diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,17933 @@ +{ + "best_global_step": 63686, + "best_metric": 3.054766364990921, + "best_model_checkpoint": "outputs/whisper-large-v3-turbo/checkpoint-63686", + "epoch": 2.0, + "eval_steps": 500, + "global_step": 63686, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 3.1404569364842585e-05, + "grad_norm": Infinity, + "learning_rate": 0.0, + "loss": 9.979303359985352, + "step": 1 + }, + { + "epoch": 0.0007851142341210646, + "grad_norm": 41.94096374511719, + "learning_rate": 9.600000000000001e-07, + "loss": 9.373766581217447, + "step": 25 + }, + { + "epoch": 0.0015702284682421292, + "grad_norm": 19.067081451416016, + "learning_rate": 1.9600000000000003e-06, + "loss": 7.542354125976562, + "step": 50 + }, + { + "epoch": 0.002355342702363194, + "grad_norm": 20.59769058227539, + "learning_rate": 2.96e-06, + "loss": 7.009388427734375, + "step": 75 + }, + { + "epoch": 0.0031404569364842584, + "grad_norm": 22.558021545410156, + "learning_rate": 3.96e-06, + "loss": 6.90596435546875, + "step": 100 + }, + { + "epoch": 0.003925571170605323, + "grad_norm": 25.78112030029297, + "learning_rate": 4.960000000000001e-06, + "loss": 6.8768798828125, + "step": 125 + }, + { + "epoch": 0.004710685404726388, + "grad_norm": 20.777931213378906, + "learning_rate": 5.9600000000000005e-06, + "loss": 6.791765747070312, + "step": 150 + }, + { + "epoch": 0.005495799638847452, + "grad_norm": 20.44936752319336, + "learning_rate": 6.96e-06, + "loss": 6.7693017578125, + "step": 175 + }, + { + "epoch": 0.006280913872968517, + "grad_norm": 19.20271873474121, + "learning_rate": 7.960000000000002e-06, + "loss": 6.772525634765625, + "step": 200 + }, + { + "epoch": 0.007066028107089581, + "grad_norm": 18.812902450561523, + "learning_rate": 8.96e-06, + "loss": 6.783408203125, + "step": 225 + }, + { + "epoch": 0.007851142341210647, + "grad_norm": 19.372535705566406, + "learning_rate": 9.960000000000001e-06, + "loss": 6.770665893554687, + "step": 250 + }, + { + "epoch": 0.008636256575331711, + "grad_norm": 17.4212589263916, + "learning_rate": 1.0960000000000002e-05, + "loss": 6.724260864257812, + "step": 275 + }, + { + "epoch": 0.009421370809452776, + "grad_norm": 16.882524490356445, + "learning_rate": 1.196e-05, + "loss": 6.825057373046875, + "step": 300 + }, + { + "epoch": 0.01020648504357384, + "grad_norm": 15.996599197387695, + "learning_rate": 1.2960000000000001e-05, + "loss": 6.773802490234375, + "step": 325 + }, + { + "epoch": 0.010991599277694905, + "grad_norm": 18.903011322021484, + "learning_rate": 1.396e-05, + "loss": 6.7702783203125, + "step": 350 + }, + { + "epoch": 0.01177671351181597, + "grad_norm": 20.501256942749023, + "learning_rate": 1.496e-05, + "loss": 6.7285888671875, + "step": 375 + }, + { + "epoch": 0.012561827745937034, + "grad_norm": 18.055997848510742, + "learning_rate": 1.5960000000000003e-05, + "loss": 6.718192749023437, + "step": 400 + }, + { + "epoch": 0.013346941980058098, + "grad_norm": 17.99781036376953, + "learning_rate": 1.696e-05, + "loss": 6.723134765625, + "step": 425 + }, + { + "epoch": 0.014132056214179163, + "grad_norm": 18.030330657958984, + "learning_rate": 1.796e-05, + "loss": 6.779344482421875, + "step": 450 + }, + { + "epoch": 0.014917170448300227, + "grad_norm": 15.898109436035156, + "learning_rate": 1.896e-05, + "loss": 6.760403442382812, + "step": 475 + }, + { + "epoch": 0.015702284682421293, + "grad_norm": 17.55674934387207, + "learning_rate": 1.9960000000000002e-05, + "loss": 6.7034454345703125, + "step": 500 + }, + { + "epoch": 0.016487398916542356, + "grad_norm": 20.509431838989258, + "learning_rate": 1.9999993592435822e-05, + "loss": 6.7798583984375, + "step": 525 + }, + { + "epoch": 0.017272513150663422, + "grad_norm": 19.132766723632812, + "learning_rate": 1.9999973290701726e-05, + "loss": 6.757376708984375, + "step": 550 + }, + { + "epoch": 0.018057627384784485, + "grad_norm": 19.585542678833008, + "learning_rate": 1.999993908370483e-05, + "loss": 6.733961181640625, + "step": 575 + }, + { + "epoch": 0.01884274161890555, + "grad_norm": 17.031105041503906, + "learning_rate": 1.9999890971497985e-05, + "loss": 6.705369873046875, + "step": 600 + }, + { + "epoch": 0.019627855853026614, + "grad_norm": 14.040271759033203, + "learning_rate": 1.9999828954155528e-05, + "loss": 6.631074829101562, + "step": 625 + }, + { + "epoch": 0.02041297008714768, + "grad_norm": 16.218955993652344, + "learning_rate": 1.999975303177328e-05, + "loss": 6.6722607421875, + "step": 650 + }, + { + "epoch": 0.021198084321268743, + "grad_norm": 14.040811538696289, + "learning_rate": 1.999966320446854e-05, + "loss": 6.6358837890625, + "step": 675 + }, + { + "epoch": 0.02198319855538981, + "grad_norm": 15.409050941467285, + "learning_rate": 1.9999559472380092e-05, + "loss": 6.658862915039062, + "step": 700 + }, + { + "epoch": 0.022768312789510872, + "grad_norm": 17.76511573791504, + "learning_rate": 1.999944183566821e-05, + "loss": 6.6335601806640625, + "step": 725 + }, + { + "epoch": 0.02355342702363194, + "grad_norm": 14.490168571472168, + "learning_rate": 1.9999310294514647e-05, + "loss": 6.631829833984375, + "step": 750 + }, + { + "epoch": 0.024338541257753005, + "grad_norm": 17.15380859375, + "learning_rate": 1.999916484912264e-05, + "loss": 6.634891967773438, + "step": 775 + }, + { + "epoch": 0.025123655491874067, + "grad_norm": 15.58073902130127, + "learning_rate": 1.9999005499716902e-05, + "loss": 6.613302612304688, + "step": 800 + }, + { + "epoch": 0.025908769725995134, + "grad_norm": 14.687374114990234, + "learning_rate": 1.9998832246543637e-05, + "loss": 6.563751831054687, + "step": 825 + }, + { + "epoch": 0.026693883960116196, + "grad_norm": 17.44748878479004, + "learning_rate": 1.999864508987052e-05, + "loss": 6.596546630859375, + "step": 850 + }, + { + "epoch": 0.027478998194237263, + "grad_norm": 12.804855346679688, + "learning_rate": 1.9998444029986727e-05, + "loss": 6.5818548583984375, + "step": 875 + }, + { + "epoch": 0.028264112428358325, + "grad_norm": 17.41668701171875, + "learning_rate": 1.999822906720289e-05, + "loss": 6.559022827148437, + "step": 900 + }, + { + "epoch": 0.02904922666247939, + "grad_norm": 12.885072708129883, + "learning_rate": 1.9998000201851144e-05, + "loss": 6.571969604492187, + "step": 925 + }, + { + "epoch": 0.029834340896600454, + "grad_norm": 14.439318656921387, + "learning_rate": 1.9997757434285084e-05, + "loss": 6.560151977539062, + "step": 950 + }, + { + "epoch": 0.03061945513072152, + "grad_norm": 15.08297348022461, + "learning_rate": 1.99975007648798e-05, + "loss": 6.557699584960938, + "step": 975 + }, + { + "epoch": 0.03140456936484259, + "grad_norm": 14.325587272644043, + "learning_rate": 1.9997230194031854e-05, + "loss": 6.520199584960937, + "step": 1000 + }, + { + "epoch": 0.032189683598963646, + "grad_norm": 13.04557991027832, + "learning_rate": 1.999694572215929e-05, + "loss": 6.5418359375, + "step": 1025 + }, + { + "epoch": 0.03297479783308471, + "grad_norm": 14.734786033630371, + "learning_rate": 1.9996647349701618e-05, + "loss": 6.519505615234375, + "step": 1050 + }, + { + "epoch": 0.03375991206720578, + "grad_norm": 11.710091590881348, + "learning_rate": 1.999633507711984e-05, + "loss": 6.452557983398438, + "step": 1075 + }, + { + "epoch": 0.034545026301326845, + "grad_norm": 11.891339302062988, + "learning_rate": 1.9996008904896426e-05, + "loss": 6.508645629882812, + "step": 1100 + }, + { + "epoch": 0.035330140535447904, + "grad_norm": 13.234967231750488, + "learning_rate": 1.9995668833535326e-05, + "loss": 6.534010009765625, + "step": 1125 + }, + { + "epoch": 0.03611525476956897, + "grad_norm": 13.071805000305176, + "learning_rate": 1.999531486356196e-05, + "loss": 6.49391357421875, + "step": 1150 + }, + { + "epoch": 0.03690036900369004, + "grad_norm": 11.06236457824707, + "learning_rate": 1.999494699552322e-05, + "loss": 6.463763427734375, + "step": 1175 + }, + { + "epoch": 0.0376854832378111, + "grad_norm": 14.228161811828613, + "learning_rate": 1.999456522998748e-05, + "loss": 6.483692016601562, + "step": 1200 + }, + { + "epoch": 0.03847059747193217, + "grad_norm": 16.29725456237793, + "learning_rate": 1.9994169567544578e-05, + "loss": 6.500346069335937, + "step": 1225 + }, + { + "epoch": 0.03925571170605323, + "grad_norm": 10.389582633972168, + "learning_rate": 1.9993760008805824e-05, + "loss": 6.461927490234375, + "step": 1250 + }, + { + "epoch": 0.040040825940174295, + "grad_norm": 10.62636661529541, + "learning_rate": 1.9993336554404005e-05, + "loss": 6.465281372070312, + "step": 1275 + }, + { + "epoch": 0.04082594017429536, + "grad_norm": 14.631693840026855, + "learning_rate": 1.9992899204993364e-05, + "loss": 6.455181274414063, + "step": 1300 + }, + { + "epoch": 0.04161105440841643, + "grad_norm": 12.233256340026855, + "learning_rate": 1.999244796124963e-05, + "loss": 6.475103149414062, + "step": 1325 + }, + { + "epoch": 0.042396168642537486, + "grad_norm": 14.816092491149902, + "learning_rate": 1.999198282386999e-05, + "loss": 6.477763671875, + "step": 1350 + }, + { + "epoch": 0.04318128287665855, + "grad_norm": 13.356731414794922, + "learning_rate": 1.999150379357309e-05, + "loss": 6.436436767578125, + "step": 1375 + }, + { + "epoch": 0.04396639711077962, + "grad_norm": 9.61710262298584, + "learning_rate": 1.999101087109905e-05, + "loss": 6.443475952148438, + "step": 1400 + }, + { + "epoch": 0.044751511344900685, + "grad_norm": 13.607565879821777, + "learning_rate": 1.999050405720945e-05, + "loss": 6.40984130859375, + "step": 1425 + }, + { + "epoch": 0.045536625579021744, + "grad_norm": 12.835264205932617, + "learning_rate": 1.998998335268734e-05, + "loss": 6.450366821289062, + "step": 1450 + }, + { + "epoch": 0.04632173981314281, + "grad_norm": 9.50557804107666, + "learning_rate": 1.9989448758337227e-05, + "loss": 6.404087524414063, + "step": 1475 + }, + { + "epoch": 0.04710685404726388, + "grad_norm": 10.875224113464355, + "learning_rate": 1.998890027498507e-05, + "loss": 6.4033935546875, + "step": 1500 + }, + { + "epoch": 0.04789196828138494, + "grad_norm": 11.079575538635254, + "learning_rate": 1.9988337903478295e-05, + "loss": 6.39769287109375, + "step": 1525 + }, + { + "epoch": 0.04867708251550601, + "grad_norm": 12.025618553161621, + "learning_rate": 1.9987761644685794e-05, + "loss": 6.419639282226562, + "step": 1550 + }, + { + "epoch": 0.04946219674962707, + "grad_norm": 10.820062637329102, + "learning_rate": 1.99871714994979e-05, + "loss": 6.41373046875, + "step": 1575 + }, + { + "epoch": 0.050247310983748135, + "grad_norm": 11.888032913208008, + "learning_rate": 1.9986567468826403e-05, + "loss": 6.375022583007812, + "step": 1600 + }, + { + "epoch": 0.0510324252178692, + "grad_norm": 12.498847961425781, + "learning_rate": 1.9985949553604558e-05, + "loss": 6.412108154296875, + "step": 1625 + }, + { + "epoch": 0.05181753945199027, + "grad_norm": 12.67115306854248, + "learning_rate": 1.9985317754787064e-05, + "loss": 6.412750244140625, + "step": 1650 + }, + { + "epoch": 0.05260265368611133, + "grad_norm": 11.6309814453125, + "learning_rate": 1.998467207335007e-05, + "loss": 6.400971069335937, + "step": 1675 + }, + { + "epoch": 0.05338776792023239, + "grad_norm": 10.761411666870117, + "learning_rate": 1.998401251029117e-05, + "loss": 6.38491455078125, + "step": 1700 + }, + { + "epoch": 0.05417288215435346, + "grad_norm": 12.329826354980469, + "learning_rate": 1.998333906662942e-05, + "loss": 6.389588623046875, + "step": 1725 + }, + { + "epoch": 0.054957996388474525, + "grad_norm": 9.342223167419434, + "learning_rate": 1.9982651743405314e-05, + "loss": 6.387445678710938, + "step": 1750 + }, + { + "epoch": 0.055743110622595585, + "grad_norm": 12.338208198547363, + "learning_rate": 1.9981950541680778e-05, + "loss": 6.388828125, + "step": 1775 + }, + { + "epoch": 0.05652822485671665, + "grad_norm": 11.37130069732666, + "learning_rate": 1.9981235462539204e-05, + "loss": 6.383390502929688, + "step": 1800 + }, + { + "epoch": 0.05731333909083772, + "grad_norm": 11.251113891601562, + "learning_rate": 1.9980506507085408e-05, + "loss": 6.395357055664062, + "step": 1825 + }, + { + "epoch": 0.05809845332495878, + "grad_norm": 10.904779434204102, + "learning_rate": 1.9979763676445655e-05, + "loss": 6.36266357421875, + "step": 1850 + }, + { + "epoch": 0.05888356755907984, + "grad_norm": 13.056145668029785, + "learning_rate": 1.9979006971767635e-05, + "loss": 6.405393676757813, + "step": 1875 + }, + { + "epoch": 0.05966868179320091, + "grad_norm": 10.779613494873047, + "learning_rate": 1.997823639422049e-05, + "loss": 6.346578369140625, + "step": 1900 + }, + { + "epoch": 0.060453796027321975, + "grad_norm": 11.322759628295898, + "learning_rate": 1.997745194499479e-05, + "loss": 6.344718627929687, + "step": 1925 + }, + { + "epoch": 0.06123891026144304, + "grad_norm": 10.215258598327637, + "learning_rate": 1.997665362530253e-05, + "loss": 6.35252197265625, + "step": 1950 + }, + { + "epoch": 0.06202402449556411, + "grad_norm": 11.057332038879395, + "learning_rate": 1.9975841436377142e-05, + "loss": 6.368969116210938, + "step": 1975 + }, + { + "epoch": 0.06280913872968517, + "grad_norm": 11.349760055541992, + "learning_rate": 1.9975015379473483e-05, + "loss": 6.363984849717882, + "step": 2000 + }, + { + "epoch": 0.06359425296380623, + "grad_norm": 11.43038558959961, + "learning_rate": 1.9974175455867844e-05, + "loss": 6.317130126953125, + "step": 2025 + }, + { + "epoch": 0.06437936719792729, + "grad_norm": 10.505550384521484, + "learning_rate": 1.9973321666857935e-05, + "loss": 6.35501953125, + "step": 2050 + }, + { + "epoch": 0.06516448143204837, + "grad_norm": 10.38772201538086, + "learning_rate": 1.9972454013762886e-05, + "loss": 6.336322631835937, + "step": 2075 + }, + { + "epoch": 0.06594959566616942, + "grad_norm": 11.543457984924316, + "learning_rate": 1.997157249792325e-05, + "loss": 6.335061645507812, + "step": 2100 + }, + { + "epoch": 0.0667347099002905, + "grad_norm": 9.45538330078125, + "learning_rate": 1.9970677120701e-05, + "loss": 6.376185302734375, + "step": 2125 + }, + { + "epoch": 0.06751982413441156, + "grad_norm": 10.921442985534668, + "learning_rate": 1.9969767883479524e-05, + "loss": 6.338211669921875, + "step": 2150 + }, + { + "epoch": 0.06830493836853262, + "grad_norm": 8.948094367980957, + "learning_rate": 1.9968844787663627e-05, + "loss": 6.331012573242187, + "step": 2175 + }, + { + "epoch": 0.06909005260265369, + "grad_norm": 10.997818946838379, + "learning_rate": 1.996790783467952e-05, + "loss": 6.336534423828125, + "step": 2200 + }, + { + "epoch": 0.06987516683677475, + "grad_norm": 12.306873321533203, + "learning_rate": 1.9966957025974824e-05, + "loss": 6.35581787109375, + "step": 2225 + }, + { + "epoch": 0.07066028107089581, + "grad_norm": 11.029871940612793, + "learning_rate": 1.9965992363018575e-05, + "loss": 6.323455810546875, + "step": 2250 + }, + { + "epoch": 0.07144539530501688, + "grad_norm": 10.700419425964355, + "learning_rate": 1.996501384730121e-05, + "loss": 6.380532836914062, + "step": 2275 + }, + { + "epoch": 0.07223050953913794, + "grad_norm": 10.489208221435547, + "learning_rate": 1.996402148033457e-05, + "loss": 6.333556518554688, + "step": 2300 + }, + { + "epoch": 0.07301562377325901, + "grad_norm": 11.169740676879883, + "learning_rate": 1.9963015263651893e-05, + "loss": 6.310536499023438, + "step": 2325 + }, + { + "epoch": 0.07380073800738007, + "grad_norm": 11.052619934082031, + "learning_rate": 1.9961995198807817e-05, + "loss": 6.325816040039062, + "step": 2350 + }, + { + "epoch": 0.07458585224150113, + "grad_norm": 9.334430694580078, + "learning_rate": 1.9960961287378378e-05, + "loss": 6.319947509765625, + "step": 2375 + }, + { + "epoch": 0.0753709664756222, + "grad_norm": 12.27335262298584, + "learning_rate": 1.9959913530961e-05, + "loss": 6.297763061523438, + "step": 2400 + }, + { + "epoch": 0.07615608070974327, + "grad_norm": 11.433621406555176, + "learning_rate": 1.9958851931174512e-05, + "loss": 6.300328369140625, + "step": 2425 + }, + { + "epoch": 0.07694119494386434, + "grad_norm": 10.371191024780273, + "learning_rate": 1.995777648965911e-05, + "loss": 6.347232666015625, + "step": 2450 + }, + { + "epoch": 0.0777263091779854, + "grad_norm": 10.479618072509766, + "learning_rate": 1.9956687208076397e-05, + "loss": 6.349697875976562, + "step": 2475 + }, + { + "epoch": 0.07851142341210646, + "grad_norm": 9.61593246459961, + "learning_rate": 1.9955584088109345e-05, + "loss": 6.279691162109375, + "step": 2500 + }, + { + "epoch": 0.07929653764622753, + "grad_norm": 9.81901741027832, + "learning_rate": 1.9954467131462312e-05, + "loss": 6.3111962890625, + "step": 2525 + }, + { + "epoch": 0.08008165188034859, + "grad_norm": 10.800393104553223, + "learning_rate": 1.9953336339861036e-05, + "loss": 6.311746826171875, + "step": 2550 + }, + { + "epoch": 0.08086676611446965, + "grad_norm": 13.160219192504883, + "learning_rate": 1.9952191715052624e-05, + "loss": 6.314322509765625, + "step": 2575 + }, + { + "epoch": 0.08165188034859072, + "grad_norm": 10.921112060546875, + "learning_rate": 1.995103325880556e-05, + "loss": 6.274252319335938, + "step": 2600 + }, + { + "epoch": 0.08243699458271178, + "grad_norm": 9.895350456237793, + "learning_rate": 1.9949860972909705e-05, + "loss": 6.302964477539063, + "step": 2625 + }, + { + "epoch": 0.08322210881683285, + "grad_norm": 10.363892555236816, + "learning_rate": 1.9948674859176272e-05, + "loss": 6.2783349609375, + "step": 2650 + }, + { + "epoch": 0.08400722305095391, + "grad_norm": 11.488625526428223, + "learning_rate": 1.9947474919437856e-05, + "loss": 6.29146484375, + "step": 2675 + }, + { + "epoch": 0.08479233728507497, + "grad_norm": 11.399894714355469, + "learning_rate": 1.9946261155548394e-05, + "loss": 6.30981689453125, + "step": 2700 + }, + { + "epoch": 0.08557745151919605, + "grad_norm": 10.254036903381348, + "learning_rate": 1.9945033569383197e-05, + "loss": 6.324871826171875, + "step": 2725 + }, + { + "epoch": 0.0863625657533171, + "grad_norm": 10.538107872009277, + "learning_rate": 1.9943792162838928e-05, + "loss": 6.3052740478515625, + "step": 2750 + }, + { + "epoch": 0.08714767998743818, + "grad_norm": 7.8543853759765625, + "learning_rate": 1.9942536937833605e-05, + "loss": 6.311158447265625, + "step": 2775 + }, + { + "epoch": 0.08793279422155924, + "grad_norm": 10.63546085357666, + "learning_rate": 1.9941267896306583e-05, + "loss": 6.249661254882812, + "step": 2800 + }, + { + "epoch": 0.0887179084556803, + "grad_norm": 9.314477920532227, + "learning_rate": 1.9939985040218586e-05, + "loss": 6.29000244140625, + "step": 2825 + }, + { + "epoch": 0.08950302268980137, + "grad_norm": 8.875669479370117, + "learning_rate": 1.9938688371551663e-05, + "loss": 6.249137573242187, + "step": 2850 + }, + { + "epoch": 0.09028813692392243, + "grad_norm": 11.796751022338867, + "learning_rate": 1.9937377892309213e-05, + "loss": 6.295302734375, + "step": 2875 + }, + { + "epoch": 0.09107325115804349, + "grad_norm": 10.234321594238281, + "learning_rate": 1.9936053604515968e-05, + "loss": 6.273316650390625, + "step": 2900 + }, + { + "epoch": 0.09185836539216456, + "grad_norm": 10.448747634887695, + "learning_rate": 1.9934715510218e-05, + "loss": 6.2974658203125, + "step": 2925 + }, + { + "epoch": 0.09264347962628562, + "grad_norm": 11.137585639953613, + "learning_rate": 1.993336361148271e-05, + "loss": 6.269514770507812, + "step": 2950 + }, + { + "epoch": 0.0934285938604067, + "grad_norm": 11.984169006347656, + "learning_rate": 1.9931997910398825e-05, + "loss": 6.2598028564453125, + "step": 2975 + }, + { + "epoch": 0.09421370809452775, + "grad_norm": 10.593313217163086, + "learning_rate": 1.99306184090764e-05, + "loss": 6.253529663085938, + "step": 3000 + }, + { + "epoch": 0.09499882232864881, + "grad_norm": 9.322269439697266, + "learning_rate": 1.9929225109646816e-05, + "loss": 6.296830444335938, + "step": 3025 + }, + { + "epoch": 0.09578393656276989, + "grad_norm": 8.563907623291016, + "learning_rate": 1.992781801426276e-05, + "loss": 6.24077392578125, + "step": 3050 + }, + { + "epoch": 0.09656905079689095, + "grad_norm": 9.148247718811035, + "learning_rate": 1.992639712509825e-05, + "loss": 6.276689453125, + "step": 3075 + }, + { + "epoch": 0.09735416503101202, + "grad_norm": 9.976004600524902, + "learning_rate": 1.9924962444348598e-05, + "loss": 6.268125610351563, + "step": 3100 + }, + { + "epoch": 0.09813927926513308, + "grad_norm": 9.7391357421875, + "learning_rate": 1.992351397423044e-05, + "loss": 6.252393188476563, + "step": 3125 + }, + { + "epoch": 0.09892439349925414, + "grad_norm": 11.162590980529785, + "learning_rate": 1.9922051716981717e-05, + "loss": 6.262725830078125, + "step": 3150 + }, + { + "epoch": 0.09970950773337521, + "grad_norm": 9.588018417358398, + "learning_rate": 1.9920575674861654e-05, + "loss": 6.244053344726563, + "step": 3175 + }, + { + "epoch": 0.10049462196749627, + "grad_norm": 10.21446704864502, + "learning_rate": 1.991908585015079e-05, + "loss": 6.255555419921875, + "step": 3200 + }, + { + "epoch": 0.10127973620161733, + "grad_norm": 10.891797065734863, + "learning_rate": 1.991758224515096e-05, + "loss": 6.2576171875, + "step": 3225 + }, + { + "epoch": 0.1020648504357384, + "grad_norm": 8.930691719055176, + "learning_rate": 1.991606486218528e-05, + "loss": 6.2703399658203125, + "step": 3250 + }, + { + "epoch": 0.10284996466985946, + "grad_norm": 10.67702579498291, + "learning_rate": 1.991453370359816e-05, + "loss": 6.241373901367187, + "step": 3275 + }, + { + "epoch": 0.10363507890398053, + "grad_norm": 8.414963722229004, + "learning_rate": 1.9912988771755287e-05, + "loss": 6.263267211914062, + "step": 3300 + }, + { + "epoch": 0.1044201931381016, + "grad_norm": 10.072351455688477, + "learning_rate": 1.991143006904364e-05, + "loss": 6.26646728515625, + "step": 3325 + }, + { + "epoch": 0.10520530737222265, + "grad_norm": 11.865885734558105, + "learning_rate": 1.990985759787146e-05, + "loss": 6.241846313476563, + "step": 3350 + }, + { + "epoch": 0.10599042160634373, + "grad_norm": 10.353376388549805, + "learning_rate": 1.9908271360668273e-05, + "loss": 6.28155517578125, + "step": 3375 + }, + { + "epoch": 0.10677553584046479, + "grad_norm": 13.077787399291992, + "learning_rate": 1.990667135988487e-05, + "loss": 6.237803955078125, + "step": 3400 + }, + { + "epoch": 0.10756065007458586, + "grad_norm": 11.286938667297363, + "learning_rate": 1.9905057597993303e-05, + "loss": 6.243350830078125, + "step": 3425 + }, + { + "epoch": 0.10834576430870692, + "grad_norm": 10.396014213562012, + "learning_rate": 1.9903430077486884e-05, + "loss": 6.2287548828125, + "step": 3450 + }, + { + "epoch": 0.10913087854282798, + "grad_norm": 8.9232816696167, + "learning_rate": 1.99017888008802e-05, + "loss": 6.235592041015625, + "step": 3475 + }, + { + "epoch": 0.10991599277694905, + "grad_norm": 8.340466499328613, + "learning_rate": 1.9900133770709068e-05, + "loss": 6.240104370117187, + "step": 3500 + }, + { + "epoch": 0.11070110701107011, + "grad_norm": 8.6317138671875, + "learning_rate": 1.989846498953057e-05, + "loss": 6.2185791015625, + "step": 3525 + }, + { + "epoch": 0.11148622124519117, + "grad_norm": 10.392827987670898, + "learning_rate": 1.989678245992303e-05, + "loss": 6.248939819335938, + "step": 3550 + }, + { + "epoch": 0.11227133547931224, + "grad_norm": 11.138606071472168, + "learning_rate": 1.989508618448601e-05, + "loss": 6.241966552734375, + "step": 3575 + }, + { + "epoch": 0.1130564497134333, + "grad_norm": 9.577104568481445, + "learning_rate": 1.9893376165840316e-05, + "loss": 6.2293634033203125, + "step": 3600 + }, + { + "epoch": 0.11384156394755438, + "grad_norm": 9.183435440063477, + "learning_rate": 1.989165240662798e-05, + "loss": 6.199210205078125, + "step": 3625 + }, + { + "epoch": 0.11462667818167543, + "grad_norm": 10.918403625488281, + "learning_rate": 1.9889914909512275e-05, + "loss": 6.210628662109375, + "step": 3650 + }, + { + "epoch": 0.1154117924157965, + "grad_norm": 10.935476303100586, + "learning_rate": 1.9888163677177687e-05, + "loss": 6.2219482421875, + "step": 3675 + }, + { + "epoch": 0.11619690664991757, + "grad_norm": 8.196861267089844, + "learning_rate": 1.9886398712329932e-05, + "loss": 6.246177368164062, + "step": 3700 + }, + { + "epoch": 0.11698202088403863, + "grad_norm": 10.488009452819824, + "learning_rate": 1.9884620017695935e-05, + "loss": 6.244432983398437, + "step": 3725 + }, + { + "epoch": 0.11776713511815969, + "grad_norm": 10.21810531616211, + "learning_rate": 1.9882827596023847e-05, + "loss": 6.220256958007813, + "step": 3750 + }, + { + "epoch": 0.11855224935228076, + "grad_norm": 11.324859619140625, + "learning_rate": 1.988102145008302e-05, + "loss": 6.214730834960937, + "step": 3775 + }, + { + "epoch": 0.11933736358640182, + "grad_norm": 8.648311614990234, + "learning_rate": 1.9879201582664e-05, + "loss": 6.231104125976563, + "step": 3800 + }, + { + "epoch": 0.12012247782052289, + "grad_norm": 10.830723762512207, + "learning_rate": 1.9877367996578556e-05, + "loss": 6.214857177734375, + "step": 3825 + }, + { + "epoch": 0.12090759205464395, + "grad_norm": 9.74167251586914, + "learning_rate": 1.9875520694659636e-05, + "loss": 6.22563232421875, + "step": 3850 + }, + { + "epoch": 0.12169270628876501, + "grad_norm": 10.081503868103027, + "learning_rate": 1.9873659679761383e-05, + "loss": 6.22583740234375, + "step": 3875 + }, + { + "epoch": 0.12247782052288608, + "grad_norm": 8.374776840209961, + "learning_rate": 1.9871784954759124e-05, + "loss": 6.203502197265625, + "step": 3900 + }, + { + "epoch": 0.12326293475700714, + "grad_norm": 9.443041801452637, + "learning_rate": 1.986989652254938e-05, + "loss": 6.21442626953125, + "step": 3925 + }, + { + "epoch": 0.12404804899112822, + "grad_norm": 8.679952621459961, + "learning_rate": 1.9867994386049842e-05, + "loss": 6.2056646728515625, + "step": 3950 + }, + { + "epoch": 0.12483316322524927, + "grad_norm": 10.583311080932617, + "learning_rate": 1.9866078548199364e-05, + "loss": 6.223202514648437, + "step": 3975 + }, + { + "epoch": 0.12561827745937035, + "grad_norm": 9.005955696105957, + "learning_rate": 1.9864149011957994e-05, + "loss": 6.213162841796875, + "step": 4000 + }, + { + "epoch": 0.1264033916934914, + "grad_norm": 11.103883743286133, + "learning_rate": 1.9862205780306922e-05, + "loss": 6.228407592773437, + "step": 4025 + }, + { + "epoch": 0.12718850592761247, + "grad_norm": 10.70343017578125, + "learning_rate": 1.9860248856248515e-05, + "loss": 6.228624877929687, + "step": 4050 + }, + { + "epoch": 0.12797362016173353, + "grad_norm": 9.76686954498291, + "learning_rate": 1.9858278242806276e-05, + "loss": 6.212389526367187, + "step": 4075 + }, + { + "epoch": 0.12875873439585458, + "grad_norm": 7.485008716583252, + "learning_rate": 1.985629394302488e-05, + "loss": 6.188821411132812, + "step": 4100 + }, + { + "epoch": 0.12954384862997567, + "grad_norm": 9.819150924682617, + "learning_rate": 1.985429595997013e-05, + "loss": 6.224051513671875, + "step": 4125 + }, + { + "epoch": 0.13032896286409673, + "grad_norm": 10.67740249633789, + "learning_rate": 1.985228429672898e-05, + "loss": 6.2152862548828125, + "step": 4150 + }, + { + "epoch": 0.1311140770982178, + "grad_norm": 10.333677291870117, + "learning_rate": 1.9850258956409513e-05, + "loss": 6.239793701171875, + "step": 4175 + }, + { + "epoch": 0.13189919133233885, + "grad_norm": 9.053677558898926, + "learning_rate": 1.9848219942140957e-05, + "loss": 6.187979736328125, + "step": 4200 + }, + { + "epoch": 0.1326843055664599, + "grad_norm": 9.030153274536133, + "learning_rate": 1.9846167257073652e-05, + "loss": 6.214325561523437, + "step": 4225 + }, + { + "epoch": 0.133469419800581, + "grad_norm": 11.282963752746582, + "learning_rate": 1.9844100904379065e-05, + "loss": 6.191294555664062, + "step": 4250 + }, + { + "epoch": 0.13425453403470206, + "grad_norm": 9.737730026245117, + "learning_rate": 1.9842020887249786e-05, + "loss": 6.213854370117187, + "step": 4275 + }, + { + "epoch": 0.13503964826882311, + "grad_norm": 8.100914001464844, + "learning_rate": 1.9839927208899504e-05, + "loss": 6.186459350585937, + "step": 4300 + }, + { + "epoch": 0.13582476250294417, + "grad_norm": 9.341636657714844, + "learning_rate": 1.9837819872563032e-05, + "loss": 6.172025756835938, + "step": 4325 + }, + { + "epoch": 0.13660987673706523, + "grad_norm": 8.412219047546387, + "learning_rate": 1.9835698881496267e-05, + "loss": 6.185524291992188, + "step": 4350 + }, + { + "epoch": 0.13739499097118632, + "grad_norm": 7.95367956161499, + "learning_rate": 1.983356423897622e-05, + "loss": 6.205936279296875, + "step": 4375 + }, + { + "epoch": 0.13818010520530738, + "grad_norm": 10.064981460571289, + "learning_rate": 1.983141594830098e-05, + "loss": 6.189014282226562, + "step": 4400 + }, + { + "epoch": 0.13896521943942844, + "grad_norm": 6.3258843421936035, + "learning_rate": 1.982925401278973e-05, + "loss": 6.1766455078125, + "step": 4425 + }, + { + "epoch": 0.1397503336735495, + "grad_norm": 10.785058975219727, + "learning_rate": 1.9827078435782732e-05, + "loss": 6.182383422851562, + "step": 4450 + }, + { + "epoch": 0.14053544790767056, + "grad_norm": 7.382954120635986, + "learning_rate": 1.9824889220641333e-05, + "loss": 6.17932373046875, + "step": 4475 + }, + { + "epoch": 0.14132056214179162, + "grad_norm": 10.448567390441895, + "learning_rate": 1.9822686370747935e-05, + "loss": 6.185806884765625, + "step": 4500 + }, + { + "epoch": 0.1421056763759127, + "grad_norm": 12.225725173950195, + "learning_rate": 1.9820469889506018e-05, + "loss": 6.211871337890625, + "step": 4525 + }, + { + "epoch": 0.14289079061003376, + "grad_norm": 10.037734031677246, + "learning_rate": 1.9818239780340127e-05, + "loss": 6.181618041992188, + "step": 4550 + }, + { + "epoch": 0.14367590484415482, + "grad_norm": 10.706823348999023, + "learning_rate": 1.9815996046695846e-05, + "loss": 6.191204833984375, + "step": 4575 + }, + { + "epoch": 0.14446101907827588, + "grad_norm": 9.014902114868164, + "learning_rate": 1.9813738692039827e-05, + "loss": 6.179193115234375, + "step": 4600 + }, + { + "epoch": 0.14524613331239694, + "grad_norm": 11.926057815551758, + "learning_rate": 1.9811467719859757e-05, + "loss": 6.181840209960938, + "step": 4625 + }, + { + "epoch": 0.14603124754651803, + "grad_norm": 9.946725845336914, + "learning_rate": 1.9809183133664367e-05, + "loss": 6.192357788085937, + "step": 4650 + }, + { + "epoch": 0.1468163617806391, + "grad_norm": 8.289819717407227, + "learning_rate": 1.9806884936983412e-05, + "loss": 6.194725341796875, + "step": 4675 + }, + { + "epoch": 0.14760147601476015, + "grad_norm": 6.509995937347412, + "learning_rate": 1.9804573133367692e-05, + "loss": 6.185567016601563, + "step": 4700 + }, + { + "epoch": 0.1483865902488812, + "grad_norm": 9.318564414978027, + "learning_rate": 1.9802247726389018e-05, + "loss": 6.179719848632812, + "step": 4725 + }, + { + "epoch": 0.14917170448300227, + "grad_norm": 11.016097068786621, + "learning_rate": 1.979990871964022e-05, + "loss": 6.17201904296875, + "step": 4750 + }, + { + "epoch": 0.14995681871712335, + "grad_norm": 8.966126441955566, + "learning_rate": 1.9797556116735148e-05, + "loss": 6.180454711914063, + "step": 4775 + }, + { + "epoch": 0.1507419329512444, + "grad_norm": 7.874191761016846, + "learning_rate": 1.979518992130865e-05, + "loss": 6.198030395507812, + "step": 4800 + }, + { + "epoch": 0.15152704718536547, + "grad_norm": 9.266504287719727, + "learning_rate": 1.9792810137016577e-05, + "loss": 6.1792059326171875, + "step": 4825 + }, + { + "epoch": 0.15231216141948653, + "grad_norm": 10.382789611816406, + "learning_rate": 1.9790416767535778e-05, + "loss": 6.19174560546875, + "step": 4850 + }, + { + "epoch": 0.1530972756536076, + "grad_norm": 8.12353229522705, + "learning_rate": 1.9788009816564087e-05, + "loss": 6.172281494140625, + "step": 4875 + }, + { + "epoch": 0.15388238988772868, + "grad_norm": 11.126033782958984, + "learning_rate": 1.9785589287820327e-05, + "loss": 6.175980834960938, + "step": 4900 + }, + { + "epoch": 0.15466750412184974, + "grad_norm": 9.489386558532715, + "learning_rate": 1.97831551850443e-05, + "loss": 6.1783819580078125, + "step": 4925 + }, + { + "epoch": 0.1554526183559708, + "grad_norm": 9.837382316589355, + "learning_rate": 1.9780707511996767e-05, + "loss": 6.17843017578125, + "step": 4950 + }, + { + "epoch": 0.15623773259009185, + "grad_norm": 9.826859474182129, + "learning_rate": 1.9778246272459476e-05, + "loss": 6.16966064453125, + "step": 4975 + }, + { + "epoch": 0.1570228468242129, + "grad_norm": 10.61966609954834, + "learning_rate": 1.977577147023512e-05, + "loss": 6.1547021484375, + "step": 5000 + }, + { + "epoch": 0.157807961058334, + "grad_norm": 8.017159461975098, + "learning_rate": 1.977328310914735e-05, + "loss": 6.16110107421875, + "step": 5025 + }, + { + "epoch": 0.15859307529245506, + "grad_norm": 9.731488227844238, + "learning_rate": 1.9770781193040776e-05, + "loss": 6.1703125, + "step": 5050 + }, + { + "epoch": 0.15937818952657612, + "grad_norm": 8.962721824645996, + "learning_rate": 1.9768265725780936e-05, + "loss": 6.16505615234375, + "step": 5075 + }, + { + "epoch": 0.16016330376069718, + "grad_norm": 9.964286804199219, + "learning_rate": 1.976573671125431e-05, + "loss": 6.175228881835937, + "step": 5100 + }, + { + "epoch": 0.16094841799481824, + "grad_norm": 8.709931373596191, + "learning_rate": 1.976319415336832e-05, + "loss": 6.1799755859375, + "step": 5125 + }, + { + "epoch": 0.1617335322289393, + "grad_norm": 8.342013359069824, + "learning_rate": 1.9760638056051293e-05, + "loss": 6.163645629882812, + "step": 5150 + }, + { + "epoch": 0.16251864646306038, + "grad_norm": 9.469314575195312, + "learning_rate": 1.9758068423252496e-05, + "loss": 6.188755493164063, + "step": 5175 + }, + { + "epoch": 0.16330376069718144, + "grad_norm": 6.644179344177246, + "learning_rate": 1.9755485258942092e-05, + "loss": 6.173428344726562, + "step": 5200 + }, + { + "epoch": 0.1640888749313025, + "grad_norm": 6.068005561828613, + "learning_rate": 1.975288856711116e-05, + "loss": 6.170798950195312, + "step": 5225 + }, + { + "epoch": 0.16487398916542356, + "grad_norm": 9.144749641418457, + "learning_rate": 1.9750278351771674e-05, + "loss": 6.147886962890625, + "step": 5250 + }, + { + "epoch": 0.16565910339954462, + "grad_norm": 8.890758514404297, + "learning_rate": 1.974765461695651e-05, + "loss": 6.161375122070313, + "step": 5275 + }, + { + "epoch": 0.1664442176336657, + "grad_norm": 7.843451023101807, + "learning_rate": 1.9745017366719416e-05, + "loss": 6.143359375, + "step": 5300 + }, + { + "epoch": 0.16722933186778677, + "grad_norm": 8.7725830078125, + "learning_rate": 1.9742366605135045e-05, + "loss": 6.173168334960938, + "step": 5325 + }, + { + "epoch": 0.16801444610190783, + "grad_norm": 8.318331718444824, + "learning_rate": 1.9739702336298906e-05, + "loss": 6.127272338867187, + "step": 5350 + }, + { + "epoch": 0.1687995603360289, + "grad_norm": 7.687984466552734, + "learning_rate": 1.9737024564327387e-05, + "loss": 6.1730194091796875, + "step": 5375 + }, + { + "epoch": 0.16958467457014995, + "grad_norm": 8.86296558380127, + "learning_rate": 1.973433329335774e-05, + "loss": 6.131113891601562, + "step": 5400 + }, + { + "epoch": 0.17036978880427103, + "grad_norm": 8.887882232666016, + "learning_rate": 1.9731628527548057e-05, + "loss": 6.142433471679688, + "step": 5425 + }, + { + "epoch": 0.1711549030383921, + "grad_norm": 10.87294864654541, + "learning_rate": 1.9728910271077307e-05, + "loss": 6.155120239257813, + "step": 5450 + }, + { + "epoch": 0.17194001727251315, + "grad_norm": 8.872554779052734, + "learning_rate": 1.972617852814528e-05, + "loss": 6.151682739257812, + "step": 5475 + }, + { + "epoch": 0.1727251315066342, + "grad_norm": 7.303031921386719, + "learning_rate": 1.972343330297261e-05, + "loss": 6.155535888671875, + "step": 5500 + }, + { + "epoch": 0.17351024574075527, + "grad_norm": 9.150452613830566, + "learning_rate": 1.9720674599800772e-05, + "loss": 6.130807495117187, + "step": 5525 + }, + { + "epoch": 0.17429535997487636, + "grad_norm": 7.366339683532715, + "learning_rate": 1.9717902422892046e-05, + "loss": 6.157325439453125, + "step": 5550 + }, + { + "epoch": 0.17508047420899742, + "grad_norm": 8.925331115722656, + "learning_rate": 1.971511677652954e-05, + "loss": 6.1735009765625, + "step": 5575 + }, + { + "epoch": 0.17586558844311848, + "grad_norm": 11.452009201049805, + "learning_rate": 1.9712317665017175e-05, + "loss": 6.158316040039063, + "step": 5600 + }, + { + "epoch": 0.17665070267723954, + "grad_norm": 8.121882438659668, + "learning_rate": 1.9709505092679674e-05, + "loss": 6.1595166015625, + "step": 5625 + }, + { + "epoch": 0.1774358169113606, + "grad_norm": 10.033334732055664, + "learning_rate": 1.9706679063862555e-05, + "loss": 6.15797119140625, + "step": 5650 + }, + { + "epoch": 0.17822093114548168, + "grad_norm": 9.641092300415039, + "learning_rate": 1.9703839582932122e-05, + "loss": 6.139176635742188, + "step": 5675 + }, + { + "epoch": 0.17900604537960274, + "grad_norm": 10.898619651794434, + "learning_rate": 1.970098665427548e-05, + "loss": 6.15917724609375, + "step": 5700 + }, + { + "epoch": 0.1797911596137238, + "grad_norm": 6.8958563804626465, + "learning_rate": 1.969812028230049e-05, + "loss": 6.164471435546875, + "step": 5725 + }, + { + "epoch": 0.18057627384784486, + "grad_norm": 11.500550270080566, + "learning_rate": 1.96952404714358e-05, + "loss": 6.1606512451171875, + "step": 5750 + }, + { + "epoch": 0.18136138808196592, + "grad_norm": 9.062318801879883, + "learning_rate": 1.9692347226130814e-05, + "loss": 6.140675659179688, + "step": 5775 + }, + { + "epoch": 0.18214650231608698, + "grad_norm": 8.850255012512207, + "learning_rate": 1.968944055085569e-05, + "loss": 6.12903076171875, + "step": 5800 + }, + { + "epoch": 0.18293161655020806, + "grad_norm": 8.567095756530762, + "learning_rate": 1.968652045010134e-05, + "loss": 6.147072143554688, + "step": 5825 + }, + { + "epoch": 0.18371673078432912, + "grad_norm": 7.051018714904785, + "learning_rate": 1.9683586928379424e-05, + "loss": 6.113607177734375, + "step": 5850 + }, + { + "epoch": 0.18450184501845018, + "grad_norm": 9.476158142089844, + "learning_rate": 1.968063999022233e-05, + "loss": 6.181278686523438, + "step": 5875 + }, + { + "epoch": 0.18528695925257124, + "grad_norm": 7.5354533195495605, + "learning_rate": 1.9677679640183174e-05, + "loss": 6.1427880859375, + "step": 5900 + }, + { + "epoch": 0.1860720734866923, + "grad_norm": 9.682645797729492, + "learning_rate": 1.9674705882835797e-05, + "loss": 6.11967529296875, + "step": 5925 + }, + { + "epoch": 0.1868571877208134, + "grad_norm": 7.102179527282715, + "learning_rate": 1.9671718722774758e-05, + "loss": 6.123211059570313, + "step": 5950 + }, + { + "epoch": 0.18764230195493445, + "grad_norm": 10.115147590637207, + "learning_rate": 1.9668718164615317e-05, + "loss": 6.202584266662598, + "step": 5975 + }, + { + "epoch": 0.1884274161890555, + "grad_norm": 7.011385917663574, + "learning_rate": 1.966570421299344e-05, + "loss": 6.12556640625, + "step": 6000 + }, + { + "epoch": 0.18921253042317657, + "grad_norm": 7.7122063636779785, + "learning_rate": 1.966267687256578e-05, + "loss": 6.1134423828125, + "step": 6025 + }, + { + "epoch": 0.18999764465729763, + "grad_norm": 7.824562072753906, + "learning_rate": 1.9659636148009686e-05, + "loss": 6.121304321289062, + "step": 6050 + }, + { + "epoch": 0.1907827588914187, + "grad_norm": 10.313433647155762, + "learning_rate": 1.9656582044023177e-05, + "loss": 6.126422119140625, + "step": 6075 + }, + { + "epoch": 0.19156787312553977, + "grad_norm": 8.119521141052246, + "learning_rate": 1.9653514565324947e-05, + "loss": 6.150328369140625, + "step": 6100 + }, + { + "epoch": 0.19235298735966083, + "grad_norm": 8.781272888183594, + "learning_rate": 1.9650433716654357e-05, + "loss": 6.125905151367188, + "step": 6125 + }, + { + "epoch": 0.1931381015937819, + "grad_norm": 9.367782592773438, + "learning_rate": 1.9647339502771416e-05, + "loss": 6.138644409179688, + "step": 6150 + }, + { + "epoch": 0.19392321582790295, + "grad_norm": 6.009584426879883, + "learning_rate": 1.9644231928456797e-05, + "loss": 6.137570190429687, + "step": 6175 + }, + { + "epoch": 0.19470833006202404, + "grad_norm": 10.180763244628906, + "learning_rate": 1.9641110998511798e-05, + "loss": 6.124027099609375, + "step": 6200 + }, + { + "epoch": 0.1954934442961451, + "grad_norm": 8.387354850769043, + "learning_rate": 1.963797671775837e-05, + "loss": 6.130537109375, + "step": 6225 + }, + { + "epoch": 0.19627855853026616, + "grad_norm": 8.385422706604004, + "learning_rate": 1.9634829091039077e-05, + "loss": 6.12103759765625, + "step": 6250 + }, + { + "epoch": 0.19706367276438722, + "grad_norm": 9.353492736816406, + "learning_rate": 1.963166812321711e-05, + "loss": 6.15225341796875, + "step": 6275 + }, + { + "epoch": 0.19784878699850827, + "grad_norm": 8.940140724182129, + "learning_rate": 1.9628493819176274e-05, + "loss": 6.13609619140625, + "step": 6300 + }, + { + "epoch": 0.19863390123262933, + "grad_norm": 7.4060869216918945, + "learning_rate": 1.9625306183820972e-05, + "loss": 6.13810791015625, + "step": 6325 + }, + { + "epoch": 0.19941901546675042, + "grad_norm": 7.979679584503174, + "learning_rate": 1.962210522207621e-05, + "loss": 6.127867431640625, + "step": 6350 + }, + { + "epoch": 0.20020412970087148, + "grad_norm": 8.318560600280762, + "learning_rate": 1.9618890938887584e-05, + "loss": 6.142042846679687, + "step": 6375 + }, + { + "epoch": 0.20098924393499254, + "grad_norm": 8.278966903686523, + "learning_rate": 1.9615663339221267e-05, + "loss": 6.136571655273437, + "step": 6400 + }, + { + "epoch": 0.2017743581691136, + "grad_norm": 8.510711669921875, + "learning_rate": 1.961242242806401e-05, + "loss": 6.141307983398438, + "step": 6425 + }, + { + "epoch": 0.20255947240323466, + "grad_norm": 10.701105117797852, + "learning_rate": 1.9609168210423134e-05, + "loss": 6.114397583007812, + "step": 6450 + }, + { + "epoch": 0.20334458663735575, + "grad_norm": 7.05831241607666, + "learning_rate": 1.9605900691326504e-05, + "loss": 6.137372436523438, + "step": 6475 + }, + { + "epoch": 0.2041297008714768, + "grad_norm": 7.726666450500488, + "learning_rate": 1.9602619875822563e-05, + "loss": 6.113399658203125, + "step": 6500 + }, + { + "epoch": 0.20491481510559786, + "grad_norm": 9.983379364013672, + "learning_rate": 1.9599325768980273e-05, + "loss": 6.1419873046875, + "step": 6525 + }, + { + "epoch": 0.20569992933971892, + "grad_norm": 7.255304336547852, + "learning_rate": 1.9596018375889142e-05, + "loss": 6.135066528320312, + "step": 6550 + }, + { + "epoch": 0.20648504357383998, + "grad_norm": 9.520317077636719, + "learning_rate": 1.95926977016592e-05, + "loss": 6.11998779296875, + "step": 6575 + }, + { + "epoch": 0.20727015780796107, + "grad_norm": 6.786963939666748, + "learning_rate": 1.9589363751421008e-05, + "loss": 6.125740356445313, + "step": 6600 + }, + { + "epoch": 0.20805527204208213, + "grad_norm": 7.7769975662231445, + "learning_rate": 1.9586016530325626e-05, + "loss": 6.1314630126953125, + "step": 6625 + }, + { + "epoch": 0.2088403862762032, + "grad_norm": 9.979001998901367, + "learning_rate": 1.958265604354463e-05, + "loss": 6.147931518554688, + "step": 6650 + }, + { + "epoch": 0.20962550051032425, + "grad_norm": 8.705126762390137, + "learning_rate": 1.9579282296270076e-05, + "loss": 6.13921630859375, + "step": 6675 + }, + { + "epoch": 0.2104106147444453, + "grad_norm": 6.68727445602417, + "learning_rate": 1.957589529371453e-05, + "loss": 6.13066650390625, + "step": 6700 + }, + { + "epoch": 0.2111957289785664, + "grad_norm": 9.349790573120117, + "learning_rate": 1.9572495041111014e-05, + "loss": 6.093824462890625, + "step": 6725 + }, + { + "epoch": 0.21198084321268745, + "grad_norm": 8.778707504272461, + "learning_rate": 1.9569081543713037e-05, + "loss": 6.105714721679687, + "step": 6750 + }, + { + "epoch": 0.2127659574468085, + "grad_norm": 8.454255104064941, + "learning_rate": 1.9565654806794568e-05, + "loss": 6.137057495117188, + "step": 6775 + }, + { + "epoch": 0.21355107168092957, + "grad_norm": 8.667468070983887, + "learning_rate": 1.9562214835650036e-05, + "loss": 6.14056396484375, + "step": 6800 + }, + { + "epoch": 0.21433618591505063, + "grad_norm": 6.602585315704346, + "learning_rate": 1.9558761635594306e-05, + "loss": 6.093595581054688, + "step": 6825 + }, + { + "epoch": 0.21512130014917172, + "grad_norm": 7.005263805389404, + "learning_rate": 1.955529521196269e-05, + "loss": 6.101895141601562, + "step": 6850 + }, + { + "epoch": 0.21590641438329278, + "grad_norm": 9.877135276794434, + "learning_rate": 1.9551815570110933e-05, + "loss": 6.129929809570313, + "step": 6875 + }, + { + "epoch": 0.21669152861741384, + "grad_norm": 8.351065635681152, + "learning_rate": 1.95483227154152e-05, + "loss": 6.109620971679687, + "step": 6900 + }, + { + "epoch": 0.2174766428515349, + "grad_norm": 8.608957290649414, + "learning_rate": 1.954481665327206e-05, + "loss": 6.141491088867188, + "step": 6925 + }, + { + "epoch": 0.21826175708565596, + "grad_norm": 7.891459941864014, + "learning_rate": 1.9541297389098507e-05, + "loss": 6.1160345458984375, + "step": 6950 + }, + { + "epoch": 0.21904687131977701, + "grad_norm": 7.396613121032715, + "learning_rate": 1.9537764928331918e-05, + "loss": 6.113394165039063, + "step": 6975 + }, + { + "epoch": 0.2198319855538981, + "grad_norm": 8.824213027954102, + "learning_rate": 1.953421927643007e-05, + "loss": 6.127730102539062, + "step": 7000 + }, + { + "epoch": 0.22061709978801916, + "grad_norm": 10.46630573272705, + "learning_rate": 1.9530660438871107e-05, + "loss": 6.093670654296875, + "step": 7025 + }, + { + "epoch": 0.22140221402214022, + "grad_norm": 9.092950820922852, + "learning_rate": 1.952708842115356e-05, + "loss": 6.115619506835937, + "step": 7050 + }, + { + "epoch": 0.22218732825626128, + "grad_norm": 6.029088973999023, + "learning_rate": 1.9523503228796317e-05, + "loss": 6.106693725585938, + "step": 7075 + }, + { + "epoch": 0.22297244249038234, + "grad_norm": 8.74901294708252, + "learning_rate": 1.9519904867338614e-05, + "loss": 6.131513061523438, + "step": 7100 + }, + { + "epoch": 0.22375755672450343, + "grad_norm": 10.972220420837402, + "learning_rate": 1.9516293342340047e-05, + "loss": 6.138427734375, + "step": 7125 + }, + { + "epoch": 0.22454267095862449, + "grad_norm": 8.7410306930542, + "learning_rate": 1.9512668659380546e-05, + "loss": 6.110509643554687, + "step": 7150 + }, + { + "epoch": 0.22532778519274554, + "grad_norm": 5.667666435241699, + "learning_rate": 1.950903082406036e-05, + "loss": 6.126967163085937, + "step": 7175 + }, + { + "epoch": 0.2261128994268666, + "grad_norm": 8.045656204223633, + "learning_rate": 1.9505379842000078e-05, + "loss": 6.12472412109375, + "step": 7200 + }, + { + "epoch": 0.22689801366098766, + "grad_norm": 7.980565071105957, + "learning_rate": 1.9501715718840587e-05, + "loss": 6.149642333984375, + "step": 7225 + }, + { + "epoch": 0.22768312789510875, + "grad_norm": 7.379549026489258, + "learning_rate": 1.9498038460243076e-05, + "loss": 6.10189453125, + "step": 7250 + }, + { + "epoch": 0.2284682421292298, + "grad_norm": 6.921355724334717, + "learning_rate": 1.949434807188904e-05, + "loss": 6.105921630859375, + "step": 7275 + }, + { + "epoch": 0.22925335636335087, + "grad_norm": 8.45422077178955, + "learning_rate": 1.949064455948025e-05, + "loss": 6.091075439453125, + "step": 7300 + }, + { + "epoch": 0.23003847059747193, + "grad_norm": 7.013403415679932, + "learning_rate": 1.948692792873876e-05, + "loss": 6.133209838867187, + "step": 7325 + }, + { + "epoch": 0.230823584831593, + "grad_norm": 7.077011585235596, + "learning_rate": 1.9483198185406887e-05, + "loss": 6.099066162109375, + "step": 7350 + }, + { + "epoch": 0.23160869906571407, + "grad_norm": 9.267788887023926, + "learning_rate": 1.9479455335247217e-05, + "loss": 6.1107330322265625, + "step": 7375 + }, + { + "epoch": 0.23239381329983513, + "grad_norm": 6.571483135223389, + "learning_rate": 1.947569938404257e-05, + "loss": 6.0654229736328125, + "step": 7400 + }, + { + "epoch": 0.2331789275339562, + "grad_norm": 8.133341789245605, + "learning_rate": 1.947193033759603e-05, + "loss": 6.097215576171875, + "step": 7425 + }, + { + "epoch": 0.23396404176807725, + "grad_norm": 8.099224090576172, + "learning_rate": 1.946814820173089e-05, + "loss": 6.097811279296875, + "step": 7450 + }, + { + "epoch": 0.2347491560021983, + "grad_norm": 7.7246575355529785, + "learning_rate": 1.9464352982290683e-05, + "loss": 6.126438598632813, + "step": 7475 + }, + { + "epoch": 0.23553427023631937, + "grad_norm": 7.32771110534668, + "learning_rate": 1.946054468513915e-05, + "loss": 6.114263916015625, + "step": 7500 + }, + { + "epoch": 0.23631938447044046, + "grad_norm": 9.025934219360352, + "learning_rate": 1.9456723316160237e-05, + "loss": 6.097006225585938, + "step": 7525 + }, + { + "epoch": 0.23710449870456152, + "grad_norm": 8.413931846618652, + "learning_rate": 1.945288888125809e-05, + "loss": 6.078002319335938, + "step": 7550 + }, + { + "epoch": 0.23788961293868258, + "grad_norm": 8.507669448852539, + "learning_rate": 1.9449041386357042e-05, + "loss": 6.11097412109375, + "step": 7575 + }, + { + "epoch": 0.23867472717280364, + "grad_norm": 7.603123188018799, + "learning_rate": 1.9445180837401603e-05, + "loss": 6.108457641601563, + "step": 7600 + }, + { + "epoch": 0.2394598414069247, + "grad_norm": 8.62441349029541, + "learning_rate": 1.9441307240356446e-05, + "loss": 6.09090576171875, + "step": 7625 + }, + { + "epoch": 0.24024495564104578, + "grad_norm": 6.84232234954834, + "learning_rate": 1.9437420601206414e-05, + "loss": 6.079618530273438, + "step": 7650 + }, + { + "epoch": 0.24103006987516684, + "grad_norm": 7.540565013885498, + "learning_rate": 1.9433520925956497e-05, + "loss": 6.145294799804687, + "step": 7675 + }, + { + "epoch": 0.2418151841092879, + "grad_norm": 8.01793098449707, + "learning_rate": 1.942960822063182e-05, + "loss": 6.105317993164062, + "step": 7700 + }, + { + "epoch": 0.24260029834340896, + "grad_norm": 6.836312294006348, + "learning_rate": 1.9425682491277653e-05, + "loss": 6.118194580078125, + "step": 7725 + }, + { + "epoch": 0.24338541257753002, + "grad_norm": 6.854023456573486, + "learning_rate": 1.942174374395937e-05, + "loss": 6.098612060546875, + "step": 7750 + }, + { + "epoch": 0.2441705268116511, + "grad_norm": 7.412289619445801, + "learning_rate": 1.9417791984762474e-05, + "loss": 6.10634765625, + "step": 7775 + }, + { + "epoch": 0.24495564104577217, + "grad_norm": 9.05181884765625, + "learning_rate": 1.9413827219792565e-05, + "loss": 6.1146923828125, + "step": 7800 + }, + { + "epoch": 0.24574075527989322, + "grad_norm": 7.55516242980957, + "learning_rate": 1.9409849455175346e-05, + "loss": 6.099641723632812, + "step": 7825 + }, + { + "epoch": 0.24652586951401428, + "grad_norm": 7.916590690612793, + "learning_rate": 1.940585869705659e-05, + "loss": 6.098927612304688, + "step": 7850 + }, + { + "epoch": 0.24731098374813534, + "grad_norm": 7.240673542022705, + "learning_rate": 1.9401854951602152e-05, + "loss": 6.111246337890625, + "step": 7875 + }, + { + "epoch": 0.24809609798225643, + "grad_norm": 6.73378849029541, + "learning_rate": 1.939783822499796e-05, + "loss": 6.067181396484375, + "step": 7900 + }, + { + "epoch": 0.2488812122163775, + "grad_norm": 7.575690746307373, + "learning_rate": 1.939380852344999e-05, + "loss": 6.0854052734375, + "step": 7925 + }, + { + "epoch": 0.24966632645049855, + "grad_norm": 8.157740592956543, + "learning_rate": 1.9389765853184263e-05, + "loss": 6.089390869140625, + "step": 7950 + }, + { + "epoch": 0.2504514406846196, + "grad_norm": 8.730521202087402, + "learning_rate": 1.9385710220446847e-05, + "loss": 6.080964769635882, + "step": 7975 + }, + { + "epoch": 0.2512365549187407, + "grad_norm": 8.453075408935547, + "learning_rate": 1.9381641631503828e-05, + "loss": 6.090482177734375, + "step": 8000 + }, + { + "epoch": 0.2520216691528617, + "grad_norm": 8.2374849319458, + "learning_rate": 1.9377560092641314e-05, + "loss": 6.093213500976563, + "step": 8025 + }, + { + "epoch": 0.2528067833869828, + "grad_norm": 7.423177719116211, + "learning_rate": 1.9373465610165423e-05, + "loss": 6.081823120117187, + "step": 8050 + }, + { + "epoch": 0.25359189762110385, + "grad_norm": 7.9487528800964355, + "learning_rate": 1.9369358190402262e-05, + "loss": 6.083873291015625, + "step": 8075 + }, + { + "epoch": 0.25437701185522493, + "grad_norm": 8.090024948120117, + "learning_rate": 1.9365237839697944e-05, + "loss": 6.120177612304688, + "step": 8100 + }, + { + "epoch": 0.255162126089346, + "grad_norm": 8.478367805480957, + "learning_rate": 1.9361104564418542e-05, + "loss": 6.0869818115234375, + "step": 8125 + }, + { + "epoch": 0.25594724032346705, + "grad_norm": 9.296222686767578, + "learning_rate": 1.9356958370950107e-05, + "loss": 6.09420654296875, + "step": 8150 + }, + { + "epoch": 0.25673235455758814, + "grad_norm": 8.544968605041504, + "learning_rate": 1.9352799265698646e-05, + "loss": 6.1034478759765625, + "step": 8175 + }, + { + "epoch": 0.25751746879170917, + "grad_norm": 7.534852027893066, + "learning_rate": 1.9348627255090125e-05, + "loss": 6.132659912109375, + "step": 8200 + }, + { + "epoch": 0.25830258302583026, + "grad_norm": 8.706232070922852, + "learning_rate": 1.9344442345570437e-05, + "loss": 6.1021826171875, + "step": 8225 + }, + { + "epoch": 0.25908769725995134, + "grad_norm": 7.513480186462402, + "learning_rate": 1.9340244543605405e-05, + "loss": 6.09001220703125, + "step": 8250 + }, + { + "epoch": 0.2598728114940724, + "grad_norm": 10.582332611083984, + "learning_rate": 1.9336033855680777e-05, + "loss": 6.080181884765625, + "step": 8275 + }, + { + "epoch": 0.26065792572819346, + "grad_norm": 8.431282043457031, + "learning_rate": 1.9331810288302216e-05, + "loss": 6.084490966796875, + "step": 8300 + }, + { + "epoch": 0.2614430399623145, + "grad_norm": 7.604475021362305, + "learning_rate": 1.9327573847995262e-05, + "loss": 6.077401123046875, + "step": 8325 + }, + { + "epoch": 0.2622281541964356, + "grad_norm": 7.361900806427002, + "learning_rate": 1.9323324541305367e-05, + "loss": 6.09489013671875, + "step": 8350 + }, + { + "epoch": 0.26301326843055667, + "grad_norm": 8.024670600891113, + "learning_rate": 1.9319062374797855e-05, + "loss": 6.065264282226562, + "step": 8375 + }, + { + "epoch": 0.2637983826646777, + "grad_norm": 9.078442573547363, + "learning_rate": 1.931478735505791e-05, + "loss": 6.1088934326171875, + "step": 8400 + }, + { + "epoch": 0.2645834968987988, + "grad_norm": 9.455032348632812, + "learning_rate": 1.9310499488690592e-05, + "loss": 6.089759521484375, + "step": 8425 + }, + { + "epoch": 0.2653686111329198, + "grad_norm": 5.793647289276123, + "learning_rate": 1.9306198782320785e-05, + "loss": 6.086008911132812, + "step": 8450 + }, + { + "epoch": 0.2661537253670409, + "grad_norm": 9.200705528259277, + "learning_rate": 1.9301885242593243e-05, + "loss": 6.063539428710937, + "step": 8475 + }, + { + "epoch": 0.266938839601162, + "grad_norm": 7.469987869262695, + "learning_rate": 1.929755887617252e-05, + "loss": 6.091107177734375, + "step": 8500 + }, + { + "epoch": 0.267723953835283, + "grad_norm": 9.012052536010742, + "learning_rate": 1.9293219689742997e-05, + "loss": 6.072770385742188, + "step": 8525 + }, + { + "epoch": 0.2685090680694041, + "grad_norm": 7.782796382904053, + "learning_rate": 1.928886769000887e-05, + "loss": 6.0926513671875, + "step": 8550 + }, + { + "epoch": 0.26929418230352514, + "grad_norm": 6.323625564575195, + "learning_rate": 1.9284502883694122e-05, + "loss": 6.08256591796875, + "step": 8575 + }, + { + "epoch": 0.27007929653764623, + "grad_norm": 7.582986354827881, + "learning_rate": 1.9280125277542526e-05, + "loss": 6.103290405273437, + "step": 8600 + }, + { + "epoch": 0.2708644107717673, + "grad_norm": 7.042757034301758, + "learning_rate": 1.927573487831763e-05, + "loss": 6.099789428710937, + "step": 8625 + }, + { + "epoch": 0.27164952500588835, + "grad_norm": 9.09696102142334, + "learning_rate": 1.9271331692802753e-05, + "loss": 6.081097412109375, + "step": 8650 + }, + { + "epoch": 0.27243463924000944, + "grad_norm": 11.186524391174316, + "learning_rate": 1.9266915727800965e-05, + "loss": 6.084293212890625, + "step": 8675 + }, + { + "epoch": 0.27321975347413047, + "grad_norm": 6.84335994720459, + "learning_rate": 1.926248699013508e-05, + "loss": 6.096376342773437, + "step": 8700 + }, + { + "epoch": 0.27400486770825155, + "grad_norm": 9.2902193069458, + "learning_rate": 1.9258045486647646e-05, + "loss": 6.06237548828125, + "step": 8725 + }, + { + "epoch": 0.27478998194237264, + "grad_norm": 5.458731651306152, + "learning_rate": 1.9253591224200938e-05, + "loss": 6.08322021484375, + "step": 8750 + }, + { + "epoch": 0.2755750961764937, + "grad_norm": 7.920042037963867, + "learning_rate": 1.924912420967694e-05, + "loss": 6.079279174804688, + "step": 8775 + }, + { + "epoch": 0.27636021041061476, + "grad_norm": 6.038084030151367, + "learning_rate": 1.9244644449977338e-05, + "loss": 6.0774169921875, + "step": 8800 + }, + { + "epoch": 0.2771453246447358, + "grad_norm": 6.85615348815918, + "learning_rate": 1.9240151952023516e-05, + "loss": 6.091692504882812, + "step": 8825 + }, + { + "epoch": 0.2779304388788569, + "grad_norm": 7.801642417907715, + "learning_rate": 1.923564672275654e-05, + "loss": 6.07525634765625, + "step": 8850 + }, + { + "epoch": 0.27871555311297797, + "grad_norm": 8.359718322753906, + "learning_rate": 1.9231128769137134e-05, + "loss": 6.0793701171875, + "step": 8875 + }, + { + "epoch": 0.279500667347099, + "grad_norm": 6.57523250579834, + "learning_rate": 1.922659809814569e-05, + "loss": 6.082147827148438, + "step": 8900 + }, + { + "epoch": 0.2802857815812201, + "grad_norm": 6.429874897003174, + "learning_rate": 1.9222054716782246e-05, + "loss": 6.071991577148437, + "step": 8925 + }, + { + "epoch": 0.2810708958153411, + "grad_norm": 6.105246543884277, + "learning_rate": 1.9217498632066483e-05, + "loss": 6.070576782226563, + "step": 8950 + }, + { + "epoch": 0.2818560100494622, + "grad_norm": 10.252945899963379, + "learning_rate": 1.9212929851037712e-05, + "loss": 6.060130069130345, + "step": 8975 + }, + { + "epoch": 0.28264112428358323, + "grad_norm": 10.666425704956055, + "learning_rate": 1.9208348380754844e-05, + "loss": 6.07262451171875, + "step": 9000 + }, + { + "epoch": 0.2834262385177043, + "grad_norm": 7.326347827911377, + "learning_rate": 1.9203754228296404e-05, + "loss": 6.089608154296875, + "step": 9025 + }, + { + "epoch": 0.2842113527518254, + "grad_norm": 6.007728576660156, + "learning_rate": 1.919914740076052e-05, + "loss": 6.067042236328125, + "step": 9050 + }, + { + "epoch": 0.28499646698594644, + "grad_norm": 6.439706802368164, + "learning_rate": 1.919452790526489e-05, + "loss": 6.100827026367187, + "step": 9075 + }, + { + "epoch": 0.2857815812200675, + "grad_norm": 7.205475807189941, + "learning_rate": 1.91898957489468e-05, + "loss": 6.0810308837890625, + "step": 9100 + }, + { + "epoch": 0.28656669545418856, + "grad_norm": 8.613924026489258, + "learning_rate": 1.918525093896308e-05, + "loss": 6.0901861572265625, + "step": 9125 + }, + { + "epoch": 0.28735180968830965, + "grad_norm": 7.643390655517578, + "learning_rate": 1.918059348249012e-05, + "loss": 6.073208618164062, + "step": 9150 + }, + { + "epoch": 0.28813692392243073, + "grad_norm": 7.871557235717773, + "learning_rate": 1.9175923386723853e-05, + "loss": 6.064640502929688, + "step": 9175 + }, + { + "epoch": 0.28892203815655176, + "grad_norm": 8.201736450195312, + "learning_rate": 1.9171240658879723e-05, + "loss": 6.089393310546875, + "step": 9200 + }, + { + "epoch": 0.28970715239067285, + "grad_norm": 9.029735565185547, + "learning_rate": 1.9166545306192714e-05, + "loss": 6.086573486328125, + "step": 9225 + }, + { + "epoch": 0.2904922666247939, + "grad_norm": 8.206645965576172, + "learning_rate": 1.9161837335917297e-05, + "loss": 6.097440795898438, + "step": 9250 + }, + { + "epoch": 0.29127738085891497, + "grad_norm": 6.268886089324951, + "learning_rate": 1.915711675532745e-05, + "loss": 6.066869506835937, + "step": 9275 + }, + { + "epoch": 0.29206249509303606, + "grad_norm": 7.023653984069824, + "learning_rate": 1.9152383571716628e-05, + "loss": 6.078419189453125, + "step": 9300 + }, + { + "epoch": 0.2928476093271571, + "grad_norm": 6.25575590133667, + "learning_rate": 1.914763779239776e-05, + "loss": 6.0618048095703125, + "step": 9325 + }, + { + "epoch": 0.2936327235612782, + "grad_norm": 7.432967185974121, + "learning_rate": 1.9142879424703233e-05, + "loss": 6.0575799560546875, + "step": 9350 + }, + { + "epoch": 0.2944178377953992, + "grad_norm": 7.492427825927734, + "learning_rate": 1.913810847598489e-05, + "loss": 6.046594848632813, + "step": 9375 + }, + { + "epoch": 0.2952029520295203, + "grad_norm": 6.218588829040527, + "learning_rate": 1.9133324953614003e-05, + "loss": 6.095703125, + "step": 9400 + }, + { + "epoch": 0.2959880662636414, + "grad_norm": 8.066258430480957, + "learning_rate": 1.912852886498128e-05, + "loss": 6.078289184570313, + "step": 9425 + }, + { + "epoch": 0.2967731804977624, + "grad_norm": 9.229392051696777, + "learning_rate": 1.9123720217496837e-05, + "loss": 6.059725952148438, + "step": 9450 + }, + { + "epoch": 0.2975582947318835, + "grad_norm": 6.735298156738281, + "learning_rate": 1.91188990185902e-05, + "loss": 6.08950439453125, + "step": 9475 + }, + { + "epoch": 0.29834340896600453, + "grad_norm": 6.997681140899658, + "learning_rate": 1.911406527571028e-05, + "loss": 6.0705413818359375, + "step": 9500 + }, + { + "epoch": 0.2991285232001256, + "grad_norm": 8.46038818359375, + "learning_rate": 1.9109218996325373e-05, + "loss": 6.0867852783203125, + "step": 9525 + }, + { + "epoch": 0.2999136374342467, + "grad_norm": 8.23660659790039, + "learning_rate": 1.910436018792315e-05, + "loss": 6.060445556640625, + "step": 9550 + }, + { + "epoch": 0.30069875166836774, + "grad_norm": 7.068746089935303, + "learning_rate": 1.909948885801063e-05, + "loss": 6.088594970703125, + "step": 9575 + }, + { + "epoch": 0.3014838659024888, + "grad_norm": 7.799080848693848, + "learning_rate": 1.9094605014114183e-05, + "loss": 6.07252685546875, + "step": 9600 + }, + { + "epoch": 0.30226898013660985, + "grad_norm": 7.389161586761475, + "learning_rate": 1.9089708663779517e-05, + "loss": 6.052852783203125, + "step": 9625 + }, + { + "epoch": 0.30305409437073094, + "grad_norm": 6.638370513916016, + "learning_rate": 1.9084799814571647e-05, + "loss": 6.065245971679688, + "step": 9650 + }, + { + "epoch": 0.30383920860485203, + "grad_norm": 9.253580093383789, + "learning_rate": 1.907987847407492e-05, + "loss": 6.095588989257813, + "step": 9675 + }, + { + "epoch": 0.30462432283897306, + "grad_norm": 6.421274185180664, + "learning_rate": 1.9074944649892977e-05, + "loss": 6.068941650390625, + "step": 9700 + }, + { + "epoch": 0.30540943707309415, + "grad_norm": 9.121893882751465, + "learning_rate": 1.9069998349648736e-05, + "loss": 6.072774658203125, + "step": 9725 + }, + { + "epoch": 0.3061945513072152, + "grad_norm": 7.841750144958496, + "learning_rate": 1.9065039580984396e-05, + "loss": 6.096015625, + "step": 9750 + }, + { + "epoch": 0.30697966554133627, + "grad_norm": 6.011407852172852, + "learning_rate": 1.906006835156143e-05, + "loss": 6.049310913085938, + "step": 9775 + }, + { + "epoch": 0.30776477977545735, + "grad_norm": 7.44486141204834, + "learning_rate": 1.9055084669060554e-05, + "loss": 6.06850830078125, + "step": 9800 + }, + { + "epoch": 0.3085498940095784, + "grad_norm": 8.295485496520996, + "learning_rate": 1.9050088541181716e-05, + "loss": 6.059293212890625, + "step": 9825 + }, + { + "epoch": 0.30933500824369947, + "grad_norm": 7.225030899047852, + "learning_rate": 1.9045079975644113e-05, + "loss": 6.0575927734375, + "step": 9850 + }, + { + "epoch": 0.3101201224778205, + "grad_norm": 5.714171409606934, + "learning_rate": 1.9040058980186143e-05, + "loss": 6.06213134765625, + "step": 9875 + }, + { + "epoch": 0.3109052367119416, + "grad_norm": 7.30633544921875, + "learning_rate": 1.9035025562565417e-05, + "loss": 6.063967895507813, + "step": 9900 + }, + { + "epoch": 0.3116903509460627, + "grad_norm": 5.63694953918457, + "learning_rate": 1.9029979730558736e-05, + "loss": 6.038084716796875, + "step": 9925 + }, + { + "epoch": 0.3124754651801837, + "grad_norm": 7.0372114181518555, + "learning_rate": 1.9024921491962074e-05, + "loss": 6.0520458984375, + "step": 9950 + }, + { + "epoch": 0.3132605794143048, + "grad_norm": 7.623991966247559, + "learning_rate": 1.9019850854590586e-05, + "loss": 6.0503997802734375, + "step": 9975 + }, + { + "epoch": 0.3140456936484258, + "grad_norm": 6.508340835571289, + "learning_rate": 1.9014767826278576e-05, + "loss": 6.064945068359375, + "step": 10000 + }, + { + "epoch": 0.3148308078825469, + "grad_norm": 6.292807579040527, + "learning_rate": 1.9009672414879498e-05, + "loss": 6.060416259765625, + "step": 10025 + }, + { + "epoch": 0.315615922116668, + "grad_norm": 8.361135482788086, + "learning_rate": 1.900456462826593e-05, + "loss": 6.035560302734375, + "step": 10050 + }, + { + "epoch": 0.31640103635078903, + "grad_norm": 6.783228397369385, + "learning_rate": 1.8999444474329577e-05, + "loss": 6.063842163085938, + "step": 10075 + }, + { + "epoch": 0.3171861505849101, + "grad_norm": 10.100295066833496, + "learning_rate": 1.8994311960981253e-05, + "loss": 6.052635498046875, + "step": 10100 + }, + { + "epoch": 0.31797126481903115, + "grad_norm": 7.355575084686279, + "learning_rate": 1.8989167096150858e-05, + "loss": 6.057730712890625, + "step": 10125 + }, + { + "epoch": 0.31875637905315224, + "grad_norm": 6.112579345703125, + "learning_rate": 1.8984009887787386e-05, + "loss": 6.071738891601562, + "step": 10150 + }, + { + "epoch": 0.3195414932872733, + "grad_norm": 5.873129844665527, + "learning_rate": 1.89788403438589e-05, + "loss": 6.049765625, + "step": 10175 + }, + { + "epoch": 0.32032660752139436, + "grad_norm": 6.607877254486084, + "learning_rate": 1.8973658472352517e-05, + "loss": 6.09537841796875, + "step": 10200 + }, + { + "epoch": 0.32111172175551544, + "grad_norm": 7.894289016723633, + "learning_rate": 1.89684642812744e-05, + "loss": 6.060230712890625, + "step": 10225 + }, + { + "epoch": 0.3218968359896365, + "grad_norm": 7.560802936553955, + "learning_rate": 1.8963257778649756e-05, + "loss": 6.086636962890625, + "step": 10250 + }, + { + "epoch": 0.32268195022375756, + "grad_norm": 7.569696426391602, + "learning_rate": 1.8958038972522807e-05, + "loss": 6.079852294921875, + "step": 10275 + }, + { + "epoch": 0.3234670644578786, + "grad_norm": 8.521283149719238, + "learning_rate": 1.895280787095678e-05, + "loss": 6.0591973876953125, + "step": 10300 + }, + { + "epoch": 0.3242521786919997, + "grad_norm": 8.364990234375, + "learning_rate": 1.894756448203391e-05, + "loss": 6.064938354492187, + "step": 10325 + }, + { + "epoch": 0.32503729292612077, + "grad_norm": 7.309362411499023, + "learning_rate": 1.8942308813855404e-05, + "loss": 6.033217163085937, + "step": 10350 + }, + { + "epoch": 0.3258224071602418, + "grad_norm": 6.4552001953125, + "learning_rate": 1.893704087454145e-05, + "loss": 6.065508422851562, + "step": 10375 + }, + { + "epoch": 0.3266075213943629, + "grad_norm": 8.206357955932617, + "learning_rate": 1.8931760672231193e-05, + "loss": 6.045038452148438, + "step": 10400 + }, + { + "epoch": 0.3273926356284839, + "grad_norm": 8.414793968200684, + "learning_rate": 1.8926468215082724e-05, + "loss": 6.039845581054688, + "step": 10425 + }, + { + "epoch": 0.328177749862605, + "grad_norm": 9.054853439331055, + "learning_rate": 1.8921163511273064e-05, + "loss": 6.053256225585938, + "step": 10450 + }, + { + "epoch": 0.3289628640967261, + "grad_norm": 7.0534844398498535, + "learning_rate": 1.8915846568998165e-05, + "loss": 6.062099609375, + "step": 10475 + }, + { + "epoch": 0.3297479783308471, + "grad_norm": 8.93610668182373, + "learning_rate": 1.891051739647288e-05, + "loss": 6.0590966796875, + "step": 10500 + }, + { + "epoch": 0.3305330925649682, + "grad_norm": 9.573548316955566, + "learning_rate": 1.890517600193096e-05, + "loss": 6.082064208984375, + "step": 10525 + }, + { + "epoch": 0.33131820679908924, + "grad_norm": 7.467121124267578, + "learning_rate": 1.889982239362504e-05, + "loss": 6.064890747070312, + "step": 10550 + }, + { + "epoch": 0.33210332103321033, + "grad_norm": 6.912933349609375, + "learning_rate": 1.8894456579826626e-05, + "loss": 6.044160766601562, + "step": 10575 + }, + { + "epoch": 0.3328884352673314, + "grad_norm": 7.090230941772461, + "learning_rate": 1.8889078568826082e-05, + "loss": 6.061206665039062, + "step": 10600 + }, + { + "epoch": 0.33367354950145245, + "grad_norm": 6.8900885581970215, + "learning_rate": 1.8883688368932617e-05, + "loss": 6.034147338867188, + "step": 10625 + }, + { + "epoch": 0.33445866373557354, + "grad_norm": 8.430006980895996, + "learning_rate": 1.887828598847427e-05, + "loss": 6.04238037109375, + "step": 10650 + }, + { + "epoch": 0.33524377796969457, + "grad_norm": 6.7249250411987305, + "learning_rate": 1.8872871435797903e-05, + "loss": 6.017744750976562, + "step": 10675 + }, + { + "epoch": 0.33602889220381565, + "grad_norm": 7.001337051391602, + "learning_rate": 1.8867444719269186e-05, + "loss": 6.073046875, + "step": 10700 + }, + { + "epoch": 0.33681400643793674, + "grad_norm": 6.054315567016602, + "learning_rate": 1.8862005847272577e-05, + "loss": 6.064963989257812, + "step": 10725 + }, + { + "epoch": 0.3375991206720578, + "grad_norm": 6.656607627868652, + "learning_rate": 1.8856554828211312e-05, + "loss": 6.040057373046875, + "step": 10750 + }, + { + "epoch": 0.33838423490617886, + "grad_norm": 7.471758842468262, + "learning_rate": 1.8851091670507404e-05, + "loss": 6.041532592773438, + "step": 10775 + }, + { + "epoch": 0.3391693491402999, + "grad_norm": 7.223377227783203, + "learning_rate": 1.8845616382601616e-05, + "loss": 6.060450439453125, + "step": 10800 + }, + { + "epoch": 0.339954463374421, + "grad_norm": 8.435572624206543, + "learning_rate": 1.8840128972953454e-05, + "loss": 6.01211181640625, + "step": 10825 + }, + { + "epoch": 0.34073957760854207, + "grad_norm": 5.700301170349121, + "learning_rate": 1.8834629450041148e-05, + "loss": 6.032303466796875, + "step": 10850 + }, + { + "epoch": 0.3415246918426631, + "grad_norm": 6.378774642944336, + "learning_rate": 1.882911782236165e-05, + "loss": 6.051644897460937, + "step": 10875 + }, + { + "epoch": 0.3423098060767842, + "grad_norm": 10.972804069519043, + "learning_rate": 1.882359409843061e-05, + "loss": 6.039440307617188, + "step": 10900 + }, + { + "epoch": 0.3430949203109052, + "grad_norm": 8.193053245544434, + "learning_rate": 1.8818058286782364e-05, + "loss": 6.0320892333984375, + "step": 10925 + }, + { + "epoch": 0.3438800345450263, + "grad_norm": 7.77625036239624, + "learning_rate": 1.8812510395969936e-05, + "loss": 6.04307861328125, + "step": 10950 + }, + { + "epoch": 0.3446651487791474, + "grad_norm": 8.063611030578613, + "learning_rate": 1.8806950434565e-05, + "loss": 6.060855712890625, + "step": 10975 + }, + { + "epoch": 0.3454502630132684, + "grad_norm": 6.406928062438965, + "learning_rate": 1.8801378411157886e-05, + "loss": 6.02157470703125, + "step": 11000 + }, + { + "epoch": 0.3462353772473895, + "grad_norm": 6.904330730438232, + "learning_rate": 1.8795794334357556e-05, + "loss": 6.047982788085937, + "step": 11025 + }, + { + "epoch": 0.34702049148151054, + "grad_norm": 6.215887069702148, + "learning_rate": 1.8790198212791596e-05, + "loss": 6.030949096679688, + "step": 11050 + }, + { + "epoch": 0.3478056057156316, + "grad_norm": 9.67748737335205, + "learning_rate": 1.8784590055106207e-05, + "loss": 6.0265655517578125, + "step": 11075 + }, + { + "epoch": 0.3485907199497527, + "grad_norm": 8.487585067749023, + "learning_rate": 1.877896986996618e-05, + "loss": 6.045703125, + "step": 11100 + }, + { + "epoch": 0.34937583418387375, + "grad_norm": 5.361997127532959, + "learning_rate": 1.8773337666054895e-05, + "loss": 6.023746948242188, + "step": 11125 + }, + { + "epoch": 0.35016094841799483, + "grad_norm": 8.568739891052246, + "learning_rate": 1.876769345207429e-05, + "loss": 6.047612915039062, + "step": 11150 + }, + { + "epoch": 0.35094606265211586, + "grad_norm": 6.879204750061035, + "learning_rate": 1.8762037236744867e-05, + "loss": 6.037516479492187, + "step": 11175 + }, + { + "epoch": 0.35173117688623695, + "grad_norm": 7.195074558258057, + "learning_rate": 1.875636902880568e-05, + "loss": 6.029046630859375, + "step": 11200 + }, + { + "epoch": 0.35251629112035804, + "grad_norm": 6.277739524841309, + "learning_rate": 1.875068883701429e-05, + "loss": 6.054810791015625, + "step": 11225 + }, + { + "epoch": 0.35330140535447907, + "grad_norm": 8.48022174835205, + "learning_rate": 1.874499667014679e-05, + "loss": 6.055784912109375, + "step": 11250 + }, + { + "epoch": 0.35408651958860016, + "grad_norm": 7.358577728271484, + "learning_rate": 1.8739292536997767e-05, + "loss": 6.0368505859375, + "step": 11275 + }, + { + "epoch": 0.3548716338227212, + "grad_norm": 6.124172687530518, + "learning_rate": 1.873357644638031e-05, + "loss": 6.033704833984375, + "step": 11300 + }, + { + "epoch": 0.3556567480568423, + "grad_norm": 7.543927192687988, + "learning_rate": 1.8727848407125958e-05, + "loss": 6.044652709960937, + "step": 11325 + }, + { + "epoch": 0.35644186229096336, + "grad_norm": 7.918491363525391, + "learning_rate": 1.8722108428084732e-05, + "loss": 6.035922241210938, + "step": 11350 + }, + { + "epoch": 0.3572269765250844, + "grad_norm": 6.090640544891357, + "learning_rate": 1.8716356518125096e-05, + "loss": 6.028107299804687, + "step": 11375 + }, + { + "epoch": 0.3580120907592055, + "grad_norm": 7.507791519165039, + "learning_rate": 1.8710592686133945e-05, + "loss": 6.053147583007813, + "step": 11400 + }, + { + "epoch": 0.3587972049933265, + "grad_norm": 7.755253791809082, + "learning_rate": 1.8704816941016584e-05, + "loss": 6.028209838867188, + "step": 11425 + }, + { + "epoch": 0.3595823192274476, + "grad_norm": 6.474155902862549, + "learning_rate": 1.8699029291696745e-05, + "loss": 6.020354614257813, + "step": 11450 + }, + { + "epoch": 0.36036743346156863, + "grad_norm": 7.208597183227539, + "learning_rate": 1.869322974711654e-05, + "loss": 6.040818481445313, + "step": 11475 + }, + { + "epoch": 0.3611525476956897, + "grad_norm": 6.3050312995910645, + "learning_rate": 1.868741831623646e-05, + "loss": 6.018604125976562, + "step": 11500 + }, + { + "epoch": 0.3619376619298108, + "grad_norm": 7.102381706237793, + "learning_rate": 1.868159500803536e-05, + "loss": 6.034991455078125, + "step": 11525 + }, + { + "epoch": 0.36272277616393184, + "grad_norm": 8.841017723083496, + "learning_rate": 1.8675759831510448e-05, + "loss": 6.02751708984375, + "step": 11550 + }, + { + "epoch": 0.3635078903980529, + "grad_norm": 8.08227252960205, + "learning_rate": 1.8669912795677274e-05, + "loss": 6.046445922851563, + "step": 11575 + }, + { + "epoch": 0.36429300463217396, + "grad_norm": 7.369481086730957, + "learning_rate": 1.8664053909569703e-05, + "loss": 6.070865478515625, + "step": 11600 + }, + { + "epoch": 0.36507811886629504, + "grad_norm": 7.1761932373046875, + "learning_rate": 1.865818318223991e-05, + "loss": 6.048270263671875, + "step": 11625 + }, + { + "epoch": 0.36586323310041613, + "grad_norm": 6.302981853485107, + "learning_rate": 1.865230062275837e-05, + "loss": 6.007454223632813, + "step": 11650 + }, + { + "epoch": 0.36664834733453716, + "grad_norm": 6.380688667297363, + "learning_rate": 1.8646406240213836e-05, + "loss": 6.001620483398438, + "step": 11675 + }, + { + "epoch": 0.36743346156865825, + "grad_norm": 7.499889850616455, + "learning_rate": 1.8640500043713323e-05, + "loss": 6.045592041015625, + "step": 11700 + }, + { + "epoch": 0.3682185758027793, + "grad_norm": 6.4084792137146, + "learning_rate": 1.863458204238211e-05, + "loss": 6.019312133789063, + "step": 11725 + }, + { + "epoch": 0.36900369003690037, + "grad_norm": 7.318563938140869, + "learning_rate": 1.862865224536371e-05, + "loss": 6.043784790039062, + "step": 11750 + }, + { + "epoch": 0.36978880427102145, + "grad_norm": 6.372910499572754, + "learning_rate": 1.862271066181985e-05, + "loss": 6.01970947265625, + "step": 11775 + }, + { + "epoch": 0.3705739185051425, + "grad_norm": 9.704291343688965, + "learning_rate": 1.861675730093049e-05, + "loss": 6.043094482421875, + "step": 11800 + }, + { + "epoch": 0.3713590327392636, + "grad_norm": 8.120965003967285, + "learning_rate": 1.8610792171893766e-05, + "loss": 6.03557373046875, + "step": 11825 + }, + { + "epoch": 0.3721441469733846, + "grad_norm": 7.747727394104004, + "learning_rate": 1.860481528392601e-05, + "loss": 6.03875, + "step": 11850 + }, + { + "epoch": 0.3729292612075057, + "grad_norm": 7.511434555053711, + "learning_rate": 1.8598826646261704e-05, + "loss": 6.04180908203125, + "step": 11875 + }, + { + "epoch": 0.3737143754416268, + "grad_norm": 6.001734256744385, + "learning_rate": 1.8592826268153504e-05, + "loss": 6.0308905029296875, + "step": 11900 + }, + { + "epoch": 0.3744994896757478, + "grad_norm": 5.927894115447998, + "learning_rate": 1.8586814158872198e-05, + "loss": 6.038502197265625, + "step": 11925 + }, + { + "epoch": 0.3752846039098689, + "grad_norm": 6.034798622131348, + "learning_rate": 1.85807903277067e-05, + "loss": 6.048773765563965, + "step": 11950 + }, + { + "epoch": 0.37606971814398993, + "grad_norm": 6.92681360244751, + "learning_rate": 1.8574754783964024e-05, + "loss": 6.027759399414062, + "step": 11975 + }, + { + "epoch": 0.376854832378111, + "grad_norm": 7.4112138748168945, + "learning_rate": 1.8568707536969296e-05, + "loss": 6.021046142578125, + "step": 12000 + }, + { + "epoch": 0.3776399466122321, + "grad_norm": 7.190526962280273, + "learning_rate": 1.8562648596065713e-05, + "loss": 6.0243902587890625, + "step": 12025 + }, + { + "epoch": 0.37842506084635313, + "grad_norm": 6.987718105316162, + "learning_rate": 1.8556577970614544e-05, + "loss": 6.018140869140625, + "step": 12050 + }, + { + "epoch": 0.3792101750804742, + "grad_norm": 8.24082088470459, + "learning_rate": 1.8550495669995114e-05, + "loss": 6.047182006835937, + "step": 12075 + }, + { + "epoch": 0.37999528931459525, + "grad_norm": 7.611581325531006, + "learning_rate": 1.854440170360478e-05, + "loss": 6.032937622070312, + "step": 12100 + }, + { + "epoch": 0.38078040354871634, + "grad_norm": 6.780263423919678, + "learning_rate": 1.853829608085892e-05, + "loss": 6.036834106445313, + "step": 12125 + }, + { + "epoch": 0.3815655177828374, + "grad_norm": 9.1497220993042, + "learning_rate": 1.8532178811190943e-05, + "loss": 6.031931762695312, + "step": 12150 + }, + { + "epoch": 0.38235063201695846, + "grad_norm": 7.544963359832764, + "learning_rate": 1.8526049904052223e-05, + "loss": 6.047650146484375, + "step": 12175 + }, + { + "epoch": 0.38313574625107955, + "grad_norm": 9.74996280670166, + "learning_rate": 1.8519909368912135e-05, + "loss": 6.048277587890625, + "step": 12200 + }, + { + "epoch": 0.3839208604852006, + "grad_norm": 7.021296977996826, + "learning_rate": 1.851375721525801e-05, + "loss": 6.027951049804687, + "step": 12225 + }, + { + "epoch": 0.38470597471932166, + "grad_norm": 7.319636344909668, + "learning_rate": 1.8507593452595138e-05, + "loss": 6.046517944335937, + "step": 12250 + }, + { + "epoch": 0.38549108895344275, + "grad_norm": 7.183573246002197, + "learning_rate": 1.8501418090446735e-05, + "loss": 6.020502319335938, + "step": 12275 + }, + { + "epoch": 0.3862762031875638, + "grad_norm": 5.594688415527344, + "learning_rate": 1.8495231138353945e-05, + "loss": 6.01718505859375, + "step": 12300 + }, + { + "epoch": 0.38706131742168487, + "grad_norm": 7.139080047607422, + "learning_rate": 1.8489032605875826e-05, + "loss": 6.040979614257813, + "step": 12325 + }, + { + "epoch": 0.3878464316558059, + "grad_norm": 6.538236618041992, + "learning_rate": 1.848282250258931e-05, + "loss": 6.025040283203125, + "step": 12350 + }, + { + "epoch": 0.388631545889927, + "grad_norm": 5.932796001434326, + "learning_rate": 1.847660083808922e-05, + "loss": 6.040492553710937, + "step": 12375 + }, + { + "epoch": 0.3894166601240481, + "grad_norm": 7.121323108673096, + "learning_rate": 1.847036762198824e-05, + "loss": 6.037491455078125, + "step": 12400 + }, + { + "epoch": 0.3902017743581691, + "grad_norm": 4.944361209869385, + "learning_rate": 1.8464122863916894e-05, + "loss": 6.050401611328125, + "step": 12425 + }, + { + "epoch": 0.3909868885922902, + "grad_norm": 7.780629634857178, + "learning_rate": 1.8457866573523554e-05, + "loss": 6.035247802734375, + "step": 12450 + }, + { + "epoch": 0.3917720028264112, + "grad_norm": 8.498658180236816, + "learning_rate": 1.845159876047439e-05, + "loss": 6.028329467773437, + "step": 12475 + }, + { + "epoch": 0.3925571170605323, + "grad_norm": 7.45485782623291, + "learning_rate": 1.8445319434453386e-05, + "loss": 6.032855224609375, + "step": 12500 + }, + { + "epoch": 0.3933422312946534, + "grad_norm": 6.773138999938965, + "learning_rate": 1.843902860516232e-05, + "loss": 6.026155395507812, + "step": 12525 + }, + { + "epoch": 0.39412734552877443, + "grad_norm": 6.737361431121826, + "learning_rate": 1.8432726282320728e-05, + "loss": 6.0356884765625, + "step": 12550 + }, + { + "epoch": 0.3949124597628955, + "grad_norm": 6.8191399574279785, + "learning_rate": 1.842641247566592e-05, + "loss": 6.0403759765625, + "step": 12575 + }, + { + "epoch": 0.39569757399701655, + "grad_norm": 6.419453144073486, + "learning_rate": 1.842008719495293e-05, + "loss": 6.029004516601563, + "step": 12600 + }, + { + "epoch": 0.39648268823113764, + "grad_norm": 7.009780406951904, + "learning_rate": 1.841375044995454e-05, + "loss": 6.020476684570313, + "step": 12625 + }, + { + "epoch": 0.39726780246525867, + "grad_norm": 6.108678340911865, + "learning_rate": 1.8407402250461227e-05, + "loss": 6.046521606445313, + "step": 12650 + }, + { + "epoch": 0.39805291669937976, + "grad_norm": 6.681120872497559, + "learning_rate": 1.8401042606281186e-05, + "loss": 6.026983642578125, + "step": 12675 + }, + { + "epoch": 0.39883803093350084, + "grad_norm": 7.839036464691162, + "learning_rate": 1.8394671527240265e-05, + "loss": 6.036826171875, + "step": 12700 + }, + { + "epoch": 0.3996231451676219, + "grad_norm": 6.829401016235352, + "learning_rate": 1.8388289023182012e-05, + "loss": 6.017455444335938, + "step": 12725 + }, + { + "epoch": 0.40040825940174296, + "grad_norm": 6.768727779388428, + "learning_rate": 1.8381895103967603e-05, + "loss": 6.0541162109375, + "step": 12750 + }, + { + "epoch": 0.401193373635864, + "grad_norm": 6.770383834838867, + "learning_rate": 1.837548977947586e-05, + "loss": 6.035471801757812, + "step": 12775 + }, + { + "epoch": 0.4019784878699851, + "grad_norm": 5.607656002044678, + "learning_rate": 1.8369073059603228e-05, + "loss": 6.0449365234375, + "step": 12800 + }, + { + "epoch": 0.40276360210410617, + "grad_norm": 7.781570911407471, + "learning_rate": 1.836264495426375e-05, + "loss": 6.017721557617188, + "step": 12825 + }, + { + "epoch": 0.4035487163382272, + "grad_norm": 10.902204513549805, + "learning_rate": 1.8356205473389075e-05, + "loss": 6.025816650390625, + "step": 12850 + }, + { + "epoch": 0.4043338305723483, + "grad_norm": 7.106377601623535, + "learning_rate": 1.8349754626928413e-05, + "loss": 6.013694458007812, + "step": 12875 + }, + { + "epoch": 0.4051189448064693, + "grad_norm": 5.889981746673584, + "learning_rate": 1.8343292424848544e-05, + "loss": 6.01642333984375, + "step": 12900 + }, + { + "epoch": 0.4059040590405904, + "grad_norm": 6.696752071380615, + "learning_rate": 1.833681887713378e-05, + "loss": 6.0293798828125, + "step": 12925 + }, + { + "epoch": 0.4066891732747115, + "grad_norm": 7.501461029052734, + "learning_rate": 1.833033399378598e-05, + "loss": 6.048368600698618, + "step": 12950 + }, + { + "epoch": 0.4074742875088325, + "grad_norm": 7.523172855377197, + "learning_rate": 1.8323837784824503e-05, + "loss": 6.053330688476563, + "step": 12975 + }, + { + "epoch": 0.4082594017429536, + "grad_norm": 6.536755561828613, + "learning_rate": 1.8317330260286215e-05, + "loss": 6.026072387695312, + "step": 13000 + }, + { + "epoch": 0.40904451597707464, + "grad_norm": 7.596675872802734, + "learning_rate": 1.8310811430225453e-05, + "loss": 6.011339721679687, + "step": 13025 + }, + { + "epoch": 0.40982963021119573, + "grad_norm": 6.989407062530518, + "learning_rate": 1.830428130471404e-05, + "loss": 6.031344604492188, + "step": 13050 + }, + { + "epoch": 0.4106147444453168, + "grad_norm": 8.470300674438477, + "learning_rate": 1.8297739893841234e-05, + "loss": 6.028606567382813, + "step": 13075 + }, + { + "epoch": 0.41139985867943785, + "grad_norm": 6.645995140075684, + "learning_rate": 1.829118720771374e-05, + "loss": 6.027051391601563, + "step": 13100 + }, + { + "epoch": 0.41218497291355893, + "grad_norm": 7.257071018218994, + "learning_rate": 1.8284623256455675e-05, + "loss": 6.024285888671875, + "step": 13125 + }, + { + "epoch": 0.41297008714767997, + "grad_norm": 7.008105278015137, + "learning_rate": 1.827804805020857e-05, + "loss": 6.007571411132813, + "step": 13150 + }, + { + "epoch": 0.41375520138180105, + "grad_norm": 7.29708194732666, + "learning_rate": 1.8271461599131336e-05, + "loss": 6.033352661132812, + "step": 13175 + }, + { + "epoch": 0.41454031561592214, + "grad_norm": 6.295455455780029, + "learning_rate": 1.8264863913400266e-05, + "loss": 6.002512817382812, + "step": 13200 + }, + { + "epoch": 0.41532542985004317, + "grad_norm": 8.981111526489258, + "learning_rate": 1.825825500320901e-05, + "loss": 6.013607177734375, + "step": 13225 + }, + { + "epoch": 0.41611054408416426, + "grad_norm": 7.749846458435059, + "learning_rate": 1.8251634878768547e-05, + "loss": 6.015499877929687, + "step": 13250 + }, + { + "epoch": 0.4168956583182853, + "grad_norm": 7.727872848510742, + "learning_rate": 1.8245003550307206e-05, + "loss": 6.0066259765625, + "step": 13275 + }, + { + "epoch": 0.4176807725524064, + "grad_norm": 4.966046333312988, + "learning_rate": 1.8238361028070602e-05, + "loss": 6.0255499267578125, + "step": 13300 + }, + { + "epoch": 0.41846588678652746, + "grad_norm": 8.59343433380127, + "learning_rate": 1.8231707322321667e-05, + "loss": 6.009500122070312, + "step": 13325 + }, + { + "epoch": 0.4192510010206485, + "grad_norm": 7.226011753082275, + "learning_rate": 1.822504244334059e-05, + "loss": 6.01266357421875, + "step": 13350 + }, + { + "epoch": 0.4200361152547696, + "grad_norm": 7.133829593658447, + "learning_rate": 1.8218366401424843e-05, + "loss": 6.027396850585937, + "step": 13375 + }, + { + "epoch": 0.4208212294888906, + "grad_norm": 9.130123138427734, + "learning_rate": 1.8211679206889135e-05, + "loss": 6.032857055664063, + "step": 13400 + }, + { + "epoch": 0.4216063437230117, + "grad_norm": 6.071317672729492, + "learning_rate": 1.8204980870065404e-05, + "loss": 6.026655883789062, + "step": 13425 + }, + { + "epoch": 0.4223914579571328, + "grad_norm": 9.760759353637695, + "learning_rate": 1.8198271401302806e-05, + "loss": 6.01298828125, + "step": 13450 + }, + { + "epoch": 0.4231765721912538, + "grad_norm": 6.463744163513184, + "learning_rate": 1.81915508109677e-05, + "loss": 6.0219091796875, + "step": 13475 + }, + { + "epoch": 0.4239616864253749, + "grad_norm": 7.907313346862793, + "learning_rate": 1.8184819109443627e-05, + "loss": 6.009920043945312, + "step": 13500 + }, + { + "epoch": 0.42474680065949594, + "grad_norm": 6.163563251495361, + "learning_rate": 1.817807630713129e-05, + "loss": 5.997871704101563, + "step": 13525 + }, + { + "epoch": 0.425531914893617, + "grad_norm": 6.326579570770264, + "learning_rate": 1.817132241444855e-05, + "loss": 6.001806640625, + "step": 13550 + }, + { + "epoch": 0.4263170291277381, + "grad_norm": 7.002986907958984, + "learning_rate": 1.816455744183039e-05, + "loss": 6.025028686523438, + "step": 13575 + }, + { + "epoch": 0.42710214336185914, + "grad_norm": 7.135458469390869, + "learning_rate": 1.815778139972893e-05, + "loss": 6.017838134765625, + "step": 13600 + }, + { + "epoch": 0.42788725759598023, + "grad_norm": 7.496002674102783, + "learning_rate": 1.8150994298613385e-05, + "loss": 6.009573364257813, + "step": 13625 + }, + { + "epoch": 0.42867237183010126, + "grad_norm": 6.810083866119385, + "learning_rate": 1.8144196148970047e-05, + "loss": 6.0144110107421875, + "step": 13650 + }, + { + "epoch": 0.42945748606422235, + "grad_norm": 7.642554759979248, + "learning_rate": 1.8137386961302298e-05, + "loss": 6.0398626708984375, + "step": 13675 + }, + { + "epoch": 0.43024260029834344, + "grad_norm": 5.227537155151367, + "learning_rate": 1.8130566746130557e-05, + "loss": 6.030364379882813, + "step": 13700 + }, + { + "epoch": 0.43102771453246447, + "grad_norm": 8.58128833770752, + "learning_rate": 1.812373551399228e-05, + "loss": 5.996229858398437, + "step": 13725 + }, + { + "epoch": 0.43181282876658555, + "grad_norm": 5.801585674285889, + "learning_rate": 1.8116893275441965e-05, + "loss": 5.98702392578125, + "step": 13750 + }, + { + "epoch": 0.4325979430007066, + "grad_norm": 8.644152641296387, + "learning_rate": 1.8110040041051093e-05, + "loss": 6.015497436523438, + "step": 13775 + }, + { + "epoch": 0.4333830572348277, + "grad_norm": 5.409648895263672, + "learning_rate": 1.8103175821408142e-05, + "loss": 5.99656982421875, + "step": 13800 + }, + { + "epoch": 0.4341681714689487, + "grad_norm": 7.467557430267334, + "learning_rate": 1.8096300627118567e-05, + "loss": 6.060056762695313, + "step": 13825 + }, + { + "epoch": 0.4349532857030698, + "grad_norm": 7.186622142791748, + "learning_rate": 1.808941446880478e-05, + "loss": 6.014412841796875, + "step": 13850 + }, + { + "epoch": 0.4357383999371909, + "grad_norm": 5.200715065002441, + "learning_rate": 1.8082517357106117e-05, + "loss": 6.021922607421875, + "step": 13875 + }, + { + "epoch": 0.4365235141713119, + "grad_norm": 5.337906360626221, + "learning_rate": 1.8075609302678857e-05, + "loss": 6.012916259765625, + "step": 13900 + }, + { + "epoch": 0.437308628405433, + "grad_norm": 7.45744514465332, + "learning_rate": 1.8068690316196174e-05, + "loss": 6.014993286132812, + "step": 13925 + }, + { + "epoch": 0.43809374263955403, + "grad_norm": 5.522850036621094, + "learning_rate": 1.8061760408348137e-05, + "loss": 5.997562408447266, + "step": 13950 + }, + { + "epoch": 0.4388788568736751, + "grad_norm": 6.472789287567139, + "learning_rate": 1.8054819589841687e-05, + "loss": 6.033994750976563, + "step": 13975 + }, + { + "epoch": 0.4396639711077962, + "grad_norm": 6.895258903503418, + "learning_rate": 1.8047867871400625e-05, + "loss": 6.0125048828125, + "step": 14000 + }, + { + "epoch": 0.44044908534191723, + "grad_norm": 7.945712566375732, + "learning_rate": 1.8040905263765588e-05, + "loss": 6.0253076171875, + "step": 14025 + }, + { + "epoch": 0.4412341995760383, + "grad_norm": 6.004476547241211, + "learning_rate": 1.8033931777694042e-05, + "loss": 6.061200561523438, + "step": 14050 + }, + { + "epoch": 0.44201931381015935, + "grad_norm": 7.918838024139404, + "learning_rate": 1.8026947423960258e-05, + "loss": 6.043032836914063, + "step": 14075 + }, + { + "epoch": 0.44280442804428044, + "grad_norm": 5.808018684387207, + "learning_rate": 1.80199522133553e-05, + "loss": 5.978490600585937, + "step": 14100 + }, + { + "epoch": 0.4435895422784015, + "grad_norm": 5.634881019592285, + "learning_rate": 1.8012946156686998e-05, + "loss": 6.008768920898437, + "step": 14125 + }, + { + "epoch": 0.44437465651252256, + "grad_norm": 6.0006794929504395, + "learning_rate": 1.8005929264779954e-05, + "loss": 6.002869873046875, + "step": 14150 + }, + { + "epoch": 0.44515977074664365, + "grad_norm": 6.0581536293029785, + "learning_rate": 1.7998901548475505e-05, + "loss": 6.020526123046875, + "step": 14175 + }, + { + "epoch": 0.4459448849807647, + "grad_norm": 7.235322952270508, + "learning_rate": 1.7991863018631698e-05, + "loss": 5.996553955078125, + "step": 14200 + }, + { + "epoch": 0.44672999921488576, + "grad_norm": 9.50629711151123, + "learning_rate": 1.7984813686123314e-05, + "loss": 6.018790893554687, + "step": 14225 + }, + { + "epoch": 0.44751511344900685, + "grad_norm": 6.713864803314209, + "learning_rate": 1.79777535618418e-05, + "loss": 6.007920532226563, + "step": 14250 + }, + { + "epoch": 0.4483002276831279, + "grad_norm": 8.611591339111328, + "learning_rate": 1.7970682656695294e-05, + "loss": 6.0057830810546875, + "step": 14275 + }, + { + "epoch": 0.44908534191724897, + "grad_norm": 5.062463283538818, + "learning_rate": 1.7963600981608575e-05, + "loss": 5.98197509765625, + "step": 14300 + }, + { + "epoch": 0.44987045615137, + "grad_norm": 6.618898391723633, + "learning_rate": 1.7956508547523068e-05, + "loss": 5.997550048828125, + "step": 14325 + }, + { + "epoch": 0.4506555703854911, + "grad_norm": 6.035687446594238, + "learning_rate": 1.7949405365396835e-05, + "loss": 6.012372436523438, + "step": 14350 + }, + { + "epoch": 0.4514406846196122, + "grad_norm": 6.016815662384033, + "learning_rate": 1.794229144620453e-05, + "loss": 6.005814819335938, + "step": 14375 + }, + { + "epoch": 0.4522257988537332, + "grad_norm": 7.440488338470459, + "learning_rate": 1.793516680093739e-05, + "loss": 6.013446044921875, + "step": 14400 + }, + { + "epoch": 0.4530109130878543, + "grad_norm": 5.334819793701172, + "learning_rate": 1.7928031440603242e-05, + "loss": 6.011575927734375, + "step": 14425 + }, + { + "epoch": 0.4537960273219753, + "grad_norm": 7.730499267578125, + "learning_rate": 1.792088537622645e-05, + "loss": 6.029010620117187, + "step": 14450 + }, + { + "epoch": 0.4545811415560964, + "grad_norm": 6.4564900398254395, + "learning_rate": 1.7913728618847934e-05, + "loss": 6.026947021484375, + "step": 14475 + }, + { + "epoch": 0.4553662557902175, + "grad_norm": 6.429846286773682, + "learning_rate": 1.790656117952512e-05, + "loss": 6.006602783203125, + "step": 14500 + }, + { + "epoch": 0.45615137002433853, + "grad_norm": 6.599667549133301, + "learning_rate": 1.7899383069331948e-05, + "loss": 6.022766776706861, + "step": 14525 + }, + { + "epoch": 0.4569364842584596, + "grad_norm": 6.249053001403809, + "learning_rate": 1.7892194299358833e-05, + "loss": 5.992688598632813, + "step": 14550 + }, + { + "epoch": 0.45772159849258065, + "grad_norm": 5.462608337402344, + "learning_rate": 1.788499488071267e-05, + "loss": 6.0088568115234375, + "step": 14575 + }, + { + "epoch": 0.45850671272670174, + "grad_norm": 7.769815921783447, + "learning_rate": 1.7877784824516816e-05, + "loss": 5.983077392578125, + "step": 14600 + }, + { + "epoch": 0.4592918269608228, + "grad_norm": 4.935760021209717, + "learning_rate": 1.787056414191104e-05, + "loss": 6.009028930664062, + "step": 14625 + }, + { + "epoch": 0.46007694119494386, + "grad_norm": 6.799657821655273, + "learning_rate": 1.786333284405154e-05, + "loss": 6.016587524414063, + "step": 14650 + }, + { + "epoch": 0.46086205542906494, + "grad_norm": 9.121086120605469, + "learning_rate": 1.7856090942110914e-05, + "loss": 6.021969604492187, + "step": 14675 + }, + { + "epoch": 0.461647169663186, + "grad_norm": 6.207000255584717, + "learning_rate": 1.7848838447278155e-05, + "loss": 6.0174102783203125, + "step": 14700 + }, + { + "epoch": 0.46243228389730706, + "grad_norm": 5.745265960693359, + "learning_rate": 1.7841575370758607e-05, + "loss": 6.0080804443359375, + "step": 14725 + }, + { + "epoch": 0.46321739813142815, + "grad_norm": 7.634740829467773, + "learning_rate": 1.783430172377397e-05, + "loss": 6.003064575195313, + "step": 14750 + }, + { + "epoch": 0.4640025123655492, + "grad_norm": 5.885915756225586, + "learning_rate": 1.7827017517562268e-05, + "loss": 5.993345947265625, + "step": 14775 + }, + { + "epoch": 0.46478762659967027, + "grad_norm": 7.342817783355713, + "learning_rate": 1.7819722763377854e-05, + "loss": 5.9821484375, + "step": 14800 + }, + { + "epoch": 0.4655727408337913, + "grad_norm": 4.743215084075928, + "learning_rate": 1.7812417472491362e-05, + "loss": 5.994312744140625, + "step": 14825 + }, + { + "epoch": 0.4663578550679124, + "grad_norm": 6.582151412963867, + "learning_rate": 1.7805101656189722e-05, + "loss": 5.984220581054688, + "step": 14850 + }, + { + "epoch": 0.4671429693020335, + "grad_norm": 6.497868537902832, + "learning_rate": 1.779777532577611e-05, + "loss": 6.01655517578125, + "step": 14875 + }, + { + "epoch": 0.4679280835361545, + "grad_norm": 6.742717742919922, + "learning_rate": 1.779043849256996e-05, + "loss": 6.00601806640625, + "step": 14900 + }, + { + "epoch": 0.4687131977702756, + "grad_norm": 6.5293731689453125, + "learning_rate": 1.7783091167906922e-05, + "loss": 6.008165893554687, + "step": 14925 + }, + { + "epoch": 0.4694983120043966, + "grad_norm": 6.881898880004883, + "learning_rate": 1.777573336313886e-05, + "loss": 6.025721010954483, + "step": 14950 + }, + { + "epoch": 0.4702834262385177, + "grad_norm": 7.383565425872803, + "learning_rate": 1.776836508963384e-05, + "loss": 5.993233642578125, + "step": 14975 + }, + { + "epoch": 0.47106854047263874, + "grad_norm": 7.164944171905518, + "learning_rate": 1.7760986358776082e-05, + "loss": 5.998822021484375, + "step": 15000 + }, + { + "epoch": 0.47185365470675983, + "grad_norm": 6.21272087097168, + "learning_rate": 1.775359718196598e-05, + "loss": 5.999829711914063, + "step": 15025 + }, + { + "epoch": 0.4726387689408809, + "grad_norm": 6.477439880371094, + "learning_rate": 1.7746197570620064e-05, + "loss": 6.002055053710937, + "step": 15050 + }, + { + "epoch": 0.47342388317500195, + "grad_norm": 5.921806812286377, + "learning_rate": 1.773878753617098e-05, + "loss": 6.00700439453125, + "step": 15075 + }, + { + "epoch": 0.47420899740912303, + "grad_norm": 8.893495559692383, + "learning_rate": 1.7731367090067484e-05, + "loss": 6.0103173828125, + "step": 15100 + }, + { + "epoch": 0.47499411164324407, + "grad_norm": 6.748020172119141, + "learning_rate": 1.7723936243774412e-05, + "loss": 5.996409301757812, + "step": 15125 + }, + { + "epoch": 0.47577922587736515, + "grad_norm": 7.704909324645996, + "learning_rate": 1.7716495008772678e-05, + "loss": 6.028046875, + "step": 15150 + }, + { + "epoch": 0.47656434011148624, + "grad_norm": 5.896365642547607, + "learning_rate": 1.7709043396559237e-05, + "loss": 6.020074462890625, + "step": 15175 + }, + { + "epoch": 0.47734945434560727, + "grad_norm": 7.135949611663818, + "learning_rate": 1.7701581418647086e-05, + "loss": 5.997328491210937, + "step": 15200 + }, + { + "epoch": 0.47813456857972836, + "grad_norm": 6.554275989532471, + "learning_rate": 1.7694109086565224e-05, + "loss": 5.9866650390625, + "step": 15225 + }, + { + "epoch": 0.4789196828138494, + "grad_norm": 6.387217998504639, + "learning_rate": 1.7686626411858666e-05, + "loss": 5.9937200927734375, + "step": 15250 + }, + { + "epoch": 0.4797047970479705, + "grad_norm": 6.937610149383545, + "learning_rate": 1.7679133406088393e-05, + "loss": 5.991025390625, + "step": 15275 + }, + { + "epoch": 0.48048991128209156, + "grad_norm": 5.617008686065674, + "learning_rate": 1.7671630080831355e-05, + "loss": 5.998310546875, + "step": 15300 + }, + { + "epoch": 0.4812750255162126, + "grad_norm": 6.899811744689941, + "learning_rate": 1.766411644768044e-05, + "loss": 6.02084228515625, + "step": 15325 + }, + { + "epoch": 0.4820601397503337, + "grad_norm": 7.302089691162109, + "learning_rate": 1.7656592518244465e-05, + "loss": 5.995045166015625, + "step": 15350 + }, + { + "epoch": 0.4828452539844547, + "grad_norm": 6.471199989318848, + "learning_rate": 1.764905830414816e-05, + "loss": 5.993697509765625, + "step": 15375 + }, + { + "epoch": 0.4836303682185758, + "grad_norm": 8.545948028564453, + "learning_rate": 1.764151381703213e-05, + "loss": 6.0272802734375, + "step": 15400 + }, + { + "epoch": 0.4844154824526969, + "grad_norm": 5.461451053619385, + "learning_rate": 1.7633959068552876e-05, + "loss": 5.991051025390625, + "step": 15425 + }, + { + "epoch": 0.4852005966868179, + "grad_norm": 6.671406269073486, + "learning_rate": 1.762639407038273e-05, + "loss": 5.985828857421875, + "step": 15450 + }, + { + "epoch": 0.485985710920939, + "grad_norm": 6.436744213104248, + "learning_rate": 1.7618818834209872e-05, + "loss": 6.006229248046875, + "step": 15475 + }, + { + "epoch": 0.48677082515506004, + "grad_norm": 5.546392917633057, + "learning_rate": 1.76112333717383e-05, + "loss": 6.017716064453125, + "step": 15500 + }, + { + "epoch": 0.4875559393891811, + "grad_norm": 4.869434833526611, + "learning_rate": 1.760363769468781e-05, + "loss": 5.9818359375, + "step": 15525 + }, + { + "epoch": 0.4883410536233022, + "grad_norm": 3.9381494522094727, + "learning_rate": 1.759603181479397e-05, + "loss": 5.999924341837565, + "step": 15550 + }, + { + "epoch": 0.48912616785742324, + "grad_norm": 7.062398910522461, + "learning_rate": 1.7588415743808136e-05, + "loss": 5.985195922851562, + "step": 15575 + }, + { + "epoch": 0.48991128209154433, + "grad_norm": 6.727392673492432, + "learning_rate": 1.7580789493497383e-05, + "loss": 5.98890625, + "step": 15600 + }, + { + "epoch": 0.49069639632566536, + "grad_norm": 5.719323635101318, + "learning_rate": 1.757315307564454e-05, + "loss": 6.015670166015625, + "step": 15625 + }, + { + "epoch": 0.49148151055978645, + "grad_norm": 5.263721942901611, + "learning_rate": 1.756550650204811e-05, + "loss": 6.005720825195312, + "step": 15650 + }, + { + "epoch": 0.49226662479390754, + "grad_norm": 6.878104209899902, + "learning_rate": 1.7557849784522323e-05, + "loss": 5.999227905273438, + "step": 15675 + }, + { + "epoch": 0.49305173902802857, + "grad_norm": 7.825394153594971, + "learning_rate": 1.755018293489706e-05, + "loss": 6.00095947265625, + "step": 15700 + }, + { + "epoch": 0.49383685326214966, + "grad_norm": 5.97175407409668, + "learning_rate": 1.7542505965017868e-05, + "loss": 5.99855712890625, + "step": 15725 + }, + { + "epoch": 0.4946219674962707, + "grad_norm": 6.147233963012695, + "learning_rate": 1.753481888674592e-05, + "loss": 5.990853271484375, + "step": 15750 + }, + { + "epoch": 0.4954070817303918, + "grad_norm": 6.691616058349609, + "learning_rate": 1.7527121711958014e-05, + "loss": 6.009390869140625, + "step": 15775 + }, + { + "epoch": 0.49619219596451286, + "grad_norm": 5.827635288238525, + "learning_rate": 1.7519414452546544e-05, + "loss": 5.983176879882812, + "step": 15800 + }, + { + "epoch": 0.4969773101986339, + "grad_norm": 7.527554512023926, + "learning_rate": 1.7511697120419485e-05, + "loss": 6.008059692382813, + "step": 15825 + }, + { + "epoch": 0.497762424432755, + "grad_norm": 5.735544204711914, + "learning_rate": 1.750396972750038e-05, + "loss": 6.0194329833984375, + "step": 15850 + }, + { + "epoch": 0.498547538666876, + "grad_norm": 6.6720967292785645, + "learning_rate": 1.7496232285728304e-05, + "loss": 6.004626187411222, + "step": 15875 + }, + { + "epoch": 0.4993326529009971, + "grad_norm": 7.539514064788818, + "learning_rate": 1.7488484807057878e-05, + "loss": 5.983084106445313, + "step": 15900 + }, + { + "epoch": 0.5000235534270236, + "eval_loss": 0.1775270402431488, + "eval_runtime": 5123.9315, + "eval_samples_per_second": 4.231, + "eval_steps_per_second": 0.264, + "eval_wer": 6.022335823076517, + "step": 15922 + }, + { + "epoch": 0.5001177671351181, + "grad_norm": 6.7327117919921875, + "learning_rate": 1.7480727303459207e-05, + "loss": 5.9751942952473955, + "step": 15925 + }, + { + "epoch": 0.5009028813692392, + "grad_norm": 4.340229511260986, + "learning_rate": 1.74729597869179e-05, + "loss": 5.9837103271484375, + "step": 15950 + }, + { + "epoch": 0.5016879956033603, + "grad_norm": 5.466999053955078, + "learning_rate": 1.746518226943504e-05, + "loss": 5.9746142578125, + "step": 15975 + }, + { + "epoch": 0.5024731098374814, + "grad_norm": 6.748511791229248, + "learning_rate": 1.745739476302714e-05, + "loss": 5.984482421875, + "step": 16000 + }, + { + "epoch": 0.5032582240716024, + "grad_norm": 5.161959171295166, + "learning_rate": 1.7449597279726168e-05, + "loss": 6.00741455078125, + "step": 16025 + }, + { + "epoch": 0.5040433383057235, + "grad_norm": 10.300546646118164, + "learning_rate": 1.74417898315795e-05, + "loss": 5.9782037353515625, + "step": 16050 + }, + { + "epoch": 0.5048284525398445, + "grad_norm": 5.949592590332031, + "learning_rate": 1.7433972430649914e-05, + "loss": 6.000860595703125, + "step": 16075 + }, + { + "epoch": 0.5056135667739656, + "grad_norm": 4.501881122589111, + "learning_rate": 1.742614508901554e-05, + "loss": 5.982913208007813, + "step": 16100 + }, + { + "epoch": 0.5063986810080867, + "grad_norm": 5.641136646270752, + "learning_rate": 1.7418307818769905e-05, + "loss": 6.000094604492188, + "step": 16125 + }, + { + "epoch": 0.5071837952422077, + "grad_norm": 6.897375583648682, + "learning_rate": 1.7410460632021842e-05, + "loss": 5.970610961914063, + "step": 16150 + }, + { + "epoch": 0.5079689094763288, + "grad_norm": 7.0756731033325195, + "learning_rate": 1.7402603540895528e-05, + "loss": 6.0036181640625, + "step": 16175 + }, + { + "epoch": 0.5087540237104499, + "grad_norm": 6.198889255523682, + "learning_rate": 1.739473655753043e-05, + "loss": 5.984674682617188, + "step": 16200 + }, + { + "epoch": 0.509539137944571, + "grad_norm": 6.3933820724487305, + "learning_rate": 1.7386859694081303e-05, + "loss": 5.977184448242188, + "step": 16225 + }, + { + "epoch": 0.510324252178692, + "grad_norm": 6.183777809143066, + "learning_rate": 1.7378972962718174e-05, + "loss": 5.993760375976563, + "step": 16250 + }, + { + "epoch": 0.511109366412813, + "grad_norm": 7.4453864097595215, + "learning_rate": 1.7371076375626296e-05, + "loss": 6.005464477539062, + "step": 16275 + }, + { + "epoch": 0.5118944806469341, + "grad_norm": 6.7400383949279785, + "learning_rate": 1.7363169945006174e-05, + "loss": 5.991670532226562, + "step": 16300 + }, + { + "epoch": 0.5126795948810552, + "grad_norm": 7.428774356842041, + "learning_rate": 1.7355253683073506e-05, + "loss": 5.993043212890625, + "step": 16325 + }, + { + "epoch": 0.5134647091151763, + "grad_norm": 6.046384334564209, + "learning_rate": 1.734732760205919e-05, + "loss": 5.988531494140625, + "step": 16350 + }, + { + "epoch": 0.5142498233492974, + "grad_norm": 6.85564661026001, + "learning_rate": 1.7339391714209278e-05, + "loss": 6.01021484375, + "step": 16375 + }, + { + "epoch": 0.5150349375834183, + "grad_norm": 8.108253479003906, + "learning_rate": 1.7331446031784997e-05, + "loss": 5.971982421875, + "step": 16400 + }, + { + "epoch": 0.5158200518175394, + "grad_norm": 8.373122215270996, + "learning_rate": 1.7323490567062687e-05, + "loss": 5.989812622070312, + "step": 16425 + }, + { + "epoch": 0.5166051660516605, + "grad_norm": 8.284374237060547, + "learning_rate": 1.7315525332333813e-05, + "loss": 5.983814086914062, + "step": 16450 + }, + { + "epoch": 0.5173902802857816, + "grad_norm": 7.153393745422363, + "learning_rate": 1.7307550339904926e-05, + "loss": 5.993232421875, + "step": 16475 + }, + { + "epoch": 0.5181753945199027, + "grad_norm": 5.523364067077637, + "learning_rate": 1.7299565602097667e-05, + "loss": 5.973712768554687, + "step": 16500 + }, + { + "epoch": 0.5189605087540237, + "grad_norm": 8.089876174926758, + "learning_rate": 1.729157113124872e-05, + "loss": 5.987085571289063, + "step": 16525 + }, + { + "epoch": 0.5197456229881448, + "grad_norm": 6.855902194976807, + "learning_rate": 1.7283566939709816e-05, + "loss": 5.976376953125, + "step": 16550 + }, + { + "epoch": 0.5205307372222658, + "grad_norm": 5.136626243591309, + "learning_rate": 1.72755530398477e-05, + "loss": 5.991744384765625, + "step": 16575 + }, + { + "epoch": 0.5213158514563869, + "grad_norm": 5.694565773010254, + "learning_rate": 1.7267529444044117e-05, + "loss": 5.962613525390625, + "step": 16600 + }, + { + "epoch": 0.522100965690508, + "grad_norm": 9.666460037231445, + "learning_rate": 1.725949616469579e-05, + "loss": 5.984867553710938, + "step": 16625 + }, + { + "epoch": 0.522886079924629, + "grad_norm": 5.078762531280518, + "learning_rate": 1.725145321421442e-05, + "loss": 5.989614868164063, + "step": 16650 + }, + { + "epoch": 0.5236711941587501, + "grad_norm": 6.477517127990723, + "learning_rate": 1.7243400605026623e-05, + "loss": 6.0136175537109375, + "step": 16675 + }, + { + "epoch": 0.5244563083928712, + "grad_norm": 4.204240322113037, + "learning_rate": 1.723533834957396e-05, + "loss": 5.96806884765625, + "step": 16700 + }, + { + "epoch": 0.5252414226269922, + "grad_norm": 7.016325950622559, + "learning_rate": 1.7227266460312883e-05, + "loss": 5.966884765625, + "step": 16725 + }, + { + "epoch": 0.5260265368611133, + "grad_norm": 5.886275291442871, + "learning_rate": 1.721918494971474e-05, + "loss": 6.015527954101563, + "step": 16750 + }, + { + "epoch": 0.5268116510952343, + "grad_norm": 5.597747802734375, + "learning_rate": 1.7211093830265733e-05, + "loss": 5.974203491210938, + "step": 16775 + }, + { + "epoch": 0.5275967653293554, + "grad_norm": 6.396210193634033, + "learning_rate": 1.7202993114466916e-05, + "loss": 5.992606811523437, + "step": 16800 + }, + { + "epoch": 0.5283818795634765, + "grad_norm": 8.385923385620117, + "learning_rate": 1.719488281483417e-05, + "loss": 5.96287841796875, + "step": 16825 + }, + { + "epoch": 0.5291669937975976, + "grad_norm": 7.3003668785095215, + "learning_rate": 1.718676294389818e-05, + "loss": 5.97940185546875, + "step": 16850 + }, + { + "epoch": 0.5299521080317187, + "grad_norm": 5.827690124511719, + "learning_rate": 1.7178633514204424e-05, + "loss": 5.982298583984375, + "step": 16875 + }, + { + "epoch": 0.5307372222658396, + "grad_norm": 5.305222511291504, + "learning_rate": 1.717049453831314e-05, + "loss": 5.996360473632812, + "step": 16900 + }, + { + "epoch": 0.5315223364999607, + "grad_norm": 5.9110331535339355, + "learning_rate": 1.7162346028799326e-05, + "loss": 5.9729681396484375, + "step": 16925 + }, + { + "epoch": 0.5323074507340818, + "grad_norm": 6.23408317565918, + "learning_rate": 1.7154187998252703e-05, + "loss": 5.97909423828125, + "step": 16950 + }, + { + "epoch": 0.5330925649682029, + "grad_norm": 6.106709957122803, + "learning_rate": 1.7146020459277703e-05, + "loss": 5.972821044921875, + "step": 16975 + }, + { + "epoch": 0.533877679202324, + "grad_norm": 6.928908824920654, + "learning_rate": 1.713784342449345e-05, + "loss": 5.999638671875, + "step": 17000 + }, + { + "epoch": 0.534662793436445, + "grad_norm": 5.384850025177002, + "learning_rate": 1.7129656906533738e-05, + "loss": 5.98822021484375, + "step": 17025 + }, + { + "epoch": 0.535447907670566, + "grad_norm": 7.443471908569336, + "learning_rate": 1.7121460918047016e-05, + "loss": 5.98213623046875, + "step": 17050 + }, + { + "epoch": 0.5362330219046871, + "grad_norm": 6.532201766967773, + "learning_rate": 1.7113255471696363e-05, + "loss": 5.99128173828125, + "step": 17075 + }, + { + "epoch": 0.5370181361388082, + "grad_norm": 6.973550319671631, + "learning_rate": 1.7105040580159468e-05, + "loss": 6.014287719726562, + "step": 17100 + }, + { + "epoch": 0.5378032503729293, + "grad_norm": 7.224700927734375, + "learning_rate": 1.709681625612862e-05, + "loss": 5.974287719726562, + "step": 17125 + }, + { + "epoch": 0.5385883646070503, + "grad_norm": 6.451601982116699, + "learning_rate": 1.708858251231067e-05, + "loss": 5.9864501953125, + "step": 17150 + }, + { + "epoch": 0.5393734788411714, + "grad_norm": 7.947690486907959, + "learning_rate": 1.7080339361427044e-05, + "loss": 5.96664794921875, + "step": 17175 + }, + { + "epoch": 0.5401585930752925, + "grad_norm": 5.501924514770508, + "learning_rate": 1.7072086816213674e-05, + "loss": 5.972179565429688, + "step": 17200 + }, + { + "epoch": 0.5409437073094135, + "grad_norm": 4.645030498504639, + "learning_rate": 1.706382488942103e-05, + "loss": 6.00225830078125, + "step": 17225 + }, + { + "epoch": 0.5417288215435346, + "grad_norm": 6.230678081512451, + "learning_rate": 1.7055553593814065e-05, + "loss": 5.969537963867188, + "step": 17250 + }, + { + "epoch": 0.5425139357776556, + "grad_norm": 5.186971664428711, + "learning_rate": 1.7047272942172208e-05, + "loss": 5.9681549072265625, + "step": 17275 + }, + { + "epoch": 0.5432990500117767, + "grad_norm": 7.244428634643555, + "learning_rate": 1.7038982947289354e-05, + "loss": 5.987784423828125, + "step": 17300 + }, + { + "epoch": 0.5440841642458978, + "grad_norm": 5.102660179138184, + "learning_rate": 1.7030683621973813e-05, + "loss": 5.952317504882813, + "step": 17325 + }, + { + "epoch": 0.5448692784800189, + "grad_norm": 5.980892181396484, + "learning_rate": 1.7022374979048335e-05, + "loss": 5.9924774169921875, + "step": 17350 + }, + { + "epoch": 0.54565439271414, + "grad_norm": 7.802984237670898, + "learning_rate": 1.701405703135004e-05, + "loss": 5.98140869140625, + "step": 17375 + }, + { + "epoch": 0.5464395069482609, + "grad_norm": 6.573966026306152, + "learning_rate": 1.7005729791730456e-05, + "loss": 5.983587036132812, + "step": 17400 + }, + { + "epoch": 0.547224621182382, + "grad_norm": 5.843371868133545, + "learning_rate": 1.6997393273055438e-05, + "loss": 6.001084594726563, + "step": 17425 + }, + { + "epoch": 0.5480097354165031, + "grad_norm": 7.278885841369629, + "learning_rate": 1.6989047488205188e-05, + "loss": 5.961531372070312, + "step": 17450 + }, + { + "epoch": 0.5487948496506242, + "grad_norm": 5.740422248840332, + "learning_rate": 1.698069245007423e-05, + "loss": 5.985166015625, + "step": 17475 + }, + { + "epoch": 0.5495799638847453, + "grad_norm": 5.239304542541504, + "learning_rate": 1.697232817157138e-05, + "loss": 5.99416259765625, + "step": 17500 + }, + { + "epoch": 0.5503650781188663, + "grad_norm": 5.947329998016357, + "learning_rate": 1.696395466561973e-05, + "loss": 6.00931884765625, + "step": 17525 + }, + { + "epoch": 0.5511501923529873, + "grad_norm": 6.159853935241699, + "learning_rate": 1.6955571945156627e-05, + "loss": 5.984913330078125, + "step": 17550 + }, + { + "epoch": 0.5519353065871084, + "grad_norm": 4.8231120109558105, + "learning_rate": 1.694718002313366e-05, + "loss": 5.997113647460938, + "step": 17575 + }, + { + "epoch": 0.5527204208212295, + "grad_norm": 4.420044898986816, + "learning_rate": 1.6938778912516635e-05, + "loss": 5.961992797851562, + "step": 17600 + }, + { + "epoch": 0.5535055350553506, + "grad_norm": 6.57444953918457, + "learning_rate": 1.6930368626285545e-05, + "loss": 5.980524291992188, + "step": 17625 + }, + { + "epoch": 0.5542906492894716, + "grad_norm": 7.454888343811035, + "learning_rate": 1.6921949177434573e-05, + "loss": 5.9910650634765625, + "step": 17650 + }, + { + "epoch": 0.5550757635235927, + "grad_norm": 6.050715923309326, + "learning_rate": 1.6913520578972046e-05, + "loss": 5.98554443359375, + "step": 17675 + }, + { + "epoch": 0.5558608777577138, + "grad_norm": 6.887601852416992, + "learning_rate": 1.6905082843920436e-05, + "loss": 5.992676391601562, + "step": 17700 + }, + { + "epoch": 0.5566459919918348, + "grad_norm": 5.23854398727417, + "learning_rate": 1.689663598531633e-05, + "loss": 5.9701025390625, + "step": 17725 + }, + { + "epoch": 0.5574311062259559, + "grad_norm": 7.124228000640869, + "learning_rate": 1.6888180016210405e-05, + "loss": 5.983826904296875, + "step": 17750 + }, + { + "epoch": 0.5582162204600769, + "grad_norm": 5.992212295532227, + "learning_rate": 1.6879714949667433e-05, + "loss": 5.969363403320313, + "step": 17775 + }, + { + "epoch": 0.559001334694198, + "grad_norm": 7.426286220550537, + "learning_rate": 1.687124079876621e-05, + "loss": 5.980809936523437, + "step": 17800 + }, + { + "epoch": 0.5597864489283191, + "grad_norm": 5.820804595947266, + "learning_rate": 1.6862757576599593e-05, + "loss": 5.9843212890625, + "step": 17825 + }, + { + "epoch": 0.5605715631624402, + "grad_norm": 6.804028511047363, + "learning_rate": 1.685426529627445e-05, + "loss": 5.976370849609375, + "step": 17850 + }, + { + "epoch": 0.5613566773965613, + "grad_norm": 5.806180953979492, + "learning_rate": 1.6845763970911634e-05, + "loss": 5.969212036132813, + "step": 17875 + }, + { + "epoch": 0.5621417916306822, + "grad_norm": 4.407258033752441, + "learning_rate": 1.6837253613645982e-05, + "loss": 5.973047485351563, + "step": 17900 + }, + { + "epoch": 0.5629269058648033, + "grad_norm": 6.541548252105713, + "learning_rate": 1.682873423762629e-05, + "loss": 5.9986669921875, + "step": 17925 + }, + { + "epoch": 0.5637120200989244, + "grad_norm": 4.314496040344238, + "learning_rate": 1.6820205856015277e-05, + "loss": 5.998218994140625, + "step": 17950 + }, + { + "epoch": 0.5644971343330455, + "grad_norm": 7.64862060546875, + "learning_rate": 1.6811668481989577e-05, + "loss": 5.957893676757813, + "step": 17975 + }, + { + "epoch": 0.5652822485671665, + "grad_norm": 4.967900276184082, + "learning_rate": 1.680312212873973e-05, + "loss": 5.9678173828125, + "step": 18000 + }, + { + "epoch": 0.5660673628012876, + "grad_norm": 5.962733268737793, + "learning_rate": 1.6794566809470133e-05, + "loss": 5.960451049804687, + "step": 18025 + }, + { + "epoch": 0.5668524770354086, + "grad_norm": 9.126944541931152, + "learning_rate": 1.6786002537399044e-05, + "loss": 5.965335693359375, + "step": 18050 + }, + { + "epoch": 0.5676375912695297, + "grad_norm": 6.437664985656738, + "learning_rate": 1.6777429325758557e-05, + "loss": 5.972857055664062, + "step": 18075 + }, + { + "epoch": 0.5684227055036508, + "grad_norm": 6.7029571533203125, + "learning_rate": 1.6768847187794573e-05, + "loss": 5.965831298828125, + "step": 18100 + }, + { + "epoch": 0.5692078197377718, + "grad_norm": 5.07132625579834, + "learning_rate": 1.676025613676678e-05, + "loss": 5.965888671875, + "step": 18125 + }, + { + "epoch": 0.5699929339718929, + "grad_norm": 7.3259968757629395, + "learning_rate": 1.6751656185948645e-05, + "loss": 5.976241455078125, + "step": 18150 + }, + { + "epoch": 0.570778048206014, + "grad_norm": 5.761317253112793, + "learning_rate": 1.6743047348627384e-05, + "loss": 5.964612426757813, + "step": 18175 + }, + { + "epoch": 0.571563162440135, + "grad_norm": 6.198018550872803, + "learning_rate": 1.6734429638103942e-05, + "loss": 5.966058349609375, + "step": 18200 + }, + { + "epoch": 0.5723482766742561, + "grad_norm": 7.538841724395752, + "learning_rate": 1.672580306769297e-05, + "loss": 5.975484619140625, + "step": 18225 + }, + { + "epoch": 0.5731333909083771, + "grad_norm": 6.728540420532227, + "learning_rate": 1.6717167650722812e-05, + "loss": 5.9976220703125, + "step": 18250 + }, + { + "epoch": 0.5739185051424982, + "grad_norm": 6.197620391845703, + "learning_rate": 1.670852340053548e-05, + "loss": 6.0005322265625, + "step": 18275 + }, + { + "epoch": 0.5747036193766193, + "grad_norm": 7.213536739349365, + "learning_rate": 1.6699870330486632e-05, + "loss": 5.969610595703125, + "step": 18300 + }, + { + "epoch": 0.5754887336107404, + "grad_norm": 8.402562141418457, + "learning_rate": 1.6691208453945552e-05, + "loss": 5.941055908203125, + "step": 18325 + }, + { + "epoch": 0.5762738478448615, + "grad_norm": 5.4232964515686035, + "learning_rate": 1.6682537784295134e-05, + "loss": 5.9587158203125, + "step": 18350 + }, + { + "epoch": 0.5770589620789824, + "grad_norm": 4.672300338745117, + "learning_rate": 1.6673858334931855e-05, + "loss": 5.967249145507813, + "step": 18375 + }, + { + "epoch": 0.5778440763131035, + "grad_norm": 7.926326751708984, + "learning_rate": 1.6665170119265758e-05, + "loss": 5.956943359375, + "step": 18400 + }, + { + "epoch": 0.5786291905472246, + "grad_norm": 6.106409549713135, + "learning_rate": 1.6656473150720424e-05, + "loss": 5.962383422851563, + "step": 18425 + }, + { + "epoch": 0.5794143047813457, + "grad_norm": 6.68604850769043, + "learning_rate": 1.664776744273297e-05, + "loss": 5.968807983398437, + "step": 18450 + }, + { + "epoch": 0.5801994190154668, + "grad_norm": 7.674676895141602, + "learning_rate": 1.663905300875401e-05, + "loss": 5.984287719726563, + "step": 18475 + }, + { + "epoch": 0.5809845332495878, + "grad_norm": 5.729357719421387, + "learning_rate": 1.6630329862247636e-05, + "loss": 5.9693670654296875, + "step": 18500 + }, + { + "epoch": 0.5817696474837089, + "grad_norm": 6.809622764587402, + "learning_rate": 1.6621598016691402e-05, + "loss": 5.973291015625, + "step": 18525 + }, + { + "epoch": 0.5825547617178299, + "grad_norm": 7.206978797912598, + "learning_rate": 1.6612857485576308e-05, + "loss": 5.964166259765625, + "step": 18550 + }, + { + "epoch": 0.583339875951951, + "grad_norm": 5.887747764587402, + "learning_rate": 1.6604108282406766e-05, + "loss": 5.995296020507812, + "step": 18575 + }, + { + "epoch": 0.5841249901860721, + "grad_norm": 7.267726898193359, + "learning_rate": 1.6595350420700594e-05, + "loss": 5.973833618164062, + "step": 18600 + }, + { + "epoch": 0.5849101044201931, + "grad_norm": 5.648615837097168, + "learning_rate": 1.6586583913988986e-05, + "loss": 5.962001953125, + "step": 18625 + }, + { + "epoch": 0.5856952186543142, + "grad_norm": 6.2857584953308105, + "learning_rate": 1.6577808775816485e-05, + "loss": 5.998330688476562, + "step": 18650 + }, + { + "epoch": 0.5864803328884353, + "grad_norm": 5.842519283294678, + "learning_rate": 1.656902501974098e-05, + "loss": 5.98353759765625, + "step": 18675 + }, + { + "epoch": 0.5872654471225563, + "grad_norm": 5.343513488769531, + "learning_rate": 1.656023265933367e-05, + "loss": 5.969229736328125, + "step": 18700 + }, + { + "epoch": 0.5880505613566774, + "grad_norm": 6.921598434448242, + "learning_rate": 1.6551431708179047e-05, + "loss": 5.963245849609375, + "step": 18725 + }, + { + "epoch": 0.5888356755907984, + "grad_norm": 7.193586349487305, + "learning_rate": 1.654262217987488e-05, + "loss": 5.961904907226563, + "step": 18750 + }, + { + "epoch": 0.5896207898249195, + "grad_norm": 6.090300559997559, + "learning_rate": 1.6533804088032187e-05, + "loss": 5.95270751953125, + "step": 18775 + }, + { + "epoch": 0.5904059040590406, + "grad_norm": 7.19282341003418, + "learning_rate": 1.652497744627522e-05, + "loss": 5.9788525390625, + "step": 18800 + }, + { + "epoch": 0.5911910182931617, + "grad_norm": 8.167314529418945, + "learning_rate": 1.651614226824144e-05, + "loss": 5.970059814453125, + "step": 18825 + }, + { + "epoch": 0.5919761325272828, + "grad_norm": 6.778092861175537, + "learning_rate": 1.650729856758149e-05, + "loss": 5.977365112304687, + "step": 18850 + }, + { + "epoch": 0.5927612467614037, + "grad_norm": 5.703238487243652, + "learning_rate": 1.6498446357959187e-05, + "loss": 5.938150634765625, + "step": 18875 + }, + { + "epoch": 0.5935463609955248, + "grad_norm": 7.178860187530518, + "learning_rate": 1.64895856530515e-05, + "loss": 5.976498413085937, + "step": 18900 + }, + { + "epoch": 0.5943314752296459, + "grad_norm": 5.8019561767578125, + "learning_rate": 1.648071646654851e-05, + "loss": 5.963931189643012, + "step": 18925 + }, + { + "epoch": 0.595116589463767, + "grad_norm": 5.785225868225098, + "learning_rate": 1.6471838812153416e-05, + "loss": 5.957450561523437, + "step": 18950 + }, + { + "epoch": 0.5959017036978881, + "grad_norm": 5.308383941650391, + "learning_rate": 1.6462952703582487e-05, + "loss": 5.967897338867187, + "step": 18975 + }, + { + "epoch": 0.5966868179320091, + "grad_norm": 5.468820571899414, + "learning_rate": 1.6454058154565065e-05, + "loss": 5.981375732421875, + "step": 19000 + }, + { + "epoch": 0.5974719321661301, + "grad_norm": 7.921565532684326, + "learning_rate": 1.6445155178843533e-05, + "loss": 5.9771875, + "step": 19025 + }, + { + "epoch": 0.5982570464002512, + "grad_norm": 5.417508602142334, + "learning_rate": 1.6436243790173284e-05, + "loss": 5.961673583984375, + "step": 19050 + }, + { + "epoch": 0.5990421606343723, + "grad_norm": 4.896191596984863, + "learning_rate": 1.6427324002322717e-05, + "loss": 5.959312744140625, + "step": 19075 + }, + { + "epoch": 0.5998272748684934, + "grad_norm": 5.8037919998168945, + "learning_rate": 1.64183958290732e-05, + "loss": 5.985181884765625, + "step": 19100 + }, + { + "epoch": 0.6006123891026144, + "grad_norm": 5.80145263671875, + "learning_rate": 1.640945928421907e-05, + "loss": 5.96064208984375, + "step": 19125 + }, + { + "epoch": 0.6013975033367355, + "grad_norm": 4.268670558929443, + "learning_rate": 1.6400514381567585e-05, + "loss": 5.960164184570313, + "step": 19150 + }, + { + "epoch": 0.6021826175708566, + "grad_norm": 7.407561302185059, + "learning_rate": 1.6391561134938922e-05, + "loss": 5.94114013671875, + "step": 19175 + }, + { + "epoch": 0.6029677318049776, + "grad_norm": 6.710731506347656, + "learning_rate": 1.6382599558166156e-05, + "loss": 5.977440185546875, + "step": 19200 + }, + { + "epoch": 0.6037528460390987, + "grad_norm": 5.982191562652588, + "learning_rate": 1.637362966509522e-05, + "loss": 5.968762817382813, + "step": 19225 + }, + { + "epoch": 0.6045379602732197, + "grad_norm": 5.094526767730713, + "learning_rate": 1.6364651469584904e-05, + "loss": 5.974467163085937, + "step": 19250 + }, + { + "epoch": 0.6053230745073408, + "grad_norm": 6.691288471221924, + "learning_rate": 1.6355664985506823e-05, + "loss": 5.954463500976562, + "step": 19275 + }, + { + "epoch": 0.6061081887414619, + "grad_norm": 7.0572099685668945, + "learning_rate": 1.6346670226745396e-05, + "loss": 5.9692919921875, + "step": 19300 + }, + { + "epoch": 0.606893302975583, + "grad_norm": 7.397186756134033, + "learning_rate": 1.6337667207197833e-05, + "loss": 5.967357177734375, + "step": 19325 + }, + { + "epoch": 0.6076784172097041, + "grad_norm": 5.563709735870361, + "learning_rate": 1.6328655940774102e-05, + "loss": 5.958339233398437, + "step": 19350 + }, + { + "epoch": 0.608463531443825, + "grad_norm": 8.807904243469238, + "learning_rate": 1.631963644139691e-05, + "loss": 5.959872436523438, + "step": 19375 + }, + { + "epoch": 0.6092486456779461, + "grad_norm": 5.945065021514893, + "learning_rate": 1.6310608723001697e-05, + "loss": 5.957438354492187, + "step": 19400 + }, + { + "epoch": 0.6100337599120672, + "grad_norm": 4.4111480712890625, + "learning_rate": 1.6301572799536588e-05, + "loss": 5.94736572265625, + "step": 19425 + }, + { + "epoch": 0.6108188741461883, + "grad_norm": 5.989535331726074, + "learning_rate": 1.629252868496239e-05, + "loss": 5.985905151367188, + "step": 19450 + }, + { + "epoch": 0.6116039883803094, + "grad_norm": 4.587847709655762, + "learning_rate": 1.628347639325256e-05, + "loss": 5.952802124023438, + "step": 19475 + }, + { + "epoch": 0.6123891026144304, + "grad_norm": 4.7660322189331055, + "learning_rate": 1.6274415938393205e-05, + "loss": 5.958681640625, + "step": 19500 + }, + { + "epoch": 0.6131742168485514, + "grad_norm": 5.339101314544678, + "learning_rate": 1.6265347334383027e-05, + "loss": 5.96294189453125, + "step": 19525 + }, + { + "epoch": 0.6139593310826725, + "grad_norm": 7.570058345794678, + "learning_rate": 1.625627059523333e-05, + "loss": 5.9756097412109375, + "step": 19550 + }, + { + "epoch": 0.6147444453167936, + "grad_norm": 6.047725200653076, + "learning_rate": 1.6247185734967976e-05, + "loss": 5.96040283203125, + "step": 19575 + }, + { + "epoch": 0.6155295595509147, + "grad_norm": 6.406645774841309, + "learning_rate": 1.623809276762339e-05, + "loss": 5.949991455078125, + "step": 19600 + }, + { + "epoch": 0.6163146737850357, + "grad_norm": 5.904462814331055, + "learning_rate": 1.6228991707248508e-05, + "loss": 5.969190063476563, + "step": 19625 + }, + { + "epoch": 0.6170997880191568, + "grad_norm": 7.003258228302002, + "learning_rate": 1.621988256790478e-05, + "loss": 5.955230102539063, + "step": 19650 + }, + { + "epoch": 0.6178849022532779, + "grad_norm": 5.269857406616211, + "learning_rate": 1.6210765363666123e-05, + "loss": 5.964899291992188, + "step": 19675 + }, + { + "epoch": 0.6186700164873989, + "grad_norm": 5.02734899520874, + "learning_rate": 1.6201640108618935e-05, + "loss": 5.943043823242188, + "step": 19700 + }, + { + "epoch": 0.61945513072152, + "grad_norm": 6.511659622192383, + "learning_rate": 1.6192506816862045e-05, + "loss": 5.953358764648438, + "step": 19725 + }, + { + "epoch": 0.620240244955641, + "grad_norm": 5.645657062530518, + "learning_rate": 1.6183365502506693e-05, + "loss": 5.9758154296875, + "step": 19750 + }, + { + "epoch": 0.6210253591897621, + "grad_norm": 5.357290267944336, + "learning_rate": 1.617421617967652e-05, + "loss": 5.958326416015625, + "step": 19775 + }, + { + "epoch": 0.6218104734238832, + "grad_norm": 4.816366195678711, + "learning_rate": 1.6165058862507535e-05, + "loss": 5.954892578125, + "step": 19800 + }, + { + "epoch": 0.6225955876580043, + "grad_norm": 6.78482723236084, + "learning_rate": 1.615589356514811e-05, + "loss": 5.9737353515625, + "step": 19825 + }, + { + "epoch": 0.6233807018921254, + "grad_norm": 7.144836902618408, + "learning_rate": 1.6146720301758937e-05, + "loss": 5.9689398193359375, + "step": 19850 + }, + { + "epoch": 0.6241658161262463, + "grad_norm": 6.895575046539307, + "learning_rate": 1.6137539086513014e-05, + "loss": 5.978463745117187, + "step": 19875 + }, + { + "epoch": 0.6249509303603674, + "grad_norm": 5.599066257476807, + "learning_rate": 1.6128349933595625e-05, + "loss": 5.953704223632813, + "step": 19900 + }, + { + "epoch": 0.6257360445944885, + "grad_norm": 5.91379976272583, + "learning_rate": 1.6119152857204334e-05, + "loss": 5.957532348632813, + "step": 19925 + }, + { + "epoch": 0.6265211588286096, + "grad_norm": 5.114150524139404, + "learning_rate": 1.6109947871548925e-05, + "loss": 5.976472778320312, + "step": 19950 + }, + { + "epoch": 0.6273062730627307, + "grad_norm": 5.2865190505981445, + "learning_rate": 1.6100734990851413e-05, + "loss": 5.976328735351562, + "step": 19975 + }, + { + "epoch": 0.6280913872968517, + "grad_norm": 4.922379016876221, + "learning_rate": 1.6091514229346017e-05, + "loss": 5.970821533203125, + "step": 20000 + }, + { + "epoch": 0.6288765015309727, + "grad_norm": 6.995778560638428, + "learning_rate": 1.6082285601279117e-05, + "loss": 5.963858642578125, + "step": 20025 + }, + { + "epoch": 0.6296616157650938, + "grad_norm": 5.878021240234375, + "learning_rate": 1.6073049120909258e-05, + "loss": 5.973790283203125, + "step": 20050 + }, + { + "epoch": 0.6304467299992149, + "grad_norm": 6.894951820373535, + "learning_rate": 1.606380480250711e-05, + "loss": 5.949996337890625, + "step": 20075 + }, + { + "epoch": 0.631231844233336, + "grad_norm": 4.92820930480957, + "learning_rate": 1.605455266035547e-05, + "loss": 5.953661499023437, + "step": 20100 + }, + { + "epoch": 0.632016958467457, + "grad_norm": 5.884142875671387, + "learning_rate": 1.6045292708749193e-05, + "loss": 5.954257202148438, + "step": 20125 + }, + { + "epoch": 0.6328020727015781, + "grad_norm": 7.254075050354004, + "learning_rate": 1.6036024961995233e-05, + "loss": 5.969572143554688, + "step": 20150 + }, + { + "epoch": 0.6335871869356992, + "grad_norm": 6.749610424041748, + "learning_rate": 1.6026749434412557e-05, + "loss": 5.99073486328125, + "step": 20175 + }, + { + "epoch": 0.6343723011698202, + "grad_norm": 5.641422271728516, + "learning_rate": 1.6017466140332184e-05, + "loss": 5.971793212890625, + "step": 20200 + }, + { + "epoch": 0.6351574154039413, + "grad_norm": 6.3761420249938965, + "learning_rate": 1.6008175094097102e-05, + "loss": 5.993739624023437, + "step": 20225 + }, + { + "epoch": 0.6359425296380623, + "grad_norm": 5.769575595855713, + "learning_rate": 1.5998876310062307e-05, + "loss": 5.952071533203125, + "step": 20250 + }, + { + "epoch": 0.6367276438721834, + "grad_norm": 5.04279899597168, + "learning_rate": 1.5989569802594718e-05, + "loss": 5.9339447021484375, + "step": 20275 + }, + { + "epoch": 0.6375127581063045, + "grad_norm": 5.488921165466309, + "learning_rate": 1.5980255586073213e-05, + "loss": 5.968342895507813, + "step": 20300 + }, + { + "epoch": 0.6382978723404256, + "grad_norm": 5.811150074005127, + "learning_rate": 1.597093367488857e-05, + "loss": 5.960331420898438, + "step": 20325 + }, + { + "epoch": 0.6390829865745467, + "grad_norm": 6.963167190551758, + "learning_rate": 1.596160408344346e-05, + "loss": 5.9563446044921875, + "step": 20350 + }, + { + "epoch": 0.6398681008086676, + "grad_norm": 5.9056901931762695, + "learning_rate": 1.5952266826152406e-05, + "loss": 5.98808349609375, + "step": 20375 + }, + { + "epoch": 0.6406532150427887, + "grad_norm": 7.787331581115723, + "learning_rate": 1.59429219174418e-05, + "loss": 5.960795288085937, + "step": 20400 + }, + { + "epoch": 0.6414383292769098, + "grad_norm": 5.391202449798584, + "learning_rate": 1.5933569371749833e-05, + "loss": 5.98278076171875, + "step": 20425 + }, + { + "epoch": 0.6422234435110309, + "grad_norm": 4.36082124710083, + "learning_rate": 1.5924209203526503e-05, + "loss": 5.922691040039062, + "step": 20450 + }, + { + "epoch": 0.6430085577451519, + "grad_norm": 5.993360996246338, + "learning_rate": 1.591484142723359e-05, + "loss": 5.959520263671875, + "step": 20475 + }, + { + "epoch": 0.643793671979273, + "grad_norm": 5.225527286529541, + "learning_rate": 1.590546605734463e-05, + "loss": 5.952388305664062, + "step": 20500 + }, + { + "epoch": 0.644578786213394, + "grad_norm": 5.704545497894287, + "learning_rate": 1.5896083108344874e-05, + "loss": 5.937107543945313, + "step": 20525 + }, + { + "epoch": 0.6453639004475151, + "grad_norm": 5.099072456359863, + "learning_rate": 1.5886692594731302e-05, + "loss": 5.9643182373046875, + "step": 20550 + }, + { + "epoch": 0.6461490146816362, + "grad_norm": 4.530437469482422, + "learning_rate": 1.5877294531012576e-05, + "loss": 5.9673193359375, + "step": 20575 + }, + { + "epoch": 0.6469341289157572, + "grad_norm": 5.158133029937744, + "learning_rate": 1.586788893170902e-05, + "loss": 5.966237182617188, + "step": 20600 + }, + { + "epoch": 0.6477192431498783, + "grad_norm": 4.317300796508789, + "learning_rate": 1.58584758113526e-05, + "loss": 5.94846923828125, + "step": 20625 + }, + { + "epoch": 0.6485043573839994, + "grad_norm": 4.5393829345703125, + "learning_rate": 1.584905518448691e-05, + "loss": 5.9521533203125, + "step": 20650 + }, + { + "epoch": 0.6492894716181205, + "grad_norm": 4.4543986320495605, + "learning_rate": 1.5839627065667134e-05, + "loss": 5.949422607421875, + "step": 20675 + }, + { + "epoch": 0.6500745858522415, + "grad_norm": 5.716519832611084, + "learning_rate": 1.5830191469460037e-05, + "loss": 5.947935791015625, + "step": 20700 + }, + { + "epoch": 0.6508597000863625, + "grad_norm": 11.801023483276367, + "learning_rate": 1.582074841044393e-05, + "loss": 5.95759521484375, + "step": 20725 + }, + { + "epoch": 0.6516448143204836, + "grad_norm": 5.441012382507324, + "learning_rate": 1.581129790320866e-05, + "loss": 5.951786499023438, + "step": 20750 + }, + { + "epoch": 0.6524299285546047, + "grad_norm": 4.929854869842529, + "learning_rate": 1.5801839962355577e-05, + "loss": 5.940153198242188, + "step": 20775 + }, + { + "epoch": 0.6532150427887258, + "grad_norm": 6.114773273468018, + "learning_rate": 1.5792374602497524e-05, + "loss": 5.953974609375, + "step": 20800 + }, + { + "epoch": 0.6540001570228469, + "grad_norm": 5.3034257888793945, + "learning_rate": 1.5782901838258804e-05, + "loss": 5.954489135742188, + "step": 20825 + }, + { + "epoch": 0.6547852712569678, + "grad_norm": 4.894354343414307, + "learning_rate": 1.577342168427515e-05, + "loss": 5.952176513671875, + "step": 20850 + }, + { + "epoch": 0.6555703854910889, + "grad_norm": 4.353710651397705, + "learning_rate": 1.5763934155193722e-05, + "loss": 5.959436645507813, + "step": 20875 + }, + { + "epoch": 0.65635549972521, + "grad_norm": 4.638297080993652, + "learning_rate": 1.5754439265673073e-05, + "loss": 5.941904296875, + "step": 20900 + }, + { + "epoch": 0.6571406139593311, + "grad_norm": 13.683270454406738, + "learning_rate": 1.5744937030383133e-05, + "loss": 5.949602661132812, + "step": 20925 + }, + { + "epoch": 0.6579257281934522, + "grad_norm": 5.719529628753662, + "learning_rate": 1.5735427464005176e-05, + "loss": 5.933261108398438, + "step": 20950 + }, + { + "epoch": 0.6587108424275732, + "grad_norm": 5.1658501625061035, + "learning_rate": 1.5725910581231798e-05, + "loss": 5.973504638671875, + "step": 20975 + }, + { + "epoch": 0.6594959566616942, + "grad_norm": 6.077699661254883, + "learning_rate": 1.5716386396766908e-05, + "loss": 5.952903442382812, + "step": 21000 + }, + { + "epoch": 0.6602810708958153, + "grad_norm": 7.679667949676514, + "learning_rate": 1.570685492532569e-05, + "loss": 5.940457153320312, + "step": 21025 + }, + { + "epoch": 0.6610661851299364, + "grad_norm": 4.2500386238098145, + "learning_rate": 1.569731618163459e-05, + "loss": 5.936610107421875, + "step": 21050 + }, + { + "epoch": 0.6618512993640575, + "grad_norm": 4.298551082611084, + "learning_rate": 1.5687770180431293e-05, + "loss": 5.93721923828125, + "step": 21075 + }, + { + "epoch": 0.6626364135981785, + "grad_norm": 6.328446388244629, + "learning_rate": 1.567821693646469e-05, + "loss": 5.96013916015625, + "step": 21100 + }, + { + "epoch": 0.6634215278322996, + "grad_norm": 6.253689765930176, + "learning_rate": 1.5668656464494868e-05, + "loss": 5.960023193359375, + "step": 21125 + }, + { + "epoch": 0.6642066420664207, + "grad_norm": 7.285246849060059, + "learning_rate": 1.5659088779293077e-05, + "loss": 5.973350830078125, + "step": 21150 + }, + { + "epoch": 0.6649917563005417, + "grad_norm": 6.608323097229004, + "learning_rate": 1.5649513895641714e-05, + "loss": 5.973410034179688, + "step": 21175 + }, + { + "epoch": 0.6657768705346628, + "grad_norm": 6.5380659103393555, + "learning_rate": 1.56399318283343e-05, + "loss": 5.942742919921875, + "step": 21200 + }, + { + "epoch": 0.6665619847687838, + "grad_norm": 5.6246018409729, + "learning_rate": 1.5630342592175448e-05, + "loss": 5.937447509765625, + "step": 21225 + }, + { + "epoch": 0.6673470990029049, + "grad_norm": 5.7414631843566895, + "learning_rate": 1.5620746201980858e-05, + "loss": 5.920745239257813, + "step": 21250 + }, + { + "epoch": 0.668132213237026, + "grad_norm": 5.64186429977417, + "learning_rate": 1.5611142672577274e-05, + "loss": 5.957138671875, + "step": 21275 + }, + { + "epoch": 0.6689173274711471, + "grad_norm": 5.522372722625732, + "learning_rate": 1.5601532018802472e-05, + "loss": 5.94736328125, + "step": 21300 + }, + { + "epoch": 0.6697024417052682, + "grad_norm": 6.00099515914917, + "learning_rate": 1.5591914255505236e-05, + "loss": 5.935474853515625, + "step": 21325 + }, + { + "epoch": 0.6704875559393891, + "grad_norm": 3.8220860958099365, + "learning_rate": 1.558228939754534e-05, + "loss": 5.94768310546875, + "step": 21350 + }, + { + "epoch": 0.6712726701735102, + "grad_norm": 6.856430530548096, + "learning_rate": 1.5572657459793512e-05, + "loss": 5.949468383789062, + "step": 21375 + }, + { + "epoch": 0.6720577844076313, + "grad_norm": 5.108688831329346, + "learning_rate": 1.5563018457131423e-05, + "loss": 5.937244262695312, + "step": 21400 + }, + { + "epoch": 0.6728428986417524, + "grad_norm": 5.812402248382568, + "learning_rate": 1.5553372404451652e-05, + "loss": 5.937905883789062, + "step": 21425 + }, + { + "epoch": 0.6736280128758735, + "grad_norm": 5.973251819610596, + "learning_rate": 1.5543719316657687e-05, + "loss": 5.982296142578125, + "step": 21450 + }, + { + "epoch": 0.6744131271099945, + "grad_norm": 5.298767566680908, + "learning_rate": 1.553405920866386e-05, + "loss": 5.955243530273438, + "step": 21475 + }, + { + "epoch": 0.6751982413441155, + "grad_norm": 6.405760288238525, + "learning_rate": 1.5524392095395376e-05, + "loss": 5.945211791992188, + "step": 21500 + }, + { + "epoch": 0.6759833555782366, + "grad_norm": 7.058891296386719, + "learning_rate": 1.5514717991788248e-05, + "loss": 5.94760986328125, + "step": 21525 + }, + { + "epoch": 0.6767684698123577, + "grad_norm": 6.27133321762085, + "learning_rate": 1.5505036912789294e-05, + "loss": 5.968565673828125, + "step": 21550 + }, + { + "epoch": 0.6775535840464788, + "grad_norm": 4.5093278884887695, + "learning_rate": 1.5495348873356104e-05, + "loss": 5.926309814453125, + "step": 21575 + }, + { + "epoch": 0.6783386982805998, + "grad_norm": 8.051047325134277, + "learning_rate": 1.5485653888457034e-05, + "loss": 5.941885986328125, + "step": 21600 + }, + { + "epoch": 0.6791238125147209, + "grad_norm": 8.223483085632324, + "learning_rate": 1.5475951973071157e-05, + "loss": 5.92613037109375, + "step": 21625 + }, + { + "epoch": 0.679908926748842, + "grad_norm": 6.628381252288818, + "learning_rate": 1.546624314218826e-05, + "loss": 5.920323486328125, + "step": 21650 + }, + { + "epoch": 0.680694040982963, + "grad_norm": 4.751639366149902, + "learning_rate": 1.5456527410808815e-05, + "loss": 5.953662109375, + "step": 21675 + }, + { + "epoch": 0.6814791552170841, + "grad_norm": 7.328832149505615, + "learning_rate": 1.5446804793943958e-05, + "loss": 5.961588745117187, + "step": 21700 + }, + { + "epoch": 0.6822642694512051, + "grad_norm": 6.9391093254089355, + "learning_rate": 1.5437075306615457e-05, + "loss": 5.94409423828125, + "step": 21725 + }, + { + "epoch": 0.6830493836853262, + "grad_norm": 5.411035537719727, + "learning_rate": 1.5427338963855705e-05, + "loss": 5.938038940429688, + "step": 21750 + }, + { + "epoch": 0.6838344979194473, + "grad_norm": 5.373900890350342, + "learning_rate": 1.541759578070767e-05, + "loss": 5.975330810546875, + "step": 21775 + }, + { + "epoch": 0.6846196121535684, + "grad_norm": 5.89310884475708, + "learning_rate": 1.5407845772224913e-05, + "loss": 5.933433837890625, + "step": 21800 + }, + { + "epoch": 0.6854047263876895, + "grad_norm": 5.958347320556641, + "learning_rate": 1.539808895347152e-05, + "loss": 5.9441357421875, + "step": 21825 + }, + { + "epoch": 0.6861898406218104, + "grad_norm": 4.54132080078125, + "learning_rate": 1.5388325339522096e-05, + "loss": 5.9487109375, + "step": 21850 + }, + { + "epoch": 0.6869749548559315, + "grad_norm": 4.488670825958252, + "learning_rate": 1.5378554945461767e-05, + "loss": 5.939719848632812, + "step": 21875 + }, + { + "epoch": 0.6877600690900526, + "grad_norm": 6.360354423522949, + "learning_rate": 1.5368777786386118e-05, + "loss": 5.959747924804687, + "step": 21900 + }, + { + "epoch": 0.6885451833241737, + "grad_norm": 6.072577953338623, + "learning_rate": 1.535899387740119e-05, + "loss": 5.952601928710937, + "step": 21925 + }, + { + "epoch": 0.6893302975582948, + "grad_norm": 8.194076538085938, + "learning_rate": 1.5349203233623447e-05, + "loss": 5.934976806640625, + "step": 21950 + }, + { + "epoch": 0.6901154117924158, + "grad_norm": 5.398212909698486, + "learning_rate": 1.5339405870179777e-05, + "loss": 5.944864501953125, + "step": 21975 + }, + { + "epoch": 0.6909005260265368, + "grad_norm": 5.118349552154541, + "learning_rate": 1.532960180220743e-05, + "loss": 5.942924194335937, + "step": 22000 + }, + { + "epoch": 0.6916856402606579, + "grad_norm": 4.518510818481445, + "learning_rate": 1.531979104485402e-05, + "loss": 5.9525830078125, + "step": 22025 + }, + { + "epoch": 0.692470754494779, + "grad_norm": 6.297440052032471, + "learning_rate": 1.5309973613277506e-05, + "loss": 5.948225708007812, + "step": 22050 + }, + { + "epoch": 0.6932558687289001, + "grad_norm": 5.943297863006592, + "learning_rate": 1.5300149522646142e-05, + "loss": 5.943578491210937, + "step": 22075 + }, + { + "epoch": 0.6940409829630211, + "grad_norm": 4.273644924163818, + "learning_rate": 1.5290318788138486e-05, + "loss": 5.9548291015625, + "step": 22100 + }, + { + "epoch": 0.6948260971971422, + "grad_norm": 6.814795017242432, + "learning_rate": 1.5280481424943353e-05, + "loss": 5.937479858398437, + "step": 22125 + }, + { + "epoch": 0.6956112114312633, + "grad_norm": 5.836712837219238, + "learning_rate": 1.527063744825981e-05, + "loss": 5.955220336914063, + "step": 22150 + }, + { + "epoch": 0.6963963256653843, + "grad_norm": 6.5769500732421875, + "learning_rate": 1.5260786873297118e-05, + "loss": 5.944104614257813, + "step": 22175 + }, + { + "epoch": 0.6971814398995054, + "grad_norm": 5.2077860832214355, + "learning_rate": 1.5250929715274757e-05, + "loss": 5.9373236083984375, + "step": 22200 + }, + { + "epoch": 0.6979665541336264, + "grad_norm": 5.728697776794434, + "learning_rate": 1.524106598942237e-05, + "loss": 5.937839965820313, + "step": 22225 + }, + { + "epoch": 0.6987516683677475, + "grad_norm": 6.814029693603516, + "learning_rate": 1.5231195710979744e-05, + "loss": 5.928355102539062, + "step": 22250 + }, + { + "epoch": 0.6995367826018686, + "grad_norm": 5.810216426849365, + "learning_rate": 1.5221318895196794e-05, + "loss": 5.952032470703125, + "step": 22275 + }, + { + "epoch": 0.7003218968359897, + "grad_norm": 4.419250011444092, + "learning_rate": 1.5211435557333534e-05, + "loss": 5.96245361328125, + "step": 22300 + }, + { + "epoch": 0.7011070110701108, + "grad_norm": 7.338109016418457, + "learning_rate": 1.5201545712660054e-05, + "loss": 5.94522216796875, + "step": 22325 + }, + { + "epoch": 0.7018921253042317, + "grad_norm": 4.209413051605225, + "learning_rate": 1.5191649376456498e-05, + "loss": 5.941893310546875, + "step": 22350 + }, + { + "epoch": 0.7026772395383528, + "grad_norm": 5.033707618713379, + "learning_rate": 1.5181746564013044e-05, + "loss": 5.9193572998046875, + "step": 22375 + }, + { + "epoch": 0.7034623537724739, + "grad_norm": 7.640013217926025, + "learning_rate": 1.5171837290629864e-05, + "loss": 5.946724243164063, + "step": 22400 + }, + { + "epoch": 0.704247468006595, + "grad_norm": 6.552804470062256, + "learning_rate": 1.5161921571617127e-05, + "loss": 5.933082275390625, + "step": 22425 + }, + { + "epoch": 0.7050325822407161, + "grad_norm": 4.747134208679199, + "learning_rate": 1.515199942229495e-05, + "loss": 5.947757568359375, + "step": 22450 + }, + { + "epoch": 0.705817696474837, + "grad_norm": 3.492997169494629, + "learning_rate": 1.5142070857993393e-05, + "loss": 5.923534545898438, + "step": 22475 + }, + { + "epoch": 0.7066028107089581, + "grad_norm": 4.520173072814941, + "learning_rate": 1.5132135894052415e-05, + "loss": 5.929746704101563, + "step": 22500 + }, + { + "epoch": 0.7073879249430792, + "grad_norm": 4.3402299880981445, + "learning_rate": 1.512219454582188e-05, + "loss": 5.928965454101562, + "step": 22525 + }, + { + "epoch": 0.7081730391772003, + "grad_norm": 4.704042434692383, + "learning_rate": 1.5112246828661504e-05, + "loss": 5.92755859375, + "step": 22550 + }, + { + "epoch": 0.7089581534113214, + "grad_norm": 4.015030860900879, + "learning_rate": 1.5102292757940844e-05, + "loss": 5.93071044921875, + "step": 22575 + }, + { + "epoch": 0.7097432676454424, + "grad_norm": 4.4564433097839355, + "learning_rate": 1.5092332349039278e-05, + "loss": 5.95046875, + "step": 22600 + }, + { + "epoch": 0.7105283818795635, + "grad_norm": 5.929630756378174, + "learning_rate": 1.5082365617345971e-05, + "loss": 5.931036376953125, + "step": 22625 + }, + { + "epoch": 0.7113134961136846, + "grad_norm": 6.370832920074463, + "learning_rate": 1.5072392578259862e-05, + "loss": 5.946773681640625, + "step": 22650 + }, + { + "epoch": 0.7120986103478056, + "grad_norm": 5.157780647277832, + "learning_rate": 1.5062413247189637e-05, + "loss": 5.931987915039063, + "step": 22675 + }, + { + "epoch": 0.7128837245819267, + "grad_norm": 4.720494747161865, + "learning_rate": 1.5052427639553698e-05, + "loss": 5.944151000976563, + "step": 22700 + }, + { + "epoch": 0.7136688388160477, + "grad_norm": 6.657490253448486, + "learning_rate": 1.5042435770780139e-05, + "loss": 5.93611328125, + "step": 22725 + }, + { + "epoch": 0.7144539530501688, + "grad_norm": 6.249160289764404, + "learning_rate": 1.5032437656306745e-05, + "loss": 5.9336614990234375, + "step": 22750 + }, + { + "epoch": 0.7152390672842899, + "grad_norm": 6.286226272583008, + "learning_rate": 1.5022433311580931e-05, + "loss": 5.944991455078125, + "step": 22775 + }, + { + "epoch": 0.716024181518411, + "grad_norm": 5.67601203918457, + "learning_rate": 1.5012422752059756e-05, + "loss": 5.93767822265625, + "step": 22800 + }, + { + "epoch": 0.7168092957525319, + "grad_norm": 4.723778247833252, + "learning_rate": 1.5002405993209872e-05, + "loss": 5.92890625, + "step": 22825 + }, + { + "epoch": 0.717594409986653, + "grad_norm": 5.168560028076172, + "learning_rate": 1.4992383050507505e-05, + "loss": 5.963192138671875, + "step": 22850 + }, + { + "epoch": 0.7183795242207741, + "grad_norm": 5.517043590545654, + "learning_rate": 1.4982353939438438e-05, + "loss": 5.925534057617187, + "step": 22875 + }, + { + "epoch": 0.7191646384548952, + "grad_norm": 7.2527546882629395, + "learning_rate": 1.4972318675497994e-05, + "loss": 5.924671630859375, + "step": 22900 + }, + { + "epoch": 0.7199497526890163, + "grad_norm": 4.404016017913818, + "learning_rate": 1.4962277274190993e-05, + "loss": 5.938770751953125, + "step": 22925 + }, + { + "epoch": 0.7207348669231373, + "grad_norm": 5.898436069488525, + "learning_rate": 1.4952229751031739e-05, + "loss": 5.925785522460938, + "step": 22950 + }, + { + "epoch": 0.7215199811572583, + "grad_norm": 4.846441268920898, + "learning_rate": 1.494217612154399e-05, + "loss": 5.927504272460937, + "step": 22975 + }, + { + "epoch": 0.7223050953913794, + "grad_norm": 5.548570156097412, + "learning_rate": 1.4932116401260948e-05, + "loss": 5.9596116202218195, + "step": 23000 + }, + { + "epoch": 0.7230902096255005, + "grad_norm": 5.576210975646973, + "learning_rate": 1.492205060572522e-05, + "loss": 5.94980224609375, + "step": 23025 + }, + { + "epoch": 0.7238753238596216, + "grad_norm": 4.417294979095459, + "learning_rate": 1.49119787504888e-05, + "loss": 5.934359130859375, + "step": 23050 + }, + { + "epoch": 0.7246604380937426, + "grad_norm": 5.5810370445251465, + "learning_rate": 1.4901900851113043e-05, + "loss": 5.952708129882812, + "step": 23075 + }, + { + "epoch": 0.7254455523278637, + "grad_norm": 6.758584499359131, + "learning_rate": 1.4891816923168645e-05, + "loss": 5.958976440429687, + "step": 23100 + }, + { + "epoch": 0.7262306665619848, + "grad_norm": 6.144657135009766, + "learning_rate": 1.4881726982235617e-05, + "loss": 5.936312255859375, + "step": 23125 + }, + { + "epoch": 0.7270157807961058, + "grad_norm": 5.675758361816406, + "learning_rate": 1.4871631043903256e-05, + "loss": 5.929334716796875, + "step": 23150 + }, + { + "epoch": 0.7278008950302269, + "grad_norm": 5.159308910369873, + "learning_rate": 1.4861529123770129e-05, + "loss": 5.9201708984375, + "step": 23175 + }, + { + "epoch": 0.7285860092643479, + "grad_norm": 6.82968807220459, + "learning_rate": 1.4851421237444046e-05, + "loss": 5.936470336914063, + "step": 23200 + }, + { + "epoch": 0.729371123498469, + "grad_norm": 5.0618720054626465, + "learning_rate": 1.4841307400542027e-05, + "loss": 5.939244384765625, + "step": 23225 + }, + { + "epoch": 0.7301562377325901, + "grad_norm": 8.101469993591309, + "learning_rate": 1.4831187628690302e-05, + "loss": 5.930047607421875, + "step": 23250 + }, + { + "epoch": 0.7309413519667112, + "grad_norm": 5.44182014465332, + "learning_rate": 1.4821061937524246e-05, + "loss": 5.935725708007812, + "step": 23275 + }, + { + "epoch": 0.7317264662008323, + "grad_norm": 7.970261573791504, + "learning_rate": 1.481093034268841e-05, + "loss": 5.948595581054687, + "step": 23300 + }, + { + "epoch": 0.7325115804349532, + "grad_norm": 3.9386982917785645, + "learning_rate": 1.4800792859836435e-05, + "loss": 5.936383666992188, + "step": 23325 + }, + { + "epoch": 0.7332966946690743, + "grad_norm": 6.579864025115967, + "learning_rate": 1.4790649504631086e-05, + "loss": 5.936826782226563, + "step": 23350 + }, + { + "epoch": 0.7340818089031954, + "grad_norm": 5.588583469390869, + "learning_rate": 1.4780500292744184e-05, + "loss": 5.956162109375, + "step": 23375 + }, + { + "epoch": 0.7348669231373165, + "grad_norm": 4.133800983428955, + "learning_rate": 1.4770345239856605e-05, + "loss": 5.931978759765625, + "step": 23400 + }, + { + "epoch": 0.7356520373714376, + "grad_norm": 6.133061408996582, + "learning_rate": 1.476018436165825e-05, + "loss": 5.925692138671875, + "step": 23425 + }, + { + "epoch": 0.7364371516055586, + "grad_norm": 6.040001392364502, + "learning_rate": 1.4750017673848017e-05, + "loss": 5.94116943359375, + "step": 23450 + }, + { + "epoch": 0.7372222658396796, + "grad_norm": 5.98701810836792, + "learning_rate": 1.4739845192133788e-05, + "loss": 5.931024780273438, + "step": 23475 + }, + { + "epoch": 0.7380073800738007, + "grad_norm": 7.4124650955200195, + "learning_rate": 1.4729666932232382e-05, + "loss": 5.915379028320313, + "step": 23500 + }, + { + "epoch": 0.7387924943079218, + "grad_norm": 5.316534996032715, + "learning_rate": 1.4719482909869557e-05, + "loss": 5.9272265625, + "step": 23525 + }, + { + "epoch": 0.7395776085420429, + "grad_norm": 5.512546062469482, + "learning_rate": 1.4709293140779982e-05, + "loss": 5.945562744140625, + "step": 23550 + }, + { + "epoch": 0.7403627227761639, + "grad_norm": 5.82699728012085, + "learning_rate": 1.4699097640707178e-05, + "loss": 5.940950927734375, + "step": 23575 + }, + { + "epoch": 0.741147837010285, + "grad_norm": 4.838285446166992, + "learning_rate": 1.468889642540355e-05, + "loss": 5.926610717773437, + "step": 23600 + }, + { + "epoch": 0.7419329512444061, + "grad_norm": 3.4617230892181396, + "learning_rate": 1.4678689510630311e-05, + "loss": 5.93740966796875, + "step": 23625 + }, + { + "epoch": 0.7427180654785271, + "grad_norm": 4.173850059509277, + "learning_rate": 1.4668476912157497e-05, + "loss": 5.9239892578125, + "step": 23650 + }, + { + "epoch": 0.7435031797126482, + "grad_norm": 4.762514591217041, + "learning_rate": 1.4658258645763912e-05, + "loss": 5.9162359619140625, + "step": 23675 + }, + { + "epoch": 0.7442882939467692, + "grad_norm": 5.5171966552734375, + "learning_rate": 1.4648034727237126e-05, + "loss": 5.944258422851562, + "step": 23700 + }, + { + "epoch": 0.7450734081808903, + "grad_norm": 5.377482891082764, + "learning_rate": 1.4637805172373439e-05, + "loss": 5.946007080078125, + "step": 23725 + }, + { + "epoch": 0.7458585224150114, + "grad_norm": 4.8370585441589355, + "learning_rate": 1.4627569996977856e-05, + "loss": 5.911834106445313, + "step": 23750 + }, + { + "epoch": 0.7466436366491325, + "grad_norm": 6.2093024253845215, + "learning_rate": 1.4617329216864068e-05, + "loss": 5.935740966796875, + "step": 23775 + }, + { + "epoch": 0.7474287508832536, + "grad_norm": 4.375349998474121, + "learning_rate": 1.460708284785443e-05, + "loss": 5.931112060546875, + "step": 23800 + }, + { + "epoch": 0.7482138651173745, + "grad_norm": 4.931936264038086, + "learning_rate": 1.4596830905779926e-05, + "loss": 5.922528076171875, + "step": 23825 + }, + { + "epoch": 0.7489989793514956, + "grad_norm": 4.663707733154297, + "learning_rate": 1.4586573406480156e-05, + "loss": 5.921451416015625, + "step": 23850 + }, + { + "epoch": 0.7497840935856167, + "grad_norm": 4.747404098510742, + "learning_rate": 1.45763103658033e-05, + "loss": 5.923754272460937, + "step": 23875 + }, + { + "epoch": 0.7500353301405355, + "eval_loss": 0.16566041111946106, + "eval_runtime": 5126.0976, + "eval_samples_per_second": 4.229, + "eval_steps_per_second": 0.264, + "eval_wer": 5.085687729670425, + "step": 23883 + }, + { + "epoch": 0.7505692078197378, + "grad_norm": 6.8687567710876465, + "learning_rate": 1.4566041799606103e-05, + "loss": 5.94312255859375, + "step": 23900 + }, + { + "epoch": 0.7513543220538589, + "grad_norm": 5.074215888977051, + "learning_rate": 1.4555767723753844e-05, + "loss": 5.9457159423828125, + "step": 23925 + }, + { + "epoch": 0.7521394362879799, + "grad_norm": 5.457148551940918, + "learning_rate": 1.4545488154120323e-05, + "loss": 5.9360986328125, + "step": 23950 + }, + { + "epoch": 0.7529245505221009, + "grad_norm": 5.027622699737549, + "learning_rate": 1.4535203106587818e-05, + "loss": 5.92920166015625, + "step": 23975 + }, + { + "epoch": 0.753709664756222, + "grad_norm": 4.596832752227783, + "learning_rate": 1.4524912597047078e-05, + "loss": 5.939633178710937, + "step": 24000 + }, + { + "epoch": 0.7544947789903431, + "grad_norm": 5.68147087097168, + "learning_rate": 1.4514616641397289e-05, + "loss": 5.927726440429687, + "step": 24025 + }, + { + "epoch": 0.7552798932244642, + "grad_norm": 5.457242012023926, + "learning_rate": 1.4504315255546044e-05, + "loss": 5.919815673828125, + "step": 24050 + }, + { + "epoch": 0.7560650074585852, + "grad_norm": 5.059640884399414, + "learning_rate": 1.4494008455409335e-05, + "loss": 5.942095947265625, + "step": 24075 + }, + { + "epoch": 0.7568501216927063, + "grad_norm": 5.992023468017578, + "learning_rate": 1.4483696256911519e-05, + "loss": 5.911986083984375, + "step": 24100 + }, + { + "epoch": 0.7576352359268274, + "grad_norm": 5.197138786315918, + "learning_rate": 1.4473378675985288e-05, + "loss": 5.919488525390625, + "step": 24125 + }, + { + "epoch": 0.7584203501609484, + "grad_norm": 7.343152046203613, + "learning_rate": 1.4463055728571658e-05, + "loss": 5.947498779296875, + "step": 24150 + }, + { + "epoch": 0.7592054643950695, + "grad_norm": 6.716486930847168, + "learning_rate": 1.4452727430619924e-05, + "loss": 5.925864868164062, + "step": 24175 + }, + { + "epoch": 0.7599905786291905, + "grad_norm": 4.447653770446777, + "learning_rate": 1.4442393798087662e-05, + "loss": 5.913412475585938, + "step": 24200 + }, + { + "epoch": 0.7607756928633116, + "grad_norm": 6.645840644836426, + "learning_rate": 1.443205484694068e-05, + "loss": 5.923351440429688, + "step": 24225 + }, + { + "epoch": 0.7615608070974327, + "grad_norm": 6.190922737121582, + "learning_rate": 1.4421710593153007e-05, + "loss": 5.921123657226563, + "step": 24250 + }, + { + "epoch": 0.7623459213315538, + "grad_norm": 6.177663803100586, + "learning_rate": 1.4411361052706863e-05, + "loss": 5.920099487304688, + "step": 24275 + }, + { + "epoch": 0.7631310355656749, + "grad_norm": 3.661764144897461, + "learning_rate": 1.4401006241592643e-05, + "loss": 5.9054803466796875, + "step": 24300 + }, + { + "epoch": 0.7639161497997958, + "grad_norm": 4.489622592926025, + "learning_rate": 1.4390646175808873e-05, + "loss": 5.920570068359375, + "step": 24325 + }, + { + "epoch": 0.7647012640339169, + "grad_norm": 5.146673202514648, + "learning_rate": 1.4380280871362214e-05, + "loss": 5.930183715820313, + "step": 24350 + }, + { + "epoch": 0.765486378268038, + "grad_norm": 6.103448390960693, + "learning_rate": 1.4369910344267402e-05, + "loss": 5.915392456054687, + "step": 24375 + }, + { + "epoch": 0.7662714925021591, + "grad_norm": 5.119777679443359, + "learning_rate": 1.4359534610547257e-05, + "loss": 5.917440185546875, + "step": 24400 + }, + { + "epoch": 0.7670566067362802, + "grad_norm": 4.482285976409912, + "learning_rate": 1.4349153686232636e-05, + "loss": 5.90766357421875, + "step": 24425 + }, + { + "epoch": 0.7678417209704012, + "grad_norm": 7.46873664855957, + "learning_rate": 1.433876758736242e-05, + "loss": 5.923113403320312, + "step": 24450 + }, + { + "epoch": 0.7686268352045222, + "grad_norm": 5.963334083557129, + "learning_rate": 1.4328376329983478e-05, + "loss": 5.92294189453125, + "step": 24475 + }, + { + "epoch": 0.7694119494386433, + "grad_norm": 6.5615668296813965, + "learning_rate": 1.4317979930150659e-05, + "loss": 5.92873046875, + "step": 24500 + }, + { + "epoch": 0.7701970636727644, + "grad_norm": 4.375472545623779, + "learning_rate": 1.430757840392675e-05, + "loss": 5.919656982421875, + "step": 24525 + }, + { + "epoch": 0.7709821779068855, + "grad_norm": 6.5642595291137695, + "learning_rate": 1.4297171767382458e-05, + "loss": 5.91892578125, + "step": 24550 + }, + { + "epoch": 0.7717672921410065, + "grad_norm": 6.7428693771362305, + "learning_rate": 1.4286760036596395e-05, + "loss": 5.931058349609375, + "step": 24575 + }, + { + "epoch": 0.7725524063751276, + "grad_norm": 3.79248046875, + "learning_rate": 1.427634322765503e-05, + "loss": 5.9193182373046875, + "step": 24600 + }, + { + "epoch": 0.7733375206092487, + "grad_norm": 6.538089752197266, + "learning_rate": 1.4265921356652686e-05, + "loss": 5.919238891601562, + "step": 24625 + }, + { + "epoch": 0.7741226348433697, + "grad_norm": 5.863292217254639, + "learning_rate": 1.425549443969151e-05, + "loss": 5.912562866210937, + "step": 24650 + }, + { + "epoch": 0.7749077490774908, + "grad_norm": 5.358902931213379, + "learning_rate": 1.4245062492881434e-05, + "loss": 5.914034423828125, + "step": 24675 + }, + { + "epoch": 0.7756928633116118, + "grad_norm": 5.0374321937561035, + "learning_rate": 1.4234625532340174e-05, + "loss": 5.900840454101562, + "step": 24700 + }, + { + "epoch": 0.7764779775457329, + "grad_norm": 5.0919952392578125, + "learning_rate": 1.4224183574193185e-05, + "loss": 5.919708862304687, + "step": 24725 + }, + { + "epoch": 0.777263091779854, + "grad_norm": 4.056941509246826, + "learning_rate": 1.4213736634573643e-05, + "loss": 5.917952270507812, + "step": 24750 + }, + { + "epoch": 0.7780482060139751, + "grad_norm": 5.336673736572266, + "learning_rate": 1.4203284729622424e-05, + "loss": 5.934959716796875, + "step": 24775 + }, + { + "epoch": 0.7788333202480962, + "grad_norm": 5.216325759887695, + "learning_rate": 1.4192827875488078e-05, + "loss": 5.922044677734375, + "step": 24800 + }, + { + "epoch": 0.7796184344822171, + "grad_norm": 6.953781604766846, + "learning_rate": 1.4182366088326794e-05, + "loss": 5.939725952148438, + "step": 24825 + }, + { + "epoch": 0.7804035487163382, + "grad_norm": 6.418949604034424, + "learning_rate": 1.4171899384302385e-05, + "loss": 5.932090454101562, + "step": 24850 + }, + { + "epoch": 0.7811886629504593, + "grad_norm": 5.152813911437988, + "learning_rate": 1.4161427779586263e-05, + "loss": 5.92225830078125, + "step": 24875 + }, + { + "epoch": 0.7819737771845804, + "grad_norm": 4.950272083282471, + "learning_rate": 1.4150951290357419e-05, + "loss": 5.90053955078125, + "step": 24900 + }, + { + "epoch": 0.7827588914187015, + "grad_norm": 5.136010646820068, + "learning_rate": 1.4140469932802372e-05, + "loss": 5.919031982421875, + "step": 24925 + }, + { + "epoch": 0.7835440056528225, + "grad_norm": 6.863821983337402, + "learning_rate": 1.412998372311518e-05, + "loss": 5.92610107421875, + "step": 24950 + }, + { + "epoch": 0.7843291198869435, + "grad_norm": 4.793138027191162, + "learning_rate": 1.4119492677497386e-05, + "loss": 5.937568359375, + "step": 24975 + }, + { + "epoch": 0.7851142341210646, + "grad_norm": 5.2514495849609375, + "learning_rate": 1.4108996812158018e-05, + "loss": 5.921182861328125, + "step": 25000 + }, + { + "epoch": 0.7858993483551857, + "grad_norm": 6.13884162902832, + "learning_rate": 1.4098496143313534e-05, + "loss": 5.92160400390625, + "step": 25025 + }, + { + "epoch": 0.7866844625893068, + "grad_norm": 5.8268656730651855, + "learning_rate": 1.4087990687187827e-05, + "loss": 5.919247436523437, + "step": 25050 + }, + { + "epoch": 0.7874695768234278, + "grad_norm": 5.355508804321289, + "learning_rate": 1.4077480460012178e-05, + "loss": 5.92015380859375, + "step": 25075 + }, + { + "epoch": 0.7882546910575489, + "grad_norm": 5.661122798919678, + "learning_rate": 1.4066965478025251e-05, + "loss": 5.913434448242188, + "step": 25100 + }, + { + "epoch": 0.78903980529167, + "grad_norm": 6.459861755371094, + "learning_rate": 1.4056445757473041e-05, + "loss": 5.92521728515625, + "step": 25125 + }, + { + "epoch": 0.789824919525791, + "grad_norm": 4.6103973388671875, + "learning_rate": 1.4045921314608873e-05, + "loss": 5.901995849609375, + "step": 25150 + }, + { + "epoch": 0.790610033759912, + "grad_norm": 5.0542521476745605, + "learning_rate": 1.4035392165693364e-05, + "loss": 5.913954467773437, + "step": 25175 + }, + { + "epoch": 0.7913951479940331, + "grad_norm": 4.584170818328857, + "learning_rate": 1.4024858326994413e-05, + "loss": 5.926514892578125, + "step": 25200 + }, + { + "epoch": 0.7921802622281542, + "grad_norm": 5.345964431762695, + "learning_rate": 1.4014319814787151e-05, + "loss": 5.93136962890625, + "step": 25225 + }, + { + "epoch": 0.7929653764622753, + "grad_norm": 5.112735271453857, + "learning_rate": 1.4003776645353937e-05, + "loss": 5.928161010742188, + "step": 25250 + }, + { + "epoch": 0.7937504906963964, + "grad_norm": 5.155666828155518, + "learning_rate": 1.399322883498432e-05, + "loss": 5.911998901367188, + "step": 25275 + }, + { + "epoch": 0.7945356049305173, + "grad_norm": 7.831445217132568, + "learning_rate": 1.3982676399975031e-05, + "loss": 5.929152221679687, + "step": 25300 + }, + { + "epoch": 0.7953207191646384, + "grad_norm": 5.197892189025879, + "learning_rate": 1.3972119356629929e-05, + "loss": 5.94556884765625, + "step": 25325 + }, + { + "epoch": 0.7961058333987595, + "grad_norm": 5.850231647491455, + "learning_rate": 1.3961557721260013e-05, + "loss": 5.928639526367188, + "step": 25350 + }, + { + "epoch": 0.7968909476328806, + "grad_norm": 5.272648811340332, + "learning_rate": 1.395099151018336e-05, + "loss": 5.935162963867188, + "step": 25375 + }, + { + "epoch": 0.7976760618670017, + "grad_norm": 6.120459079742432, + "learning_rate": 1.3940420739725122e-05, + "loss": 5.953248291015625, + "step": 25400 + }, + { + "epoch": 0.7984611761011227, + "grad_norm": 5.081854820251465, + "learning_rate": 1.39298454262175e-05, + "loss": 5.96111572265625, + "step": 25425 + }, + { + "epoch": 0.7992462903352437, + "grad_norm": 5.883688926696777, + "learning_rate": 1.3919265585999712e-05, + "loss": 5.929671020507812, + "step": 25450 + }, + { + "epoch": 0.8000314045693648, + "grad_norm": 6.279791355133057, + "learning_rate": 1.3908681235417965e-05, + "loss": 5.905565795898437, + "step": 25475 + }, + { + "epoch": 0.8008165188034859, + "grad_norm": 3.954092025756836, + "learning_rate": 1.389809239082544e-05, + "loss": 5.913531494140625, + "step": 25500 + }, + { + "epoch": 0.801601633037607, + "grad_norm": 5.0571770668029785, + "learning_rate": 1.3887499068582258e-05, + "loss": 5.93718994140625, + "step": 25525 + }, + { + "epoch": 0.802386747271728, + "grad_norm": 7.146028995513916, + "learning_rate": 1.3876901285055462e-05, + "loss": 5.94028564453125, + "step": 25550 + }, + { + "epoch": 0.8031718615058491, + "grad_norm": 5.1267218589782715, + "learning_rate": 1.3866299056618984e-05, + "loss": 5.932505493164062, + "step": 25575 + }, + { + "epoch": 0.8039569757399702, + "grad_norm": 5.322806358337402, + "learning_rate": 1.3855692399653624e-05, + "loss": 5.9270086669921875, + "step": 25600 + }, + { + "epoch": 0.8047420899740912, + "grad_norm": 4.6873321533203125, + "learning_rate": 1.384508133054703e-05, + "loss": 5.919036865234375, + "step": 25625 + }, + { + "epoch": 0.8055272042082123, + "grad_norm": 5.431657791137695, + "learning_rate": 1.3834465865693656e-05, + "loss": 5.930333862304687, + "step": 25650 + }, + { + "epoch": 0.8063123184423333, + "grad_norm": 3.826162815093994, + "learning_rate": 1.3823846021494761e-05, + "loss": 5.901649780273438, + "step": 25675 + }, + { + "epoch": 0.8070974326764544, + "grad_norm": 6.513355255126953, + "learning_rate": 1.3813221814358352e-05, + "loss": 5.918021240234375, + "step": 25700 + }, + { + "epoch": 0.8078825469105755, + "grad_norm": 4.628607749938965, + "learning_rate": 1.3802593260699202e-05, + "loss": 5.94511474609375, + "step": 25725 + }, + { + "epoch": 0.8086676611446966, + "grad_norm": 5.282809257507324, + "learning_rate": 1.3791960376938776e-05, + "loss": 5.927496337890625, + "step": 25750 + }, + { + "epoch": 0.8094527753788177, + "grad_norm": 6.59267520904541, + "learning_rate": 1.378132317950524e-05, + "loss": 5.910077514648438, + "step": 25775 + }, + { + "epoch": 0.8102378896129386, + "grad_norm": 5.2645978927612305, + "learning_rate": 1.3770681684833427e-05, + "loss": 5.8995123291015625, + "step": 25800 + }, + { + "epoch": 0.8110230038470597, + "grad_norm": 6.111353874206543, + "learning_rate": 1.3760035909364804e-05, + "loss": 5.926395874023438, + "step": 25825 + }, + { + "epoch": 0.8118081180811808, + "grad_norm": 7.12476110458374, + "learning_rate": 1.3749385869547452e-05, + "loss": 5.9258740234375, + "step": 25850 + }, + { + "epoch": 0.8125932323153019, + "grad_norm": 6.301643371582031, + "learning_rate": 1.3738731581836043e-05, + "loss": 5.917979125976562, + "step": 25875 + }, + { + "epoch": 0.813378346549423, + "grad_norm": 6.049911975860596, + "learning_rate": 1.3728073062691811e-05, + "loss": 5.9285595703125, + "step": 25900 + }, + { + "epoch": 0.814163460783544, + "grad_norm": 6.385380268096924, + "learning_rate": 1.371741032858253e-05, + "loss": 5.941363525390625, + "step": 25925 + }, + { + "epoch": 0.814948575017665, + "grad_norm": 5.579838275909424, + "learning_rate": 1.3706743395982485e-05, + "loss": 5.91442626953125, + "step": 25950 + }, + { + "epoch": 0.8157336892517861, + "grad_norm": 5.533994197845459, + "learning_rate": 1.3696072281372445e-05, + "loss": 5.904838256835937, + "step": 25975 + }, + { + "epoch": 0.8165188034859072, + "grad_norm": 5.021455764770508, + "learning_rate": 1.3685397001239644e-05, + "loss": 5.9096630859375, + "step": 26000 + }, + { + "epoch": 0.8173039177200283, + "grad_norm": 5.315699577331543, + "learning_rate": 1.3674717572077751e-05, + "loss": 5.920933227539063, + "step": 26025 + }, + { + "epoch": 0.8180890319541493, + "grad_norm": 4.016049385070801, + "learning_rate": 1.366403401038684e-05, + "loss": 5.9052099609375, + "step": 26050 + }, + { + "epoch": 0.8188741461882704, + "grad_norm": 5.34450101852417, + "learning_rate": 1.365334633267338e-05, + "loss": 5.9308404541015625, + "step": 26075 + }, + { + "epoch": 0.8196592604223915, + "grad_norm": 7.506278991699219, + "learning_rate": 1.3642654555450195e-05, + "loss": 5.908517456054687, + "step": 26100 + }, + { + "epoch": 0.8204443746565125, + "grad_norm": 4.486503601074219, + "learning_rate": 1.3631958695236438e-05, + "loss": 5.915509643554688, + "step": 26125 + }, + { + "epoch": 0.8212294888906336, + "grad_norm": 7.123664855957031, + "learning_rate": 1.3621258768557575e-05, + "loss": 5.91797119140625, + "step": 26150 + }, + { + "epoch": 0.8220146031247546, + "grad_norm": 6.063544750213623, + "learning_rate": 1.3610554791945359e-05, + "loss": 5.895484619140625, + "step": 26175 + }, + { + "epoch": 0.8227997173588757, + "grad_norm": 3.6790363788604736, + "learning_rate": 1.3599846781937787e-05, + "loss": 5.9004583740234375, + "step": 26200 + }, + { + "epoch": 0.8235848315929968, + "grad_norm": 4.883180618286133, + "learning_rate": 1.3589134755079106e-05, + "loss": 5.922233276367187, + "step": 26225 + }, + { + "epoch": 0.8243699458271179, + "grad_norm": 4.843576908111572, + "learning_rate": 1.3578418727919748e-05, + "loss": 5.909412231445312, + "step": 26250 + }, + { + "epoch": 0.825155060061239, + "grad_norm": 4.847573280334473, + "learning_rate": 1.356769871701635e-05, + "loss": 5.918565673828125, + "step": 26275 + }, + { + "epoch": 0.8259401742953599, + "grad_norm": 4.738499164581299, + "learning_rate": 1.3556974738931682e-05, + "loss": 5.916605224609375, + "step": 26300 + }, + { + "epoch": 0.826725288529481, + "grad_norm": 3.9779415130615234, + "learning_rate": 1.3546246810234653e-05, + "loss": 5.899671020507813, + "step": 26325 + }, + { + "epoch": 0.8275104027636021, + "grad_norm": 5.2473978996276855, + "learning_rate": 1.3535514947500278e-05, + "loss": 5.904078369140625, + "step": 26350 + }, + { + "epoch": 0.8282955169977232, + "grad_norm": 5.515202522277832, + "learning_rate": 1.352477916730965e-05, + "loss": 5.92246826171875, + "step": 26375 + }, + { + "epoch": 0.8290806312318443, + "grad_norm": 4.978670597076416, + "learning_rate": 1.3514039486249905e-05, + "loss": 5.906882934570312, + "step": 26400 + }, + { + "epoch": 0.8298657454659653, + "grad_norm": 3.7597055435180664, + "learning_rate": 1.3503295920914218e-05, + "loss": 5.910196533203125, + "step": 26425 + }, + { + "epoch": 0.8306508597000863, + "grad_norm": 4.8066911697387695, + "learning_rate": 1.349254848790176e-05, + "loss": 5.9197900390625, + "step": 26450 + }, + { + "epoch": 0.8314359739342074, + "grad_norm": 6.4747538566589355, + "learning_rate": 1.348179720381768e-05, + "loss": 5.9114013671875, + "step": 26475 + }, + { + "epoch": 0.8322210881683285, + "grad_norm": 5.243331432342529, + "learning_rate": 1.3471042085273067e-05, + "loss": 5.91856201171875, + "step": 26500 + }, + { + "epoch": 0.8330062024024496, + "grad_norm": 5.765937328338623, + "learning_rate": 1.3460283148884952e-05, + "loss": 5.9151544189453125, + "step": 26525 + }, + { + "epoch": 0.8337913166365706, + "grad_norm": 6.960680961608887, + "learning_rate": 1.3449520411276252e-05, + "loss": 5.920487670898438, + "step": 26550 + }, + { + "epoch": 0.8345764308706917, + "grad_norm": 5.239001274108887, + "learning_rate": 1.3438753889075755e-05, + "loss": 5.928582763671875, + "step": 26575 + }, + { + "epoch": 0.8353615451048128, + "grad_norm": 3.7424862384796143, + "learning_rate": 1.342798359891811e-05, + "loss": 5.9039373779296875, + "step": 26600 + }, + { + "epoch": 0.8361466593389338, + "grad_norm": 8.112526893615723, + "learning_rate": 1.3417209557443774e-05, + "loss": 5.916763916015625, + "step": 26625 + }, + { + "epoch": 0.8369317735730549, + "grad_norm": 5.667616844177246, + "learning_rate": 1.3406431781299005e-05, + "loss": 5.920863647460937, + "step": 26650 + }, + { + "epoch": 0.8377168878071759, + "grad_norm": 5.168137073516846, + "learning_rate": 1.3395650287135835e-05, + "loss": 5.898240966796875, + "step": 26675 + }, + { + "epoch": 0.838502002041297, + "grad_norm": 6.185590744018555, + "learning_rate": 1.3384865091612031e-05, + "loss": 5.901964721679687, + "step": 26700 + }, + { + "epoch": 0.8392871162754181, + "grad_norm": 5.919241905212402, + "learning_rate": 1.3374076211391091e-05, + "loss": 5.917716674804687, + "step": 26725 + }, + { + "epoch": 0.8400722305095392, + "grad_norm": 7.538485050201416, + "learning_rate": 1.3363283663142192e-05, + "loss": 5.92005859375, + "step": 26750 + }, + { + "epoch": 0.8408573447436603, + "grad_norm": 4.79243803024292, + "learning_rate": 1.3352487463540196e-05, + "loss": 5.89739990234375, + "step": 26775 + }, + { + "epoch": 0.8416424589777812, + "grad_norm": 5.843585968017578, + "learning_rate": 1.3341687629265587e-05, + "loss": 5.910722045898438, + "step": 26800 + }, + { + "epoch": 0.8424275732119023, + "grad_norm": 4.522217273712158, + "learning_rate": 1.3330884177004478e-05, + "loss": 5.9146356201171875, + "step": 26825 + }, + { + "epoch": 0.8432126874460234, + "grad_norm": 5.405519962310791, + "learning_rate": 1.3320077123448562e-05, + "loss": 5.902208251953125, + "step": 26850 + }, + { + "epoch": 0.8439978016801445, + "grad_norm": 5.014639377593994, + "learning_rate": 1.3309266485295113e-05, + "loss": 5.916913452148438, + "step": 26875 + }, + { + "epoch": 0.8447829159142656, + "grad_norm": 5.063216686248779, + "learning_rate": 1.3298452279246922e-05, + "loss": 5.922537231445313, + "step": 26900 + }, + { + "epoch": 0.8455680301483866, + "grad_norm": 6.688914775848389, + "learning_rate": 1.3287634522012307e-05, + "loss": 5.9293060302734375, + "step": 26925 + }, + { + "epoch": 0.8463531443825076, + "grad_norm": 6.3757195472717285, + "learning_rate": 1.3276813230305065e-05, + "loss": 5.919682006835938, + "step": 26950 + }, + { + "epoch": 0.8471382586166287, + "grad_norm": 6.317503929138184, + "learning_rate": 1.3265988420844464e-05, + "loss": 5.919302368164063, + "step": 26975 + }, + { + "epoch": 0.8479233728507498, + "grad_norm": 5.400466442108154, + "learning_rate": 1.3255160110355197e-05, + "loss": 5.893981323242188, + "step": 27000 + }, + { + "epoch": 0.8487084870848709, + "grad_norm": 5.0842366218566895, + "learning_rate": 1.3244328315567367e-05, + "loss": 5.91895263671875, + "step": 27025 + }, + { + "epoch": 0.8494936013189919, + "grad_norm": 4.91605806350708, + "learning_rate": 1.3233493053216462e-05, + "loss": 5.920592651367188, + "step": 27050 + }, + { + "epoch": 0.850278715553113, + "grad_norm": 4.487494468688965, + "learning_rate": 1.3222654340043335e-05, + "loss": 5.908385009765625, + "step": 27075 + }, + { + "epoch": 0.851063829787234, + "grad_norm": 5.367815971374512, + "learning_rate": 1.3211812192794157e-05, + "loss": 5.8929638671875, + "step": 27100 + }, + { + "epoch": 0.8518489440213551, + "grad_norm": 5.166168689727783, + "learning_rate": 1.320096662822042e-05, + "loss": 5.917147827148438, + "step": 27125 + }, + { + "epoch": 0.8526340582554762, + "grad_norm": 5.5328874588012695, + "learning_rate": 1.3190117663078881e-05, + "loss": 5.91535400390625, + "step": 27150 + }, + { + "epoch": 0.8534191724895972, + "grad_norm": 6.045344829559326, + "learning_rate": 1.3179265314131556e-05, + "loss": 5.910491333007813, + "step": 27175 + }, + { + "epoch": 0.8542042867237183, + "grad_norm": 3.950676918029785, + "learning_rate": 1.3168409598145698e-05, + "loss": 5.921619873046875, + "step": 27200 + }, + { + "epoch": 0.8549894009578394, + "grad_norm": 5.867423057556152, + "learning_rate": 1.315755053189375e-05, + "loss": 5.923235473632812, + "step": 27225 + }, + { + "epoch": 0.8557745151919605, + "grad_norm": 6.442917346954346, + "learning_rate": 1.3146688132153333e-05, + "loss": 5.926492309570312, + "step": 27250 + }, + { + "epoch": 0.8565596294260815, + "grad_norm": 4.493001461029053, + "learning_rate": 1.3135822415707234e-05, + "loss": 5.907914428710938, + "step": 27275 + }, + { + "epoch": 0.8573447436602025, + "grad_norm": 6.029769420623779, + "learning_rate": 1.312495339934334e-05, + "loss": 5.910380249023437, + "step": 27300 + }, + { + "epoch": 0.8581298578943236, + "grad_norm": 4.512490749359131, + "learning_rate": 1.3114081099854653e-05, + "loss": 5.923084716796875, + "step": 27325 + }, + { + "epoch": 0.8589149721284447, + "grad_norm": 6.066438674926758, + "learning_rate": 1.3103205534039239e-05, + "loss": 5.9048919677734375, + "step": 27350 + }, + { + "epoch": 0.8597000863625658, + "grad_norm": 4.040054798126221, + "learning_rate": 1.3092326718700222e-05, + "loss": 5.90914306640625, + "step": 27375 + }, + { + "epoch": 0.8604852005966869, + "grad_norm": 4.419363498687744, + "learning_rate": 1.308144467064573e-05, + "loss": 5.8962451171875, + "step": 27400 + }, + { + "epoch": 0.8612703148308078, + "grad_norm": 4.494662761688232, + "learning_rate": 1.307055940668891e-05, + "loss": 5.919818115234375, + "step": 27425 + }, + { + "epoch": 0.8620554290649289, + "grad_norm": 4.534942150115967, + "learning_rate": 1.3059670943647844e-05, + "loss": 5.911969604492188, + "step": 27450 + }, + { + "epoch": 0.86284054329905, + "grad_norm": 6.330618858337402, + "learning_rate": 1.3048779298345589e-05, + "loss": 5.920072631835938, + "step": 27475 + }, + { + "epoch": 0.8636256575331711, + "grad_norm": 4.869382858276367, + "learning_rate": 1.3037884487610097e-05, + "loss": 5.91664306640625, + "step": 27500 + }, + { + "epoch": 0.8644107717672922, + "grad_norm": 5.838260173797607, + "learning_rate": 1.3026986528274223e-05, + "loss": 5.90989013671875, + "step": 27525 + }, + { + "epoch": 0.8651958860014132, + "grad_norm": 4.5731329917907715, + "learning_rate": 1.3016085437175682e-05, + "loss": 5.921937866210937, + "step": 27550 + }, + { + "epoch": 0.8659810002355343, + "grad_norm": 4.020991802215576, + "learning_rate": 1.300518123115703e-05, + "loss": 5.932952880859375, + "step": 27575 + }, + { + "epoch": 0.8667661144696553, + "grad_norm": 7.618587493896484, + "learning_rate": 1.2994273927065624e-05, + "loss": 5.933410034179688, + "step": 27600 + }, + { + "epoch": 0.8675512287037764, + "grad_norm": 8.426789283752441, + "learning_rate": 1.2983363541753627e-05, + "loss": 5.910284423828125, + "step": 27625 + }, + { + "epoch": 0.8683363429378974, + "grad_norm": 5.590900897979736, + "learning_rate": 1.2972450092077951e-05, + "loss": 5.9137725830078125, + "step": 27650 + }, + { + "epoch": 0.8691214571720185, + "grad_norm": 4.979513168334961, + "learning_rate": 1.2961533594900244e-05, + "loss": 5.927318115234375, + "step": 27675 + }, + { + "epoch": 0.8699065714061396, + "grad_norm": 7.231375694274902, + "learning_rate": 1.2950614067086857e-05, + "loss": 5.916806030273437, + "step": 27700 + }, + { + "epoch": 0.8706916856402607, + "grad_norm": 7.3363471031188965, + "learning_rate": 1.2939691525508842e-05, + "loss": 5.91991943359375, + "step": 27725 + }, + { + "epoch": 0.8714767998743818, + "grad_norm": 6.643073081970215, + "learning_rate": 1.2928765987041883e-05, + "loss": 5.896051635742188, + "step": 27750 + }, + { + "epoch": 0.8722619141085027, + "grad_norm": 6.982113361358643, + "learning_rate": 1.2917837468566312e-05, + "loss": 5.900677490234375, + "step": 27775 + }, + { + "epoch": 0.8730470283426238, + "grad_norm": 6.371608257293701, + "learning_rate": 1.2906905986967058e-05, + "loss": 5.90343994140625, + "step": 27800 + }, + { + "epoch": 0.8738321425767449, + "grad_norm": 4.57841157913208, + "learning_rate": 1.289597155913363e-05, + "loss": 5.892197875976563, + "step": 27825 + }, + { + "epoch": 0.874617256810866, + "grad_norm": 9.691904067993164, + "learning_rate": 1.2885034201960087e-05, + "loss": 5.914972534179688, + "step": 27850 + }, + { + "epoch": 0.8754023710449871, + "grad_norm": 4.313063144683838, + "learning_rate": 1.2874093932345018e-05, + "loss": 5.898465576171875, + "step": 27875 + }, + { + "epoch": 0.8761874852791081, + "grad_norm": 5.968775272369385, + "learning_rate": 1.2863150767191504e-05, + "loss": 5.898539428710937, + "step": 27900 + }, + { + "epoch": 0.8769725995132291, + "grad_norm": 4.778985500335693, + "learning_rate": 1.2852204723407111e-05, + "loss": 5.919385375976563, + "step": 27925 + }, + { + "epoch": 0.8777577137473502, + "grad_norm": 5.1884074211120605, + "learning_rate": 1.2841255817903839e-05, + "loss": 5.911961669921875, + "step": 27950 + }, + { + "epoch": 0.8785428279814713, + "grad_norm": 4.247082710266113, + "learning_rate": 1.2830304067598127e-05, + "loss": 5.908561401367187, + "step": 27975 + }, + { + "epoch": 0.8793279422155924, + "grad_norm": 4.083285331726074, + "learning_rate": 1.281934948941079e-05, + "loss": 5.879266357421875, + "step": 28000 + }, + { + "epoch": 0.8801130564497134, + "grad_norm": 5.785336494445801, + "learning_rate": 1.280839210026703e-05, + "loss": 5.912898559570312, + "step": 28025 + }, + { + "epoch": 0.8808981706838345, + "grad_norm": 8.496610641479492, + "learning_rate": 1.2797431917096372e-05, + "loss": 5.927062377929688, + "step": 28050 + }, + { + "epoch": 0.8816832849179556, + "grad_norm": 5.949460029602051, + "learning_rate": 1.2786468956832679e-05, + "loss": 5.912607421875, + "step": 28075 + }, + { + "epoch": 0.8824683991520766, + "grad_norm": 5.411488056182861, + "learning_rate": 1.2775503236414093e-05, + "loss": 5.88436767578125, + "step": 28100 + }, + { + "epoch": 0.8832535133861977, + "grad_norm": 4.449924468994141, + "learning_rate": 1.2764534772783019e-05, + "loss": 5.917919921875, + "step": 28125 + }, + { + "epoch": 0.8840386276203187, + "grad_norm": 6.629367828369141, + "learning_rate": 1.2753563582886105e-05, + "loss": 5.908243408203125, + "step": 28150 + }, + { + "epoch": 0.8848237418544398, + "grad_norm": 4.803131580352783, + "learning_rate": 1.2742589683674216e-05, + "loss": 5.9053564453125, + "step": 28175 + }, + { + "epoch": 0.8856088560885609, + "grad_norm": 4.59758996963501, + "learning_rate": 1.2731613092102391e-05, + "loss": 5.894130859375, + "step": 28200 + }, + { + "epoch": 0.886393970322682, + "grad_norm": 4.784117698669434, + "learning_rate": 1.2720633825129836e-05, + "loss": 5.888395385742188, + "step": 28225 + }, + { + "epoch": 0.887179084556803, + "grad_norm": 3.891033172607422, + "learning_rate": 1.2709651899719886e-05, + "loss": 5.912936401367188, + "step": 28250 + }, + { + "epoch": 0.887964198790924, + "grad_norm": 7.627559185028076, + "learning_rate": 1.2698667332839993e-05, + "loss": 5.896973266601562, + "step": 28275 + }, + { + "epoch": 0.8887493130250451, + "grad_norm": 4.375471591949463, + "learning_rate": 1.2687680141461682e-05, + "loss": 5.896114501953125, + "step": 28300 + }, + { + "epoch": 0.8895344272591662, + "grad_norm": 6.47393798828125, + "learning_rate": 1.2676690342560533e-05, + "loss": 5.89478759765625, + "step": 28325 + }, + { + "epoch": 0.8903195414932873, + "grad_norm": 4.697755336761475, + "learning_rate": 1.2665697953116158e-05, + "loss": 5.8997344970703125, + "step": 28350 + }, + { + "epoch": 0.8911046557274084, + "grad_norm": 7.295173645019531, + "learning_rate": 1.2654702990112169e-05, + "loss": 5.903268432617187, + "step": 28375 + }, + { + "epoch": 0.8918897699615294, + "grad_norm": 3.4301259517669678, + "learning_rate": 1.2643705470536161e-05, + "loss": 5.879957275390625, + "step": 28400 + }, + { + "epoch": 0.8926748841956504, + "grad_norm": 5.3680572509765625, + "learning_rate": 1.2632705411379664e-05, + "loss": 5.897317504882812, + "step": 28425 + }, + { + "epoch": 0.8934599984297715, + "grad_norm": 6.315160274505615, + "learning_rate": 1.2621702829638143e-05, + "loss": 5.88599609375, + "step": 28450 + }, + { + "epoch": 0.8942451126638926, + "grad_norm": 4.5944318771362305, + "learning_rate": 1.2610697742310967e-05, + "loss": 5.88390625, + "step": 28475 + }, + { + "epoch": 0.8950302268980137, + "grad_norm": 6.811062335968018, + "learning_rate": 1.2599690166401355e-05, + "loss": 5.890977783203125, + "step": 28500 + }, + { + "epoch": 0.8958153411321347, + "grad_norm": 6.246709823608398, + "learning_rate": 1.25886801189164e-05, + "loss": 5.908681640625, + "step": 28525 + }, + { + "epoch": 0.8966004553662558, + "grad_norm": 5.361160755157471, + "learning_rate": 1.2577667616866978e-05, + "loss": 5.91365966796875, + "step": 28550 + }, + { + "epoch": 0.8973855696003769, + "grad_norm": 5.704426288604736, + "learning_rate": 1.256665267726779e-05, + "loss": 5.914246826171875, + "step": 28575 + }, + { + "epoch": 0.8981706838344979, + "grad_norm": 4.4250807762146, + "learning_rate": 1.2555635317137285e-05, + "loss": 5.8962554931640625, + "step": 28600 + }, + { + "epoch": 0.898955798068619, + "grad_norm": 6.3878278732299805, + "learning_rate": 1.2544615553497658e-05, + "loss": 5.911901245117187, + "step": 28625 + }, + { + "epoch": 0.89974091230274, + "grad_norm": 4.954400062561035, + "learning_rate": 1.253359340337482e-05, + "loss": 5.880250854492187, + "step": 28650 + }, + { + "epoch": 0.9005260265368611, + "grad_norm": 6.084632396697998, + "learning_rate": 1.2522568883798358e-05, + "loss": 5.9216986083984375, + "step": 28675 + }, + { + "epoch": 0.9013111407709822, + "grad_norm": 4.896385669708252, + "learning_rate": 1.2511542011801529e-05, + "loss": 5.885166015625, + "step": 28700 + }, + { + "epoch": 0.9020962550051033, + "grad_norm": 6.50496244430542, + "learning_rate": 1.2500512804421228e-05, + "loss": 5.896882934570312, + "step": 28725 + }, + { + "epoch": 0.9028813692392244, + "grad_norm": 6.642240047454834, + "learning_rate": 1.2489481278697947e-05, + "loss": 5.905081787109375, + "step": 28750 + }, + { + "epoch": 0.9036664834733453, + "grad_norm": 4.8952250480651855, + "learning_rate": 1.247844745167577e-05, + "loss": 5.900316162109375, + "step": 28775 + }, + { + "epoch": 0.9044515977074664, + "grad_norm": 5.617021560668945, + "learning_rate": 1.2467411340402328e-05, + "loss": 5.893380126953125, + "step": 28800 + }, + { + "epoch": 0.9052367119415875, + "grad_norm": 6.7723774909973145, + "learning_rate": 1.245637296192879e-05, + "loss": 5.892852172851563, + "step": 28825 + }, + { + "epoch": 0.9060218261757086, + "grad_norm": 5.420709133148193, + "learning_rate": 1.2445332333309825e-05, + "loss": 5.904044799804687, + "step": 28850 + }, + { + "epoch": 0.9068069404098297, + "grad_norm": 5.207626819610596, + "learning_rate": 1.2434289471603571e-05, + "loss": 5.935516357421875, + "step": 28875 + }, + { + "epoch": 0.9075920546439507, + "grad_norm": 4.423824310302734, + "learning_rate": 1.2423244393871627e-05, + "loss": 5.896692504882813, + "step": 28900 + }, + { + "epoch": 0.9083771688780717, + "grad_norm": 5.383023738861084, + "learning_rate": 1.241219711717901e-05, + "loss": 5.902133178710938, + "step": 28925 + }, + { + "epoch": 0.9091622831121928, + "grad_norm": 4.236893177032471, + "learning_rate": 1.240114765859414e-05, + "loss": 5.888338623046875, + "step": 28950 + }, + { + "epoch": 0.9099473973463139, + "grad_norm": 5.010431289672852, + "learning_rate": 1.2390096035188794e-05, + "loss": 5.875116577148438, + "step": 28975 + }, + { + "epoch": 0.910732511580435, + "grad_norm": 6.412112712860107, + "learning_rate": 1.2379042264038111e-05, + "loss": 5.897857055664063, + "step": 29000 + }, + { + "epoch": 0.911517625814556, + "grad_norm": 5.494471073150635, + "learning_rate": 1.2367986362220539e-05, + "loss": 5.903126831054688, + "step": 29025 + }, + { + "epoch": 0.9123027400486771, + "grad_norm": 7.001485824584961, + "learning_rate": 1.2356928346817817e-05, + "loss": 5.905589599609375, + "step": 29050 + }, + { + "epoch": 0.9130878542827982, + "grad_norm": 4.531250476837158, + "learning_rate": 1.2345868234914961e-05, + "loss": 5.907861938476563, + "step": 29075 + }, + { + "epoch": 0.9138729685169192, + "grad_norm": 5.657148361206055, + "learning_rate": 1.2334806043600208e-05, + "loss": 5.886134033203125, + "step": 29100 + }, + { + "epoch": 0.9146580827510403, + "grad_norm": 4.89274263381958, + "learning_rate": 1.2323741789965023e-05, + "loss": 5.901471557617188, + "step": 29125 + }, + { + "epoch": 0.9154431969851613, + "grad_norm": 5.567114353179932, + "learning_rate": 1.2312675491104046e-05, + "loss": 5.902862548828125, + "step": 29150 + }, + { + "epoch": 0.9162283112192824, + "grad_norm": 4.4444074630737305, + "learning_rate": 1.2301607164115092e-05, + "loss": 5.903074951171875, + "step": 29175 + }, + { + "epoch": 0.9170134254534035, + "grad_norm": 7.408876895904541, + "learning_rate": 1.2290536826099093e-05, + "loss": 5.89898193359375, + "step": 29200 + }, + { + "epoch": 0.9177985396875246, + "grad_norm": 5.334888458251953, + "learning_rate": 1.2279464494160099e-05, + "loss": 5.901865844726562, + "step": 29225 + }, + { + "epoch": 0.9185836539216456, + "grad_norm": 5.233527660369873, + "learning_rate": 1.2268390185405228e-05, + "loss": 5.8923492431640625, + "step": 29250 + }, + { + "epoch": 0.9193687681557666, + "grad_norm": 4.669017314910889, + "learning_rate": 1.2257313916944673e-05, + "loss": 5.8918798828125, + "step": 29275 + }, + { + "epoch": 0.9201538823898877, + "grad_norm": 7.741791725158691, + "learning_rate": 1.2246235705891636e-05, + "loss": 5.905147094726562, + "step": 29300 + }, + { + "epoch": 0.9209389966240088, + "grad_norm": 5.840625762939453, + "learning_rate": 1.2235155569362331e-05, + "loss": 5.919288330078125, + "step": 29325 + }, + { + "epoch": 0.9217241108581299, + "grad_norm": 4.661303997039795, + "learning_rate": 1.2224073524475942e-05, + "loss": 5.901409301757813, + "step": 29350 + }, + { + "epoch": 0.922509225092251, + "grad_norm": 6.765843391418457, + "learning_rate": 1.22129895883546e-05, + "loss": 5.901305541992188, + "step": 29375 + }, + { + "epoch": 0.923294339326372, + "grad_norm": 5.668766021728516, + "learning_rate": 1.2201903778123365e-05, + "loss": 5.912753295898438, + "step": 29400 + }, + { + "epoch": 0.924079453560493, + "grad_norm": 4.302783012390137, + "learning_rate": 1.2190816110910186e-05, + "loss": 5.888531494140625, + "step": 29425 + }, + { + "epoch": 0.9248645677946141, + "grad_norm": 6.733384609222412, + "learning_rate": 1.2179726603845882e-05, + "loss": 5.917396850585938, + "step": 29450 + }, + { + "epoch": 0.9256496820287352, + "grad_norm": 5.473190784454346, + "learning_rate": 1.2168635274064122e-05, + "loss": 5.8955126953125, + "step": 29475 + }, + { + "epoch": 0.9264347962628563, + "grad_norm": 6.383914947509766, + "learning_rate": 1.2157542138701379e-05, + "loss": 5.889956665039063, + "step": 29500 + }, + { + "epoch": 0.9272199104969773, + "grad_norm": 4.029685020446777, + "learning_rate": 1.2146447214896926e-05, + "loss": 5.900350341796875, + "step": 29525 + }, + { + "epoch": 0.9280050247310984, + "grad_norm": 5.633557319641113, + "learning_rate": 1.213535051979279e-05, + "loss": 5.878663940429687, + "step": 29550 + }, + { + "epoch": 0.9287901389652194, + "grad_norm": 5.284363746643066, + "learning_rate": 1.2124252070533744e-05, + "loss": 5.902693481445312, + "step": 29575 + }, + { + "epoch": 0.9295752531993405, + "grad_norm": 4.838602542877197, + "learning_rate": 1.2113151884267264e-05, + "loss": 5.879899291992188, + "step": 29600 + }, + { + "epoch": 0.9303603674334616, + "grad_norm": 6.804969310760498, + "learning_rate": 1.2102049978143521e-05, + "loss": 5.904154052734375, + "step": 29625 + }, + { + "epoch": 0.9311454816675826, + "grad_norm": 4.918222427368164, + "learning_rate": 1.2090946369315324e-05, + "loss": 5.869876098632813, + "step": 29650 + }, + { + "epoch": 0.9319305959017037, + "grad_norm": 7.121532917022705, + "learning_rate": 1.2079841074938133e-05, + "loss": 5.904607543945312, + "step": 29675 + }, + { + "epoch": 0.9327157101358248, + "grad_norm": 5.135168075561523, + "learning_rate": 1.2068734112169993e-05, + "loss": 5.878980102539063, + "step": 29700 + }, + { + "epoch": 0.9335008243699459, + "grad_norm": 4.123371124267578, + "learning_rate": 1.2057625498171547e-05, + "loss": 5.914134521484375, + "step": 29725 + }, + { + "epoch": 0.934285938604067, + "grad_norm": 5.819398403167725, + "learning_rate": 1.2046515250105973e-05, + "loss": 5.912734375, + "step": 29750 + }, + { + "epoch": 0.9350710528381879, + "grad_norm": 5.66091251373291, + "learning_rate": 1.2035403385138981e-05, + "loss": 5.8867340087890625, + "step": 29775 + }, + { + "epoch": 0.935856167072309, + "grad_norm": 5.8618645668029785, + "learning_rate": 1.2024289920438777e-05, + "loss": 5.884973754882813, + "step": 29800 + }, + { + "epoch": 0.9366412813064301, + "grad_norm": 5.066692352294922, + "learning_rate": 1.2013174873176044e-05, + "loss": 5.88591552734375, + "step": 29825 + }, + { + "epoch": 0.9374263955405512, + "grad_norm": 4.894474983215332, + "learning_rate": 1.2002058260523897e-05, + "loss": 5.91037109375, + "step": 29850 + }, + { + "epoch": 0.9382115097746723, + "grad_norm": 5.2974395751953125, + "learning_rate": 1.1990940099657887e-05, + "loss": 5.902234497070313, + "step": 29875 + }, + { + "epoch": 0.9389966240087932, + "grad_norm": 4.21222448348999, + "learning_rate": 1.197982040775594e-05, + "loss": 5.877948608398437, + "step": 29900 + }, + { + "epoch": 0.9397817382429143, + "grad_norm": 5.144382953643799, + "learning_rate": 1.1968699201998364e-05, + "loss": 5.884660034179688, + "step": 29925 + }, + { + "epoch": 0.9405668524770354, + "grad_norm": 4.32300329208374, + "learning_rate": 1.195757649956779e-05, + "loss": 5.889692993164062, + "step": 29950 + }, + { + "epoch": 0.9413519667111565, + "grad_norm": 3.6288626194000244, + "learning_rate": 1.1946452317649175e-05, + "loss": 5.890128173828125, + "step": 29975 + }, + { + "epoch": 0.9421370809452775, + "grad_norm": 4.684532165527344, + "learning_rate": 1.193532667342975e-05, + "loss": 5.882379760742188, + "step": 30000 + }, + { + "epoch": 0.9429221951793986, + "grad_norm": 6.159887790679932, + "learning_rate": 1.1924199584099018e-05, + "loss": 5.888679809570313, + "step": 30025 + }, + { + "epoch": 0.9437073094135197, + "grad_norm": 7.43757438659668, + "learning_rate": 1.1913071066848703e-05, + "loss": 5.899791259765625, + "step": 30050 + }, + { + "epoch": 0.9444924236476407, + "grad_norm": 3.7081398963928223, + "learning_rate": 1.1901941138872743e-05, + "loss": 5.885630493164062, + "step": 30075 + }, + { + "epoch": 0.9452775378817618, + "grad_norm": 3.068618059158325, + "learning_rate": 1.1890809817367249e-05, + "loss": 5.891066284179687, + "step": 30100 + }, + { + "epoch": 0.9460626521158828, + "grad_norm": 4.667649745941162, + "learning_rate": 1.1879677119530492e-05, + "loss": 5.899071044921875, + "step": 30125 + }, + { + "epoch": 0.9468477663500039, + "grad_norm": 5.238046646118164, + "learning_rate": 1.1868543062562866e-05, + "loss": 5.888322143554688, + "step": 30150 + }, + { + "epoch": 0.947632880584125, + "grad_norm": 4.612542152404785, + "learning_rate": 1.1857407663666867e-05, + "loss": 5.898612060546875, + "step": 30175 + }, + { + "epoch": 0.9484179948182461, + "grad_norm": 6.182141304016113, + "learning_rate": 1.1846270940047053e-05, + "loss": 5.897420043945313, + "step": 30200 + }, + { + "epoch": 0.9492031090523672, + "grad_norm": 4.700918197631836, + "learning_rate": 1.1835132908910051e-05, + "loss": 5.912654418945312, + "step": 30225 + }, + { + "epoch": 0.9499882232864881, + "grad_norm": 3.4082067012786865, + "learning_rate": 1.1823993587464486e-05, + "loss": 5.901951293945313, + "step": 30250 + }, + { + "epoch": 0.9507733375206092, + "grad_norm": 4.555963039398193, + "learning_rate": 1.1812852992920998e-05, + "loss": 5.907822265625, + "step": 30275 + }, + { + "epoch": 0.9515584517547303, + "grad_norm": 4.745087146759033, + "learning_rate": 1.1801711142492164e-05, + "loss": 5.891825561523437, + "step": 30300 + }, + { + "epoch": 0.9523435659888514, + "grad_norm": 6.651101112365723, + "learning_rate": 1.1790568053392535e-05, + "loss": 5.883617553710938, + "step": 30325 + }, + { + "epoch": 0.9531286802229725, + "grad_norm": 6.230369567871094, + "learning_rate": 1.1779423742838551e-05, + "loss": 5.881992797851563, + "step": 30350 + }, + { + "epoch": 0.9539137944570935, + "grad_norm": 4.444051265716553, + "learning_rate": 1.1768278228048553e-05, + "loss": 5.8847503662109375, + "step": 30375 + }, + { + "epoch": 0.9546989086912145, + "grad_norm": 4.834921360015869, + "learning_rate": 1.1757131526242735e-05, + "loss": 5.8880419921875, + "step": 30400 + }, + { + "epoch": 0.9554840229253356, + "grad_norm": 6.850707530975342, + "learning_rate": 1.1745983654643131e-05, + "loss": 5.9068670654296875, + "step": 30425 + }, + { + "epoch": 0.9562691371594567, + "grad_norm": 3.955620765686035, + "learning_rate": 1.1734834630473575e-05, + "loss": 5.883507080078125, + "step": 30450 + }, + { + "epoch": 0.9570542513935778, + "grad_norm": 5.340014457702637, + "learning_rate": 1.1723684470959692e-05, + "loss": 5.877402954101562, + "step": 30475 + }, + { + "epoch": 0.9578393656276988, + "grad_norm": 7.473015308380127, + "learning_rate": 1.171253319332885e-05, + "loss": 5.8789111328125, + "step": 30500 + }, + { + "epoch": 0.9586244798618199, + "grad_norm": 4.978881359100342, + "learning_rate": 1.1701380814810152e-05, + "loss": 5.884993286132812, + "step": 30525 + }, + { + "epoch": 0.959409594095941, + "grad_norm": 4.17647123336792, + "learning_rate": 1.1690227352634398e-05, + "loss": 5.8927587890625, + "step": 30550 + }, + { + "epoch": 0.960194708330062, + "grad_norm": 6.330019474029541, + "learning_rate": 1.1679072824034063e-05, + "loss": 5.89285400390625, + "step": 30575 + }, + { + "epoch": 0.9609798225641831, + "grad_norm": 6.325782775878906, + "learning_rate": 1.166791724624328e-05, + "loss": 5.894048461914062, + "step": 30600 + }, + { + "epoch": 0.9617649367983041, + "grad_norm": 4.8818511962890625, + "learning_rate": 1.1656760636497777e-05, + "loss": 5.8911279296875, + "step": 30625 + }, + { + "epoch": 0.9625500510324252, + "grad_norm": 5.705923080444336, + "learning_rate": 1.1645603012034906e-05, + "loss": 5.8723388671875, + "step": 30650 + }, + { + "epoch": 0.9633351652665463, + "grad_norm": 6.274566173553467, + "learning_rate": 1.1634444390093569e-05, + "loss": 5.889078979492187, + "step": 30675 + }, + { + "epoch": 0.9641202795006674, + "grad_norm": 5.460311412811279, + "learning_rate": 1.1623284787914214e-05, + "loss": 5.876781616210938, + "step": 30700 + }, + { + "epoch": 0.9649053937347885, + "grad_norm": 5.316038608551025, + "learning_rate": 1.1612124222738807e-05, + "loss": 5.88171142578125, + "step": 30725 + }, + { + "epoch": 0.9656905079689094, + "grad_norm": 4.721954345703125, + "learning_rate": 1.160096271181079e-05, + "loss": 5.898412475585937, + "step": 30750 + }, + { + "epoch": 0.9664756222030305, + "grad_norm": 6.401038646697998, + "learning_rate": 1.1589800272375085e-05, + "loss": 5.880723876953125, + "step": 30775 + }, + { + "epoch": 0.9672607364371516, + "grad_norm": 5.049510955810547, + "learning_rate": 1.1578636921678029e-05, + "loss": 5.897097778320313, + "step": 30800 + }, + { + "epoch": 0.9680458506712727, + "grad_norm": 4.422399520874023, + "learning_rate": 1.1567472676967384e-05, + "loss": 5.897332763671875, + "step": 30825 + }, + { + "epoch": 0.9688309649053938, + "grad_norm": 5.713922500610352, + "learning_rate": 1.1556307555492276e-05, + "loss": 5.883538818359375, + "step": 30850 + }, + { + "epoch": 0.9696160791395148, + "grad_norm": 5.122622966766357, + "learning_rate": 1.1545141574503203e-05, + "loss": 5.883052978515625, + "step": 30875 + }, + { + "epoch": 0.9704011933736358, + "grad_norm": 5.051434516906738, + "learning_rate": 1.1533974751251977e-05, + "loss": 5.879835205078125, + "step": 30900 + }, + { + "epoch": 0.9711863076077569, + "grad_norm": 4.754411220550537, + "learning_rate": 1.1522807102991722e-05, + "loss": 5.866260986328125, + "step": 30925 + }, + { + "epoch": 0.971971421841878, + "grad_norm": 5.359310626983643, + "learning_rate": 1.151163864697683e-05, + "loss": 5.895767822265625, + "step": 30950 + }, + { + "epoch": 0.9727565360759991, + "grad_norm": 5.119156360626221, + "learning_rate": 1.1500469400462944e-05, + "loss": 5.907258911132812, + "step": 30975 + }, + { + "epoch": 0.9735416503101201, + "grad_norm": 5.069839000701904, + "learning_rate": 1.1489299380706921e-05, + "loss": 5.886080847846137, + "step": 31000 + }, + { + "epoch": 0.9743267645442412, + "grad_norm": 5.304202556610107, + "learning_rate": 1.1478128604966829e-05, + "loss": 5.892762451171875, + "step": 31025 + }, + { + "epoch": 0.9751118787783623, + "grad_norm": 4.514579772949219, + "learning_rate": 1.1466957090501889e-05, + "loss": 5.893353881835938, + "step": 31050 + }, + { + "epoch": 0.9758969930124833, + "grad_norm": 5.197580337524414, + "learning_rate": 1.1455784854572465e-05, + "loss": 5.87187255859375, + "step": 31075 + }, + { + "epoch": 0.9766821072466044, + "grad_norm": 7.033134937286377, + "learning_rate": 1.1444611914440043e-05, + "loss": 5.875283203125, + "step": 31100 + }, + { + "epoch": 0.9774672214807254, + "grad_norm": 4.527016639709473, + "learning_rate": 1.1433438287367191e-05, + "loss": 5.902391967773437, + "step": 31125 + }, + { + "epoch": 0.9782523357148465, + "grad_norm": 5.6208577156066895, + "learning_rate": 1.1422263990617544e-05, + "loss": 5.890577697753907, + "step": 31150 + }, + { + "epoch": 0.9790374499489676, + "grad_norm": 4.472509384155273, + "learning_rate": 1.1411089041455766e-05, + "loss": 5.875953369140625, + "step": 31175 + }, + { + "epoch": 0.9798225641830887, + "grad_norm": 4.124829292297363, + "learning_rate": 1.1399913457147525e-05, + "loss": 5.8969091796875, + "step": 31200 + }, + { + "epoch": 0.9806076784172097, + "grad_norm": 5.888116836547852, + "learning_rate": 1.1388737254959483e-05, + "loss": 5.864260864257813, + "step": 31225 + }, + { + "epoch": 0.9813927926513307, + "grad_norm": 3.933732748031616, + "learning_rate": 1.137756045215925e-05, + "loss": 5.90355712890625, + "step": 31250 + }, + { + "epoch": 0.9821779068854518, + "grad_norm": 4.915975093841553, + "learning_rate": 1.1366383066015365e-05, + "loss": 5.902286987304688, + "step": 31275 + }, + { + "epoch": 0.9829630211195729, + "grad_norm": 4.640792369842529, + "learning_rate": 1.135520511379726e-05, + "loss": 5.893407592773437, + "step": 31300 + }, + { + "epoch": 0.983748135353694, + "grad_norm": 4.571191787719727, + "learning_rate": 1.1344026612775258e-05, + "loss": 5.889780883789062, + "step": 31325 + }, + { + "epoch": 0.9845332495878151, + "grad_norm": 4.917351245880127, + "learning_rate": 1.1332847580220511e-05, + "loss": 5.871256103515625, + "step": 31350 + }, + { + "epoch": 0.985318363821936, + "grad_norm": 5.146409034729004, + "learning_rate": 1.1321668033405014e-05, + "loss": 5.887171020507813, + "step": 31375 + }, + { + "epoch": 0.9861034780560571, + "grad_norm": 4.695837020874023, + "learning_rate": 1.1310487989601532e-05, + "loss": 5.889156494140625, + "step": 31400 + }, + { + "epoch": 0.9868885922901782, + "grad_norm": 4.468144416809082, + "learning_rate": 1.129930746608362e-05, + "loss": 5.888562622070313, + "step": 31425 + }, + { + "epoch": 0.9876737065242993, + "grad_norm": 4.083787441253662, + "learning_rate": 1.1288126480125558e-05, + "loss": 5.913250732421875, + "step": 31450 + }, + { + "epoch": 0.9884588207584204, + "grad_norm": 6.47160005569458, + "learning_rate": 1.127694504900235e-05, + "loss": 5.885429077148437, + "step": 31475 + }, + { + "epoch": 0.9892439349925414, + "grad_norm": 3.1641154289245605, + "learning_rate": 1.1265763189989689e-05, + "loss": 5.882817993164062, + "step": 31500 + }, + { + "epoch": 0.9900290492266625, + "grad_norm": 5.562784671783447, + "learning_rate": 1.1254580920363918e-05, + "loss": 5.866495361328125, + "step": 31525 + }, + { + "epoch": 0.9908141634607835, + "grad_norm": 4.698551654815674, + "learning_rate": 1.1243398257402021e-05, + "loss": 5.88704833984375, + "step": 31550 + }, + { + "epoch": 0.9915992776949046, + "grad_norm": 5.829992294311523, + "learning_rate": 1.1232215218381597e-05, + "loss": 5.90143310546875, + "step": 31575 + }, + { + "epoch": 0.9923843919290257, + "grad_norm": 4.75034236907959, + "learning_rate": 1.1221031820580814e-05, + "loss": 5.896345825195312, + "step": 31600 + }, + { + "epoch": 0.9931695061631467, + "grad_norm": 4.485020637512207, + "learning_rate": 1.12098480812784e-05, + "loss": 5.887971801757812, + "step": 31625 + }, + { + "epoch": 0.9939546203972678, + "grad_norm": 4.868165969848633, + "learning_rate": 1.119866401775361e-05, + "loss": 5.877945556640625, + "step": 31650 + }, + { + "epoch": 0.9947397346313889, + "grad_norm": 4.648759365081787, + "learning_rate": 1.1187479647286204e-05, + "loss": 5.8821868896484375, + "step": 31675 + }, + { + "epoch": 0.99552484886551, + "grad_norm": 5.286471843719482, + "learning_rate": 1.1176294987156409e-05, + "loss": 5.89136962890625, + "step": 31700 + }, + { + "epoch": 0.996309963099631, + "grad_norm": 5.576229572296143, + "learning_rate": 1.11651100546449e-05, + "loss": 5.902265930175782, + "step": 31725 + }, + { + "epoch": 0.997095077333752, + "grad_norm": 3.987130641937256, + "learning_rate": 1.1153924867032777e-05, + "loss": 5.8894580078125, + "step": 31750 + }, + { + "epoch": 0.9978801915678731, + "grad_norm": 5.2408270835876465, + "learning_rate": 1.1142739441601538e-05, + "loss": 5.885211791992187, + "step": 31775 + }, + { + "epoch": 0.9986653058019942, + "grad_norm": 4.77084493637085, + "learning_rate": 1.1131553795633036e-05, + "loss": 5.869826049804687, + "step": 31800 + }, + { + "epoch": 0.9994504200361153, + "grad_norm": 4.0268683433532715, + "learning_rate": 1.1120367946409478e-05, + "loss": 5.887615966796875, + "step": 31825 + }, + { + "epoch": 1.0, + "eval_loss": 0.15892985463142395, + "eval_runtime": 5099.4777, + "eval_samples_per_second": 4.251, + "eval_steps_per_second": 0.266, + "eval_wer": 4.302831983058886, + "step": 31843 + }, + { + "epoch": 1.000219831985554, + "grad_norm": 6.201948642730713, + "learning_rate": 1.1109181911213374e-05, + "loss": 5.754544067382812, + "step": 31850 + }, + { + "epoch": 1.001004946219675, + "grad_norm": 5.235509872436523, + "learning_rate": 1.1097995707327532e-05, + "loss": 5.83971435546875, + "step": 31875 + }, + { + "epoch": 1.001790060453796, + "grad_norm": 4.28889799118042, + "learning_rate": 1.1086809352035004e-05, + "loss": 5.853460083007812, + "step": 31900 + }, + { + "epoch": 1.002575174687917, + "grad_norm": 3.670464515686035, + "learning_rate": 1.1075622862619102e-05, + "loss": 5.834912719726563, + "step": 31925 + }, + { + "epoch": 1.003360288922038, + "grad_norm": 4.201482772827148, + "learning_rate": 1.1064436256363318e-05, + "loss": 5.847487182617187, + "step": 31950 + }, + { + "epoch": 1.0041454031561592, + "grad_norm": 6.409285068511963, + "learning_rate": 1.1053249550551341e-05, + "loss": 5.830325444539388, + "step": 31975 + }, + { + "epoch": 1.0049305173902803, + "grad_norm": 3.319579601287842, + "learning_rate": 1.1042062762467009e-05, + "loss": 5.872825317382812, + "step": 32000 + }, + { + "epoch": 1.0057156316244014, + "grad_norm": 5.832342147827148, + "learning_rate": 1.1030875909394292e-05, + "loss": 5.857715454101562, + "step": 32025 + }, + { + "epoch": 1.0065007458585224, + "grad_norm": 6.230907917022705, + "learning_rate": 1.101968900861725e-05, + "loss": 5.829619140625, + "step": 32050 + }, + { + "epoch": 1.0072858600926435, + "grad_norm": 5.264492034912109, + "learning_rate": 1.1008502077420023e-05, + "loss": 5.844229736328125, + "step": 32075 + }, + { + "epoch": 1.0080709743267646, + "grad_norm": 3.6380555629730225, + "learning_rate": 1.0997315133086801e-05, + "loss": 5.853093872070312, + "step": 32100 + }, + { + "epoch": 1.0088560885608857, + "grad_norm": 3.9729838371276855, + "learning_rate": 1.0986128192901786e-05, + "loss": 5.85333984375, + "step": 32125 + }, + { + "epoch": 1.0096412027950066, + "grad_norm": 6.982206344604492, + "learning_rate": 1.0974941274149184e-05, + "loss": 5.837183837890625, + "step": 32150 + }, + { + "epoch": 1.0104263170291277, + "grad_norm": 3.7375454902648926, + "learning_rate": 1.0963754394113159e-05, + "loss": 5.852042236328125, + "step": 32175 + }, + { + "epoch": 1.0112114312632488, + "grad_norm": 4.6317973136901855, + "learning_rate": 1.0952567570077821e-05, + "loss": 5.840907592773437, + "step": 32200 + }, + { + "epoch": 1.0119965454973698, + "grad_norm": 3.499211549758911, + "learning_rate": 1.0941380819327188e-05, + "loss": 5.860418090820312, + "step": 32225 + }, + { + "epoch": 1.012781659731491, + "grad_norm": 4.53197717666626, + "learning_rate": 1.0930194159145171e-05, + "loss": 5.853024291992187, + "step": 32250 + }, + { + "epoch": 1.013566773965612, + "grad_norm": 6.043356895446777, + "learning_rate": 1.0919007606815538e-05, + "loss": 5.870966796875, + "step": 32275 + }, + { + "epoch": 1.014351888199733, + "grad_norm": 3.850526809692383, + "learning_rate": 1.0907821179621886e-05, + "loss": 5.8407177734375, + "step": 32300 + }, + { + "epoch": 1.0151370024338542, + "grad_norm": 5.243224143981934, + "learning_rate": 1.0896634894847625e-05, + "loss": 5.8623828125, + "step": 32325 + }, + { + "epoch": 1.0159221166679753, + "grad_norm": 4.869398593902588, + "learning_rate": 1.0885448769775945e-05, + "loss": 5.85316162109375, + "step": 32350 + }, + { + "epoch": 1.0167072309020964, + "grad_norm": 3.4251067638397217, + "learning_rate": 1.0874262821689789e-05, + "loss": 5.85655029296875, + "step": 32375 + }, + { + "epoch": 1.0174923451362172, + "grad_norm": 6.494776248931885, + "learning_rate": 1.0863077067871818e-05, + "loss": 5.866388549804688, + "step": 32400 + }, + { + "epoch": 1.0182774593703383, + "grad_norm": 4.833017826080322, + "learning_rate": 1.08518915256044e-05, + "loss": 5.85609130859375, + "step": 32425 + }, + { + "epoch": 1.0190625736044594, + "grad_norm": 5.464127540588379, + "learning_rate": 1.0840706212169582e-05, + "loss": 5.834376220703125, + "step": 32450 + }, + { + "epoch": 1.0198476878385805, + "grad_norm": NaN, + "learning_rate": 1.0829521144849054e-05, + "loss": 5.844287719726562, + "step": 32475 + }, + { + "epoch": 1.0206328020727016, + "grad_norm": 4.227790832519531, + "learning_rate": 1.0818336340924112e-05, + "loss": 5.842628173828125, + "step": 32500 + }, + { + "epoch": 1.0214179163068227, + "grad_norm": 4.91751766204834, + "learning_rate": 1.0807151817675654e-05, + "loss": 5.842296142578125, + "step": 32525 + }, + { + "epoch": 1.0222030305409437, + "grad_norm": 4.032611846923828, + "learning_rate": 1.079596759238416e-05, + "loss": 5.865907592773437, + "step": 32550 + }, + { + "epoch": 1.0229881447750648, + "grad_norm": 3.5838093757629395, + "learning_rate": 1.078478368232963e-05, + "loss": 5.834075317382813, + "step": 32575 + }, + { + "epoch": 1.023773259009186, + "grad_norm": 5.341316223144531, + "learning_rate": 1.0773600104791582e-05, + "loss": 5.851273803710938, + "step": 32600 + }, + { + "epoch": 1.024558373243307, + "grad_norm": 3.4830095767974854, + "learning_rate": 1.0762416877049016e-05, + "loss": 5.8428955078125, + "step": 32625 + }, + { + "epoch": 1.0253434874774279, + "grad_norm": 4.629174709320068, + "learning_rate": 1.0751234016380405e-05, + "loss": 5.851924438476562, + "step": 32650 + }, + { + "epoch": 1.026128601711549, + "grad_norm": 4.708155632019043, + "learning_rate": 1.0740051540063643e-05, + "loss": 5.8463983154296875, + "step": 32675 + }, + { + "epoch": 1.02691371594567, + "grad_norm": 5.8721022605896, + "learning_rate": 1.0728869465376036e-05, + "loss": 5.852442626953125, + "step": 32700 + }, + { + "epoch": 1.0276988301797911, + "grad_norm": 5.075923919677734, + "learning_rate": 1.0717687809594259e-05, + "loss": 5.867118530273437, + "step": 32725 + }, + { + "epoch": 1.0284839444139122, + "grad_norm": 8.236954689025879, + "learning_rate": 1.0706506589994362e-05, + "loss": 5.852444458007812, + "step": 32750 + }, + { + "epoch": 1.0292690586480333, + "grad_norm": 4.746253967285156, + "learning_rate": 1.0695325823851694e-05, + "loss": 5.851539306640625, + "step": 32775 + }, + { + "epoch": 1.0300541728821544, + "grad_norm": 4.221921443939209, + "learning_rate": 1.0684145528440923e-05, + "loss": 5.851951293945312, + "step": 32800 + }, + { + "epoch": 1.0308392871162755, + "grad_norm": 4.457822322845459, + "learning_rate": 1.0672965721035982e-05, + "loss": 5.852950439453125, + "step": 32825 + }, + { + "epoch": 1.0316244013503966, + "grad_norm": 4.568160533905029, + "learning_rate": 1.066178641891005e-05, + "loss": 5.848480834960937, + "step": 32850 + }, + { + "epoch": 1.0324095155845177, + "grad_norm": 4.394083499908447, + "learning_rate": 1.0650607639335529e-05, + "loss": 5.84037353515625, + "step": 32875 + }, + { + "epoch": 1.0331946298186385, + "grad_norm": 5.1555094718933105, + "learning_rate": 1.063942939958401e-05, + "loss": 5.849656982421875, + "step": 32900 + }, + { + "epoch": 1.0339797440527596, + "grad_norm": 6.078909397125244, + "learning_rate": 1.0628251716926246e-05, + "loss": 5.842965087890625, + "step": 32925 + }, + { + "epoch": 1.0347648582868807, + "grad_norm": 3.9934213161468506, + "learning_rate": 1.061707460863214e-05, + "loss": 5.852908325195313, + "step": 32950 + }, + { + "epoch": 1.0355499725210018, + "grad_norm": 3.692906141281128, + "learning_rate": 1.0605898091970703e-05, + "loss": 5.847462158203125, + "step": 32975 + }, + { + "epoch": 1.0363350867551229, + "grad_norm": 4.510336875915527, + "learning_rate": 1.0594722184210029e-05, + "loss": 5.847506103515625, + "step": 33000 + }, + { + "epoch": 1.037120200989244, + "grad_norm": 5.821335315704346, + "learning_rate": 1.0583546902617272e-05, + "loss": 5.864319458007812, + "step": 33025 + }, + { + "epoch": 1.037905315223365, + "grad_norm": 5.763269901275635, + "learning_rate": 1.0572372264458618e-05, + "loss": 5.839691162109375, + "step": 33050 + }, + { + "epoch": 1.0386904294574861, + "grad_norm": 5.298308372497559, + "learning_rate": 1.0561198286999268e-05, + "loss": 5.860303344726563, + "step": 33075 + }, + { + "epoch": 1.0394755436916072, + "grad_norm": 4.343894958496094, + "learning_rate": 1.0550024987503386e-05, + "loss": 5.8540496826171875, + "step": 33100 + }, + { + "epoch": 1.040260657925728, + "grad_norm": 4.157199382781982, + "learning_rate": 1.05388523832341e-05, + "loss": 5.855509033203125, + "step": 33125 + }, + { + "epoch": 1.0410457721598492, + "grad_norm": 4.165585994720459, + "learning_rate": 1.0527680491453462e-05, + "loss": 5.844832763671875, + "step": 33150 + }, + { + "epoch": 1.0418308863939703, + "grad_norm": 6.406728744506836, + "learning_rate": 1.0516509329422424e-05, + "loss": 5.833931884765625, + "step": 33175 + }, + { + "epoch": 1.0426160006280913, + "grad_norm": 6.856272220611572, + "learning_rate": 1.0505338914400806e-05, + "loss": 5.83986572265625, + "step": 33200 + }, + { + "epoch": 1.0434011148622124, + "grad_norm": 4.098104953765869, + "learning_rate": 1.0494169263647276e-05, + "loss": 5.852251586914062, + "step": 33225 + }, + { + "epoch": 1.0441862290963335, + "grad_norm": 3.7272002696990967, + "learning_rate": 1.0483000394419324e-05, + "loss": 5.834419555664063, + "step": 33250 + }, + { + "epoch": 1.0449713433304546, + "grad_norm": 4.725160121917725, + "learning_rate": 1.0471832323973232e-05, + "loss": 5.844205322265625, + "step": 33275 + }, + { + "epoch": 1.0457564575645757, + "grad_norm": 5.677024841308594, + "learning_rate": 1.0460665069564042e-05, + "loss": 5.8617791748046875, + "step": 33300 + }, + { + "epoch": 1.0465415717986968, + "grad_norm": 5.23048210144043, + "learning_rate": 1.0449498648445537e-05, + "loss": 5.83403076171875, + "step": 33325 + }, + { + "epoch": 1.0473266860328179, + "grad_norm": 4.655245304107666, + "learning_rate": 1.0438333077870224e-05, + "loss": 5.853594970703125, + "step": 33350 + }, + { + "epoch": 1.048111800266939, + "grad_norm": 3.4963579177856445, + "learning_rate": 1.042716837508928e-05, + "loss": 5.831986694335938, + "step": 33375 + }, + { + "epoch": 1.0488969145010598, + "grad_norm": 4.421491622924805, + "learning_rate": 1.041600455735255e-05, + "loss": 5.841192016601562, + "step": 33400 + }, + { + "epoch": 1.049682028735181, + "grad_norm": 4.6999030113220215, + "learning_rate": 1.0404841641908506e-05, + "loss": 5.85521484375, + "step": 33425 + }, + { + "epoch": 1.050467142969302, + "grad_norm": 4.245143890380859, + "learning_rate": 1.0393679646004237e-05, + "loss": 5.849547729492188, + "step": 33450 + }, + { + "epoch": 1.051252257203423, + "grad_norm": 4.18108606338501, + "learning_rate": 1.0382518586885397e-05, + "loss": 5.86058349609375, + "step": 33475 + }, + { + "epoch": 1.0520373714375442, + "grad_norm": 5.060965061187744, + "learning_rate": 1.0371358481796205e-05, + "loss": 5.8409521484375, + "step": 33500 + }, + { + "epoch": 1.0528224856716653, + "grad_norm": 5.54477596282959, + "learning_rate": 1.0360199347979392e-05, + "loss": 5.839140625, + "step": 33525 + }, + { + "epoch": 1.0536075999057863, + "grad_norm": 4.286438941955566, + "learning_rate": 1.0349041202676205e-05, + "loss": 5.868643188476563, + "step": 33550 + }, + { + "epoch": 1.0543927141399074, + "grad_norm": 7.1363325119018555, + "learning_rate": 1.0337884063126353e-05, + "loss": 5.844595336914063, + "step": 33575 + }, + { + "epoch": 1.0551778283740285, + "grad_norm": 4.307097434997559, + "learning_rate": 1.0326727946567993e-05, + "loss": 5.83691650390625, + "step": 33600 + }, + { + "epoch": 1.0559629426081494, + "grad_norm": 5.828007698059082, + "learning_rate": 1.03155728702377e-05, + "loss": 5.8560888671875, + "step": 33625 + }, + { + "epoch": 1.0567480568422705, + "grad_norm": 3.7977066040039062, + "learning_rate": 1.0304418851370447e-05, + "loss": 5.86068115234375, + "step": 33650 + }, + { + "epoch": 1.0575331710763916, + "grad_norm": 4.777566909790039, + "learning_rate": 1.029326590719957e-05, + "loss": 5.84009521484375, + "step": 33675 + }, + { + "epoch": 1.0583182853105126, + "grad_norm": 3.3671932220458984, + "learning_rate": 1.0282114054956744e-05, + "loss": 5.85011474609375, + "step": 33700 + }, + { + "epoch": 1.0591033995446337, + "grad_norm": 4.755646228790283, + "learning_rate": 1.0270963311871954e-05, + "loss": 5.83819091796875, + "step": 33725 + }, + { + "epoch": 1.0598885137787548, + "grad_norm": 4.3460798263549805, + "learning_rate": 1.0259813695173477e-05, + "loss": 5.831466064453125, + "step": 33750 + }, + { + "epoch": 1.060673628012876, + "grad_norm": 6.6369709968566895, + "learning_rate": 1.0248665222087853e-05, + "loss": 5.857848510742188, + "step": 33775 + }, + { + "epoch": 1.061458742246997, + "grad_norm": 3.505983591079712, + "learning_rate": 1.0237517909839845e-05, + "loss": 5.84196044921875, + "step": 33800 + }, + { + "epoch": 1.062243856481118, + "grad_norm": 6.325646877288818, + "learning_rate": 1.022637177565242e-05, + "loss": 5.849762573242187, + "step": 33825 + }, + { + "epoch": 1.0630289707152392, + "grad_norm": 4.175459861755371, + "learning_rate": 1.0215226836746743e-05, + "loss": 5.8562493324279785, + "step": 33850 + }, + { + "epoch": 1.06381408494936, + "grad_norm": 7.236473560333252, + "learning_rate": 1.0204083110342116e-05, + "loss": 5.845634155273437, + "step": 33875 + }, + { + "epoch": 1.0645991991834811, + "grad_norm": 3.805389642715454, + "learning_rate": 1.0192940613655975e-05, + "loss": 5.848778076171875, + "step": 33900 + }, + { + "epoch": 1.0653843134176022, + "grad_norm": 5.397750377655029, + "learning_rate": 1.0181799363903849e-05, + "loss": 5.85615966796875, + "step": 33925 + }, + { + "epoch": 1.0661694276517233, + "grad_norm": 5.490453243255615, + "learning_rate": 1.0170659378299353e-05, + "loss": 5.8550341796875, + "step": 33950 + }, + { + "epoch": 1.0669545418858444, + "grad_norm": 7.175689697265625, + "learning_rate": 1.0159520674054133e-05, + "loss": 5.833981323242187, + "step": 33975 + }, + { + "epoch": 1.0677396561199655, + "grad_norm": 4.0517258644104, + "learning_rate": 1.014838326837787e-05, + "loss": 5.836512451171875, + "step": 34000 + }, + { + "epoch": 1.0685247703540866, + "grad_norm": 5.39613676071167, + "learning_rate": 1.0137247178478226e-05, + "loss": 5.844761962890625, + "step": 34025 + }, + { + "epoch": 1.0693098845882076, + "grad_norm": 4.433953285217285, + "learning_rate": 1.0126112421560843e-05, + "loss": 5.82681396484375, + "step": 34050 + }, + { + "epoch": 1.0700949988223287, + "grad_norm": 3.3792924880981445, + "learning_rate": 1.0114979014829292e-05, + "loss": 5.8620166015625, + "step": 34075 + }, + { + "epoch": 1.0708801130564498, + "grad_norm": 3.7765183448791504, + "learning_rate": 1.0103846975485064e-05, + "loss": 5.85434326171875, + "step": 34100 + }, + { + "epoch": 1.0716652272905707, + "grad_norm": 5.131790637969971, + "learning_rate": 1.0092716320727531e-05, + "loss": 5.852839965820312, + "step": 34125 + }, + { + "epoch": 1.0724503415246918, + "grad_norm": 4.397039413452148, + "learning_rate": 1.008158706775394e-05, + "loss": 5.845491333007812, + "step": 34150 + }, + { + "epoch": 1.0732354557588129, + "grad_norm": 4.678694725036621, + "learning_rate": 1.0070459233759358e-05, + "loss": 5.852223510742188, + "step": 34175 + }, + { + "epoch": 1.074020569992934, + "grad_norm": 8.122096061706543, + "learning_rate": 1.0059332835936664e-05, + "loss": 5.847528076171875, + "step": 34200 + }, + { + "epoch": 1.074805684227055, + "grad_norm": 3.4901371002197266, + "learning_rate": 1.0048207891476516e-05, + "loss": 5.8442926025390625, + "step": 34225 + }, + { + "epoch": 1.0755907984611761, + "grad_norm": 4.018471717834473, + "learning_rate": 1.0037084417567333e-05, + "loss": 5.834841918945313, + "step": 34250 + }, + { + "epoch": 1.0763759126952972, + "grad_norm": 3.8245162963867188, + "learning_rate": 1.0025962431395257e-05, + "loss": 5.855857543945312, + "step": 34275 + }, + { + "epoch": 1.0771610269294183, + "grad_norm": 5.099288463592529, + "learning_rate": 1.0014841950144135e-05, + "loss": 5.85105224609375, + "step": 34300 + }, + { + "epoch": 1.0779461411635394, + "grad_norm": 3.918813705444336, + "learning_rate": 1.000372299099548e-05, + "loss": 5.839308471679687, + "step": 34325 + }, + { + "epoch": 1.0787312553976605, + "grad_norm": 4.0834197998046875, + "learning_rate": 9.992605571128467e-06, + "loss": 5.845870361328125, + "step": 34350 + }, + { + "epoch": 1.0795163696317813, + "grad_norm": 3.366459369659424, + "learning_rate": 9.981489707719877e-06, + "loss": 5.844082641601562, + "step": 34375 + }, + { + "epoch": 1.0803014838659024, + "grad_norm": 4.665246486663818, + "learning_rate": 9.970375417944103e-06, + "loss": 5.838689575195312, + "step": 34400 + }, + { + "epoch": 1.0810865981000235, + "grad_norm": 3.7909910678863525, + "learning_rate": 9.959262718973091e-06, + "loss": 5.858175048828125, + "step": 34425 + }, + { + "epoch": 1.0818717123341446, + "grad_norm": 6.114875316619873, + "learning_rate": 9.948151627976341e-06, + "loss": 5.852548217773437, + "step": 34450 + }, + { + "epoch": 1.0826568265682657, + "grad_norm": 3.9292078018188477, + "learning_rate": 9.937042162120862e-06, + "loss": 5.854252319335938, + "step": 34475 + }, + { + "epoch": 1.0834419408023868, + "grad_norm": 4.651977062225342, + "learning_rate": 9.925934338571155e-06, + "loss": 5.841199951171875, + "step": 34500 + }, + { + "epoch": 1.0842270550365078, + "grad_norm": 4.5292768478393555, + "learning_rate": 9.914828174489179e-06, + "loss": 5.841875610351562, + "step": 34525 + }, + { + "epoch": 1.085012169270629, + "grad_norm": 3.3002374172210693, + "learning_rate": 9.903723687034342e-06, + "loss": 5.841260986328125, + "step": 34550 + }, + { + "epoch": 1.08579728350475, + "grad_norm": 4.900968551635742, + "learning_rate": 9.892620893363444e-06, + "loss": 5.8454296875, + "step": 34575 + }, + { + "epoch": 1.086582397738871, + "grad_norm": 5.019837379455566, + "learning_rate": 9.881519810630677e-06, + "loss": 5.851170043945313, + "step": 34600 + }, + { + "epoch": 1.087367511972992, + "grad_norm": 4.494377613067627, + "learning_rate": 9.870420455987588e-06, + "loss": 5.8331451416015625, + "step": 34625 + }, + { + "epoch": 1.088152626207113, + "grad_norm": 9.16425609588623, + "learning_rate": 9.859322846583062e-06, + "loss": 5.870549926757812, + "step": 34650 + }, + { + "epoch": 1.0889377404412341, + "grad_norm": 4.704226016998291, + "learning_rate": 9.848226999563277e-06, + "loss": 5.84917724609375, + "step": 34675 + }, + { + "epoch": 1.0897228546753552, + "grad_norm": 5.916869163513184, + "learning_rate": 9.83713293207169e-06, + "loss": 5.842286376953125, + "step": 34700 + }, + { + "epoch": 1.0905079689094763, + "grad_norm": 5.392207145690918, + "learning_rate": 9.826040661249003e-06, + "loss": 5.844016723632812, + "step": 34725 + }, + { + "epoch": 1.0912930831435974, + "grad_norm": 4.64647912979126, + "learning_rate": 9.814950204233162e-06, + "loss": 5.8439208984375, + "step": 34750 + }, + { + "epoch": 1.0920781973777185, + "grad_norm": 5.687864780426025, + "learning_rate": 9.803861578159295e-06, + "loss": 5.848070678710937, + "step": 34775 + }, + { + "epoch": 1.0928633116118396, + "grad_norm": 5.121543884277344, + "learning_rate": 9.792774800159698e-06, + "loss": 5.839305419921875, + "step": 34800 + }, + { + "epoch": 1.0936484258459607, + "grad_norm": 4.261571407318115, + "learning_rate": 9.78168988736382e-06, + "loss": 5.845883178710937, + "step": 34825 + }, + { + "epoch": 1.0944335400800815, + "grad_norm": 4.041290760040283, + "learning_rate": 9.770606856898225e-06, + "loss": 5.830824584960937, + "step": 34850 + }, + { + "epoch": 1.0952186543142026, + "grad_norm": 3.8960187435150146, + "learning_rate": 9.759525725886575e-06, + "loss": 5.84013427734375, + "step": 34875 + }, + { + "epoch": 1.0960037685483237, + "grad_norm": 4.56716775894165, + "learning_rate": 9.748446511449592e-06, + "loss": 5.8437957763671875, + "step": 34900 + }, + { + "epoch": 1.0967888827824448, + "grad_norm": 4.612364768981934, + "learning_rate": 9.737369230705023e-06, + "loss": 5.834293212890625, + "step": 34925 + }, + { + "epoch": 1.0975739970165659, + "grad_norm": 4.841100692749023, + "learning_rate": 9.726293900767656e-06, + "loss": 5.850547485351562, + "step": 34950 + }, + { + "epoch": 1.098359111250687, + "grad_norm": 3.6838135719299316, + "learning_rate": 9.715220538749244e-06, + "loss": 5.854037475585938, + "step": 34975 + }, + { + "epoch": 1.099144225484808, + "grad_norm": 3.8636696338653564, + "learning_rate": 9.704149161758507e-06, + "loss": 5.850225219726562, + "step": 35000 + }, + { + "epoch": 1.0999293397189291, + "grad_norm": 3.8689072132110596, + "learning_rate": 9.693079786901092e-06, + "loss": 5.835115356445312, + "step": 35025 + }, + { + "epoch": 1.1007144539530502, + "grad_norm": 2.9639275074005127, + "learning_rate": 9.682012431279562e-06, + "loss": 5.840496215820313, + "step": 35050 + }, + { + "epoch": 1.1014995681871713, + "grad_norm": 4.371021270751953, + "learning_rate": 9.670947111993356e-06, + "loss": 5.84619384765625, + "step": 35075 + }, + { + "epoch": 1.1022846824212924, + "grad_norm": 4.025989532470703, + "learning_rate": 9.659883846138762e-06, + "loss": 5.845103759765625, + "step": 35100 + }, + { + "epoch": 1.1030697966554133, + "grad_norm": 4.776800155639648, + "learning_rate": 9.6488226508089e-06, + "loss": 5.823587646484375, + "step": 35125 + }, + { + "epoch": 1.1038549108895344, + "grad_norm": 5.064579963684082, + "learning_rate": 9.637763543093697e-06, + "loss": 5.831324462890625, + "step": 35150 + }, + { + "epoch": 1.1046400251236554, + "grad_norm": 4.657581329345703, + "learning_rate": 9.62670654007984e-06, + "loss": 5.843070068359375, + "step": 35175 + }, + { + "epoch": 1.1054251393577765, + "grad_norm": 4.387777805328369, + "learning_rate": 9.615651658850776e-06, + "loss": 5.832176513671875, + "step": 35200 + }, + { + "epoch": 1.1062102535918976, + "grad_norm": 3.2371129989624023, + "learning_rate": 9.604598916486666e-06, + "loss": 5.827352905273438, + "step": 35225 + }, + { + "epoch": 1.1069953678260187, + "grad_norm": 5.427971839904785, + "learning_rate": 9.593548330064377e-06, + "loss": 5.846790771484375, + "step": 35250 + }, + { + "epoch": 1.1077804820601398, + "grad_norm": 5.401098728179932, + "learning_rate": 9.582499916657435e-06, + "loss": 5.8313446044921875, + "step": 35275 + }, + { + "epoch": 1.1085655962942609, + "grad_norm": 4.998412132263184, + "learning_rate": 9.571453693336013e-06, + "loss": 5.834010620117187, + "step": 35300 + }, + { + "epoch": 1.109350710528382, + "grad_norm": 4.066189289093018, + "learning_rate": 9.560409677166895e-06, + "loss": 5.847559814453125, + "step": 35325 + }, + { + "epoch": 1.1101358247625028, + "grad_norm": 4.130155086517334, + "learning_rate": 9.549367885213466e-06, + "loss": 5.835978393554687, + "step": 35350 + }, + { + "epoch": 1.110920938996624, + "grad_norm": 4.5989460945129395, + "learning_rate": 9.53832833453566e-06, + "loss": 5.855612182617188, + "step": 35375 + }, + { + "epoch": 1.111706053230745, + "grad_norm": 4.297070026397705, + "learning_rate": 9.527291042189964e-06, + "loss": 5.856536865234375, + "step": 35400 + }, + { + "epoch": 1.112491167464866, + "grad_norm": 4.212048053741455, + "learning_rate": 9.516256025229358e-06, + "loss": 5.851868286132812, + "step": 35425 + }, + { + "epoch": 1.1132762816989872, + "grad_norm": 5.364826202392578, + "learning_rate": 9.505223300703326e-06, + "loss": 5.8524609375, + "step": 35450 + }, + { + "epoch": 1.1140613959331083, + "grad_norm": 5.790790557861328, + "learning_rate": 9.494192885657793e-06, + "loss": 5.84050537109375, + "step": 35475 + }, + { + "epoch": 1.1148465101672294, + "grad_norm": 4.9367852210998535, + "learning_rate": 9.483164797135127e-06, + "loss": 5.845260009765625, + "step": 35500 + }, + { + "epoch": 1.1156316244013504, + "grad_norm": 4.282532215118408, + "learning_rate": 9.472139052174092e-06, + "loss": 5.85559814453125, + "step": 35525 + }, + { + "epoch": 1.1164167386354715, + "grad_norm": 5.091096878051758, + "learning_rate": 9.461115667809847e-06, + "loss": 5.8360107421875, + "step": 35550 + }, + { + "epoch": 1.1172018528695926, + "grad_norm": 5.125621795654297, + "learning_rate": 9.45009466107388e-06, + "loss": 5.842407836914062, + "step": 35575 + }, + { + "epoch": 1.1179869671037137, + "grad_norm": 4.736443042755127, + "learning_rate": 9.439076048994029e-06, + "loss": 5.837450561523437, + "step": 35600 + }, + { + "epoch": 1.1187720813378346, + "grad_norm": 3.466925859451294, + "learning_rate": 9.428059848594413e-06, + "loss": 5.821846923828125, + "step": 35625 + }, + { + "epoch": 1.1195571955719557, + "grad_norm": 5.278817653656006, + "learning_rate": 9.417046076895446e-06, + "loss": 5.837213745117188, + "step": 35650 + }, + { + "epoch": 1.1203423098060767, + "grad_norm": 4.175888538360596, + "learning_rate": 9.406034750913767e-06, + "loss": 5.847453002929687, + "step": 35675 + }, + { + "epoch": 1.1211274240401978, + "grad_norm": 6.482365608215332, + "learning_rate": 9.39502588766225e-06, + "loss": 5.83150390625, + "step": 35700 + }, + { + "epoch": 1.121912538274319, + "grad_norm": 4.151716709136963, + "learning_rate": 9.384019504149957e-06, + "loss": 5.834659423828125, + "step": 35725 + }, + { + "epoch": 1.12269765250844, + "grad_norm": 3.213770866394043, + "learning_rate": 9.373015617382133e-06, + "loss": 5.839141235351563, + "step": 35750 + }, + { + "epoch": 1.123482766742561, + "grad_norm": 4.012185096740723, + "learning_rate": 9.362014244360145e-06, + "loss": 5.847507934570313, + "step": 35775 + }, + { + "epoch": 1.1242678809766822, + "grad_norm": 4.534530162811279, + "learning_rate": 9.351015402081488e-06, + "loss": 5.830602416992187, + "step": 35800 + }, + { + "epoch": 1.1250529952108033, + "grad_norm": 3.3402926921844482, + "learning_rate": 9.340019107539746e-06, + "loss": 5.850203247070312, + "step": 35825 + }, + { + "epoch": 1.1258381094449241, + "grad_norm": 5.739969730377197, + "learning_rate": 9.329025377724563e-06, + "loss": 5.8403955078125, + "step": 35850 + }, + { + "epoch": 1.1266232236790452, + "grad_norm": 6.560964107513428, + "learning_rate": 9.31803422962163e-06, + "loss": 5.829802856445313, + "step": 35875 + }, + { + "epoch": 1.1274083379131663, + "grad_norm": 4.672070503234863, + "learning_rate": 9.307045680212636e-06, + "loss": 5.847775268554687, + "step": 35900 + }, + { + "epoch": 1.1281934521472874, + "grad_norm": 4.267703533172607, + "learning_rate": 9.296059746475261e-06, + "loss": 5.838240966796875, + "step": 35925 + }, + { + "epoch": 1.1289785663814085, + "grad_norm": 5.147006034851074, + "learning_rate": 9.285076445383144e-06, + "loss": 5.856603393554687, + "step": 35950 + }, + { + "epoch": 1.1297636806155296, + "grad_norm": 4.877357006072998, + "learning_rate": 9.27409579390586e-06, + "loss": 5.833148193359375, + "step": 35975 + }, + { + "epoch": 1.1305487948496507, + "grad_norm": 3.4559009075164795, + "learning_rate": 9.263117809008885e-06, + "loss": 5.83130615234375, + "step": 36000 + }, + { + "epoch": 1.1313339090837717, + "grad_norm": 5.171026706695557, + "learning_rate": 9.252142507653577e-06, + "loss": 5.828690185546875, + "step": 36025 + }, + { + "epoch": 1.1321190233178928, + "grad_norm": 4.279566287994385, + "learning_rate": 9.241169906797138e-06, + "loss": 5.85171875, + "step": 36050 + }, + { + "epoch": 1.132904137552014, + "grad_norm": 5.133225440979004, + "learning_rate": 9.230200023392621e-06, + "loss": 5.8380908203125, + "step": 36075 + }, + { + "epoch": 1.133689251786135, + "grad_norm": 3.5686590671539307, + "learning_rate": 9.219232874388861e-06, + "loss": 5.825465698242187, + "step": 36100 + }, + { + "epoch": 1.1344743660202559, + "grad_norm": 4.3523030281066895, + "learning_rate": 9.208268476730475e-06, + "loss": 5.832857055664062, + "step": 36125 + }, + { + "epoch": 1.135259480254377, + "grad_norm": 3.987222671508789, + "learning_rate": 9.19730684735782e-06, + "loss": 5.843688354492188, + "step": 36150 + }, + { + "epoch": 1.136044594488498, + "grad_norm": 4.16348934173584, + "learning_rate": 9.186348003206992e-06, + "loss": 5.835902709960937, + "step": 36175 + }, + { + "epoch": 1.1368297087226191, + "grad_norm": 4.874325752258301, + "learning_rate": 9.175391961209776e-06, + "loss": 5.835609130859375, + "step": 36200 + }, + { + "epoch": 1.1376148229567402, + "grad_norm": 5.2950005531311035, + "learning_rate": 9.164438738293626e-06, + "loss": 5.847503051757813, + "step": 36225 + }, + { + "epoch": 1.1383999371908613, + "grad_norm": 4.716102123260498, + "learning_rate": 9.153488351381634e-06, + "loss": 5.838817749023438, + "step": 36250 + }, + { + "epoch": 1.1391850514249824, + "grad_norm": 4.168239593505859, + "learning_rate": 9.142540817392527e-06, + "loss": 5.838504638671875, + "step": 36275 + }, + { + "epoch": 1.1399701656591035, + "grad_norm": 3.634369134902954, + "learning_rate": 9.13159615324061e-06, + "loss": 5.829432373046875, + "step": 36300 + }, + { + "epoch": 1.1407552798932246, + "grad_norm": 4.732321262359619, + "learning_rate": 9.120654375835762e-06, + "loss": 5.8203271484375, + "step": 36325 + }, + { + "epoch": 1.1415403941273454, + "grad_norm": 5.148401737213135, + "learning_rate": 9.109715502083394e-06, + "loss": 5.8402978515625, + "step": 36350 + }, + { + "epoch": 1.1423255083614665, + "grad_norm": 5.036197662353516, + "learning_rate": 9.09877954888444e-06, + "loss": 5.84857666015625, + "step": 36375 + }, + { + "epoch": 1.1431106225955876, + "grad_norm": 4.464617729187012, + "learning_rate": 9.087846533135318e-06, + "loss": 5.8379888916015625, + "step": 36400 + }, + { + "epoch": 1.1438957368297087, + "grad_norm": 4.654612064361572, + "learning_rate": 9.076916471727903e-06, + "loss": 5.83460205078125, + "step": 36425 + }, + { + "epoch": 1.1446808510638298, + "grad_norm": 5.686317443847656, + "learning_rate": 9.065989381549509e-06, + "loss": 5.859210815429687, + "step": 36450 + }, + { + "epoch": 1.1454659652979509, + "grad_norm": 5.124882221221924, + "learning_rate": 9.055065279482861e-06, + "loss": 5.8325347900390625, + "step": 36475 + }, + { + "epoch": 1.146251079532072, + "grad_norm": 3.8178493976593018, + "learning_rate": 9.04414418240607e-06, + "loss": 5.82905517578125, + "step": 36500 + }, + { + "epoch": 1.147036193766193, + "grad_norm": 3.6566321849823, + "learning_rate": 9.033226107192592e-06, + "loss": 5.835380249023437, + "step": 36525 + }, + { + "epoch": 1.1478213080003141, + "grad_norm": 4.456484794616699, + "learning_rate": 9.022311070711227e-06, + "loss": 5.8515167236328125, + "step": 36550 + }, + { + "epoch": 1.148606422234435, + "grad_norm": 3.654580593109131, + "learning_rate": 9.011399089826074e-06, + "loss": 5.850076904296875, + "step": 36575 + }, + { + "epoch": 1.1493915364685563, + "grad_norm": 4.064891815185547, + "learning_rate": 9.000490181396511e-06, + "loss": 5.843077392578125, + "step": 36600 + }, + { + "epoch": 1.1501766507026772, + "grad_norm": 6.231684684753418, + "learning_rate": 8.989584362277173e-06, + "loss": 5.828570556640625, + "step": 36625 + }, + { + "epoch": 1.1509617649367982, + "grad_norm": 4.250801086425781, + "learning_rate": 8.978681649317916e-06, + "loss": 5.84646240234375, + "step": 36650 + }, + { + "epoch": 1.1517468791709193, + "grad_norm": 4.88543176651001, + "learning_rate": 8.9677820593638e-06, + "loss": 5.832628173828125, + "step": 36675 + }, + { + "epoch": 1.1525319934050404, + "grad_norm": 3.9674055576324463, + "learning_rate": 8.956885609255064e-06, + "loss": 5.823825073242188, + "step": 36700 + }, + { + "epoch": 1.1533171076391615, + "grad_norm": 4.803829193115234, + "learning_rate": 8.945992315827085e-06, + "loss": 5.853135986328125, + "step": 36725 + }, + { + "epoch": 1.1541022218732826, + "grad_norm": 6.630747318267822, + "learning_rate": 8.935102195910375e-06, + "loss": 5.836453247070312, + "step": 36750 + }, + { + "epoch": 1.1548873361074037, + "grad_norm": 4.103579998016357, + "learning_rate": 8.924215266330534e-06, + "loss": 5.834352416992187, + "step": 36775 + }, + { + "epoch": 1.1556724503415248, + "grad_norm": 4.198637008666992, + "learning_rate": 8.913331543908237e-06, + "loss": 5.830870971679688, + "step": 36800 + }, + { + "epoch": 1.1564575645756459, + "grad_norm": 6.361374378204346, + "learning_rate": 8.902451045459204e-06, + "loss": 5.846996459960938, + "step": 36825 + }, + { + "epoch": 1.1572426788097667, + "grad_norm": 5.386788845062256, + "learning_rate": 8.891573787794168e-06, + "loss": 5.8507232666015625, + "step": 36850 + }, + { + "epoch": 1.1580277930438878, + "grad_norm": 5.066763877868652, + "learning_rate": 8.880699787718866e-06, + "loss": 5.833203125, + "step": 36875 + }, + { + "epoch": 1.158812907278009, + "grad_norm": 4.132023811340332, + "learning_rate": 8.869829062033991e-06, + "loss": 5.835829467773437, + "step": 36900 + }, + { + "epoch": 1.15959802151213, + "grad_norm": 4.925024032592773, + "learning_rate": 8.858961627535185e-06, + "loss": 5.831136474609375, + "step": 36925 + }, + { + "epoch": 1.160383135746251, + "grad_norm": 3.536458969116211, + "learning_rate": 8.848097501012996e-06, + "loss": 5.84334228515625, + "step": 36950 + }, + { + "epoch": 1.1611682499803722, + "grad_norm": 5.168023586273193, + "learning_rate": 8.837236699252873e-06, + "loss": 5.831126708984375, + "step": 36975 + }, + { + "epoch": 1.1619533642144932, + "grad_norm": 3.4447929859161377, + "learning_rate": 8.826379239035118e-06, + "loss": 5.816966552734375, + "step": 37000 + }, + { + "epoch": 1.1627384784486143, + "grad_norm": 7.216976165771484, + "learning_rate": 8.815525137134877e-06, + "loss": 5.839212646484375, + "step": 37025 + }, + { + "epoch": 1.1635235926827354, + "grad_norm": 4.9169158935546875, + "learning_rate": 8.804674410322097e-06, + "loss": 5.866154174804688, + "step": 37050 + }, + { + "epoch": 1.1643087069168563, + "grad_norm": 4.403421401977539, + "learning_rate": 8.793827075361523e-06, + "loss": 5.835376586914062, + "step": 37075 + }, + { + "epoch": 1.1650938211509774, + "grad_norm": 3.4567816257476807, + "learning_rate": 8.782983149012656e-06, + "loss": 5.840734252929687, + "step": 37100 + }, + { + "epoch": 1.1658789353850985, + "grad_norm": 5.025951862335205, + "learning_rate": 8.772142648029724e-06, + "loss": 5.835670166015625, + "step": 37125 + }, + { + "epoch": 1.1666640496192195, + "grad_norm": 5.5395965576171875, + "learning_rate": 8.761305589161667e-06, + "loss": 5.8222589111328125, + "step": 37150 + }, + { + "epoch": 1.1674491638533406, + "grad_norm": 5.933688163757324, + "learning_rate": 8.750471989152112e-06, + "loss": 5.8224859619140625, + "step": 37175 + }, + { + "epoch": 1.1682342780874617, + "grad_norm": 3.6461875438690186, + "learning_rate": 8.739641864739335e-06, + "loss": 5.837327880859375, + "step": 37200 + }, + { + "epoch": 1.1690193923215828, + "grad_norm": 4.214407444000244, + "learning_rate": 8.72881523265624e-06, + "loss": 5.824078979492188, + "step": 37225 + }, + { + "epoch": 1.169804506555704, + "grad_norm": 4.370615482330322, + "learning_rate": 8.717992109630343e-06, + "loss": 5.8347705078125, + "step": 37250 + }, + { + "epoch": 1.170589620789825, + "grad_norm": 5.157212257385254, + "learning_rate": 8.707172512383735e-06, + "loss": 5.81637451171875, + "step": 37275 + }, + { + "epoch": 1.171374735023946, + "grad_norm": 3.435497522354126, + "learning_rate": 8.696356457633057e-06, + "loss": 5.83412841796875, + "step": 37300 + }, + { + "epoch": 1.1721598492580672, + "grad_norm": 5.077123165130615, + "learning_rate": 8.685543962089486e-06, + "loss": 5.838905029296875, + "step": 37325 + }, + { + "epoch": 1.172944963492188, + "grad_norm": 4.9990315437316895, + "learning_rate": 8.67473504245868e-06, + "loss": 5.849012451171875, + "step": 37350 + }, + { + "epoch": 1.173730077726309, + "grad_norm": 4.617331027984619, + "learning_rate": 8.663929715440798e-06, + "loss": 5.851357421875, + "step": 37375 + }, + { + "epoch": 1.1745151919604302, + "grad_norm": 4.231682777404785, + "learning_rate": 8.653127997730428e-06, + "loss": 5.8496728515625, + "step": 37400 + }, + { + "epoch": 1.1753003061945513, + "grad_norm": 4.029294013977051, + "learning_rate": 8.642329906016591e-06, + "loss": 5.827471313476562, + "step": 37425 + }, + { + "epoch": 1.1760854204286724, + "grad_norm": 5.980893611907959, + "learning_rate": 8.6315354569827e-06, + "loss": 5.833710327148437, + "step": 37450 + }, + { + "epoch": 1.1768705346627935, + "grad_norm": 4.82021427154541, + "learning_rate": 8.620744667306551e-06, + "loss": 5.840830688476562, + "step": 37475 + }, + { + "epoch": 1.1776556488969145, + "grad_norm": 6.041388988494873, + "learning_rate": 8.609957553660273e-06, + "loss": 5.8339996337890625, + "step": 37500 + }, + { + "epoch": 1.1784407631310356, + "grad_norm": 20.752763748168945, + "learning_rate": 8.599174132710321e-06, + "loss": 5.840697631835938, + "step": 37525 + }, + { + "epoch": 1.1792258773651567, + "grad_norm": 5.057047367095947, + "learning_rate": 8.588394421117445e-06, + "loss": 5.827725830078125, + "step": 37550 + }, + { + "epoch": 1.1800109915992776, + "grad_norm": 3.1863958835601807, + "learning_rate": 8.577618435536667e-06, + "loss": 5.8270751953125, + "step": 37575 + }, + { + "epoch": 1.1807961058333987, + "grad_norm": 4.257750511169434, + "learning_rate": 8.566846192617244e-06, + "loss": 5.853928833007813, + "step": 37600 + }, + { + "epoch": 1.1815812200675198, + "grad_norm": 5.2106099128723145, + "learning_rate": 8.556077709002662e-06, + "loss": 5.819297485351562, + "step": 37625 + }, + { + "epoch": 1.1823663343016408, + "grad_norm": 5.439002513885498, + "learning_rate": 8.545313001330582e-06, + "loss": 5.832151489257813, + "step": 37650 + }, + { + "epoch": 1.183151448535762, + "grad_norm": 4.59678840637207, + "learning_rate": 8.534552086232848e-06, + "loss": 5.832298583984375, + "step": 37675 + }, + { + "epoch": 1.183936562769883, + "grad_norm": 4.613393306732178, + "learning_rate": 8.523794980335438e-06, + "loss": 5.848527221679688, + "step": 37700 + }, + { + "epoch": 1.184721677004004, + "grad_norm": 4.536832332611084, + "learning_rate": 8.513041700258441e-06, + "loss": 5.8279351806640625, + "step": 37725 + }, + { + "epoch": 1.1855067912381252, + "grad_norm": 3.2508468627929688, + "learning_rate": 8.50229226261604e-06, + "loss": 5.826113891601563, + "step": 37750 + }, + { + "epoch": 1.1862919054722463, + "grad_norm": 4.999807357788086, + "learning_rate": 8.491546684016477e-06, + "loss": 5.827764282226562, + "step": 37775 + }, + { + "epoch": 1.1870770197063674, + "grad_norm": 4.805812835693359, + "learning_rate": 8.480804981062035e-06, + "loss": 5.829117431640625, + "step": 37800 + }, + { + "epoch": 1.1878621339404885, + "grad_norm": 5.440676689147949, + "learning_rate": 8.47006717034901e-06, + "loss": 5.821284790039062, + "step": 37825 + }, + { + "epoch": 1.1886472481746093, + "grad_norm": 3.7920405864715576, + "learning_rate": 8.459333268467678e-06, + "loss": 5.838054809570313, + "step": 37850 + }, + { + "epoch": 1.1894323624087304, + "grad_norm": 4.252841472625732, + "learning_rate": 8.448603292002285e-06, + "loss": 5.834248046875, + "step": 37875 + }, + { + "epoch": 1.1902174766428515, + "grad_norm": 3.8843419551849365, + "learning_rate": 8.437877257531003e-06, + "loss": 5.835272216796875, + "step": 37900 + }, + { + "epoch": 1.1910025908769726, + "grad_norm": 4.0430474281311035, + "learning_rate": 8.42715518162592e-06, + "loss": 5.821538696289062, + "step": 37925 + }, + { + "epoch": 1.1917877051110937, + "grad_norm": 3.9783449172973633, + "learning_rate": 8.416437080853004e-06, + "loss": 5.840226440429688, + "step": 37950 + }, + { + "epoch": 1.1925728193452148, + "grad_norm": 4.256082057952881, + "learning_rate": 8.405722971772086e-06, + "loss": 5.828871459960937, + "step": 37975 + }, + { + "epoch": 1.1933579335793358, + "grad_norm": 3.7039988040924072, + "learning_rate": 8.395012870936823e-06, + "loss": 5.825353393554687, + "step": 38000 + }, + { + "epoch": 1.194143047813457, + "grad_norm": 5.093902111053467, + "learning_rate": 8.384306794894685e-06, + "loss": 5.837097778320312, + "step": 38025 + }, + { + "epoch": 1.194928162047578, + "grad_norm": 4.561976432800293, + "learning_rate": 8.373604760186915e-06, + "loss": 5.834966430664062, + "step": 38050 + }, + { + "epoch": 1.1957132762816989, + "grad_norm": 4.786259651184082, + "learning_rate": 8.362906783348532e-06, + "loss": 5.8189892578125, + "step": 38075 + }, + { + "epoch": 1.19649839051582, + "grad_norm": 6.600461006164551, + "learning_rate": 8.352212880908259e-06, + "loss": 5.839473266601562, + "step": 38100 + }, + { + "epoch": 1.197283504749941, + "grad_norm": 4.072597503662109, + "learning_rate": 8.34152306938854e-06, + "loss": 5.84893310546875, + "step": 38125 + }, + { + "epoch": 1.1980686189840621, + "grad_norm": 4.464570999145508, + "learning_rate": 8.330837365305492e-06, + "loss": 5.844085693359375, + "step": 38150 + }, + { + "epoch": 1.1988537332181832, + "grad_norm": 4.982614517211914, + "learning_rate": 8.3201557851689e-06, + "loss": 5.829598388671875, + "step": 38175 + }, + { + "epoch": 1.1996388474523043, + "grad_norm": 4.606203556060791, + "learning_rate": 8.309478345482154e-06, + "loss": 5.82502685546875, + "step": 38200 + }, + { + "epoch": 1.2004239616864254, + "grad_norm": 4.746890068054199, + "learning_rate": 8.298805062742263e-06, + "loss": 5.838676147460937, + "step": 38225 + }, + { + "epoch": 1.2012090759205465, + "grad_norm": 3.6237220764160156, + "learning_rate": 8.288135953439808e-06, + "loss": 5.838619384765625, + "step": 38250 + }, + { + "epoch": 1.2019941901546676, + "grad_norm": 4.018848419189453, + "learning_rate": 8.277471034058928e-06, + "loss": 5.839178466796875, + "step": 38275 + }, + { + "epoch": 1.2027793043887887, + "grad_norm": 4.456884860992432, + "learning_rate": 8.266810321077284e-06, + "loss": 5.835109252929687, + "step": 38300 + }, + { + "epoch": 1.2035644186229097, + "grad_norm": 5.686465740203857, + "learning_rate": 8.256153830966032e-06, + "loss": 5.819737548828125, + "step": 38325 + }, + { + "epoch": 1.2043495328570306, + "grad_norm": 3.1492576599121094, + "learning_rate": 8.245501580189807e-06, + "loss": 5.822366943359375, + "step": 38350 + }, + { + "epoch": 1.2051346470911517, + "grad_norm": 5.302074909210205, + "learning_rate": 8.234853585206708e-06, + "loss": 5.846635131835938, + "step": 38375 + }, + { + "epoch": 1.2059197613252728, + "grad_norm": 3.7273895740509033, + "learning_rate": 8.22420986246824e-06, + "loss": 5.8269891357421875, + "step": 38400 + }, + { + "epoch": 1.2067048755593939, + "grad_norm": 4.921469211578369, + "learning_rate": 8.213570428419312e-06, + "loss": 5.841266479492187, + "step": 38425 + }, + { + "epoch": 1.207489989793515, + "grad_norm": 4.32681131362915, + "learning_rate": 8.202935299498212e-06, + "loss": 5.817271118164062, + "step": 38450 + }, + { + "epoch": 1.208275104027636, + "grad_norm": 5.008928298950195, + "learning_rate": 8.192304492136577e-06, + "loss": 5.829351196289062, + "step": 38475 + }, + { + "epoch": 1.2090602182617571, + "grad_norm": 4.4875264167785645, + "learning_rate": 8.181678022759353e-06, + "loss": 5.833065795898437, + "step": 38500 + }, + { + "epoch": 1.2098453324958782, + "grad_norm": 5.443745136260986, + "learning_rate": 8.171055907784805e-06, + "loss": 5.839254150390625, + "step": 38525 + }, + { + "epoch": 1.2106304467299993, + "grad_norm": 5.175171852111816, + "learning_rate": 8.160438163624456e-06, + "loss": 5.8294189453125, + "step": 38550 + }, + { + "epoch": 1.2114155609641202, + "grad_norm": 5.344850540161133, + "learning_rate": 8.149824806683078e-06, + "loss": 5.82380859375, + "step": 38575 + }, + { + "epoch": 1.2122006751982413, + "grad_norm": 4.938891887664795, + "learning_rate": 8.139215853358667e-06, + "loss": 5.840344848632813, + "step": 38600 + }, + { + "epoch": 1.2129857894323623, + "grad_norm": 4.5618438720703125, + "learning_rate": 8.128611320042413e-06, + "loss": 5.8257989501953125, + "step": 38625 + }, + { + "epoch": 1.2137709036664834, + "grad_norm": 3.6780197620391846, + "learning_rate": 8.118011223118678e-06, + "loss": 5.833349609375, + "step": 38650 + }, + { + "epoch": 1.2145560179006045, + "grad_norm": 3.0416276454925537, + "learning_rate": 8.107415578964974e-06, + "loss": 5.837657470703125, + "step": 38675 + }, + { + "epoch": 1.2153411321347256, + "grad_norm": 5.079187870025635, + "learning_rate": 8.096824403951926e-06, + "loss": 5.833922729492188, + "step": 38700 + }, + { + "epoch": 1.2161262463688467, + "grad_norm": 4.052175045013428, + "learning_rate": 8.08623771444326e-06, + "loss": 5.814656982421875, + "step": 38725 + }, + { + "epoch": 1.2169113606029678, + "grad_norm": 3.389702320098877, + "learning_rate": 8.075655526795763e-06, + "loss": 5.836128540039063, + "step": 38750 + }, + { + "epoch": 1.2176964748370889, + "grad_norm": 4.893049716949463, + "learning_rate": 8.065077857359275e-06, + "loss": 5.83212646484375, + "step": 38775 + }, + { + "epoch": 1.2184815890712097, + "grad_norm": 3.238645553588867, + "learning_rate": 8.054504722476659e-06, + "loss": 5.832366943359375, + "step": 38800 + }, + { + "epoch": 1.219266703305331, + "grad_norm": 4.4812397956848145, + "learning_rate": 8.043936138483757e-06, + "loss": 5.829714965820313, + "step": 38825 + }, + { + "epoch": 1.220051817539452, + "grad_norm": 6.916929244995117, + "learning_rate": 8.03337212170939e-06, + "loss": 5.823140019955843, + "step": 38850 + }, + { + "epoch": 1.220836931773573, + "grad_norm": 3.9350388050079346, + "learning_rate": 8.022812688475325e-06, + "loss": 5.819668579101562, + "step": 38875 + }, + { + "epoch": 1.221622046007694, + "grad_norm": 4.532763481140137, + "learning_rate": 8.012257855096242e-06, + "loss": 5.82296630859375, + "step": 38900 + }, + { + "epoch": 1.2224071602418152, + "grad_norm": 4.324350357055664, + "learning_rate": 8.001707637879712e-06, + "loss": 5.822684326171875, + "step": 38925 + }, + { + "epoch": 1.2231922744759363, + "grad_norm": 4.51919412612915, + "learning_rate": 7.991162053126176e-06, + "loss": 5.824349365234375, + "step": 38950 + }, + { + "epoch": 1.2239773887100573, + "grad_norm": 4.068173885345459, + "learning_rate": 7.980621117128926e-06, + "loss": 5.861052856445313, + "step": 38975 + }, + { + "epoch": 1.2247625029441784, + "grad_norm": 3.7237331867218018, + "learning_rate": 7.970084846174057e-06, + "loss": 5.824208374023438, + "step": 39000 + }, + { + "epoch": 1.2255476171782995, + "grad_norm": 4.1946702003479, + "learning_rate": 7.95955325654047e-06, + "loss": 5.819202270507812, + "step": 39025 + }, + { + "epoch": 1.2263327314124206, + "grad_norm": 5.305007457733154, + "learning_rate": 7.94902636449982e-06, + "loss": 5.840626220703125, + "step": 39050 + }, + { + "epoch": 1.2271178456465415, + "grad_norm": 4.323069095611572, + "learning_rate": 7.938504186316514e-06, + "loss": 5.836240844726563, + "step": 39075 + }, + { + "epoch": 1.2279029598806626, + "grad_norm": 4.745602607727051, + "learning_rate": 7.927986738247673e-06, + "loss": 5.8335943603515625, + "step": 39100 + }, + { + "epoch": 1.2286880741147836, + "grad_norm": 4.625556945800781, + "learning_rate": 7.917474036543111e-06, + "loss": 5.830066528320312, + "step": 39125 + }, + { + "epoch": 1.2294731883489047, + "grad_norm": 4.449573516845703, + "learning_rate": 7.906966097445304e-06, + "loss": 5.837135620117188, + "step": 39150 + }, + { + "epoch": 1.2302583025830258, + "grad_norm": 5.611043453216553, + "learning_rate": 7.896462937189375e-06, + "loss": 5.818214111328125, + "step": 39175 + }, + { + "epoch": 1.231043416817147, + "grad_norm": 3.5684187412261963, + "learning_rate": 7.885964572003062e-06, + "loss": 5.822053833007812, + "step": 39200 + }, + { + "epoch": 1.231828531051268, + "grad_norm": 3.816650152206421, + "learning_rate": 7.875471018106693e-06, + "loss": 5.820888671875, + "step": 39225 + }, + { + "epoch": 1.232613645285389, + "grad_norm": 4.082398891448975, + "learning_rate": 7.864982291713157e-06, + "loss": 5.832744750976563, + "step": 39250 + }, + { + "epoch": 1.2333987595195102, + "grad_norm": 3.929248809814453, + "learning_rate": 7.854498409027906e-06, + "loss": 5.827926025390625, + "step": 39275 + }, + { + "epoch": 1.234183873753631, + "grad_norm": 3.8883156776428223, + "learning_rate": 7.84401938624888e-06, + "loss": 5.835979614257813, + "step": 39300 + }, + { + "epoch": 1.2349689879877523, + "grad_norm": 3.2671151161193848, + "learning_rate": 7.833545239566529e-06, + "loss": 5.81873291015625, + "step": 39325 + }, + { + "epoch": 1.2357541022218732, + "grad_norm": 4.993469715118408, + "learning_rate": 7.823075985163758e-06, + "loss": 5.823692016601562, + "step": 39350 + }, + { + "epoch": 1.2365392164559943, + "grad_norm": 4.232470512390137, + "learning_rate": 7.812611639215921e-06, + "loss": 5.83196533203125, + "step": 39375 + }, + { + "epoch": 1.2373243306901154, + "grad_norm": 3.6027092933654785, + "learning_rate": 7.802152217890794e-06, + "loss": 5.829478759765625, + "step": 39400 + }, + { + "epoch": 1.2381094449242365, + "grad_norm": 3.994748592376709, + "learning_rate": 7.791697737348527e-06, + "loss": 5.831668090820313, + "step": 39425 + }, + { + "epoch": 1.2388945591583576, + "grad_norm": 5.290883541107178, + "learning_rate": 7.781248213741643e-06, + "loss": 5.809736328125, + "step": 39450 + }, + { + "epoch": 1.2396796733924786, + "grad_norm": 4.232672214508057, + "learning_rate": 7.77080366321502e-06, + "loss": 5.83046875, + "step": 39475 + }, + { + "epoch": 1.2404647876265997, + "grad_norm": 4.946388244628906, + "learning_rate": 7.760364101905835e-06, + "loss": 5.826399536132812, + "step": 39500 + }, + { + "epoch": 1.2412499018607208, + "grad_norm": 3.647505283355713, + "learning_rate": 7.749929545943566e-06, + "loss": 5.82225341796875, + "step": 39525 + }, + { + "epoch": 1.242035016094842, + "grad_norm": 5.198247909545898, + "learning_rate": 7.739500011449947e-06, + "loss": 5.81729248046875, + "step": 39550 + }, + { + "epoch": 1.2428201303289628, + "grad_norm": 5.563745498657227, + "learning_rate": 7.729075514538968e-06, + "loss": 5.825994873046875, + "step": 39575 + }, + { + "epoch": 1.2436052445630839, + "grad_norm": 3.933283567428589, + "learning_rate": 7.718656071316827e-06, + "loss": 5.812549438476562, + "step": 39600 + }, + { + "epoch": 1.244390358797205, + "grad_norm": 6.521614074707031, + "learning_rate": 7.708241697881922e-06, + "loss": 5.81947265625, + "step": 39625 + }, + { + "epoch": 1.245175473031326, + "grad_norm": 3.431220293045044, + "learning_rate": 7.697832410324796e-06, + "loss": 5.833071899414063, + "step": 39650 + }, + { + "epoch": 1.2459605872654471, + "grad_norm": 4.17876672744751, + "learning_rate": 7.687428224728166e-06, + "loss": 5.8194512939453125, + "step": 39675 + }, + { + "epoch": 1.2467457014995682, + "grad_norm": 5.538009166717529, + "learning_rate": 7.677029157166842e-06, + "loss": 5.812825927734375, + "step": 39700 + }, + { + "epoch": 1.2475308157336893, + "grad_norm": 9.500683784484863, + "learning_rate": 7.666635223707738e-06, + "loss": 5.838218383789062, + "step": 39725 + }, + { + "epoch": 1.2483159299678104, + "grad_norm": 3.5675814151763916, + "learning_rate": 7.656246440409831e-06, + "loss": 5.822977294921875, + "step": 39750 + }, + { + "epoch": 1.2491010442019315, + "grad_norm": 4.076515197753906, + "learning_rate": 7.645862823324141e-06, + "loss": 5.834183959960938, + "step": 39775 + }, + { + "epoch": 1.2498861584360523, + "grad_norm": 2.4608380794525146, + "learning_rate": 7.63548438849371e-06, + "loss": 5.812723388671875, + "step": 39800 + }, + { + "epoch": 1.250011776713512, + "eval_loss": 0.14855222404003143, + "eval_runtime": 5108.268, + "eval_samples_per_second": 4.244, + "eval_steps_per_second": 0.265, + "eval_wer": 3.7997729047589393, + "step": 39804 + }, + { + "epoch": 1.2506712726701736, + "grad_norm": 3.5847396850585938, + "learning_rate": 7.625111151953571e-06, + "loss": 5.8174462890625, + "step": 39825 + }, + { + "epoch": 1.2514563869042945, + "grad_norm": 3.5503106117248535, + "learning_rate": 7.614743129730717e-06, + "loss": 5.82653564453125, + "step": 39850 + }, + { + "epoch": 1.2522415011384156, + "grad_norm": 4.539660453796387, + "learning_rate": 7.604380337844102e-06, + "loss": 5.82011962890625, + "step": 39875 + }, + { + "epoch": 1.2530266153725367, + "grad_norm": 4.292988300323486, + "learning_rate": 7.594022792304587e-06, + "loss": 5.824702758789062, + "step": 39900 + }, + { + "epoch": 1.2538117296066578, + "grad_norm": 4.833174705505371, + "learning_rate": 7.583670509114928e-06, + "loss": 5.8071184158325195, + "step": 39925 + }, + { + "epoch": 1.2545968438407789, + "grad_norm": 4.150866508483887, + "learning_rate": 7.573323504269752e-06, + "loss": 5.833430786132812, + "step": 39950 + }, + { + "epoch": 1.2553819580749, + "grad_norm": 3.2644827365875244, + "learning_rate": 7.562981793755533e-06, + "loss": 5.823411865234375, + "step": 39975 + }, + { + "epoch": 1.256167072309021, + "grad_norm": 5.762414455413818, + "learning_rate": 7.552645393550561e-06, + "loss": 5.8249853515625, + "step": 40000 + }, + { + "epoch": 1.256952186543142, + "grad_norm": 2.706501007080078, + "learning_rate": 7.542314319624928e-06, + "loss": 5.832597045898438, + "step": 40025 + }, + { + "epoch": 1.2577373007772632, + "grad_norm": 6.314685821533203, + "learning_rate": 7.531988587940486e-06, + "loss": 5.828173828125, + "step": 40050 + }, + { + "epoch": 1.258522415011384, + "grad_norm": 4.4891581535339355, + "learning_rate": 7.521668214450846e-06, + "loss": 5.831097412109375, + "step": 40075 + }, + { + "epoch": 1.2593075292455052, + "grad_norm": 4.213101863861084, + "learning_rate": 7.511353215101327e-06, + "loss": 5.836502685546875, + "step": 40100 + }, + { + "epoch": 1.2600926434796262, + "grad_norm": 4.540609836578369, + "learning_rate": 7.501043605828957e-06, + "loss": 5.8298187255859375, + "step": 40125 + }, + { + "epoch": 1.2608777577137473, + "grad_norm": 4.314133167266846, + "learning_rate": 7.490739402562426e-06, + "loss": 5.82147216796875, + "step": 40150 + }, + { + "epoch": 1.2616628719478684, + "grad_norm": 6.821467399597168, + "learning_rate": 7.480440621222078e-06, + "loss": 5.826658325195313, + "step": 40175 + }, + { + "epoch": 1.2624479861819895, + "grad_norm": 4.550699710845947, + "learning_rate": 7.470147277719882e-06, + "loss": 5.8049249267578125, + "step": 40200 + }, + { + "epoch": 1.2632331004161106, + "grad_norm": 6.934886455535889, + "learning_rate": 7.459859387959394e-06, + "loss": 5.8225732421875, + "step": 40225 + }, + { + "epoch": 1.2640182146502317, + "grad_norm": 5.311056613922119, + "learning_rate": 7.449576967835756e-06, + "loss": 5.839495849609375, + "step": 40250 + }, + { + "epoch": 1.2648033288843528, + "grad_norm": 5.215572834014893, + "learning_rate": 7.439300033235654e-06, + "loss": 5.829530029296875, + "step": 40275 + }, + { + "epoch": 1.2655884431184736, + "grad_norm": 4.744568347930908, + "learning_rate": 7.429028600037302e-06, + "loss": 5.821338500976562, + "step": 40300 + }, + { + "epoch": 1.266373557352595, + "grad_norm": 3.3285861015319824, + "learning_rate": 7.418762684110408e-06, + "loss": 5.836581420898438, + "step": 40325 + }, + { + "epoch": 1.2671586715867158, + "grad_norm": 4.334656238555908, + "learning_rate": 7.408502301316154e-06, + "loss": 5.821019287109375, + "step": 40350 + }, + { + "epoch": 1.2679437858208369, + "grad_norm": 4.456195831298828, + "learning_rate": 7.398247467507187e-06, + "loss": 5.8372344970703125, + "step": 40375 + }, + { + "epoch": 1.268728900054958, + "grad_norm": 3.6377251148223877, + "learning_rate": 7.387998198527568e-06, + "loss": 5.820863647460937, + "step": 40400 + }, + { + "epoch": 1.269514014289079, + "grad_norm": 4.30137825012207, + "learning_rate": 7.377754510212765e-06, + "loss": 5.826814575195312, + "step": 40425 + }, + { + "epoch": 1.2702991285232001, + "grad_norm": 5.484102249145508, + "learning_rate": 7.367516418389615e-06, + "loss": 5.827919311523438, + "step": 40450 + }, + { + "epoch": 1.2710842427573212, + "grad_norm": 3.710092067718506, + "learning_rate": 7.357283938876327e-06, + "loss": 5.819649047851563, + "step": 40475 + }, + { + "epoch": 1.2718693569914423, + "grad_norm": 3.9449498653411865, + "learning_rate": 7.347057087482425e-06, + "loss": 5.826300048828125, + "step": 40500 + }, + { + "epoch": 1.2726544712255632, + "grad_norm": 3.7348852157592773, + "learning_rate": 7.336835880008735e-06, + "loss": 5.823142700195312, + "step": 40525 + }, + { + "epoch": 1.2734395854596845, + "grad_norm": 4.6351847648620605, + "learning_rate": 7.3266203322473674e-06, + "loss": 5.84142822265625, + "step": 40550 + }, + { + "epoch": 1.2742246996938054, + "grad_norm": 3.9608492851257324, + "learning_rate": 7.316410459981696e-06, + "loss": 5.82736083984375, + "step": 40575 + }, + { + "epoch": 1.2750098139279264, + "grad_norm": 5.346611499786377, + "learning_rate": 7.3062062789863145e-06, + "loss": 5.832288818359375, + "step": 40600 + }, + { + "epoch": 1.2757949281620475, + "grad_norm": 4.816490173339844, + "learning_rate": 7.296007805027023e-06, + "loss": 5.826905517578125, + "step": 40625 + }, + { + "epoch": 1.2765800423961686, + "grad_norm": 4.341994762420654, + "learning_rate": 7.285815053860812e-06, + "loss": 5.825653686523437, + "step": 40650 + }, + { + "epoch": 1.2773651566302897, + "grad_norm": 5.028903007507324, + "learning_rate": 7.275628041235829e-06, + "loss": 5.8166259765625, + "step": 40675 + }, + { + "epoch": 1.2781502708644108, + "grad_norm": 5.220634937286377, + "learning_rate": 7.26544678289135e-06, + "loss": 5.80900634765625, + "step": 40700 + }, + { + "epoch": 1.2789353850985319, + "grad_norm": 3.384140968322754, + "learning_rate": 7.2552712945577645e-06, + "loss": 5.823921508789063, + "step": 40725 + }, + { + "epoch": 1.279720499332653, + "grad_norm": 3.378743886947632, + "learning_rate": 7.245101591956541e-06, + "loss": 5.800816040039063, + "step": 40750 + }, + { + "epoch": 1.280505613566774, + "grad_norm": 6.762293815612793, + "learning_rate": 7.234937690800221e-06, + "loss": 5.819578247070313, + "step": 40775 + }, + { + "epoch": 1.281290727800895, + "grad_norm": 3.8215651512145996, + "learning_rate": 7.2247796067923694e-06, + "loss": 5.80933837890625, + "step": 40800 + }, + { + "epoch": 1.2820758420350162, + "grad_norm": 5.185006141662598, + "learning_rate": 7.214627355627571e-06, + "loss": 5.843076782226563, + "step": 40825 + }, + { + "epoch": 1.282860956269137, + "grad_norm": 4.325466632843018, + "learning_rate": 7.2044809529913925e-06, + "loss": 5.842747192382813, + "step": 40850 + }, + { + "epoch": 1.2836460705032582, + "grad_norm": 3.9615390300750732, + "learning_rate": 7.194340414560381e-06, + "loss": 5.804140625, + "step": 40875 + }, + { + "epoch": 1.2844311847373793, + "grad_norm": 3.4685866832733154, + "learning_rate": 7.184205756002003e-06, + "loss": 5.8174676513671875, + "step": 40900 + }, + { + "epoch": 1.2852162989715004, + "grad_norm": 5.193195819854736, + "learning_rate": 7.174076992974645e-06, + "loss": 5.827684326171875, + "step": 40925 + }, + { + "epoch": 1.2860014132056214, + "grad_norm": 5.100523471832275, + "learning_rate": 7.163954141127592e-06, + "loss": 5.82667236328125, + "step": 40950 + }, + { + "epoch": 1.2867865274397425, + "grad_norm": 3.5903098583221436, + "learning_rate": 7.153837216100994e-06, + "loss": 5.832816772460937, + "step": 40975 + }, + { + "epoch": 1.2875716416738636, + "grad_norm": 6.0330376625061035, + "learning_rate": 7.143726233525836e-06, + "loss": 5.827779541015625, + "step": 41000 + }, + { + "epoch": 1.2883567559079845, + "grad_norm": 4.527500152587891, + "learning_rate": 7.133621209023935e-06, + "loss": 5.819263305664062, + "step": 41025 + }, + { + "epoch": 1.2891418701421058, + "grad_norm": 3.273733615875244, + "learning_rate": 7.123522158207887e-06, + "loss": 5.824338989257813, + "step": 41050 + }, + { + "epoch": 1.2899269843762267, + "grad_norm": 4.512317180633545, + "learning_rate": 7.113429096681077e-06, + "loss": 5.809480590820312, + "step": 41075 + }, + { + "epoch": 1.2907120986103477, + "grad_norm": 4.38275146484375, + "learning_rate": 7.1033420400376196e-06, + "loss": 5.824080810546875, + "step": 41100 + }, + { + "epoch": 1.2914972128444688, + "grad_norm": 5.97402811050415, + "learning_rate": 7.093261003862357e-06, + "loss": 5.823933715820313, + "step": 41125 + }, + { + "epoch": 1.29228232707859, + "grad_norm": 5.016282081604004, + "learning_rate": 7.0831860037308306e-06, + "loss": 5.826375122070313, + "step": 41150 + }, + { + "epoch": 1.293067441312711, + "grad_norm": 3.7956621646881104, + "learning_rate": 7.073117055209262e-06, + "loss": 5.831461181640625, + "step": 41175 + }, + { + "epoch": 1.293852555546832, + "grad_norm": 5.361743450164795, + "learning_rate": 7.063054173854513e-06, + "loss": 5.82039794921875, + "step": 41200 + }, + { + "epoch": 1.2946376697809532, + "grad_norm": 4.026911735534668, + "learning_rate": 7.052997375214077e-06, + "loss": 5.825764770507813, + "step": 41225 + }, + { + "epoch": 1.2954227840150743, + "grad_norm": 3.030885696411133, + "learning_rate": 7.042946674826036e-06, + "loss": 5.8297662353515625, + "step": 41250 + }, + { + "epoch": 1.2962078982491954, + "grad_norm": 4.973598003387451, + "learning_rate": 7.03290208821908e-06, + "loss": 5.81445068359375, + "step": 41275 + }, + { + "epoch": 1.2969930124833162, + "grad_norm": 4.953793048858643, + "learning_rate": 7.022863630912422e-06, + "loss": 5.844222412109375, + "step": 41300 + }, + { + "epoch": 1.2977781267174373, + "grad_norm": 4.0591607093811035, + "learning_rate": 7.012831318415827e-06, + "loss": 5.828668823242188, + "step": 41325 + }, + { + "epoch": 1.2985632409515584, + "grad_norm": 3.4998648166656494, + "learning_rate": 7.002805166229547e-06, + "loss": 5.821820068359375, + "step": 41350 + }, + { + "epoch": 1.2993483551856795, + "grad_norm": 4.473606586456299, + "learning_rate": 6.99278518984433e-06, + "loss": 5.830493774414062, + "step": 41375 + }, + { + "epoch": 1.3001334694198006, + "grad_norm": 4.404461860656738, + "learning_rate": 6.982771404741383e-06, + "loss": 5.820552368164062, + "step": 41400 + }, + { + "epoch": 1.3009185836539217, + "grad_norm": 3.800593852996826, + "learning_rate": 6.972763826392333e-06, + "loss": 5.829497680664063, + "step": 41425 + }, + { + "epoch": 1.3017036978880427, + "grad_norm": 4.422252178192139, + "learning_rate": 6.9627624702592345e-06, + "loss": 5.81260009765625, + "step": 41450 + }, + { + "epoch": 1.3024888121221638, + "grad_norm": 5.908450126647949, + "learning_rate": 6.952767351794518e-06, + "loss": 5.826061401367188, + "step": 41475 + }, + { + "epoch": 1.303273926356285, + "grad_norm": 2.8411171436309814, + "learning_rate": 6.9427784864409754e-06, + "loss": 5.824048461914063, + "step": 41500 + }, + { + "epoch": 1.3040590405904058, + "grad_norm": 3.1454696655273438, + "learning_rate": 6.932795889631749e-06, + "loss": 5.82474853515625, + "step": 41525 + }, + { + "epoch": 1.304844154824527, + "grad_norm": 2.9233968257904053, + "learning_rate": 6.922819576790279e-06, + "loss": 5.814248657226562, + "step": 41550 + }, + { + "epoch": 1.305629269058648, + "grad_norm": 4.560819625854492, + "learning_rate": 6.912849563330313e-06, + "loss": 5.837764282226562, + "step": 41575 + }, + { + "epoch": 1.306414383292769, + "grad_norm": 4.884252548217773, + "learning_rate": 6.902885864655858e-06, + "loss": 5.814729614257812, + "step": 41600 + }, + { + "epoch": 1.3071994975268901, + "grad_norm": 3.7128405570983887, + "learning_rate": 6.892928496161159e-06, + "loss": 5.823046264648437, + "step": 41625 + }, + { + "epoch": 1.3079846117610112, + "grad_norm": 3.770001173019409, + "learning_rate": 6.882977473230691e-06, + "loss": 5.82371337890625, + "step": 41650 + }, + { + "epoch": 1.3087697259951323, + "grad_norm": 5.529378414154053, + "learning_rate": 6.873032811239124e-06, + "loss": 5.82896728515625, + "step": 41675 + }, + { + "epoch": 1.3095548402292534, + "grad_norm": 4.418456077575684, + "learning_rate": 6.8630945255512905e-06, + "loss": 5.796940307617188, + "step": 41700 + }, + { + "epoch": 1.3103399544633745, + "grad_norm": 4.187838554382324, + "learning_rate": 6.8531626315221835e-06, + "loss": 5.8142938232421875, + "step": 41725 + }, + { + "epoch": 1.3111250686974956, + "grad_norm": 3.998932123184204, + "learning_rate": 6.8432371444969085e-06, + "loss": 5.826372680664062, + "step": 41750 + }, + { + "epoch": 1.3119101829316167, + "grad_norm": 3.334317684173584, + "learning_rate": 6.83331807981068e-06, + "loss": 5.810306396484375, + "step": 41775 + }, + { + "epoch": 1.3126952971657375, + "grad_norm": 3.360288619995117, + "learning_rate": 6.823405452788796e-06, + "loss": 5.807302856445313, + "step": 41800 + }, + { + "epoch": 1.3134804113998586, + "grad_norm": 3.519850492477417, + "learning_rate": 6.813499278746587e-06, + "loss": 5.823236694335938, + "step": 41825 + }, + { + "epoch": 1.3142655256339797, + "grad_norm": 4.309455871582031, + "learning_rate": 6.803599572989435e-06, + "loss": 5.827537841796875, + "step": 41850 + }, + { + "epoch": 1.3150506398681008, + "grad_norm": 5.174105644226074, + "learning_rate": 6.793706350812722e-06, + "loss": 5.826810913085938, + "step": 41875 + }, + { + "epoch": 1.3158357541022219, + "grad_norm": 4.137282371520996, + "learning_rate": 6.783819627501802e-06, + "loss": 5.800485229492187, + "step": 41900 + }, + { + "epoch": 1.316620868336343, + "grad_norm": 6.2611894607543945, + "learning_rate": 6.773939418332005e-06, + "loss": 5.812197875976563, + "step": 41925 + }, + { + "epoch": 1.317405982570464, + "grad_norm": 2.984330177307129, + "learning_rate": 6.764065738568583e-06, + "loss": 5.813395385742187, + "step": 41950 + }, + { + "epoch": 1.3181910968045851, + "grad_norm": 4.957423210144043, + "learning_rate": 6.754198603466701e-06, + "loss": 5.845759887695312, + "step": 41975 + }, + { + "epoch": 1.3189762110387062, + "grad_norm": 3.8578410148620605, + "learning_rate": 6.744338028271429e-06, + "loss": 5.83157958984375, + "step": 42000 + }, + { + "epoch": 1.319761325272827, + "grad_norm": 4.698902606964111, + "learning_rate": 6.734484028217676e-06, + "loss": 5.818756103515625, + "step": 42025 + }, + { + "epoch": 1.3205464395069484, + "grad_norm": 3.999647855758667, + "learning_rate": 6.724636618530209e-06, + "loss": 5.81864013671875, + "step": 42050 + }, + { + "epoch": 1.3213315537410693, + "grad_norm": 2.8529045581817627, + "learning_rate": 6.7147958144236155e-06, + "loss": 5.828797607421875, + "step": 42075 + }, + { + "epoch": 1.3221166679751903, + "grad_norm": 3.3405768871307373, + "learning_rate": 6.704961631102258e-06, + "loss": 5.826842041015625, + "step": 42100 + }, + { + "epoch": 1.3229017822093114, + "grad_norm": 4.919105052947998, + "learning_rate": 6.695134083760294e-06, + "loss": 5.8137445068359375, + "step": 42125 + }, + { + "epoch": 1.3236868964434325, + "grad_norm": 5.17466402053833, + "learning_rate": 6.685313187581602e-06, + "loss": 5.818268432617187, + "step": 42150 + }, + { + "epoch": 1.3244720106775536, + "grad_norm": 4.288014888763428, + "learning_rate": 6.675498957739816e-06, + "loss": 5.818073120117187, + "step": 42175 + }, + { + "epoch": 1.3252571249116747, + "grad_norm": 3.0581088066101074, + "learning_rate": 6.66569140939824e-06, + "loss": 5.822312622070313, + "step": 42200 + }, + { + "epoch": 1.3260422391457958, + "grad_norm": 4.412501335144043, + "learning_rate": 6.655890557709871e-06, + "loss": 5.822937622070312, + "step": 42225 + }, + { + "epoch": 1.3268273533799166, + "grad_norm": 3.894340991973877, + "learning_rate": 6.6460964178173536e-06, + "loss": 5.811527709960938, + "step": 42250 + }, + { + "epoch": 1.327612467614038, + "grad_norm": 4.584700107574463, + "learning_rate": 6.63630900485297e-06, + "loss": 5.816731567382813, + "step": 42275 + }, + { + "epoch": 1.3283975818481588, + "grad_norm": 4.942855358123779, + "learning_rate": 6.626528333938603e-06, + "loss": 5.821326904296875, + "step": 42300 + }, + { + "epoch": 1.32918269608228, + "grad_norm": 4.092140197753906, + "learning_rate": 6.616754420185718e-06, + "loss": 5.8320379638671875, + "step": 42325 + }, + { + "epoch": 1.329967810316401, + "grad_norm": 5.201463222503662, + "learning_rate": 6.606987278695335e-06, + "loss": 5.833919677734375, + "step": 42350 + }, + { + "epoch": 1.330752924550522, + "grad_norm": 3.5328218936920166, + "learning_rate": 6.597226924558031e-06, + "loss": 5.808099975585938, + "step": 42375 + }, + { + "epoch": 1.3315380387846432, + "grad_norm": 5.559754848480225, + "learning_rate": 6.587473372853874e-06, + "loss": 5.8243560791015625, + "step": 42400 + }, + { + "epoch": 1.3323231530187643, + "grad_norm": 4.31860876083374, + "learning_rate": 6.5777266386524364e-06, + "loss": 5.827918701171875, + "step": 42425 + }, + { + "epoch": 1.3331082672528853, + "grad_norm": 7.18262243270874, + "learning_rate": 6.567986737012744e-06, + "loss": 5.817261352539062, + "step": 42450 + }, + { + "epoch": 1.3338933814870064, + "grad_norm": 3.6293444633483887, + "learning_rate": 6.558253682983279e-06, + "loss": 5.820407322474888, + "step": 42475 + }, + { + "epoch": 1.3346784957211275, + "grad_norm": 3.5676779747009277, + "learning_rate": 6.548527491601941e-06, + "loss": 5.826690673828125, + "step": 42500 + }, + { + "epoch": 1.3354636099552484, + "grad_norm": 3.89274263381958, + "learning_rate": 6.538808177896019e-06, + "loss": 5.810488891601563, + "step": 42525 + }, + { + "epoch": 1.3362487241893697, + "grad_norm": 4.014410495758057, + "learning_rate": 6.529095756882183e-06, + "loss": 5.81024169921875, + "step": 42550 + }, + { + "epoch": 1.3370338384234906, + "grad_norm": 3.332129955291748, + "learning_rate": 6.519390243566457e-06, + "loss": 5.8187933349609375, + "step": 42575 + }, + { + "epoch": 1.3378189526576116, + "grad_norm": 3.7073850631713867, + "learning_rate": 6.509691652944178e-06, + "loss": 5.838155517578125, + "step": 42600 + }, + { + "epoch": 1.3386040668917327, + "grad_norm": 3.2493534088134766, + "learning_rate": 6.500000000000003e-06, + "loss": 5.81275390625, + "step": 42625 + }, + { + "epoch": 1.3393891811258538, + "grad_norm": 5.187699317932129, + "learning_rate": 6.490315299707857e-06, + "loss": 5.824697265625, + "step": 42650 + }, + { + "epoch": 1.340174295359975, + "grad_norm": 4.123056411743164, + "learning_rate": 6.480637567030934e-06, + "loss": 5.826127319335938, + "step": 42675 + }, + { + "epoch": 1.340959409594096, + "grad_norm": 7.6056227684021, + "learning_rate": 6.470966816921658e-06, + "loss": 5.829129028320312, + "step": 42700 + }, + { + "epoch": 1.341744523828217, + "grad_norm": 4.605141639709473, + "learning_rate": 6.461303064321661e-06, + "loss": 5.815722045898437, + "step": 42725 + }, + { + "epoch": 1.342529638062338, + "grad_norm": 4.434563159942627, + "learning_rate": 6.45164632416177e-06, + "loss": 5.810322875976563, + "step": 42750 + }, + { + "epoch": 1.3433147522964592, + "grad_norm": 7.5077643394470215, + "learning_rate": 6.441996611361978e-06, + "loss": 5.81827392578125, + "step": 42775 + }, + { + "epoch": 1.3440998665305801, + "grad_norm": 3.5003912448883057, + "learning_rate": 6.432353940831408e-06, + "loss": 5.814085083007813, + "step": 42800 + }, + { + "epoch": 1.3448849807647012, + "grad_norm": 4.550718784332275, + "learning_rate": 6.4227183274683224e-06, + "loss": 5.823692626953125, + "step": 42825 + }, + { + "epoch": 1.3456700949988223, + "grad_norm": 4.338807106018066, + "learning_rate": 6.413089786160063e-06, + "loss": 5.834568481445313, + "step": 42850 + }, + { + "epoch": 1.3464552092329434, + "grad_norm": 4.641574859619141, + "learning_rate": 6.403468331783049e-06, + "loss": 5.819386596679688, + "step": 42875 + }, + { + "epoch": 1.3472403234670645, + "grad_norm": 5.582952499389648, + "learning_rate": 6.393853979202762e-06, + "loss": 5.810989990234375, + "step": 42900 + }, + { + "epoch": 1.3480254377011855, + "grad_norm": 3.573458671569824, + "learning_rate": 6.384246743273691e-06, + "loss": 5.806709594726563, + "step": 42925 + }, + { + "epoch": 1.3488105519353066, + "grad_norm": 4.333592891693115, + "learning_rate": 6.374646638839346e-06, + "loss": 5.816863403320313, + "step": 42950 + }, + { + "epoch": 1.3495956661694277, + "grad_norm": 4.073453426361084, + "learning_rate": 6.365053680732215e-06, + "loss": 5.816244506835938, + "step": 42975 + }, + { + "epoch": 1.3503807804035488, + "grad_norm": 3.7137627601623535, + "learning_rate": 6.355467883773735e-06, + "loss": 5.804434814453125, + "step": 43000 + }, + { + "epoch": 1.3511658946376697, + "grad_norm": 4.826092720031738, + "learning_rate": 6.345889262774294e-06, + "loss": 5.822174682617187, + "step": 43025 + }, + { + "epoch": 1.351951008871791, + "grad_norm": 3.633890151977539, + "learning_rate": 6.336317832533177e-06, + "loss": 5.815968627929688, + "step": 43050 + }, + { + "epoch": 1.3527361231059118, + "grad_norm": 4.194397449493408, + "learning_rate": 6.326753607838572e-06, + "loss": 5.804359741210938, + "step": 43075 + }, + { + "epoch": 1.353521237340033, + "grad_norm": 4.220597743988037, + "learning_rate": 6.3171966034675324e-06, + "loss": 5.803433227539062, + "step": 43100 + }, + { + "epoch": 1.354306351574154, + "grad_norm": 5.55937385559082, + "learning_rate": 6.307646834185943e-06, + "loss": 5.816494750976562, + "step": 43125 + }, + { + "epoch": 1.355091465808275, + "grad_norm": 5.396577835083008, + "learning_rate": 6.298104314748527e-06, + "loss": 5.798345336914062, + "step": 43150 + }, + { + "epoch": 1.3558765800423962, + "grad_norm": 4.004359722137451, + "learning_rate": 6.288569059898804e-06, + "loss": 5.822540893554687, + "step": 43175 + }, + { + "epoch": 1.3566616942765173, + "grad_norm": 7.067991256713867, + "learning_rate": 6.2790410843690555e-06, + "loss": 5.829288330078125, + "step": 43200 + }, + { + "epoch": 1.3574468085106384, + "grad_norm": 4.157956600189209, + "learning_rate": 6.2695204028803315e-06, + "loss": 5.809461669921875, + "step": 43225 + }, + { + "epoch": 1.3582319227447592, + "grad_norm": 3.974282741546631, + "learning_rate": 6.260007030142397e-06, + "loss": 5.83349609375, + "step": 43250 + }, + { + "epoch": 1.3590170369788805, + "grad_norm": 3.8068041801452637, + "learning_rate": 6.2505009808537485e-06, + "loss": 5.81498291015625, + "step": 43275 + }, + { + "epoch": 1.3598021512130014, + "grad_norm": 3.107210636138916, + "learning_rate": 6.241002269701547e-06, + "loss": 5.830946655273437, + "step": 43300 + }, + { + "epoch": 1.3605872654471225, + "grad_norm": 3.7908620834350586, + "learning_rate": 6.231510911361618e-06, + "loss": 5.81794189453125, + "step": 43325 + }, + { + "epoch": 1.3613723796812436, + "grad_norm": 4.354649543762207, + "learning_rate": 6.222026920498434e-06, + "loss": 5.80828857421875, + "step": 43350 + }, + { + "epoch": 1.3621574939153647, + "grad_norm": 3.417980909347534, + "learning_rate": 6.212550311765082e-06, + "loss": 5.821104736328125, + "step": 43375 + }, + { + "epoch": 1.3629426081494858, + "grad_norm": 4.75178337097168, + "learning_rate": 6.203081099803247e-06, + "loss": 5.818112182617187, + "step": 43400 + }, + { + "epoch": 1.3637277223836068, + "grad_norm": 5.436915397644043, + "learning_rate": 6.193619299243176e-06, + "loss": 5.809725341796875, + "step": 43425 + }, + { + "epoch": 1.364512836617728, + "grad_norm": 4.785301208496094, + "learning_rate": 6.1841649247036636e-06, + "loss": 5.816985473632813, + "step": 43450 + }, + { + "epoch": 1.365297950851849, + "grad_norm": 3.4740970134735107, + "learning_rate": 6.174717990792053e-06, + "loss": 5.821357421875, + "step": 43475 + }, + { + "epoch": 1.36608306508597, + "grad_norm": 5.624035358428955, + "learning_rate": 6.165278512104166e-06, + "loss": 5.81565673828125, + "step": 43500 + }, + { + "epoch": 1.366868179320091, + "grad_norm": 4.148986339569092, + "learning_rate": 6.155846503224321e-06, + "loss": 5.82021484375, + "step": 43525 + }, + { + "epoch": 1.367653293554212, + "grad_norm": 4.207647800445557, + "learning_rate": 6.1464219787252855e-06, + "loss": 5.800231323242188, + "step": 43550 + }, + { + "epoch": 1.3684384077883331, + "grad_norm": 3.8121895790100098, + "learning_rate": 6.137004953168271e-06, + "loss": 5.819781494140625, + "step": 43575 + }, + { + "epoch": 1.3692235220224542, + "grad_norm": 4.465154647827148, + "learning_rate": 6.127595441102905e-06, + "loss": 5.799212646484375, + "step": 43600 + }, + { + "epoch": 1.3700086362565753, + "grad_norm": 2.9391591548919678, + "learning_rate": 6.1181934570671905e-06, + "loss": 5.805802001953125, + "step": 43625 + }, + { + "epoch": 1.3707937504906964, + "grad_norm": 5.08090877532959, + "learning_rate": 6.108799015587518e-06, + "loss": 5.799700317382812, + "step": 43650 + }, + { + "epoch": 1.3715788647248175, + "grad_norm": 3.939491033554077, + "learning_rate": 6.099412131178619e-06, + "loss": 5.819354858398437, + "step": 43675 + }, + { + "epoch": 1.3723639789589386, + "grad_norm": 4.693300247192383, + "learning_rate": 6.090032818343541e-06, + "loss": 5.798568115234375, + "step": 43700 + }, + { + "epoch": 1.3731490931930597, + "grad_norm": 4.306169033050537, + "learning_rate": 6.080661091573645e-06, + "loss": 5.796414794921875, + "step": 43725 + }, + { + "epoch": 1.3739342074271805, + "grad_norm": 4.354258060455322, + "learning_rate": 6.071296965348559e-06, + "loss": 5.806763305664062, + "step": 43750 + }, + { + "epoch": 1.3747193216613018, + "grad_norm": 3.833468437194824, + "learning_rate": 6.061940454136177e-06, + "loss": 5.80314453125, + "step": 43775 + }, + { + "epoch": 1.3755044358954227, + "grad_norm": 3.0796730518341064, + "learning_rate": 6.05259157239263e-06, + "loss": 5.815092163085938, + "step": 43800 + }, + { + "epoch": 1.3762895501295438, + "grad_norm": 3.867314338684082, + "learning_rate": 6.043250334562249e-06, + "loss": 5.803670654296875, + "step": 43825 + }, + { + "epoch": 1.3770746643636649, + "grad_norm": 2.946002960205078, + "learning_rate": 6.033916755077564e-06, + "loss": 5.807271118164063, + "step": 43850 + }, + { + "epoch": 1.377859778597786, + "grad_norm": 3.8855676651000977, + "learning_rate": 6.0245908483592755e-06, + "loss": 5.818943481445313, + "step": 43875 + }, + { + "epoch": 1.378644892831907, + "grad_norm": 3.209009885787964, + "learning_rate": 6.015272628816215e-06, + "loss": 5.803363037109375, + "step": 43900 + }, + { + "epoch": 1.3794300070660281, + "grad_norm": 4.465564727783203, + "learning_rate": 6.005962110845359e-06, + "loss": 5.828009033203125, + "step": 43925 + }, + { + "epoch": 1.3802151213001492, + "grad_norm": 4.764284610748291, + "learning_rate": 5.99665930883176e-06, + "loss": 5.81720703125, + "step": 43950 + }, + { + "epoch": 1.3810002355342703, + "grad_norm": 3.6677627563476562, + "learning_rate": 5.987364237148567e-06, + "loss": 5.812550659179688, + "step": 43975 + }, + { + "epoch": 1.3817853497683914, + "grad_norm": 4.403491497039795, + "learning_rate": 5.978076910156984e-06, + "loss": 5.80472900390625, + "step": 44000 + }, + { + "epoch": 1.3825704640025123, + "grad_norm": 3.4093682765960693, + "learning_rate": 5.968797342206235e-06, + "loss": 5.807542114257813, + "step": 44025 + }, + { + "epoch": 1.3833555782366334, + "grad_norm": 2.978846788406372, + "learning_rate": 5.959525547633574e-06, + "loss": 5.8168115234375, + "step": 44050 + }, + { + "epoch": 1.3841406924707544, + "grad_norm": 2.8826873302459717, + "learning_rate": 5.950261540764235e-06, + "loss": 5.804735107421875, + "step": 44075 + }, + { + "epoch": 1.3849258067048755, + "grad_norm": 3.556987762451172, + "learning_rate": 5.941005335911419e-06, + "loss": 5.817007446289063, + "step": 44100 + }, + { + "epoch": 1.3857109209389966, + "grad_norm": 3.985995292663574, + "learning_rate": 5.93175694737628e-06, + "loss": 5.813240356445313, + "step": 44125 + }, + { + "epoch": 1.3864960351731177, + "grad_norm": 5.052744388580322, + "learning_rate": 5.922516389447885e-06, + "loss": 5.811167602539062, + "step": 44150 + }, + { + "epoch": 1.3872811494072388, + "grad_norm": 4.0221967697143555, + "learning_rate": 5.913283676403212e-06, + "loss": 5.807158203125, + "step": 44175 + }, + { + "epoch": 1.3880662636413599, + "grad_norm": 3.835662841796875, + "learning_rate": 5.904058822507116e-06, + "loss": 5.801024169921875, + "step": 44200 + }, + { + "epoch": 1.388851377875481, + "grad_norm": 3.531071901321411, + "learning_rate": 5.894841842012302e-06, + "loss": 5.8204248046875, + "step": 44225 + }, + { + "epoch": 1.3896364921096018, + "grad_norm": 3.057039260864258, + "learning_rate": 5.88563274915932e-06, + "loss": 5.7984130859375, + "step": 44250 + }, + { + "epoch": 1.3904216063437231, + "grad_norm": 4.762048721313477, + "learning_rate": 5.876431558176534e-06, + "loss": 5.823186645507812, + "step": 44275 + }, + { + "epoch": 1.391206720577844, + "grad_norm": 4.485466957092285, + "learning_rate": 5.867238283280085e-06, + "loss": 5.829664306640625, + "step": 44300 + }, + { + "epoch": 1.391991834811965, + "grad_norm": 8.325467109680176, + "learning_rate": 5.8580529386739045e-06, + "loss": 5.82308837890625, + "step": 44325 + }, + { + "epoch": 1.3927769490460862, + "grad_norm": 3.861259937286377, + "learning_rate": 5.8488755385496465e-06, + "loss": 5.809869995117188, + "step": 44350 + }, + { + "epoch": 1.3935620632802073, + "grad_norm": 3.178537130355835, + "learning_rate": 5.839706097086719e-06, + "loss": 5.798770751953125, + "step": 44375 + }, + { + "epoch": 1.3943471775143284, + "grad_norm": 3.1674389839172363, + "learning_rate": 5.830544628452213e-06, + "loss": 5.79826904296875, + "step": 44400 + }, + { + "epoch": 1.3951322917484494, + "grad_norm": 4.407901287078857, + "learning_rate": 5.821391146800901e-06, + "loss": 5.813531494140625, + "step": 44425 + }, + { + "epoch": 1.3959174059825705, + "grad_norm": 4.875741958618164, + "learning_rate": 5.812245666275227e-06, + "loss": 5.803872680664062, + "step": 44450 + }, + { + "epoch": 1.3967025202166914, + "grad_norm": 3.8608405590057373, + "learning_rate": 5.8031082010052685e-06, + "loss": 5.81066162109375, + "step": 44475 + }, + { + "epoch": 1.3974876344508127, + "grad_norm": 3.9360785484313965, + "learning_rate": 5.793978765108719e-06, + "loss": 5.819261474609375, + "step": 44500 + }, + { + "epoch": 1.3982727486849336, + "grad_norm": 4.581697940826416, + "learning_rate": 5.7848573726908615e-06, + "loss": 5.807943115234375, + "step": 44525 + }, + { + "epoch": 1.3990578629190547, + "grad_norm": 3.0425655841827393, + "learning_rate": 5.775744037844552e-06, + "loss": 5.801187744140625, + "step": 44550 + }, + { + "epoch": 1.3998429771531757, + "grad_norm": 4.58329963684082, + "learning_rate": 5.766638774650211e-06, + "loss": 5.8167572021484375, + "step": 44575 + }, + { + "epoch": 1.4006280913872968, + "grad_norm": 4.512588977813721, + "learning_rate": 5.757541597175769e-06, + "loss": 5.815403442382813, + "step": 44600 + }, + { + "epoch": 1.401413205621418, + "grad_norm": 2.694904327392578, + "learning_rate": 5.748452519476681e-06, + "loss": 5.800839233398437, + "step": 44625 + }, + { + "epoch": 1.402198319855539, + "grad_norm": 6.163465976715088, + "learning_rate": 5.73937155559587e-06, + "loss": 5.815359497070313, + "step": 44650 + }, + { + "epoch": 1.40298343408966, + "grad_norm": 3.6442959308624268, + "learning_rate": 5.730298719563734e-06, + "loss": 5.807352294921875, + "step": 44675 + }, + { + "epoch": 1.4037685483237812, + "grad_norm": 3.763200521469116, + "learning_rate": 5.721234025398117e-06, + "loss": 5.798366088867187, + "step": 44700 + }, + { + "epoch": 1.4045536625579023, + "grad_norm": 3.445002317428589, + "learning_rate": 5.712177487104271e-06, + "loss": 5.823878173828125, + "step": 44725 + }, + { + "epoch": 1.4053387767920231, + "grad_norm": 4.994740009307861, + "learning_rate": 5.703129118674857e-06, + "loss": 5.796717529296875, + "step": 44750 + }, + { + "epoch": 1.4061238910261444, + "grad_norm": 4.494963645935059, + "learning_rate": 5.694088934089914e-06, + "loss": 5.8135498046875, + "step": 44775 + }, + { + "epoch": 1.4069090052602653, + "grad_norm": 4.248727321624756, + "learning_rate": 5.685056947316822e-06, + "loss": 5.806055297851563, + "step": 44800 + }, + { + "epoch": 1.4076941194943864, + "grad_norm": 5.061432838439941, + "learning_rate": 5.6760331723103165e-06, + "loss": 5.821351928710937, + "step": 44825 + }, + { + "epoch": 1.4084792337285075, + "grad_norm": 3.9054107666015625, + "learning_rate": 5.667017623012427e-06, + "loss": 5.8182177734375, + "step": 44850 + }, + { + "epoch": 1.4092643479626286, + "grad_norm": 2.8923909664154053, + "learning_rate": 5.658010313352485e-06, + "loss": 5.800775756835938, + "step": 44875 + }, + { + "epoch": 1.4100494621967496, + "grad_norm": 3.783634662628174, + "learning_rate": 5.64901125724709e-06, + "loss": 5.811065673828125, + "step": 44900 + }, + { + "epoch": 1.4108345764308707, + "grad_norm": 4.71266508102417, + "learning_rate": 5.640020468600085e-06, + "loss": 5.8030328369140625, + "step": 44925 + }, + { + "epoch": 1.4116196906649918, + "grad_norm": 3.371321439743042, + "learning_rate": 5.6310379613025414e-06, + "loss": 5.793353271484375, + "step": 44950 + }, + { + "epoch": 1.4124048048991127, + "grad_norm": 4.706638336181641, + "learning_rate": 5.62206374923274e-06, + "loss": 5.8124456787109375, + "step": 44975 + }, + { + "epoch": 1.413189919133234, + "grad_norm": 4.701776504516602, + "learning_rate": 5.613097846256137e-06, + "loss": 5.805025024414062, + "step": 45000 + }, + { + "epoch": 1.4139750333673549, + "grad_norm": 3.5699567794799805, + "learning_rate": 5.604140266225363e-06, + "loss": 5.8093701171875, + "step": 45025 + }, + { + "epoch": 1.414760147601476, + "grad_norm": 4.312647819519043, + "learning_rate": 5.595191022980172e-06, + "loss": 5.817752075195313, + "step": 45050 + }, + { + "epoch": 1.415545261835597, + "grad_norm": 5.65584135055542, + "learning_rate": 5.586250130347452e-06, + "loss": 5.8382568359375, + "step": 45075 + }, + { + "epoch": 1.4163303760697181, + "grad_norm": 5.0007548332214355, + "learning_rate": 5.577317602141188e-06, + "loss": 5.826161499023438, + "step": 45100 + }, + { + "epoch": 1.4171154903038392, + "grad_norm": 3.273911714553833, + "learning_rate": 5.568393452162431e-06, + "loss": 5.820038452148437, + "step": 45125 + }, + { + "epoch": 1.4179006045379603, + "grad_norm": 5.498136520385742, + "learning_rate": 5.559477694199296e-06, + "loss": 5.810863647460938, + "step": 45150 + }, + { + "epoch": 1.4186857187720814, + "grad_norm": 4.320960521697998, + "learning_rate": 5.5505703420269365e-06, + "loss": 5.804503784179688, + "step": 45175 + }, + { + "epoch": 1.4194708330062025, + "grad_norm": 4.2402825355529785, + "learning_rate": 5.541671409407504e-06, + "loss": 5.811564127604167, + "step": 45200 + }, + { + "epoch": 1.4202559472403236, + "grad_norm": 3.479201555252075, + "learning_rate": 5.532780910090157e-06, + "loss": 5.8166876220703125, + "step": 45225 + }, + { + "epoch": 1.4210410614744444, + "grad_norm": 3.3140487670898438, + "learning_rate": 5.52389885781101e-06, + "loss": 5.801007080078125, + "step": 45250 + }, + { + "epoch": 1.4218261757085657, + "grad_norm": 4.312045097351074, + "learning_rate": 5.515025266293137e-06, + "loss": 5.809956665039063, + "step": 45275 + }, + { + "epoch": 1.4226112899426866, + "grad_norm": 3.9315264225006104, + "learning_rate": 5.5061601492465426e-06, + "loss": 5.812565307617188, + "step": 45300 + }, + { + "epoch": 1.4233964041768077, + "grad_norm": 4.782158851623535, + "learning_rate": 5.497303520368122e-06, + "loss": 5.805308837890625, + "step": 45325 + }, + { + "epoch": 1.4241815184109288, + "grad_norm": 3.3097758293151855, + "learning_rate": 5.488455393341673e-06, + "loss": 5.81678466796875, + "step": 45350 + }, + { + "epoch": 1.4249666326450499, + "grad_norm": 3.606321334838867, + "learning_rate": 5.479615781837851e-06, + "loss": 5.801299438476563, + "step": 45375 + }, + { + "epoch": 1.425751746879171, + "grad_norm": 5.053475856781006, + "learning_rate": 5.470784699514149e-06, + "loss": 5.805849609375, + "step": 45400 + }, + { + "epoch": 1.426536861113292, + "grad_norm": 5.017884731292725, + "learning_rate": 5.461962160014893e-06, + "loss": 5.814133911132813, + "step": 45425 + }, + { + "epoch": 1.4273219753474131, + "grad_norm": 3.636611223220825, + "learning_rate": 5.4531481769711984e-06, + "loss": 5.8207177734375, + "step": 45450 + }, + { + "epoch": 1.428107089581534, + "grad_norm": 3.368312120437622, + "learning_rate": 5.444342764000976e-06, + "loss": 5.802736206054687, + "step": 45475 + }, + { + "epoch": 1.4288922038156553, + "grad_norm": 5.420502185821533, + "learning_rate": 5.435545934708881e-06, + "loss": 5.808641357421875, + "step": 45500 + }, + { + "epoch": 1.4296773180497762, + "grad_norm": 5.192573070526123, + "learning_rate": 5.42675770268631e-06, + "loss": 5.803773193359375, + "step": 45525 + }, + { + "epoch": 1.4304624322838972, + "grad_norm": 5.398906230926514, + "learning_rate": 5.417978081511379e-06, + "loss": 5.820648193359375, + "step": 45550 + }, + { + "epoch": 1.4312475465180183, + "grad_norm": 3.164841890335083, + "learning_rate": 5.4092070847489e-06, + "loss": 5.806771850585937, + "step": 45575 + }, + { + "epoch": 1.4320326607521394, + "grad_norm": 5.4601898193359375, + "learning_rate": 5.400444725950365e-06, + "loss": 5.8117437744140625, + "step": 45600 + }, + { + "epoch": 1.4328177749862605, + "grad_norm": 3.4484071731567383, + "learning_rate": 5.391691018653908e-06, + "loss": 5.813043823242188, + "step": 45625 + }, + { + "epoch": 1.4336028892203816, + "grad_norm": 4.177372455596924, + "learning_rate": 5.382945976384299e-06, + "loss": 5.823253173828125, + "step": 45650 + }, + { + "epoch": 1.4343880034545027, + "grad_norm": 3.326119899749756, + "learning_rate": 5.374209612652934e-06, + "loss": 5.81277587890625, + "step": 45675 + }, + { + "epoch": 1.4351731176886238, + "grad_norm": 3.617121696472168, + "learning_rate": 5.365481940957783e-06, + "loss": 5.801365966796875, + "step": 45700 + }, + { + "epoch": 1.4359582319227449, + "grad_norm": 4.412740230560303, + "learning_rate": 5.356762974783396e-06, + "loss": 5.799016723632812, + "step": 45725 + }, + { + "epoch": 1.4367433461568657, + "grad_norm": 4.498205661773682, + "learning_rate": 5.34805272760087e-06, + "loss": 5.8119580078125, + "step": 45750 + }, + { + "epoch": 1.4375284603909868, + "grad_norm": 5.404195785522461, + "learning_rate": 5.339351212867828e-06, + "loss": 5.800308837890625, + "step": 45775 + }, + { + "epoch": 1.438313574625108, + "grad_norm": 2.986529588699341, + "learning_rate": 5.330658444028413e-06, + "loss": 5.801492309570312, + "step": 45800 + }, + { + "epoch": 1.439098688859229, + "grad_norm": 4.476986408233643, + "learning_rate": 5.321974434513239e-06, + "loss": 5.810015869140625, + "step": 45825 + }, + { + "epoch": 1.43988380309335, + "grad_norm": 4.673818588256836, + "learning_rate": 5.31329919773939e-06, + "loss": 5.80731201171875, + "step": 45850 + }, + { + "epoch": 1.4406689173274712, + "grad_norm": 3.4040024280548096, + "learning_rate": 5.304632747110409e-06, + "loss": 5.805257568359375, + "step": 45875 + }, + { + "epoch": 1.4414540315615922, + "grad_norm": 3.631460189819336, + "learning_rate": 5.295975096016254e-06, + "loss": 5.819400634765625, + "step": 45900 + }, + { + "epoch": 1.4422391457957133, + "grad_norm": 4.559212684631348, + "learning_rate": 5.287326257833285e-06, + "loss": 5.816569213867187, + "step": 45925 + }, + { + "epoch": 1.4430242600298344, + "grad_norm": 5.349948406219482, + "learning_rate": 5.278686245924249e-06, + "loss": 5.792520141601562, + "step": 45950 + }, + { + "epoch": 1.4438093742639553, + "grad_norm": 4.818277835845947, + "learning_rate": 5.2700550736382585e-06, + "loss": 5.802567138671875, + "step": 45975 + }, + { + "epoch": 1.4445944884980766, + "grad_norm": 4.882867813110352, + "learning_rate": 5.261432754310767e-06, + "loss": 5.8112640380859375, + "step": 46000 + }, + { + "epoch": 1.4453796027321975, + "grad_norm": 2.8960342407226562, + "learning_rate": 5.252819301263545e-06, + "loss": 5.803134155273438, + "step": 46025 + }, + { + "epoch": 1.4461647169663185, + "grad_norm": 4.030308246612549, + "learning_rate": 5.244214727804675e-06, + "loss": 5.811726684570313, + "step": 46050 + }, + { + "epoch": 1.4469498312004396, + "grad_norm": 4.752276420593262, + "learning_rate": 5.235619047228506e-06, + "loss": 5.807528686523438, + "step": 46075 + }, + { + "epoch": 1.4477349454345607, + "grad_norm": 3.252326250076294, + "learning_rate": 5.22703227281566e-06, + "loss": 5.8058624267578125, + "step": 46100 + }, + { + "epoch": 1.4485200596686818, + "grad_norm": 3.9688398838043213, + "learning_rate": 5.218454417832992e-06, + "loss": 5.8082952880859375, + "step": 46125 + }, + { + "epoch": 1.449305173902803, + "grad_norm": 3.126335382461548, + "learning_rate": 5.209885495533578e-06, + "loss": 5.807700805664062, + "step": 46150 + }, + { + "epoch": 1.450090288136924, + "grad_norm": 3.261836528778076, + "learning_rate": 5.201325519156691e-06, + "loss": 5.801231689453125, + "step": 46175 + }, + { + "epoch": 1.450875402371045, + "grad_norm": 5.625180721282959, + "learning_rate": 5.1927745019277884e-06, + "loss": 5.801173706054687, + "step": 46200 + }, + { + "epoch": 1.4516605166051662, + "grad_norm": 4.276218891143799, + "learning_rate": 5.1842324570584755e-06, + "loss": 5.8236932373046875, + "step": 46225 + }, + { + "epoch": 1.452445630839287, + "grad_norm": 3.827385425567627, + "learning_rate": 5.175699397746505e-06, + "loss": 5.819913940429688, + "step": 46250 + }, + { + "epoch": 1.453230745073408, + "grad_norm": 3.1872832775115967, + "learning_rate": 5.167175337175736e-06, + "loss": 5.797150268554687, + "step": 46275 + }, + { + "epoch": 1.4540158593075292, + "grad_norm": 4.119002819061279, + "learning_rate": 5.15866028851613e-06, + "loss": 5.813787841796875, + "step": 46300 + }, + { + "epoch": 1.4548009735416503, + "grad_norm": 4.365518093109131, + "learning_rate": 5.150154264923735e-06, + "loss": 5.817274169921875, + "step": 46325 + }, + { + "epoch": 1.4555860877757714, + "grad_norm": 3.785271406173706, + "learning_rate": 5.141657279540633e-06, + "loss": 5.81601806640625, + "step": 46350 + }, + { + "epoch": 1.4563712020098925, + "grad_norm": 3.9039387702941895, + "learning_rate": 5.13316934549496e-06, + "loss": 5.80700439453125, + "step": 46375 + }, + { + "epoch": 1.4571563162440135, + "grad_norm": 3.050629138946533, + "learning_rate": 5.12469047590086e-06, + "loss": 5.809136962890625, + "step": 46400 + }, + { + "epoch": 1.4579414304781346, + "grad_norm": 4.810796737670898, + "learning_rate": 5.116220683858468e-06, + "loss": 5.802843017578125, + "step": 46425 + }, + { + "epoch": 1.4587265447122557, + "grad_norm": 3.815103769302368, + "learning_rate": 5.107759982453908e-06, + "loss": 5.82484619140625, + "step": 46450 + }, + { + "epoch": 1.4595116589463766, + "grad_norm": 3.6164090633392334, + "learning_rate": 5.099308384759242e-06, + "loss": 5.80244384765625, + "step": 46475 + }, + { + "epoch": 1.4602967731804979, + "grad_norm": 4.872348785400391, + "learning_rate": 5.090865903832476e-06, + "loss": 5.81367431640625, + "step": 46500 + }, + { + "epoch": 1.4610818874146188, + "grad_norm": 3.8753013610839844, + "learning_rate": 5.082432552717531e-06, + "loss": 5.8145208740234375, + "step": 46525 + }, + { + "epoch": 1.4618670016487398, + "grad_norm": 5.176597595214844, + "learning_rate": 5.074008344444214e-06, + "loss": 5.807797241210937, + "step": 46550 + }, + { + "epoch": 1.462652115882861, + "grad_norm": 3.469132900238037, + "learning_rate": 5.065593292028216e-06, + "loss": 5.787489013671875, + "step": 46575 + }, + { + "epoch": 1.463437230116982, + "grad_norm": 3.323320150375366, + "learning_rate": 5.057187408471078e-06, + "loss": 5.809095458984375, + "step": 46600 + }, + { + "epoch": 1.464222344351103, + "grad_norm": 3.2163000106811523, + "learning_rate": 5.048790706760169e-06, + "loss": 5.803060302734375, + "step": 46625 + }, + { + "epoch": 1.4650074585852242, + "grad_norm": 4.360834121704102, + "learning_rate": 5.040403199868682e-06, + "loss": 5.810787353515625, + "step": 46650 + }, + { + "epoch": 1.4657925728193453, + "grad_norm": 3.755800485610962, + "learning_rate": 5.03202490075559e-06, + "loss": 5.807811279296875, + "step": 46675 + }, + { + "epoch": 1.4665776870534664, + "grad_norm": 4.816633701324463, + "learning_rate": 5.0236558223656615e-06, + "loss": 5.802965087890625, + "step": 46700 + }, + { + "epoch": 1.4673628012875874, + "grad_norm": 3.0495412349700928, + "learning_rate": 5.015295977629397e-06, + "loss": 5.811248168945313, + "step": 46725 + }, + { + "epoch": 1.4681479155217083, + "grad_norm": 5.4774322509765625, + "learning_rate": 5.006945379463039e-06, + "loss": 5.826179809570313, + "step": 46750 + }, + { + "epoch": 1.4689330297558294, + "grad_norm": 9.656095504760742, + "learning_rate": 4.9986040407685445e-06, + "loss": 5.81126953125, + "step": 46775 + }, + { + "epoch": 1.4697181439899505, + "grad_norm": 6.818397045135498, + "learning_rate": 4.990271974433567e-06, + "loss": 5.815377197265625, + "step": 46800 + }, + { + "epoch": 1.4705032582240716, + "grad_norm": 2.904303789138794, + "learning_rate": 4.981949193331424e-06, + "loss": 5.81883056640625, + "step": 46825 + }, + { + "epoch": 1.4712883724581927, + "grad_norm": 4.488351821899414, + "learning_rate": 4.973635710321101e-06, + "loss": 5.809859619140625, + "step": 46850 + }, + { + "epoch": 1.4720734866923137, + "grad_norm": 5.303575038909912, + "learning_rate": 4.9653315382472e-06, + "loss": 5.810093383789063, + "step": 46875 + }, + { + "epoch": 1.4728586009264348, + "grad_norm": 4.886533737182617, + "learning_rate": 4.957036689939961e-06, + "loss": 5.808573608398437, + "step": 46900 + }, + { + "epoch": 1.473643715160556, + "grad_norm": 5.1588826179504395, + "learning_rate": 4.948751178215195e-06, + "loss": 5.815363159179688, + "step": 46925 + }, + { + "epoch": 1.474428829394677, + "grad_norm": 4.504428386688232, + "learning_rate": 4.9404750158743e-06, + "loss": 5.808534545898437, + "step": 46950 + }, + { + "epoch": 1.4752139436287979, + "grad_norm": 2.626024007797241, + "learning_rate": 4.932208215704225e-06, + "loss": 5.821071166992187, + "step": 46975 + }, + { + "epoch": 1.4759990578629192, + "grad_norm": 3.6005520820617676, + "learning_rate": 4.923950790477457e-06, + "loss": 5.803356323242188, + "step": 47000 + }, + { + "epoch": 1.47678417209704, + "grad_norm": 4.191322326660156, + "learning_rate": 4.915702752952001e-06, + "loss": 5.8200732421875, + "step": 47025 + }, + { + "epoch": 1.4775692863311611, + "grad_norm": 3.170454978942871, + "learning_rate": 4.907464115871349e-06, + "loss": 5.80082275390625, + "step": 47050 + }, + { + "epoch": 1.4783544005652822, + "grad_norm": 4.435052394866943, + "learning_rate": 4.899234891964469e-06, + "loss": 5.810223388671875, + "step": 47075 + }, + { + "epoch": 1.4791395147994033, + "grad_norm": 4.7515549659729, + "learning_rate": 4.8910150939458e-06, + "loss": 5.814357299804687, + "step": 47100 + }, + { + "epoch": 1.4799246290335244, + "grad_norm": 4.163820266723633, + "learning_rate": 4.8828047345152015e-06, + "loss": 5.795509643554688, + "step": 47125 + }, + { + "epoch": 1.4807097432676455, + "grad_norm": 4.123980522155762, + "learning_rate": 4.8746038263579595e-06, + "loss": 5.801663818359375, + "step": 47150 + }, + { + "epoch": 1.4814948575017666, + "grad_norm": 3.203594923019409, + "learning_rate": 4.866412382144749e-06, + "loss": 5.810164794921875, + "step": 47175 + }, + { + "epoch": 1.4822799717358874, + "grad_norm": 2.509615182876587, + "learning_rate": 4.8582304145316325e-06, + "loss": 5.805584716796875, + "step": 47200 + }, + { + "epoch": 1.4830650859700087, + "grad_norm": 4.635387897491455, + "learning_rate": 4.8500579361600285e-06, + "loss": 5.797359619140625, + "step": 47225 + }, + { + "epoch": 1.4838502002041296, + "grad_norm": 2.8707964420318604, + "learning_rate": 4.841894959656685e-06, + "loss": 5.7976837158203125, + "step": 47250 + }, + { + "epoch": 1.4846353144382507, + "grad_norm": 4.088871002197266, + "learning_rate": 4.833741497633681e-06, + "loss": 5.807738647460938, + "step": 47275 + }, + { + "epoch": 1.4854204286723718, + "grad_norm": 3.1679604053497314, + "learning_rate": 4.825597562688392e-06, + "loss": 5.803985595703125, + "step": 47300 + }, + { + "epoch": 1.4862055429064929, + "grad_norm": 5.167155742645264, + "learning_rate": 4.8174631674034705e-06, + "loss": 5.814492797851562, + "step": 47325 + }, + { + "epoch": 1.486990657140614, + "grad_norm": 4.09507417678833, + "learning_rate": 4.809338324346834e-06, + "loss": 5.804129028320313, + "step": 47350 + }, + { + "epoch": 1.487775771374735, + "grad_norm": 5.278029918670654, + "learning_rate": 4.801223046071635e-06, + "loss": 5.795820922851562, + "step": 47375 + }, + { + "epoch": 1.4885608856088561, + "grad_norm": 3.268155097961426, + "learning_rate": 4.793117345116255e-06, + "loss": 5.8060009765625, + "step": 47400 + }, + { + "epoch": 1.4893459998429772, + "grad_norm": 4.783140182495117, + "learning_rate": 4.785021234004279e-06, + "loss": 5.797471313476563, + "step": 47425 + }, + { + "epoch": 1.4901311140770983, + "grad_norm": 1.8614799976348877, + "learning_rate": 4.776934725244468e-06, + "loss": 5.803027954101562, + "step": 47450 + }, + { + "epoch": 1.4909162283112192, + "grad_norm": 5.835747241973877, + "learning_rate": 4.768857831330752e-06, + "loss": 5.813692626953125, + "step": 47475 + }, + { + "epoch": 1.4917013425453405, + "grad_norm": 4.276222229003906, + "learning_rate": 4.760790564742209e-06, + "loss": 5.802783813476562, + "step": 47500 + }, + { + "epoch": 1.4924864567794613, + "grad_norm": 2.9051780700683594, + "learning_rate": 4.752732937943033e-06, + "loss": 5.78793212890625, + "step": 47525 + }, + { + "epoch": 1.4932715710135824, + "grad_norm": 5.623703956604004, + "learning_rate": 4.744684963382533e-06, + "loss": 5.805918579101562, + "step": 47550 + }, + { + "epoch": 1.4940566852477035, + "grad_norm": 4.607259273529053, + "learning_rate": 4.736646653495099e-06, + "loss": 5.806642456054687, + "step": 47575 + }, + { + "epoch": 1.4948417994818246, + "grad_norm": 4.110782146453857, + "learning_rate": 4.728618020700191e-06, + "loss": 5.804598388671875, + "step": 47600 + }, + { + "epoch": 1.4956269137159457, + "grad_norm": 3.3246264457702637, + "learning_rate": 4.720599077402325e-06, + "loss": 5.805532836914063, + "step": 47625 + }, + { + "epoch": 1.4964120279500668, + "grad_norm": 2.9042410850524902, + "learning_rate": 4.7125898359910295e-06, + "loss": 5.792052001953125, + "step": 47650 + }, + { + "epoch": 1.4971971421841879, + "grad_norm": 3.220515489578247, + "learning_rate": 4.704590308840858e-06, + "loss": 5.824701538085938, + "step": 47675 + }, + { + "epoch": 1.4979822564183087, + "grad_norm": 3.847501754760742, + "learning_rate": 4.696600508311352e-06, + "loss": 5.813218994140625, + "step": 47700 + }, + { + "epoch": 1.49876737065243, + "grad_norm": 4.128615856170654, + "learning_rate": 4.68862044674702e-06, + "loss": 5.799762573242187, + "step": 47725 + }, + { + "epoch": 1.499552484886551, + "grad_norm": 4.460413932800293, + "learning_rate": 4.68065013647733e-06, + "loss": 5.805140991210937, + "step": 47750 + }, + { + "epoch": 1.5000235534270236, + "eval_loss": 0.1449582427740097, + "eval_runtime": 5080.5413, + "eval_samples_per_second": 4.267, + "eval_steps_per_second": 0.267, + "eval_wer": 3.422239044091931, + "step": 47765 + }, + { + "epoch": 1.5003375991206722, + "grad_norm": 4.308821201324463, + "learning_rate": 4.672689589816675e-06, + "loss": 5.8106494140625, + "step": 47775 + }, + { + "epoch": 1.501122713354793, + "grad_norm": 3.5418992042541504, + "learning_rate": 4.664738819064372e-06, + "loss": 5.794514770507813, + "step": 47800 + }, + { + "epoch": 1.5019078275889142, + "grad_norm": 5.12945032119751, + "learning_rate": 4.656797836504633e-06, + "loss": 5.8130224609375, + "step": 47825 + }, + { + "epoch": 1.5026929418230353, + "grad_norm": 3.6653265953063965, + "learning_rate": 4.648866654406539e-06, + "loss": 5.803592529296875, + "step": 47850 + }, + { + "epoch": 1.5034780560571563, + "grad_norm": 2.855783700942993, + "learning_rate": 4.640945285024037e-06, + "loss": 5.801455688476563, + "step": 47875 + }, + { + "epoch": 1.5042631702912774, + "grad_norm": 4.742208003997803, + "learning_rate": 4.633033740595914e-06, + "loss": 5.81910400390625, + "step": 47900 + }, + { + "epoch": 1.5050482845253983, + "grad_norm": 5.099589824676514, + "learning_rate": 4.625132033345766e-06, + "loss": 5.804302978515625, + "step": 47925 + }, + { + "epoch": 1.5058333987595196, + "grad_norm": 4.124429225921631, + "learning_rate": 4.617240175482004e-06, + "loss": 5.8045849609375, + "step": 47950 + }, + { + "epoch": 1.5066185129936405, + "grad_norm": 4.627848148345947, + "learning_rate": 4.609358179197809e-06, + "loss": 5.7966650390625, + "step": 47975 + }, + { + "epoch": 1.5074036272277618, + "grad_norm": 4.399104595184326, + "learning_rate": 4.601486056671139e-06, + "loss": 5.792603759765625, + "step": 48000 + }, + { + "epoch": 1.5081887414618826, + "grad_norm": 4.538598537445068, + "learning_rate": 4.5936238200646845e-06, + "loss": 5.807698364257813, + "step": 48025 + }, + { + "epoch": 1.5089738556960037, + "grad_norm": 4.9471635818481445, + "learning_rate": 4.5857714815258656e-06, + "loss": 5.8064453125, + "step": 48050 + }, + { + "epoch": 1.5097589699301248, + "grad_norm": 3.4664576053619385, + "learning_rate": 4.577929053186812e-06, + "loss": 5.8002557373046875, + "step": 48075 + }, + { + "epoch": 1.510544084164246, + "grad_norm": 3.4974417686462402, + "learning_rate": 4.570096547164342e-06, + "loss": 5.792763671875, + "step": 48100 + }, + { + "epoch": 1.511329198398367, + "grad_norm": 3.5879738330841064, + "learning_rate": 4.5622739755599435e-06, + "loss": 5.81141845703125, + "step": 48125 + }, + { + "epoch": 1.512114312632488, + "grad_norm": 3.951632022857666, + "learning_rate": 4.554461350459751e-06, + "loss": 5.804617919921875, + "step": 48150 + }, + { + "epoch": 1.5128994268666092, + "grad_norm": 3.418093681335449, + "learning_rate": 4.546658683934529e-06, + "loss": 5.7985986328125, + "step": 48175 + }, + { + "epoch": 1.51368454110073, + "grad_norm": 5.200983047485352, + "learning_rate": 4.538865988039672e-06, + "loss": 5.8123779296875, + "step": 48200 + }, + { + "epoch": 1.5144696553348513, + "grad_norm": 4.20051908493042, + "learning_rate": 4.5310832748151504e-06, + "loss": 5.78998046875, + "step": 48225 + }, + { + "epoch": 1.5152547695689722, + "grad_norm": 2.581824541091919, + "learning_rate": 4.523310556285524e-06, + "loss": 5.793152465820312, + "step": 48250 + }, + { + "epoch": 1.5160398838030935, + "grad_norm": 4.623429298400879, + "learning_rate": 4.515547844459901e-06, + "loss": 5.801927490234375, + "step": 48275 + }, + { + "epoch": 1.5168249980372144, + "grad_norm": 3.9095213413238525, + "learning_rate": 4.5077951513319354e-06, + "loss": 5.791971435546875, + "step": 48300 + }, + { + "epoch": 1.5176101122713355, + "grad_norm": 5.25413179397583, + "learning_rate": 4.500052488879803e-06, + "loss": 5.81707763671875, + "step": 48325 + }, + { + "epoch": 1.5183952265054566, + "grad_norm": 7.758753776550293, + "learning_rate": 4.492319869066173e-06, + "loss": 5.8115478515625, + "step": 48350 + }, + { + "epoch": 1.5191803407395776, + "grad_norm": 3.5600297451019287, + "learning_rate": 4.484597303838207e-06, + "loss": 5.788359985351563, + "step": 48375 + }, + { + "epoch": 1.5199654549736987, + "grad_norm": 3.6756205558776855, + "learning_rate": 4.476884805127533e-06, + "loss": 5.819822387695313, + "step": 48400 + }, + { + "epoch": 1.5207505692078196, + "grad_norm": 5.288193225860596, + "learning_rate": 4.469182384850216e-06, + "loss": 5.805006103515625, + "step": 48425 + }, + { + "epoch": 1.521535683441941, + "grad_norm": 4.92435884475708, + "learning_rate": 4.46149005490676e-06, + "loss": 5.810961303710937, + "step": 48450 + }, + { + "epoch": 1.5223207976760618, + "grad_norm": 4.995941162109375, + "learning_rate": 4.453807827182071e-06, + "loss": 5.803041381835937, + "step": 48475 + }, + { + "epoch": 1.523105911910183, + "grad_norm": 4.107648849487305, + "learning_rate": 4.446135713545451e-06, + "loss": 5.817176513671875, + "step": 48500 + }, + { + "epoch": 1.523891026144304, + "grad_norm": 3.527723789215088, + "learning_rate": 4.43847372585058e-06, + "loss": 5.797156982421875, + "step": 48525 + }, + { + "epoch": 1.524676140378425, + "grad_norm": 3.9134833812713623, + "learning_rate": 4.430821875935479e-06, + "loss": 5.7935595703125, + "step": 48550 + }, + { + "epoch": 1.5254612546125461, + "grad_norm": 3.4801197052001953, + "learning_rate": 4.42318017562252e-06, + "loss": 5.789678344726562, + "step": 48575 + }, + { + "epoch": 1.5262463688466672, + "grad_norm": 3.15828013420105, + "learning_rate": 4.415548636718392e-06, + "loss": 5.80240966796875, + "step": 48600 + }, + { + "epoch": 1.5270314830807883, + "grad_norm": 4.717751502990723, + "learning_rate": 4.407927271014072e-06, + "loss": 5.79595458984375, + "step": 48625 + }, + { + "epoch": 1.5278165973149094, + "grad_norm": 3.3242313861846924, + "learning_rate": 4.400316090284834e-06, + "loss": 5.799410400390625, + "step": 48650 + }, + { + "epoch": 1.5286017115490305, + "grad_norm": 3.475637674331665, + "learning_rate": 4.392715106290205e-06, + "loss": 5.802703247070313, + "step": 48675 + }, + { + "epoch": 1.5293868257831513, + "grad_norm": 4.224560737609863, + "learning_rate": 4.385124330773964e-06, + "loss": 5.796054077148438, + "step": 48700 + }, + { + "epoch": 1.5301719400172726, + "grad_norm": 3.253397226333618, + "learning_rate": 4.377543775464118e-06, + "loss": 5.8087944984436035, + "step": 48725 + }, + { + "epoch": 1.5309570542513935, + "grad_norm": 4.4590744972229, + "learning_rate": 4.369973452072876e-06, + "loss": 5.800333862304687, + "step": 48750 + }, + { + "epoch": 1.5317421684855146, + "grad_norm": 3.7819671630859375, + "learning_rate": 4.362413372296646e-06, + "loss": 5.800781860351562, + "step": 48775 + }, + { + "epoch": 1.5325272827196357, + "grad_norm": 5.034989356994629, + "learning_rate": 4.354863547816009e-06, + "loss": 5.8006439208984375, + "step": 48800 + }, + { + "epoch": 1.5333123969537568, + "grad_norm": 2.922006607055664, + "learning_rate": 4.347323990295693e-06, + "loss": 5.796203002929688, + "step": 48825 + }, + { + "epoch": 1.5340975111878778, + "grad_norm": 3.930546760559082, + "learning_rate": 4.339794711384577e-06, + "loss": 5.7989697265625, + "step": 48850 + }, + { + "epoch": 1.534882625421999, + "grad_norm": 3.7057647705078125, + "learning_rate": 4.332275722715644e-06, + "loss": 5.797864379882813, + "step": 48875 + }, + { + "epoch": 1.53566773965612, + "grad_norm": 4.315715789794922, + "learning_rate": 4.324767035905989e-06, + "loss": 5.809759521484375, + "step": 48900 + }, + { + "epoch": 1.5364528538902409, + "grad_norm": 2.6745822429656982, + "learning_rate": 4.317268662556788e-06, + "loss": 5.794406127929688, + "step": 48925 + }, + { + "epoch": 1.5372379681243622, + "grad_norm": 5.32199764251709, + "learning_rate": 4.309780614253278e-06, + "loss": 5.804632568359375, + "step": 48950 + }, + { + "epoch": 1.538023082358483, + "grad_norm": 3.549057960510254, + "learning_rate": 4.3023029025647465e-06, + "loss": 5.797032470703125, + "step": 48975 + }, + { + "epoch": 1.5388081965926044, + "grad_norm": 5.501411437988281, + "learning_rate": 4.294835539044515e-06, + "loss": 5.791925048828125, + "step": 49000 + }, + { + "epoch": 1.5395933108267252, + "grad_norm": 3.4964258670806885, + "learning_rate": 4.287378535229908e-06, + "loss": 5.811981201171875, + "step": 49025 + }, + { + "epoch": 1.5403784250608463, + "grad_norm": 6.616456985473633, + "learning_rate": 4.279931902642254e-06, + "loss": 5.791806030273437, + "step": 49050 + }, + { + "epoch": 1.5411635392949674, + "grad_norm": 3.3583388328552246, + "learning_rate": 4.27249565278684e-06, + "loss": 5.798595581054688, + "step": 49075 + }, + { + "epoch": 1.5419486535290885, + "grad_norm": 4.477759838104248, + "learning_rate": 4.265069797152938e-06, + "loss": 5.8063525390625, + "step": 49100 + }, + { + "epoch": 1.5427337677632096, + "grad_norm": 4.019205570220947, + "learning_rate": 4.257654347213739e-06, + "loss": 5.801917724609375, + "step": 49125 + }, + { + "epoch": 1.5435188819973304, + "grad_norm": 4.459771156311035, + "learning_rate": 4.25024931442636e-06, + "loss": 5.808381958007812, + "step": 49150 + }, + { + "epoch": 1.5443039962314518, + "grad_norm": 3.7705111503601074, + "learning_rate": 4.242854710231832e-06, + "loss": 5.816163330078125, + "step": 49175 + }, + { + "epoch": 1.5450891104655726, + "grad_norm": 4.069270133972168, + "learning_rate": 4.235470546055066e-06, + "loss": 5.79068115234375, + "step": 49200 + }, + { + "epoch": 1.545874224699694, + "grad_norm": 4.601371765136719, + "learning_rate": 4.228096833304849e-06, + "loss": 5.800584106445313, + "step": 49225 + }, + { + "epoch": 1.5466593389338148, + "grad_norm": 5.955254077911377, + "learning_rate": 4.220733583373813e-06, + "loss": 5.785252075195313, + "step": 49250 + }, + { + "epoch": 1.5474444531679359, + "grad_norm": 3.9174818992614746, + "learning_rate": 4.213380807638424e-06, + "loss": 5.79844482421875, + "step": 49275 + }, + { + "epoch": 1.548229567402057, + "grad_norm": 4.743170738220215, + "learning_rate": 4.206038517458977e-06, + "loss": 5.814091796875, + "step": 49300 + }, + { + "epoch": 1.549014681636178, + "grad_norm": 3.9399921894073486, + "learning_rate": 4.198706724179554e-06, + "loss": 5.796376953125, + "step": 49325 + }, + { + "epoch": 1.5497997958702991, + "grad_norm": 3.090674638748169, + "learning_rate": 4.191385439128024e-06, + "loss": 5.795886840820312, + "step": 49350 + }, + { + "epoch": 1.5505849101044202, + "grad_norm": 2.4802112579345703, + "learning_rate": 4.1840746736160185e-06, + "loss": 5.799581909179688, + "step": 49375 + }, + { + "epoch": 1.5513700243385413, + "grad_norm": 5.521714210510254, + "learning_rate": 4.1767744389389184e-06, + "loss": 5.80541748046875, + "step": 49400 + }, + { + "epoch": 1.5521551385726622, + "grad_norm": 4.181097030639648, + "learning_rate": 4.1694847463758344e-06, + "loss": 5.789967651367188, + "step": 49425 + }, + { + "epoch": 1.5529402528067835, + "grad_norm": 4.555105209350586, + "learning_rate": 4.162205607189586e-06, + "loss": 5.818929443359375, + "step": 49450 + }, + { + "epoch": 1.5537253670409044, + "grad_norm": 3.933619976043701, + "learning_rate": 4.154937032626687e-06, + "loss": 5.8026055908203125, + "step": 49475 + }, + { + "epoch": 1.5545104812750257, + "grad_norm": 4.84575080871582, + "learning_rate": 4.1476790339173364e-06, + "loss": 5.798392333984375, + "step": 49500 + }, + { + "epoch": 1.5552955955091465, + "grad_norm": 3.3546805381774902, + "learning_rate": 4.140431622275382e-06, + "loss": 5.785770874023438, + "step": 49525 + }, + { + "epoch": 1.5560807097432676, + "grad_norm": 3.6085240840911865, + "learning_rate": 4.1331948088983245e-06, + "loss": 5.799102783203125, + "step": 49550 + }, + { + "epoch": 1.5568658239773887, + "grad_norm": 5.256474018096924, + "learning_rate": 4.125968604967278e-06, + "loss": 5.803284301757812, + "step": 49575 + }, + { + "epoch": 1.5576509382115098, + "grad_norm": 3.824007749557495, + "learning_rate": 4.118753021646977e-06, + "loss": 5.816071166992187, + "step": 49600 + }, + { + "epoch": 1.5584360524456309, + "grad_norm": 4.072995185852051, + "learning_rate": 4.111548070085745e-06, + "loss": 5.794486083984375, + "step": 49625 + }, + { + "epoch": 1.5592211666797517, + "grad_norm": 4.562678337097168, + "learning_rate": 4.104353761415466e-06, + "loss": 5.800682373046875, + "step": 49650 + }, + { + "epoch": 1.560006280913873, + "grad_norm": 3.9635262489318848, + "learning_rate": 4.097170106751595e-06, + "loss": 5.805027465820313, + "step": 49675 + }, + { + "epoch": 1.560791395147994, + "grad_norm": 4.6036882400512695, + "learning_rate": 4.0899971171931255e-06, + "loss": 5.817686157226563, + "step": 49700 + }, + { + "epoch": 1.5615765093821152, + "grad_norm": 4.044723033905029, + "learning_rate": 4.0828348038225616e-06, + "loss": 5.795219116210937, + "step": 49725 + }, + { + "epoch": 1.562361623616236, + "grad_norm": 3.1099600791931152, + "learning_rate": 4.075683177705923e-06, + "loss": 5.795703125, + "step": 49750 + }, + { + "epoch": 1.5631467378503572, + "grad_norm": 3.8120906352996826, + "learning_rate": 4.068542249892713e-06, + "loss": 5.797279663085938, + "step": 49775 + }, + { + "epoch": 1.5639318520844783, + "grad_norm": 2.5287041664123535, + "learning_rate": 4.061412031415904e-06, + "loss": 5.785037231445313, + "step": 49800 + }, + { + "epoch": 1.5647169663185994, + "grad_norm": 4.744734764099121, + "learning_rate": 4.054292533291932e-06, + "loss": 5.813002319335937, + "step": 49825 + }, + { + "epoch": 1.5655020805527204, + "grad_norm": 2.928680896759033, + "learning_rate": 4.047183766520652e-06, + "loss": 5.799598999023438, + "step": 49850 + }, + { + "epoch": 1.5662871947868415, + "grad_norm": 5.562773704528809, + "learning_rate": 4.0400857420853545e-06, + "loss": 5.787500610351563, + "step": 49875 + }, + { + "epoch": 1.5670723090209626, + "grad_norm": 5.248204231262207, + "learning_rate": 4.032998470952728e-06, + "loss": 5.808291625976563, + "step": 49900 + }, + { + "epoch": 1.5678574232550835, + "grad_norm": 8.129254341125488, + "learning_rate": 4.025921964072842e-06, + "loss": 5.795855102539062, + "step": 49925 + }, + { + "epoch": 1.5686425374892048, + "grad_norm": 4.619943141937256, + "learning_rate": 4.018856232379143e-06, + "loss": 5.87205696105957, + "step": 49950 + }, + { + "epoch": 1.5694276517233257, + "grad_norm": 3.950857400894165, + "learning_rate": 4.011801286788422e-06, + "loss": 5.79917724609375, + "step": 49975 + }, + { + "epoch": 1.570212765957447, + "grad_norm": 4.116634845733643, + "learning_rate": 4.0047571382008064e-06, + "loss": 5.788277587890625, + "step": 50000 + }, + { + "epoch": 1.5709978801915678, + "grad_norm": 3.7676749229431152, + "learning_rate": 3.997723797499752e-06, + "loss": 5.81627685546875, + "step": 50025 + }, + { + "epoch": 1.571782994425689, + "grad_norm": 3.3878965377807617, + "learning_rate": 3.990701275551998e-06, + "loss": 5.79866455078125, + "step": 50050 + }, + { + "epoch": 1.57256810865981, + "grad_norm": 3.9051897525787354, + "learning_rate": 3.983689583207584e-06, + "loss": 5.7900390625, + "step": 50075 + }, + { + "epoch": 1.573353222893931, + "grad_norm": 4.049507141113281, + "learning_rate": 3.976688731299812e-06, + "loss": 5.80328857421875, + "step": 50100 + }, + { + "epoch": 1.5741383371280522, + "grad_norm": 3.169278860092163, + "learning_rate": 3.96969873064523e-06, + "loss": 5.794063110351562, + "step": 50125 + }, + { + "epoch": 1.574923451362173, + "grad_norm": 3.8539087772369385, + "learning_rate": 3.96271959204363e-06, + "loss": 5.805430908203125, + "step": 50150 + }, + { + "epoch": 1.5757085655962944, + "grad_norm": 3.972229480743408, + "learning_rate": 3.955751326278011e-06, + "loss": 5.797606201171875, + "step": 50175 + }, + { + "epoch": 1.5764936798304152, + "grad_norm": 3.0209920406341553, + "learning_rate": 3.948793944114586e-06, + "loss": 5.814797973632812, + "step": 50200 + }, + { + "epoch": 1.5772787940645365, + "grad_norm": 3.0055081844329834, + "learning_rate": 3.941847456302744e-06, + "loss": 5.776929321289063, + "step": 50225 + }, + { + "epoch": 1.5780639082986574, + "grad_norm": 3.2291371822357178, + "learning_rate": 3.934911873575039e-06, + "loss": 5.787539672851563, + "step": 50250 + }, + { + "epoch": 1.5788490225327785, + "grad_norm": 3.6499080657958984, + "learning_rate": 3.927987206647184e-06, + "loss": 5.790692749023438, + "step": 50275 + }, + { + "epoch": 1.5796341367668996, + "grad_norm": 3.6828227043151855, + "learning_rate": 3.921073466218022e-06, + "loss": 5.813136596679687, + "step": 50300 + }, + { + "epoch": 1.5804192510010207, + "grad_norm": 4.607122421264648, + "learning_rate": 3.9141706629695175e-06, + "loss": 5.80171875, + "step": 50325 + }, + { + "epoch": 1.5812043652351417, + "grad_norm": 2.6995022296905518, + "learning_rate": 3.907278807566736e-06, + "loss": 5.788217163085937, + "step": 50350 + }, + { + "epoch": 1.5819894794692628, + "grad_norm": 2.7353358268737793, + "learning_rate": 3.900397910657817e-06, + "loss": 5.78747802734375, + "step": 50375 + }, + { + "epoch": 1.582774593703384, + "grad_norm": 3.577000856399536, + "learning_rate": 3.8935279828739925e-06, + "loss": 5.8153448486328125, + "step": 50400 + }, + { + "epoch": 1.5835597079375048, + "grad_norm": 4.148398399353027, + "learning_rate": 3.886669034829524e-06, + "loss": 5.815330200195312, + "step": 50425 + }, + { + "epoch": 1.584344822171626, + "grad_norm": 3.7801356315612793, + "learning_rate": 3.8798210771217235e-06, + "loss": 5.79373291015625, + "step": 50450 + }, + { + "epoch": 1.585129936405747, + "grad_norm": 3.0653574466705322, + "learning_rate": 3.872984120330911e-06, + "loss": 5.791068725585937, + "step": 50475 + }, + { + "epoch": 1.5859150506398683, + "grad_norm": 4.347634315490723, + "learning_rate": 3.866158175020419e-06, + "loss": 5.786807250976563, + "step": 50500 + }, + { + "epoch": 1.5867001648739891, + "grad_norm": 3.9783742427825928, + "learning_rate": 3.859343251736566e-06, + "loss": 5.798259887695313, + "step": 50525 + }, + { + "epoch": 1.5874852791081102, + "grad_norm": 3.458160400390625, + "learning_rate": 3.852539361008634e-06, + "loss": 5.808035888671875, + "step": 50550 + }, + { + "epoch": 1.5882703933422313, + "grad_norm": 5.4045867919921875, + "learning_rate": 3.845746513348866e-06, + "loss": 5.785310668945312, + "step": 50575 + }, + { + "epoch": 1.5890555075763524, + "grad_norm": 3.6321609020233154, + "learning_rate": 3.838964719252444e-06, + "loss": 5.794059448242187, + "step": 50600 + }, + { + "epoch": 1.5898406218104735, + "grad_norm": 4.486666202545166, + "learning_rate": 3.832193989197463e-06, + "loss": 5.7962939453125, + "step": 50625 + }, + { + "epoch": 1.5906257360445943, + "grad_norm": 3.2979328632354736, + "learning_rate": 3.825434333644937e-06, + "loss": 5.8048828125, + "step": 50650 + }, + { + "epoch": 1.5914108502787156, + "grad_norm": 3.7277791500091553, + "learning_rate": 3.818685763038756e-06, + "loss": 5.79448486328125, + "step": 50675 + }, + { + "epoch": 1.5921959645128365, + "grad_norm": 4.430881023406982, + "learning_rate": 3.811948287805691e-06, + "loss": 5.796631469726562, + "step": 50700 + }, + { + "epoch": 1.5929810787469578, + "grad_norm": 3.419097423553467, + "learning_rate": 3.805221918355372e-06, + "loss": 5.805740966796875, + "step": 50725 + }, + { + "epoch": 1.5937661929810787, + "grad_norm": 5.074674129486084, + "learning_rate": 3.798506665080263e-06, + "loss": 5.801837768554687, + "step": 50750 + }, + { + "epoch": 1.5945513072151998, + "grad_norm": 2.712246894836426, + "learning_rate": 3.7918025383556576e-06, + "loss": 5.797415771484375, + "step": 50775 + }, + { + "epoch": 1.5953364214493209, + "grad_norm": 3.161252498626709, + "learning_rate": 3.785109548539663e-06, + "loss": 5.8008349609375, + "step": 50800 + }, + { + "epoch": 1.596121535683442, + "grad_norm": 3.076082229614258, + "learning_rate": 3.778427705973169e-06, + "loss": 5.803912353515625, + "step": 50825 + }, + { + "epoch": 1.596906649917563, + "grad_norm": 4.224883556365967, + "learning_rate": 3.7717570209798515e-06, + "loss": 5.803405151367188, + "step": 50850 + }, + { + "epoch": 1.5976917641516841, + "grad_norm": 5.6546478271484375, + "learning_rate": 3.765097503866141e-06, + "loss": 5.794392700195313, + "step": 50875 + }, + { + "epoch": 1.5984768783858052, + "grad_norm": 6.64470911026001, + "learning_rate": 3.7584491649212186e-06, + "loss": 5.80628662109375, + "step": 50900 + }, + { + "epoch": 1.599261992619926, + "grad_norm": 2.5986669063568115, + "learning_rate": 3.7518120144169943e-06, + "loss": 5.799351806640625, + "step": 50925 + }, + { + "epoch": 1.6000471068540474, + "grad_norm": 4.745136260986328, + "learning_rate": 3.7451860626080847e-06, + "loss": 5.796547241210938, + "step": 50950 + }, + { + "epoch": 1.6008322210881683, + "grad_norm": 3.961850881576538, + "learning_rate": 3.738571319731813e-06, + "loss": 5.774854125976563, + "step": 50975 + }, + { + "epoch": 1.6016173353222896, + "grad_norm": 5.10260009765625, + "learning_rate": 3.7319677960081824e-06, + "loss": 5.788701782226562, + "step": 51000 + }, + { + "epoch": 1.6024024495564104, + "grad_norm": 4.172293186187744, + "learning_rate": 3.725375501639854e-06, + "loss": 5.7841845703125, + "step": 51025 + }, + { + "epoch": 1.6031875637905315, + "grad_norm": 6.4775309562683105, + "learning_rate": 3.7187944468121515e-06, + "loss": 5.7960357666015625, + "step": 51050 + }, + { + "epoch": 1.6039726780246526, + "grad_norm": 3.742861747741699, + "learning_rate": 3.7122246416930234e-06, + "loss": 5.805542602539062, + "step": 51075 + }, + { + "epoch": 1.6047577922587737, + "grad_norm": 5.948304176330566, + "learning_rate": 3.70566609643304e-06, + "loss": 5.798612670898438, + "step": 51100 + }, + { + "epoch": 1.6055429064928948, + "grad_norm": 3.3397841453552246, + "learning_rate": 3.6991188211653796e-06, + "loss": 5.800460205078125, + "step": 51125 + }, + { + "epoch": 1.6063280207270156, + "grad_norm": 2.6725990772247314, + "learning_rate": 3.692582826005798e-06, + "loss": 5.796121826171875, + "step": 51150 + }, + { + "epoch": 1.607113134961137, + "grad_norm": 4.699281215667725, + "learning_rate": 3.6860581210526327e-06, + "loss": 5.804586181640625, + "step": 51175 + }, + { + "epoch": 1.6078982491952578, + "grad_norm": 3.4811160564422607, + "learning_rate": 3.6795447163867727e-06, + "loss": 5.79287841796875, + "step": 51200 + }, + { + "epoch": 1.6086833634293791, + "grad_norm": 3.9069652557373047, + "learning_rate": 3.6730426220716455e-06, + "loss": 5.807425537109375, + "step": 51225 + }, + { + "epoch": 1.6094684776635, + "grad_norm": 3.905623197555542, + "learning_rate": 3.6665518481532113e-06, + "loss": 5.801078491210937, + "step": 51250 + }, + { + "epoch": 1.610253591897621, + "grad_norm": 3.5513505935668945, + "learning_rate": 3.6600724046599278e-06, + "loss": 5.800455932617187, + "step": 51275 + }, + { + "epoch": 1.6110387061317422, + "grad_norm": 4.965606212615967, + "learning_rate": 3.6536043016027633e-06, + "loss": 5.804674072265625, + "step": 51300 + }, + { + "epoch": 1.6118238203658632, + "grad_norm": 3.2908143997192383, + "learning_rate": 3.6471475489751495e-06, + "loss": 5.7995538330078125, + "step": 51325 + }, + { + "epoch": 1.6126089345999843, + "grad_norm": 2.7769670486450195, + "learning_rate": 3.6407021567529876e-06, + "loss": 5.79237060546875, + "step": 51350 + }, + { + "epoch": 1.6133940488341052, + "grad_norm": 3.28472900390625, + "learning_rate": 3.634268134894627e-06, + "loss": 5.77873291015625, + "step": 51375 + }, + { + "epoch": 1.6141791630682265, + "grad_norm": 3.2067043781280518, + "learning_rate": 3.6278454933408503e-06, + "loss": 5.796770629882812, + "step": 51400 + }, + { + "epoch": 1.6149642773023474, + "grad_norm": 3.248952865600586, + "learning_rate": 3.6214342420148558e-06, + "loss": 5.795384521484375, + "step": 51425 + }, + { + "epoch": 1.6157493915364687, + "grad_norm": 2.877992868423462, + "learning_rate": 3.615034390822243e-06, + "loss": 5.80084716796875, + "step": 51450 + }, + { + "epoch": 1.6165345057705895, + "grad_norm": 2.805896282196045, + "learning_rate": 3.6086459496509924e-06, + "loss": 5.79056640625, + "step": 51475 + }, + { + "epoch": 1.6173196200047106, + "grad_norm": 2.6245462894439697, + "learning_rate": 3.602268928371472e-06, + "loss": 5.797354736328125, + "step": 51500 + }, + { + "epoch": 1.6181047342388317, + "grad_norm": 4.103006362915039, + "learning_rate": 3.595903336836386e-06, + "loss": 5.801563110351562, + "step": 51525 + }, + { + "epoch": 1.6188898484729528, + "grad_norm": 2.5940170288085938, + "learning_rate": 3.589549184880795e-06, + "loss": 5.802990112304688, + "step": 51550 + }, + { + "epoch": 1.619674962707074, + "grad_norm": 2.9274685382843018, + "learning_rate": 3.5832064823220735e-06, + "loss": 5.781011962890625, + "step": 51575 + }, + { + "epoch": 1.620460076941195, + "grad_norm": 3.8081278800964355, + "learning_rate": 3.5768752389599126e-06, + "loss": 5.80527587890625, + "step": 51600 + }, + { + "epoch": 1.621245191175316, + "grad_norm": 4.534651756286621, + "learning_rate": 3.570555464576301e-06, + "loss": 5.7780615234375, + "step": 51625 + }, + { + "epoch": 1.622030305409437, + "grad_norm": 4.019165992736816, + "learning_rate": 3.5642471689354997e-06, + "loss": 5.787950439453125, + "step": 51650 + }, + { + "epoch": 1.6228154196435582, + "grad_norm": 2.739419937133789, + "learning_rate": 3.5579503617840348e-06, + "loss": 5.802354125976563, + "step": 51675 + }, + { + "epoch": 1.623600533877679, + "grad_norm": 5.103831768035889, + "learning_rate": 3.5516650528506956e-06, + "loss": 5.792191162109375, + "step": 51700 + }, + { + "epoch": 1.6243856481118004, + "grad_norm": 3.2540924549102783, + "learning_rate": 3.5453912518464897e-06, + "loss": 5.795818481445313, + "step": 51725 + }, + { + "epoch": 1.6251707623459213, + "grad_norm": 4.4360504150390625, + "learning_rate": 3.5391289684646558e-06, + "loss": 5.810909423828125, + "step": 51750 + }, + { + "epoch": 1.6259558765800424, + "grad_norm": 4.132941246032715, + "learning_rate": 3.532878212380628e-06, + "loss": 5.813388061523438, + "step": 51775 + }, + { + "epoch": 1.6267409908141635, + "grad_norm": 4.491993427276611, + "learning_rate": 3.5266389932520395e-06, + "loss": 5.7979052734375, + "step": 51800 + }, + { + "epoch": 1.6275261050482845, + "grad_norm": 3.7917466163635254, + "learning_rate": 3.5204113207186975e-06, + "loss": 5.800252075195313, + "step": 51825 + }, + { + "epoch": 1.6283112192824056, + "grad_norm": 3.878314256668091, + "learning_rate": 3.514195204402559e-06, + "loss": 5.78463134765625, + "step": 51850 + }, + { + "epoch": 1.6290963335165265, + "grad_norm": 3.469482660293579, + "learning_rate": 3.5079906539077385e-06, + "loss": 5.784945068359375, + "step": 51875 + }, + { + "epoch": 1.6298814477506478, + "grad_norm": 3.7619130611419678, + "learning_rate": 3.5017976788204775e-06, + "loss": 5.8057318115234375, + "step": 51900 + }, + { + "epoch": 1.6306665619847687, + "grad_norm": 2.8409385681152344, + "learning_rate": 3.4956162887091273e-06, + "loss": 5.798825073242187, + "step": 51925 + }, + { + "epoch": 1.63145167621889, + "grad_norm": 3.999506950378418, + "learning_rate": 3.489446493124148e-06, + "loss": 5.7862176513671875, + "step": 51950 + }, + { + "epoch": 1.6322367904530108, + "grad_norm": 4.165113925933838, + "learning_rate": 3.4832883015980794e-06, + "loss": 5.790623779296875, + "step": 51975 + }, + { + "epoch": 1.633021904687132, + "grad_norm": 4.279726028442383, + "learning_rate": 3.4771417236455373e-06, + "loss": 5.797593994140625, + "step": 52000 + }, + { + "epoch": 1.633807018921253, + "grad_norm": 4.084758758544922, + "learning_rate": 3.4710067687631943e-06, + "loss": 5.788948974609375, + "step": 52025 + }, + { + "epoch": 1.634592133155374, + "grad_norm": 5.0263566970825195, + "learning_rate": 3.4648834464297578e-06, + "loss": 5.786517333984375, + "step": 52050 + }, + { + "epoch": 1.6353772473894952, + "grad_norm": 2.74605393409729, + "learning_rate": 3.4587717661059707e-06, + "loss": 5.804745483398437, + "step": 52075 + }, + { + "epoch": 1.6361623616236163, + "grad_norm": 3.7980713844299316, + "learning_rate": 3.452671737234588e-06, + "loss": 5.792777709960937, + "step": 52100 + }, + { + "epoch": 1.6369474758577374, + "grad_norm": 1.8956425189971924, + "learning_rate": 3.446583369240355e-06, + "loss": 5.785962524414063, + "step": 52125 + }, + { + "epoch": 1.6377325900918582, + "grad_norm": 5.045468807220459, + "learning_rate": 3.440506671530012e-06, + "loss": 5.798984985351563, + "step": 52150 + }, + { + "epoch": 1.6385177043259795, + "grad_norm": 4.237487316131592, + "learning_rate": 3.434441653492255e-06, + "loss": 5.7989739990234375, + "step": 52175 + }, + { + "epoch": 1.6393028185601004, + "grad_norm": 4.43571138381958, + "learning_rate": 3.428388324497748e-06, + "loss": 5.787532958984375, + "step": 52200 + }, + { + "epoch": 1.6400879327942217, + "grad_norm": 4.360158920288086, + "learning_rate": 3.4223466938990866e-06, + "loss": 5.80435302734375, + "step": 52225 + }, + { + "epoch": 1.6408730470283426, + "grad_norm": 3.1161060333251953, + "learning_rate": 3.4163167710307932e-06, + "loss": 5.798643798828125, + "step": 52250 + }, + { + "epoch": 1.6416581612624637, + "grad_norm": 3.3289191722869873, + "learning_rate": 3.4102985652093035e-06, + "loss": 5.802236938476563, + "step": 52275 + }, + { + "epoch": 1.6424432754965848, + "grad_norm": 5.154234409332275, + "learning_rate": 3.404292085732951e-06, + "loss": 5.796903076171875, + "step": 52300 + }, + { + "epoch": 1.6432283897307058, + "grad_norm": 3.691981792449951, + "learning_rate": 3.3982973418819444e-06, + "loss": 5.791055297851562, + "step": 52325 + }, + { + "epoch": 1.644013503964827, + "grad_norm": 4.618564128875732, + "learning_rate": 3.392314342918371e-06, + "loss": 5.781434936523437, + "step": 52350 + }, + { + "epoch": 1.6447986181989478, + "grad_norm": 3.254326820373535, + "learning_rate": 3.386343098086159e-06, + "loss": 5.807782592773438, + "step": 52375 + }, + { + "epoch": 1.645583732433069, + "grad_norm": 4.9267497062683105, + "learning_rate": 3.3803836166110916e-06, + "loss": 5.805394897460937, + "step": 52400 + }, + { + "epoch": 1.64636884666719, + "grad_norm": 4.787402629852295, + "learning_rate": 3.3744359077007653e-06, + "loss": 5.7934716796875, + "step": 52425 + }, + { + "epoch": 1.6471539609013113, + "grad_norm": 4.1648030281066895, + "learning_rate": 3.3684999805445896e-06, + "loss": 5.799417114257812, + "step": 52450 + }, + { + "epoch": 1.6479390751354321, + "grad_norm": 3.340914011001587, + "learning_rate": 3.362575844313774e-06, + "loss": 5.796717166900635, + "step": 52475 + }, + { + "epoch": 1.6487241893695532, + "grad_norm": 3.3443055152893066, + "learning_rate": 3.3566635081613095e-06, + "loss": 5.79302734375, + "step": 52500 + }, + { + "epoch": 1.6495093036036743, + "grad_norm": 2.5274951457977295, + "learning_rate": 3.3507629812219565e-06, + "loss": 5.7960205078125, + "step": 52525 + }, + { + "epoch": 1.6502944178377954, + "grad_norm": 4.929781913757324, + "learning_rate": 3.3448742726122293e-06, + "loss": 5.79507568359375, + "step": 52550 + }, + { + "epoch": 1.6510795320719165, + "grad_norm": 3.5698354244232178, + "learning_rate": 3.338997391430373e-06, + "loss": 5.78392333984375, + "step": 52575 + }, + { + "epoch": 1.6518646463060376, + "grad_norm": 2.812836170196533, + "learning_rate": 3.3331323467563794e-06, + "loss": 5.793589477539062, + "step": 52600 + }, + { + "epoch": 1.6526497605401587, + "grad_norm": 1.9206541776657104, + "learning_rate": 3.3272791476519343e-06, + "loss": 5.804886474609375, + "step": 52625 + }, + { + "epoch": 1.6534348747742795, + "grad_norm": 3.4247820377349854, + "learning_rate": 3.3214378031604273e-06, + "loss": 5.778856201171875, + "step": 52650 + }, + { + "epoch": 1.6542199890084008, + "grad_norm": 3.3723719120025635, + "learning_rate": 3.3156083223069346e-06, + "loss": 5.799539184570312, + "step": 52675 + }, + { + "epoch": 1.6550051032425217, + "grad_norm": 3.918879508972168, + "learning_rate": 3.3097907140982017e-06, + "loss": 5.786864013671875, + "step": 52700 + }, + { + "epoch": 1.655790217476643, + "grad_norm": 4.236285209655762, + "learning_rate": 3.3039849875226302e-06, + "loss": 5.79086181640625, + "step": 52725 + }, + { + "epoch": 1.6565753317107639, + "grad_norm": 3.7409722805023193, + "learning_rate": 3.298191151550264e-06, + "loss": 5.782718505859375, + "step": 52750 + }, + { + "epoch": 1.657360445944885, + "grad_norm": 4.355134963989258, + "learning_rate": 3.2924092151327704e-06, + "loss": 5.794862670898437, + "step": 52775 + }, + { + "epoch": 1.658145560179006, + "grad_norm": 4.211248397827148, + "learning_rate": 3.2866391872034433e-06, + "loss": 5.796520385742188, + "step": 52800 + }, + { + "epoch": 1.6589306744131271, + "grad_norm": 4.470012187957764, + "learning_rate": 3.280881076677167e-06, + "loss": 5.790586547851563, + "step": 52825 + }, + { + "epoch": 1.6597157886472482, + "grad_norm": 3.8508198261260986, + "learning_rate": 3.275134892450419e-06, + "loss": 5.78813720703125, + "step": 52850 + }, + { + "epoch": 1.660500902881369, + "grad_norm": 4.225559711456299, + "learning_rate": 3.269400643401245e-06, + "loss": 5.787883911132813, + "step": 52875 + }, + { + "epoch": 1.6612860171154904, + "grad_norm": 3.799661159515381, + "learning_rate": 3.2636783383892554e-06, + "loss": 5.798718872070313, + "step": 52900 + }, + { + "epoch": 1.6620711313496113, + "grad_norm": 3.7599282264709473, + "learning_rate": 3.257967986255606e-06, + "loss": 5.791277465820312, + "step": 52925 + }, + { + "epoch": 1.6628562455837326, + "grad_norm": 4.519002437591553, + "learning_rate": 3.252269595822982e-06, + "loss": 5.798148193359375, + "step": 52950 + }, + { + "epoch": 1.6636413598178534, + "grad_norm": 3.1788482666015625, + "learning_rate": 3.246583175895588e-06, + "loss": 5.792970581054687, + "step": 52975 + }, + { + "epoch": 1.6644264740519745, + "grad_norm": 3.5583693981170654, + "learning_rate": 3.2409087352591393e-06, + "loss": 5.784022827148437, + "step": 53000 + }, + { + "epoch": 1.6652115882860956, + "grad_norm": 3.9589362144470215, + "learning_rate": 3.235246282680834e-06, + "loss": 5.779111328125, + "step": 53025 + }, + { + "epoch": 1.6659967025202167, + "grad_norm": 4.636691093444824, + "learning_rate": 3.2295958269093554e-06, + "loss": 5.8014373779296875, + "step": 53050 + }, + { + "epoch": 1.6667818167543378, + "grad_norm": 3.0693392753601074, + "learning_rate": 3.223957376674845e-06, + "loss": 5.807294921875, + "step": 53075 + }, + { + "epoch": 1.6675669309884589, + "grad_norm": 3.6302435398101807, + "learning_rate": 3.218330940688902e-06, + "loss": 5.790675048828125, + "step": 53100 + }, + { + "epoch": 1.66835204522258, + "grad_norm": 4.392565727233887, + "learning_rate": 3.2127165276445617e-06, + "loss": 5.787136840820312, + "step": 53125 + }, + { + "epoch": 1.6691371594567008, + "grad_norm": 3.778292655944824, + "learning_rate": 3.2071141462162768e-06, + "loss": 5.790072631835938, + "step": 53150 + }, + { + "epoch": 1.6699222736908221, + "grad_norm": 4.017690658569336, + "learning_rate": 3.2015238050599196e-06, + "loss": 5.795682373046875, + "step": 53175 + }, + { + "epoch": 1.670707387924943, + "grad_norm": 3.844534158706665, + "learning_rate": 3.195945512812757e-06, + "loss": 5.8134228515625, + "step": 53200 + }, + { + "epoch": 1.6714925021590643, + "grad_norm": 4.966689586639404, + "learning_rate": 3.1903792780934345e-06, + "loss": 5.797835083007812, + "step": 53225 + }, + { + "epoch": 1.6722776163931852, + "grad_norm": 3.36676025390625, + "learning_rate": 3.1848251095019783e-06, + "loss": 5.789369506835937, + "step": 53250 + }, + { + "epoch": 1.6730627306273063, + "grad_norm": 3.6670477390289307, + "learning_rate": 3.1792830156197614e-06, + "loss": 5.788646850585938, + "step": 53275 + }, + { + "epoch": 1.6738478448614273, + "grad_norm": 3.4687070846557617, + "learning_rate": 3.173753005009507e-06, + "loss": 5.78541259765625, + "step": 53300 + }, + { + "epoch": 1.6746329590955484, + "grad_norm": 3.2948906421661377, + "learning_rate": 3.1682350862152726e-06, + "loss": 5.797109985351563, + "step": 53325 + }, + { + "epoch": 1.6754180733296695, + "grad_norm": 5.167768478393555, + "learning_rate": 3.1627292677624256e-06, + "loss": 5.79853271484375, + "step": 53350 + }, + { + "epoch": 1.6762031875637904, + "grad_norm": 3.2556800842285156, + "learning_rate": 3.157235558157641e-06, + "loss": 5.798134155273438, + "step": 53375 + }, + { + "epoch": 1.6769883017979117, + "grad_norm": 2.5559847354888916, + "learning_rate": 3.151753965888888e-06, + "loss": 5.7810009765625, + "step": 53400 + }, + { + "epoch": 1.6777734160320326, + "grad_norm": 6.909741401672363, + "learning_rate": 3.1462844994254105e-06, + "loss": 5.778453826904297, + "step": 53425 + }, + { + "epoch": 1.6785585302661539, + "grad_norm": 4.012985706329346, + "learning_rate": 3.1408271672177193e-06, + "loss": 5.784638061523437, + "step": 53450 + }, + { + "epoch": 1.6793436445002747, + "grad_norm": 3.675302267074585, + "learning_rate": 3.1353819776975742e-06, + "loss": 5.77371337890625, + "step": 53475 + }, + { + "epoch": 1.6801287587343958, + "grad_norm": 4.200839519500732, + "learning_rate": 3.1299489392779836e-06, + "loss": 5.781553955078125, + "step": 53500 + }, + { + "epoch": 1.680913872968517, + "grad_norm": 5.0064568519592285, + "learning_rate": 3.1245280603531713e-06, + "loss": 5.800042724609375, + "step": 53525 + }, + { + "epoch": 1.681698987202638, + "grad_norm": 4.023277282714844, + "learning_rate": 3.119119349298577e-06, + "loss": 5.779158935546875, + "step": 53550 + }, + { + "epoch": 1.682484101436759, + "grad_norm": 4.6343159675598145, + "learning_rate": 3.1137228144708435e-06, + "loss": 5.787694702148437, + "step": 53575 + }, + { + "epoch": 1.68326921567088, + "grad_norm": 4.726887226104736, + "learning_rate": 3.108338464207802e-06, + "loss": 5.789046630859375, + "step": 53600 + }, + { + "epoch": 1.6840543299050013, + "grad_norm": 3.3134701251983643, + "learning_rate": 3.1029663068284505e-06, + "loss": 5.788505859375, + "step": 53625 + }, + { + "epoch": 1.6848394441391221, + "grad_norm": 4.763998508453369, + "learning_rate": 3.0976063506329582e-06, + "loss": 5.80382080078125, + "step": 53650 + }, + { + "epoch": 1.6856245583732434, + "grad_norm": 2.619739532470703, + "learning_rate": 3.092258603902633e-06, + "loss": 5.789880981445313, + "step": 53675 + }, + { + "epoch": 1.6864096726073643, + "grad_norm": 6.013173580169678, + "learning_rate": 3.086923074899931e-06, + "loss": 5.78770263671875, + "step": 53700 + }, + { + "epoch": 1.6871947868414854, + "grad_norm": 3.612316370010376, + "learning_rate": 3.081599771868423e-06, + "loss": 5.7882421875, + "step": 53725 + }, + { + "epoch": 1.6879799010756065, + "grad_norm": 3.328181266784668, + "learning_rate": 3.076288703032788e-06, + "loss": 5.797979125976562, + "step": 53750 + }, + { + "epoch": 1.6887650153097276, + "grad_norm": 3.0525639057159424, + "learning_rate": 3.07098987659881e-06, + "loss": 5.787008056640625, + "step": 53775 + }, + { + "epoch": 1.6895501295438486, + "grad_norm": 2.8377997875213623, + "learning_rate": 3.065703300753353e-06, + "loss": 5.784354248046875, + "step": 53800 + }, + { + "epoch": 1.6903352437779697, + "grad_norm": 4.609061241149902, + "learning_rate": 3.0604289836643603e-06, + "loss": 5.783494262695313, + "step": 53825 + }, + { + "epoch": 1.6911203580120908, + "grad_norm": 3.664961814880371, + "learning_rate": 3.055166933480825e-06, + "loss": 5.795774536132813, + "step": 53850 + }, + { + "epoch": 1.6919054722462117, + "grad_norm": 2.870007038116455, + "learning_rate": 3.0499171583327908e-06, + "loss": 5.8123095703125, + "step": 53875 + }, + { + "epoch": 1.692690586480333, + "grad_norm": 3.254917860031128, + "learning_rate": 3.044679666331343e-06, + "loss": 5.797506713867188, + "step": 53900 + }, + { + "epoch": 1.6934757007144539, + "grad_norm": 4.540425777435303, + "learning_rate": 3.0394544655685804e-06, + "loss": 5.787763671875, + "step": 53925 + }, + { + "epoch": 1.6942608149485752, + "grad_norm": 3.2745890617370605, + "learning_rate": 3.0342415641176174e-06, + "loss": 5.796687622070312, + "step": 53950 + }, + { + "epoch": 1.695045929182696, + "grad_norm": 4.555976867675781, + "learning_rate": 3.0290409700325594e-06, + "loss": 5.791295166015625, + "step": 53975 + }, + { + "epoch": 1.6958310434168171, + "grad_norm": 3.2083778381347656, + "learning_rate": 3.023852691348501e-06, + "loss": 5.788763427734375, + "step": 54000 + }, + { + "epoch": 1.6966161576509382, + "grad_norm": 3.855635404586792, + "learning_rate": 3.0186767360815107e-06, + "loss": 5.793141479492188, + "step": 54025 + }, + { + "epoch": 1.6974012718850593, + "grad_norm": 8.728151321411133, + "learning_rate": 3.013513112228611e-06, + "loss": 5.802308349609375, + "step": 54050 + }, + { + "epoch": 1.6981863861191804, + "grad_norm": 4.769170761108398, + "learning_rate": 3.0083618277677756e-06, + "loss": 5.7904638671875, + "step": 54075 + }, + { + "epoch": 1.6989715003533012, + "grad_norm": 3.8085601329803467, + "learning_rate": 3.003222890657916e-06, + "loss": 5.77656005859375, + "step": 54100 + }, + { + "epoch": 1.6997566145874226, + "grad_norm": 4.54194974899292, + "learning_rate": 2.9980963088388602e-06, + "loss": 5.801517333984375, + "step": 54125 + }, + { + "epoch": 1.7005417288215434, + "grad_norm": 3.703043222427368, + "learning_rate": 2.9929820902313527e-06, + "loss": 5.7862310791015625, + "step": 54150 + }, + { + "epoch": 1.7013268430556647, + "grad_norm": 4.9394073486328125, + "learning_rate": 2.9878802427370317e-06, + "loss": 5.786273803710937, + "step": 54175 + }, + { + "epoch": 1.7021119572897856, + "grad_norm": 2.9924726486206055, + "learning_rate": 2.9827907742384244e-06, + "loss": 5.78591796875, + "step": 54200 + }, + { + "epoch": 1.7028970715239067, + "grad_norm": 3.799065589904785, + "learning_rate": 2.9777136925989353e-06, + "loss": 5.802099609375, + "step": 54225 + }, + { + "epoch": 1.7036821857580278, + "grad_norm": 4.732241630554199, + "learning_rate": 2.9726490056628223e-06, + "loss": 5.807063598632812, + "step": 54250 + }, + { + "epoch": 1.7044672999921489, + "grad_norm": 3.383070945739746, + "learning_rate": 2.9675967212551975e-06, + "loss": 5.791766357421875, + "step": 54275 + }, + { + "epoch": 1.70525241422627, + "grad_norm": 5.5677103996276855, + "learning_rate": 2.9625568471820177e-06, + "loss": 5.80056640625, + "step": 54300 + }, + { + "epoch": 1.706037528460391, + "grad_norm": 4.111900329589844, + "learning_rate": 2.957529391230051e-06, + "loss": 5.797904052734375, + "step": 54325 + }, + { + "epoch": 1.7068226426945121, + "grad_norm": 3.7623276710510254, + "learning_rate": 2.952514361166891e-06, + "loss": 5.795888061523438, + "step": 54350 + }, + { + "epoch": 1.707607756928633, + "grad_norm": 3.2053778171539307, + "learning_rate": 2.9475117647409265e-06, + "loss": 5.7895574951171875, + "step": 54375 + }, + { + "epoch": 1.7083928711627543, + "grad_norm": 3.157865524291992, + "learning_rate": 2.942521609681338e-06, + "loss": 5.779329833984375, + "step": 54400 + }, + { + "epoch": 1.7091779853968752, + "grad_norm": 7.712612152099609, + "learning_rate": 2.937543903698087e-06, + "loss": 5.801334838867188, + "step": 54425 + }, + { + "epoch": 1.7099630996309965, + "grad_norm": 5.730584621429443, + "learning_rate": 2.932578654481894e-06, + "loss": 5.78336181640625, + "step": 54450 + }, + { + "epoch": 1.7107482138651173, + "grad_norm": 3.8324596881866455, + "learning_rate": 2.9276258697042358e-06, + "loss": 5.791301879882813, + "step": 54475 + }, + { + "epoch": 1.7115333280992384, + "grad_norm": 2.9169509410858154, + "learning_rate": 2.9226855570173375e-06, + "loss": 5.794508056640625, + "step": 54500 + }, + { + "epoch": 1.7123184423333595, + "grad_norm": 4.839617729187012, + "learning_rate": 2.9177577240541426e-06, + "loss": 5.799713745117187, + "step": 54525 + }, + { + "epoch": 1.7131035565674806, + "grad_norm": 4.52235746383667, + "learning_rate": 2.9128423784283256e-06, + "loss": 5.785316162109375, + "step": 54550 + }, + { + "epoch": 1.7138886708016017, + "grad_norm": 3.603635311126709, + "learning_rate": 2.907939527734255e-06, + "loss": 5.788912963867188, + "step": 54575 + }, + { + "epoch": 1.7146737850357225, + "grad_norm": 3.5103304386138916, + "learning_rate": 2.9030491795470077e-06, + "loss": 5.8051708984375, + "step": 54600 + }, + { + "epoch": 1.7154588992698439, + "grad_norm": 3.333629608154297, + "learning_rate": 2.8981713414223333e-06, + "loss": 5.78905029296875, + "step": 54625 + }, + { + "epoch": 1.7162440135039647, + "grad_norm": 3.6196796894073486, + "learning_rate": 2.8933060208966544e-06, + "loss": 5.784867553710938, + "step": 54650 + }, + { + "epoch": 1.717029127738086, + "grad_norm": 4.840312480926514, + "learning_rate": 2.8884532254870588e-06, + "loss": 5.80423828125, + "step": 54675 + }, + { + "epoch": 1.717814241972207, + "grad_norm": 4.034519195556641, + "learning_rate": 2.88361296269128e-06, + "loss": 5.803225708007813, + "step": 54700 + }, + { + "epoch": 1.718599356206328, + "grad_norm": 3.2703158855438232, + "learning_rate": 2.8787852399876847e-06, + "loss": 5.79246337890625, + "step": 54725 + }, + { + "epoch": 1.719384470440449, + "grad_norm": 3.8150782585144043, + "learning_rate": 2.87397006483527e-06, + "loss": 5.78220947265625, + "step": 54750 + }, + { + "epoch": 1.7201695846745702, + "grad_norm": 3.3181285858154297, + "learning_rate": 2.8691674446736394e-06, + "loss": 5.801692504882812, + "step": 54775 + }, + { + "epoch": 1.7209546989086912, + "grad_norm": 3.8962817192077637, + "learning_rate": 2.8643773869230118e-06, + "loss": 5.786429443359375, + "step": 54800 + }, + { + "epoch": 1.7217398131428123, + "grad_norm": 3.37101674079895, + "learning_rate": 2.8595998989841828e-06, + "loss": 5.792263793945312, + "step": 54825 + }, + { + "epoch": 1.7225249273769334, + "grad_norm": 3.432154893875122, + "learning_rate": 2.8548349882385313e-06, + "loss": 5.8006829833984375, + "step": 54850 + }, + { + "epoch": 1.7233100416110543, + "grad_norm": 3.4587669372558594, + "learning_rate": 2.850082662048007e-06, + "loss": 5.793782958984375, + "step": 54875 + }, + { + "epoch": 1.7240951558451756, + "grad_norm": 3.298506021499634, + "learning_rate": 2.845342927755116e-06, + "loss": 5.782908325195312, + "step": 54900 + }, + { + "epoch": 1.7248802700792965, + "grad_norm": 3.2372870445251465, + "learning_rate": 2.840615792682908e-06, + "loss": 5.788709716796875, + "step": 54925 + }, + { + "epoch": 1.7256653843134178, + "grad_norm": 3.5261876583099365, + "learning_rate": 2.8359012641349647e-06, + "loss": 5.789381713867187, + "step": 54950 + }, + { + "epoch": 1.7264504985475386, + "grad_norm": 3.829169988632202, + "learning_rate": 2.83119934939539e-06, + "loss": 5.773970947265625, + "step": 54975 + }, + { + "epoch": 1.7272356127816597, + "grad_norm": 3.3297226428985596, + "learning_rate": 2.826510055728806e-06, + "loss": 5.782086181640625, + "step": 55000 + }, + { + "epoch": 1.7280207270157808, + "grad_norm": 2.840602159500122, + "learning_rate": 2.8218333903803263e-06, + "loss": 5.788776245117187, + "step": 55025 + }, + { + "epoch": 1.7288058412499019, + "grad_norm": 4.069692611694336, + "learning_rate": 2.817169360575559e-06, + "loss": 5.785874633789063, + "step": 55050 + }, + { + "epoch": 1.729590955484023, + "grad_norm": 4.161444664001465, + "learning_rate": 2.8125179735205852e-06, + "loss": 5.78714599609375, + "step": 55075 + }, + { + "epoch": 1.7303760697181438, + "grad_norm": 3.645965576171875, + "learning_rate": 2.8078792364019564e-06, + "loss": 5.798280639648437, + "step": 55100 + }, + { + "epoch": 1.7311611839522651, + "grad_norm": 3.9164159297943115, + "learning_rate": 2.80325315638668e-06, + "loss": 5.802951049804688, + "step": 55125 + }, + { + "epoch": 1.731946298186386, + "grad_norm": 2.9530913829803467, + "learning_rate": 2.798639740622202e-06, + "loss": 5.794804077148438, + "step": 55150 + }, + { + "epoch": 1.7327314124205073, + "grad_norm": 4.062150478363037, + "learning_rate": 2.7940389962364084e-06, + "loss": 5.786390380859375, + "step": 55175 + }, + { + "epoch": 1.7335165266546282, + "grad_norm": 2.291429042816162, + "learning_rate": 2.7894509303376055e-06, + "loss": 5.799359130859375, + "step": 55200 + }, + { + "epoch": 1.7343016408887493, + "grad_norm": 4.043982028961182, + "learning_rate": 2.7848755500145065e-06, + "loss": 5.7982861328125, + "step": 55225 + }, + { + "epoch": 1.7350867551228704, + "grad_norm": 3.8498106002807617, + "learning_rate": 2.7803128623362337e-06, + "loss": 5.785423583984375, + "step": 55250 + }, + { + "epoch": 1.7358718693569914, + "grad_norm": 4.7921037673950195, + "learning_rate": 2.775762874352288e-06, + "loss": 5.788812255859375, + "step": 55275 + }, + { + "epoch": 1.7366569835911125, + "grad_norm": 4.145835876464844, + "learning_rate": 2.771225593092557e-06, + "loss": 5.7924566650390625, + "step": 55300 + }, + { + "epoch": 1.7374420978252336, + "grad_norm": 3.3712410926818848, + "learning_rate": 2.7667010255672964e-06, + "loss": 5.783648071289062, + "step": 55325 + }, + { + "epoch": 1.7382272120593547, + "grad_norm": 3.053072929382324, + "learning_rate": 2.762189178767112e-06, + "loss": 5.7918023681640625, + "step": 55350 + }, + { + "epoch": 1.7390123262934756, + "grad_norm": 2.815648317337036, + "learning_rate": 2.7576900596629595e-06, + "loss": 5.786428833007813, + "step": 55375 + }, + { + "epoch": 1.7397974405275969, + "grad_norm": 2.670663833618164, + "learning_rate": 2.753203675206134e-06, + "loss": 5.793081665039063, + "step": 55400 + }, + { + "epoch": 1.7405825547617177, + "grad_norm": 4.056495666503906, + "learning_rate": 2.7487300323282474e-06, + "loss": 5.786751708984375, + "step": 55425 + }, + { + "epoch": 1.741367668995839, + "grad_norm": 2.9202921390533447, + "learning_rate": 2.7442691379412305e-06, + "loss": 5.77890380859375, + "step": 55450 + }, + { + "epoch": 1.74215278322996, + "grad_norm": 3.901839017868042, + "learning_rate": 2.7398209989373137e-06, + "loss": 5.791027221679688, + "step": 55475 + }, + { + "epoch": 1.742937897464081, + "grad_norm": 2.9990553855895996, + "learning_rate": 2.735385622189025e-06, + "loss": 5.80008056640625, + "step": 55500 + }, + { + "epoch": 1.743723011698202, + "grad_norm": 3.204742193222046, + "learning_rate": 2.730963014549171e-06, + "loss": 5.79429931640625, + "step": 55525 + }, + { + "epoch": 1.7445081259323232, + "grad_norm": 3.166318655014038, + "learning_rate": 2.726553182850827e-06, + "loss": 5.77638053894043, + "step": 55550 + }, + { + "epoch": 1.7452932401664443, + "grad_norm": 4.488119125366211, + "learning_rate": 2.7221561339073345e-06, + "loss": 5.79408447265625, + "step": 55575 + }, + { + "epoch": 1.7460783544005651, + "grad_norm": 3.983168601989746, + "learning_rate": 2.7177718745122837e-06, + "loss": 5.7857208251953125, + "step": 55600 + }, + { + "epoch": 1.7468634686346864, + "grad_norm": 3.380866289138794, + "learning_rate": 2.7134004114395e-06, + "loss": 5.789664306640625, + "step": 55625 + }, + { + "epoch": 1.7476485828688073, + "grad_norm": 4.034645080566406, + "learning_rate": 2.7090417514430446e-06, + "loss": 5.778468627929687, + "step": 55650 + }, + { + "epoch": 1.7484336971029286, + "grad_norm": 4.583098888397217, + "learning_rate": 2.7046959012571926e-06, + "loss": 5.798292846679687, + "step": 55675 + }, + { + "epoch": 1.7492188113370495, + "grad_norm": 3.6379969120025635, + "learning_rate": 2.7003628675964295e-06, + "loss": 5.792781982421875, + "step": 55700 + }, + { + "epoch": 1.7500039255711706, + "grad_norm": 2.7330336570739746, + "learning_rate": 2.6960426571554417e-06, + "loss": 5.795667724609375, + "step": 55725 + }, + { + "epoch": 1.7500353301405354, + "eval_loss": 0.14121590554714203, + "eval_runtime": 5073.9292, + "eval_samples_per_second": 4.273, + "eval_steps_per_second": 0.267, + "eval_wer": 3.1759796476670035, + "step": 55726 + }, + { + "epoch": 1.7507890398052917, + "grad_norm": 3.364391565322876, + "learning_rate": 2.6917352766090955e-06, + "loss": 5.788433227539063, + "step": 55750 + }, + { + "epoch": 1.7515741540394127, + "grad_norm": 3.355116844177246, + "learning_rate": 2.6874407326124423e-06, + "loss": 5.779723510742188, + "step": 55775 + }, + { + "epoch": 1.7523592682735338, + "grad_norm": 3.1245226860046387, + "learning_rate": 2.683159031800698e-06, + "loss": 5.78419189453125, + "step": 55800 + }, + { + "epoch": 1.753144382507655, + "grad_norm": 3.59114408493042, + "learning_rate": 2.678890180789233e-06, + "loss": 5.797167358398437, + "step": 55825 + }, + { + "epoch": 1.753929496741776, + "grad_norm": 2.549438238143921, + "learning_rate": 2.674634186173568e-06, + "loss": 5.786692504882812, + "step": 55850 + }, + { + "epoch": 1.7547146109758969, + "grad_norm": 2.533783435821533, + "learning_rate": 2.6703910545293565e-06, + "loss": 5.79389892578125, + "step": 55875 + }, + { + "epoch": 1.7554997252100182, + "grad_norm": 3.9065845012664795, + "learning_rate": 2.666160792412384e-06, + "loss": 5.768716430664062, + "step": 55900 + }, + { + "epoch": 1.756284839444139, + "grad_norm": 3.6107354164123535, + "learning_rate": 2.661943406358546e-06, + "loss": 5.789741821289063, + "step": 55925 + }, + { + "epoch": 1.7570699536782601, + "grad_norm": 4.3172478675842285, + "learning_rate": 2.657738902883844e-06, + "loss": 5.797694091796875, + "step": 55950 + }, + { + "epoch": 1.7578550679123812, + "grad_norm": 3.8097493648529053, + "learning_rate": 2.6535472884843814e-06, + "loss": 5.7917822265625, + "step": 55975 + }, + { + "epoch": 1.7586401821465023, + "grad_norm": 6.03124475479126, + "learning_rate": 2.6493685696363435e-06, + "loss": 5.779965209960937, + "step": 56000 + }, + { + "epoch": 1.7594252963806234, + "grad_norm": 4.280158996582031, + "learning_rate": 2.6452027527959915e-06, + "loss": 5.787221069335938, + "step": 56025 + }, + { + "epoch": 1.7602104106147445, + "grad_norm": 3.304903268814087, + "learning_rate": 2.6410498443996545e-06, + "loss": 5.79031982421875, + "step": 56050 + }, + { + "epoch": 1.7609955248488656, + "grad_norm": 3.691190481185913, + "learning_rate": 2.636909850863713e-06, + "loss": 5.78988525390625, + "step": 56075 + }, + { + "epoch": 1.7617806390829864, + "grad_norm": 2.9102067947387695, + "learning_rate": 2.6327827785846006e-06, + "loss": 5.784320068359375, + "step": 56100 + }, + { + "epoch": 1.7625657533171077, + "grad_norm": 4.923401832580566, + "learning_rate": 2.62866863393878e-06, + "loss": 5.79582763671875, + "step": 56125 + }, + { + "epoch": 1.7633508675512286, + "grad_norm": 2.782412528991699, + "learning_rate": 2.624567423282748e-06, + "loss": 5.776354370117187, + "step": 56150 + }, + { + "epoch": 1.76413598178535, + "grad_norm": 5.672576904296875, + "learning_rate": 2.620479152953009e-06, + "loss": 5.7943182373046875, + "step": 56175 + }, + { + "epoch": 1.7649210960194708, + "grad_norm": 5.128127098083496, + "learning_rate": 2.616403829266081e-06, + "loss": 5.792017822265625, + "step": 56200 + }, + { + "epoch": 1.7657062102535919, + "grad_norm": 4.13068962097168, + "learning_rate": 2.612341458518478e-06, + "loss": 5.793185424804688, + "step": 56225 + }, + { + "epoch": 1.766491324487713, + "grad_norm": 2.8265223503112793, + "learning_rate": 2.608292046986698e-06, + "loss": 5.784890747070312, + "step": 56250 + }, + { + "epoch": 1.767276438721834, + "grad_norm": 3.5440585613250732, + "learning_rate": 2.604255600927219e-06, + "loss": 5.779779052734375, + "step": 56275 + }, + { + "epoch": 1.7680615529559551, + "grad_norm": 7.149130344390869, + "learning_rate": 2.600232126576487e-06, + "loss": 5.789750366210938, + "step": 56300 + }, + { + "epoch": 1.768846667190076, + "grad_norm": 3.3122153282165527, + "learning_rate": 2.596221630150906e-06, + "loss": 5.777681274414062, + "step": 56325 + }, + { + "epoch": 1.7696317814241973, + "grad_norm": 5.084323406219482, + "learning_rate": 2.592224117846829e-06, + "loss": 5.80149169921875, + "step": 56350 + }, + { + "epoch": 1.7704168956583182, + "grad_norm": 3.9701216220855713, + "learning_rate": 2.5882395958405453e-06, + "loss": 5.803560180664062, + "step": 56375 + }, + { + "epoch": 1.7712020098924395, + "grad_norm": 4.529189586639404, + "learning_rate": 2.5842680702882766e-06, + "loss": 5.785661010742188, + "step": 56400 + }, + { + "epoch": 1.7719871241265603, + "grad_norm": 3.6762421131134033, + "learning_rate": 2.580309547326167e-06, + "loss": 5.792308959960938, + "step": 56425 + }, + { + "epoch": 1.7727722383606814, + "grad_norm": 4.787751197814941, + "learning_rate": 2.576364033070264e-06, + "loss": 5.782610473632812, + "step": 56450 + }, + { + "epoch": 1.7735573525948025, + "grad_norm": 4.09505033493042, + "learning_rate": 2.5724315336165253e-06, + "loss": 5.786259155273438, + "step": 56475 + }, + { + "epoch": 1.7743424668289236, + "grad_norm": 3.6521527767181396, + "learning_rate": 2.5685120550407904e-06, + "loss": 5.7947540283203125, + "step": 56500 + }, + { + "epoch": 1.7751275810630447, + "grad_norm": 2.441530704498291, + "learning_rate": 2.5646056033987883e-06, + "loss": 5.779439697265625, + "step": 56525 + }, + { + "epoch": 1.7759126952971658, + "grad_norm": 2.9152140617370605, + "learning_rate": 2.5607121847261195e-06, + "loss": 5.788770141601563, + "step": 56550 + }, + { + "epoch": 1.7766978095312869, + "grad_norm": 2.918607711791992, + "learning_rate": 2.5568318050382464e-06, + "loss": 5.787138671875, + "step": 56575 + }, + { + "epoch": 1.7774829237654077, + "grad_norm": 2.930283546447754, + "learning_rate": 2.552964470330487e-06, + "loss": 5.780997314453125, + "step": 56600 + }, + { + "epoch": 1.778268037999529, + "grad_norm": 3.4797418117523193, + "learning_rate": 2.5491101865780045e-06, + "loss": 5.785594482421875, + "step": 56625 + }, + { + "epoch": 1.77905315223365, + "grad_norm": 2.5417637825012207, + "learning_rate": 2.5452689597357965e-06, + "loss": 5.79296630859375, + "step": 56650 + }, + { + "epoch": 1.7798382664677712, + "grad_norm": 2.4895529747009277, + "learning_rate": 2.54144079573869e-06, + "loss": 5.800777587890625, + "step": 56675 + }, + { + "epoch": 1.780623380701892, + "grad_norm": 4.3347392082214355, + "learning_rate": 2.537625700501325e-06, + "loss": 5.782273559570313, + "step": 56700 + }, + { + "epoch": 1.7814084949360132, + "grad_norm": 3.424748182296753, + "learning_rate": 2.5338236799181556e-06, + "loss": 5.7925830078125, + "step": 56725 + }, + { + "epoch": 1.7821936091701343, + "grad_norm": 4.845651149749756, + "learning_rate": 2.53003473986343e-06, + "loss": 5.791981811523438, + "step": 56750 + }, + { + "epoch": 1.7829787234042553, + "grad_norm": 3.9231224060058594, + "learning_rate": 2.5262588861911885e-06, + "loss": 5.802079467773438, + "step": 56775 + }, + { + "epoch": 1.7837638376383764, + "grad_norm": 3.408733367919922, + "learning_rate": 2.522496124735253e-06, + "loss": 5.766301879882812, + "step": 56800 + }, + { + "epoch": 1.7845489518724973, + "grad_norm": 4.129566669464111, + "learning_rate": 2.5187464613092187e-06, + "loss": 5.795888671875, + "step": 56825 + }, + { + "epoch": 1.7853340661066186, + "grad_norm": 4.915196895599365, + "learning_rate": 2.5150099017064406e-06, + "loss": 5.777572021484375, + "step": 56850 + }, + { + "epoch": 1.7861191803407395, + "grad_norm": 3.2676990032196045, + "learning_rate": 2.5112864517000296e-06, + "loss": 5.777117919921875, + "step": 56875 + }, + { + "epoch": 1.7869042945748608, + "grad_norm": 5.32418966293335, + "learning_rate": 2.5075761170428412e-06, + "loss": 5.79288330078125, + "step": 56900 + }, + { + "epoch": 1.7876894088089816, + "grad_norm": 4.582352161407471, + "learning_rate": 2.503878903467469e-06, + "loss": 5.78538449605306, + "step": 56925 + }, + { + "epoch": 1.7884745230431027, + "grad_norm": 3.987060308456421, + "learning_rate": 2.500194816686234e-06, + "loss": 5.791098022460938, + "step": 56950 + }, + { + "epoch": 1.7892596372772238, + "grad_norm": 3.131958246231079, + "learning_rate": 2.4965238623911718e-06, + "loss": 5.781668090820313, + "step": 56975 + }, + { + "epoch": 1.790044751511345, + "grad_norm": 3.7891573905944824, + "learning_rate": 2.492866046254031e-06, + "loss": 5.783619384765625, + "step": 57000 + }, + { + "epoch": 1.790829865745466, + "grad_norm": 4.714060306549072, + "learning_rate": 2.4892213739262646e-06, + "loss": 5.779414672851562, + "step": 57025 + }, + { + "epoch": 1.791614979979587, + "grad_norm": 3.9150326251983643, + "learning_rate": 2.4855898510390106e-06, + "loss": 5.79236083984375, + "step": 57050 + }, + { + "epoch": 1.7924000942137082, + "grad_norm": 3.274944305419922, + "learning_rate": 2.481971483203098e-06, + "loss": 5.779938354492187, + "step": 57075 + }, + { + "epoch": 1.793185208447829, + "grad_norm": 3.646406412124634, + "learning_rate": 2.478366276009022e-06, + "loss": 5.783805541992187, + "step": 57100 + }, + { + "epoch": 1.7939703226819503, + "grad_norm": 6.259413242340088, + "learning_rate": 2.4747742350269562e-06, + "loss": 5.786651611328125, + "step": 57125 + }, + { + "epoch": 1.7947554369160712, + "grad_norm": 4.198552131652832, + "learning_rate": 2.4711953658067224e-06, + "loss": 5.787659301757812, + "step": 57150 + }, + { + "epoch": 1.7955405511501925, + "grad_norm": 3.543794631958008, + "learning_rate": 2.4676296738777924e-06, + "loss": 5.8025732421875, + "step": 57175 + }, + { + "epoch": 1.7963256653843134, + "grad_norm": 3.018110513687134, + "learning_rate": 2.464077164749283e-06, + "loss": 5.780305786132812, + "step": 57200 + }, + { + "epoch": 1.7971107796184345, + "grad_norm": 2.677635908126831, + "learning_rate": 2.4605378439099393e-06, + "loss": 5.783151245117187, + "step": 57225 + }, + { + "epoch": 1.7978958938525555, + "grad_norm": 4.003279209136963, + "learning_rate": 2.4570117168281345e-06, + "loss": 5.783790893554688, + "step": 57250 + }, + { + "epoch": 1.7986810080866766, + "grad_norm": 3.4230806827545166, + "learning_rate": 2.4534987889518517e-06, + "loss": 5.794078979492188, + "step": 57275 + }, + { + "epoch": 1.7994661223207977, + "grad_norm": 3.4749460220336914, + "learning_rate": 2.4499990657086813e-06, + "loss": 5.784822998046875, + "step": 57300 + }, + { + "epoch": 1.8002512365549186, + "grad_norm": 2.7835826873779297, + "learning_rate": 2.4465125525058193e-06, + "loss": 5.79635498046875, + "step": 57325 + }, + { + "epoch": 1.80103635078904, + "grad_norm": 4.266292095184326, + "learning_rate": 2.4430392547300412e-06, + "loss": 5.778224487304687, + "step": 57350 + }, + { + "epoch": 1.8018214650231608, + "grad_norm": 5.103662014007568, + "learning_rate": 2.4395791777477133e-06, + "loss": 5.782383422851563, + "step": 57375 + }, + { + "epoch": 1.802606579257282, + "grad_norm": 3.1946451663970947, + "learning_rate": 2.436132326904769e-06, + "loss": 5.780348510742187, + "step": 57400 + }, + { + "epoch": 1.803391693491403, + "grad_norm": 3.102651596069336, + "learning_rate": 2.4326987075267076e-06, + "loss": 5.774244995117187, + "step": 57425 + }, + { + "epoch": 1.804176807725524, + "grad_norm": 4.263691425323486, + "learning_rate": 2.429278324918593e-06, + "loss": 5.785714111328125, + "step": 57450 + }, + { + "epoch": 1.804961921959645, + "grad_norm": 4.787829875946045, + "learning_rate": 2.4258711843650276e-06, + "loss": 5.780347900390625, + "step": 57475 + }, + { + "epoch": 1.8057470361937662, + "grad_norm": 5.553043842315674, + "learning_rate": 2.4224772911301568e-06, + "loss": 5.7987841796875, + "step": 57500 + }, + { + "epoch": 1.8065321504278873, + "grad_norm": 3.440094232559204, + "learning_rate": 2.4190966504576656e-06, + "loss": 5.769808959960938, + "step": 57525 + }, + { + "epoch": 1.8073172646620084, + "grad_norm": 4.034043312072754, + "learning_rate": 2.415729267570754e-06, + "loss": 5.7893359375, + "step": 57550 + }, + { + "epoch": 1.8081023788961295, + "grad_norm": 3.6898655891418457, + "learning_rate": 2.4123751476721445e-06, + "loss": 5.765208129882812, + "step": 57575 + }, + { + "epoch": 1.8088874931302503, + "grad_norm": 3.910935640335083, + "learning_rate": 2.4090342959440637e-06, + "loss": 5.78477294921875, + "step": 57600 + }, + { + "epoch": 1.8096726073643716, + "grad_norm": 3.28654146194458, + "learning_rate": 2.405706717548242e-06, + "loss": 5.788392333984375, + "step": 57625 + }, + { + "epoch": 1.8104577215984925, + "grad_norm": 4.983091831207275, + "learning_rate": 2.4023924176259e-06, + "loss": 5.7857470703125, + "step": 57650 + }, + { + "epoch": 1.8112428358326138, + "grad_norm": 3.1230952739715576, + "learning_rate": 2.3990914012977427e-06, + "loss": 5.785372924804688, + "step": 57675 + }, + { + "epoch": 1.8120279500667347, + "grad_norm": 3.2945268154144287, + "learning_rate": 2.3958036736639524e-06, + "loss": 5.792018432617187, + "step": 57700 + }, + { + "epoch": 1.8128130643008558, + "grad_norm": 2.338304042816162, + "learning_rate": 2.3925292398041797e-06, + "loss": 5.783242797851562, + "step": 57725 + }, + { + "epoch": 1.8135981785349768, + "grad_norm": 4.440317630767822, + "learning_rate": 2.3892681047775344e-06, + "loss": 5.782334594726563, + "step": 57750 + }, + { + "epoch": 1.814383292769098, + "grad_norm": 2.765134811401367, + "learning_rate": 2.3860202736225818e-06, + "loss": 5.789317626953125, + "step": 57775 + }, + { + "epoch": 1.815168407003219, + "grad_norm": 4.170756816864014, + "learning_rate": 2.3827857513573284e-06, + "loss": 5.7811517333984375, + "step": 57800 + }, + { + "epoch": 1.8159535212373399, + "grad_norm": 4.07201623916626, + "learning_rate": 2.3795645429792217e-06, + "loss": 5.7814013671875, + "step": 57825 + }, + { + "epoch": 1.8167386354714612, + "grad_norm": 4.814392566680908, + "learning_rate": 2.3763566534651384e-06, + "loss": 5.811619873046875, + "step": 57850 + }, + { + "epoch": 1.817523749705582, + "grad_norm": 3.9908409118652344, + "learning_rate": 2.3731620877713758e-06, + "loss": 5.794703369140625, + "step": 57875 + }, + { + "epoch": 1.8183088639397034, + "grad_norm": 5.778538703918457, + "learning_rate": 2.3699808508336434e-06, + "loss": 5.781085815429687, + "step": 57900 + }, + { + "epoch": 1.8190939781738242, + "grad_norm": 5.17758846282959, + "learning_rate": 2.366812947567065e-06, + "loss": 5.804376831054688, + "step": 57925 + }, + { + "epoch": 1.8198790924079453, + "grad_norm": 5.38642692565918, + "learning_rate": 2.363658382866152e-06, + "loss": 5.80313232421875, + "step": 57950 + }, + { + "epoch": 1.8206642066420664, + "grad_norm": 4.87246561050415, + "learning_rate": 2.360517161604818e-06, + "loss": 5.791321411132812, + "step": 57975 + }, + { + "epoch": 1.8214493208761875, + "grad_norm": 3.3646600246429443, + "learning_rate": 2.3573892886363536e-06, + "loss": 5.786720581054688, + "step": 58000 + }, + { + "epoch": 1.8222344351103086, + "grad_norm": 2.6734704971313477, + "learning_rate": 2.354274768793428e-06, + "loss": 5.791036376953125, + "step": 58025 + }, + { + "epoch": 1.8230195493444297, + "grad_norm": 2.505638360977173, + "learning_rate": 2.351173606888082e-06, + "loss": 5.785167846679688, + "step": 58050 + }, + { + "epoch": 1.8238046635785508, + "grad_norm": 4.570799827575684, + "learning_rate": 2.348085807711713e-06, + "loss": 5.79658447265625, + "step": 58075 + }, + { + "epoch": 1.8245897778126716, + "grad_norm": 3.4782326221466064, + "learning_rate": 2.3450113760350762e-06, + "loss": 5.786616821289062, + "step": 58100 + }, + { + "epoch": 1.825374892046793, + "grad_norm": 4.405145168304443, + "learning_rate": 2.341950316608274e-06, + "loss": 5.770802001953125, + "step": 58125 + }, + { + "epoch": 1.8261600062809138, + "grad_norm": 3.6037471294403076, + "learning_rate": 2.338902634160743e-06, + "loss": 5.783311767578125, + "step": 58150 + }, + { + "epoch": 1.826945120515035, + "grad_norm": 3.260993003845215, + "learning_rate": 2.3358683334012583e-06, + "loss": 5.79878173828125, + "step": 58175 + }, + { + "epoch": 1.827730234749156, + "grad_norm": 4.4021220207214355, + "learning_rate": 2.3328474190179144e-06, + "loss": 5.78266845703125, + "step": 58200 + }, + { + "epoch": 1.828515348983277, + "grad_norm": 3.6197638511657715, + "learning_rate": 2.329839895678131e-06, + "loss": 5.80041748046875, + "step": 58225 + }, + { + "epoch": 1.8293004632173981, + "grad_norm": 4.246242523193359, + "learning_rate": 2.3268457680286288e-06, + "loss": 5.783023071289063, + "step": 58250 + }, + { + "epoch": 1.8300855774515192, + "grad_norm": 3.528808116912842, + "learning_rate": 2.3238650406954366e-06, + "loss": 5.783646850585938, + "step": 58275 + }, + { + "epoch": 1.8308706916856403, + "grad_norm": 4.730537414550781, + "learning_rate": 2.32089771828388e-06, + "loss": 5.788618774414062, + "step": 58300 + }, + { + "epoch": 1.8316558059197612, + "grad_norm": 4.1898579597473145, + "learning_rate": 2.317943805378573e-06, + "loss": 5.773575439453125, + "step": 58325 + }, + { + "epoch": 1.8324409201538825, + "grad_norm": 4.67256498336792, + "learning_rate": 2.315003306543411e-06, + "loss": 5.797017822265625, + "step": 58350 + }, + { + "epoch": 1.8332260343880034, + "grad_norm": 3.5924487113952637, + "learning_rate": 2.3120762263215627e-06, + "loss": 5.79921142578125, + "step": 58375 + }, + { + "epoch": 1.8340111486221247, + "grad_norm": 2.035011053085327, + "learning_rate": 2.3091625692354636e-06, + "loss": 5.784267578125, + "step": 58400 + }, + { + "epoch": 1.8347962628562455, + "grad_norm": 2.923826217651367, + "learning_rate": 2.3062623397868175e-06, + "loss": 5.77276611328125, + "step": 58425 + }, + { + "epoch": 1.8355813770903666, + "grad_norm": 4.960667610168457, + "learning_rate": 2.303375542456574e-06, + "loss": 5.794833374023438, + "step": 58450 + }, + { + "epoch": 1.8363664913244877, + "grad_norm": 3.762519121170044, + "learning_rate": 2.3005021817049334e-06, + "loss": 5.775562744140625, + "step": 58475 + }, + { + "epoch": 1.8371516055586088, + "grad_norm": 2.9825613498687744, + "learning_rate": 2.297642261971334e-06, + "loss": 5.780364990234375, + "step": 58500 + }, + { + "epoch": 1.8379367197927299, + "grad_norm": 3.4780750274658203, + "learning_rate": 2.294795787674449e-06, + "loss": 5.78034912109375, + "step": 58525 + }, + { + "epoch": 1.8387218340268507, + "grad_norm": 3.4250328540802, + "learning_rate": 2.291962763212178e-06, + "loss": 5.78702392578125, + "step": 58550 + }, + { + "epoch": 1.839506948260972, + "grad_norm": 5.635530948638916, + "learning_rate": 2.2891431929616393e-06, + "loss": 5.787509765625, + "step": 58575 + }, + { + "epoch": 1.840292062495093, + "grad_norm": 3.18450927734375, + "learning_rate": 2.2863370812791615e-06, + "loss": 5.7843359375, + "step": 58600 + }, + { + "epoch": 1.8410771767292142, + "grad_norm": 2.879680871963501, + "learning_rate": 2.283544432500286e-06, + "loss": 5.791638793945313, + "step": 58625 + }, + { + "epoch": 1.841862290963335, + "grad_norm": 3.5358824729919434, + "learning_rate": 2.280765250939748e-06, + "loss": 5.77949462890625, + "step": 58650 + }, + { + "epoch": 1.8426474051974562, + "grad_norm": 4.449724197387695, + "learning_rate": 2.277999540891477e-06, + "loss": 5.785150146484375, + "step": 58675 + }, + { + "epoch": 1.8434325194315773, + "grad_norm": 3.453212261199951, + "learning_rate": 2.275247306628587e-06, + "loss": 5.781573486328125, + "step": 58700 + }, + { + "epoch": 1.8442176336656984, + "grad_norm": 5.285698413848877, + "learning_rate": 2.2725085524033744e-06, + "loss": 5.7803717041015625, + "step": 58725 + }, + { + "epoch": 1.8450027478998194, + "grad_norm": 2.9167189598083496, + "learning_rate": 2.269783282447307e-06, + "loss": 5.797994384765625, + "step": 58750 + }, + { + "epoch": 1.8457878621339405, + "grad_norm": 5.716033458709717, + "learning_rate": 2.2670715009710185e-06, + "loss": 5.775316772460937, + "step": 58775 + }, + { + "epoch": 1.8465729763680616, + "grad_norm": 2.6297805309295654, + "learning_rate": 2.264373212164303e-06, + "loss": 5.781688842773438, + "step": 58800 + }, + { + "epoch": 1.8473580906021825, + "grad_norm": 3.4627342224121094, + "learning_rate": 2.2616884201961083e-06, + "loss": 5.7897509765625, + "step": 58825 + }, + { + "epoch": 1.8481432048363038, + "grad_norm": 4.144840240478516, + "learning_rate": 2.259017129214528e-06, + "loss": 5.786851196289063, + "step": 58850 + }, + { + "epoch": 1.8489283190704247, + "grad_norm": 3.671869993209839, + "learning_rate": 2.2563593433467983e-06, + "loss": 5.78335693359375, + "step": 58875 + }, + { + "epoch": 1.849713433304546, + "grad_norm": 2.381124496459961, + "learning_rate": 2.2537150666992858e-06, + "loss": 5.7755194091796875, + "step": 58900 + }, + { + "epoch": 1.8504985475386668, + "grad_norm": 3.9017984867095947, + "learning_rate": 2.251084303357488e-06, + "loss": 5.79106689453125, + "step": 58925 + }, + { + "epoch": 1.851283661772788, + "grad_norm": 3.1406571865081787, + "learning_rate": 2.2484670573860247e-06, + "loss": 5.778856549944196, + "step": 58950 + }, + { + "epoch": 1.852068776006909, + "grad_norm": 3.477116107940674, + "learning_rate": 2.2458633328286285e-06, + "loss": 5.795448608398438, + "step": 58975 + }, + { + "epoch": 1.85285389024103, + "grad_norm": 3.5605623722076416, + "learning_rate": 2.243273133708141e-06, + "loss": 5.795104370117188, + "step": 59000 + }, + { + "epoch": 1.8536390044751512, + "grad_norm": 4.19479513168335, + "learning_rate": 2.2406964640265095e-06, + "loss": 5.772466430664062, + "step": 59025 + }, + { + "epoch": 1.854424118709272, + "grad_norm": 2.3583836555480957, + "learning_rate": 2.238133327764772e-06, + "loss": 5.78193359375, + "step": 59050 + }, + { + "epoch": 1.8552092329433933, + "grad_norm": 2.3124489784240723, + "learning_rate": 2.235583728883065e-06, + "loss": 5.797161865234375, + "step": 59075 + }, + { + "epoch": 1.8559943471775142, + "grad_norm": 5.676886558532715, + "learning_rate": 2.233047671320601e-06, + "loss": 5.7967724609375, + "step": 59100 + }, + { + "epoch": 1.8567794614116355, + "grad_norm": 4.253902912139893, + "learning_rate": 2.2305251589956763e-06, + "loss": 5.7898876953125, + "step": 59125 + }, + { + "epoch": 1.8575645756457564, + "grad_norm": 4.704356670379639, + "learning_rate": 2.2280161958056597e-06, + "loss": 5.781226806640625, + "step": 59150 + }, + { + "epoch": 1.8583496898798775, + "grad_norm": 4.7895731925964355, + "learning_rate": 2.22552078562698e-06, + "loss": 5.773465576171875, + "step": 59175 + }, + { + "epoch": 1.8591348041139986, + "grad_norm": 4.225935935974121, + "learning_rate": 2.223038932315133e-06, + "loss": 5.779439086914063, + "step": 59200 + }, + { + "epoch": 1.8599199183481196, + "grad_norm": 4.086517810821533, + "learning_rate": 2.2205706397046665e-06, + "loss": 5.794036254882813, + "step": 59225 + }, + { + "epoch": 1.8607050325822407, + "grad_norm": 3.264148473739624, + "learning_rate": 2.2181159116091736e-06, + "loss": 5.7904150390625, + "step": 59250 + }, + { + "epoch": 1.8614901468163618, + "grad_norm": 5.057831287384033, + "learning_rate": 2.2156747518212935e-06, + "loss": 5.783369750976562, + "step": 59275 + }, + { + "epoch": 1.862275261050483, + "grad_norm": 3.8166027069091797, + "learning_rate": 2.2132471641126997e-06, + "loss": 5.7827825927734375, + "step": 59300 + }, + { + "epoch": 1.8630603752846038, + "grad_norm": 2.9191529750823975, + "learning_rate": 2.2108331522340984e-06, + "loss": 5.763616943359375, + "step": 59325 + }, + { + "epoch": 1.863845489518725, + "grad_norm": 4.630315780639648, + "learning_rate": 2.20843271991522e-06, + "loss": 5.789846801757813, + "step": 59350 + }, + { + "epoch": 1.864630603752846, + "grad_norm": 2.87978196144104, + "learning_rate": 2.2060458708648104e-06, + "loss": 5.781611938476562, + "step": 59375 + }, + { + "epoch": 1.8654157179869673, + "grad_norm": 4.191797256469727, + "learning_rate": 2.203672608770635e-06, + "loss": 5.802743530273437, + "step": 59400 + }, + { + "epoch": 1.8662008322210881, + "grad_norm": 3.7984495162963867, + "learning_rate": 2.201312937299463e-06, + "loss": 5.788426513671875, + "step": 59425 + }, + { + "epoch": 1.8669859464552092, + "grad_norm": 4.2166290283203125, + "learning_rate": 2.198966860097067e-06, + "loss": 5.797529296875, + "step": 59450 + }, + { + "epoch": 1.8677710606893303, + "grad_norm": 3.049044132232666, + "learning_rate": 2.1966343807882155e-06, + "loss": 5.7952728271484375, + "step": 59475 + }, + { + "epoch": 1.8685561749234514, + "grad_norm": 2.9743266105651855, + "learning_rate": 2.1943155029766654e-06, + "loss": 5.789248046875, + "step": 59500 + }, + { + "epoch": 1.8693412891575725, + "grad_norm": 2.7810287475585938, + "learning_rate": 2.192010230245166e-06, + "loss": 5.79384521484375, + "step": 59525 + }, + { + "epoch": 1.8701264033916933, + "grad_norm": 3.7113232612609863, + "learning_rate": 2.1897185661554384e-06, + "loss": 5.8027581787109375, + "step": 59550 + }, + { + "epoch": 1.8709115176258146, + "grad_norm": 3.822075605392456, + "learning_rate": 2.1874405142481805e-06, + "loss": 5.777032470703125, + "step": 59575 + }, + { + "epoch": 1.8716966318599355, + "grad_norm": 3.7263662815093994, + "learning_rate": 2.1851760780430603e-06, + "loss": 5.7768408203125, + "step": 59600 + }, + { + "epoch": 1.8724817460940568, + "grad_norm": 4.195851802825928, + "learning_rate": 2.1829252610387085e-06, + "loss": 5.777986450195312, + "step": 59625 + }, + { + "epoch": 1.8732668603281777, + "grad_norm": 2.6194324493408203, + "learning_rate": 2.1806880667127137e-06, + "loss": 5.785612182617188, + "step": 59650 + }, + { + "epoch": 1.8740519745622988, + "grad_norm": 4.039409637451172, + "learning_rate": 2.1784644985216174e-06, + "loss": 5.789314575195313, + "step": 59675 + }, + { + "epoch": 1.8748370887964199, + "grad_norm": 2.8238441944122314, + "learning_rate": 2.1762545599009036e-06, + "loss": 5.775830078125, + "step": 59700 + }, + { + "epoch": 1.875622203030541, + "grad_norm": 6.607328414916992, + "learning_rate": 2.174058254265008e-06, + "loss": 5.78297119140625, + "step": 59725 + }, + { + "epoch": 1.876407317264662, + "grad_norm": 3.6201274394989014, + "learning_rate": 2.1718755850072927e-06, + "loss": 5.785714721679687, + "step": 59750 + }, + { + "epoch": 1.8771924314987831, + "grad_norm": 4.098516941070557, + "learning_rate": 2.1697065555000577e-06, + "loss": 5.784662475585938, + "step": 59775 + }, + { + "epoch": 1.8779775457329042, + "grad_norm": 3.4030988216400146, + "learning_rate": 2.167551169094525e-06, + "loss": 5.784811401367188, + "step": 59800 + }, + { + "epoch": 1.878762659967025, + "grad_norm": 3.757260322570801, + "learning_rate": 2.165409429120841e-06, + "loss": 5.786997680664062, + "step": 59825 + }, + { + "epoch": 1.8795477742011464, + "grad_norm": 4.464357376098633, + "learning_rate": 2.1632813388880653e-06, + "loss": 5.789347534179687, + "step": 59850 + }, + { + "epoch": 1.8803328884352672, + "grad_norm": 3.848306655883789, + "learning_rate": 2.1611669016841704e-06, + "loss": 5.7890625, + "step": 59875 + }, + { + "epoch": 1.8811180026693886, + "grad_norm": 3.6775693893432617, + "learning_rate": 2.159066120776031e-06, + "loss": 5.80010498046875, + "step": 59900 + }, + { + "epoch": 1.8819031169035094, + "grad_norm": 3.8425791263580322, + "learning_rate": 2.156978999409427e-06, + "loss": 5.795027465820312, + "step": 59925 + }, + { + "epoch": 1.8826882311376305, + "grad_norm": 3.9309451580047607, + "learning_rate": 2.1549055408090305e-06, + "loss": 5.779423217773438, + "step": 59950 + }, + { + "epoch": 1.8834733453717516, + "grad_norm": 4.011724472045898, + "learning_rate": 2.1528457481784045e-06, + "loss": 5.778952026367188, + "step": 59975 + }, + { + "epoch": 1.8842584596058727, + "grad_norm": 7.163304328918457, + "learning_rate": 2.1507996247e-06, + "loss": 5.78554931640625, + "step": 60000 + }, + { + "epoch": 1.8850435738399938, + "grad_norm": 8.378658294677734, + "learning_rate": 2.1487671735351435e-06, + "loss": 5.790424194335937, + "step": 60025 + }, + { + "epoch": 1.8858286880741146, + "grad_norm": 3.287416696548462, + "learning_rate": 2.1467483978240443e-06, + "loss": 5.79379638671875, + "step": 60050 + }, + { + "epoch": 1.886613802308236, + "grad_norm": 4.962676525115967, + "learning_rate": 2.1447433006857777e-06, + "loss": 5.78060546875, + "step": 60075 + }, + { + "epoch": 1.8873989165423568, + "grad_norm": 2.7218706607818604, + "learning_rate": 2.142751885218286e-06, + "loss": 5.77807861328125, + "step": 60100 + }, + { + "epoch": 1.8881840307764781, + "grad_norm": 4.824492931365967, + "learning_rate": 2.1407741544983747e-06, + "loss": 5.785328979492188, + "step": 60125 + }, + { + "epoch": 1.888969145010599, + "grad_norm": 3.6048169136047363, + "learning_rate": 2.138810111581702e-06, + "loss": 5.7989434814453125, + "step": 60150 + }, + { + "epoch": 1.88975425924472, + "grad_norm": 2.204124927520752, + "learning_rate": 2.1368597595027844e-06, + "loss": 5.792523193359375, + "step": 60175 + }, + { + "epoch": 1.8905393734788412, + "grad_norm": 6.602739334106445, + "learning_rate": 2.1349231012749777e-06, + "loss": 5.78777587890625, + "step": 60200 + }, + { + "epoch": 1.8913244877129622, + "grad_norm": 3.595128059387207, + "learning_rate": 2.1330001398904878e-06, + "loss": 5.787883911132813, + "step": 60225 + }, + { + "epoch": 1.8921096019470833, + "grad_norm": 3.647548198699951, + "learning_rate": 2.131090878320352e-06, + "loss": 5.785348510742187, + "step": 60250 + }, + { + "epoch": 1.8928947161812044, + "grad_norm": 4.918752193450928, + "learning_rate": 2.1291953195144462e-06, + "loss": 5.791661376953125, + "step": 60275 + }, + { + "epoch": 1.8936798304153255, + "grad_norm": 5.96439790725708, + "learning_rate": 2.127313466401472e-06, + "loss": 5.79294189453125, + "step": 60300 + }, + { + "epoch": 1.8944649446494464, + "grad_norm": 3.9335436820983887, + "learning_rate": 2.1254453218889556e-06, + "loss": 5.77436279296875, + "step": 60325 + }, + { + "epoch": 1.8952500588835677, + "grad_norm": 4.609477996826172, + "learning_rate": 2.1235908888632437e-06, + "loss": 5.777764892578125, + "step": 60350 + }, + { + "epoch": 1.8960351731176885, + "grad_norm": 4.48944616317749, + "learning_rate": 2.121750170189499e-06, + "loss": 5.7913427734375, + "step": 60375 + }, + { + "epoch": 1.8968202873518099, + "grad_norm": 4.636702537536621, + "learning_rate": 2.1199231687116915e-06, + "loss": 5.775206298828125, + "step": 60400 + }, + { + "epoch": 1.8976054015859307, + "grad_norm": 3.537494421005249, + "learning_rate": 2.118109887252603e-06, + "loss": 5.781174926757813, + "step": 60425 + }, + { + "epoch": 1.8983905158200518, + "grad_norm": 4.737220764160156, + "learning_rate": 2.116310328613813e-06, + "loss": 5.779356079101563, + "step": 60450 + }, + { + "epoch": 1.899175630054173, + "grad_norm": 3.298790216445923, + "learning_rate": 2.1145244955757006e-06, + "loss": 5.779755249023437, + "step": 60475 + }, + { + "epoch": 1.899960744288294, + "grad_norm": 3.706453561782837, + "learning_rate": 2.1127523908974384e-06, + "loss": 5.7891943359375, + "step": 60500 + }, + { + "epoch": 1.900745858522415, + "grad_norm": 5.8668928146362305, + "learning_rate": 2.110994017316989e-06, + "loss": 5.786451416015625, + "step": 60525 + }, + { + "epoch": 1.901530972756536, + "grad_norm": 3.6240322589874268, + "learning_rate": 2.1092493775510974e-06, + "loss": 5.779509887695313, + "step": 60550 + }, + { + "epoch": 1.9023160869906572, + "grad_norm": 5.013444900512695, + "learning_rate": 2.107518474295292e-06, + "loss": 5.78287353515625, + "step": 60575 + }, + { + "epoch": 1.903101201224778, + "grad_norm": 3.8221323490142822, + "learning_rate": 2.1058013102238743e-06, + "loss": 5.78815185546875, + "step": 60600 + }, + { + "epoch": 1.9038863154588994, + "grad_norm": 3.975430488586426, + "learning_rate": 2.104097887989925e-06, + "loss": 5.788541870117188, + "step": 60625 + }, + { + "epoch": 1.9046714296930203, + "grad_norm": 2.6323115825653076, + "learning_rate": 2.102408210225286e-06, + "loss": 5.7849578857421875, + "step": 60650 + }, + { + "epoch": 1.9054565439271414, + "grad_norm": 2.6785826683044434, + "learning_rate": 2.100732279540568e-06, + "loss": 5.788777465820313, + "step": 60675 + }, + { + "epoch": 1.9062416581612625, + "grad_norm": 3.802520751953125, + "learning_rate": 2.0990700985251398e-06, + "loss": 5.785486450195313, + "step": 60700 + }, + { + "epoch": 1.9070267723953835, + "grad_norm": 9.879276275634766, + "learning_rate": 2.097421669747127e-06, + "loss": 5.800463256835937, + "step": 60725 + }, + { + "epoch": 1.9078118866295046, + "grad_norm": 2.5163230895996094, + "learning_rate": 2.095786995753409e-06, + "loss": 5.785342407226563, + "step": 60750 + }, + { + "epoch": 1.9085970008636255, + "grad_norm": 4.3876190185546875, + "learning_rate": 2.09416607906961e-06, + "loss": 5.787952880859375, + "step": 60775 + }, + { + "epoch": 1.9093821150977468, + "grad_norm": 4.441567420959473, + "learning_rate": 2.0925589222001026e-06, + "loss": 5.7884814453125, + "step": 60800 + }, + { + "epoch": 1.9101672293318677, + "grad_norm": 3.4357216358184814, + "learning_rate": 2.090965527628e-06, + "loss": 5.791012573242187, + "step": 60825 + }, + { + "epoch": 1.910952343565989, + "grad_norm": 4.300515174865723, + "learning_rate": 2.0893858978151475e-06, + "loss": 5.785883178710938, + "step": 60850 + }, + { + "epoch": 1.9117374578001098, + "grad_norm": 2.846773386001587, + "learning_rate": 2.0878200352021293e-06, + "loss": 5.77925537109375, + "step": 60875 + }, + { + "epoch": 1.912522572034231, + "grad_norm": 3.041958808898926, + "learning_rate": 2.086267942208253e-06, + "loss": 5.792296752929688, + "step": 60900 + }, + { + "epoch": 1.913307686268352, + "grad_norm": 4.894898414611816, + "learning_rate": 2.084729621231557e-06, + "loss": 5.776842041015625, + "step": 60925 + }, + { + "epoch": 1.914092800502473, + "grad_norm": 4.310283184051514, + "learning_rate": 2.0832050746488e-06, + "loss": 5.783323974609375, + "step": 60950 + }, + { + "epoch": 1.9148779147365942, + "grad_norm": 3.4523653984069824, + "learning_rate": 2.081694304815456e-06, + "loss": 5.790947265625, + "step": 60975 + }, + { + "epoch": 1.9156630289707153, + "grad_norm": 3.7091033458709717, + "learning_rate": 2.080197314065715e-06, + "loss": 5.779934692382812, + "step": 61000 + }, + { + "epoch": 1.9164481432048364, + "grad_norm": 4.69081449508667, + "learning_rate": 2.0787141047124814e-06, + "loss": 5.774959106445312, + "step": 61025 + }, + { + "epoch": 1.9172332574389572, + "grad_norm": 3.4976179599761963, + "learning_rate": 2.0772446790473603e-06, + "loss": 5.77817138671875, + "step": 61050 + }, + { + "epoch": 1.9180183716730785, + "grad_norm": 4.179015636444092, + "learning_rate": 2.075789039340666e-06, + "loss": 5.774219970703125, + "step": 61075 + }, + { + "epoch": 1.9188034859071994, + "grad_norm": 3.8303868770599365, + "learning_rate": 2.07434718784141e-06, + "loss": 5.789683227539062, + "step": 61100 + }, + { + "epoch": 1.9195886001413207, + "grad_norm": 3.7448995113372803, + "learning_rate": 2.072919126777302e-06, + "loss": 5.783723754882812, + "step": 61125 + }, + { + "epoch": 1.9203737143754416, + "grad_norm": 4.499347686767578, + "learning_rate": 2.0715048583547455e-06, + "loss": 5.78147705078125, + "step": 61150 + }, + { + "epoch": 1.9211588286095627, + "grad_norm": 3.2793524265289307, + "learning_rate": 2.070104384758831e-06, + "loss": 5.780831298828125, + "step": 61175 + }, + { + "epoch": 1.9219439428436838, + "grad_norm": 4.54244327545166, + "learning_rate": 2.0687177081533393e-06, + "loss": 5.789022827148438, + "step": 61200 + }, + { + "epoch": 1.9227290570778048, + "grad_norm": 6.825347423553467, + "learning_rate": 2.067344830680731e-06, + "loss": 5.78580810546875, + "step": 61225 + }, + { + "epoch": 1.923514171311926, + "grad_norm": 3.4214630126953125, + "learning_rate": 2.0659857544621505e-06, + "loss": 5.782777709960937, + "step": 61250 + }, + { + "epoch": 1.9242992855460468, + "grad_norm": 4.202260971069336, + "learning_rate": 2.064640481597415e-06, + "loss": 5.785875244140625, + "step": 61275 + }, + { + "epoch": 1.925084399780168, + "grad_norm": 4.406137943267822, + "learning_rate": 2.0633090141650157e-06, + "loss": 5.7816796875, + "step": 61300 + }, + { + "epoch": 1.925869514014289, + "grad_norm": 5.518707275390625, + "learning_rate": 2.0619913542221163e-06, + "loss": 5.777359619140625, + "step": 61325 + }, + { + "epoch": 1.9266546282484103, + "grad_norm": 4.180536270141602, + "learning_rate": 2.060687503804545e-06, + "loss": 5.781214599609375, + "step": 61350 + }, + { + "epoch": 1.9274397424825311, + "grad_norm": 4.059115409851074, + "learning_rate": 2.0593974649267945e-06, + "loss": 5.783255615234375, + "step": 61375 + }, + { + "epoch": 1.9282248567166522, + "grad_norm": 4.425353050231934, + "learning_rate": 2.0581212395820195e-06, + "loss": 5.782518920898437, + "step": 61400 + }, + { + "epoch": 1.9290099709507733, + "grad_norm": 4.238407611846924, + "learning_rate": 2.0568588297420316e-06, + "loss": 5.777192993164062, + "step": 61425 + }, + { + "epoch": 1.9297950851848944, + "grad_norm": 3.9501936435699463, + "learning_rate": 2.0556102373572946e-06, + "loss": 5.780035400390625, + "step": 61450 + }, + { + "epoch": 1.9305801994190155, + "grad_norm": 3.738571882247925, + "learning_rate": 2.054375464356929e-06, + "loss": 5.790166015625, + "step": 61475 + }, + { + "epoch": 1.9313653136531366, + "grad_norm": 4.869884014129639, + "learning_rate": 2.0531545126486966e-06, + "loss": 5.789022216796875, + "step": 61500 + }, + { + "epoch": 1.9321504278872577, + "grad_norm": 3.0519678592681885, + "learning_rate": 2.051947384119013e-06, + "loss": 5.7828125, + "step": 61525 + }, + { + "epoch": 1.9329355421213785, + "grad_norm": 2.7393877506256104, + "learning_rate": 2.0507540806329308e-06, + "loss": 5.774522094726563, + "step": 61550 + }, + { + "epoch": 1.9337206563554998, + "grad_norm": 3.5009636878967285, + "learning_rate": 2.049574604034143e-06, + "loss": 5.775469360351562, + "step": 61575 + }, + { + "epoch": 1.9345057705896207, + "grad_norm": 3.3845021724700928, + "learning_rate": 2.0484089561449833e-06, + "loss": 5.773892211914062, + "step": 61600 + }, + { + "epoch": 1.935290884823742, + "grad_norm": 3.335068702697754, + "learning_rate": 2.047257138766416e-06, + "loss": 5.795314331054687, + "step": 61625 + }, + { + "epoch": 1.9360759990578629, + "grad_norm": 3.045802354812622, + "learning_rate": 2.0461191536780373e-06, + "loss": 5.792161254882813, + "step": 61650 + }, + { + "epoch": 1.936861113291984, + "grad_norm": 3.1388752460479736, + "learning_rate": 2.044995002638074e-06, + "loss": 5.789420776367187, + "step": 61675 + }, + { + "epoch": 1.937646227526105, + "grad_norm": 3.2305843830108643, + "learning_rate": 2.0438846873833752e-06, + "loss": 5.7780908203125, + "step": 61700 + }, + { + "epoch": 1.9384313417602261, + "grad_norm": 3.571904182434082, + "learning_rate": 2.042788209629418e-06, + "loss": 5.775611572265625, + "step": 61725 + }, + { + "epoch": 1.9392164559943472, + "grad_norm": 3.4981415271759033, + "learning_rate": 2.0417055710702955e-06, + "loss": 5.775521850585937, + "step": 61750 + }, + { + "epoch": 1.940001570228468, + "grad_norm": 3.311546802520752, + "learning_rate": 2.0406367733787223e-06, + "loss": 5.782902221679688, + "step": 61775 + }, + { + "epoch": 1.9407866844625894, + "grad_norm": 3.3298513889312744, + "learning_rate": 2.0395818182060266e-06, + "loss": 5.7849658203125, + "step": 61800 + }, + { + "epoch": 1.9415717986967103, + "grad_norm": 2.769050121307373, + "learning_rate": 2.038540707182149e-06, + "loss": 5.78597900390625, + "step": 61825 + }, + { + "epoch": 1.9423569129308316, + "grad_norm": 4.5914435386657715, + "learning_rate": 2.0375134419156417e-06, + "loss": 5.77887451171875, + "step": 61850 + }, + { + "epoch": 1.9431420271649524, + "grad_norm": 4.043377876281738, + "learning_rate": 2.0365000239936655e-06, + "loss": 5.7737054443359375, + "step": 61875 + }, + { + "epoch": 1.9439271413990735, + "grad_norm": 3.947849750518799, + "learning_rate": 2.0355004549819822e-06, + "loss": 5.79206787109375, + "step": 61900 + }, + { + "epoch": 1.9447122556331946, + "grad_norm": 3.781388998031616, + "learning_rate": 2.034514736424963e-06, + "loss": 5.778006591796875, + "step": 61925 + }, + { + "epoch": 1.9454973698673157, + "grad_norm": 3.867705821990967, + "learning_rate": 2.0335428698455744e-06, + "loss": 5.769979248046875, + "step": 61950 + }, + { + "epoch": 1.9462824841014368, + "grad_norm": 5.753755569458008, + "learning_rate": 2.0325848567453847e-06, + "loss": 5.790403442382813, + "step": 61975 + }, + { + "epoch": 1.9470675983355579, + "grad_norm": 4.019261837005615, + "learning_rate": 2.031640698604555e-06, + "loss": 5.780509643554687, + "step": 62000 + }, + { + "epoch": 1.947852712569679, + "grad_norm": 5.093113422393799, + "learning_rate": 2.030710396881841e-06, + "loss": 5.773137817382812, + "step": 62025 + }, + { + "epoch": 1.9486378268037998, + "grad_norm": 3.4105749130249023, + "learning_rate": 2.029793953014593e-06, + "loss": 5.7886419677734375, + "step": 62050 + }, + { + "epoch": 1.9494229410379211, + "grad_norm": 3.835343599319458, + "learning_rate": 2.028891368418745e-06, + "loss": 5.773190307617187, + "step": 62075 + }, + { + "epoch": 1.950208055272042, + "grad_norm": 5.637033939361572, + "learning_rate": 2.0280026444888216e-06, + "loss": 5.786654663085938, + "step": 62100 + }, + { + "epoch": 1.9509931695061633, + "grad_norm": 3.5364420413970947, + "learning_rate": 2.0271277825979322e-06, + "loss": 5.788189086914063, + "step": 62125 + }, + { + "epoch": 1.9517782837402842, + "grad_norm": 5.1593403816223145, + "learning_rate": 2.026266784097767e-06, + "loss": 5.817431640625, + "step": 62150 + }, + { + "epoch": 1.9525633979744053, + "grad_norm": 2.0856986045837402, + "learning_rate": 2.025419650318598e-06, + "loss": 5.761876220703125, + "step": 62175 + }, + { + "epoch": 1.9533485122085263, + "grad_norm": 3.3942999839782715, + "learning_rate": 2.0245863825692756e-06, + "loss": 5.785904541015625, + "step": 62200 + }, + { + "epoch": 1.9541336264426474, + "grad_norm": 3.7550201416015625, + "learning_rate": 2.0237669821372264e-06, + "loss": 5.781552124023437, + "step": 62225 + }, + { + "epoch": 1.9549187406767685, + "grad_norm": 3.946678876876831, + "learning_rate": 2.022961450288452e-06, + "loss": 5.771177368164063, + "step": 62250 + }, + { + "epoch": 1.9557038549108894, + "grad_norm": 6.150500297546387, + "learning_rate": 2.022169788267526e-06, + "loss": 5.77107177734375, + "step": 62275 + }, + { + "epoch": 1.9564889691450107, + "grad_norm": 3.1442553997039795, + "learning_rate": 2.021391997297593e-06, + "loss": 5.785365600585937, + "step": 62300 + }, + { + "epoch": 1.9572740833791316, + "grad_norm": 3.1751978397369385, + "learning_rate": 2.0206280785803673e-06, + "loss": 5.784210815429687, + "step": 62325 + }, + { + "epoch": 1.9580591976132529, + "grad_norm": 4.0022430419921875, + "learning_rate": 2.019878033296128e-06, + "loss": 5.7871826171875, + "step": 62350 + }, + { + "epoch": 1.9588443118473737, + "grad_norm": 4.053112506866455, + "learning_rate": 2.01914186260372e-06, + "loss": 5.780485229492188, + "step": 62375 + }, + { + "epoch": 1.9596294260814948, + "grad_norm": 4.368366241455078, + "learning_rate": 2.018419567640554e-06, + "loss": 5.783101196289063, + "step": 62400 + }, + { + "epoch": 1.960414540315616, + "grad_norm": 3.72163987159729, + "learning_rate": 2.0177111495225976e-06, + "loss": 5.790060424804688, + "step": 62425 + }, + { + "epoch": 1.961199654549737, + "grad_norm": 3.8638598918914795, + "learning_rate": 2.0170166093443828e-06, + "loss": 5.789242553710937, + "step": 62450 + }, + { + "epoch": 1.961984768783858, + "grad_norm": 3.8291313648223877, + "learning_rate": 2.016335948178996e-06, + "loss": 5.78851318359375, + "step": 62475 + }, + { + "epoch": 1.9627698830179792, + "grad_norm": 4.495358943939209, + "learning_rate": 2.0156691670780813e-06, + "loss": 5.77971923828125, + "step": 62500 + }, + { + "epoch": 1.9635549972521003, + "grad_norm": 3.2389588356018066, + "learning_rate": 2.015016267071841e-06, + "loss": 5.765732421875, + "step": 62525 + }, + { + "epoch": 1.9643401114862211, + "grad_norm": 2.8751864433288574, + "learning_rate": 2.0143772491690255e-06, + "loss": 5.765968017578125, + "step": 62550 + }, + { + "epoch": 1.9651252257203424, + "grad_norm": 5.628393173217773, + "learning_rate": 2.0137521143569387e-06, + "loss": 5.7785491943359375, + "step": 62575 + }, + { + "epoch": 1.9659103399544633, + "grad_norm": 3.7507827281951904, + "learning_rate": 2.013140863601437e-06, + "loss": 5.782030029296875, + "step": 62600 + }, + { + "epoch": 1.9666954541885846, + "grad_norm": 3.572502851486206, + "learning_rate": 2.0125434978469206e-06, + "loss": 5.794946899414063, + "step": 62625 + }, + { + "epoch": 1.9674805684227055, + "grad_norm": 4.890468120574951, + "learning_rate": 2.0119600180163422e-06, + "loss": 5.786099243164062, + "step": 62650 + }, + { + "epoch": 1.9682656826568266, + "grad_norm": 3.5127391815185547, + "learning_rate": 2.011390425011197e-06, + "loss": 5.788460083007813, + "step": 62675 + }, + { + "epoch": 1.9690507968909476, + "grad_norm": 5.765087604522705, + "learning_rate": 2.0108347197115248e-06, + "loss": 5.7733642578125, + "step": 62700 + }, + { + "epoch": 1.9698359111250687, + "grad_norm": 3.939879894256592, + "learning_rate": 2.0102929029759106e-06, + "loss": 5.7734136962890625, + "step": 62725 + }, + { + "epoch": 1.9706210253591898, + "grad_norm": 4.255554676055908, + "learning_rate": 2.009764975641477e-06, + "loss": 5.787247924804688, + "step": 62750 + }, + { + "epoch": 1.9714061395933107, + "grad_norm": 8.5287504196167, + "learning_rate": 2.0092509385238925e-06, + "loss": 5.785989379882812, + "step": 62775 + }, + { + "epoch": 1.972191253827432, + "grad_norm": 3.518280267715454, + "learning_rate": 2.0087507924173606e-06, + "loss": 5.785432739257812, + "step": 62800 + }, + { + "epoch": 1.9729763680615529, + "grad_norm": 4.965644359588623, + "learning_rate": 2.0082645380946235e-06, + "loss": 5.788304443359375, + "step": 62825 + }, + { + "epoch": 1.9737614822956742, + "grad_norm": 3.230278730392456, + "learning_rate": 2.007792176306961e-06, + "loss": 5.778438720703125, + "step": 62850 + }, + { + "epoch": 1.974546596529795, + "grad_norm": 3.305490732192993, + "learning_rate": 2.007333707784187e-06, + "loss": 5.7770361328125, + "step": 62875 + }, + { + "epoch": 1.9753317107639161, + "grad_norm": 2.441220998764038, + "learning_rate": 2.0068891332346528e-06, + "loss": 5.77703369140625, + "step": 62900 + }, + { + "epoch": 1.9761168249980372, + "grad_norm": 3.250880718231201, + "learning_rate": 2.00645845334524e-06, + "loss": 5.783766479492187, + "step": 62925 + }, + { + "epoch": 1.9769019392321583, + "grad_norm": 4.854565620422363, + "learning_rate": 2.006041668781364e-06, + "loss": 5.795638427734375, + "step": 62950 + }, + { + "epoch": 1.9776870534662794, + "grad_norm": 3.2064599990844727, + "learning_rate": 2.00563878018697e-06, + "loss": 5.7755224609375, + "step": 62975 + }, + { + "epoch": 1.9784721677004005, + "grad_norm": 3.4355716705322266, + "learning_rate": 2.0052497881845347e-06, + "loss": 5.776912231445312, + "step": 63000 + }, + { + "epoch": 1.9792572819345216, + "grad_norm": 3.208225965499878, + "learning_rate": 2.004874693375065e-06, + "loss": 5.777654418945312, + "step": 63025 + }, + { + "epoch": 1.9800423961686424, + "grad_norm": 3.9131510257720947, + "learning_rate": 2.0045134963380948e-06, + "loss": 5.790040283203125, + "step": 63050 + }, + { + "epoch": 1.9808275104027637, + "grad_norm": 3.998945951461792, + "learning_rate": 2.0041661976316845e-06, + "loss": 5.772805786132812, + "step": 63075 + }, + { + "epoch": 1.9816126246368846, + "grad_norm": 2.9812707901000977, + "learning_rate": 2.003832797792424e-06, + "loss": 5.778757934570312, + "step": 63100 + }, + { + "epoch": 1.9823977388710057, + "grad_norm": 3.8830528259277344, + "learning_rate": 2.003513297335426e-06, + "loss": 5.777207641601563, + "step": 63125 + }, + { + "epoch": 1.9831828531051268, + "grad_norm": 4.639061450958252, + "learning_rate": 2.00320769675433e-06, + "loss": 5.768572998046875, + "step": 63150 + }, + { + "epoch": 1.9839679673392479, + "grad_norm": 4.861164569854736, + "learning_rate": 2.0029159965212987e-06, + "loss": 5.801976928710937, + "step": 63175 + }, + { + "epoch": 1.984753081573369, + "grad_norm": 4.996809482574463, + "learning_rate": 2.002638197087021e-06, + "loss": 5.783984985351562, + "step": 63200 + }, + { + "epoch": 1.98553819580749, + "grad_norm": 2.604578733444214, + "learning_rate": 2.002374298880703e-06, + "loss": 5.789267578125, + "step": 63225 + }, + { + "epoch": 1.9863233100416111, + "grad_norm": 3.670734167098999, + "learning_rate": 2.0021243023100786e-06, + "loss": 5.7835009765625, + "step": 63250 + }, + { + "epoch": 1.987108424275732, + "grad_norm": 3.3542206287384033, + "learning_rate": 2.0018882077614004e-06, + "loss": 5.78148193359375, + "step": 63275 + }, + { + "epoch": 1.9878935385098533, + "grad_norm": 3.5213093757629395, + "learning_rate": 2.0016660155994426e-06, + "loss": 5.7762841796875, + "step": 63300 + }, + { + "epoch": 1.9886786527439742, + "grad_norm": 6.5668721199035645, + "learning_rate": 2.0014577261674995e-06, + "loss": 5.776617431640625, + "step": 63325 + }, + { + "epoch": 1.9894637669780955, + "grad_norm": 2.660402536392212, + "learning_rate": 2.001263339787385e-06, + "loss": 5.7809716796875, + "step": 63350 + }, + { + "epoch": 1.9902488812122163, + "grad_norm": 2.714001178741455, + "learning_rate": 2.0010828567594335e-06, + "loss": 5.774329833984375, + "step": 63375 + }, + { + "epoch": 1.9910339954463374, + "grad_norm": 3.748929977416992, + "learning_rate": 2.000916277362497e-06, + "loss": 5.78482421875, + "step": 63400 + }, + { + "epoch": 1.9918191096804585, + "grad_norm": 3.7747581005096436, + "learning_rate": 2.000763601853945e-06, + "loss": 5.776339721679688, + "step": 63425 + }, + { + "epoch": 1.9926042239145796, + "grad_norm": 3.380877733230591, + "learning_rate": 2.0006248304696684e-06, + "loss": 5.807557373046875, + "step": 63450 + }, + { + "epoch": 1.9933893381487007, + "grad_norm": 3.650698184967041, + "learning_rate": 2.0004999634240723e-06, + "loss": 5.781338500976562, + "step": 63475 + }, + { + "epoch": 1.9941744523828215, + "grad_norm": 3.0879533290863037, + "learning_rate": 2.00038900091008e-06, + "loss": 5.7684088134765625, + "step": 63500 + }, + { + "epoch": 1.9949595666169428, + "grad_norm": 3.995515823364258, + "learning_rate": 2.000291943099135e-06, + "loss": 5.777334594726563, + "step": 63525 + }, + { + "epoch": 1.9957446808510637, + "grad_norm": 3.4350199699401855, + "learning_rate": 2.0002087901411913e-06, + "loss": 5.77222412109375, + "step": 63550 + }, + { + "epoch": 1.996529795085185, + "grad_norm": 5.14459228515625, + "learning_rate": 2.0001395421647265e-06, + "loss": 5.76921142578125, + "step": 63575 + }, + { + "epoch": 1.9973149093193059, + "grad_norm": 3.0829570293426514, + "learning_rate": 2.0000841992767283e-06, + "loss": 5.788284912109375, + "step": 63600 + }, + { + "epoch": 1.998100023553427, + "grad_norm": 3.0940370559692383, + "learning_rate": 2.000042761562705e-06, + "loss": 5.77649658203125, + "step": 63625 + }, + { + "epoch": 1.998885137787548, + "grad_norm": 5.1160807609558105, + "learning_rate": 2.000015229086678e-06, + "loss": 5.7973175048828125, + "step": 63650 + }, + { + "epoch": 1.9996702520216691, + "grad_norm": 4.021876811981201, + "learning_rate": 2.0000016018911868e-06, + "loss": 5.787115478515625, + "step": 63675 + }, + { + "epoch": 2.0, + "eval_loss": 0.14045928418636322, + "eval_runtime": 5084.6062, + "eval_samples_per_second": 4.264, + "eval_steps_per_second": 0.266, + "eval_wer": 3.054766364990921, + "step": 63686 + } + ], + "logging_steps": 25, + "max_steps": 63686, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 6.949104696073912e+21, + "train_batch_size": 16, + "trial_name": null, + "trial_params": null +}