{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6114260238201389, "eval_steps": 10000.0, "global_step": 6000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0010190433730335648, "grad_norm": 39.0, "learning_rate": 6.779661016949153e-07, "loss": 8.6035, "step": 10 }, { "epoch": 0.0020380867460671297, "grad_norm": 36.5, "learning_rate": 1.3559322033898307e-06, "loss": 8.3336, "step": 20 }, { "epoch": 0.003057130119100694, "grad_norm": 28.0, "learning_rate": 2.033898305084746e-06, "loss": 7.3424, "step": 30 }, { "epoch": 0.004076173492134259, "grad_norm": 9.5, "learning_rate": 2.7118644067796613e-06, "loss": 5.2891, "step": 40 }, { "epoch": 0.005095216865167823, "grad_norm": 7.625, "learning_rate": 3.3898305084745763e-06, "loss": 3.1259, "step": 50 }, { "epoch": 0.006114260238201388, "grad_norm": 1.2265625, "learning_rate": 4.067796610169492e-06, "loss": 1.9534, "step": 60 }, { "epoch": 0.007133303611234953, "grad_norm": 0.984375, "learning_rate": 4.745762711864408e-06, "loss": 1.7291, "step": 70 }, { "epoch": 0.008152346984268519, "grad_norm": 0.88671875, "learning_rate": 5.423728813559323e-06, "loss": 1.6523, "step": 80 }, { "epoch": 0.009171390357302084, "grad_norm": 0.80859375, "learning_rate": 6.1016949152542385e-06, "loss": 1.5714, "step": 90 }, { "epoch": 0.010190433730335647, "grad_norm": 0.80078125, "learning_rate": 6.779661016949153e-06, "loss": 1.5387, "step": 100 }, { "epoch": 0.011209477103369211, "grad_norm": 0.6953125, "learning_rate": 7.4576271186440685e-06, "loss": 1.5065, "step": 110 }, { "epoch": 0.012228520476402776, "grad_norm": 0.69140625, "learning_rate": 8.135593220338983e-06, "loss": 1.469, "step": 120 }, { "epoch": 0.013247563849436341, "grad_norm": 0.81640625, "learning_rate": 8.8135593220339e-06, "loss": 1.4463, "step": 130 }, { "epoch": 0.014266607222469906, "grad_norm": 0.95703125, "learning_rate": 9.491525423728815e-06, "loss": 1.3843, "step": 140 }, { "epoch": 0.01528565059550347, "grad_norm": 1.09375, "learning_rate": 1.016949152542373e-05, "loss": 1.3006, "step": 150 }, { "epoch": 0.016304693968537037, "grad_norm": 1.1171875, "learning_rate": 1.0847457627118645e-05, "loss": 1.1446, "step": 160 }, { "epoch": 0.0173237373415706, "grad_norm": 1.0859375, "learning_rate": 1.1525423728813561e-05, "loss": 0.9626, "step": 170 }, { "epoch": 0.018342780714604167, "grad_norm": 1.2109375, "learning_rate": 1.2203389830508477e-05, "loss": 0.789, "step": 180 }, { "epoch": 0.01936182408763773, "grad_norm": 1.0, "learning_rate": 1.288135593220339e-05, "loss": 0.5722, "step": 190 }, { "epoch": 0.020380867460671293, "grad_norm": 0.8359375, "learning_rate": 1.3559322033898305e-05, "loss": 0.387, "step": 200 }, { "epoch": 0.02139991083370486, "grad_norm": 0.6796875, "learning_rate": 1.4237288135593221e-05, "loss": 0.2307, "step": 210 }, { "epoch": 0.022418954206738423, "grad_norm": 0.470703125, "learning_rate": 1.4915254237288137e-05, "loss": 0.1231, "step": 220 }, { "epoch": 0.02343799757977199, "grad_norm": 0.27734375, "learning_rate": 1.5593220338983053e-05, "loss": 0.0755, "step": 230 }, { "epoch": 0.024457040952805553, "grad_norm": 0.357421875, "learning_rate": 1.6271186440677967e-05, "loss": 0.0474, "step": 240 }, { "epoch": 0.02547608432583912, "grad_norm": 0.4375, "learning_rate": 1.694915254237288e-05, "loss": 0.042, "step": 250 }, { "epoch": 0.026495127698872682, "grad_norm": 0.228515625, "learning_rate": 1.76271186440678e-05, "loss": 0.0411, "step": 260 }, { "epoch": 0.02751417107190625, "grad_norm": 0.123046875, "learning_rate": 1.8305084745762713e-05, "loss": 0.0303, "step": 270 }, { "epoch": 0.028533214444939812, "grad_norm": 0.146484375, "learning_rate": 1.898305084745763e-05, "loss": 0.0381, "step": 280 }, { "epoch": 0.02955225781797338, "grad_norm": 0.0947265625, "learning_rate": 1.9661016949152545e-05, "loss": 0.0315, "step": 290 }, { "epoch": 0.03057130119100694, "grad_norm": 0.1005859375, "learning_rate": 1.999998638184544e-05, "loss": 0.0337, "step": 300 }, { "epoch": 0.03159034456404051, "grad_norm": 0.11279296875, "learning_rate": 1.999987743683151e-05, "loss": 0.0331, "step": 310 }, { "epoch": 0.032609387937074075, "grad_norm": 0.0849609375, "learning_rate": 1.9999659547990556e-05, "loss": 0.0243, "step": 320 }, { "epoch": 0.033628431310107634, "grad_norm": 0.08447265625, "learning_rate": 1.9999332717696364e-05, "loss": 0.0315, "step": 330 }, { "epoch": 0.0346474746831412, "grad_norm": 0.07568359375, "learning_rate": 1.99988969495096e-05, "loss": 0.0263, "step": 340 }, { "epoch": 0.03566651805617477, "grad_norm": 0.1083984375, "learning_rate": 1.9998352248177743e-05, "loss": 0.0302, "step": 350 }, { "epoch": 0.036685561429208334, "grad_norm": 0.13671875, "learning_rate": 1.999769861963505e-05, "loss": 0.0298, "step": 360 }, { "epoch": 0.037704604802241894, "grad_norm": 0.078125, "learning_rate": 1.999693607100249e-05, "loss": 0.0248, "step": 370 }, { "epoch": 0.03872364817527546, "grad_norm": 0.10205078125, "learning_rate": 1.9996064610587666e-05, "loss": 0.0286, "step": 380 }, { "epoch": 0.03974269154830903, "grad_norm": 0.10546875, "learning_rate": 1.9995084247884704e-05, "loss": 0.0254, "step": 390 }, { "epoch": 0.040761734921342586, "grad_norm": 0.1474609375, "learning_rate": 1.9993994993574196e-05, "loss": 0.0267, "step": 400 }, { "epoch": 0.04178077829437615, "grad_norm": 0.0869140625, "learning_rate": 1.9992796859523033e-05, "loss": 0.0261, "step": 410 }, { "epoch": 0.04279982166740972, "grad_norm": 0.197265625, "learning_rate": 1.9991489858784306e-05, "loss": 0.0242, "step": 420 }, { "epoch": 0.043818865040443286, "grad_norm": 0.126953125, "learning_rate": 1.999007400559716e-05, "loss": 0.0268, "step": 430 }, { "epoch": 0.044837908413476846, "grad_norm": 0.1357421875, "learning_rate": 1.9988549315386624e-05, "loss": 0.0247, "step": 440 }, { "epoch": 0.04585695178651041, "grad_norm": 0.271484375, "learning_rate": 1.998691580476347e-05, "loss": 0.0257, "step": 450 }, { "epoch": 0.04687599515954398, "grad_norm": 0.11083984375, "learning_rate": 1.9985173491523996e-05, "loss": 0.0269, "step": 460 }, { "epoch": 0.047895038532577545, "grad_norm": 0.06982421875, "learning_rate": 1.9983322394649868e-05, "loss": 0.0241, "step": 470 }, { "epoch": 0.048914081905611105, "grad_norm": 0.10791015625, "learning_rate": 1.9981362534307886e-05, "loss": 0.0243, "step": 480 }, { "epoch": 0.04993312527864467, "grad_norm": 0.05419921875, "learning_rate": 1.9979293931849786e-05, "loss": 0.023, "step": 490 }, { "epoch": 0.05095216865167824, "grad_norm": 0.134765625, "learning_rate": 1.9977116609811985e-05, "loss": 0.0202, "step": 500 }, { "epoch": 0.051971212024711805, "grad_norm": 0.16015625, "learning_rate": 1.997483059191536e-05, "loss": 0.0261, "step": 510 }, { "epoch": 0.052990255397745364, "grad_norm": 0.06689453125, "learning_rate": 1.9972435903064964e-05, "loss": 0.0197, "step": 520 }, { "epoch": 0.05400929877077893, "grad_norm": 0.05615234375, "learning_rate": 1.9969932569349778e-05, "loss": 0.0235, "step": 530 }, { "epoch": 0.0550283421438125, "grad_norm": 0.095703125, "learning_rate": 1.996732061804241e-05, "loss": 0.0228, "step": 540 }, { "epoch": 0.056047385516846064, "grad_norm": 0.07373046875, "learning_rate": 1.9964600077598804e-05, "loss": 0.0214, "step": 550 }, { "epoch": 0.057066428889879624, "grad_norm": 0.07080078125, "learning_rate": 1.9961770977657937e-05, "loss": 0.0236, "step": 560 }, { "epoch": 0.05808547226291319, "grad_norm": 0.061279296875, "learning_rate": 1.9958833349041477e-05, "loss": 0.02, "step": 570 }, { "epoch": 0.05910451563594676, "grad_norm": 0.1572265625, "learning_rate": 1.9955787223753474e-05, "loss": 0.0264, "step": 580 }, { "epoch": 0.060123559008980317, "grad_norm": 0.0576171875, "learning_rate": 1.9952632634979984e-05, "loss": 0.0231, "step": 590 }, { "epoch": 0.06114260238201388, "grad_norm": 0.0673828125, "learning_rate": 1.9949369617088728e-05, "loss": 0.0206, "step": 600 }, { "epoch": 0.06216164575504745, "grad_norm": 0.0576171875, "learning_rate": 1.9945998205628706e-05, "loss": 0.0238, "step": 610 }, { "epoch": 0.06318068912808102, "grad_norm": 0.0498046875, "learning_rate": 1.994251843732982e-05, "loss": 0.0183, "step": 620 }, { "epoch": 0.06419973250111458, "grad_norm": 0.06396484375, "learning_rate": 1.9938930350102453e-05, "loss": 0.0222, "step": 630 }, { "epoch": 0.06521877587414815, "grad_norm": 0.10009765625, "learning_rate": 1.993523398303709e-05, "loss": 0.0232, "step": 640 }, { "epoch": 0.06623781924718171, "grad_norm": 0.07177734375, "learning_rate": 1.9931429376403854e-05, "loss": 0.0226, "step": 650 }, { "epoch": 0.06725686262021527, "grad_norm": 0.050537109375, "learning_rate": 1.9927516571652096e-05, "loss": 0.0229, "step": 660 }, { "epoch": 0.06827590599324884, "grad_norm": 0.045166015625, "learning_rate": 1.9923495611409933e-05, "loss": 0.0184, "step": 670 }, { "epoch": 0.0692949493662824, "grad_norm": 0.04638671875, "learning_rate": 1.9919366539483776e-05, "loss": 0.0236, "step": 680 }, { "epoch": 0.07031399273931596, "grad_norm": 0.11376953125, "learning_rate": 1.9915129400857872e-05, "loss": 0.0214, "step": 690 }, { "epoch": 0.07133303611234953, "grad_norm": 0.12060546875, "learning_rate": 1.9910784241693796e-05, "loss": 0.0188, "step": 700 }, { "epoch": 0.0723520794853831, "grad_norm": 0.09423828125, "learning_rate": 1.9906331109329952e-05, "loss": 0.02, "step": 710 }, { "epoch": 0.07337112285841667, "grad_norm": 0.043701171875, "learning_rate": 1.9901770052281067e-05, "loss": 0.0201, "step": 720 }, { "epoch": 0.07439016623145023, "grad_norm": 0.068359375, "learning_rate": 1.9897101120237644e-05, "loss": 0.02, "step": 730 }, { "epoch": 0.07540920960448379, "grad_norm": 0.1025390625, "learning_rate": 1.9892324364065445e-05, "loss": 0.02, "step": 740 }, { "epoch": 0.07642825297751736, "grad_norm": 0.0859375, "learning_rate": 1.9887439835804918e-05, "loss": 0.0185, "step": 750 }, { "epoch": 0.07744729635055092, "grad_norm": 0.064453125, "learning_rate": 1.988244758867063e-05, "loss": 0.0215, "step": 760 }, { "epoch": 0.07846633972358448, "grad_norm": 0.08251953125, "learning_rate": 1.9877347677050706e-05, "loss": 0.018, "step": 770 }, { "epoch": 0.07948538309661805, "grad_norm": 0.0546875, "learning_rate": 1.987214015650621e-05, "loss": 0.019, "step": 780 }, { "epoch": 0.08050442646965161, "grad_norm": 0.0498046875, "learning_rate": 1.9866825083770558e-05, "loss": 0.023, "step": 790 }, { "epoch": 0.08152346984268517, "grad_norm": 0.058837890625, "learning_rate": 1.9861402516748897e-05, "loss": 0.0176, "step": 800 }, { "epoch": 0.08254251321571875, "grad_norm": 0.07861328125, "learning_rate": 1.985587251451748e-05, "loss": 0.0225, "step": 810 }, { "epoch": 0.0835615565887523, "grad_norm": 0.05712890625, "learning_rate": 1.9850235137322995e-05, "loss": 0.018, "step": 820 }, { "epoch": 0.08458059996178588, "grad_norm": 0.06396484375, "learning_rate": 1.9844490446581945e-05, "loss": 0.0217, "step": 830 }, { "epoch": 0.08559964333481944, "grad_norm": 0.1083984375, "learning_rate": 1.9838638504879955e-05, "loss": 0.0187, "step": 840 }, { "epoch": 0.086618686707853, "grad_norm": 0.0859375, "learning_rate": 1.98326793759711e-05, "loss": 0.0188, "step": 850 }, { "epoch": 0.08763773008088657, "grad_norm": 0.07080078125, "learning_rate": 1.9826613124777206e-05, "loss": 0.0203, "step": 860 }, { "epoch": 0.08865677345392013, "grad_norm": 0.12353515625, "learning_rate": 1.9820439817387144e-05, "loss": 0.0192, "step": 870 }, { "epoch": 0.08967581682695369, "grad_norm": 0.05908203125, "learning_rate": 1.981415952105611e-05, "loss": 0.0184, "step": 880 }, { "epoch": 0.09069486019998727, "grad_norm": 0.142578125, "learning_rate": 1.98077723042049e-05, "loss": 0.0203, "step": 890 }, { "epoch": 0.09171390357302082, "grad_norm": 0.2255859375, "learning_rate": 1.9801278236419147e-05, "loss": 0.0193, "step": 900 }, { "epoch": 0.0927329469460544, "grad_norm": 0.068359375, "learning_rate": 1.9794677388448584e-05, "loss": 0.0192, "step": 910 }, { "epoch": 0.09375199031908796, "grad_norm": 0.1123046875, "learning_rate": 1.9787969832206245e-05, "loss": 0.0177, "step": 920 }, { "epoch": 0.09477103369212152, "grad_norm": 0.09716796875, "learning_rate": 1.978115564076772e-05, "loss": 0.0205, "step": 930 }, { "epoch": 0.09579007706515509, "grad_norm": 0.103515625, "learning_rate": 1.9774234888370324e-05, "loss": 0.0185, "step": 940 }, { "epoch": 0.09680912043818865, "grad_norm": 0.267578125, "learning_rate": 1.976720765041231e-05, "loss": 0.0167, "step": 950 }, { "epoch": 0.09782816381122221, "grad_norm": 0.059814453125, "learning_rate": 1.9760074003452028e-05, "loss": 0.0197, "step": 960 }, { "epoch": 0.09884720718425578, "grad_norm": 0.0537109375, "learning_rate": 1.9752834025207116e-05, "loss": 0.0166, "step": 970 }, { "epoch": 0.09986625055728934, "grad_norm": 0.068359375, "learning_rate": 1.974548779455364e-05, "loss": 0.021, "step": 980 }, { "epoch": 0.1008852939303229, "grad_norm": 0.053955078125, "learning_rate": 1.973803539152522e-05, "loss": 0.0186, "step": 990 }, { "epoch": 0.10190433730335648, "grad_norm": 0.10546875, "learning_rate": 1.9730476897312187e-05, "loss": 0.0163, "step": 1000 }, { "epoch": 0.10292338067639004, "grad_norm": 0.11572265625, "learning_rate": 1.9722812394260674e-05, "loss": 0.0208, "step": 1010 }, { "epoch": 0.10394242404942361, "grad_norm": 0.07373046875, "learning_rate": 1.9715041965871743e-05, "loss": 0.0179, "step": 1020 }, { "epoch": 0.10496146742245717, "grad_norm": 0.06103515625, "learning_rate": 1.970716569680044e-05, "loss": 0.0182, "step": 1030 }, { "epoch": 0.10598051079549073, "grad_norm": 0.05126953125, "learning_rate": 1.9699183672854916e-05, "loss": 0.0177, "step": 1040 }, { "epoch": 0.1069995541685243, "grad_norm": 0.08642578125, "learning_rate": 1.9691095980995455e-05, "loss": 0.0175, "step": 1050 }, { "epoch": 0.10801859754155786, "grad_norm": 0.0859375, "learning_rate": 1.9682902709333552e-05, "loss": 0.0198, "step": 1060 }, { "epoch": 0.10903764091459142, "grad_norm": 0.09912109375, "learning_rate": 1.967460394713093e-05, "loss": 0.0169, "step": 1070 }, { "epoch": 0.110056684287625, "grad_norm": 0.068359375, "learning_rate": 1.96661997847986e-05, "loss": 0.0179, "step": 1080 }, { "epoch": 0.11107572766065855, "grad_norm": 0.125, "learning_rate": 1.9657690313895834e-05, "loss": 0.0188, "step": 1090 }, { "epoch": 0.11209477103369213, "grad_norm": 0.09033203125, "learning_rate": 1.9649075627129208e-05, "loss": 0.0159, "step": 1100 }, { "epoch": 0.11311381440672569, "grad_norm": 0.045654296875, "learning_rate": 1.964035581835156e-05, "loss": 0.0172, "step": 1110 }, { "epoch": 0.11413285777975925, "grad_norm": 0.04541015625, "learning_rate": 1.9631530982560996e-05, "loss": 0.0183, "step": 1120 }, { "epoch": 0.11515190115279282, "grad_norm": 0.05859375, "learning_rate": 1.962260121589983e-05, "loss": 0.0193, "step": 1130 }, { "epoch": 0.11617094452582638, "grad_norm": 0.056884765625, "learning_rate": 1.9613566615653546e-05, "loss": 0.0193, "step": 1140 }, { "epoch": 0.11718998789885994, "grad_norm": 0.103515625, "learning_rate": 1.960442728024975e-05, "loss": 0.0163, "step": 1150 }, { "epoch": 0.11820903127189351, "grad_norm": 0.048828125, "learning_rate": 1.9595183309257074e-05, "loss": 0.0186, "step": 1160 }, { "epoch": 0.11922807464492707, "grad_norm": 0.064453125, "learning_rate": 1.958583480338411e-05, "loss": 0.0181, "step": 1170 }, { "epoch": 0.12024711801796063, "grad_norm": 0.0400390625, "learning_rate": 1.957638186447831e-05, "loss": 0.0187, "step": 1180 }, { "epoch": 0.1212661613909942, "grad_norm": 0.055908203125, "learning_rate": 1.956682459552487e-05, "loss": 0.0195, "step": 1190 }, { "epoch": 0.12228520476402777, "grad_norm": 0.048828125, "learning_rate": 1.9557163100645612e-05, "loss": 0.016, "step": 1200 }, { "epoch": 0.12330424813706134, "grad_norm": 0.1044921875, "learning_rate": 1.9547397485097846e-05, "loss": 0.0194, "step": 1210 }, { "epoch": 0.1243232915100949, "grad_norm": 0.05322265625, "learning_rate": 1.953752785527323e-05, "loss": 0.0162, "step": 1220 }, { "epoch": 0.12534233488312846, "grad_norm": 0.06982421875, "learning_rate": 1.9527554318696607e-05, "loss": 0.0193, "step": 1230 }, { "epoch": 0.12636137825616203, "grad_norm": 0.045654296875, "learning_rate": 1.9517476984024838e-05, "loss": 0.0171, "step": 1240 }, { "epoch": 0.1273804216291956, "grad_norm": 0.08349609375, "learning_rate": 1.95072959610456e-05, "loss": 0.0158, "step": 1250 }, { "epoch": 0.12839946500222915, "grad_norm": 0.042724609375, "learning_rate": 1.949701136067622e-05, "loss": 0.0211, "step": 1260 }, { "epoch": 0.12941850837526273, "grad_norm": 0.06689453125, "learning_rate": 1.948662329496244e-05, "loss": 0.0168, "step": 1270 }, { "epoch": 0.1304375517482963, "grad_norm": 0.06591796875, "learning_rate": 1.9476131877077205e-05, "loss": 0.0175, "step": 1280 }, { "epoch": 0.13145659512132984, "grad_norm": 0.05078125, "learning_rate": 1.9465537221319453e-05, "loss": 0.0182, "step": 1290 }, { "epoch": 0.13247563849436342, "grad_norm": 0.11474609375, "learning_rate": 1.945483944311282e-05, "loss": 0.0153, "step": 1300 }, { "epoch": 0.133494681867397, "grad_norm": 0.04150390625, "learning_rate": 1.944403865900443e-05, "loss": 0.0211, "step": 1310 }, { "epoch": 0.13451372524043054, "grad_norm": 0.0732421875, "learning_rate": 1.9433134986663596e-05, "loss": 0.0153, "step": 1320 }, { "epoch": 0.1355327686134641, "grad_norm": 0.0703125, "learning_rate": 1.9422128544880562e-05, "loss": 0.0177, "step": 1330 }, { "epoch": 0.13655181198649768, "grad_norm": 0.0517578125, "learning_rate": 1.9411019453565174e-05, "loss": 0.0187, "step": 1340 }, { "epoch": 0.13757085535953123, "grad_norm": 0.07861328125, "learning_rate": 1.9399807833745622e-05, "loss": 0.0151, "step": 1350 }, { "epoch": 0.1385898987325648, "grad_norm": 0.05615234375, "learning_rate": 1.9388493807567067e-05, "loss": 0.0183, "step": 1360 }, { "epoch": 0.13960894210559838, "grad_norm": 0.087890625, "learning_rate": 1.937707749829036e-05, "loss": 0.0171, "step": 1370 }, { "epoch": 0.14062798547863192, "grad_norm": 0.0546875, "learning_rate": 1.936555903029066e-05, "loss": 0.0158, "step": 1380 }, { "epoch": 0.1416470288516655, "grad_norm": 0.057861328125, "learning_rate": 1.935393852905611e-05, "loss": 0.0183, "step": 1390 }, { "epoch": 0.14266607222469907, "grad_norm": 0.11376953125, "learning_rate": 1.9342216121186447e-05, "loss": 0.0158, "step": 1400 }, { "epoch": 0.14368511559773262, "grad_norm": 0.043212890625, "learning_rate": 1.9330391934391635e-05, "loss": 0.02, "step": 1410 }, { "epoch": 0.1447041589707662, "grad_norm": 0.06201171875, "learning_rate": 1.931846609749047e-05, "loss": 0.0177, "step": 1420 }, { "epoch": 0.14572320234379976, "grad_norm": 0.07421875, "learning_rate": 1.9306438740409168e-05, "loss": 0.0184, "step": 1430 }, { "epoch": 0.14674224571683334, "grad_norm": 0.048095703125, "learning_rate": 1.9294309994179972e-05, "loss": 0.0172, "step": 1440 }, { "epoch": 0.14776128908986688, "grad_norm": 0.05322265625, "learning_rate": 1.9282079990939707e-05, "loss": 0.0159, "step": 1450 }, { "epoch": 0.14878033246290046, "grad_norm": 0.08984375, "learning_rate": 1.9269748863928338e-05, "loss": 0.0176, "step": 1460 }, { "epoch": 0.14979937583593403, "grad_norm": 0.052978515625, "learning_rate": 1.9257316747487528e-05, "loss": 0.0169, "step": 1470 }, { "epoch": 0.15081841920896757, "grad_norm": 0.05859375, "learning_rate": 1.9244783777059175e-05, "loss": 0.016, "step": 1480 }, { "epoch": 0.15183746258200115, "grad_norm": 0.07763671875, "learning_rate": 1.9232150089183925e-05, "loss": 0.018, "step": 1490 }, { "epoch": 0.15285650595503472, "grad_norm": 0.051513671875, "learning_rate": 1.9219415821499697e-05, "loss": 0.0169, "step": 1500 }, { "epoch": 0.15387554932806827, "grad_norm": 0.060546875, "learning_rate": 1.9206581112740175e-05, "loss": 0.0199, "step": 1510 }, { "epoch": 0.15489459270110184, "grad_norm": 0.052734375, "learning_rate": 1.9193646102733307e-05, "loss": 0.0166, "step": 1520 }, { "epoch": 0.15591363607413541, "grad_norm": 0.06689453125, "learning_rate": 1.918061093239976e-05, "loss": 0.0192, "step": 1530 }, { "epoch": 0.15693267944716896, "grad_norm": 0.0556640625, "learning_rate": 1.916747574375142e-05, "loss": 0.0187, "step": 1540 }, { "epoch": 0.15795172282020253, "grad_norm": 0.04833984375, "learning_rate": 1.9154240679889806e-05, "loss": 0.0161, "step": 1550 }, { "epoch": 0.1589707661932361, "grad_norm": 0.033447265625, "learning_rate": 1.914090588500454e-05, "loss": 0.0182, "step": 1560 }, { "epoch": 0.15998980956626965, "grad_norm": 0.049560546875, "learning_rate": 1.912747150437176e-05, "loss": 0.0167, "step": 1570 }, { "epoch": 0.16100885293930323, "grad_norm": 0.1748046875, "learning_rate": 1.9113937684352536e-05, "loss": 0.0194, "step": 1580 }, { "epoch": 0.1620278963123368, "grad_norm": 0.04345703125, "learning_rate": 1.91003045723913e-05, "loss": 0.021, "step": 1590 }, { "epoch": 0.16304693968537035, "grad_norm": 0.05029296875, "learning_rate": 1.908657231701421e-05, "loss": 0.0173, "step": 1600 }, { "epoch": 0.16406598305840392, "grad_norm": 0.09619140625, "learning_rate": 1.9072741067827542e-05, "loss": 0.0208, "step": 1610 }, { "epoch": 0.1650850264314375, "grad_norm": 0.061279296875, "learning_rate": 1.9058810975516067e-05, "loss": 0.0197, "step": 1620 }, { "epoch": 0.16610406980447107, "grad_norm": 0.045654296875, "learning_rate": 1.90447821918414e-05, "loss": 0.0177, "step": 1630 }, { "epoch": 0.1671231131775046, "grad_norm": 0.0615234375, "learning_rate": 1.9030654869640356e-05, "loss": 0.0185, "step": 1640 }, { "epoch": 0.16814215655053819, "grad_norm": 0.125, "learning_rate": 1.901642916282327e-05, "loss": 0.0141, "step": 1650 }, { "epoch": 0.16916119992357176, "grad_norm": 0.091796875, "learning_rate": 1.900210522637234e-05, "loss": 0.0212, "step": 1660 }, { "epoch": 0.1701802432966053, "grad_norm": 0.04833984375, "learning_rate": 1.8987683216339924e-05, "loss": 0.0173, "step": 1670 }, { "epoch": 0.17119928666963888, "grad_norm": 0.060546875, "learning_rate": 1.897316328984684e-05, "loss": 0.019, "step": 1680 }, { "epoch": 0.17221833004267245, "grad_norm": 0.033935546875, "learning_rate": 1.895854560508067e-05, "loss": 0.0201, "step": 1690 }, { "epoch": 0.173237373415706, "grad_norm": 0.049560546875, "learning_rate": 1.8943830321294015e-05, "loss": 0.0149, "step": 1700 }, { "epoch": 0.17425641678873957, "grad_norm": 0.043212890625, "learning_rate": 1.892901759880277e-05, "loss": 0.0194, "step": 1710 }, { "epoch": 0.17527546016177314, "grad_norm": 0.0556640625, "learning_rate": 1.8914107598984385e-05, "loss": 0.0156, "step": 1720 }, { "epoch": 0.1762945035348067, "grad_norm": 0.083984375, "learning_rate": 1.889910048427609e-05, "loss": 0.0173, "step": 1730 }, { "epoch": 0.17731354690784026, "grad_norm": 0.044921875, "learning_rate": 1.8883996418173142e-05, "loss": 0.019, "step": 1740 }, { "epoch": 0.17833259028087384, "grad_norm": 0.08251953125, "learning_rate": 1.886879556522704e-05, "loss": 0.0167, "step": 1750 }, { "epoch": 0.17935163365390738, "grad_norm": 0.043701171875, "learning_rate": 1.8853498091043715e-05, "loss": 0.0203, "step": 1760 }, { "epoch": 0.18037067702694096, "grad_norm": 0.06396484375, "learning_rate": 1.883810416228175e-05, "loss": 0.0185, "step": 1770 }, { "epoch": 0.18138972039997453, "grad_norm": 0.06982421875, "learning_rate": 1.8822613946650555e-05, "loss": 0.0178, "step": 1780 }, { "epoch": 0.18240876377300808, "grad_norm": 0.080078125, "learning_rate": 1.880702761290853e-05, "loss": 0.0162, "step": 1790 }, { "epoch": 0.18342780714604165, "grad_norm": 0.046875, "learning_rate": 1.8791345330861244e-05, "loss": 0.0172, "step": 1800 }, { "epoch": 0.18444685051907522, "grad_norm": 0.0361328125, "learning_rate": 1.8775567271359575e-05, "loss": 0.0192, "step": 1810 }, { "epoch": 0.1854658938921088, "grad_norm": 0.04345703125, "learning_rate": 1.875969360629785e-05, "loss": 0.0168, "step": 1820 }, { "epoch": 0.18648493726514234, "grad_norm": 0.044677734375, "learning_rate": 1.8743724508611962e-05, "loss": 0.0155, "step": 1830 }, { "epoch": 0.18750398063817592, "grad_norm": 0.040771484375, "learning_rate": 1.872766015227751e-05, "loss": 0.0174, "step": 1840 }, { "epoch": 0.1885230240112095, "grad_norm": 0.050537109375, "learning_rate": 1.871150071230789e-05, "loss": 0.0162, "step": 1850 }, { "epoch": 0.18954206738424303, "grad_norm": 0.036865234375, "learning_rate": 1.8695246364752376e-05, "loss": 0.019, "step": 1860 }, { "epoch": 0.1905611107572766, "grad_norm": 0.05029296875, "learning_rate": 1.8678897286694222e-05, "loss": 0.016, "step": 1870 }, { "epoch": 0.19158015413031018, "grad_norm": 0.0673828125, "learning_rate": 1.8662453656248725e-05, "loss": 0.0161, "step": 1880 }, { "epoch": 0.19259919750334373, "grad_norm": 0.044921875, "learning_rate": 1.8645915652561286e-05, "loss": 0.0166, "step": 1890 }, { "epoch": 0.1936182408763773, "grad_norm": 0.09814453125, "learning_rate": 1.862928345580545e-05, "loss": 0.0159, "step": 1900 }, { "epoch": 0.19463728424941087, "grad_norm": 0.0908203125, "learning_rate": 1.861255724718096e-05, "loss": 0.019, "step": 1910 }, { "epoch": 0.19565632762244442, "grad_norm": 0.040771484375, "learning_rate": 1.8595737208911762e-05, "loss": 0.0168, "step": 1920 }, { "epoch": 0.196675370995478, "grad_norm": 0.05029296875, "learning_rate": 1.857882352424404e-05, "loss": 0.0169, "step": 1930 }, { "epoch": 0.19769441436851157, "grad_norm": 0.068359375, "learning_rate": 1.8561816377444206e-05, "loss": 0.0197, "step": 1940 }, { "epoch": 0.1987134577415451, "grad_norm": 0.03662109375, "learning_rate": 1.8544715953796893e-05, "loss": 0.0162, "step": 1950 }, { "epoch": 0.1997325011145787, "grad_norm": 0.03466796875, "learning_rate": 1.8527522439602946e-05, "loss": 0.0217, "step": 1960 }, { "epoch": 0.20075154448761226, "grad_norm": 0.03662109375, "learning_rate": 1.851023602217739e-05, "loss": 0.0164, "step": 1970 }, { "epoch": 0.2017705878606458, "grad_norm": 0.04638671875, "learning_rate": 1.849285688984737e-05, "loss": 0.0176, "step": 1980 }, { "epoch": 0.20278963123367938, "grad_norm": 0.053955078125, "learning_rate": 1.847538523195013e-05, "loss": 0.0201, "step": 1990 }, { "epoch": 0.20380867460671295, "grad_norm": 0.0400390625, "learning_rate": 1.8457821238830933e-05, "loss": 0.0155, "step": 2000 }, { "epoch": 0.20482771797974653, "grad_norm": 0.032958984375, "learning_rate": 1.8440165101840985e-05, "loss": 0.0177, "step": 2010 }, { "epoch": 0.20584676135278007, "grad_norm": 0.06787109375, "learning_rate": 1.8422417013335353e-05, "loss": 0.0159, "step": 2020 }, { "epoch": 0.20686580472581365, "grad_norm": 0.048828125, "learning_rate": 1.8404577166670883e-05, "loss": 0.0165, "step": 2030 }, { "epoch": 0.20788484809884722, "grad_norm": 0.043701171875, "learning_rate": 1.838664575620407e-05, "loss": 0.0166, "step": 2040 }, { "epoch": 0.20890389147188076, "grad_norm": 0.07568359375, "learning_rate": 1.8368622977288955e-05, "loss": 0.0151, "step": 2050 }, { "epoch": 0.20992293484491434, "grad_norm": 0.06298828125, "learning_rate": 1.8350509026274997e-05, "loss": 0.0168, "step": 2060 }, { "epoch": 0.2109419782179479, "grad_norm": 0.03564453125, "learning_rate": 1.833230410050493e-05, "loss": 0.0157, "step": 2070 }, { "epoch": 0.21196102159098146, "grad_norm": 0.11572265625, "learning_rate": 1.831400839831261e-05, "loss": 0.0141, "step": 2080 }, { "epoch": 0.21298006496401503, "grad_norm": 0.09326171875, "learning_rate": 1.8295622119020867e-05, "loss": 0.0179, "step": 2090 }, { "epoch": 0.2139991083370486, "grad_norm": 0.10400390625, "learning_rate": 1.8277145462939313e-05, "loss": 0.0152, "step": 2100 }, { "epoch": 0.21501815171008215, "grad_norm": 0.08154296875, "learning_rate": 1.8258578631362182e-05, "loss": 0.0187, "step": 2110 }, { "epoch": 0.21603719508311572, "grad_norm": 0.04345703125, "learning_rate": 1.823992182656612e-05, "loss": 0.017, "step": 2120 }, { "epoch": 0.2170562384561493, "grad_norm": 0.05908203125, "learning_rate": 1.822117525180799e-05, "loss": 0.016, "step": 2130 }, { "epoch": 0.21807528182918284, "grad_norm": 0.05712890625, "learning_rate": 1.8202339111322657e-05, "loss": 0.017, "step": 2140 }, { "epoch": 0.21909432520221642, "grad_norm": 0.0654296875, "learning_rate": 1.8183413610320756e-05, "loss": 0.0151, "step": 2150 }, { "epoch": 0.22011336857525, "grad_norm": 0.03857421875, "learning_rate": 1.8164398954986463e-05, "loss": 0.0178, "step": 2160 }, { "epoch": 0.22113241194828354, "grad_norm": 0.06396484375, "learning_rate": 1.8145295352475258e-05, "loss": 0.0183, "step": 2170 }, { "epoch": 0.2221514553213171, "grad_norm": 0.06201171875, "learning_rate": 1.812610301091164e-05, "loss": 0.018, "step": 2180 }, { "epoch": 0.22317049869435068, "grad_norm": 0.045654296875, "learning_rate": 1.8106822139386886e-05, "loss": 0.0188, "step": 2190 }, { "epoch": 0.22418954206738426, "grad_norm": 0.0849609375, "learning_rate": 1.808745294795677e-05, "loss": 0.0152, "step": 2200 }, { "epoch": 0.2252085854404178, "grad_norm": 0.057373046875, "learning_rate": 1.8067995647639257e-05, "loss": 0.0186, "step": 2210 }, { "epoch": 0.22622762881345138, "grad_norm": 0.050537109375, "learning_rate": 1.804845045041222e-05, "loss": 0.0185, "step": 2220 }, { "epoch": 0.22724667218648495, "grad_norm": 0.0537109375, "learning_rate": 1.8028817569211137e-05, "loss": 0.0161, "step": 2230 }, { "epoch": 0.2282657155595185, "grad_norm": 0.064453125, "learning_rate": 1.8009097217926736e-05, "loss": 0.0178, "step": 2240 }, { "epoch": 0.22928475893255207, "grad_norm": 0.055908203125, "learning_rate": 1.7989289611402717e-05, "loss": 0.0154, "step": 2250 }, { "epoch": 0.23030380230558564, "grad_norm": 0.07080078125, "learning_rate": 1.7969394965433365e-05, "loss": 0.0181, "step": 2260 }, { "epoch": 0.2313228456786192, "grad_norm": 0.037841796875, "learning_rate": 1.7949413496761222e-05, "loss": 0.0178, "step": 2270 }, { "epoch": 0.23234188905165276, "grad_norm": 0.04541015625, "learning_rate": 1.792934542307473e-05, "loss": 0.017, "step": 2280 }, { "epoch": 0.23336093242468633, "grad_norm": 0.06201171875, "learning_rate": 1.7909190963005833e-05, "loss": 0.0197, "step": 2290 }, { "epoch": 0.23437997579771988, "grad_norm": 0.09228515625, "learning_rate": 1.788895033612763e-05, "loss": 0.0139, "step": 2300 }, { "epoch": 0.23539901917075345, "grad_norm": 0.042724609375, "learning_rate": 1.7868623762951957e-05, "loss": 0.0172, "step": 2310 }, { "epoch": 0.23641806254378703, "grad_norm": 0.03759765625, "learning_rate": 1.7848211464927e-05, "loss": 0.0162, "step": 2320 }, { "epoch": 0.23743710591682057, "grad_norm": 0.036865234375, "learning_rate": 1.7827713664434864e-05, "loss": 0.017, "step": 2330 }, { "epoch": 0.23845614928985415, "grad_norm": 0.040771484375, "learning_rate": 1.7807130584789175e-05, "loss": 0.0161, "step": 2340 }, { "epoch": 0.23947519266288772, "grad_norm": 0.06103515625, "learning_rate": 1.7786462450232624e-05, "loss": 0.0135, "step": 2350 }, { "epoch": 0.24049423603592127, "grad_norm": 0.043212890625, "learning_rate": 1.7765709485934538e-05, "loss": 0.0187, "step": 2360 }, { "epoch": 0.24151327940895484, "grad_norm": 0.048828125, "learning_rate": 1.7744871917988424e-05, "loss": 0.0161, "step": 2370 }, { "epoch": 0.2425323227819884, "grad_norm": 0.0693359375, "learning_rate": 1.772394997340951e-05, "loss": 0.0165, "step": 2380 }, { "epoch": 0.243551366155022, "grad_norm": 0.051513671875, "learning_rate": 1.7702943880132254e-05, "loss": 0.0171, "step": 2390 }, { "epoch": 0.24457040952805553, "grad_norm": 0.052978515625, "learning_rate": 1.7681853867007884e-05, "loss": 0.0164, "step": 2400 }, { "epoch": 0.2455894529010891, "grad_norm": 0.045654296875, "learning_rate": 1.7660680163801885e-05, "loss": 0.0171, "step": 2410 }, { "epoch": 0.24660849627412268, "grad_norm": 0.05517578125, "learning_rate": 1.7639423001191517e-05, "loss": 0.0182, "step": 2420 }, { "epoch": 0.24762753964715623, "grad_norm": 0.07666015625, "learning_rate": 1.7618082610763275e-05, "loss": 0.0159, "step": 2430 }, { "epoch": 0.2486465830201898, "grad_norm": 0.04345703125, "learning_rate": 1.75966592250104e-05, "loss": 0.0169, "step": 2440 }, { "epoch": 0.24966562639322337, "grad_norm": 0.047119140625, "learning_rate": 1.7575153077330305e-05, "loss": 0.013, "step": 2450 }, { "epoch": 0.2506846697662569, "grad_norm": 0.06591796875, "learning_rate": 1.7553564402022073e-05, "loss": 0.0181, "step": 2460 }, { "epoch": 0.25170371313929046, "grad_norm": 0.046142578125, "learning_rate": 1.753189343428387e-05, "loss": 0.0186, "step": 2470 }, { "epoch": 0.25272275651232406, "grad_norm": 0.06005859375, "learning_rate": 1.7510140410210416e-05, "loss": 0.0157, "step": 2480 }, { "epoch": 0.2537417998853576, "grad_norm": 0.05029296875, "learning_rate": 1.7488305566790372e-05, "loss": 0.0162, "step": 2490 }, { "epoch": 0.2547608432583912, "grad_norm": 0.042724609375, "learning_rate": 1.7466389141903805e-05, "loss": 0.013, "step": 2500 }, { "epoch": 0.25577988663142476, "grad_norm": 0.056640625, "learning_rate": 1.744439137431955e-05, "loss": 0.0162, "step": 2510 }, { "epoch": 0.2567989300044583, "grad_norm": 0.03173828125, "learning_rate": 1.742231250369266e-05, "loss": 0.0157, "step": 2520 }, { "epoch": 0.2578179733774919, "grad_norm": 0.06689453125, "learning_rate": 1.740015277056173e-05, "loss": 0.0154, "step": 2530 }, { "epoch": 0.25883701675052545, "grad_norm": 0.047607421875, "learning_rate": 1.7377912416346348e-05, "loss": 0.0181, "step": 2540 }, { "epoch": 0.259856060123559, "grad_norm": 0.06298828125, "learning_rate": 1.7355591683344405e-05, "loss": 0.0126, "step": 2550 }, { "epoch": 0.2608751034965926, "grad_norm": 0.08935546875, "learning_rate": 1.7333190814729494e-05, "loss": 0.0162, "step": 2560 }, { "epoch": 0.26189414686962614, "grad_norm": 0.06298828125, "learning_rate": 1.7310710054548235e-05, "loss": 0.0178, "step": 2570 }, { "epoch": 0.2629131902426597, "grad_norm": 0.0546875, "learning_rate": 1.7288149647717637e-05, "loss": 0.0166, "step": 2580 }, { "epoch": 0.2639322336156933, "grad_norm": 0.05078125, "learning_rate": 1.7265509840022425e-05, "loss": 0.0163, "step": 2590 }, { "epoch": 0.26495127698872684, "grad_norm": 0.07861328125, "learning_rate": 1.7242790878112344e-05, "loss": 0.0139, "step": 2600 }, { "epoch": 0.2659703203617604, "grad_norm": 0.05078125, "learning_rate": 1.72199930094995e-05, "loss": 0.0182, "step": 2610 }, { "epoch": 0.266989363734794, "grad_norm": 0.03857421875, "learning_rate": 1.7197116482555634e-05, "loss": 0.0155, "step": 2620 }, { "epoch": 0.26800840710782753, "grad_norm": 0.038330078125, "learning_rate": 1.7174161546509448e-05, "loss": 0.0174, "step": 2630 }, { "epoch": 0.2690274504808611, "grad_norm": 0.04150390625, "learning_rate": 1.715112845144386e-05, "loss": 0.0195, "step": 2640 }, { "epoch": 0.2700464938538947, "grad_norm": 0.0595703125, "learning_rate": 1.7128017448293298e-05, "loss": 0.0149, "step": 2650 }, { "epoch": 0.2710655372269282, "grad_norm": 0.0732421875, "learning_rate": 1.7104828788840966e-05, "loss": 0.0173, "step": 2660 }, { "epoch": 0.27208458059996177, "grad_norm": 0.055419921875, "learning_rate": 1.7081562725716086e-05, "loss": 0.0167, "step": 2670 }, { "epoch": 0.27310362397299537, "grad_norm": 0.04052734375, "learning_rate": 1.7058219512391162e-05, "loss": 0.0147, "step": 2680 }, { "epoch": 0.2741226673460289, "grad_norm": 0.060546875, "learning_rate": 1.7034799403179208e-05, "loss": 0.0164, "step": 2690 }, { "epoch": 0.27514171071906246, "grad_norm": 0.035400390625, "learning_rate": 1.701130265323099e-05, "loss": 0.0148, "step": 2700 }, { "epoch": 0.27616075409209606, "grad_norm": 0.051513671875, "learning_rate": 1.6987729518532224e-05, "loss": 0.0185, "step": 2710 }, { "epoch": 0.2771797974651296, "grad_norm": 0.0771484375, "learning_rate": 1.696408025590081e-05, "loss": 0.0153, "step": 2720 }, { "epoch": 0.27819884083816315, "grad_norm": 0.034423828125, "learning_rate": 1.6940355122984032e-05, "loss": 0.0163, "step": 2730 }, { "epoch": 0.27921788421119675, "grad_norm": 0.052978515625, "learning_rate": 1.6916554378255726e-05, "loss": 0.0167, "step": 2740 }, { "epoch": 0.2802369275842303, "grad_norm": 0.051025390625, "learning_rate": 1.6892678281013495e-05, "loss": 0.0147, "step": 2750 }, { "epoch": 0.28125597095726385, "grad_norm": 0.07763671875, "learning_rate": 1.686872709137587e-05, "loss": 0.0176, "step": 2760 }, { "epoch": 0.28227501433029745, "grad_norm": 0.044677734375, "learning_rate": 1.684470107027947e-05, "loss": 0.0149, "step": 2770 }, { "epoch": 0.283294057703331, "grad_norm": 0.0439453125, "learning_rate": 1.6820600479476176e-05, "loss": 0.0148, "step": 2780 }, { "epoch": 0.28431310107636454, "grad_norm": 0.042236328125, "learning_rate": 1.6796425581530266e-05, "loss": 0.0164, "step": 2790 }, { "epoch": 0.28533214444939814, "grad_norm": 0.03759765625, "learning_rate": 1.6772176639815555e-05, "loss": 0.0149, "step": 2800 }, { "epoch": 0.2863511878224317, "grad_norm": 0.0849609375, "learning_rate": 1.6747853918512534e-05, "loss": 0.0184, "step": 2810 }, { "epoch": 0.28737023119546523, "grad_norm": 0.03515625, "learning_rate": 1.6723457682605488e-05, "loss": 0.0148, "step": 2820 }, { "epoch": 0.28838927456849883, "grad_norm": 0.07080078125, "learning_rate": 1.6698988197879596e-05, "loss": 0.0143, "step": 2830 }, { "epoch": 0.2894083179415324, "grad_norm": 0.034423828125, "learning_rate": 1.6674445730918066e-05, "loss": 0.0171, "step": 2840 }, { "epoch": 0.2904273613145659, "grad_norm": 0.03759765625, "learning_rate": 1.6649830549099198e-05, "loss": 0.0131, "step": 2850 }, { "epoch": 0.2914464046875995, "grad_norm": 0.0419921875, "learning_rate": 1.662514292059349e-05, "loss": 0.0157, "step": 2860 }, { "epoch": 0.29246544806063307, "grad_norm": 0.0634765625, "learning_rate": 1.660038311436071e-05, "loss": 0.0151, "step": 2870 }, { "epoch": 0.29348449143366667, "grad_norm": 0.06591796875, "learning_rate": 1.6575551400146974e-05, "loss": 0.0138, "step": 2880 }, { "epoch": 0.2945035348067002, "grad_norm": 0.02880859375, "learning_rate": 1.6550648048481793e-05, "loss": 0.0156, "step": 2890 }, { "epoch": 0.29552257817973376, "grad_norm": 0.049560546875, "learning_rate": 1.6525673330675133e-05, "loss": 0.0149, "step": 2900 }, { "epoch": 0.29654162155276736, "grad_norm": 0.052490234375, "learning_rate": 1.6500627518814466e-05, "loss": 0.0163, "step": 2910 }, { "epoch": 0.2975606649258009, "grad_norm": 0.04052734375, "learning_rate": 1.6475510885761793e-05, "loss": 0.0167, "step": 2920 }, { "epoch": 0.29857970829883446, "grad_norm": 0.078125, "learning_rate": 1.6450323705150684e-05, "loss": 0.0163, "step": 2930 }, { "epoch": 0.29959875167186806, "grad_norm": 0.034423828125, "learning_rate": 1.6425066251383285e-05, "loss": 0.0163, "step": 2940 }, { "epoch": 0.3006177950449016, "grad_norm": 0.06494140625, "learning_rate": 1.6399738799627334e-05, "loss": 0.0136, "step": 2950 }, { "epoch": 0.30163683841793515, "grad_norm": 0.076171875, "learning_rate": 1.6374341625813165e-05, "loss": 0.018, "step": 2960 }, { "epoch": 0.30265588179096875, "grad_norm": 0.11767578125, "learning_rate": 1.6348875006630706e-05, "loss": 0.0169, "step": 2970 }, { "epoch": 0.3036749251640023, "grad_norm": 0.1015625, "learning_rate": 1.6323339219526447e-05, "loss": 0.0158, "step": 2980 }, { "epoch": 0.30469396853703584, "grad_norm": 0.0341796875, "learning_rate": 1.6297734542700438e-05, "loss": 0.0173, "step": 2990 }, { "epoch": 0.30571301191006944, "grad_norm": 0.06884765625, "learning_rate": 1.6272061255103247e-05, "loss": 0.012, "step": 3000 }, { "epoch": 0.306732055283103, "grad_norm": 0.043701171875, "learning_rate": 1.624631963643292e-05, "loss": 0.0165, "step": 3010 }, { "epoch": 0.30775109865613653, "grad_norm": 0.046142578125, "learning_rate": 1.622050996713194e-05, "loss": 0.0145, "step": 3020 }, { "epoch": 0.30877014202917014, "grad_norm": 0.0791015625, "learning_rate": 1.6194632528384167e-05, "loss": 0.0144, "step": 3030 }, { "epoch": 0.3097891854022037, "grad_norm": 0.07275390625, "learning_rate": 1.616868760211178e-05, "loss": 0.0157, "step": 3040 }, { "epoch": 0.3108082287752372, "grad_norm": 0.07861328125, "learning_rate": 1.6142675470972198e-05, "loss": 0.0129, "step": 3050 }, { "epoch": 0.31182727214827083, "grad_norm": 0.05224609375, "learning_rate": 1.6116596418355007e-05, "loss": 0.0186, "step": 3060 }, { "epoch": 0.3128463155213044, "grad_norm": 0.037841796875, "learning_rate": 1.6090450728378865e-05, "loss": 0.0154, "step": 3070 }, { "epoch": 0.3138653588943379, "grad_norm": 0.059326171875, "learning_rate": 1.6064238685888424e-05, "loss": 0.0168, "step": 3080 }, { "epoch": 0.3148844022673715, "grad_norm": 0.039306640625, "learning_rate": 1.6037960576451198e-05, "loss": 0.0148, "step": 3090 }, { "epoch": 0.31590344564040507, "grad_norm": 0.058837890625, "learning_rate": 1.6011616686354478e-05, "loss": 0.0148, "step": 3100 }, { "epoch": 0.3169224890134386, "grad_norm": 0.0625, "learning_rate": 1.5985207302602205e-05, "loss": 0.0163, "step": 3110 }, { "epoch": 0.3179415323864722, "grad_norm": 0.03466796875, "learning_rate": 1.595873271291184e-05, "loss": 0.0159, "step": 3120 }, { "epoch": 0.31896057575950576, "grad_norm": 0.041748046875, "learning_rate": 1.593219320571123e-05, "loss": 0.0162, "step": 3130 }, { "epoch": 0.3199796191325393, "grad_norm": 0.035400390625, "learning_rate": 1.590558907013546e-05, "loss": 0.0142, "step": 3140 }, { "epoch": 0.3209986625055729, "grad_norm": 0.034912109375, "learning_rate": 1.587892059602372e-05, "loss": 0.0131, "step": 3150 }, { "epoch": 0.32201770587860645, "grad_norm": 0.044921875, "learning_rate": 1.5852188073916145e-05, "loss": 0.0165, "step": 3160 }, { "epoch": 0.32303674925164, "grad_norm": 0.042236328125, "learning_rate": 1.5825391795050615e-05, "loss": 0.0161, "step": 3170 }, { "epoch": 0.3240557926246736, "grad_norm": 0.044189453125, "learning_rate": 1.5798532051359628e-05, "loss": 0.0135, "step": 3180 }, { "epoch": 0.32507483599770715, "grad_norm": 0.04931640625, "learning_rate": 1.577160913546711e-05, "loss": 0.0153, "step": 3190 }, { "epoch": 0.3260938793707407, "grad_norm": 0.0517578125, "learning_rate": 1.574462334068519e-05, "loss": 0.0133, "step": 3200 }, { "epoch": 0.3271129227437743, "grad_norm": 0.050537109375, "learning_rate": 1.5717574961011058e-05, "loss": 0.0175, "step": 3210 }, { "epoch": 0.32813196611680784, "grad_norm": 0.054931640625, "learning_rate": 1.5690464291123716e-05, "loss": 0.0147, "step": 3220 }, { "epoch": 0.3291510094898414, "grad_norm": 0.08984375, "learning_rate": 1.5663291626380805e-05, "loss": 0.0133, "step": 3230 }, { "epoch": 0.330170052862875, "grad_norm": 0.055419921875, "learning_rate": 1.5636057262815357e-05, "loss": 0.0155, "step": 3240 }, { "epoch": 0.33118909623590853, "grad_norm": 0.02587890625, "learning_rate": 1.560876149713259e-05, "loss": 0.0132, "step": 3250 }, { "epoch": 0.33220813960894213, "grad_norm": 0.059326171875, "learning_rate": 1.558140462670667e-05, "loss": 0.0181, "step": 3260 }, { "epoch": 0.3332271829819757, "grad_norm": 0.0390625, "learning_rate": 1.555398694957746e-05, "loss": 0.0169, "step": 3270 }, { "epoch": 0.3342462263550092, "grad_norm": 0.0361328125, "learning_rate": 1.552650876444729e-05, "loss": 0.0136, "step": 3280 }, { "epoch": 0.3352652697280428, "grad_norm": 0.04638671875, "learning_rate": 1.5498970370677694e-05, "loss": 0.0164, "step": 3290 }, { "epoch": 0.33628431310107637, "grad_norm": 0.06982421875, "learning_rate": 1.547137206828615e-05, "loss": 0.0122, "step": 3300 }, { "epoch": 0.3373033564741099, "grad_norm": 0.07861328125, "learning_rate": 1.544371415794281e-05, "loss": 0.0167, "step": 3310 }, { "epoch": 0.3383223998471435, "grad_norm": 0.06982421875, "learning_rate": 1.541599694096723e-05, "loss": 0.0174, "step": 3320 }, { "epoch": 0.33934144322017706, "grad_norm": 0.08642578125, "learning_rate": 1.538822071932508e-05, "loss": 0.0146, "step": 3330 }, { "epoch": 0.3403604865932106, "grad_norm": 0.03759765625, "learning_rate": 1.5360385795624853e-05, "loss": 0.0178, "step": 3340 }, { "epoch": 0.3413795299662442, "grad_norm": 0.043701171875, "learning_rate": 1.5332492473114584e-05, "loss": 0.0139, "step": 3350 }, { "epoch": 0.34239857333927776, "grad_norm": 0.047607421875, "learning_rate": 1.530454105567852e-05, "loss": 0.0154, "step": 3360 }, { "epoch": 0.3434176167123113, "grad_norm": 0.0595703125, "learning_rate": 1.5276531847833834e-05, "loss": 0.0145, "step": 3370 }, { "epoch": 0.3444366600853449, "grad_norm": 0.046875, "learning_rate": 1.52484651547273e-05, "loss": 0.0148, "step": 3380 }, { "epoch": 0.34545570345837845, "grad_norm": 0.0361328125, "learning_rate": 1.5220341282131963e-05, "loss": 0.0183, "step": 3390 }, { "epoch": 0.346474746831412, "grad_norm": 0.055908203125, "learning_rate": 1.5192160536443798e-05, "loss": 0.0142, "step": 3400 }, { "epoch": 0.3474937902044456, "grad_norm": 0.047607421875, "learning_rate": 1.5163923224678407e-05, "loss": 0.0179, "step": 3410 }, { "epoch": 0.34851283357747914, "grad_norm": 0.09326171875, "learning_rate": 1.5135629654467639e-05, "loss": 0.0154, "step": 3420 }, { "epoch": 0.3495318769505127, "grad_norm": 0.0712890625, "learning_rate": 1.5107280134056251e-05, "loss": 0.0148, "step": 3430 }, { "epoch": 0.3505509203235463, "grad_norm": 0.0732421875, "learning_rate": 1.5078874972298554e-05, "loss": 0.0157, "step": 3440 }, { "epoch": 0.35156996369657983, "grad_norm": 0.05419921875, "learning_rate": 1.5050414478655046e-05, "loss": 0.0134, "step": 3450 }, { "epoch": 0.3525890070696134, "grad_norm": 0.051025390625, "learning_rate": 1.5021898963189037e-05, "loss": 0.0159, "step": 3460 }, { "epoch": 0.353608050442647, "grad_norm": 0.0625, "learning_rate": 1.4993328736563267e-05, "loss": 0.0154, "step": 3470 }, { "epoch": 0.3546270938156805, "grad_norm": 0.048828125, "learning_rate": 1.4964704110036541e-05, "loss": 0.0148, "step": 3480 }, { "epoch": 0.3556461371887141, "grad_norm": 0.045654296875, "learning_rate": 1.4936025395460314e-05, "loss": 0.016, "step": 3490 }, { "epoch": 0.3566651805617477, "grad_norm": 0.07080078125, "learning_rate": 1.490729290527531e-05, "loss": 0.0126, "step": 3500 }, { "epoch": 0.3576842239347812, "grad_norm": 0.0498046875, "learning_rate": 1.4878506952508103e-05, "loss": 0.0171, "step": 3510 }, { "epoch": 0.35870326730781477, "grad_norm": 0.05029296875, "learning_rate": 1.484966785076773e-05, "loss": 0.016, "step": 3520 }, { "epoch": 0.35972231068084837, "grad_norm": 0.03369140625, "learning_rate": 1.4820775914242249e-05, "loss": 0.015, "step": 3530 }, { "epoch": 0.3607413540538819, "grad_norm": 0.06103515625, "learning_rate": 1.4791831457695333e-05, "loss": 0.0155, "step": 3540 }, { "epoch": 0.36176039742691546, "grad_norm": 0.03955078125, "learning_rate": 1.4762834796462832e-05, "loss": 0.0143, "step": 3550 }, { "epoch": 0.36277944079994906, "grad_norm": 0.05712890625, "learning_rate": 1.4733786246449346e-05, "loss": 0.0166, "step": 3560 }, { "epoch": 0.3637984841729826, "grad_norm": 0.08740234375, "learning_rate": 1.4704686124124767e-05, "loss": 0.0151, "step": 3570 }, { "epoch": 0.36481752754601615, "grad_norm": 0.06689453125, "learning_rate": 1.4675534746520858e-05, "loss": 0.016, "step": 3580 }, { "epoch": 0.36583657091904975, "grad_norm": 0.06298828125, "learning_rate": 1.4646332431227775e-05, "loss": 0.0164, "step": 3590 }, { "epoch": 0.3668556142920833, "grad_norm": 0.047607421875, "learning_rate": 1.461707949639061e-05, "loss": 0.0126, "step": 3600 }, { "epoch": 0.36787465766511684, "grad_norm": 0.09619140625, "learning_rate": 1.4587776260705937e-05, "loss": 0.0174, "step": 3610 }, { "epoch": 0.36889370103815045, "grad_norm": 0.03515625, "learning_rate": 1.4558423043418337e-05, "loss": 0.0165, "step": 3620 }, { "epoch": 0.369912744411184, "grad_norm": 0.06787109375, "learning_rate": 1.4529020164316913e-05, "loss": 0.0137, "step": 3630 }, { "epoch": 0.3709317877842176, "grad_norm": 0.04150390625, "learning_rate": 1.4499567943731803e-05, "loss": 0.017, "step": 3640 }, { "epoch": 0.37195083115725114, "grad_norm": 0.06396484375, "learning_rate": 1.4470066702530705e-05, "loss": 0.014, "step": 3650 }, { "epoch": 0.3729698745302847, "grad_norm": 0.0537109375, "learning_rate": 1.4440516762115372e-05, "loss": 0.0167, "step": 3660 }, { "epoch": 0.3739889179033183, "grad_norm": 0.0380859375, "learning_rate": 1.4410918444418106e-05, "loss": 0.0151, "step": 3670 }, { "epoch": 0.37500796127635183, "grad_norm": 0.10498046875, "learning_rate": 1.4381272071898263e-05, "loss": 0.0142, "step": 3680 }, { "epoch": 0.3760270046493854, "grad_norm": 0.03515625, "learning_rate": 1.4351577967538726e-05, "loss": 0.0172, "step": 3690 }, { "epoch": 0.377046048022419, "grad_norm": 0.057861328125, "learning_rate": 1.4321836454842399e-05, "loss": 0.0145, "step": 3700 }, { "epoch": 0.3780650913954525, "grad_norm": 0.0498046875, "learning_rate": 1.4292047857828672e-05, "loss": 0.0167, "step": 3710 }, { "epoch": 0.37908413476848607, "grad_norm": 0.060546875, "learning_rate": 1.4262212501029904e-05, "loss": 0.0157, "step": 3720 }, { "epoch": 0.38010317814151967, "grad_norm": 0.04931640625, "learning_rate": 1.4232330709487871e-05, "loss": 0.0149, "step": 3730 }, { "epoch": 0.3811222215145532, "grad_norm": 0.048095703125, "learning_rate": 1.4202402808750235e-05, "loss": 0.0184, "step": 3740 }, { "epoch": 0.38214126488758676, "grad_norm": 0.05029296875, "learning_rate": 1.4172429124866993e-05, "loss": 0.0137, "step": 3750 }, { "epoch": 0.38316030826062036, "grad_norm": 0.046875, "learning_rate": 1.4142409984386937e-05, "loss": 0.0167, "step": 3760 }, { "epoch": 0.3841793516336539, "grad_norm": 0.052978515625, "learning_rate": 1.411234571435408e-05, "loss": 0.0146, "step": 3770 }, { "epoch": 0.38519839500668746, "grad_norm": 0.043212890625, "learning_rate": 1.4082236642304095e-05, "loss": 0.0146, "step": 3780 }, { "epoch": 0.38621743837972106, "grad_norm": 0.037109375, "learning_rate": 1.4052083096260758e-05, "loss": 0.0174, "step": 3790 }, { "epoch": 0.3872364817527546, "grad_norm": 0.09912109375, "learning_rate": 1.4021885404732367e-05, "loss": 0.0129, "step": 3800 }, { "epoch": 0.38825552512578815, "grad_norm": 0.09326171875, "learning_rate": 1.3991643896708163e-05, "loss": 0.0167, "step": 3810 }, { "epoch": 0.38927456849882175, "grad_norm": 0.091796875, "learning_rate": 1.3961358901654743e-05, "loss": 0.0153, "step": 3820 }, { "epoch": 0.3902936118718553, "grad_norm": 0.0771484375, "learning_rate": 1.393103074951248e-05, "loss": 0.0179, "step": 3830 }, { "epoch": 0.39131265524488884, "grad_norm": 0.05224609375, "learning_rate": 1.3900659770691918e-05, "loss": 0.0165, "step": 3840 }, { "epoch": 0.39233169861792244, "grad_norm": 0.0439453125, "learning_rate": 1.3870246296070176e-05, "loss": 0.013, "step": 3850 }, { "epoch": 0.393350741990956, "grad_norm": 0.041015625, "learning_rate": 1.3839790656987355e-05, "loss": 0.0162, "step": 3860 }, { "epoch": 0.39436978536398953, "grad_norm": 0.044189453125, "learning_rate": 1.3809293185242903e-05, "loss": 0.0158, "step": 3870 }, { "epoch": 0.39538882873702313, "grad_norm": 0.048095703125, "learning_rate": 1.3778754213092019e-05, "loss": 0.0144, "step": 3880 }, { "epoch": 0.3964078721100567, "grad_norm": 0.05322265625, "learning_rate": 1.3748174073242035e-05, "loss": 0.0153, "step": 3890 }, { "epoch": 0.3974269154830902, "grad_norm": 0.052978515625, "learning_rate": 1.3717553098848781e-05, "loss": 0.0128, "step": 3900 }, { "epoch": 0.3984459588561238, "grad_norm": 0.051513671875, "learning_rate": 1.368689162351296e-05, "loss": 0.0171, "step": 3910 }, { "epoch": 0.3994650022291574, "grad_norm": 0.044921875, "learning_rate": 1.3656189981276507e-05, "loss": 0.0158, "step": 3920 }, { "epoch": 0.4004840456021909, "grad_norm": 0.06591796875, "learning_rate": 1.3625448506618967e-05, "loss": 0.0147, "step": 3930 }, { "epoch": 0.4015030889752245, "grad_norm": 0.0517578125, "learning_rate": 1.3594667534453834e-05, "loss": 0.0172, "step": 3940 }, { "epoch": 0.40252213234825807, "grad_norm": 0.04833984375, "learning_rate": 1.3563847400124906e-05, "loss": 0.0149, "step": 3950 }, { "epoch": 0.4035411757212916, "grad_norm": 0.055908203125, "learning_rate": 1.353298843940264e-05, "loss": 0.017, "step": 3960 }, { "epoch": 0.4045602190943252, "grad_norm": 0.036865234375, "learning_rate": 1.3502090988480483e-05, "loss": 0.015, "step": 3970 }, { "epoch": 0.40557926246735876, "grad_norm": 0.06591796875, "learning_rate": 1.3471155383971215e-05, "loss": 0.0145, "step": 3980 }, { "epoch": 0.4065983058403923, "grad_norm": 0.052001953125, "learning_rate": 1.344018196290328e-05, "loss": 0.0156, "step": 3990 }, { "epoch": 0.4076173492134259, "grad_norm": 0.035888671875, "learning_rate": 1.340917106271712e-05, "loss": 0.0147, "step": 4000 }, { "epoch": 0.40863639258645945, "grad_norm": 0.0576171875, "learning_rate": 1.337812302126149e-05, "loss": 0.0156, "step": 4010 }, { "epoch": 0.40965543595949305, "grad_norm": 0.047119140625, "learning_rate": 1.3347038176789778e-05, "loss": 0.0143, "step": 4020 }, { "epoch": 0.4106744793325266, "grad_norm": 0.0400390625, "learning_rate": 1.3315916867956329e-05, "loss": 0.0153, "step": 4030 }, { "epoch": 0.41169352270556014, "grad_norm": 0.0498046875, "learning_rate": 1.3284759433812748e-05, "loss": 0.0161, "step": 4040 }, { "epoch": 0.41271256607859375, "grad_norm": 0.0546875, "learning_rate": 1.3253566213804208e-05, "loss": 0.0135, "step": 4050 }, { "epoch": 0.4137316094516273, "grad_norm": 0.059326171875, "learning_rate": 1.3222337547765743e-05, "loss": 0.0152, "step": 4060 }, { "epoch": 0.41475065282466084, "grad_norm": 0.047119140625, "learning_rate": 1.3191073775918571e-05, "loss": 0.0158, "step": 4070 }, { "epoch": 0.41576969619769444, "grad_norm": 0.12060546875, "learning_rate": 1.315977523886636e-05, "loss": 0.0132, "step": 4080 }, { "epoch": 0.416788739570728, "grad_norm": 0.07421875, "learning_rate": 1.3128442277591522e-05, "loss": 0.0162, "step": 4090 }, { "epoch": 0.41780778294376153, "grad_norm": 0.083984375, "learning_rate": 1.3097075233451523e-05, "loss": 0.0142, "step": 4100 }, { "epoch": 0.41882682631679513, "grad_norm": 0.08349609375, "learning_rate": 1.3065674448175129e-05, "loss": 0.0179, "step": 4110 }, { "epoch": 0.4198458696898287, "grad_norm": 0.046630859375, "learning_rate": 1.3034240263858706e-05, "loss": 0.0161, "step": 4120 }, { "epoch": 0.4208649130628622, "grad_norm": 0.038330078125, "learning_rate": 1.3002773022962485e-05, "loss": 0.0143, "step": 4130 }, { "epoch": 0.4218839564358958, "grad_norm": 0.03662109375, "learning_rate": 1.297127306830683e-05, "loss": 0.0161, "step": 4140 }, { "epoch": 0.42290299980892937, "grad_norm": 0.060302734375, "learning_rate": 1.2939740743068515e-05, "loss": 0.0134, "step": 4150 }, { "epoch": 0.4239220431819629, "grad_norm": 0.0595703125, "learning_rate": 1.2908176390776958e-05, "loss": 0.0193, "step": 4160 }, { "epoch": 0.4249410865549965, "grad_norm": 0.035400390625, "learning_rate": 1.287658035531052e-05, "loss": 0.0163, "step": 4170 }, { "epoch": 0.42596012992803006, "grad_norm": 0.039306640625, "learning_rate": 1.2844952980892715e-05, "loss": 0.0151, "step": 4180 }, { "epoch": 0.4269791733010636, "grad_norm": 0.048095703125, "learning_rate": 1.2813294612088485e-05, "loss": 0.0153, "step": 4190 }, { "epoch": 0.4279982166740972, "grad_norm": 0.049072265625, "learning_rate": 1.2781605593800449e-05, "loss": 0.0125, "step": 4200 }, { "epoch": 0.42901726004713076, "grad_norm": 0.05126953125, "learning_rate": 1.2749886271265127e-05, "loss": 0.0163, "step": 4210 }, { "epoch": 0.4300363034201643, "grad_norm": 0.0390625, "learning_rate": 1.2718136990049193e-05, "loss": 0.0136, "step": 4220 }, { "epoch": 0.4310553467931979, "grad_norm": 0.06494140625, "learning_rate": 1.2686358096045706e-05, "loss": 0.0147, "step": 4230 }, { "epoch": 0.43207439016623145, "grad_norm": 0.0439453125, "learning_rate": 1.2654549935470346e-05, "loss": 0.0168, "step": 4240 }, { "epoch": 0.433093433539265, "grad_norm": 0.05810546875, "learning_rate": 1.2622712854857632e-05, "loss": 0.012, "step": 4250 }, { "epoch": 0.4341124769122986, "grad_norm": 0.06201171875, "learning_rate": 1.2590847201057155e-05, "loss": 0.0156, "step": 4260 }, { "epoch": 0.43513152028533214, "grad_norm": 0.049072265625, "learning_rate": 1.2558953321229799e-05, "loss": 0.016, "step": 4270 }, { "epoch": 0.4361505636583657, "grad_norm": 0.043212890625, "learning_rate": 1.2527031562843954e-05, "loss": 0.0161, "step": 4280 }, { "epoch": 0.4371696070313993, "grad_norm": 0.07275390625, "learning_rate": 1.2495082273671736e-05, "loss": 0.0181, "step": 4290 }, { "epoch": 0.43818865040443283, "grad_norm": 0.06787109375, "learning_rate": 1.2463105801785196e-05, "loss": 0.0128, "step": 4300 }, { "epoch": 0.4392076937774664, "grad_norm": 0.0419921875, "learning_rate": 1.2431102495552524e-05, "loss": 0.0176, "step": 4310 }, { "epoch": 0.4402267371505, "grad_norm": 0.057861328125, "learning_rate": 1.2399072703634257e-05, "loss": 0.0168, "step": 4320 }, { "epoch": 0.4412457805235335, "grad_norm": 0.06982421875, "learning_rate": 1.236701677497948e-05, "loss": 0.0146, "step": 4330 }, { "epoch": 0.44226482389656707, "grad_norm": 0.0693359375, "learning_rate": 1.2334935058822042e-05, "loss": 0.0156, "step": 4340 }, { "epoch": 0.4432838672696007, "grad_norm": 0.03466796875, "learning_rate": 1.2302827904676706e-05, "loss": 0.0114, "step": 4350 }, { "epoch": 0.4443029106426342, "grad_norm": 0.05126953125, "learning_rate": 1.2270695662335393e-05, "loss": 0.0162, "step": 4360 }, { "epoch": 0.44532195401566776, "grad_norm": 0.043212890625, "learning_rate": 1.223853868186333e-05, "loss": 0.0152, "step": 4370 }, { "epoch": 0.44634099738870137, "grad_norm": 0.099609375, "learning_rate": 1.220635731359527e-05, "loss": 0.0143, "step": 4380 }, { "epoch": 0.4473600407617349, "grad_norm": 0.03564453125, "learning_rate": 1.217415190813165e-05, "loss": 0.0152, "step": 4390 }, { "epoch": 0.4483790841347685, "grad_norm": 0.051513671875, "learning_rate": 1.2141922816334779e-05, "loss": 0.0144, "step": 4400 }, { "epoch": 0.44939812750780206, "grad_norm": 0.050048828125, "learning_rate": 1.2109670389325028e-05, "loss": 0.0167, "step": 4410 }, { "epoch": 0.4504171708808356, "grad_norm": 0.07275390625, "learning_rate": 1.207739497847698e-05, "loss": 0.0161, "step": 4420 }, { "epoch": 0.4514362142538692, "grad_norm": 0.0498046875, "learning_rate": 1.2045096935415623e-05, "loss": 0.0135, "step": 4430 }, { "epoch": 0.45245525762690275, "grad_norm": 0.04150390625, "learning_rate": 1.2012776612012511e-05, "loss": 0.0158, "step": 4440 }, { "epoch": 0.4534743009999363, "grad_norm": 0.055908203125, "learning_rate": 1.1980434360381935e-05, "loss": 0.0154, "step": 4450 }, { "epoch": 0.4544933443729699, "grad_norm": 0.05029296875, "learning_rate": 1.1948070532877083e-05, "loss": 0.0169, "step": 4460 }, { "epoch": 0.45551238774600344, "grad_norm": 0.03955078125, "learning_rate": 1.1915685482086187e-05, "loss": 0.0169, "step": 4470 }, { "epoch": 0.456531431119037, "grad_norm": 0.047607421875, "learning_rate": 1.1883279560828722e-05, "loss": 0.0134, "step": 4480 }, { "epoch": 0.4575504744920706, "grad_norm": 0.038818359375, "learning_rate": 1.1850853122151516e-05, "loss": 0.0155, "step": 4490 }, { "epoch": 0.45856951786510414, "grad_norm": 0.06005859375, "learning_rate": 1.1818406519324933e-05, "loss": 0.0119, "step": 4500 }, { "epoch": 0.4595885612381377, "grad_norm": 0.055419921875, "learning_rate": 1.1785940105839013e-05, "loss": 0.0169, "step": 4510 }, { "epoch": 0.4606076046111713, "grad_norm": 0.04150390625, "learning_rate": 1.1753454235399625e-05, "loss": 0.0156, "step": 4520 }, { "epoch": 0.46162664798420483, "grad_norm": 0.038330078125, "learning_rate": 1.1720949261924614e-05, "loss": 0.0124, "step": 4530 }, { "epoch": 0.4626456913572384, "grad_norm": 0.06396484375, "learning_rate": 1.1688425539539936e-05, "loss": 0.0149, "step": 4540 }, { "epoch": 0.463664734730272, "grad_norm": 0.048828125, "learning_rate": 1.1655883422575818e-05, "loss": 0.0137, "step": 4550 }, { "epoch": 0.4646837781033055, "grad_norm": 0.07861328125, "learning_rate": 1.1623323265562876e-05, "loss": 0.0165, "step": 4560 }, { "epoch": 0.46570282147633907, "grad_norm": 0.06494140625, "learning_rate": 1.1590745423228273e-05, "loss": 0.0149, "step": 4570 }, { "epoch": 0.46672186484937267, "grad_norm": 0.042236328125, "learning_rate": 1.1558150250491843e-05, "loss": 0.0147, "step": 4580 }, { "epoch": 0.4677409082224062, "grad_norm": 0.03955078125, "learning_rate": 1.1525538102462217e-05, "loss": 0.0135, "step": 4590 }, { "epoch": 0.46875995159543976, "grad_norm": 0.09521484375, "learning_rate": 1.1492909334432978e-05, "loss": 0.0137, "step": 4600 }, { "epoch": 0.46977899496847336, "grad_norm": 0.06640625, "learning_rate": 1.1460264301878762e-05, "loss": 0.016, "step": 4610 }, { "epoch": 0.4707980383415069, "grad_norm": 0.06396484375, "learning_rate": 1.1427603360451415e-05, "loss": 0.0146, "step": 4620 }, { "epoch": 0.47181708171454045, "grad_norm": 0.06982421875, "learning_rate": 1.1394926865976088e-05, "loss": 0.0141, "step": 4630 }, { "epoch": 0.47283612508757406, "grad_norm": 0.05810546875, "learning_rate": 1.1362235174447378e-05, "loss": 0.0164, "step": 4640 }, { "epoch": 0.4738551684606076, "grad_norm": 0.0400390625, "learning_rate": 1.1329528642025453e-05, "loss": 0.0132, "step": 4650 }, { "epoch": 0.47487421183364115, "grad_norm": 0.0556640625, "learning_rate": 1.1296807625032162e-05, "loss": 0.0177, "step": 4660 }, { "epoch": 0.47589325520667475, "grad_norm": 0.03759765625, "learning_rate": 1.1264072479947153e-05, "loss": 0.0151, "step": 4670 }, { "epoch": 0.4769122985797083, "grad_norm": 0.05224609375, "learning_rate": 1.1231323563403994e-05, "loss": 0.0133, "step": 4680 }, { "epoch": 0.47793134195274184, "grad_norm": 0.060302734375, "learning_rate": 1.1198561232186291e-05, "loss": 0.0144, "step": 4690 }, { "epoch": 0.47895038532577544, "grad_norm": 0.056640625, "learning_rate": 1.1165785843223792e-05, "loss": 0.0124, "step": 4700 }, { "epoch": 0.479969428698809, "grad_norm": 0.060791015625, "learning_rate": 1.1132997753588503e-05, "loss": 0.0167, "step": 4710 }, { "epoch": 0.48098847207184253, "grad_norm": 0.04931640625, "learning_rate": 1.1100197320490802e-05, "loss": 0.016, "step": 4720 }, { "epoch": 0.48200751544487613, "grad_norm": 0.2216796875, "learning_rate": 1.1067384901275538e-05, "loss": 0.0129, "step": 4730 }, { "epoch": 0.4830265588179097, "grad_norm": 0.0400390625, "learning_rate": 1.1034560853418144e-05, "loss": 0.0157, "step": 4740 }, { "epoch": 0.4840456021909432, "grad_norm": 0.0654296875, "learning_rate": 1.100172553452074e-05, "loss": 0.0116, "step": 4750 }, { "epoch": 0.4850646455639768, "grad_norm": 0.0361328125, "learning_rate": 1.0968879302308244e-05, "loss": 0.0166, "step": 4760 }, { "epoch": 0.48608368893701037, "grad_norm": 0.057861328125, "learning_rate": 1.0936022514624464e-05, "loss": 0.0141, "step": 4770 }, { "epoch": 0.487102732310044, "grad_norm": 0.04345703125, "learning_rate": 1.0903155529428207e-05, "loss": 0.0152, "step": 4780 }, { "epoch": 0.4881217756830775, "grad_norm": 0.0712890625, "learning_rate": 1.087027870478938e-05, "loss": 0.017, "step": 4790 }, { "epoch": 0.48914081905611106, "grad_norm": 0.05224609375, "learning_rate": 1.0837392398885078e-05, "loss": 0.014, "step": 4800 }, { "epoch": 0.49015986242914467, "grad_norm": 0.046875, "learning_rate": 1.0804496969995694e-05, "loss": 0.0172, "step": 4810 }, { "epoch": 0.4911789058021782, "grad_norm": 0.03759765625, "learning_rate": 1.0771592776501017e-05, "loss": 0.0157, "step": 4820 }, { "epoch": 0.49219794917521176, "grad_norm": 0.041259765625, "learning_rate": 1.0738680176876315e-05, "loss": 0.0149, "step": 4830 }, { "epoch": 0.49321699254824536, "grad_norm": 0.061279296875, "learning_rate": 1.0705759529688435e-05, "loss": 0.0141, "step": 4840 }, { "epoch": 0.4942360359212789, "grad_norm": 0.060302734375, "learning_rate": 1.0672831193591905e-05, "loss": 0.0147, "step": 4850 }, { "epoch": 0.49525507929431245, "grad_norm": 0.035400390625, "learning_rate": 1.0639895527325018e-05, "loss": 0.0171, "step": 4860 }, { "epoch": 0.49627412266734605, "grad_norm": 0.04345703125, "learning_rate": 1.0606952889705922e-05, "loss": 0.0137, "step": 4870 }, { "epoch": 0.4972931660403796, "grad_norm": 0.08203125, "learning_rate": 1.0574003639628718e-05, "loss": 0.016, "step": 4880 }, { "epoch": 0.49831220941341314, "grad_norm": 0.06982421875, "learning_rate": 1.0541048136059547e-05, "loss": 0.0158, "step": 4890 }, { "epoch": 0.49933125278644674, "grad_norm": 0.049560546875, "learning_rate": 1.0508086738032679e-05, "loss": 0.014, "step": 4900 }, { "epoch": 0.5003502961594802, "grad_norm": 0.053466796875, "learning_rate": 1.0475119804646596e-05, "loss": 0.0174, "step": 4910 }, { "epoch": 0.5013693395325138, "grad_norm": 0.045166015625, "learning_rate": 1.044214769506009e-05, "loss": 0.0159, "step": 4920 }, { "epoch": 0.5023883829055474, "grad_norm": 0.034912109375, "learning_rate": 1.0409170768488344e-05, "loss": 0.0139, "step": 4930 }, { "epoch": 0.5034074262785809, "grad_norm": 0.04541015625, "learning_rate": 1.0376189384199023e-05, "loss": 0.0174, "step": 4940 }, { "epoch": 0.5044264696516145, "grad_norm": 0.08154296875, "learning_rate": 1.0343203901508346e-05, "loss": 0.0143, "step": 4950 }, { "epoch": 0.5054455130246481, "grad_norm": 0.05615234375, "learning_rate": 1.0310214679777198e-05, "loss": 0.0178, "step": 4960 }, { "epoch": 0.5064645563976817, "grad_norm": 0.037109375, "learning_rate": 1.0277222078407186e-05, "loss": 0.0144, "step": 4970 }, { "epoch": 0.5074835997707152, "grad_norm": 0.047607421875, "learning_rate": 1.024422645683674e-05, "loss": 0.0144, "step": 4980 }, { "epoch": 0.5085026431437488, "grad_norm": 0.059326171875, "learning_rate": 1.0211228174537195e-05, "loss": 0.0158, "step": 4990 }, { "epoch": 0.5095216865167824, "grad_norm": 0.06298828125, "learning_rate": 1.0178227591008878e-05, "loss": 0.0126, "step": 5000 }, { "epoch": 0.5105407298898159, "grad_norm": 0.055908203125, "learning_rate": 1.0145225065777178e-05, "loss": 0.0153, "step": 5010 }, { "epoch": 0.5115597732628495, "grad_norm": 0.035888671875, "learning_rate": 1.011222095838864e-05, "loss": 0.0155, "step": 5020 }, { "epoch": 0.5125788166358831, "grad_norm": 0.07568359375, "learning_rate": 1.0079215628407048e-05, "loss": 0.0161, "step": 5030 }, { "epoch": 0.5135978600089166, "grad_norm": 0.064453125, "learning_rate": 1.0046209435409511e-05, "loss": 0.0155, "step": 5040 }, { "epoch": 0.5146169033819502, "grad_norm": 0.04150390625, "learning_rate": 1.0013202738982527e-05, "loss": 0.012, "step": 5050 }, { "epoch": 0.5156359467549838, "grad_norm": 0.057373046875, "learning_rate": 9.98019589871809e-06, "loss": 0.0181, "step": 5060 }, { "epoch": 0.5166549901280173, "grad_norm": 0.05712890625, "learning_rate": 9.94718927420976e-06, "loss": 0.0151, "step": 5070 }, { "epoch": 0.5176740335010509, "grad_norm": 0.051025390625, "learning_rate": 9.914183225048732e-06, "loss": 0.0149, "step": 5080 }, { "epoch": 0.5186930768740845, "grad_norm": 0.046142578125, "learning_rate": 9.881178110819954e-06, "loss": 0.0156, "step": 5090 }, { "epoch": 0.519712120247118, "grad_norm": 0.051025390625, "learning_rate": 9.848174291098175e-06, "loss": 0.0126, "step": 5100 }, { "epoch": 0.5207311636201516, "grad_norm": 0.0625, "learning_rate": 9.815172125444052e-06, "loss": 0.0173, "step": 5110 }, { "epoch": 0.5217502069931852, "grad_norm": 0.052978515625, "learning_rate": 9.7821719734002e-06, "loss": 0.0156, "step": 5120 }, { "epoch": 0.5227692503662187, "grad_norm": 0.068359375, "learning_rate": 9.749174194487323e-06, "loss": 0.0137, "step": 5130 }, { "epoch": 0.5237882937392523, "grad_norm": 0.06640625, "learning_rate": 9.716179148200252e-06, "loss": 0.0162, "step": 5140 }, { "epoch": 0.5248073371122859, "grad_norm": 0.0732421875, "learning_rate": 9.683187194004062e-06, "loss": 0.0137, "step": 5150 }, { "epoch": 0.5258263804853194, "grad_norm": 0.060302734375, "learning_rate": 9.650198691330123e-06, "loss": 0.0159, "step": 5160 }, { "epoch": 0.526845423858353, "grad_norm": 0.033447265625, "learning_rate": 9.617213999572223e-06, "loss": 0.0141, "step": 5170 }, { "epoch": 0.5278644672313866, "grad_norm": 0.04052734375, "learning_rate": 9.584233478082615e-06, "loss": 0.015, "step": 5180 }, { "epoch": 0.5288835106044201, "grad_norm": 0.04736328125, "learning_rate": 9.551257486168128e-06, "loss": 0.0157, "step": 5190 }, { "epoch": 0.5299025539774537, "grad_norm": 0.045166015625, "learning_rate": 9.518286383086238e-06, "loss": 0.0124, "step": 5200 }, { "epoch": 0.5309215973504873, "grad_norm": 0.037109375, "learning_rate": 9.485320528041171e-06, "loss": 0.0157, "step": 5210 }, { "epoch": 0.5319406407235208, "grad_norm": 0.034423828125, "learning_rate": 9.45236028017996e-06, "loss": 0.0128, "step": 5220 }, { "epoch": 0.5329596840965544, "grad_norm": 0.130859375, "learning_rate": 9.419405998588565e-06, "loss": 0.0154, "step": 5230 }, { "epoch": 0.533978727469588, "grad_norm": 0.055419921875, "learning_rate": 9.386458042287941e-06, "loss": 0.0144, "step": 5240 }, { "epoch": 0.5349977708426215, "grad_norm": 0.051513671875, "learning_rate": 9.353516770230134e-06, "loss": 0.0127, "step": 5250 }, { "epoch": 0.5360168142156551, "grad_norm": 0.0888671875, "learning_rate": 9.320582541294365e-06, "loss": 0.0162, "step": 5260 }, { "epoch": 0.5370358575886887, "grad_norm": 0.05078125, "learning_rate": 9.287655714283126e-06, "loss": 0.0141, "step": 5270 }, { "epoch": 0.5380549009617221, "grad_norm": 0.045166015625, "learning_rate": 9.254736647918266e-06, "loss": 0.0137, "step": 5280 }, { "epoch": 0.5390739443347558, "grad_norm": 0.03955078125, "learning_rate": 9.221825700837097e-06, "loss": 0.0159, "step": 5290 }, { "epoch": 0.5400929877077894, "grad_norm": 0.05859375, "learning_rate": 9.188923231588451e-06, "loss": 0.0125, "step": 5300 }, { "epoch": 0.5411120310808228, "grad_norm": 0.044677734375, "learning_rate": 9.156029598628829e-06, "loss": 0.0168, "step": 5310 }, { "epoch": 0.5421310744538564, "grad_norm": 0.05322265625, "learning_rate": 9.123145160318441e-06, "loss": 0.0154, "step": 5320 }, { "epoch": 0.54315011782689, "grad_norm": 0.03857421875, "learning_rate": 9.090270274917331e-06, "loss": 0.0161, "step": 5330 }, { "epoch": 0.5441691611999235, "grad_norm": 0.034423828125, "learning_rate": 9.05740530058148e-06, "loss": 0.0146, "step": 5340 }, { "epoch": 0.5451882045729571, "grad_norm": 0.041748046875, "learning_rate": 9.024550595358874e-06, "loss": 0.0124, "step": 5350 }, { "epoch": 0.5462072479459907, "grad_norm": 0.07080078125, "learning_rate": 8.991706517185641e-06, "loss": 0.0172, "step": 5360 }, { "epoch": 0.5472262913190242, "grad_norm": 0.08544921875, "learning_rate": 8.958873423882118e-06, "loss": 0.0158, "step": 5370 }, { "epoch": 0.5482453346920578, "grad_norm": 0.0625, "learning_rate": 8.92605167314898e-06, "loss": 0.0146, "step": 5380 }, { "epoch": 0.5492643780650914, "grad_norm": 0.0625, "learning_rate": 8.893241622563314e-06, "loss": 0.0173, "step": 5390 }, { "epoch": 0.5502834214381249, "grad_norm": 0.047119140625, "learning_rate": 8.860443629574755e-06, "loss": 0.0113, "step": 5400 }, { "epoch": 0.5513024648111585, "grad_norm": 0.040283203125, "learning_rate": 8.827658051501565e-06, "loss": 0.0166, "step": 5410 }, { "epoch": 0.5523215081841921, "grad_norm": 0.0311279296875, "learning_rate": 8.794885245526764e-06, "loss": 0.0169, "step": 5420 }, { "epoch": 0.5533405515572256, "grad_norm": 0.059814453125, "learning_rate": 8.76212556869421e-06, "loss": 0.0129, "step": 5430 }, { "epoch": 0.5543595949302592, "grad_norm": 0.057373046875, "learning_rate": 8.72937937790474e-06, "loss": 0.0166, "step": 5440 }, { "epoch": 0.5553786383032928, "grad_norm": 0.07373046875, "learning_rate": 8.696647029912255e-06, "loss": 0.0136, "step": 5450 }, { "epoch": 0.5563976816763263, "grad_norm": 0.034423828125, "learning_rate": 8.66392888131986e-06, "loss": 0.0177, "step": 5460 }, { "epoch": 0.5574167250493599, "grad_norm": 0.06689453125, "learning_rate": 8.631225288575946e-06, "loss": 0.0143, "step": 5470 }, { "epoch": 0.5584357684223935, "grad_norm": 0.0546875, "learning_rate": 8.598536607970345e-06, "loss": 0.0144, "step": 5480 }, { "epoch": 0.559454811795427, "grad_norm": 0.05029296875, "learning_rate": 8.56586319563041e-06, "loss": 0.016, "step": 5490 }, { "epoch": 0.5604738551684606, "grad_norm": 0.0517578125, "learning_rate": 8.533205407517167e-06, "loss": 0.0129, "step": 5500 }, { "epoch": 0.5614928985414942, "grad_norm": 0.047119140625, "learning_rate": 8.500563599421413e-06, "loss": 0.0157, "step": 5510 }, { "epoch": 0.5625119419145277, "grad_norm": 0.078125, "learning_rate": 8.467938126959867e-06, "loss": 0.0148, "step": 5520 }, { "epoch": 0.5635309852875613, "grad_norm": 0.055908203125, "learning_rate": 8.435329345571256e-06, "loss": 0.0156, "step": 5530 }, { "epoch": 0.5645500286605949, "grad_norm": 0.038330078125, "learning_rate": 8.402737610512484e-06, "loss": 0.0151, "step": 5540 }, { "epoch": 0.5655690720336284, "grad_norm": 0.05078125, "learning_rate": 8.370163276854737e-06, "loss": 0.015, "step": 5550 }, { "epoch": 0.566588115406662, "grad_norm": 0.05908203125, "learning_rate": 8.337606699479624e-06, "loss": 0.0177, "step": 5560 }, { "epoch": 0.5676071587796956, "grad_norm": 0.039794921875, "learning_rate": 8.3050682330753e-06, "loss": 0.0131, "step": 5570 }, { "epoch": 0.5686262021527291, "grad_norm": 0.04541015625, "learning_rate": 8.272548232132619e-06, "loss": 0.0169, "step": 5580 }, { "epoch": 0.5696452455257627, "grad_norm": 0.043701171875, "learning_rate": 8.240047050941257e-06, "loss": 0.0164, "step": 5590 }, { "epoch": 0.5706642888987963, "grad_norm": 0.0849609375, "learning_rate": 8.207565043585866e-06, "loss": 0.013, "step": 5600 }, { "epoch": 0.5716833322718298, "grad_norm": 0.033447265625, "learning_rate": 8.17510256394219e-06, "loss": 0.0148, "step": 5610 }, { "epoch": 0.5727023756448634, "grad_norm": 0.0830078125, "learning_rate": 8.142659965673256e-06, "loss": 0.0146, "step": 5620 }, { "epoch": 0.573721419017897, "grad_norm": 0.047607421875, "learning_rate": 8.110237602225464e-06, "loss": 0.0124, "step": 5630 }, { "epoch": 0.5747404623909305, "grad_norm": 0.034423828125, "learning_rate": 8.077835826824786e-06, "loss": 0.0155, "step": 5640 }, { "epoch": 0.5757595057639641, "grad_norm": 0.04541015625, "learning_rate": 8.04545499247289e-06, "loss": 0.0123, "step": 5650 }, { "epoch": 0.5767785491369977, "grad_norm": 0.03515625, "learning_rate": 8.013095451943297e-06, "loss": 0.0161, "step": 5660 }, { "epoch": 0.5777975925100312, "grad_norm": 0.05517578125, "learning_rate": 7.980757557777549e-06, "loss": 0.0141, "step": 5670 }, { "epoch": 0.5788166358830648, "grad_norm": 0.07763671875, "learning_rate": 7.948441662281359e-06, "loss": 0.0137, "step": 5680 }, { "epoch": 0.5798356792560984, "grad_norm": 0.06298828125, "learning_rate": 7.916148117520776e-06, "loss": 0.0154, "step": 5690 }, { "epoch": 0.5808547226291318, "grad_norm": 0.037841796875, "learning_rate": 7.883877275318345e-06, "loss": 0.0129, "step": 5700 }, { "epoch": 0.5818737660021654, "grad_norm": 0.08935546875, "learning_rate": 7.851629487249281e-06, "loss": 0.0163, "step": 5710 }, { "epoch": 0.582892809375199, "grad_norm": 0.050048828125, "learning_rate": 7.819405104637634e-06, "loss": 0.014, "step": 5720 }, { "epoch": 0.5839118527482327, "grad_norm": 0.0634765625, "learning_rate": 7.787204478552468e-06, "loss": 0.0134, "step": 5730 }, { "epoch": 0.5849308961212661, "grad_norm": 0.060546875, "learning_rate": 7.755027959804019e-06, "loss": 0.0165, "step": 5740 }, { "epoch": 0.5859499394942997, "grad_norm": 0.06298828125, "learning_rate": 7.722875898939897e-06, "loss": 0.0126, "step": 5750 }, { "epoch": 0.5869689828673333, "grad_norm": 0.09423828125, "learning_rate": 7.690748646241248e-06, "loss": 0.0146, "step": 5760 }, { "epoch": 0.5879880262403668, "grad_norm": 0.04052734375, "learning_rate": 7.658646551718953e-06, "loss": 0.0153, "step": 5770 }, { "epoch": 0.5890070696134004, "grad_norm": 0.047119140625, "learning_rate": 7.626569965109798e-06, "loss": 0.0128, "step": 5780 }, { "epoch": 0.590026112986434, "grad_norm": 0.0654296875, "learning_rate": 7.59451923587268e-06, "loss": 0.0149, "step": 5790 }, { "epoch": 0.5910451563594675, "grad_norm": 0.06982421875, "learning_rate": 7.5624947131847826e-06, "loss": 0.0128, "step": 5800 }, { "epoch": 0.5920641997325011, "grad_norm": 0.049072265625, "learning_rate": 7.530496745937793e-06, "loss": 0.0176, "step": 5810 }, { "epoch": 0.5930832431055347, "grad_norm": 0.040283203125, "learning_rate": 7.498525682734081e-06, "loss": 0.0142, "step": 5820 }, { "epoch": 0.5941022864785682, "grad_norm": 0.03564453125, "learning_rate": 7.466581871882921e-06, "loss": 0.0149, "step": 5830 }, { "epoch": 0.5951213298516018, "grad_norm": 0.045654296875, "learning_rate": 7.434665661396667e-06, "loss": 0.0145, "step": 5840 }, { "epoch": 0.5961403732246354, "grad_norm": 0.051025390625, "learning_rate": 7.402777398987002e-06, "loss": 0.0114, "step": 5850 }, { "epoch": 0.5971594165976689, "grad_norm": 0.0634765625, "learning_rate": 7.370917432061113e-06, "loss": 0.0144, "step": 5860 }, { "epoch": 0.5981784599707025, "grad_norm": 0.041748046875, "learning_rate": 7.3390861077179335e-06, "loss": 0.0136, "step": 5870 }, { "epoch": 0.5991975033437361, "grad_norm": 0.052490234375, "learning_rate": 7.307283772744334e-06, "loss": 0.0133, "step": 5880 }, { "epoch": 0.6002165467167696, "grad_norm": 0.046875, "learning_rate": 7.275510773611374e-06, "loss": 0.0137, "step": 5890 }, { "epoch": 0.6012355900898032, "grad_norm": 0.0693359375, "learning_rate": 7.243767456470505e-06, "loss": 0.0113, "step": 5900 }, { "epoch": 0.6022546334628368, "grad_norm": 0.041748046875, "learning_rate": 7.21205416714982e-06, "loss": 0.0149, "step": 5910 }, { "epoch": 0.6032736768358703, "grad_norm": 0.06591796875, "learning_rate": 7.180371251150249e-06, "loss": 0.0132, "step": 5920 }, { "epoch": 0.6042927202089039, "grad_norm": 0.046630859375, "learning_rate": 7.1487190536418505e-06, "loss": 0.0137, "step": 5930 }, { "epoch": 0.6053117635819375, "grad_norm": 0.036376953125, "learning_rate": 7.117097919459992e-06, "loss": 0.0143, "step": 5940 }, { "epoch": 0.606330806954971, "grad_norm": 0.0927734375, "learning_rate": 7.085508193101639e-06, "loss": 0.0129, "step": 5950 }, { "epoch": 0.6073498503280046, "grad_norm": 0.0625, "learning_rate": 7.053950218721575e-06, "loss": 0.0173, "step": 5960 }, { "epoch": 0.6083688937010382, "grad_norm": 0.03271484375, "learning_rate": 7.02242434012867e-06, "loss": 0.0142, "step": 5970 }, { "epoch": 0.6093879370740717, "grad_norm": 0.061279296875, "learning_rate": 6.990930900782112e-06, "loss": 0.0138, "step": 5980 }, { "epoch": 0.6104069804471053, "grad_norm": 0.047607421875, "learning_rate": 6.95947024378769e-06, "loss": 0.0145, "step": 5990 }, { "epoch": 0.6114260238201389, "grad_norm": 0.05224609375, "learning_rate": 6.928042711894047e-06, "loss": 0.0122, "step": 6000 } ], "logging_steps": 10, "max_steps": 9813, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.8788170935536976e+19, "train_batch_size": 3, "trial_name": null, "trial_params": null }