{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.1986591815948486, "learning_rate": 1.8e-06, "loss": 1.3067, "step": 10 }, { "grad_norm": 1.2062952518463135, "learning_rate": 3.8e-06, "loss": 1.2829, "step": 20 }, { "grad_norm": 0.657501220703125, "learning_rate": 5.8e-06, "loss": 1.2232, "step": 30 }, { "grad_norm": 0.22315914928913116, "learning_rate": 7.8e-06, "loss": 1.182, "step": 40 }, { "grad_norm": 0.25485023856163025, "learning_rate": 9.800000000000001e-06, "loss": 1.1853, "step": 50 }, { "grad_norm": 0.3476347029209137, "learning_rate": 1.18e-05, "loss": 1.1855, "step": 60 }, { "grad_norm": 0.24359121918678284, "learning_rate": 1.3800000000000002e-05, "loss": 1.1801, "step": 70 }, { "grad_norm": 0.28171366453170776, "learning_rate": 1.58e-05, "loss": 1.1827, "step": 80 }, { "grad_norm": 0.21789725124835968, "learning_rate": 1.78e-05, "loss": 1.1729, "step": 90 }, { "grad_norm": 0.2512347400188446, "learning_rate": 1.9800000000000004e-05, "loss": 1.1702, "step": 100 }, { "grad_norm": 0.2558528184890747, "learning_rate": 2.18e-05, "loss": 1.1641, "step": 110 }, { "grad_norm": 0.7746357321739197, "learning_rate": 2.38e-05, "loss": 1.1497, "step": 120 }, { "grad_norm": 0.48968103528022766, "learning_rate": 2.58e-05, "loss": 1.1053, "step": 130 }, { "grad_norm": 0.8197448253631592, "learning_rate": 2.7800000000000005e-05, "loss": 1.086, "step": 140 }, { "grad_norm": 1.2424033880233765, "learning_rate": 2.98e-05, "loss": 1.0245, "step": 150 }, { "grad_norm": 1.6148344278335571, "learning_rate": 3.18e-05, "loss": 0.9457, "step": 160 }, { "grad_norm": 1.6781394481658936, "learning_rate": 3.38e-05, "loss": 0.8176, "step": 170 }, { "grad_norm": 1.3674410581588745, "learning_rate": 3.58e-05, "loss": 0.6892, "step": 180 }, { "grad_norm": 1.997373342514038, "learning_rate": 3.7800000000000004e-05, "loss": 0.6027, "step": 190 }, { "grad_norm": 1.6941936016082764, "learning_rate": 3.9800000000000005e-05, "loss": 0.5073, "step": 200 }, { "grad_norm": 1.415013074874878, "learning_rate": 4.18e-05, "loss": 0.4276, "step": 210 }, { "grad_norm": 2.05452299118042, "learning_rate": 4.38e-05, "loss": 0.3673, "step": 220 }, { "grad_norm": 1.516156792640686, "learning_rate": 4.58e-05, "loss": 0.3332, "step": 230 }, { "grad_norm": 1.7693427801132202, "learning_rate": 4.78e-05, "loss": 0.3117, "step": 240 }, { "grad_norm": 1.2460193634033203, "learning_rate": 4.9800000000000004e-05, "loss": 0.2891, "step": 250 }, { "grad_norm": 1.157357096672058, "learning_rate": 5.1800000000000005e-05, "loss": 0.2835, "step": 260 }, { "grad_norm": 1.2472822666168213, "learning_rate": 5.380000000000001e-05, "loss": 0.2785, "step": 270 }, { "grad_norm": 1.238992691040039, "learning_rate": 5.580000000000001e-05, "loss": 0.2738, "step": 280 }, { "grad_norm": 1.061076045036316, "learning_rate": 5.7799999999999995e-05, "loss": 0.2663, "step": 290 }, { "grad_norm": 0.8667409420013428, "learning_rate": 5.9800000000000003e-05, "loss": 0.2661, "step": 300 }, { "grad_norm": 0.9989050030708313, "learning_rate": 6.18e-05, "loss": 0.26, "step": 310 }, { "grad_norm": 0.9462183117866516, "learning_rate": 6.38e-05, "loss": 0.244, "step": 320 }, { "grad_norm": 1.010033130645752, "learning_rate": 6.58e-05, "loss": 0.2466, "step": 330 }, { "grad_norm": 0.9539633989334106, "learning_rate": 6.780000000000001e-05, "loss": 0.2578, "step": 340 }, { "grad_norm": 0.8161935806274414, "learning_rate": 6.98e-05, "loss": 0.2426, "step": 350 }, { "grad_norm": 0.7835903763771057, "learning_rate": 7.18e-05, "loss": 0.247, "step": 360 }, { "grad_norm": 0.7458878755569458, "learning_rate": 7.38e-05, "loss": 0.2325, "step": 370 }, { "grad_norm": 0.7540352940559387, "learning_rate": 7.58e-05, "loss": 0.2328, "step": 380 }, { "grad_norm": 0.6814438104629517, "learning_rate": 7.780000000000001e-05, "loss": 0.2306, "step": 390 }, { "grad_norm": 0.7077115774154663, "learning_rate": 7.98e-05, "loss": 0.2261, "step": 400 }, { "grad_norm": 0.6930910348892212, "learning_rate": 8.18e-05, "loss": 0.2274, "step": 410 }, { "grad_norm": 0.7360082268714905, "learning_rate": 8.38e-05, "loss": 0.2173, "step": 420 }, { "grad_norm": 0.7077138423919678, "learning_rate": 8.58e-05, "loss": 0.2274, "step": 430 }, { "grad_norm": 0.7573286294937134, "learning_rate": 8.78e-05, "loss": 0.215, "step": 440 }, { "grad_norm": 0.6667159795761108, "learning_rate": 8.98e-05, "loss": 0.2092, "step": 450 }, { "grad_norm": 0.5598796010017395, "learning_rate": 9.180000000000001e-05, "loss": 0.2059, "step": 460 }, { "grad_norm": 0.661995530128479, "learning_rate": 9.38e-05, "loss": 0.1984, "step": 470 }, { "grad_norm": 0.7282671332359314, "learning_rate": 9.58e-05, "loss": 0.2014, "step": 480 }, { "grad_norm": 0.7656590342521667, "learning_rate": 9.78e-05, "loss": 0.2022, "step": 490 }, { "grad_norm": 0.74053955078125, "learning_rate": 9.98e-05, "loss": 0.2051, "step": 500 }, { "grad_norm": 0.4436301589012146, "learning_rate": 9.9999778549206e-05, "loss": 0.2027, "step": 510 }, { "grad_norm": 0.5675735473632812, "learning_rate": 9.999901304280685e-05, "loss": 0.2044, "step": 520 }, { "grad_norm": 0.7628726363182068, "learning_rate": 9.999770075521164e-05, "loss": 0.2031, "step": 530 }, { "grad_norm": 0.531736433506012, "learning_rate": 9.99958417007713e-05, "loss": 0.204, "step": 540 }, { "grad_norm": 0.6587666869163513, "learning_rate": 9.999343589981615e-05, "loss": 0.1898, "step": 550 }, { "grad_norm": 0.4870506823062897, "learning_rate": 9.999048337865568e-05, "loss": 0.1998, "step": 560 }, { "grad_norm": 0.5031449198722839, "learning_rate": 9.998698416957815e-05, "loss": 0.2029, "step": 570 }, { "grad_norm": 0.661207377910614, "learning_rate": 9.998293831085037e-05, "loss": 0.1933, "step": 580 }, { "grad_norm": 0.5457364320755005, "learning_rate": 9.997834584671719e-05, "loss": 0.1943, "step": 590 }, { "grad_norm": 0.5063786506652832, "learning_rate": 9.997320682740107e-05, "loss": 0.1911, "step": 600 }, { "grad_norm": 0.5209915041923523, "learning_rate": 9.996752130910149e-05, "loss": 0.1844, "step": 610 }, { "grad_norm": 0.6069880723953247, "learning_rate": 9.99612893539944e-05, "loss": 0.1883, "step": 620 }, { "grad_norm": 0.4907151460647583, "learning_rate": 9.995451103023144e-05, "loss": 0.1866, "step": 630 }, { "grad_norm": 0.5463731288909912, "learning_rate": 9.994718641193928e-05, "loss": 0.1919, "step": 640 }, { "grad_norm": 0.6598379611968994, "learning_rate": 9.993931557921874e-05, "loss": 0.1898, "step": 650 }, { "grad_norm": 0.5706124901771545, "learning_rate": 9.993089861814402e-05, "loss": 0.1884, "step": 660 }, { "grad_norm": 0.48418113589286804, "learning_rate": 9.992193562076166e-05, "loss": 0.1813, "step": 670 }, { "grad_norm": 0.5193236470222473, "learning_rate": 9.991242668508954e-05, "loss": 0.1767, "step": 680 }, { "grad_norm": 0.5084429383277893, "learning_rate": 9.990237191511587e-05, "loss": 0.1751, "step": 690 }, { "grad_norm": 0.5268884301185608, "learning_rate": 9.989177142079802e-05, "loss": 0.1761, "step": 700 }, { "grad_norm": 0.5338177680969238, "learning_rate": 9.988062531806126e-05, "loss": 0.1802, "step": 710 }, { "grad_norm": 0.46959829330444336, "learning_rate": 9.986893372879762e-05, "loss": 0.1791, "step": 720 }, { "grad_norm": 0.4675990045070648, "learning_rate": 9.985669678086443e-05, "loss": 0.1848, "step": 730 }, { "grad_norm": 0.41440287232398987, "learning_rate": 9.984391460808298e-05, "loss": 0.1857, "step": 740 }, { "grad_norm": 0.5096134543418884, "learning_rate": 9.983058735023709e-05, "loss": 0.1816, "step": 750 }, { "grad_norm": 0.46873027086257935, "learning_rate": 9.98167151530715e-05, "loss": 0.1762, "step": 760 }, { "grad_norm": 0.4754364490509033, "learning_rate": 9.980229816829034e-05, "loss": 0.1818, "step": 770 }, { "grad_norm": 0.46662041544914246, "learning_rate": 9.978733655355544e-05, "loss": 0.185, "step": 780 }, { "grad_norm": 0.4606422781944275, "learning_rate": 9.977183047248464e-05, "loss": 0.1772, "step": 790 }, { "grad_norm": 0.5338506698608398, "learning_rate": 9.975578009464992e-05, "loss": 0.1756, "step": 800 }, { "grad_norm": 0.4376639723777771, "learning_rate": 9.97391855955757e-05, "loss": 0.1703, "step": 810 }, { "grad_norm": 0.5105850696563721, "learning_rate": 9.972204715673669e-05, "loss": 0.1737, "step": 820 }, { "grad_norm": 0.47884154319763184, "learning_rate": 9.970436496555617e-05, "loss": 0.1687, "step": 830 }, { "grad_norm": 0.46137410402297974, "learning_rate": 9.968613921540373e-05, "loss": 0.1778, "step": 840 }, { "grad_norm": 0.4971647560596466, "learning_rate": 9.966737010559326e-05, "loss": 0.1751, "step": 850 }, { "grad_norm": 0.5285759568214417, "learning_rate": 9.964805784138072e-05, "loss": 0.1735, "step": 860 }, { "grad_norm": 0.3871442675590515, "learning_rate": 9.962820263396195e-05, "loss": 0.1714, "step": 870 }, { "grad_norm": 0.5244999527931213, "learning_rate": 9.960780470047033e-05, "loss": 0.1743, "step": 880 }, { "grad_norm": 0.42402705550193787, "learning_rate": 9.958686426397437e-05, "loss": 0.1732, "step": 890 }, { "grad_norm": 0.5233232975006104, "learning_rate": 9.956538155347534e-05, "loss": 0.1748, "step": 900 }, { "grad_norm": 0.38042786717414856, "learning_rate": 9.95433568039047e-05, "loss": 0.1645, "step": 910 }, { "grad_norm": 0.32170557975769043, "learning_rate": 9.952079025612162e-05, "loss": 0.1608, "step": 920 }, { "grad_norm": 0.4248397946357727, "learning_rate": 9.949768215691022e-05, "loss": 0.1699, "step": 930 }, { "grad_norm": 0.3056546449661255, "learning_rate": 9.9474032758977e-05, "loss": 0.1679, "step": 940 }, { "grad_norm": 0.4052722156047821, "learning_rate": 9.944984232094794e-05, "loss": 0.1647, "step": 950 }, { "grad_norm": 0.40288034081459045, "learning_rate": 9.942511110736584e-05, "loss": 0.1694, "step": 960 }, { "grad_norm": 0.3972863554954529, "learning_rate": 9.939983938868726e-05, "loss": 0.1633, "step": 970 }, { "grad_norm": 0.37959742546081543, "learning_rate": 9.93740274412797e-05, "loss": 0.1621, "step": 980 }, { "grad_norm": 0.4162011742591858, "learning_rate": 9.934767554741846e-05, "loss": 0.1739, "step": 990 }, { "grad_norm": 0.4448692798614502, "learning_rate": 9.932078399528361e-05, "loss": 0.1638, "step": 1000 }, { "grad_norm": 0.31835252046585083, "learning_rate": 9.929335307895689e-05, "loss": 0.1612, "step": 1010 }, { "grad_norm": 0.5261848568916321, "learning_rate": 9.926538309841839e-05, "loss": 0.1722, "step": 1020 }, { "grad_norm": 0.3229111135005951, "learning_rate": 9.923687435954334e-05, "loss": 0.1632, "step": 1030 }, { "grad_norm": 0.36171478033065796, "learning_rate": 9.920782717409873e-05, "loss": 0.1594, "step": 1040 }, { "grad_norm": 0.38773488998413086, "learning_rate": 9.917824185973994e-05, "loss": 0.1549, "step": 1050 }, { "grad_norm": 0.309159517288208, "learning_rate": 9.914811874000723e-05, "loss": 0.156, "step": 1060 }, { "grad_norm": 0.31718602776527405, "learning_rate": 9.911745814432218e-05, "loss": 0.1619, "step": 1070 }, { "grad_norm": 0.4561702311038971, "learning_rate": 9.90862604079842e-05, "loss": 0.1629, "step": 1080 }, { "grad_norm": 0.4013567864894867, "learning_rate": 9.90545258721667e-05, "loss": 0.1658, "step": 1090 }, { "grad_norm": 0.4345211386680603, "learning_rate": 9.90222548839135e-05, "loss": 0.1579, "step": 1100 }, { "grad_norm": 0.38550132513046265, "learning_rate": 9.898944779613495e-05, "loss": 0.16, "step": 1110 }, { "grad_norm": 0.28444838523864746, "learning_rate": 9.89561049676041e-05, "loss": 0.1597, "step": 1120 }, { "grad_norm": 0.4140962064266205, "learning_rate": 9.89222267629528e-05, "loss": 0.1602, "step": 1130 }, { "grad_norm": 0.37575763463974, "learning_rate": 9.888781355266763e-05, "loss": 0.1559, "step": 1140 }, { "grad_norm": 0.32500800490379333, "learning_rate": 9.885286571308598e-05, "loss": 0.1558, "step": 1150 }, { "grad_norm": 0.3155668079853058, "learning_rate": 9.881738362639182e-05, "loss": 0.1626, "step": 1160 }, { "grad_norm": 0.3958005905151367, "learning_rate": 9.878136768061154e-05, "loss": 0.1666, "step": 1170 }, { "grad_norm": 0.36953893303871155, "learning_rate": 9.874481826960979e-05, "loss": 0.1631, "step": 1180 }, { "grad_norm": 0.40385204553604126, "learning_rate": 9.870773579308503e-05, "loss": 0.1584, "step": 1190 }, { "grad_norm": 0.476495623588562, "learning_rate": 9.867012065656533e-05, "loss": 0.1633, "step": 1200 }, { "grad_norm": 0.4282894432544708, "learning_rate": 9.863197327140376e-05, "loss": 0.1575, "step": 1210 }, { "grad_norm": 0.34781745076179504, "learning_rate": 9.859329405477403e-05, "loss": 0.1588, "step": 1220 }, { "grad_norm": 0.45303016901016235, "learning_rate": 9.855408342966585e-05, "loss": 0.1631, "step": 1230 }, { "grad_norm": 0.3727198839187622, "learning_rate": 9.851434182488033e-05, "loss": 0.1578, "step": 1240 }, { "grad_norm": 0.3594115376472473, "learning_rate": 9.84740696750253e-05, "loss": 0.1568, "step": 1250 }, { "grad_norm": 0.2778436243534088, "learning_rate": 9.843326742051055e-05, "loss": 0.1628, "step": 1260 }, { "grad_norm": 0.28701916337013245, "learning_rate": 9.839193550754297e-05, "loss": 0.1568, "step": 1270 }, { "grad_norm": 0.37518635392189026, "learning_rate": 9.835007438812177e-05, "loss": 0.1627, "step": 1280 }, { "grad_norm": 0.3559496998786926, "learning_rate": 9.830768452003341e-05, "loss": 0.1615, "step": 1290 }, { "grad_norm": 0.3004937171936035, "learning_rate": 9.826476636684671e-05, "loss": 0.1621, "step": 1300 }, { "grad_norm": 0.3872029185295105, "learning_rate": 9.822132039790773e-05, "loss": 0.1557, "step": 1310 }, { "grad_norm": 0.2521299719810486, "learning_rate": 9.817734708833461e-05, "loss": 0.1521, "step": 1320 }, { "grad_norm": 0.30966511368751526, "learning_rate": 9.813284691901243e-05, "loss": 0.1626, "step": 1330 }, { "grad_norm": 0.4167332053184509, "learning_rate": 9.808782037658792e-05, "loss": 0.1635, "step": 1340 }, { "grad_norm": 0.3960261344909668, "learning_rate": 9.804226795346411e-05, "loss": 0.1631, "step": 1350 }, { "grad_norm": 0.32885199785232544, "learning_rate": 9.799619014779503e-05, "loss": 0.161, "step": 1360 }, { "grad_norm": 0.4660717844963074, "learning_rate": 9.794958746348013e-05, "loss": 0.1608, "step": 1370 }, { "grad_norm": 0.33331605792045593, "learning_rate": 9.790246041015896e-05, "loss": 0.1541, "step": 1380 }, { "grad_norm": 0.35680457949638367, "learning_rate": 9.785480950320538e-05, "loss": 0.1587, "step": 1390 }, { "grad_norm": 0.37307751178741455, "learning_rate": 9.78066352637221e-05, "loss": 0.1566, "step": 1400 }, { "grad_norm": 0.3227611482143402, "learning_rate": 9.775793821853488e-05, "loss": 0.155, "step": 1410 }, { "grad_norm": 0.3211585283279419, "learning_rate": 9.77087189001868e-05, "loss": 0.1562, "step": 1420 }, { "grad_norm": 0.38690608739852905, "learning_rate": 9.765897784693243e-05, "loss": 0.1554, "step": 1430 }, { "grad_norm": 0.3528446853160858, "learning_rate": 9.760871560273197e-05, "loss": 0.1516, "step": 1440 }, { "grad_norm": 0.26060885190963745, "learning_rate": 9.755793271724526e-05, "loss": 0.1515, "step": 1450 }, { "grad_norm": 0.35703030228614807, "learning_rate": 9.750662974582584e-05, "loss": 0.1616, "step": 1460 }, { "grad_norm": 0.44114935398101807, "learning_rate": 9.745480724951473e-05, "loss": 0.1538, "step": 1470 }, { "grad_norm": 0.3952416479587555, "learning_rate": 9.740246579503447e-05, "loss": 0.155, "step": 1480 }, { "grad_norm": 0.2751481831073761, "learning_rate": 9.734960595478284e-05, "loss": 0.1535, "step": 1490 }, { "grad_norm": 0.3489803969860077, "learning_rate": 9.729622830682657e-05, "loss": 0.1531, "step": 1500 }, { "grad_norm": 0.2933256924152374, "learning_rate": 9.724233343489504e-05, "loss": 0.1575, "step": 1510 }, { "grad_norm": 0.31148776412010193, "learning_rate": 9.718792192837396e-05, "loss": 0.1519, "step": 1520 }, { "grad_norm": 0.38123375177383423, "learning_rate": 9.713299438229886e-05, "loss": 0.1585, "step": 1530 }, { "grad_norm": 0.36282679438591003, "learning_rate": 9.707755139734855e-05, "loss": 0.1583, "step": 1540 }, { "grad_norm": 0.336394339799881, "learning_rate": 9.702159357983866e-05, "loss": 0.1572, "step": 1550 }, { "grad_norm": 0.35583966970443726, "learning_rate": 9.696512154171492e-05, "loss": 0.1518, "step": 1560 }, { "grad_norm": 0.3243831396102905, "learning_rate": 9.690813590054645e-05, "loss": 0.159, "step": 1570 }, { "grad_norm": 0.4028131365776062, "learning_rate": 9.685063727951914e-05, "loss": 0.1553, "step": 1580 }, { "grad_norm": 0.24144324660301208, "learning_rate": 9.679262630742865e-05, "loss": 0.1569, "step": 1590 }, { "grad_norm": 0.40314143896102905, "learning_rate": 9.673410361867373e-05, "loss": 0.1552, "step": 1600 }, { "grad_norm": 0.35892507433891296, "learning_rate": 9.667506985324909e-05, "loss": 0.1515, "step": 1610 }, { "grad_norm": 0.35124096274375916, "learning_rate": 9.661552565673855e-05, "loss": 0.1497, "step": 1620 }, { "grad_norm": 0.44414857029914856, "learning_rate": 9.655547168030789e-05, "loss": 0.1564, "step": 1630 }, { "grad_norm": 0.27231210470199585, "learning_rate": 9.649490858069777e-05, "loss": 0.1534, "step": 1640 }, { "grad_norm": 0.33163806796073914, "learning_rate": 9.643383702021658e-05, "loss": 0.1538, "step": 1650 }, { "grad_norm": 0.2961295545101166, "learning_rate": 9.637225766673307e-05, "loss": 0.1521, "step": 1660 }, { "grad_norm": 0.35117107629776, "learning_rate": 9.631017119366922e-05, "loss": 0.1541, "step": 1670 }, { "grad_norm": 0.3611936867237091, "learning_rate": 9.624757827999273e-05, "loss": 0.1505, "step": 1680 }, { "grad_norm": 0.3903159201145172, "learning_rate": 9.618447961020971e-05, "loss": 0.161, "step": 1690 }, { "grad_norm": 0.23246657848358154, "learning_rate": 9.612087587435707e-05, "loss": 0.1595, "step": 1700 }, { "grad_norm": 0.2525351643562317, "learning_rate": 9.605676776799508e-05, "loss": 0.1572, "step": 1710 }, { "grad_norm": 0.3065750002861023, "learning_rate": 9.599215599219973e-05, "loss": 0.1549, "step": 1720 }, { "grad_norm": 0.33358967304229736, "learning_rate": 9.592704125355505e-05, "loss": 0.1518, "step": 1730 }, { "grad_norm": 0.3254912495613098, "learning_rate": 9.586142426414538e-05, "loss": 0.1546, "step": 1740 }, { "grad_norm": 0.3136656582355499, "learning_rate": 9.57953057415476e-05, "loss": 0.156, "step": 1750 }, { "grad_norm": 0.25366419553756714, "learning_rate": 9.572868640882328e-05, "loss": 0.1547, "step": 1760 }, { "grad_norm": 0.4185545742511749, "learning_rate": 9.56615669945108e-05, "loss": 0.1552, "step": 1770 }, { "grad_norm": 0.3697797358036041, "learning_rate": 9.55939482326173e-05, "loss": 0.1585, "step": 1780 }, { "grad_norm": 0.29024067521095276, "learning_rate": 9.552583086261069e-05, "loss": 0.1586, "step": 1790 }, { "grad_norm": 0.3730272948741913, "learning_rate": 9.545721562941168e-05, "loss": 0.1587, "step": 1800 }, { "grad_norm": 0.4166305363178253, "learning_rate": 9.538810328338543e-05, "loss": 0.1522, "step": 1810 }, { "grad_norm": 0.40496861934661865, "learning_rate": 9.531849458033349e-05, "loss": 0.154, "step": 1820 }, { "grad_norm": 0.3961986005306244, "learning_rate": 9.524839028148547e-05, "loss": 0.1525, "step": 1830 }, { "grad_norm": 0.29068803787231445, "learning_rate": 9.517779115349077e-05, "loss": 0.1508, "step": 1840 }, { "grad_norm": 0.37395983934402466, "learning_rate": 9.510669796841014e-05, "loss": 0.1524, "step": 1850 }, { "grad_norm": 0.35216084122657776, "learning_rate": 9.503511150370727e-05, "loss": 0.1569, "step": 1860 }, { "grad_norm": 0.32542961835861206, "learning_rate": 9.496303254224024e-05, "loss": 0.152, "step": 1870 }, { "grad_norm": 0.26481327414512634, "learning_rate": 9.489046187225306e-05, "loss": 0.1494, "step": 1880 }, { "grad_norm": 0.3297596871852875, "learning_rate": 9.481740028736692e-05, "loss": 0.1481, "step": 1890 }, { "grad_norm": 0.2952992618083954, "learning_rate": 9.474384858657164e-05, "loss": 0.147, "step": 1900 }, { "grad_norm": 0.30430904030799866, "learning_rate": 9.466980757421679e-05, "loss": 0.149, "step": 1910 }, { "grad_norm": 0.26173660159111023, "learning_rate": 9.459527806000305e-05, "loss": 0.1506, "step": 1920 }, { "grad_norm": 0.2292925864458084, "learning_rate": 9.452026085897325e-05, "loss": 0.1506, "step": 1930 }, { "grad_norm": 0.2502480745315552, "learning_rate": 9.444475679150348e-05, "loss": 0.1522, "step": 1940 }, { "grad_norm": 0.4015428125858307, "learning_rate": 9.436876668329411e-05, "loss": 0.1522, "step": 1950 }, { "grad_norm": 0.34276536107063293, "learning_rate": 9.429229136536079e-05, "loss": 0.1504, "step": 1960 }, { "grad_norm": 0.29634934663772583, "learning_rate": 9.421533167402534e-05, "loss": 0.1536, "step": 1970 }, { "grad_norm": 0.3181280195713043, "learning_rate": 9.413788845090666e-05, "loss": 0.1505, "step": 1980 }, { "grad_norm": 0.29984375834465027, "learning_rate": 9.405996254291136e-05, "loss": 0.151, "step": 1990 }, { "grad_norm": 0.3330492675304413, "learning_rate": 9.398155480222474e-05, "loss": 0.149, "step": 2000 }, { "grad_norm": 0.39731672406196594, "learning_rate": 9.390266608630128e-05, "loss": 0.1486, "step": 2010 }, { "grad_norm": 0.3324965536594391, "learning_rate": 9.38232972578553e-05, "loss": 0.1521, "step": 2020 }, { "grad_norm": 0.3332907557487488, "learning_rate": 9.374344918485164e-05, "loss": 0.1501, "step": 2030 }, { "grad_norm": 0.2761757969856262, "learning_rate": 9.366312274049602e-05, "loss": 0.1468, "step": 2040 }, { "grad_norm": 0.2493954598903656, "learning_rate": 9.358231880322554e-05, "loss": 0.1517, "step": 2050 }, { "grad_norm": 0.26896241307258606, "learning_rate": 9.350103825669916e-05, "loss": 0.1512, "step": 2060 }, { "grad_norm": 0.30001798272132874, "learning_rate": 9.341928198978787e-05, "loss": 0.1493, "step": 2070 }, { "grad_norm": 0.24016854166984558, "learning_rate": 9.333705089656512e-05, "loss": 0.143, "step": 2080 }, { "grad_norm": 0.28328436613082886, "learning_rate": 9.325434587629698e-05, "loss": 0.1473, "step": 2090 }, { "grad_norm": 0.2471616119146347, "learning_rate": 9.31711678334323e-05, "loss": 0.1486, "step": 2100 }, { "grad_norm": 0.3372943103313446, "learning_rate": 9.308751767759282e-05, "loss": 0.1465, "step": 2110 }, { "grad_norm": 0.2913314402103424, "learning_rate": 9.300339632356325e-05, "loss": 0.1431, "step": 2120 }, { "grad_norm": 0.23359070718288422, "learning_rate": 9.291880469128124e-05, "loss": 0.1528, "step": 2130 }, { "grad_norm": 0.22234484553337097, "learning_rate": 9.283374370582732e-05, "loss": 0.1488, "step": 2140 }, { "grad_norm": 0.25127702951431274, "learning_rate": 9.274821429741482e-05, "loss": 0.1471, "step": 2150 }, { "grad_norm": 0.29147952795028687, "learning_rate": 9.266221740137961e-05, "loss": 0.1478, "step": 2160 }, { "grad_norm": 0.282123863697052, "learning_rate": 9.257575395817001e-05, "loss": 0.1495, "step": 2170 }, { "grad_norm": 0.2752014100551605, "learning_rate": 9.248882491333637e-05, "loss": 0.1484, "step": 2180 }, { "grad_norm": 0.27321657538414, "learning_rate": 9.240143121752076e-05, "loss": 0.1492, "step": 2190 }, { "grad_norm": 0.3750593364238739, "learning_rate": 9.23135738264467e-05, "loss": 0.1565, "step": 2200 }, { "grad_norm": 0.3431914150714874, "learning_rate": 9.222525370090849e-05, "loss": 0.1488, "step": 2210 }, { "grad_norm": 0.2743963599205017, "learning_rate": 9.213647180676088e-05, "loss": 0.1529, "step": 2220 }, { "grad_norm": 0.24094721674919128, "learning_rate": 9.204722911490846e-05, "loss": 0.1437, "step": 2230 }, { "grad_norm": 0.30099794268608093, "learning_rate": 9.1957526601295e-05, "loss": 0.1452, "step": 2240 }, { "grad_norm": 0.22911037504673004, "learning_rate": 9.186736524689281e-05, "loss": 0.1482, "step": 2250 }, { "grad_norm": 0.32866403460502625, "learning_rate": 9.177674603769204e-05, "loss": 0.1514, "step": 2260 }, { "grad_norm": 0.2690809965133667, "learning_rate": 9.168566996468983e-05, "loss": 0.1473, "step": 2270 }, { "grad_norm": 0.36028704047203064, "learning_rate": 9.159413802387951e-05, "loss": 0.144, "step": 2280 }, { "grad_norm": 0.3449890613555908, "learning_rate": 9.150215121623974e-05, "loss": 0.1455, "step": 2290 }, { "grad_norm": 0.3125399947166443, "learning_rate": 9.140971054772349e-05, "loss": 0.1454, "step": 2300 }, { "grad_norm": 0.23733501136302948, "learning_rate": 9.131681702924713e-05, "loss": 0.148, "step": 2310 }, { "grad_norm": 0.34111151099205017, "learning_rate": 9.122347167667926e-05, "loss": 0.1498, "step": 2320 }, { "grad_norm": 0.30981823801994324, "learning_rate": 9.112967551082973e-05, "loss": 0.1531, "step": 2330 }, { "grad_norm": 0.3011240065097809, "learning_rate": 9.103542955743835e-05, "loss": 0.1426, "step": 2340 }, { "grad_norm": 0.16210812330245972, "learning_rate": 9.094073484716381e-05, "loss": 0.1527, "step": 2350 }, { "grad_norm": 0.19529588520526886, "learning_rate": 9.084559241557226e-05, "loss": 0.1488, "step": 2360 }, { "grad_norm": 0.277685284614563, "learning_rate": 9.075000330312608e-05, "loss": 0.1457, "step": 2370 }, { "grad_norm": 0.3037843108177185, "learning_rate": 9.065396855517253e-05, "loss": 0.1511, "step": 2380 }, { "grad_norm": 0.23564369976520538, "learning_rate": 9.055748922193219e-05, "loss": 0.1524, "step": 2390 }, { "grad_norm": 0.22768235206604004, "learning_rate": 9.046056635848761e-05, "loss": 0.1476, "step": 2400 }, { "grad_norm": 0.29369890689849854, "learning_rate": 9.036320102477169e-05, "loss": 0.1455, "step": 2410 }, { "grad_norm": 0.31512168049812317, "learning_rate": 9.02653942855561e-05, "loss": 0.1544, "step": 2420 }, { "grad_norm": 0.2561895549297333, "learning_rate": 9.016714721043971e-05, "loss": 0.1487, "step": 2430 }, { "grad_norm": 0.28954461216926575, "learning_rate": 9.006846087383675e-05, "loss": 0.1471, "step": 2440 }, { "grad_norm": 0.3062959909439087, "learning_rate": 8.996933635496523e-05, "loss": 0.1482, "step": 2450 }, { "grad_norm": 0.24846193194389343, "learning_rate": 8.986977473783498e-05, "loss": 0.1463, "step": 2460 }, { "grad_norm": 0.26708948612213135, "learning_rate": 8.97697771112359e-05, "loss": 0.1474, "step": 2470 }, { "grad_norm": 0.21605928242206573, "learning_rate": 8.966934456872602e-05, "loss": 0.1446, "step": 2480 }, { "grad_norm": 0.26997026801109314, "learning_rate": 8.95684782086195e-05, "loss": 0.1511, "step": 2490 }, { "grad_norm": 0.348294198513031, "learning_rate": 8.946717913397476e-05, "loss": 0.1478, "step": 2500 }, { "grad_norm": 0.23830530047416687, "learning_rate": 8.93654484525822e-05, "loss": 0.145, "step": 2510 }, { "grad_norm": 0.2281816452741623, "learning_rate": 8.926328727695226e-05, "loss": 0.1452, "step": 2520 }, { "grad_norm": 0.23404772579669952, "learning_rate": 8.916069672430319e-05, "loss": 0.1448, "step": 2530 }, { "grad_norm": 0.28105345368385315, "learning_rate": 8.905767791654884e-05, "loss": 0.1446, "step": 2540 }, { "grad_norm": 0.34967371821403503, "learning_rate": 8.895423198028638e-05, "loss": 0.1475, "step": 2550 }, { "grad_norm": 0.33839184045791626, "learning_rate": 8.885036004678402e-05, "loss": 0.1519, "step": 2560 }, { "grad_norm": 0.32090646028518677, "learning_rate": 8.874606325196857e-05, "loss": 0.1439, "step": 2570 }, { "grad_norm": 0.2824375331401825, "learning_rate": 8.864134273641304e-05, "loss": 0.144, "step": 2580 }, { "grad_norm": 0.27718451619148254, "learning_rate": 8.853619964532427e-05, "loss": 0.1437, "step": 2590 }, { "grad_norm": 0.28382250666618347, "learning_rate": 8.843063512853019e-05, "loss": 0.1507, "step": 2600 }, { "grad_norm": 0.30156874656677246, "learning_rate": 8.832465034046749e-05, "loss": 0.1479, "step": 2610 }, { "grad_norm": 0.2413700670003891, "learning_rate": 8.821824644016882e-05, "loss": 0.1488, "step": 2620 }, { "grad_norm": 0.22377872467041016, "learning_rate": 8.811142459125019e-05, "loss": 0.1456, "step": 2630 }, { "grad_norm": 0.34218090772628784, "learning_rate": 8.800418596189822e-05, "loss": 0.148, "step": 2640 }, { "grad_norm": 0.26538988947868347, "learning_rate": 8.789653172485737e-05, "loss": 0.1474, "step": 2650 }, { "grad_norm": 0.30965787172317505, "learning_rate": 8.778846305741715e-05, "loss": 0.1493, "step": 2660 }, { "grad_norm": 0.24406598508358002, "learning_rate": 8.767998114139918e-05, "loss": 0.1462, "step": 2670 }, { "grad_norm": 0.3095422387123108, "learning_rate": 8.757108716314429e-05, "loss": 0.1459, "step": 2680 }, { "grad_norm": 0.2716175615787506, "learning_rate": 8.746178231349962e-05, "loss": 0.1492, "step": 2690 }, { "grad_norm": 0.25041866302490234, "learning_rate": 8.735206778780549e-05, "loss": 0.1421, "step": 2700 }, { "grad_norm": 0.24789278209209442, "learning_rate": 8.724194478588234e-05, "loss": 0.1445, "step": 2710 }, { "grad_norm": 0.3536326587200165, "learning_rate": 8.713141451201772e-05, "loss": 0.1463, "step": 2720 }, { "grad_norm": 0.18370474874973297, "learning_rate": 8.702047817495295e-05, "loss": 0.145, "step": 2730 }, { "grad_norm": 0.3408331274986267, "learning_rate": 8.69091369878701e-05, "loss": 0.1452, "step": 2740 }, { "grad_norm": 0.3321828246116638, "learning_rate": 8.679739216837849e-05, "loss": 0.1439, "step": 2750 }, { "grad_norm": 0.1822550892829895, "learning_rate": 8.66852449385016e-05, "loss": 0.1478, "step": 2760 }, { "grad_norm": 0.27523133158683777, "learning_rate": 8.657269652466356e-05, "loss": 0.1471, "step": 2770 }, { "grad_norm": 0.21025203168392181, "learning_rate": 8.645974815767577e-05, "loss": 0.1405, "step": 2780 }, { "grad_norm": 0.21099801361560822, "learning_rate": 8.634640107272351e-05, "loss": 0.1426, "step": 2790 }, { "grad_norm": 0.28931620717048645, "learning_rate": 8.623265650935234e-05, "loss": 0.1461, "step": 2800 }, { "grad_norm": 0.28504449129104614, "learning_rate": 8.611851571145456e-05, "loss": 0.1496, "step": 2810 }, { "grad_norm": 0.2725865840911865, "learning_rate": 8.600397992725566e-05, "loss": 0.1503, "step": 2820 }, { "grad_norm": 0.2519656717777252, "learning_rate": 8.588905040930061e-05, "loss": 0.1415, "step": 2830 }, { "grad_norm": 0.23929940164089203, "learning_rate": 8.577372841444022e-05, "loss": 0.1422, "step": 2840 }, { "grad_norm": 0.2376219630241394, "learning_rate": 8.565801520381736e-05, "loss": 0.1464, "step": 2850 }, { "grad_norm": 0.2880883514881134, "learning_rate": 8.554191204285313e-05, "loss": 0.1507, "step": 2860 }, { "grad_norm": 0.19880278408527374, "learning_rate": 8.542542020123315e-05, "loss": 0.1473, "step": 2870 }, { "grad_norm": 0.20648367702960968, "learning_rate": 8.530854095289347e-05, "loss": 0.1437, "step": 2880 }, { "grad_norm": 0.2092444896697998, "learning_rate": 8.519127557600688e-05, "loss": 0.1468, "step": 2890 }, { "grad_norm": 0.241817906498909, "learning_rate": 8.507362535296871e-05, "loss": 0.1459, "step": 2900 }, { "grad_norm": 0.26988479495048523, "learning_rate": 8.495559157038299e-05, "loss": 0.1442, "step": 2910 }, { "grad_norm": 0.25270703434944153, "learning_rate": 8.483717551904823e-05, "loss": 0.1489, "step": 2920 }, { "grad_norm": 0.2316846251487732, "learning_rate": 8.47183784939434e-05, "loss": 0.1453, "step": 2930 }, { "grad_norm": 0.2821042537689209, "learning_rate": 8.459920179421374e-05, "loss": 0.1488, "step": 2940 }, { "grad_norm": 0.30260536074638367, "learning_rate": 8.447964672315656e-05, "loss": 0.1449, "step": 2950 }, { "grad_norm": 0.24286040663719177, "learning_rate": 8.435971458820692e-05, "loss": 0.1491, "step": 2960 }, { "grad_norm": 0.2317284792661667, "learning_rate": 8.423940670092345e-05, "loss": 0.1433, "step": 2970 }, { "grad_norm": 0.2558586299419403, "learning_rate": 8.411872437697394e-05, "loss": 0.1429, "step": 2980 }, { "grad_norm": 0.21183688938617706, "learning_rate": 8.399766893612096e-05, "loss": 0.1438, "step": 2990 }, { "grad_norm": 0.2230009138584137, "learning_rate": 8.38762417022074e-05, "loss": 0.1433, "step": 3000 }, { "grad_norm": 0.2001793533563614, "learning_rate": 8.375444400314204e-05, "loss": 0.1432, "step": 3010 }, { "grad_norm": 0.31928494572639465, "learning_rate": 8.3632277170885e-05, "loss": 0.1455, "step": 3020 }, { "grad_norm": 0.20921824872493744, "learning_rate": 8.350974254143318e-05, "loss": 0.1428, "step": 3030 }, { "grad_norm": 0.27949658036231995, "learning_rate": 8.338684145480566e-05, "loss": 0.1464, "step": 3040 }, { "grad_norm": 0.19098623096942902, "learning_rate": 8.326357525502904e-05, "loss": 0.1415, "step": 3050 }, { "grad_norm": 0.22426141798496246, "learning_rate": 8.313994529012273e-05, "loss": 0.1458, "step": 3060 }, { "grad_norm": 0.17794322967529297, "learning_rate": 8.301595291208422e-05, "loss": 0.1466, "step": 3070 }, { "grad_norm": 0.2082102745771408, "learning_rate": 8.289159947687427e-05, "loss": 0.1456, "step": 3080 }, { "grad_norm": 0.21185842156410217, "learning_rate": 8.276688634440216e-05, "loss": 0.1464, "step": 3090 }, { "grad_norm": 0.24942827224731445, "learning_rate": 8.26418148785107e-05, "loss": 0.1453, "step": 3100 }, { "grad_norm": 0.17011144757270813, "learning_rate": 8.251638644696141e-05, "loss": 0.146, "step": 3110 }, { "grad_norm": 0.19342312216758728, "learning_rate": 8.23906024214195e-05, "loss": 0.1442, "step": 3120 }, { "grad_norm": 0.15668097138404846, "learning_rate": 8.226446417743897e-05, "loss": 0.1473, "step": 3130 }, { "grad_norm": 0.25087910890579224, "learning_rate": 8.213797309444742e-05, "loss": 0.1447, "step": 3140 }, { "grad_norm": 0.18921996653079987, "learning_rate": 8.201113055573105e-05, "loss": 0.1456, "step": 3150 }, { "grad_norm": 0.2648988962173462, "learning_rate": 8.188393794841958e-05, "loss": 0.1444, "step": 3160 }, { "grad_norm": 0.25288838148117065, "learning_rate": 8.175639666347094e-05, "loss": 0.1438, "step": 3170 }, { "grad_norm": 0.2881231904029846, "learning_rate": 8.162850809565623e-05, "loss": 0.1418, "step": 3180 }, { "grad_norm": 0.27005308866500854, "learning_rate": 8.150027364354431e-05, "loss": 0.1425, "step": 3190 }, { "grad_norm": 0.20269057154655457, "learning_rate": 8.137169470948662e-05, "loss": 0.1424, "step": 3200 }, { "grad_norm": 0.2176142930984497, "learning_rate": 8.124277269960179e-05, "loss": 0.1436, "step": 3210 }, { "grad_norm": 0.3340862989425659, "learning_rate": 8.111350902376023e-05, "loss": 0.1421, "step": 3220 }, { "grad_norm": 0.2911006510257721, "learning_rate": 8.098390509556883e-05, "loss": 0.1414, "step": 3230 }, { "grad_norm": 0.2560000717639923, "learning_rate": 8.085396233235536e-05, "loss": 0.1458, "step": 3240 }, { "grad_norm": 0.29439622163772583, "learning_rate": 8.072368215515306e-05, "loss": 0.1444, "step": 3250 }, { "grad_norm": 0.23155541718006134, "learning_rate": 8.059306598868506e-05, "loss": 0.1471, "step": 3260 }, { "grad_norm": 0.20351189374923706, "learning_rate": 8.046211526134888e-05, "loss": 0.1443, "step": 3270 }, { "grad_norm": 0.2211005687713623, "learning_rate": 8.033083140520065e-05, "loss": 0.1436, "step": 3280 }, { "grad_norm": 0.262359619140625, "learning_rate": 8.019921585593962e-05, "loss": 0.1505, "step": 3290 }, { "grad_norm": 0.2653680145740509, "learning_rate": 8.006727005289232e-05, "loss": 0.1397, "step": 3300 }, { "grad_norm": 0.23254480957984924, "learning_rate": 7.993499543899692e-05, "loss": 0.1417, "step": 3310 }, { "grad_norm": 0.30542925000190735, "learning_rate": 7.980239346078742e-05, "loss": 0.1454, "step": 3320 }, { "grad_norm": 0.20530977845191956, "learning_rate": 7.966946556837778e-05, "loss": 0.1411, "step": 3330 }, { "grad_norm": 0.21580870449543, "learning_rate": 7.953621321544616e-05, "loss": 0.141, "step": 3340 }, { "grad_norm": 0.24497996270656586, "learning_rate": 7.940263785921896e-05, "loss": 0.1395, "step": 3350 }, { "grad_norm": 0.2318846583366394, "learning_rate": 7.926874096045482e-05, "loss": 0.1417, "step": 3360 }, { "grad_norm": 0.26386377215385437, "learning_rate": 7.913452398342881e-05, "loss": 0.1421, "step": 3370 }, { "grad_norm": 0.24405337870121002, "learning_rate": 7.89999883959163e-05, "loss": 0.1401, "step": 3380 }, { "grad_norm": 0.21852239966392517, "learning_rate": 7.886513566917687e-05, "loss": 0.1369, "step": 3390 }, { "grad_norm": 0.22262151539325714, "learning_rate": 7.872996727793838e-05, "loss": 0.1447, "step": 3400 }, { "grad_norm": 0.2687032222747803, "learning_rate": 7.859448470038069e-05, "loss": 0.1387, "step": 3410 }, { "grad_norm": 0.2897404432296753, "learning_rate": 7.845868941811956e-05, "loss": 0.1411, "step": 3420 }, { "grad_norm": 0.23512592911720276, "learning_rate": 7.832258291619043e-05, "loss": 0.1432, "step": 3430 }, { "grad_norm": 0.23354469239711761, "learning_rate": 7.81861666830322e-05, "loss": 0.1454, "step": 3440 }, { "grad_norm": 0.2600792944431305, "learning_rate": 7.804944221047097e-05, "loss": 0.1406, "step": 3450 }, { "grad_norm": 0.18907132744789124, "learning_rate": 7.791241099370364e-05, "loss": 0.1425, "step": 3460 }, { "grad_norm": 0.20533791184425354, "learning_rate": 7.777507453128163e-05, "loss": 0.1399, "step": 3470 }, { "grad_norm": 0.2803758978843689, "learning_rate": 7.763743432509451e-05, "loss": 0.1432, "step": 3480 }, { "grad_norm": 0.2402222603559494, "learning_rate": 7.749949188035353e-05, "loss": 0.1413, "step": 3490 }, { "grad_norm": 0.1919722706079483, "learning_rate": 7.736124870557516e-05, "loss": 0.142, "step": 3500 }, { "grad_norm": 0.1812194585800171, "learning_rate": 7.722270631256459e-05, "loss": 0.138, "step": 3510 }, { "grad_norm": 0.21413545310497284, "learning_rate": 7.708386621639925e-05, "loss": 0.141, "step": 3520 }, { "grad_norm": 0.26781123876571655, "learning_rate": 7.694472993541219e-05, "loss": 0.1428, "step": 3530 }, { "grad_norm": 0.19944126904010773, "learning_rate": 7.680529899117547e-05, "loss": 0.139, "step": 3540 }, { "grad_norm": 0.22467485070228577, "learning_rate": 7.666557490848358e-05, "loss": 0.1401, "step": 3550 }, { "grad_norm": 0.2753590941429138, "learning_rate": 7.65255592153367e-05, "loss": 0.1411, "step": 3560 }, { "grad_norm": 0.211239293217659, "learning_rate": 7.638525344292402e-05, "loss": 0.141, "step": 3570 }, { "grad_norm": 0.23313218355178833, "learning_rate": 7.624465912560697e-05, "loss": 0.1407, "step": 3580 }, { "grad_norm": 0.22403769195079803, "learning_rate": 7.610377780090249e-05, "loss": 0.1377, "step": 3590 }, { "grad_norm": 0.19674132764339447, "learning_rate": 7.596261100946618e-05, "loss": 0.1422, "step": 3600 }, { "grad_norm": 0.18445558845996857, "learning_rate": 7.582116029507542e-05, "loss": 0.1398, "step": 3610 }, { "grad_norm": 0.16082623600959778, "learning_rate": 7.56794272046126e-05, "loss": 0.138, "step": 3620 }, { "grad_norm": 0.21798865497112274, "learning_rate": 7.55374132880481e-05, "loss": 0.1386, "step": 3630 }, { "grad_norm": 0.24314968287944794, "learning_rate": 7.539512009842333e-05, "loss": 0.1392, "step": 3640 }, { "grad_norm": 0.2277681827545166, "learning_rate": 7.525254919183382e-05, "loss": 0.1405, "step": 3650 }, { "grad_norm": 0.1823507845401764, "learning_rate": 7.510970212741215e-05, "loss": 0.1417, "step": 3660 }, { "grad_norm": 0.21112054586410522, "learning_rate": 7.496658046731096e-05, "loss": 0.1417, "step": 3670 }, { "grad_norm": 0.24363280832767487, "learning_rate": 7.482318577668578e-05, "loss": 0.1393, "step": 3680 }, { "grad_norm": 0.22302944958209991, "learning_rate": 7.467951962367796e-05, "loss": 0.1403, "step": 3690 }, { "grad_norm": 0.16851010918617249, "learning_rate": 7.453558357939755e-05, "loss": 0.1397, "step": 3700 }, { "grad_norm": 0.24148724973201752, "learning_rate": 7.439137921790606e-05, "loss": 0.1377, "step": 3710 }, { "grad_norm": 0.23120324313640594, "learning_rate": 7.42469081161993e-05, "loss": 0.139, "step": 3720 }, { "grad_norm": 0.18809053301811218, "learning_rate": 7.410217185419006e-05, "loss": 0.1407, "step": 3730 }, { "grad_norm": 0.1773308366537094, "learning_rate": 7.395717201469095e-05, "loss": 0.1391, "step": 3740 }, { "grad_norm": 0.18020175397396088, "learning_rate": 7.381191018339696e-05, "loss": 0.1403, "step": 3750 }, { "grad_norm": 0.28343868255615234, "learning_rate": 7.36663879488682e-05, "loss": 0.1397, "step": 3760 }, { "grad_norm": 0.27247682213783264, "learning_rate": 7.352060690251254e-05, "loss": 0.1437, "step": 3770 }, { "grad_norm": 0.19435696303844452, "learning_rate": 7.337456863856811e-05, "loss": 0.1375, "step": 3780 }, { "grad_norm": 0.22533005475997925, "learning_rate": 7.3228274754086e-05, "loss": 0.1413, "step": 3790 }, { "grad_norm": 0.22039999067783356, "learning_rate": 7.308172684891267e-05, "loss": 0.1396, "step": 3800 }, { "grad_norm": 0.24114154279232025, "learning_rate": 7.293492652567255e-05, "loss": 0.1395, "step": 3810 }, { "grad_norm": 0.23901355266571045, "learning_rate": 7.278787538975043e-05, "loss": 0.14, "step": 3820 }, { "grad_norm": 0.19949999451637268, "learning_rate": 7.2640575049274e-05, "loss": 0.1421, "step": 3830 }, { "grad_norm": 0.18923339247703552, "learning_rate": 7.249302711509616e-05, "loss": 0.138, "step": 3840 }, { "grad_norm": 0.24646790325641632, "learning_rate": 7.23452332007775e-05, "loss": 0.1396, "step": 3850 }, { "grad_norm": 0.1837541162967682, "learning_rate": 7.219719492256858e-05, "loss": 0.1409, "step": 3860 }, { "grad_norm": 0.18584759533405304, "learning_rate": 7.20489138993923e-05, "loss": 0.1427, "step": 3870 }, { "grad_norm": 0.24849654734134674, "learning_rate": 7.190039175282614e-05, "loss": 0.1399, "step": 3880 }, { "grad_norm": 0.17534513771533966, "learning_rate": 7.175163010708455e-05, "loss": 0.1396, "step": 3890 }, { "grad_norm": 0.156320258975029, "learning_rate": 7.1602630589001e-05, "loss": 0.1404, "step": 3900 }, { "grad_norm": 0.24140328168869019, "learning_rate": 7.14533948280104e-05, "loss": 0.1431, "step": 3910 }, { "grad_norm": 0.2953331768512726, "learning_rate": 7.130392445613109e-05, "loss": 0.1443, "step": 3920 }, { "grad_norm": 0.18554715812206268, "learning_rate": 7.115422110794711e-05, "loss": 0.1377, "step": 3930 }, { "grad_norm": 0.2688322961330414, "learning_rate": 7.100428642059033e-05, "loss": 0.1402, "step": 3940 }, { "grad_norm": 0.21256540715694427, "learning_rate": 7.08541220337224e-05, "loss": 0.1401, "step": 3950 }, { "grad_norm": 0.16598273813724518, "learning_rate": 7.070372958951706e-05, "loss": 0.1434, "step": 3960 }, { "grad_norm": 0.19348745048046112, "learning_rate": 7.055311073264194e-05, "loss": 0.1414, "step": 3970 }, { "grad_norm": 0.2642860412597656, "learning_rate": 7.040226711024077e-05, "loss": 0.1467, "step": 3980 }, { "grad_norm": 0.18460509181022644, "learning_rate": 7.02512003719152e-05, "loss": 0.1467, "step": 3990 }, { "grad_norm": 0.18276984989643097, "learning_rate": 7.00999121697069e-05, "loss": 0.1436, "step": 4000 }, { "grad_norm": 0.24415366351604462, "learning_rate": 6.99484041580794e-05, "loss": 0.1405, "step": 4010 }, { "grad_norm": 0.1465180665254593, "learning_rate": 6.979667799390004e-05, "loss": 0.1391, "step": 4020 }, { "grad_norm": 0.20695193111896515, "learning_rate": 6.964473533642185e-05, "loss": 0.1404, "step": 4030 }, { "grad_norm": 0.17077915370464325, "learning_rate": 6.949257784726539e-05, "loss": 0.139, "step": 4040 }, { "grad_norm": 0.1790696233510971, "learning_rate": 6.934020719040056e-05, "loss": 0.1371, "step": 4050 }, { "grad_norm": 0.18274666368961334, "learning_rate": 6.918762503212848e-05, "loss": 0.141, "step": 4060 }, { "grad_norm": 0.21777990460395813, "learning_rate": 6.903483304106319e-05, "loss": 0.1419, "step": 4070 }, { "grad_norm": 0.17992500960826874, "learning_rate": 6.888183288811341e-05, "loss": 0.1394, "step": 4080 }, { "grad_norm": 0.2523497939109802, "learning_rate": 6.87286262464643e-05, "loss": 0.1421, "step": 4090 }, { "grad_norm": 0.18036724627017975, "learning_rate": 6.857521479155915e-05, "loss": 0.1425, "step": 4100 }, { "grad_norm": 0.21799318492412567, "learning_rate": 6.842160020108104e-05, "loss": 0.1439, "step": 4110 }, { "grad_norm": 0.21253177523612976, "learning_rate": 6.826778415493455e-05, "loss": 0.1408, "step": 4120 }, { "grad_norm": 0.21397800743579865, "learning_rate": 6.811376833522729e-05, "loss": 0.1371, "step": 4130 }, { "grad_norm": 0.2591836750507355, "learning_rate": 6.795955442625159e-05, "loss": 0.1445, "step": 4140 }, { "grad_norm": 0.18533650040626526, "learning_rate": 6.780514411446608e-05, "loss": 0.1403, "step": 4150 }, { "grad_norm": 0.1537729799747467, "learning_rate": 6.765053908847716e-05, "loss": 0.1448, "step": 4160 }, { "grad_norm": 0.19125142693519592, "learning_rate": 6.749574103902064e-05, "loss": 0.1404, "step": 4170 }, { "grad_norm": 0.22955143451690674, "learning_rate": 6.734075165894317e-05, "loss": 0.1383, "step": 4180 }, { "grad_norm": 0.15027308464050293, "learning_rate": 6.71855726431838e-05, "loss": 0.1373, "step": 4190 }, { "grad_norm": 0.2132178395986557, "learning_rate": 6.703020568875538e-05, "loss": 0.1397, "step": 4200 }, { "grad_norm": 0.15905748307704926, "learning_rate": 6.687465249472603e-05, "loss": 0.1386, "step": 4210 }, { "grad_norm": 0.18354040384292603, "learning_rate": 6.671891476220055e-05, "loss": 0.1407, "step": 4220 }, { "grad_norm": 0.18853023648262024, "learning_rate": 6.656299419430183e-05, "loss": 0.1403, "step": 4230 }, { "grad_norm": 0.19970139861106873, "learning_rate": 6.640689249615223e-05, "loss": 0.1429, "step": 4240 }, { "grad_norm": 0.24953103065490723, "learning_rate": 6.625061137485491e-05, "loss": 0.1414, "step": 4250 }, { "grad_norm": 0.23096860945224762, "learning_rate": 6.609415253947517e-05, "loss": 0.1423, "step": 4260 }, { "grad_norm": 0.14040957391262054, "learning_rate": 6.593751770102178e-05, "loss": 0.1369, "step": 4270 }, { "grad_norm": 0.1717834621667862, "learning_rate": 6.578070857242823e-05, "loss": 0.1404, "step": 4280 }, { "grad_norm": 0.18914955854415894, "learning_rate": 6.562372686853402e-05, "loss": 0.1389, "step": 4290 }, { "grad_norm": 0.20225109159946442, "learning_rate": 6.546657430606593e-05, "loss": 0.1389, "step": 4300 }, { "grad_norm": 0.2370501607656479, "learning_rate": 6.530925260361918e-05, "loss": 0.1427, "step": 4310 }, { "grad_norm": 0.22026988863945007, "learning_rate": 6.515176348163871e-05, "loss": 0.1373, "step": 4320 }, { "grad_norm": 0.18674996495246887, "learning_rate": 6.499410866240032e-05, "loss": 0.1393, "step": 4330 }, { "grad_norm": 0.30642351508140564, "learning_rate": 6.48362898699919e-05, "loss": 0.1398, "step": 4340 }, { "grad_norm": 0.17516835033893585, "learning_rate": 6.467830883029443e-05, "loss": 0.1378, "step": 4350 }, { "grad_norm": 0.2644605338573456, "learning_rate": 6.452016727096326e-05, "loss": 0.1373, "step": 4360 }, { "grad_norm": 0.21681728959083557, "learning_rate": 6.436186692140916e-05, "loss": 0.142, "step": 4370 }, { "grad_norm": 0.27583640813827515, "learning_rate": 6.420340951277938e-05, "loss": 0.1402, "step": 4380 }, { "grad_norm": 0.178934708237648, "learning_rate": 6.404479677793874e-05, "loss": 0.1412, "step": 4390 }, { "grad_norm": 0.23508059978485107, "learning_rate": 6.388603045145075e-05, "loss": 0.1408, "step": 4400 }, { "grad_norm": 0.22743992507457733, "learning_rate": 6.372711226955843e-05, "loss": 0.1421, "step": 4410 }, { "grad_norm": 0.21188682317733765, "learning_rate": 6.356804397016564e-05, "loss": 0.1369, "step": 4420 }, { "grad_norm": 0.1756334751844406, "learning_rate": 6.340882729281779e-05, "loss": 0.1422, "step": 4430 }, { "grad_norm": 0.152481809258461, "learning_rate": 6.324946397868294e-05, "loss": 0.137, "step": 4440 }, { "grad_norm": 0.21525263786315918, "learning_rate": 6.308995577053276e-05, "loss": 0.1387, "step": 4450 }, { "grad_norm": 0.1848566085100174, "learning_rate": 6.293030441272347e-05, "loss": 0.1342, "step": 4460 }, { "grad_norm": 0.17765450477600098, "learning_rate": 6.277051165117677e-05, "loss": 0.1341, "step": 4470 }, { "grad_norm": 0.2028893679380417, "learning_rate": 6.261057923336064e-05, "loss": 0.1413, "step": 4480 }, { "grad_norm": 0.17806218564510345, "learning_rate": 6.245050890827042e-05, "loss": 0.1373, "step": 4490 }, { "grad_norm": 0.21595104038715363, "learning_rate": 6.229030242640952e-05, "loss": 0.139, "step": 4500 }, { "grad_norm": 0.19062687456607819, "learning_rate": 6.212996153977037e-05, "loss": 0.1382, "step": 4510 }, { "grad_norm": 0.17274987697601318, "learning_rate": 6.196948800181523e-05, "loss": 0.1381, "step": 4520 }, { "grad_norm": 0.19183765351772308, "learning_rate": 6.180888356745695e-05, "loss": 0.1391, "step": 4530 }, { "grad_norm": 0.1931641846895218, "learning_rate": 6.164814999303995e-05, "loss": 0.1392, "step": 4540 }, { "grad_norm": 0.18116290867328644, "learning_rate": 6.148728903632081e-05, "loss": 0.1383, "step": 4550 }, { "grad_norm": 0.1921987533569336, "learning_rate": 6.132630245644921e-05, "loss": 0.142, "step": 4560 }, { "grad_norm": 0.2170371413230896, "learning_rate": 6.116519201394857e-05, "loss": 0.1357, "step": 4570 }, { "grad_norm": 0.19295448064804077, "learning_rate": 6.10039594706969e-05, "loss": 0.1449, "step": 4580 }, { "grad_norm": 0.2575233280658722, "learning_rate": 6.084260658990744e-05, "loss": 0.1436, "step": 4590 }, { "grad_norm": 0.20364496111869812, "learning_rate": 6.068113513610943e-05, "loss": 0.1375, "step": 4600 }, { "grad_norm": 0.17163479328155518, "learning_rate": 6.0519546875128876e-05, "loss": 0.1391, "step": 4610 }, { "grad_norm": 0.4000489115715027, "learning_rate": 6.035784357406906e-05, "loss": 0.1412, "step": 4620 }, { "grad_norm": 0.3028452694416046, "learning_rate": 6.01960270012914e-05, "loss": 0.1444, "step": 4630 }, { "grad_norm": 0.22404395043849945, "learning_rate": 6.003409892639599e-05, "loss": 0.1425, "step": 4640 }, { "grad_norm": 0.15880945324897766, "learning_rate": 5.9872061120202336e-05, "loss": 0.1412, "step": 4650 }, { "grad_norm": 0.18611487746238708, "learning_rate": 5.9709915354729914e-05, "loss": 0.1401, "step": 4660 }, { "grad_norm": 0.15304820239543915, "learning_rate": 5.9547663403178824e-05, "loss": 0.1424, "step": 4670 }, { "grad_norm": 0.20641250908374786, "learning_rate": 5.9385307039910445e-05, "loss": 0.1395, "step": 4680 }, { "grad_norm": 0.1620214581489563, "learning_rate": 5.922284804042792e-05, "loss": 0.1446, "step": 4690 }, { "grad_norm": 0.1421450674533844, "learning_rate": 5.906028818135687e-05, "loss": 0.138, "step": 4700 }, { "grad_norm": 0.17642006278038025, "learning_rate": 5.889762924042585e-05, "loss": 0.1397, "step": 4710 }, { "grad_norm": 0.24790439009666443, "learning_rate": 5.873487299644699e-05, "loss": 0.1422, "step": 4720 }, { "grad_norm": 0.276292085647583, "learning_rate": 5.857202122929649e-05, "loss": 0.1405, "step": 4730 }, { "grad_norm": 0.22956880927085876, "learning_rate": 5.840907571989518e-05, "loss": 0.1395, "step": 4740 }, { "grad_norm": 0.20441822707653046, "learning_rate": 5.824603825018904e-05, "loss": 0.1407, "step": 4750 }, { "grad_norm": 0.21084265410900116, "learning_rate": 5.808291060312975e-05, "loss": 0.1397, "step": 4760 }, { "grad_norm": 0.16395530104637146, "learning_rate": 5.7919694562655083e-05, "loss": 0.1411, "step": 4770 }, { "grad_norm": 0.224909245967865, "learning_rate": 5.775639191366954e-05, "loss": 0.1426, "step": 4780 }, { "grad_norm": 0.19728395342826843, "learning_rate": 5.75930044420247e-05, "loss": 0.1423, "step": 4790 }, { "grad_norm": 0.20519207417964935, "learning_rate": 5.74295339344998e-05, "loss": 0.1403, "step": 4800 }, { "grad_norm": 0.26532208919525146, "learning_rate": 5.726598217878211e-05, "loss": 0.1457, "step": 4810 }, { "grad_norm": 0.2440793514251709, "learning_rate": 5.71023509634474e-05, "loss": 0.142, "step": 4820 }, { "grad_norm": 0.20775945484638214, "learning_rate": 5.693864207794049e-05, "loss": 0.1396, "step": 4830 }, { "grad_norm": 0.18267016112804413, "learning_rate": 5.677485731255545e-05, "loss": 0.1393, "step": 4840 }, { "grad_norm": 0.1298273801803589, "learning_rate": 5.6610998458416296e-05, "loss": 0.1383, "step": 4850 }, { "grad_norm": 0.14296481013298035, "learning_rate": 5.644706730745716e-05, "loss": 0.144, "step": 4860 }, { "grad_norm": 0.18198657035827637, "learning_rate": 5.628306565240287e-05, "loss": 0.1399, "step": 4870 }, { "grad_norm": 0.18292468786239624, "learning_rate": 5.611899528674923e-05, "loss": 0.1403, "step": 4880 }, { "grad_norm": 0.160342738032341, "learning_rate": 5.595485800474349e-05, "loss": 0.1379, "step": 4890 }, { "grad_norm": 0.18790368735790253, "learning_rate": 5.579065560136467e-05, "loss": 0.1397, "step": 4900 }, { "grad_norm": 0.21585240960121155, "learning_rate": 5.562638987230392e-05, "loss": 0.1385, "step": 4910 }, { "grad_norm": 0.15999537706375122, "learning_rate": 5.546206261394498e-05, "loss": 0.1431, "step": 4920 }, { "grad_norm": 0.18990777432918549, "learning_rate": 5.529767562334437e-05, "loss": 0.1418, "step": 4930 }, { "grad_norm": 0.18967507779598236, "learning_rate": 5.5133230698211926e-05, "loss": 0.1393, "step": 4940 }, { "grad_norm": 0.1792706549167633, "learning_rate": 5.496872963689096e-05, "loss": 0.1417, "step": 4950 }, { "grad_norm": 0.1524370163679123, "learning_rate": 5.4804174238338756e-05, "loss": 0.1421, "step": 4960 }, { "grad_norm": 0.16905544698238373, "learning_rate": 5.463956630210678e-05, "loss": 0.1374, "step": 4970 }, { "grad_norm": 0.16646763682365417, "learning_rate": 5.4474907628321046e-05, "loss": 0.1408, "step": 4980 }, { "grad_norm": 0.18539665639400482, "learning_rate": 5.431020001766244e-05, "loss": 0.1386, "step": 4990 }, { "grad_norm": 0.17078949511051178, "learning_rate": 5.4145445271346986e-05, "loss": 0.1416, "step": 5000 }, { "grad_norm": 0.23945964872837067, "learning_rate": 5.398064519110622e-05, "loss": 0.1422, "step": 5010 }, { "grad_norm": 0.19377689063549042, "learning_rate": 5.3815801579167394e-05, "loss": 0.1426, "step": 5020 }, { "grad_norm": 0.17530454695224762, "learning_rate": 5.365091623823382e-05, "loss": 0.1367, "step": 5030 }, { "grad_norm": 0.1962692141532898, "learning_rate": 5.348599097146521e-05, "loss": 0.1386, "step": 5040 }, { "grad_norm": 0.18964730203151703, "learning_rate": 5.3321027582457836e-05, "loss": 0.1371, "step": 5050 }, { "grad_norm": 0.14053449034690857, "learning_rate": 5.315602787522491e-05, "loss": 0.1386, "step": 5060 }, { "grad_norm": 0.17130205035209656, "learning_rate": 5.299099365417678e-05, "loss": 0.1393, "step": 5070 }, { "grad_norm": 0.21795575320720673, "learning_rate": 5.2825926724101236e-05, "loss": 0.1396, "step": 5080 }, { "grad_norm": 0.1834845393896103, "learning_rate": 5.26608288901438e-05, "loss": 0.1404, "step": 5090 }, { "grad_norm": 0.1859319508075714, "learning_rate": 5.24957019577879e-05, "loss": 0.1376, "step": 5100 }, { "grad_norm": 0.24217164516448975, "learning_rate": 5.2330547732835266e-05, "loss": 0.1388, "step": 5110 }, { "grad_norm": 0.18788227438926697, "learning_rate": 5.2165368021385996e-05, "loss": 0.1387, "step": 5120 }, { "grad_norm": 0.19587846100330353, "learning_rate": 5.200016462981897e-05, "loss": 0.1394, "step": 5130 }, { "grad_norm": 0.21457888185977936, "learning_rate": 5.1834939364772015e-05, "loss": 0.1424, "step": 5140 }, { "grad_norm": 0.20325326919555664, "learning_rate": 5.166969403312214e-05, "loss": 0.1405, "step": 5150 }, { "grad_norm": 0.19539128243923187, "learning_rate": 5.1504430441965844e-05, "loss": 0.1412, "step": 5160 }, { "grad_norm": 0.20316605269908905, "learning_rate": 5.133915039859923e-05, "loss": 0.1393, "step": 5170 }, { "grad_norm": 0.12033970654010773, "learning_rate": 5.1173855710498444e-05, "loss": 0.1362, "step": 5180 }, { "grad_norm": 0.1520044058561325, "learning_rate": 5.100854818529967e-05, "loss": 0.1405, "step": 5190 }, { "grad_norm": 0.19174933433532715, "learning_rate": 5.084322963077951e-05, "loss": 0.1402, "step": 5200 }, { "grad_norm": 0.2099388688802719, "learning_rate": 5.067790185483522e-05, "loss": 0.1397, "step": 5210 }, { "grad_norm": 0.18337737023830414, "learning_rate": 5.0512566665464844e-05, "loss": 0.1391, "step": 5220 }, { "grad_norm": 0.20584355294704437, "learning_rate": 5.034722587074755e-05, "loss": 0.1388, "step": 5230 }, { "grad_norm": 0.18631185591220856, "learning_rate": 5.018188127882375e-05, "loss": 0.1391, "step": 5240 }, { "grad_norm": 0.2188747525215149, "learning_rate": 5.0016534697875417e-05, "loss": 0.1384, "step": 5250 }, { "grad_norm": 0.18918076157569885, "learning_rate": 4.9851187936106294e-05, "loss": 0.1393, "step": 5260 }, { "grad_norm": 0.15814457833766937, "learning_rate": 4.968584280172206e-05, "loss": 0.1406, "step": 5270 }, { "grad_norm": 0.1531916707754135, "learning_rate": 4.95205011029106e-05, "loss": 0.1342, "step": 5280 }, { "grad_norm": 0.16841033101081848, "learning_rate": 4.935516464782227e-05, "loss": 0.1396, "step": 5290 }, { "grad_norm": 0.1942206174135208, "learning_rate": 4.918983524455003e-05, "loss": 0.14, "step": 5300 }, { "grad_norm": 0.19472914934158325, "learning_rate": 4.9024514701109766e-05, "loss": 0.1368, "step": 5310 }, { "grad_norm": 0.17008107900619507, "learning_rate": 4.885920482542043e-05, "loss": 0.139, "step": 5320 }, { "grad_norm": 0.1290515661239624, "learning_rate": 4.869390742528438e-05, "loss": 0.1371, "step": 5330 }, { "grad_norm": 0.16929982602596283, "learning_rate": 4.852862430836744e-05, "loss": 0.1401, "step": 5340 }, { "grad_norm": 0.19875632226467133, "learning_rate": 4.836335728217933e-05, "loss": 0.1356, "step": 5350 }, { "grad_norm": 0.15715964138507843, "learning_rate": 4.819810815405379e-05, "loss": 0.1386, "step": 5360 }, { "grad_norm": 0.16490112245082855, "learning_rate": 4.803287873112877e-05, "loss": 0.1366, "step": 5370 }, { "grad_norm": 0.17743565142154694, "learning_rate": 4.786767082032681e-05, "loss": 0.1367, "step": 5380 }, { "grad_norm": 0.1623411774635315, "learning_rate": 4.77024862283351e-05, "loss": 0.1356, "step": 5390 }, { "grad_norm": 0.22918647527694702, "learning_rate": 4.753732676158593e-05, "loss": 0.1368, "step": 5400 }, { "grad_norm": 0.17592835426330566, "learning_rate": 4.737219422623672e-05, "loss": 0.1407, "step": 5410 }, { "grad_norm": 0.24048466980457306, "learning_rate": 4.720709042815044e-05, "loss": 0.1372, "step": 5420 }, { "grad_norm": 0.15495198965072632, "learning_rate": 4.704201717287578e-05, "loss": 0.1391, "step": 5430 }, { "grad_norm": 0.16570685803890228, "learning_rate": 4.6876976265627404e-05, "loss": 0.1369, "step": 5440 }, { "grad_norm": 0.1928415298461914, "learning_rate": 4.671196951126626e-05, "loss": 0.137, "step": 5450 }, { "grad_norm": 0.1447875201702118, "learning_rate": 4.654699871427971e-05, "loss": 0.1344, "step": 5460 }, { "grad_norm": 0.18527117371559143, "learning_rate": 4.6382065678762034e-05, "loss": 0.1376, "step": 5470 }, { "grad_norm": 0.1755950003862381, "learning_rate": 4.6217172208394424e-05, "loss": 0.1384, "step": 5480 }, { "grad_norm": 0.22146707773208618, "learning_rate": 4.605232010642549e-05, "loss": 0.1375, "step": 5490 }, { "grad_norm": 0.2151176482439041, "learning_rate": 4.588751117565142e-05, "loss": 0.1418, "step": 5500 }, { "grad_norm": 0.14217908680438995, "learning_rate": 4.5722747218396214e-05, "loss": 0.1357, "step": 5510 }, { "grad_norm": 0.16911311447620392, "learning_rate": 4.5558030036492194e-05, "loss": 0.1375, "step": 5520 }, { "grad_norm": 0.16833887994289398, "learning_rate": 4.539336143125999e-05, "loss": 0.1363, "step": 5530 }, { "grad_norm": 0.1799604594707489, "learning_rate": 4.522874320348916e-05, "loss": 0.14, "step": 5540 }, { "grad_norm": 0.18645404279232025, "learning_rate": 4.506417715341821e-05, "loss": 0.1381, "step": 5550 }, { "grad_norm": 0.1807125210762024, "learning_rate": 4.489966508071511e-05, "loss": 0.1397, "step": 5560 }, { "grad_norm": 0.16233235597610474, "learning_rate": 4.4735208784457575e-05, "loss": 0.136, "step": 5570 }, { "grad_norm": 0.15166056156158447, "learning_rate": 4.457081006311325e-05, "loss": 0.1411, "step": 5580 }, { "grad_norm": 0.13611136376857758, "learning_rate": 4.440647071452027e-05, "loss": 0.1376, "step": 5590 }, { "grad_norm": 0.15664826333522797, "learning_rate": 4.424219253586737e-05, "loss": 0.1437, "step": 5600 }, { "grad_norm": 0.17359979450702667, "learning_rate": 4.407797732367443e-05, "loss": 0.1408, "step": 5610 }, { "grad_norm": 0.14235416054725647, "learning_rate": 4.391382687377268e-05, "loss": 0.1391, "step": 5620 }, { "grad_norm": 0.18614311516284943, "learning_rate": 4.374974298128512e-05, "loss": 0.1395, "step": 5630 }, { "grad_norm": 0.1562073826789856, "learning_rate": 4.358572744060699e-05, "loss": 0.1355, "step": 5640 }, { "grad_norm": 0.18601353466510773, "learning_rate": 4.342178204538588e-05, "loss": 0.1385, "step": 5650 }, { "grad_norm": 0.14688754081726074, "learning_rate": 4.325790858850241e-05, "loss": 0.1368, "step": 5660 }, { "grad_norm": 0.16731791198253632, "learning_rate": 4.309410886205043e-05, "loss": 0.1406, "step": 5670 }, { "grad_norm": 0.14388041198253632, "learning_rate": 4.293038465731752e-05, "loss": 0.1381, "step": 5680 }, { "grad_norm": 0.19782860577106476, "learning_rate": 4.276673776476533e-05, "loss": 0.1417, "step": 5690 }, { "grad_norm": 0.15570643544197083, "learning_rate": 4.260316997401007e-05, "loss": 0.1349, "step": 5700 }, { "grad_norm": 0.2079993039369583, "learning_rate": 4.243968307380293e-05, "loss": 0.1397, "step": 5710 }, { "grad_norm": 0.20648974180221558, "learning_rate": 4.22762788520104e-05, "loss": 0.1356, "step": 5720 }, { "grad_norm": 0.18002009391784668, "learning_rate": 4.211295909559491e-05, "loss": 0.1446, "step": 5730 }, { "grad_norm": 0.17395684123039246, "learning_rate": 4.194972559059511e-05, "loss": 0.1404, "step": 5740 }, { "grad_norm": 0.17982737720012665, "learning_rate": 4.178658012210651e-05, "loss": 0.137, "step": 5750 }, { "grad_norm": 0.15348301827907562, "learning_rate": 4.162352447426177e-05, "loss": 0.1403, "step": 5760 }, { "grad_norm": 0.1701090931892395, "learning_rate": 4.146056043021135e-05, "loss": 0.1382, "step": 5770 }, { "grad_norm": 0.18298955261707306, "learning_rate": 4.1297689772103944e-05, "loss": 0.1362, "step": 5780 }, { "grad_norm": 0.15671363472938538, "learning_rate": 4.113491428106694e-05, "loss": 0.1357, "step": 5790 }, { "grad_norm": 0.17840313911437988, "learning_rate": 4.0972235737187055e-05, "loss": 0.1389, "step": 5800 }, { "grad_norm": 0.12834087014198303, "learning_rate": 4.080965591949076e-05, "loss": 0.1387, "step": 5810 }, { "grad_norm": 0.10675042122602463, "learning_rate": 4.0647176605924924e-05, "loss": 0.1363, "step": 5820 }, { "grad_norm": 0.14996156096458435, "learning_rate": 4.0484799573337255e-05, "loss": 0.1352, "step": 5830 }, { "grad_norm": 0.1723569631576538, "learning_rate": 4.032252659745699e-05, "loss": 0.1382, "step": 5840 }, { "grad_norm": 0.16011595726013184, "learning_rate": 4.016035945287539e-05, "loss": 0.1399, "step": 5850 }, { "grad_norm": 0.20151974260807037, "learning_rate": 3.999829991302635e-05, "loss": 0.1398, "step": 5860 }, { "grad_norm": 0.14940814673900604, "learning_rate": 3.983634975016707e-05, "loss": 0.1364, "step": 5870 }, { "grad_norm": 0.14781658351421356, "learning_rate": 3.967451073535854e-05, "loss": 0.1388, "step": 5880 }, { "grad_norm": 0.1312480866909027, "learning_rate": 3.951278463844633e-05, "loss": 0.1349, "step": 5890 }, { "grad_norm": 0.1748630851507187, "learning_rate": 3.935117322804111e-05, "loss": 0.1378, "step": 5900 }, { "grad_norm": 0.19062404334545135, "learning_rate": 3.918967827149938e-05, "loss": 0.1376, "step": 5910 }, { "grad_norm": 0.17294155061244965, "learning_rate": 3.9028301534904094e-05, "loss": 0.1372, "step": 5920 }, { "grad_norm": 0.21876798570156097, "learning_rate": 3.88670447830454e-05, "loss": 0.1411, "step": 5930 }, { "grad_norm": 0.1384684145450592, "learning_rate": 3.870590977940132e-05, "loss": 0.1381, "step": 5940 }, { "grad_norm": 0.15588685870170593, "learning_rate": 3.8544898286118404e-05, "loss": 0.1424, "step": 5950 }, { "grad_norm": 0.1409302055835724, "learning_rate": 3.838401206399257e-05, "loss": 0.1371, "step": 5960 }, { "grad_norm": 0.1550574004650116, "learning_rate": 3.822325287244975e-05, "loss": 0.1366, "step": 5970 }, { "grad_norm": 0.1400359869003296, "learning_rate": 3.8062622469526725e-05, "loss": 0.14, "step": 5980 }, { "grad_norm": 0.17648111283779144, "learning_rate": 3.790212261185183e-05, "loss": 0.1399, "step": 5990 }, { "grad_norm": 0.14976288378238678, "learning_rate": 3.7741755054625794e-05, "loss": 0.1356, "step": 6000 }, { "grad_norm": 0.1836836040019989, "learning_rate": 3.758152155160255e-05, "loss": 0.1405, "step": 6010 }, { "grad_norm": 0.15409569442272186, "learning_rate": 3.742142385506999e-05, "loss": 0.1378, "step": 6020 }, { "grad_norm": 0.2239159196615219, "learning_rate": 3.72614637158309e-05, "loss": 0.142, "step": 6030 }, { "grad_norm": 0.1695176213979721, "learning_rate": 3.710164288318371e-05, "loss": 0.1382, "step": 6040 }, { "grad_norm": 0.129371777176857, "learning_rate": 3.694196310490345e-05, "loss": 0.1381, "step": 6050 }, { "grad_norm": 0.17858611047267914, "learning_rate": 3.678242612722259e-05, "loss": 0.1367, "step": 6060 }, { "grad_norm": 0.13880248367786407, "learning_rate": 3.6623033694811953e-05, "loss": 0.1385, "step": 6070 }, { "grad_norm": 0.12002900242805481, "learning_rate": 3.6463787550761665e-05, "loss": 0.1411, "step": 6080 }, { "grad_norm": 0.11256267130374908, "learning_rate": 3.630468943656202e-05, "loss": 0.1388, "step": 6090 }, { "grad_norm": 0.16008077561855316, "learning_rate": 3.6145741092084523e-05, "loss": 0.1385, "step": 6100 }, { "grad_norm": 0.21591132879257202, "learning_rate": 3.598694425556278e-05, "loss": 0.1408, "step": 6110 }, { "grad_norm": 0.18286028504371643, "learning_rate": 3.58283006635736e-05, "loss": 0.1409, "step": 6120 }, { "grad_norm": 0.16577377915382385, "learning_rate": 3.566981205101781e-05, "loss": 0.1382, "step": 6130 }, { "grad_norm": 0.1730106770992279, "learning_rate": 3.5511480151101556e-05, "loss": 0.1399, "step": 6140 }, { "grad_norm": 0.1321248859167099, "learning_rate": 3.5353306695317104e-05, "loss": 0.1361, "step": 6150 }, { "grad_norm": 0.16622547805309296, "learning_rate": 3.519529341342402e-05, "loss": 0.1386, "step": 6160 }, { "grad_norm": 0.16883622109889984, "learning_rate": 3.503744203343026e-05, "loss": 0.1355, "step": 6170 }, { "grad_norm": 0.15355727076530457, "learning_rate": 3.487975428157318e-05, "loss": 0.1366, "step": 6180 }, { "grad_norm": 0.20334392786026, "learning_rate": 3.472223188230083e-05, "loss": 0.1374, "step": 6190 }, { "grad_norm": 0.2035192847251892, "learning_rate": 3.4564876558252866e-05, "loss": 0.1386, "step": 6200 }, { "grad_norm": 0.13637188076972961, "learning_rate": 3.440769003024195e-05, "loss": 0.1399, "step": 6210 }, { "grad_norm": 0.1509222537279129, "learning_rate": 3.425067401723477e-05, "loss": 0.1403, "step": 6220 }, { "grad_norm": 0.208926260471344, "learning_rate": 3.409383023633325e-05, "loss": 0.1387, "step": 6230 }, { "grad_norm": 0.17525267601013184, "learning_rate": 3.3937160402755894e-05, "loss": 0.1377, "step": 6240 }, { "grad_norm": 0.17071472108364105, "learning_rate": 3.378066622981885e-05, "loss": 0.137, "step": 6250 }, { "grad_norm": 0.16263912618160248, "learning_rate": 3.362434942891738e-05, "loss": 0.1344, "step": 6260 }, { "grad_norm": 0.14737029373645782, "learning_rate": 3.346821170950693e-05, "loss": 0.1371, "step": 6270 }, { "grad_norm": 0.10153527557849884, "learning_rate": 3.3312254779084585e-05, "loss": 0.1364, "step": 6280 }, { "grad_norm": 0.1524280160665512, "learning_rate": 3.315648034317039e-05, "loss": 0.14, "step": 6290 }, { "grad_norm": 0.17136512696743011, "learning_rate": 3.3000890105288564e-05, "loss": 0.1358, "step": 6300 }, { "grad_norm": 0.10505557060241699, "learning_rate": 3.284548576694908e-05, "loss": 0.1358, "step": 6310 }, { "grad_norm": 0.17435956001281738, "learning_rate": 3.2690269027628815e-05, "loss": 0.1381, "step": 6320 }, { "grad_norm": 0.14125145971775055, "learning_rate": 3.253524158475324e-05, "loss": 0.1394, "step": 6330 }, { "grad_norm": 0.15284080803394318, "learning_rate": 3.238040513367757e-05, "loss": 0.1375, "step": 6340 }, { "grad_norm": 0.19492118060588837, "learning_rate": 3.222576136766843e-05, "loss": 0.1415, "step": 6350 }, { "grad_norm": 0.18184015154838562, "learning_rate": 3.2071311977885324e-05, "loss": 0.137, "step": 6360 }, { "grad_norm": 0.14534787833690643, "learning_rate": 3.191705865336197e-05, "loss": 0.1393, "step": 6370 }, { "grad_norm": 0.1608627289533615, "learning_rate": 3.1763003080988075e-05, "loss": 0.1374, "step": 6380 }, { "grad_norm": 0.1854849010705948, "learning_rate": 3.160914694549063e-05, "loss": 0.1379, "step": 6390 }, { "grad_norm": 0.19459804892539978, "learning_rate": 3.145549192941573e-05, "loss": 0.1397, "step": 6400 }, { "grad_norm": 0.17401166260242462, "learning_rate": 3.130203971310999e-05, "loss": 0.1401, "step": 6410 }, { "grad_norm": 0.133311465382576, "learning_rate": 3.114879197470225e-05, "loss": 0.1392, "step": 6420 }, { "grad_norm": 0.14735054969787598, "learning_rate": 3.0995750390085285e-05, "loss": 0.1366, "step": 6430 }, { "grad_norm": 0.15537108480930328, "learning_rate": 3.084291663289728e-05, "loss": 0.1379, "step": 6440 }, { "grad_norm": 0.11261343955993652, "learning_rate": 3.069029237450375e-05, "loss": 0.1366, "step": 6450 }, { "grad_norm": 0.18856383860111237, "learning_rate": 3.053787928397911e-05, "loss": 0.1372, "step": 6460 }, { "grad_norm": 0.16627824306488037, "learning_rate": 3.0385679028088526e-05, "loss": 0.1338, "step": 6470 }, { "grad_norm": 0.18101073801517487, "learning_rate": 3.023369327126959e-05, "loss": 0.1366, "step": 6480 }, { "grad_norm": 0.1641179770231247, "learning_rate": 3.0081923675614198e-05, "loss": 0.1383, "step": 6490 }, { "grad_norm": 0.20374365150928497, "learning_rate": 2.993037190085034e-05, "loss": 0.1377, "step": 6500 }, { "grad_norm": 0.14802561700344086, "learning_rate": 2.977903960432392e-05, "loss": 0.1386, "step": 6510 }, { "grad_norm": 0.13251923024654388, "learning_rate": 2.9627928440980722e-05, "loss": 0.1383, "step": 6520 }, { "grad_norm": 0.16976147890090942, "learning_rate": 2.9477040063348183e-05, "loss": 0.1366, "step": 6530 }, { "grad_norm": 0.13962304592132568, "learning_rate": 2.9326376121517456e-05, "loss": 0.1348, "step": 6540 }, { "grad_norm": 0.1282155066728592, "learning_rate": 2.9175938263125236e-05, "loss": 0.1386, "step": 6550 }, { "grad_norm": 0.17909574508666992, "learning_rate": 2.9025728133335873e-05, "loss": 0.1375, "step": 6560 }, { "grad_norm": 0.1561977118253708, "learning_rate": 2.8875747374823288e-05, "loss": 0.1395, "step": 6570 }, { "grad_norm": 0.1899399608373642, "learning_rate": 2.872599762775298e-05, "loss": 0.1352, "step": 6580 }, { "grad_norm": 0.13712243735790253, "learning_rate": 2.857648052976425e-05, "loss": 0.137, "step": 6590 }, { "grad_norm": 0.14623521268367767, "learning_rate": 2.8427197715952047e-05, "loss": 0.1387, "step": 6600 }, { "grad_norm": 0.14085721969604492, "learning_rate": 2.8278150818849393e-05, "loss": 0.1372, "step": 6610 }, { "grad_norm": 0.14713753759860992, "learning_rate": 2.812934146840922e-05, "loss": 0.1401, "step": 6620 }, { "grad_norm": 0.15419937670230865, "learning_rate": 2.7980771291986764e-05, "loss": 0.1392, "step": 6630 }, { "grad_norm": 0.1897653043270111, "learning_rate": 2.783244191432167e-05, "loss": 0.1381, "step": 6640 }, { "grad_norm": 0.19099217653274536, "learning_rate": 2.768435495752022e-05, "loss": 0.1405, "step": 6650 }, { "grad_norm": 0.165348619222641, "learning_rate": 2.753651204103771e-05, "loss": 0.1359, "step": 6660 }, { "grad_norm": 0.1346958875656128, "learning_rate": 2.7388914781660523e-05, "loss": 0.137, "step": 6670 }, { "grad_norm": 0.19271200895309448, "learning_rate": 2.7241564793488693e-05, "loss": 0.1349, "step": 6680 }, { "grad_norm": 0.15344682335853577, "learning_rate": 2.7094463687918037e-05, "loss": 0.1382, "step": 6690 }, { "grad_norm": 0.12133382260799408, "learning_rate": 2.694761307362268e-05, "loss": 0.1356, "step": 6700 }, { "grad_norm": 0.16535556316375732, "learning_rate": 2.6801014556537467e-05, "loss": 0.1388, "step": 6710 }, { "grad_norm": 0.18111754953861237, "learning_rate": 2.6654669739840243e-05, "loss": 0.1404, "step": 6720 }, { "grad_norm": 0.1553724706172943, "learning_rate": 2.650858022393451e-05, "loss": 0.1383, "step": 6730 }, { "grad_norm": 0.15397483110427856, "learning_rate": 2.6362747606431747e-05, "loss": 0.1382, "step": 6740 }, { "grad_norm": 0.11917636543512344, "learning_rate": 2.6217173482134172e-05, "loss": 0.1403, "step": 6750 }, { "grad_norm": 0.15791700780391693, "learning_rate": 2.6071859443017044e-05, "loss": 0.1388, "step": 6760 }, { "grad_norm": 0.14545629918575287, "learning_rate": 2.5926807078211414e-05, "loss": 0.1405, "step": 6770 }, { "grad_norm": 0.1955435574054718, "learning_rate": 2.5782017973986728e-05, "loss": 0.1387, "step": 6780 }, { "grad_norm": 0.1505528688430786, "learning_rate": 2.5637493713733374e-05, "loss": 0.1361, "step": 6790 }, { "grad_norm": 0.15487198531627655, "learning_rate": 2.549323587794559e-05, "loss": 0.1385, "step": 6800 }, { "grad_norm": 0.14913375675678253, "learning_rate": 2.5349246044203895e-05, "loss": 0.1386, "step": 6810 }, { "grad_norm": 0.11209173500537872, "learning_rate": 2.520552578715808e-05, "loss": 0.1387, "step": 6820 }, { "grad_norm": 0.1637064516544342, "learning_rate": 2.506207667850981e-05, "loss": 0.1377, "step": 6830 }, { "grad_norm": 0.18081381916999817, "learning_rate": 2.4918900286995555e-05, "loss": 0.1371, "step": 6840 }, { "grad_norm": 0.12709639966487885, "learning_rate": 2.4775998178369458e-05, "loss": 0.1361, "step": 6850 }, { "grad_norm": 0.17958208918571472, "learning_rate": 2.4633371915386017e-05, "loss": 0.1368, "step": 6860 }, { "grad_norm": 0.14783644676208496, "learning_rate": 2.4491023057783235e-05, "loss": 0.1382, "step": 6870 }, { "grad_norm": 0.15946941077709198, "learning_rate": 2.4348953162265375e-05, "loss": 0.138, "step": 6880 }, { "grad_norm": 0.08897168189287186, "learning_rate": 2.420716378248607e-05, "loss": 0.1355, "step": 6890 }, { "grad_norm": 0.15280269086360931, "learning_rate": 2.4065656469031266e-05, "loss": 0.1404, "step": 6900 }, { "grad_norm": 0.19151926040649414, "learning_rate": 2.3924432769402268e-05, "loss": 0.1364, "step": 6910 }, { "grad_norm": 0.12708789110183716, "learning_rate": 2.3783494227998844e-05, "loss": 0.1379, "step": 6920 }, { "grad_norm": 0.15859632194042206, "learning_rate": 2.3642842386102264e-05, "loss": 0.1377, "step": 6930 }, { "grad_norm": 0.1212797686457634, "learning_rate": 2.3502478781858567e-05, "loss": 0.1346, "step": 6940 }, { "grad_norm": 0.17150767147541046, "learning_rate": 2.3362404950261628e-05, "loss": 0.1388, "step": 6950 }, { "grad_norm": 0.20891523361206055, "learning_rate": 2.3222622423136458e-05, "loss": 0.1378, "step": 6960 }, { "grad_norm": 0.14243164658546448, "learning_rate": 2.3083132729122332e-05, "loss": 0.1367, "step": 6970 }, { "grad_norm": 0.18947196006774902, "learning_rate": 2.294393739365621e-05, "loss": 0.1405, "step": 6980 }, { "grad_norm": 0.15855497121810913, "learning_rate": 2.2805037938956e-05, "loss": 0.1343, "step": 6990 }, { "grad_norm": 0.15863677859306335, "learning_rate": 2.266643588400386e-05, "loss": 0.1366, "step": 7000 }, { "grad_norm": 0.18190181255340576, "learning_rate": 2.252813274452969e-05, "loss": 0.1379, "step": 7010 }, { "grad_norm": 0.17462483048439026, "learning_rate": 2.2390130032994427e-05, "loss": 0.138, "step": 7020 }, { "grad_norm": 0.16822536289691925, "learning_rate": 2.2252429258573633e-05, "loss": 0.1342, "step": 7030 }, { "grad_norm": 0.13760408759117126, "learning_rate": 2.2115031927140904e-05, "loss": 0.1361, "step": 7040 }, { "grad_norm": 0.16947977244853973, "learning_rate": 2.1977939541251463e-05, "loss": 0.1383, "step": 7050 }, { "grad_norm": 0.17488440871238708, "learning_rate": 2.1841153600125684e-05, "loss": 0.1345, "step": 7060 }, { "grad_norm": 0.16400085389614105, "learning_rate": 2.170467559963267e-05, "loss": 0.1397, "step": 7070 }, { "grad_norm": 0.15305553376674652, "learning_rate": 2.1568507032273982e-05, "loss": 0.1362, "step": 7080 }, { "grad_norm": 0.2422412931919098, "learning_rate": 2.1432649387167264e-05, "loss": 0.1372, "step": 7090 }, { "grad_norm": 0.13273100554943085, "learning_rate": 2.1297104150029973e-05, "loss": 0.1373, "step": 7100 }, { "grad_norm": 0.158936008810997, "learning_rate": 2.116187280316307e-05, "loss": 0.1372, "step": 7110 }, { "grad_norm": 0.12613752484321594, "learning_rate": 2.1026956825434908e-05, "loss": 0.1365, "step": 7120 }, { "grad_norm": 0.16345149278640747, "learning_rate": 2.0892357692265017e-05, "loss": 0.1366, "step": 7130 }, { "grad_norm": 0.08742272108793259, "learning_rate": 2.0758076875607947e-05, "loss": 0.1363, "step": 7140 }, { "grad_norm": 0.1709565818309784, "learning_rate": 2.0624115843937207e-05, "loss": 0.1357, "step": 7150 }, { "grad_norm": 0.20790213346481323, "learning_rate": 2.0490476062229157e-05, "loss": 0.1389, "step": 7160 }, { "grad_norm": 0.17578555643558502, "learning_rate": 2.035715899194704e-05, "loss": 0.1385, "step": 7170 }, { "grad_norm": 0.1889265775680542, "learning_rate": 2.022416609102499e-05, "loss": 0.1356, "step": 7180 }, { "grad_norm": 0.11250610649585724, "learning_rate": 2.009149881385205e-05, "loss": 0.1389, "step": 7190 }, { "grad_norm": 0.17334629595279694, "learning_rate": 1.995915861125634e-05, "loss": 0.139, "step": 7200 }, { "grad_norm": 0.1644093096256256, "learning_rate": 1.9827146930489065e-05, "loss": 0.1386, "step": 7210 }, { "grad_norm": 0.16419662535190582, "learning_rate": 1.9695465215208848e-05, "loss": 0.1321, "step": 7220 }, { "grad_norm": 0.16467627882957458, "learning_rate": 1.9564114905465813e-05, "loss": 0.1402, "step": 7230 }, { "grad_norm": 0.10572504252195358, "learning_rate": 1.9433097437685936e-05, "loss": 0.1406, "step": 7240 }, { "grad_norm": 0.13015249371528625, "learning_rate": 1.930241424465521e-05, "loss": 0.1344, "step": 7250 }, { "grad_norm": 0.13751178979873657, "learning_rate": 1.9172066755504115e-05, "loss": 0.136, "step": 7260 }, { "grad_norm": 0.15921659767627716, "learning_rate": 1.9042056395691914e-05, "loss": 0.1365, "step": 7270 }, { "grad_norm": 0.17845675349235535, "learning_rate": 1.8912384586991066e-05, "loss": 0.1375, "step": 7280 }, { "grad_norm": 0.1042633056640625, "learning_rate": 1.8783052747471717e-05, "loss": 0.138, "step": 7290 }, { "grad_norm": 0.08751855045557022, "learning_rate": 1.865406229148611e-05, "loss": 0.1368, "step": 7300 }, { "grad_norm": 0.16895350813865662, "learning_rate": 1.8525414629653233e-05, "loss": 0.1371, "step": 7310 }, { "grad_norm": 0.19929863512516022, "learning_rate": 1.8397111168843255e-05, "loss": 0.1407, "step": 7320 }, { "grad_norm": 0.10168038308620453, "learning_rate": 1.8269153312162323e-05, "loss": 0.1362, "step": 7330 }, { "grad_norm": 0.1354413777589798, "learning_rate": 1.8141542458937054e-05, "loss": 0.1356, "step": 7340 }, { "grad_norm": 0.12250649183988571, "learning_rate": 1.8014280004699268e-05, "loss": 0.1371, "step": 7350 }, { "grad_norm": 0.1782764047384262, "learning_rate": 1.788736734117078e-05, "loss": 0.1431, "step": 7360 }, { "grad_norm": 0.18234555423259735, "learning_rate": 1.7760805856248152e-05, "loss": 0.1417, "step": 7370 }, { "grad_norm": 0.14732302725315094, "learning_rate": 1.7634596933987518e-05, "loss": 0.137, "step": 7380 }, { "grad_norm": 0.1159769669175148, "learning_rate": 1.7508741954589404e-05, "loss": 0.1384, "step": 7390 }, { "grad_norm": 0.11164289712905884, "learning_rate": 1.7383242294383717e-05, "loss": 0.1353, "step": 7400 }, { "grad_norm": 0.15245497226715088, "learning_rate": 1.7258099325814632e-05, "loss": 0.1359, "step": 7410 }, { "grad_norm": 0.13275210559368134, "learning_rate": 1.7133314417425594e-05, "loss": 0.1356, "step": 7420 }, { "grad_norm": 0.12145767360925674, "learning_rate": 1.7008888933844408e-05, "loss": 0.1366, "step": 7430 }, { "grad_norm": 0.10657090693712234, "learning_rate": 1.6884824235768172e-05, "loss": 0.1349, "step": 7440 }, { "grad_norm": 0.096003957092762, "learning_rate": 1.6761121679948592e-05, "loss": 0.1351, "step": 7450 }, { "grad_norm": 0.1958954632282257, "learning_rate": 1.663778261917695e-05, "loss": 0.1411, "step": 7460 }, { "grad_norm": 0.18981142342090607, "learning_rate": 1.651480840226952e-05, "loss": 0.1372, "step": 7470 }, { "grad_norm": 0.11200006306171417, "learning_rate": 1.639220037405258e-05, "loss": 0.1371, "step": 7480 }, { "grad_norm": 0.10354040563106537, "learning_rate": 1.6269959875347906e-05, "loss": 0.1341, "step": 7490 }, { "grad_norm": 0.09363014996051788, "learning_rate": 1.614808824295802e-05, "loss": 0.136, "step": 7500 }, { "grad_norm": 0.09933657199144363, "learning_rate": 1.602658680965152e-05, "loss": 0.1343, "step": 7510 }, { "grad_norm": 0.10720168799161911, "learning_rate": 1.5905456904148686e-05, "loss": 0.1388, "step": 7520 }, { "grad_norm": 0.11268419772386551, "learning_rate": 1.57846998511067e-05, "loss": 0.1366, "step": 7530 }, { "grad_norm": 0.08695314079523087, "learning_rate": 1.566431697110538e-05, "loss": 0.1359, "step": 7540 }, { "grad_norm": 0.17678678035736084, "learning_rate": 1.554430958063259e-05, "loss": 0.135, "step": 7550 }, { "grad_norm": 0.1328829526901245, "learning_rate": 1.5424678992069912e-05, "loss": 0.1392, "step": 7560 }, { "grad_norm": 0.11767739802598953, "learning_rate": 1.5305426513678362e-05, "loss": 0.1359, "step": 7570 }, { "grad_norm": 0.10485532879829407, "learning_rate": 1.518655344958388e-05, "loss": 0.1347, "step": 7580 }, { "grad_norm": 0.1387827843427658, "learning_rate": 1.5068061099763275e-05, "loss": 0.135, "step": 7590 }, { "grad_norm": 0.14451664686203003, "learning_rate": 1.494995076002988e-05, "loss": 0.1384, "step": 7600 }, { "grad_norm": 0.12369150668382645, "learning_rate": 1.4832223722019456e-05, "loss": 0.135, "step": 7610 }, { "grad_norm": 0.1272563636302948, "learning_rate": 1.4714881273176035e-05, "loss": 0.1364, "step": 7620 }, { "grad_norm": 0.14464479684829712, "learning_rate": 1.4597924696737835e-05, "loss": 0.1363, "step": 7630 }, { "grad_norm": 0.15380112826824188, "learning_rate": 1.4481355271723252e-05, "loss": 0.1377, "step": 7640 }, { "grad_norm": 0.1084940955042839, "learning_rate": 1.4365174272916809e-05, "loss": 0.1369, "step": 7650 }, { "grad_norm": 0.1955633908510208, "learning_rate": 1.4249382970855319e-05, "loss": 0.1399, "step": 7660 }, { "grad_norm": 0.1308707296848297, "learning_rate": 1.4133982631813903e-05, "loss": 0.1372, "step": 7670 }, { "grad_norm": 0.14867974817752838, "learning_rate": 1.4018974517792194e-05, "loss": 0.1348, "step": 7680 }, { "grad_norm": 0.11717529594898224, "learning_rate": 1.390435988650048e-05, "loss": 0.1358, "step": 7690 }, { "grad_norm": 0.16398921608924866, "learning_rate": 1.3790139991346006e-05, "loss": 0.139, "step": 7700 }, { "grad_norm": 0.12998664379119873, "learning_rate": 1.367631608141926e-05, "loss": 0.1389, "step": 7710 }, { "grad_norm": 0.1321418434381485, "learning_rate": 1.3562889401480278e-05, "loss": 0.1348, "step": 7720 }, { "grad_norm": 0.1413685828447342, "learning_rate": 1.3449861191945074e-05, "loss": 0.1376, "step": 7730 }, { "grad_norm": 0.13793465495109558, "learning_rate": 1.3337232688872009e-05, "loss": 0.1364, "step": 7740 }, { "grad_norm": 0.12822720408439636, "learning_rate": 1.3225005123948364e-05, "loss": 0.1374, "step": 7750 }, { "grad_norm": 0.11293677240610123, "learning_rate": 1.311317972447681e-05, "loss": 0.1367, "step": 7760 }, { "grad_norm": 0.1273677796125412, "learning_rate": 1.3001757713361996e-05, "loss": 0.1404, "step": 7770 }, { "grad_norm": 0.0994945764541626, "learning_rate": 1.2890740309097204e-05, "loss": 0.1366, "step": 7780 }, { "grad_norm": 0.11453263461589813, "learning_rate": 1.2780128725750944e-05, "loss": 0.1361, "step": 7790 }, { "grad_norm": 0.10319621115922928, "learning_rate": 1.266992417295379e-05, "loss": 0.1377, "step": 7800 }, { "grad_norm": 0.11687227338552475, "learning_rate": 1.2560127855885073e-05, "loss": 0.1374, "step": 7810 }, { "grad_norm": 0.13098742067813873, "learning_rate": 1.2450740975259745e-05, "loss": 0.1354, "step": 7820 }, { "grad_norm": 0.11969354003667831, "learning_rate": 1.234176472731517e-05, "loss": 0.1378, "step": 7830 }, { "grad_norm": 0.11732219904661179, "learning_rate": 1.2233200303798158e-05, "loss": 0.1353, "step": 7840 }, { "grad_norm": 0.12243682146072388, "learning_rate": 1.2125048891951846e-05, "loss": 0.1367, "step": 7850 }, { "grad_norm": 0.12654563784599304, "learning_rate": 1.2017311674502745e-05, "loss": 0.1368, "step": 7860 }, { "grad_norm": 0.12742966413497925, "learning_rate": 1.1909989829647822e-05, "loss": 0.1355, "step": 7870 }, { "grad_norm": 0.1389680653810501, "learning_rate": 1.1803084531041553e-05, "loss": 0.1369, "step": 7880 }, { "grad_norm": 0.14346174895763397, "learning_rate": 1.1696596947783162e-05, "loss": 0.1384, "step": 7890 }, { "grad_norm": 0.211452454328537, "learning_rate": 1.1590528244403803e-05, "loss": 0.1388, "step": 7900 }, { "grad_norm": 0.12816539406776428, "learning_rate": 1.148487958085382e-05, "loss": 0.1365, "step": 7910 }, { "grad_norm": 0.1138722151517868, "learning_rate": 1.1379652112490086e-05, "loss": 0.1363, "step": 7920 }, { "grad_norm": 0.13151313364505768, "learning_rate": 1.1274846990063315e-05, "loss": 0.1363, "step": 7930 }, { "grad_norm": 0.1647166609764099, "learning_rate": 1.117046535970554e-05, "loss": 0.1368, "step": 7940 }, { "grad_norm": 0.1125076413154602, "learning_rate": 1.106650836291755e-05, "loss": 0.1372, "step": 7950 }, { "grad_norm": 0.12292832881212234, "learning_rate": 1.0962977136556418e-05, "loss": 0.139, "step": 7960 }, { "grad_norm": 0.1617049276828766, "learning_rate": 1.0859872812823024e-05, "loss": 0.1374, "step": 7970 }, { "grad_norm": 0.13567329943180084, "learning_rate": 1.0757196519249747e-05, "loss": 0.1387, "step": 7980 }, { "grad_norm": 0.15606868267059326, "learning_rate": 1.0654949378688077e-05, "loss": 0.1361, "step": 7990 }, { "grad_norm": 0.10712672024965286, "learning_rate": 1.0553132509296376e-05, "loss": 0.135, "step": 8000 }, { "grad_norm": 0.09787201136350632, "learning_rate": 1.0451747024527613e-05, "loss": 0.1337, "step": 8010 }, { "grad_norm": 0.13271142542362213, "learning_rate": 1.0350794033117189e-05, "loss": 0.1371, "step": 8020 }, { "grad_norm": 0.10905721783638, "learning_rate": 1.0250274639070856e-05, "loss": 0.1355, "step": 8030 }, { "grad_norm": 0.1483287364244461, "learning_rate": 1.0150189941652599e-05, "loss": 0.1369, "step": 8040 }, { "grad_norm": 0.114744633436203, "learning_rate": 1.0050541035372635e-05, "loss": 0.1346, "step": 8050 }, { "grad_norm": 0.13219349086284637, "learning_rate": 9.951329009975458e-06, "loss": 0.1354, "step": 8060 }, { "grad_norm": 0.1264726221561432, "learning_rate": 9.852554950427845e-06, "loss": 0.1382, "step": 8070 }, { "grad_norm": 0.11584024131298065, "learning_rate": 9.754219936907105e-06, "loss": 0.1378, "step": 8080 }, { "grad_norm": 0.13420680165290833, "learning_rate": 9.656325044789194e-06, "loss": 0.1377, "step": 8090 }, { "grad_norm": 0.10516663640737534, "learning_rate": 9.55887134463697e-06, "loss": 0.1379, "step": 8100 }, { "grad_norm": 0.10943170636892319, "learning_rate": 9.461859902188475e-06, "loss": 0.1335, "step": 8110 }, { "grad_norm": 0.1395423263311386, "learning_rate": 9.365291778345303e-06, "loss": 0.1383, "step": 8120 }, { "grad_norm": 0.12454972416162491, "learning_rate": 9.269168029160991e-06, "loss": 0.1361, "step": 8130 }, { "grad_norm": 0.11044501513242722, "learning_rate": 9.173489705829447e-06, "loss": 0.1346, "step": 8140 }, { "grad_norm": 0.0971987247467041, "learning_rate": 9.078257854673516e-06, "loss": 0.1369, "step": 8150 }, { "grad_norm": 0.11135916411876678, "learning_rate": 8.983473517133429e-06, "loss": 0.1369, "step": 8160 }, { "grad_norm": 0.09740894287824631, "learning_rate": 8.889137729755537e-06, "loss": 0.1355, "step": 8170 }, { "grad_norm": 0.10679884254932404, "learning_rate": 8.79525152418087e-06, "loss": 0.1379, "step": 8180 }, { "grad_norm": 0.13310901820659637, "learning_rate": 8.701815927133961e-06, "loss": 0.1389, "step": 8190 }, { "grad_norm": 0.10424364358186722, "learning_rate": 8.608831960411534e-06, "loss": 0.1365, "step": 8200 }, { "grad_norm": 0.11388423293828964, "learning_rate": 8.516300640871321e-06, "loss": 0.1374, "step": 8210 }, { "grad_norm": 0.1239543929696083, "learning_rate": 8.424222980421038e-06, "loss": 0.137, "step": 8220 }, { "grad_norm": 0.12759342789649963, "learning_rate": 8.332599986007184e-06, "loss": 0.1336, "step": 8230 }, { "grad_norm": 0.13210158050060272, "learning_rate": 8.241432659604203e-06, "loss": 0.1362, "step": 8240 }, { "grad_norm": 0.10422317683696747, "learning_rate": 8.150721998203331e-06, "loss": 0.1373, "step": 8250 }, { "grad_norm": 0.09743431210517883, "learning_rate": 8.06046899380184e-06, "loss": 0.1351, "step": 8260 }, { "grad_norm": 0.1178847923874855, "learning_rate": 7.970674633392133e-06, "loss": 0.1356, "step": 8270 }, { "grad_norm": 0.09533163160085678, "learning_rate": 7.881339898950924e-06, "loss": 0.1348, "step": 8280 }, { "grad_norm": 0.1511492282152176, "learning_rate": 7.792465767428597e-06, "loss": 0.1363, "step": 8290 }, { "grad_norm": 0.13032913208007812, "learning_rate": 7.704053210738376e-06, "loss": 0.1352, "step": 8300 }, { "grad_norm": 0.12915371358394623, "learning_rate": 7.6161031957458494e-06, "loss": 0.1396, "step": 8310 }, { "grad_norm": 0.11599449068307877, "learning_rate": 7.5286166842582605e-06, "loss": 0.1349, "step": 8320 }, { "grad_norm": 0.08395397663116455, "learning_rate": 7.4415946330140814e-06, "loss": 0.1357, "step": 8330 }, { "grad_norm": 0.10078112781047821, "learning_rate": 7.3550379936725644e-06, "loss": 0.1361, "step": 8340 }, { "grad_norm": 0.16352300345897675, "learning_rate": 7.2689477128032035e-06, "loss": 0.1388, "step": 8350 }, { "grad_norm": 0.0999695435166359, "learning_rate": 7.183324731875551e-06, "loss": 0.1388, "step": 8360 }, { "grad_norm": 0.06937150657176971, "learning_rate": 7.098169987248782e-06, "loss": 0.1335, "step": 8370 }, { "grad_norm": 0.11373361945152283, "learning_rate": 7.013484410161553e-06, "loss": 0.1359, "step": 8380 }, { "grad_norm": 0.08429645001888275, "learning_rate": 6.92926892672176e-06, "loss": 0.1374, "step": 8390 }, { "grad_norm": 0.06724434345960617, "learning_rate": 6.845524457896446e-06, "loss": 0.1363, "step": 8400 }, { "grad_norm": 0.14312094449996948, "learning_rate": 6.7622519195017165e-06, "loss": 0.1392, "step": 8410 }, { "grad_norm": 0.17622436583042145, "learning_rate": 6.679452222192684e-06, "loss": 0.1362, "step": 8420 }, { "grad_norm": 0.1358077973127365, "learning_rate": 6.597126271453579e-06, "loss": 0.1367, "step": 8430 }, { "grad_norm": 0.1365446150302887, "learning_rate": 6.51527496758782e-06, "loss": 0.1374, "step": 8440 }, { "grad_norm": 0.1713290810585022, "learning_rate": 6.433899205708155e-06, "loss": 0.1442, "step": 8450 }, { "grad_norm": 0.11505335569381714, "learning_rate": 6.352999875726856e-06, "loss": 0.1359, "step": 8460 }, { "grad_norm": 0.13365548849105835, "learning_rate": 6.272577862346052e-06, "loss": 0.1363, "step": 8470 }, { "grad_norm": 0.11833521723747253, "learning_rate": 6.192634045047996e-06, "loss": 0.1359, "step": 8480 }, { "grad_norm": 0.11864876002073288, "learning_rate": 6.113169298085458e-06, "loss": 0.1372, "step": 8490 }, { "grad_norm": 0.15620292723178864, "learning_rate": 6.034184490472195e-06, "loss": 0.1397, "step": 8500 }, { "grad_norm": 0.0901854932308197, "learning_rate": 5.955680485973386e-06, "loss": 0.1381, "step": 8510 }, { "grad_norm": 0.10469307005405426, "learning_rate": 5.877658143096265e-06, "loss": 0.1364, "step": 8520 }, { "grad_norm": 0.11437556147575378, "learning_rate": 5.800118315080661e-06, "loss": 0.1351, "step": 8530 }, { "grad_norm": 0.10484666377305984, "learning_rate": 5.723061849889716e-06, "loss": 0.1385, "step": 8540 }, { "grad_norm": 0.12785661220550537, "learning_rate": 5.646489590200604e-06, "loss": 0.1349, "step": 8550 }, { "grad_norm": 0.12281761318445206, "learning_rate": 5.570402373395256e-06, "loss": 0.1365, "step": 8560 }, { "grad_norm": 0.08600356429815292, "learning_rate": 5.494801031551305e-06, "loss": 0.1399, "step": 8570 }, { "grad_norm": 0.14602923393249512, "learning_rate": 5.41968639143291e-06, "loss": 0.1365, "step": 8580 }, { "grad_norm": 0.16058653593063354, "learning_rate": 5.345059274481751e-06, "loss": 0.1349, "step": 8590 }, { "grad_norm": 0.11875080317258835, "learning_rate": 5.270920496808002e-06, "loss": 0.1357, "step": 8600 }, { "grad_norm": 0.09683624655008316, "learning_rate": 5.1972708691814695e-06, "loss": 0.1372, "step": 8610 }, { "grad_norm": 0.11797384917736053, "learning_rate": 5.124111197022674e-06, "loss": 0.1367, "step": 8620 }, { "grad_norm": 0.1560738980770111, "learning_rate": 5.051442280394081e-06, "loss": 0.1386, "step": 8630 }, { "grad_norm": 0.15634621679782867, "learning_rate": 4.979264913991322e-06, "loss": 0.1368, "step": 8640 }, { "grad_norm": 0.13562452793121338, "learning_rate": 4.907579887134489e-06, "loss": 0.1368, "step": 8650 }, { "grad_norm": 0.09580609202384949, "learning_rate": 4.836387983759572e-06, "loss": 0.1363, "step": 8660 }, { "grad_norm": 0.10818251222372055, "learning_rate": 4.765689982409816e-06, "loss": 0.1391, "step": 8670 }, { "grad_norm": 0.1145770251750946, "learning_rate": 4.695486656227233e-06, "loss": 0.138, "step": 8680 }, { "grad_norm": 0.09436694532632828, "learning_rate": 4.625778772944156e-06, "loss": 0.1354, "step": 8690 }, { "grad_norm": 0.11100834608078003, "learning_rate": 4.556567094874825e-06, "loss": 0.1356, "step": 8700 }, { "grad_norm": 0.13518892228603363, "learning_rate": 4.487852378907059e-06, "loss": 0.1353, "step": 8710 }, { "grad_norm": 0.11741168051958084, "learning_rate": 4.419635376493986e-06, "loss": 0.1355, "step": 8720 }, { "grad_norm": 0.17966045439243317, "learning_rate": 4.351916833645825e-06, "loss": 0.1363, "step": 8730 }, { "grad_norm": 0.10683520883321762, "learning_rate": 4.284697490921691e-06, "loss": 0.1368, "step": 8740 }, { "grad_norm": 0.0942920595407486, "learning_rate": 4.2179780834215585e-06, "loss": 0.134, "step": 8750 }, { "grad_norm": 0.09258046001195908, "learning_rate": 4.151759340778178e-06, "loss": 0.1362, "step": 8760 }, { "grad_norm": 0.10762850195169449, "learning_rate": 4.086041987149109e-06, "loss": 0.1392, "step": 8770 }, { "grad_norm": 0.12010587751865387, "learning_rate": 4.020826741208811e-06, "loss": 0.1387, "step": 8780 }, { "grad_norm": 0.1469656527042389, "learning_rate": 3.956114316140746e-06, "loss": 0.1375, "step": 8790 }, { "grad_norm": 0.07427356392145157, "learning_rate": 3.891905419629643e-06, "loss": 0.1347, "step": 8800 }, { "grad_norm": 0.07937679439783096, "learning_rate": 3.8282007538536946e-06, "loss": 0.1347, "step": 8810 }, { "grad_norm": 0.11501304060220718, "learning_rate": 3.7650010154769265e-06, "loss": 0.1368, "step": 8820 }, { "grad_norm": 0.12938648462295532, "learning_rate": 3.7023068956415608e-06, "loss": 0.1371, "step": 8830 }, { "grad_norm": 0.10112431645393372, "learning_rate": 3.6401190799604303e-06, "loss": 0.1399, "step": 8840 }, { "grad_norm": 0.11886759102344513, "learning_rate": 3.578438248509536e-06, "loss": 0.1389, "step": 8850 }, { "grad_norm": 0.16394320130348206, "learning_rate": 3.5172650758205583e-06, "loss": 0.1381, "step": 8860 }, { "grad_norm": 0.13859078288078308, "learning_rate": 3.45660023087353e-06, "loss": 0.1381, "step": 8870 }, { "grad_norm": 0.11666133254766464, "learning_rate": 3.3964443770894528e-06, "loss": 0.1345, "step": 8880 }, { "grad_norm": 0.08694508671760559, "learning_rate": 3.3367981723231245e-06, "loss": 0.1388, "step": 8890 }, { "grad_norm": 0.09554547071456909, "learning_rate": 3.2776622688558746e-06, "loss": 0.1379, "step": 8900 }, { "grad_norm": 0.13130256533622742, "learning_rate": 3.2190373133884677e-06, "loss": 0.1362, "step": 8910 }, { "grad_norm": 0.13983257114887238, "learning_rate": 3.1609239470340446e-06, "loss": 0.1397, "step": 8920 }, { "grad_norm": 0.11059857159852982, "learning_rate": 3.1033228053110373e-06, "loss": 0.1406, "step": 8930 }, { "grad_norm": 0.12198301404714584, "learning_rate": 3.0462345181363314e-06, "loss": 0.1368, "step": 8940 }, { "grad_norm": 0.09540163725614548, "learning_rate": 2.9896597098182654e-06, "loss": 0.1366, "step": 8950 }, { "grad_norm": 0.08712207525968552, "learning_rate": 2.933598999049891e-06, "loss": 0.1367, "step": 8960 }, { "grad_norm": 0.10449322313070297, "learning_rate": 2.8780529989021697e-06, "loss": 0.1389, "step": 8970 }, { "grad_norm": 0.1560509353876114, "learning_rate": 2.823022316817242e-06, "loss": 0.1363, "step": 8980 }, { "grad_norm": 0.11265828460454941, "learning_rate": 2.7685075546018456e-06, "loss": 0.14, "step": 8990 }, { "grad_norm": 0.15037453174591064, "learning_rate": 2.7145093084206598e-06, "loss": 0.1378, "step": 9000 }, { "grad_norm": 0.11329244077205658, "learning_rate": 2.661028168789892e-06, "loss": 0.1382, "step": 9010 }, { "grad_norm": 0.07544533908367157, "learning_rate": 2.6080647205706855e-06, "loss": 0.1387, "step": 9020 }, { "grad_norm": 0.06872090697288513, "learning_rate": 2.555619542962834e-06, "loss": 0.1396, "step": 9030 }, { "grad_norm": 0.1194721907377243, "learning_rate": 2.503693209498409e-06, "loss": 0.1358, "step": 9040 }, { "grad_norm": 0.09280619025230408, "learning_rate": 2.452286288035449e-06, "loss": 0.1369, "step": 9050 }, { "grad_norm": 0.1303427815437317, "learning_rate": 2.4013993407518363e-06, "loss": 0.1354, "step": 9060 }, { "grad_norm": 0.14007259905338287, "learning_rate": 2.351032924139063e-06, "loss": 0.1385, "step": 9070 }, { "grad_norm": 0.1144586056470871, "learning_rate": 2.30118758899619e-06, "loss": 0.1361, "step": 9080 }, { "grad_norm": 0.17019011080265045, "learning_rate": 2.2518638804238157e-06, "loss": 0.1357, "step": 9090 }, { "grad_norm": 0.09577572345733643, "learning_rate": 2.203062337818118e-06, "loss": 0.1347, "step": 9100 }, { "grad_norm": 0.13716194033622742, "learning_rate": 2.1547834948649483e-06, "loss": 0.1355, "step": 9110 }, { "grad_norm": 0.11890782415866852, "learning_rate": 2.1070278795340017e-06, "loss": 0.1379, "step": 9120 }, { "grad_norm": 0.06272739171981812, "learning_rate": 2.059796014073029e-06, "loss": 0.1366, "step": 9130 }, { "grad_norm": 0.06440910696983337, "learning_rate": 2.01308841500214e-06, "loss": 0.1359, "step": 9140 }, { "grad_norm": 0.13654431700706482, "learning_rate": 1.9669055931081704e-06, "loss": 0.1371, "step": 9150 }, { "grad_norm": 0.07405722141265869, "learning_rate": 1.9212480534390507e-06, "loss": 0.1357, "step": 9160 }, { "grad_norm": 0.10964131355285645, "learning_rate": 1.8761162952983246e-06, "loss": 0.1374, "step": 9170 }, { "grad_norm": 0.07633847743272781, "learning_rate": 1.8315108122396618e-06, "loss": 0.1363, "step": 9180 }, { "grad_norm": 0.1478925198316574, "learning_rate": 1.787432092061475e-06, "loss": 0.1377, "step": 9190 }, { "grad_norm": 0.07084264606237411, "learning_rate": 1.743880616801602e-06, "loss": 0.1361, "step": 9200 }, { "grad_norm": 0.07236980646848679, "learning_rate": 1.7008568627319865e-06, "loss": 0.1379, "step": 9210 }, { "grad_norm": 0.12144934386014938, "learning_rate": 1.6583613003535226e-06, "loss": 0.136, "step": 9220 }, { "grad_norm": 0.09035366028547287, "learning_rate": 1.6163943943908522e-06, "loss": 0.1369, "step": 9230 }, { "grad_norm": 0.14795254170894623, "learning_rate": 1.5749566037873476e-06, "loss": 0.1349, "step": 9240 }, { "grad_norm": 0.09773296862840652, "learning_rate": 1.5340483817000428e-06, "loss": 0.1375, "step": 9250 }, { "grad_norm": 0.09650065749883652, "learning_rate": 1.4936701754947101e-06, "loss": 0.1369, "step": 9260 }, { "grad_norm": 0.1255313605070114, "learning_rate": 1.4538224267409361e-06, "loss": 0.1386, "step": 9270 }, { "grad_norm": 0.09544844925403595, "learning_rate": 1.414505571207314e-06, "loss": 0.1369, "step": 9280 }, { "grad_norm": 0.12606243789196014, "learning_rate": 1.3757200388566816e-06, "loss": 0.1356, "step": 9290 }, { "grad_norm": 0.1130032017827034, "learning_rate": 1.3374662538414074e-06, "loss": 0.1353, "step": 9300 }, { "grad_norm": 0.13287268579006195, "learning_rate": 1.2997446344987617e-06, "loss": 0.1362, "step": 9310 }, { "grad_norm": 0.07898017019033432, "learning_rate": 1.262555593346315e-06, "loss": 0.1361, "step": 9320 }, { "grad_norm": 0.12233997136354446, "learning_rate": 1.2258995370774685e-06, "loss": 0.1366, "step": 9330 }, { "grad_norm": 0.07027930021286011, "learning_rate": 1.1897768665569798e-06, "loss": 0.1407, "step": 9340 }, { "grad_norm": 0.12403062731027603, "learning_rate": 1.1541879768165954e-06, "loss": 0.137, "step": 9350 }, { "grad_norm": 0.09209410101175308, "learning_rate": 1.1191332570507085e-06, "loss": 0.1345, "step": 9360 }, { "grad_norm": 0.08304783701896667, "learning_rate": 1.0846130906121132e-06, "loss": 0.1349, "step": 9370 }, { "grad_norm": 0.10310087352991104, "learning_rate": 1.0506278550078131e-06, "loss": 0.1352, "step": 9380 }, { "grad_norm": 0.12324751913547516, "learning_rate": 1.0171779218949185e-06, "loss": 0.1367, "step": 9390 }, { "grad_norm": 0.11066099256277084, "learning_rate": 9.842636570765174e-07, "loss": 0.1393, "step": 9400 }, { "grad_norm": 0.1072869822382927, "learning_rate": 9.518854204977612e-07, "loss": 0.1367, "step": 9410 }, { "grad_norm": 0.07194332778453827, "learning_rate": 9.200435662418349e-07, "loss": 0.1322, "step": 9420 }, { "grad_norm": 0.08593005686998367, "learning_rate": 8.887384425261658e-07, "loss": 0.1348, "step": 9430 }, { "grad_norm": 0.13954207301139832, "learning_rate": 8.579703916985648e-07, "loss": 0.1377, "step": 9440 }, { "grad_norm": 0.11076579988002777, "learning_rate": 8.277397502335194e-07, "loss": 0.1341, "step": 9450 }, { "grad_norm": 0.09968645125627518, "learning_rate": 7.980468487284675e-07, "loss": 0.1369, "step": 9460 }, { "grad_norm": 0.05683652684092522, "learning_rate": 7.688920119002297e-07, "loss": 0.1363, "step": 9470 }, { "grad_norm": 0.08242069184780121, "learning_rate": 7.402755585814269e-07, "loss": 0.1356, "step": 9480 }, { "grad_norm": 0.0722484216094017, "learning_rate": 7.121978017170073e-07, "loss": 0.1358, "step": 9490 }, { "grad_norm": 0.06697814911603928, "learning_rate": 6.846590483608306e-07, "loss": 0.1345, "step": 9500 }, { "grad_norm": 0.12655308842658997, "learning_rate": 6.576595996722834e-07, "loss": 0.1346, "step": 9510 }, { "grad_norm": 0.08260273188352585, "learning_rate": 6.311997509130141e-07, "loss": 0.1368, "step": 9520 }, { "grad_norm": 0.10377214103937149, "learning_rate": 6.052797914436803e-07, "loss": 0.136, "step": 9530 }, { "grad_norm": 0.13638345897197723, "learning_rate": 5.799000047208181e-07, "loss": 0.1366, "step": 9540 }, { "grad_norm": 0.163876011967659, "learning_rate": 5.550606682937054e-07, "loss": 0.1353, "step": 9550 }, { "grad_norm": 0.07725352048873901, "learning_rate": 5.307620538013481e-07, "loss": 0.1344, "step": 9560 }, { "grad_norm": 0.07766133546829224, "learning_rate": 5.070044269694874e-07, "loss": 0.1369, "step": 9570 }, { "grad_norm": 0.10108968615531921, "learning_rate": 4.837880476077417e-07, "loss": 0.1353, "step": 9580 }, { "grad_norm": 0.10944552719593048, "learning_rate": 4.6111316960670835e-07, "loss": 0.1367, "step": 9590 }, { "grad_norm": 0.08827920258045197, "learning_rate": 4.389800409352218e-07, "loss": 0.1363, "step": 9600 }, { "grad_norm": 0.09800177812576294, "learning_rate": 4.173889036376277e-07, "loss": 0.133, "step": 9610 }, { "grad_norm": 0.08752765506505966, "learning_rate": 3.963399938311463e-07, "loss": 0.1355, "step": 9620 }, { "grad_norm": 0.09335701167583466, "learning_rate": 3.7583354170328545e-07, "loss": 0.1365, "step": 9630 }, { "grad_norm": 0.12018894404172897, "learning_rate": 3.558697715093207e-07, "loss": 0.1369, "step": 9640 }, { "grad_norm": 0.1076042428612709, "learning_rate": 3.3644890156983576e-07, "loss": 0.1357, "step": 9650 }, { "grad_norm": 0.10124968737363815, "learning_rate": 3.175711442683638e-07, "loss": 0.1342, "step": 9660 }, { "grad_norm": 0.05939052253961563, "learning_rate": 2.9923670604902197e-07, "loss": 0.1381, "step": 9670 }, { "grad_norm": 0.09335540235042572, "learning_rate": 2.814457874143028e-07, "loss": 0.1356, "step": 9680 }, { "grad_norm": 0.10526993125677109, "learning_rate": 2.641985829228366e-07, "loss": 0.1349, "step": 9690 }, { "grad_norm": 0.08347027003765106, "learning_rate": 2.474952811872877e-07, "loss": 0.1352, "step": 9700 }, { "grad_norm": 0.08271834999322891, "learning_rate": 2.3133606487228397e-07, "loss": 0.1378, "step": 9710 }, { "grad_norm": 0.0988229513168335, "learning_rate": 2.157211106924295e-07, "loss": 0.136, "step": 9720 }, { "grad_norm": 0.10028582811355591, "learning_rate": 2.006505894103672e-07, "loss": 0.1349, "step": 9730 }, { "grad_norm": 0.10500890016555786, "learning_rate": 1.8612466583489696e-07, "loss": 0.1399, "step": 9740 }, { "grad_norm": 0.1178087517619133, "learning_rate": 1.7214349881918834e-07, "loss": 0.1384, "step": 9750 }, { "grad_norm": 0.04222963750362396, "learning_rate": 1.5870724125904845e-07, "loss": 0.1365, "step": 9760 }, { "grad_norm": 0.12475298345088959, "learning_rate": 1.4581604009124006e-07, "loss": 0.1366, "step": 9770 }, { "grad_norm": 0.10404250770807266, "learning_rate": 1.334700362918717e-07, "loss": 0.1391, "step": 9780 }, { "grad_norm": 0.09343831241130829, "learning_rate": 1.2166936487486015e-07, "loss": 0.137, "step": 9790 }, { "grad_norm": 0.12810835242271423, "learning_rate": 1.1041415489045914e-07, "loss": 0.1349, "step": 9800 }, { "grad_norm": 0.09085965156555176, "learning_rate": 9.970452942384412e-08, "loss": 0.1351, "step": 9810 }, { "grad_norm": 0.06877006590366364, "learning_rate": 8.954060559375754e-08, "loss": 0.1349, "step": 9820 }, { "grad_norm": 0.11306249350309372, "learning_rate": 7.99224945512489e-08, "loss": 0.1353, "step": 9830 }, { "grad_norm": 0.05656269192695618, "learning_rate": 7.085030147843675e-08, "loss": 0.1358, "step": 9840 }, { "grad_norm": 0.09066634625196457, "learning_rate": 6.232412558736523e-08, "loss": 0.138, "step": 9850 }, { "grad_norm": 0.07955700904130936, "learning_rate": 5.434406011893822e-08, "loss": 0.1363, "step": 9860 }, { "grad_norm": 0.1255960464477539, "learning_rate": 4.6910192341864664e-08, "loss": 0.138, "step": 9870 }, { "grad_norm": 0.10786505043506622, "learning_rate": 4.0022603551737035e-08, "loss": 0.1375, "step": 9880 }, { "grad_norm": 0.056810472160577774, "learning_rate": 3.3681369070120985e-08, "loss": 0.1376, "step": 9890 }, { "grad_norm": 0.0823664590716362, "learning_rate": 2.7886558243744866e-08, "loss": 0.1375, "step": 9900 }, { "grad_norm": 0.13539406657218933, "learning_rate": 2.2638234443722596e-08, "loss": 0.1362, "step": 9910 }, { "grad_norm": 0.057054609060287476, "learning_rate": 1.7936455064887504e-08, "loss": 0.1385, "step": 9920 }, { "grad_norm": 0.10754342377185822, "learning_rate": 1.378127152514841e-08, "loss": 0.1393, "step": 9930 }, { "grad_norm": 0.07002149522304535, "learning_rate": 1.0172729264917857e-08, "loss": 0.1366, "step": 9940 }, { "grad_norm": 0.07984930276870728, "learning_rate": 7.1108677466458215e-09, "loss": 0.1368, "step": 9950 }, { "grad_norm": 0.1085810512304306, "learning_rate": 4.595720454353414e-09, "loss": 0.1357, "step": 9960 }, { "grad_norm": 0.07886433601379395, "learning_rate": 2.627314893294264e-09, "loss": 0.1341, "step": 9970 }, { "grad_norm": 0.0967990905046463, "learning_rate": 1.2056725896270048e-09, "loss": 0.1358, "step": 9980 }, { "grad_norm": 0.08520369976758957, "learning_rate": 3.308090902098826e-10, "loss": 0.1343, "step": 9990 }, { "grad_norm": 0.06365083158016205, "learning_rate": 2.7339624120159555e-12, "loss": 0.1361, "step": 10000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }