "loss": 0.0058, "step": 8530 }, { "epoch": 7.849264705882353, "grad_norm": 0.29254844784736633, "learning_rate": 5.723061849889716e-06, "loss": 0.0059, "step": 8540 }, { "epoch": 7.858455882352941, "grad_norm": 0.36842775344848633, "learning_rate": 5.646489590200604e-06, "loss": 0.0073, "step": 8550 }, { "epoch": 7.867647058823529, "grad_norm": 0.276644229888916, "learning_rate": 5.570402373395256e-06, "loss": 0.0059, "step": 8560 }, { "epoch": 7.876838235294118, "grad_norm": 0.21737320721149445, "learning_rate": 5.494801031551305e-06, "loss": 0.0079, "step": 8570 }, { "epoch": 7.886029411764706, "grad_norm": 0.174285426735878, "learning_rate": 5.41968639143291e-06, "loss": 0.006, "step": 8580 }, { "epoch": 7.895220588235294, "grad_norm": 0.20091643929481506, "learning_rate": 5.345059274481751e-06, "loss": 0.0059, "step": 8590 }, { "epoch": 7.904411764705882, "grad_norm": 0.24660427868366241, "learning_rate": 5.270920496808002e-06, "loss": 0.0077, "step": 8600 }, { "epoch": 7.913602941176471, "grad_norm": 0.2320578247308731, "learning_rate": 5.1972708691814695e-06, "loss": 0.0075, "step": 8610 }, { "epoch": 7.922794117647059, "grad_norm": 0.22498774528503418, "learning_rate": 5.124111197022674e-06, "loss": 0.0073, "step": 8620 }, { "epoch": 7.931985294117647, "grad_norm": 0.33274200558662415, "learning_rate": 5.051442280394081e-06, "loss": 0.0061, "step": 8630 }, { "epoch": 7.9411764705882355, "grad_norm": 0.2908051311969757, "learning_rate": 4.979264913991322e-06, "loss": 0.006, "step": 8640 }, { "epoch": 7.950367647058823, "grad_norm": 0.1951013058423996, "learning_rate": 4.907579887134489e-06, "loss": 0.008, "step": 8650 }, { "epoch": 7.959558823529412, "grad_norm": 0.3692641854286194, "learning_rate": 4.836387983759572e-06, "loss": 0.0062, "step": 8660 }, { "epoch": 7.96875, "grad_norm": 0.2372119426727295, "learning_rate": 4.765689982409816e-06, "loss": 0.0071, "step": 8670 }, { "epoch": 7.977941176470588, "grad_norm": 0.23715977370738983, "learning_rate": 4.695486656227233e-06, "loss": 0.0085, "step": 8680 }, { "epoch": 7.987132352941177, "grad_norm": 0.1801896095275879, "learning_rate": 4.625778772944156e-06, "loss": 0.006, "step": 8690 }, { "epoch": 7.9963235294117645, "grad_norm": 0.18388265371322632, "learning_rate": 4.556567094874825e-06, "loss": 0.0072, "step": 8700 }, { "epoch": 8.005514705882353, "grad_norm": 0.19362954795360565, "learning_rate": 4.487852378907059e-06, "loss": 0.0051, "step": 8710 }, { "epoch": 8.014705882352942, "grad_norm": 0.1550358682870865, "learning_rate": 4.419635376493986e-06, "loss": 0.0063, "step": 8720 }, { "epoch": 8.023897058823529, "grad_norm": 0.17197181284427643, "learning_rate": 4.351916833645825e-06, "loss": 0.0069, "step": 8730 }, { "epoch": 8.033088235294118, "grad_norm": 0.268187552690506, "learning_rate": 4.284697490921691e-06, "loss": 0.0066, "step": 8740 }, { "epoch": 8.042279411764707, "grad_norm": 0.20384307205677032, "learning_rate": 4.2179780834215585e-06, "loss": 0.0071, "step": 8750 }, { "epoch": 8.051470588235293, "grad_norm": 0.2538691461086273, "learning_rate": 4.151759340778178e-06, "loss": 0.0072, "step": 8760 }, { "epoch": 8.060661764705882, "grad_norm": 0.19308795034885406, "learning_rate": 4.086041987149109e-06, "loss": 0.0059, "step": 8770 }, { "epoch": 8.069852941176471, "grad_norm": 0.2545320391654968, "learning_rate": 4.020826741208811e-06, "loss": 0.0077, "step": 8780 }, { "epoch": 8.079044117647058, "grad_norm": 0.2672986388206482, "learning_rate": 3.956114316140746e-06, "loss": 0.0076, "step": 8790 }, { "epoch": 8.088235294117647, "grad_norm": 0.3118535280227661, "learning_rate": 3.891905419629643e-06, "loss": 0.0065, "step": 8800 }, { "epoch": 8.097426470588236, "grad_norm": 0.43252572417259216, "learning_rate": 3.8282007538536946e-06, "loss": 0.0084, "step": 8810 }, { "epoch": 8.106617647058824, "grad_norm": 0.25299420952796936, "learning_rate": 3.7650010154769265e-06, "loss": 0.0076, "step": 8820 }, { "epoch": 8.115808823529411, "grad_norm": 0.24529822170734406, "learning_rate": 3.7023068956415608e-06, "loss": 0.0064, "step": 8830 }, { "epoch": 8.125, "grad_norm": 0.15536293387413025, "learning_rate": 3.6401190799604303e-06, "loss": 0.0056, "step": 8840 }, { "epoch": 8.134191176470589, "grad_norm": 0.25395727157592773, "learning_rate": 3.578438248509536e-06, "loss": 0.0074, "step": 8850 }, { "epoch": 8.143382352941176, "grad_norm": 0.26248422265052795, "learning_rate": 3.5172650758205583e-06, "loss": 0.0072, "step": 8860 }, { "epoch": 8.152573529411764, "grad_norm": 0.2564133107662201, "learning_rate": 3.45660023087353e-06, "loss": 0.0056, "step": 8870 }, { "epoch": 8.161764705882353, "grad_norm": 0.18524330854415894, "learning_rate": 3.3964443770894528e-06, "loss": 0.0071, "step": 8880 }, { "epoch": 8.170955882352942, "grad_norm": 0.26521211862564087, "learning_rate": 3.3367981723231245e-06, "loss": 0.0057, "step": 8890 }, { "epoch": 8.180147058823529, "grad_norm": 0.21821390092372894, "learning_rate": 3.2776622688558746e-06, "loss": 0.0085, "step": 8900 }, { "epoch": 8.189338235294118, "grad_norm": 0.2756965160369873, "learning_rate": 3.2190373133884677e-06, "loss": 0.0056, "step": 8910 }, { "epoch": 8.198529411764707, "grad_norm": 0.20703698694705963, "learning_rate": 3.1609239470340446e-06, "loss": 0.0056, "step": 8920 }, { "epoch": 8.207720588235293, "grad_norm": 0.19805669784545898, "learning_rate": 3.1033228053110373e-06, "loss": 0.0055, "step": 8930 }, { "epoch": 8.216911764705882, "grad_norm": 0.2527613639831543, "learning_rate": 3.0462345181363314e-06, "loss": 0.0059, "step": 8940 }, { "epoch": 8.226102941176471, "grad_norm": 0.3347202241420746, "learning_rate": 2.9896597098182654e-06, "loss": 0.0076, "step": 8950 }, { "epoch": 8.235294117647058, "grad_norm": 0.20862768590450287, "learning_rate": 2.933598999049891e-06, "loss": 0.0072, "step": 8960 }, { "epoch": 8.244485294117647, "grad_norm": 0.2757716178894043, "learning_rate": 2.8780529989021697e-06, "loss": 0.0053, "step": 8970 }, { "epoch": 8.253676470588236, "grad_norm": 0.20975635945796967, "learning_rate": 2.823022316817242e-06, "loss": 0.0092, "step": 8980 }, { "epoch": 8.262867647058824, "grad_norm": 0.23042984306812286, "learning_rate": 2.7685075546018456e-06, "loss": 0.0063, "step": 8990 }, { "epoch": 8.272058823529411, "grad_norm": 0.3666793704032898, "learning_rate": 2.7145093084206598e-06, "loss": 0.0084, "step": 9000 }, { "epoch": 8.28125, "grad_norm": 0.22860491275787354, "learning_rate": 2.661028168789892e-06, "loss": 0.0058, "step": 9010 }, { "epoch": 8.290441176470589, "grad_norm": 0.1939830183982849, "learning_rate": 2.6080647205706855e-06, "loss": 0.0054, "step": 9020 }, { "epoch": 8.299632352941176, "grad_norm": 0.34478938579559326, "learning_rate": 2.555619542962834e-06, "loss": 0.0062, "step": 9030 }, { "epoch": 8.308823529411764, "grad_norm": 0.21270525455474854, "learning_rate": 2.503693209498409e-06, "loss": 0.0056, "step": 9040 }, { "epoch": 8.318014705882353, "grad_norm": 0.33734065294265747, "learning_rate": 2.452286288035449e-06, "loss": 0.0059, "step": 9050 }, { "epoch": 8.327205882352942, "grad_norm": 0.20809118449687958, "learning_rate": 2.4013993407518363e-06, "loss": 0.0068, "step": 9060 }, { "epoch": 8.336397058823529, "grad_norm": 0.19125983119010925, "learning_rate": 2.351032924139063e-06, "loss": 0.006, "step": 9070 }, { "epoch": 8.345588235294118, "grad_norm": 0.3663831055164337, "learning_rate": 2.30118758899619e-06, "loss": 0.0076, "step": 9080 }, { "epoch": 8.354779411764707, "grad_norm": 0.16556689143180847, "learning_rate": 2.2518638804238157e-06, "loss": 0.0059, "step": 9090 }, { "epoch": 8.363970588235293, "grad_norm": 0.24895450472831726, "learning_rate": 2.203062337818118e-06, "loss": 0.0064, "step": 9100 }, { "epoch": 8.373161764705882, "grad_norm": 0.15893439948558807, "learning_rate": 2.1547834948649483e-06, "loss": 0.0069, "step": 9110 }, { "epoch": 8.382352941176471, "grad_norm": 0.3203609585762024, "learning_rate": 2.1070278795340017e-06, "loss": 0.0062, "step": 9120 }, { "epoch": 8.391544117647058, "grad_norm": 0.16880406439304352, "learning_rate": 2.059796014073029e-06, "loss": 0.0045, "step": 9130 }, { "epoch": 8.400735294117647, "grad_norm": 0.22788073122501373, "learning_rate": 2.01308841500214e-06, "loss": 0.0063, "step": 9140 }, { "epoch": 8.409926470588236, "grad_norm": 0.22897785902023315, "learning_rate": 1.9669055931081704e-06, "loss": 0.008, "step": 9150 }, { "epoch": 8.419117647058824, "grad_norm": 0.17304888367652893, "learning_rate": 1.9212480534390507e-06, "loss": 0.0058, "step": 9160 }, { "epoch": 8.428308823529411, "grad_norm": 0.20100751519203186, "learning_rate": 1.8761162952983246e-06, "loss": 0.0076, "step": 9170 }, { "epoch": 8.4375, "grad_norm": 0.21167223155498505, "learning_rate": 1.8315108122396618e-06, "loss": 0.0069, "step": 9180 }, { "epoch": 8.446691176470589, "grad_norm": 0.19408510625362396, "learning_rate": 1.787432092061475e-06, "loss": 0.0056, "step": 9190 }, { "epoch": 8.455882352941176, "grad_norm": 0.15020735561847687, "learning_rate": 1.743880616801602e-06, "loss": 0.0072, "step": 9200 }, { "epoch": 8.465073529411764, "grad_norm": 0.19424644112586975, "learning_rate": 1.7008568627319865e-06, "loss": 0.0058, "step": 9210 }, { "epoch": 8.474264705882353, "grad_norm": 0.24864591658115387, "learning_rate": 1.6583613003535226e-06, "loss": 0.0087, "step": 9220 }, { "epoch": 8.483455882352942, "grad_norm": 0.28398922085762024, "learning_rate": 1.6163943943908522e-06, "loss": 0.0062, "step": 9230 }, { "epoch": 8.492647058823529, "grad_norm": 0.18913626670837402, "learning_rate": 1.5749566037873476e-06, "loss": 0.006, "step": 9240 }, { "epoch": 8.501838235294118, "grad_norm": 0.2456756830215454, "learning_rate": 1.5340483817000428e-06, "loss": 0.0073, "step": 9250 }, { "epoch": 8.511029411764707, "grad_norm": 0.2928996980190277, "learning_rate": 1.4936701754947101e-06, "loss": 0.0056, "step": 9260 }, { "epoch": 8.520220588235293, "grad_norm": 0.19216085970401764, "learning_rate": 1.4538224267409361e-06, "loss": 0.0064, "step": 9270 }, { "epoch": 8.529411764705882, "grad_norm": 0.20526443421840668, "learning_rate": 1.414505571207314e-06, "loss": 0.0053, "step": 9280 }, { "epoch": 8.538602941176471, "grad_norm": 0.19458536803722382, "learning_rate": 1.3757200388566816e-06, "loss": 0.0063, "step": 9290 }, { "epoch": 8.547794117647058, "grad_norm": 0.21073277294635773, "learning_rate": 1.3374662538414074e-06, "loss": 0.0086, "step": 9300 }, { "epoch": 8.556985294117647, "grad_norm": 0.21046152710914612, "learning_rate": 1.2997446344987617e-06, "loss": 0.0063, "step": 9310 }, { "epoch": 8.566176470588236, "grad_norm": 0.21999268233776093, "learning_rate": 1.262555593346315e-06, "loss": 0.0069, "step": 9320 }, { "epoch": 8.575367647058824, "grad_norm": 0.23019997775554657, "learning_rate": 1.2258995370774685e-06, "loss": 0.0069, "step": 9330 }, { "epoch": 8.584558823529411, "grad_norm": 0.25326743721961975, "learning_rate": 1.1897768665569798e-06, "loss": 0.0066, "step": 9340 }, { "epoch": 8.59375, "grad_norm": 0.15901057422161102, "learning_rate": 1.1541879768165954e-06, "loss": 0.005, "step": 9350 }, { "epoch": 8.602941176470589, "grad_norm": 0.23218585550785065, "learning_rate": 1.1191332570507085e-06, "loss": 0.0087, "step": 9360 }, { "epoch": 8.612132352941176, "grad_norm": 0.23443476855754852, "learning_rate": 1.0846130906121132e-06, "loss": 0.0085, "step": 9370 }, { "epoch": 8.621323529411764, "grad_norm": 0.23904192447662354, "learning_rate": 1.0506278550078131e-06, "loss": 0.0078, "step": 9380 }, { "epoch": 8.630514705882353, "grad_norm": 0.2619757652282715, "learning_rate": 1.0171779218949185e-06, "loss": 0.0073, "step": 9390 }, { "epoch": 8.639705882352942, "grad_norm": 0.21205805242061615, "learning_rate": 9.842636570765174e-07, "loss": 0.0051, "step": 9400 }, { "epoch": 8.648897058823529, "grad_norm": 0.331827312707901, "learning_rate": 9.518854204977612e-07, "loss": 0.0055, "step": 9410 }, { "epoch": 8.658088235294118, "grad_norm": 0.2562917470932007, "learning_rate": 9.200435662418349e-07, "loss": 0.0065, "step": 9420 }, { "epoch": 8.667279411764707, "grad_norm": 0.17624002695083618, "learning_rate": 8.887384425261658e-07, "loss": 0.0058, "step": 9430 }, { "epoch": 8.676470588235293, "grad_norm": 0.18016794323921204, "learning_rate": 8.579703916985648e-07, "loss": 0.0079, "step": 9440 }, { "epoch": 8.685661764705882, "grad_norm": 0.13987427949905396, "learning_rate": 8.277397502335194e-07, "loss": 0.0055, "step": 9450 }, { "epoch": 8.694852941176471, "grad_norm": 0.22785551846027374, "learning_rate": 7.980468487284675e-07, "loss": 0.0068, "step": 9460 }, { "epoch": 8.704044117647058, "grad_norm": 0.15202787518501282, "learning_rate": 7.688920119002297e-07, "loss": 0.0063, "step": 9470 }, { "epoch": 8.713235294117647, "grad_norm": 0.1441149264574051, "learning_rate": 7.402755585814269e-07, "loss": 0.0056, "step": 9480 }, { "epoch": 8.722426470588236, "grad_norm": 0.17426654696464539, "learning_rate": 7.121978017170073e-07, "loss": 0.0069, "step": 9490 }, { "epoch": 8.731617647058824, "grad_norm": 0.2608022391796112, "learning_rate": 6.846590483608306e-07, "loss": 0.0086, "step": 9500 }, { "epoch": 8.740808823529411, "grad_norm": 0.19900059700012207, "learning_rate": 6.576595996722834e-07, "loss": 0.0068, "step": 9510 }, { "epoch": 8.75, "grad_norm": 0.20335809886455536, "learning_rate": 6.311997509130141e-07, "loss": 0.0084, "step": 9520 }, { "epoch": 8.759191176470589, "grad_norm": 0.2050287127494812, "learning_rate": 6.052797914436803e-07, "loss": 0.006, "step": 9530 }, { "epoch": 8.768382352941176, "grad_norm": 0.1771957278251648, "learning_rate": 5.799000047208181e-07, "loss": 0.0082, "step": 9540 }, { "epoch": 8.777573529411764, "grad_norm": 0.23284511268138885, "learning_rate": 5.550606682937054e-07, "loss": 0.0072, "step": 9550 }, { "epoch": 8.786764705882353, "grad_norm": 0.17588572204113007, "learning_rate": 5.307620538013481e-07, "loss": 0.0069, "step": 9560 }, { "epoch": 8.795955882352942, "grad_norm": 0.2722502052783966, "learning_rate": 5.070044269694874e-07, "loss": 0.0066, "step": 9570 }, { "epoch": 8.805147058823529, "grad_norm": 0.19602122902870178, "learning_rate": 4.837880476077417e-07, "loss": 0.0075, "step": 9580 }, { "epoch": 8.814338235294118, "grad_norm": 0.1927773654460907, "learning_rate": 4.6111316960670835e-07, "loss": 0.0048, "step": 9590 }, { "epoch": 8.823529411764707, "grad_norm": 0.16688086092472076, "learning_rate": 4.389800409352218e-07, "loss": 0.0066, "step": 9600 }, { "epoch": 8.832720588235293, "grad_norm": 0.20968252420425415, "learning_rate": 4.173889036376277e-07, "loss": 0.0085, "step": 9610 }, { "epoch": 8.841911764705882, "grad_norm": 0.19717751443386078, "learning_rate": 3.963399938311463e-07, "loss": 0.0067, "step": 9620 }, { "epoch": 8.851102941176471, "grad_norm": 0.20746855437755585, "learning_rate": 3.7583354170328545e-07, "loss": 0.0081, "step": 9630 }, { "epoch": 8.860294117647058, "grad_norm": 0.14649520814418793, "learning_rate": 3.558697715093207e-07, "loss": 0.0054, "step": 9640 }, { "epoch": 8.869485294117647, "grad_norm": 0.29624661803245544, "learning_rate": 3.3644890156983576e-07, "loss": 0.0073, "step": 9650 }, { "epoch": 8.878676470588236, "grad_norm": 0.15716412663459778, "learning_rate": 3.175711442683638e-07, "loss": 0.006, "step": 9660 }, { "epoch": 8.887867647058824, "grad_norm": 0.3463781774044037, "learning_rate": 2.9923670604902197e-07, "loss": 0.0064, "step": 9670 }, { "epoch": 8.897058823529411, "grad_norm": 0.25711876153945923, "learning_rate": 2.814457874143028e-07, "loss": 0.0073, "step": 9680 }, { "epoch": 8.90625, "grad_norm": 0.13100850582122803, "learning_rate": 2.641985829228366e-07, "loss": 0.0053, "step": 9690 }, { "epoch": 8.915441176470589, "grad_norm": 0.14132870733737946, "learning_rate": 2.474952811872877e-07, "loss": 0.0055, "step": 9700 }, { "epoch": 8.924632352941176, "grad_norm": 0.2465519905090332, "learning_rate": 2.3133606487228397e-07, "loss": 0.007, "step": 9710 }, { "epoch": 8.933823529411764, "grad_norm": 0.1988314837217331, "learning_rate": 2.157211106924295e-07, "loss": 0.0069, "step": 9720 }, { "epoch": 8.943014705882353, "grad_norm": 0.1303059607744217, "learning_rate": 2.006505894103672e-07, "loss": 0.0068, "step": 9730 }, { "epoch": 8.952205882352942, "grad_norm": 0.20611578226089478, "learning_rate": 1.8612466583489696e-07, "loss": 0.0056, "step": 9740 }, { "epoch": 8.961397058823529, "grad_norm": 0.20655281841754913, "learning_rate": 1.7214349881918834e-07, "loss": 0.006, "step": 9750 }, { "epoch": 8.970588235294118, "grad_norm": 0.19210897386074066, "learning_rate": 1.5870724125904845e-07, "loss": 0.0055, "step": 9760 }, { "epoch": 8.979779411764707, "grad_norm": 0.257587194442749, "learning_rate": 1.4581604009124006e-07, "loss": 0.0058, "step": 9770 }, { "epoch": 8.988970588235293, "grad_norm": 0.2370702028274536, "learning_rate": 1.334700362918717e-07, "loss": 0.0056, "step": 9780 }, { "epoch": 8.998161764705882, "grad_norm": 0.135923832654953, "learning_rate": 1.2166936487486015e-07, "loss": 0.0074, "step": 9790 }, { "epoch": 9.007352941176471, "grad_norm": 0.17796720564365387, "learning_rate": 1.1041415489045914e-07, "loss": 0.0048, "step": 9800 }, { "epoch": 9.016544117647058, "grad_norm": 0.1701475828886032, "learning_rate": 9.970452942384412e-08, "loss": 0.0065, "step": 9810 }, { "epoch": 9.025735294117647, "grad_norm": 0.18860657513141632, "learning_rate": 8.954060559375754e-08, "loss": 0.0049, "step": 9820 }, { "epoch": 9.034926470588236, "grad_norm": 0.21800945699214935, "learning_rate": 7.99224945512489e-08, "loss": 0.0086, "step": 9830 }, { "epoch": 9.044117647058824, "grad_norm": 0.17556744813919067, "learning_rate": 7.085030147843675e-08, "loss": 0.0071, "step": 9840 }, { "epoch": 9.053308823529411, "grad_norm": 0.18153336644172668, "learning_rate": 6.232412558736523e-08, "loss": 0.0076, "step": 9850 }, { "epoch": 9.0625, "grad_norm": 0.16894005239009857, "learning_rate": 5.434406011893822e-08, "loss": 0.0081, "step": 9860 }, { "epoch": 9.071691176470589, "grad_norm": 0.21346765756607056, "learning_rate": 4.6910192341864664e-08, "loss": 0.0055, "step": 9870 }, { "epoch": 9.080882352941176, "grad_norm": 0.16597215831279755, "learning_rate": 4.0022603551737035e-08, "loss": 0.0075, "step": 9880 }, { "epoch": 9.090073529411764, "grad_norm": 0.15955832600593567, "learning_rate": 3.3681369070120985e-08, "loss": 0.0079, "step": 9890 }, { "epoch": 9.099264705882353, "grad_norm": 0.3604470193386078, "learning_rate": 2.7886558243744866e-08, "loss": 0.0065, "step": 9900 }, { "epoch": 9.108455882352942, "grad_norm": 0.18860377371311188, "learning_rate": 2.2638234443722596e-08, "loss": 0.0071, "step": 9910 }, { "epoch": 9.117647058823529, "grad_norm": 0.18186844885349274, "learning_rate": 1.7936455064887504e-08, "loss": 0.0087, "step": 9920 }, { "epoch": 9.126838235294118, "grad_norm": 0.17973589897155762, "learning_rate": 1.378127152514841e-08, "loss": 0.0046, "step": 9930 }, { "epoch": 9.136029411764707, "grad_norm": 0.2021789401769638, "learning_rate": 1.0172729264917857e-08, "loss": 0.0075, "step": 9940 }, { "epoch": 9.145220588235293, "grad_norm": 0.20491167902946472, "learning_rate": 7.1108677466458215e-09, "loss": 0.0073, "step": 9950 }, { "epoch": 9.154411764705882, "grad_norm": 0.20612673461437225, "learning_rate": 4.595720454353414e-09, "loss": 0.0065, "step": 9960 }, { "epoch": 9.163602941176471, "grad_norm": 0.2125847190618515, "learning_rate": 2.627314893294264e-09, "loss": 0.0059, "step": 9970 }, { "epoch": 9.172794117647058, "grad_norm": 0.1722206026315689, "learning_rate": 1.2056725896270048e-09, "loss": 0.0058, "step": 9980 }, { "epoch": 9.181985294117647, "grad_norm": 0.21713954210281372, "learning_rate": 3.308090902098826e-10, "loss": 0.0066, "step": 9990 }, { "epoch": 9.191176470588236, "grad_norm": 0.22238269448280334, "learning_rate": 2.7339624120159555e-12, "loss": 0.0078, "step": 10000 }, { "epoch": 9.191176470588236, "step": 10000, "total_flos": 0.0, "train_loss": 0.03981423426233232, "train_runtime": 10936.3917, "train_samples_per_second": 234.081, "train_steps_per_second": 0.914 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 10, "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": 32, "trial_name": null, "trial_params": null }